Class _^JL Book 'f~'/ Z (7?¥ n siht Electrical Engineer's pocket-book: A BAND-BOOK OF USEFUL DATA FOR ELECTRICIANS AND ELECTRICAL ENGINEERS. BY | HORATIO A. FOSTER, Fellow A. I. E. E.; Mem. A. S. M. E. Consulting Engineer. Author of " Engineering Valuation of Public Utilities and Factories." With the collaboration of eminent specialists. SEVEJVT^f^Qf^p^ HE VISED mymm£nTpnm r FOR OFFIQA^JJSE C ACCOUNTED FOR IN tfKBfcwiTH PAR. 693 A. h. NEW YORK: D. VAN NO-STRAND COMPANY 1918. Copyrighted, 1902, 1908, 1913, by D. VAN NOSTRAND COMPANY, New York. gj 3IHT By Exchange Mlddtetown Depot Supply March 10,188! -4 r PREFACE TO THE FIFTH EDITION. ^ £ In appreciation of the very cordial reception accorded the earlier editions of this book, and in recognition of the fact that vast changes and advances have occurred in every branch of electrical engineering since the original publication, the author feels called upon to issue the present revised and enlarged edition. The book as now presented, exceeds the previous editions in magnitude by about 600 pages, while the subject matter of every section has been either com- pletely revised and brought up to date, or entirely re-written. The aim throughout has been to supply in exhaustive and condensed form, the data essential to the engineer engaged in any of the branches of the vast domain of electrical engineering. While our concep- tion of the fundamental principles of electrical science can of necessity have undergone no very considerable alteration, those esse&ttal details which in effect con- stitute thet ^working -data .Qf the practicing engineer have so altered and g^wn : that books published only a few years ago are already obsolete. It is believed that a stage in the progress of electrical engineering standardization has now been reached wherein a com- pilation such as the present can be accepted as embody- ing the vital element to which future advmces will appear to a degree in the relation of superficial alter- ations. The original plan of dividing the subject into a number of sections and having each revised by an iii iV PREFACE TO THE FIFTH EDITION. eminent specialist in that particular field has again been followed. Aside from the easy accessibility afforded, this plan of construction is valuable only in proportion to the weightiness of the authorities entrusted with the revision of the several divisions, and it is confidently believed that a perusal of the names heading the sections will lead to the conviction that a more approved and authoritative organization could not have been wished for. The several con- tributors are widely known and recognized as among the first of their respective specialties, and it is be- lieved that the general average of excellence assured by their collaboration surpasses that of any compila- tion of the kind previously attempted. Each section is complete in itself, but needless repetition has been avoided by the free use of cross references through the medium of the very extensive index. Attention is directed to the large quantity of new matter, appearing for the first time in print, in the several sections. In the section on Conductors, e.g., the tables of Inductance, Capacity and Impedance, will be found new and original. Many sections, e.g., Street Railways, Photometry, Conductors, Lighting, Roentgen Rays, etc., are pointed out as examples of exhaustive though condensed presentation. The mechanical section has been treated with the same care and attention as the electrical. The matter has been confined to the requirements of the electrical trades and sciences, the inclusion of the usual mathematical tables and data found in the commonly used handbooks having been avoided. These tables being easily accessible, and the present PREFACE TO THE FIFTH EDITION. V edition being already of great magnitude, this exclu- sion will be appreciated. An important feature of the present volume will be found in the voluminous and studiously developed index and table of contents. The index is as com- plete as the limitations of manipulative facility will permit, and is calculated to render the finding of the particular phase of the subject sought a matter of least possible labor. The table of contents is designed to supplement and extend the use of the index, and in conjunction with the marginal thumb-index will render instantaneous the location of sections and subdivisions. The careful and lengthy work of revision and search leads the author to believe that the number of errors cannot be large, and he ventures to express the hope that readers discovering any will have the kindness to bring them to his attention. In conclusion the author begs to express his grati- tude to the many contributors for their cooperation, and to the publishers for their painstaking effort and generosity in making so handsome and substantial a volume. HORATIO A. FOSTER. 100 Broadway, New York. June 1, 1908. PREFACE TO THE SEVENTH EDITION. No attempt has been made in this edition to add new matter nor to make radical changes in the old, but in a few cases substitution has been made, as in the latest revision of the Standardization Rules of the Am. Inst. E. E., and some changes in the text and cuts in the chapter on Switchboards. A number of typographical and other errors have been corrected. HOEATIO A. FOSTER. 43 Exchange Place, New York, April 1, 1913. LIST OF CONTRIBUTORS. I Section. Symbols, Units, Instruments Measurements Revised by ( W. N. Goodwin, Jr. \ J. Frank Stevens. jW. N. Goodwin, Jr. I Prof. Samuel Sheldon. Magnetic Properties of Iron j Townsend Wolcott. Electromagnets r magnetic induction. J£, Magnetizing force. ^, Magnetomotive force. (ft, Keluctance, Magnetic le- sistance. m, Magnetic permeability. K , Magnetic susceptibility. v, Reluctivity (specific mag- netic resistance). Derived electromagnetic. D, Resistance, Ohm. do, megohm. Electromotive force, volt. Difference of potential, volt. Intensity of current, Ampere. Quantity of electricity, Am- pere-hour ; Coulomb. Capacity. Farad. Electric Energy, Watt-hour ; Joule. Electric Power, Watt ; Kilo- watt. Resistivity (specific resis- tance), Ohm-centimeter. Conductance, Mho. Conductivity (specific con- ductivity/ Admittance, mho. Impedance, ohm. Reactance, ohm. Susceptance, mho. Inductance (coefficient of Induction), Henry. Ratio of electro-magnetic to electrostatic unit of quan- tity =■ 3 X 10 10 centimeters per second approximately. Symbols in general use. Diameter. Radius. Temperature. Deflection of galvanometer needle. SYMBOLS, UNITS, INSTRUMENTS. N, n, ^/, G, S, N, n, S, s, A.M. V.M. A.C. D.C. P.D. O.G.S. B. &S. B.W.G Number of anything. Circumference -^- diameter : 3.141592. 2tt N = 6.2831 X frequency, in alternating current. Frequency, periodicity, cy- cles per second. Galvanometer. Shunt. North pole of a magnet. South pole of a magnet. Ammeter. Voltmeter. Alternating current. Direct current. Potential difference. Centimeter, Gramme, Second system. Brown & Sharpe wire gauge. ,, Birmingham Wire gauge. R.p.m., Revolutions per minute C.P. Candlepower. — o — Incandescent lamp. I X Arc lamp. HhoR-S- , Condenser. ♦ Battery of cells. >sC Dynamo or motor, d.c. s@ Dynamo or motor, a.c. M Converter. UaajuJ •VWVWVW Static transformer. Inductive resistance. Non-inductive resistance. CHAPTER II. ELECTRICAL ENGINEERING UNITS. Index Notation. Electrical units and values oftentimes require the use of large numbers of many figures both as whole numbers and in decimals. In order to avoid this to a great extent the index method of notation is in universal use in connection with all electrical computations. In indicating a large number, for example, say, a million, instead of writ- ing 1,000,000, it would by the index method be written 10 6 ; and 35,000,000 would be written 35 X 10 6 . A decimal is written with a minus sign before the exponent, or, T £ = .01 = 10" 2 ; and .00048 is written 48 x 10~ 5 . The velocity of light is 30,000,000,000 cms. per sec, and is written 3 x 10 10 . In multiplying numbers expressed in this notation the significant figures are multiplied, and to their product is annexed 10, with an index equal to the sum of the indices of the two numbers. In dividing, the significant figures are divided, and 10, with an index equal to the difference of the two indices of the numbers is annexed to the divi- dend. Fundamental Units. The phvsical qualities, such as force, velocity, momentum, etc., are ex- pressed in terms of length, mass, time, and for electricity the system of terms in universal use is that known as the C. G. S. system, viz. : The unit of length is the Centimeter. The unit of mass is the Gramme. The unit of time is the Second. Expressed in more familiar units, the Centimeter is equal to .3937 inch in length ; the Gramme is equal to 15.432 grains, and represents the mass or quantity of a cubic centimeter of water at 4° C, or 39.2° Fah. ; the Second is the HS i^Tffs part of a sidereal day, or the „|„„ part of a mean solar day. These units are also often called absolute units. I>erived Geometric Units. The unit of area or surface is the square centimeter. The unit of volume is the cubic centimeter. Derived IVIechanical Units. Velocity is the rate of change of position, and is uniform velocity when equal distances are passed over in equal spaces of time ; unit velocity is a rate of change of one centimeter per second. ELECTRICAL ENGINEERING UNITS. 5 Angular Velocity is the angular distance about a center passed through in one second of time. Unit angular velocity is the velocity of a body moving in a circular path, whose radius is unity, and which would traverse a unit angle in unit time. Unit angle is 57°, 17', 44.8" approximately ; i.e., an angle whose arc equals its radius. Momentum is the quantity of motion in a body, and equals the mass times the velocity. Acceleration is the rate at which velocity changes ; the unit is an accel- eration of one centimeter per second per second. The acceleration due to gravity is the increment in velocity imparted to falling bodies by gravity, and is usually taken as 32.2 feet per second, or 981 centimeters per second. This value differs somewhat at different localities. At the North Pole g = 983.1 ; at the equator g = 978.1 ; and at Greenwich it is 981.1. Force acts to change a body's condition of rest or motion. It is that which tends to produce, alter, or destroy motion, and is measured by the time rate of change of momentum produced. The unit of force is that force which, acting for one second on a mass of one gramme, gives the mass a velocity of one centimeter per second ; this unit is called a dyne. The force of gravity or weight of a mass in dynes may be found by multiplying the mass in grammes by the value of g at the par- ticular place where the force is exerted. The pull of gravity on one pound in the United States may be taken as 445,000 dynes. Work is the product of a force into the distance through which it acts. The unit is the erg, and equals the work done in pushing a mass through a distance of one centimeter against a force of one dyne. As the " weight" of one gramme is 1 X 981, or 981 dynes, the work done in raising a weight of one gramme through a height of one centimeter against the force of gravity, or 981 dynes, equals 1 X 981 = 981 ergs. One kilogramme- meter = 100000 x 981 ergs. Kinetic energy is the work a body is able to do by reason of its motion. Potential energy is the work a body is able to do by reason of its position. The unit of energy is the erg. Power is the rate of working, and the unit is the watt = W 7 ergs per sec. Horse-power is the unit of power in common use and, although a somewhat arbitrary unit, it is difficult to compel people to change from it to any other. It equals 33,000 lbs. raised one foot high in one minute, or 550 foot-pounds per second. 1 ft.-lb. = 1.356 X 10 7 ergs. 1 watt = 10 7 ergs per second. 1 horse-power = 550 x 1.356 X 10 7 ergs = 746 watts. If a current of / am- peres flow through P ohms under a pressure of E volts, then — = — -— — E 2 • g represents the horse-power involved. The French "force de cheval" =736 watts =542.48 ft. lbs. per sec.= .9863 H. P., and 1 H.P. = 1.01389 "force de cheval." Heat. The Joule WJ= 10 7 ergs, and is the work done, or heat generated, by a watt second, or ampere flowing for a second through a resistance of an ohm. If i/ = heat generated in gramme calories, i= current in amperes, 2£ = e.m.f. in volts, R=z resistance in ohms, and 2 = time in seconds, then ^=0.24/2^ = 0.24 Elt. gramme calories or therms. EH Then IEt = I 2 Pt = -- — EQ— Joules. or, as 1 horse-power = 550 foot-pounds of work per second, Joules = ft%EQ = .7373 E Q f t. lbs. If eat Units. The British Thermal Unit is the amount of heat required to raise the temperature of one pound of water one deg. F. at or near its temp, of max. density, 39.1°; = 1 pound-degree-Fah. = 251 .9 French calories. The Calorie is the amount of heat required to raise the temperature of a 4 SYMBOLS, UNITS, INSTRUMENTS. mass of 1 gramme of water from 4° C. to 5° C. = 1 gramme-degree-centi- grade. Water at 4° C. is at its maximum density. Joules equivalent, 7, is the amount of energy equal to a heat unit. For a B.T.U., or pound-degree-Fali., 7=1.07 X 10 10 ergs., or = 778 foot- pounds. For one pound-degree — Centigrade, 7= 1.93 x 10 10 ergs. For a calorie .7=4.189 X 10 7 ergs. The heat generated in t seconds of time is £!|? = ^? t whe re 7=4.189 X 10 7 , J J and 7, R, and E are expressed in practical units. Electrical Units. There are two sets of electrical units derived from the fundamental C. G. S. units; viz., the electrostatic and the electromagnetic. The first is based on the force exerted between two quantities of electricity, and the sec- ond upon the force exerted between a current and a magnetic pole. The ratio of the electrostatic to the electromagnetic units has been carefully de- termined by a number of authorities, and is found to be some multiple or sub-multiple of a quantity represented by v, whose value is approximately 3 X 10 10 centimeters per second. Convenient rules for changing from one to the other set of units will be stated later on in this chapter. Electrostatic Units. As yet there have been no names assigned to these. Their values are as follows : The unit of quantity is that quantity of electricity which repels with a force of one dyne a similar and equal quantity of electricity placed at unit distance (one centimeter) in air. Unit of current is that which conveys a unit of quantity along a conduc- tor in unit time (one second). Unit difference of potential or unit electro-motive force exists between two points when one erg of work is required to pass a unit quantity of electricity from one point to the other. Unit of resistance is possessed by that conductor through which unit cur- rent will pass under unit electro-motive force at its ends. Unit of capacity is that which, when charged by unit potential, will hold one unit of electricity ; or that capacity which, when charged with one unit of electricity, has a unit difference of potential. Specific inductive capacity of a substance is the ratio between the capacity of a condenser having that substance as a dielectric to the capacity of the same condenser using dry air at 0° C. and a pressure of 76 centimeters as the dielectric. Magnetic Units. Unit Strength of Pole (symbol m) is that which repels another similar and equal pole with unit force (one dyne) when placed at unit distance (one centimeter) from it. Magnetic Moment (symbol 9K ) is tne product of the strength of either pole into the distance between the two poles. Intensity of Magnetization is the magnetic moment of a magnet divided by its volume, (symbol Q). Intensity of Magnetic Field (symbol J£ ) is measured by the force it exerts upon a unit magnetic pole, and therefore the unit is that intensity of field which acts on a unit pole with a unit force (one dyne). Magnetic Induction (symbol (&) is the magnetic flux or the number of magnetic lines per unit area of cross-section of magnetized material, the area being at every point perpendicular to the direction of flux. It is equal to the magnetizing force or field intensity J£ multiplied by the permeability ft: the unit is the gauss. Magnetic Flux (symbol $) is equal to the average field intensity multiplied by the area. Its unit is the maxwell. Magnetizing Force (symbol J£ ) per unit of length of a solenoid equals ELECTRICAL ENGINEERING UNITS. 4 n NI -7- L where 2V= the number of turns of wire on the solenoid ; L = the length of the solenoid in cms., and I = the current in absolute units. Magnetomotive Force (symbol 9F ) is the total magnetizing force developed in a magnetic circuit by a coil, equals 4 n AT, and the unit is the gil- bert. Reluctance, or Magnetic Resistance (symbol (ft), is the resistance offered to the magnetic flux by the material magnetized, and is the ratio of magneto- motive force to magnetic flux; that is, unit magnetomotive force will generate a unit of magnetic flux through unit reluctance : the unit is the oersted; i.e., the reluctance ottered by a cubic centimeter of vacuum. Magnetic Permeability (symbol /*) is the ratio of the magnetic induction (ft to the magnetizing force J£, that is ^ = ft. Magnetic Susceptibility (symbol k) is the ratio of the intensity of mag- netization to the magnetizing force, or k = ^ • Reluctivity , or Specific Magnetic Resistance (symbol v), is the reluctance per unit of length and of unit cross-section that a material offers to being magnetized. Electromag-netic Units. Resistance (symbol R) is that property of a material that opposes the flow of a current of electricity through it; and the unit is that resistance which, with an electro-motive force or pressure between its ends of one unit, will permit the flow of a unit of current. The practical unit is the ohm, and its value in C.S.G. units is 10 9 . The standard unit is a column of pure mercury at 0° C, of uniform cross-section, 106.3 centimeters long, and 14.4521 grammes weight. For convenience in use for very high resistances the prefix meg is used; and the megohm, or million ohms, becomes the unit for use in expressing the insulation resistances of submarine cables and all other high resistances. Electro-motive Force (symbol E) is the electric pressure which forces the current through a resistance, and unit E.M.F. is that pressure which will force a unit current one ampere through a unit resistance. The unit is the volt, and the practical standard adopted by the international congress of elec- tricians at Chicago in 1893 is the Clark cell, directions for making which will be given farther on. The E.M.F. of a Clark cell is 1.434 volt at 15° C. The value of the volt in C.G.S. units is 10 8 . For small E.M.F's. the unit millivolt, or one-thousandth volt, is used. The International Volt is 1.1358 B. A. volts; and the ratio of B. A. volt to the International volt is .9866. Difference of Potential, as the name indicates, is simply a difference of electric pressure between two points. The unit is the volt. Current (symbol /) is the intensity of the electric current that flows through a circuit. A unit current will flow through a resistance of one ohm, with an electro-motive force of one volt between its ends. The unit is the ampere, and is practically represented by the current that will electro- lytically deposit silver at the rate of .001118 gramme per second. Its value in C.G.S. units is 10 -1 . For small values the milliampere is used, and it equals one-thousandth of an ampere. The Quantity of Electricity (symbol Q) which passes through a given cross- section of an individual circuit in t seconds when a current of /amperes is flowing is equal to It units. The unit is therefore the ampere-second. Its name is the Coulomb, and its value in C.G.S. units is 10 -1 . Capacity (symbol C) is the property of a material condenser for holding a charge of electricity. A condenser of unit capacity is one which will be charged to a potential of one volt by a quantity of 1 coulomb. The unit is the farad, its C.G.S. value is 10~ 9 ; and this being so much larger than ever obtains in practical work, its millionth part, or the micro-farad, is used as the practical unit, and its value in absolute units is 10 _ 15 . A condenser of one-third micro-farad capacity is the size in most common use in the U. S. Electric Energy (symbol W) is represented by the work done in a circuit or conductor by a current flowing through it. The unit is the Joule, its absolute value is 10 7 ergs, and it reprepresents the work done by the flow for one second of unit current (1 ampere) through 1 ohm. Electric Power (symbol P) is measured in watts, and is represented by a current of 1 ampere under a pressure of 1 volt, or 1 Joule per second. The SYMBOLS, UNITS, MEASUREMENTS. > do © rx • © p Jf .2 5 ,0 .« C p ;■«§! -O P e8 ©P <2 flfl © QQ oP id *6 2- « SO ai-tf QQ ^ ^P p 00 S3 ,4 p be... bC-, . tig ©*s •► 2.s COM c3^ hi* 2 2 -* II £§ ft S/ v X S « S o . 'u © %!> bfi"£ © ®, br, © p* g Am © $ ®2 5 fcb p © B--* © IS ©-© £5 6 2 S « bcu. o © a ©< M as © © £ exp b^ © "1*H © B © © 00 Q GO X ©1 © o P © 5 © 5 © © a,© . © . © . © £ a c bX) * Sfl M P< •~-p bC ^*< S|8| ^ ^ fe sis S mi it ii it it mi i! lS*tf ?fe ft, 3 * b£ CO © c rt a 5*2 -w © i?p co © • 5 2 £ a - a -a © -d * a? © 35 O ee C > S © bo © 0) S3 & C so a s .d O O > m & ■s * o r ^ Hn Hn Si Si nJ? «|N W|N I ^ S *q ^ L ^ ^ 51 -In Nn|n He, H- S* -In C S| Hn ^ Sj ^ ^ Sj -4n "^ Hn Hn ^h | W|N | I HfN ^ S ^ ^ ^ fc < a 1 ^ PH Hn II II ei^^l^ii ^S ii ii ^ ^ * 6 & & g «5 feJ a,^ £ ^ O & * g * •5 +3 ^v « a 3 60 O R o ^3 be •S « fa bo ^ bo ^S3 oq 53 a o bD © 00 S3 S3 S3 © bo © 2 o © © © a EL s be bp bo c3 e« S S fafaS£g« s ® • — •!h is © © a S3 bD bD ri eS © fe *j s >« 3? © S3 M « W SYMBOLS, UNITS, INSTRUMENTS. ,4 £ * I I I »- ci {.I'd !► © s s o o ,S»gog ©^ *6 « ^ '3+a < o o 02 CO © AG'S 2 as w ^ k t & &* ^ t, ^ ^ 2, 2, ^ ^ H ft - t £ t, ?, ^ fc v ^ ^ t s ° Sis II II II II II II II II II II b^0»0^^ <*■ Cb ^ ^ ~to + > & ^ H K5 3 ■«* C* O k to" s~ c£ o fe ^ - g J* csj ^ ft; p 04 o ^ CD M o W Ph to 5 5J w V O4 CP © q 33 .22 2 -5 O . M .5 ec a ca o a -p-p iB-P s eS 3 3 « 3 § £ a © © i o 0^*5 O U P^ "< « 02 INTERNATIONAL ELECTRICAL UNITS. 9 watt equals 10 7 absolute units, and 746 watts equals 1 horse-power. In elec- tric lighting and power the unit kilowatt, or 1000 watts, is considerably used to avoid the use of large numbers. Resistivity (symbol p) is the specific resistance of a substance, and is the resistance in ohms of a centimeter cube of the material to a flow of cur- rent between opposite faces. Conductance (symbol G) is that property of a metal or substance by which it conducts an electric current, and equals the reciprocal of its resistance. The unit proposed for conductance is the Mho, but it has not come into prominent use as yet. Conductivity (symbol v) is the specific conductance of a material, and is therefore the reciprocal of its resistivity. It is often expressed in compari- son with the conductivity of some standard metal such as silver or copper, and is then stated as a percentage. Inductance (symbol L), or coefficient of self-induction, of a circuit is that coefficient by which the time rate of change of the current in the circuit must be multiplied in order to give the E.M.F. of self-induction in the circuit. The practical unit is the henry, which equals 10 9 absolute units, and exists in a circuit when a current varying 1 ampere per second produces 2l volt of electro-motive force in that circuit. As the henry is so large as to be seldom met with in practice, 1 thousandth of it, or the milli-henry , is the unit most in use. Below will be found a few rules for reducing values stated in electrostatic units to units in the electro-magnetic system. To reduce electrostatic potential to volts, multiply by 300 ; 44 capacity to micro-farads, divide by 900,000 ; 44 quantity to coulombs, divide by 3 x 10 9 ; 44 current to amperes, divide by 3 X 10 9 ; 44 resistance to ohms, multiply by 9 x 10 11 . ¥I¥TEIt]¥AMO]¥AJL EIECIRICAL UMTS. At the International Congress of Electricians, held at Chicago, August 21, 1893, the following resolutions met with unanimous approval, and being approved for publication by the Treasury Department of the United States Government, Dec. 27, 1893, and legalized by act of Congress and approved by the President, July 12, 1894, are now recognized as the International units of value for their respective purposes. RESOL VED, That the several governments represented by the delegates of the International Congress of Electricians be, and they are hereby, recommended to formally adopt as legal units of electrical measure ttie following : 1. As a unit of resistance, the International ohm, which is based upon the ohm equal to 10 9 units of resistance of the C.G.S. system of electro-magnetic units, and is represented by the resistance offered' to an unvarying electric current by a column of mercury at a temperature of melting ice, 14.4[»21 grammes in mass, of a constant cross-sectional area, and of the length 10^.3 centimeters. 2. As a unit of current, the International ampere, which is one-tenth of the unit of current of the C.G.S. system of electro-magnetic units, and which is represented sufficiently well for practical use by the unvarying current which, when passed through a solution of nitrate of silver in water, in accordance with the accompanying specification (A) deposits silver at t lie rate of 0.001118 gramme per second. 3. As a unit of electro-motive force the international volt which is the E.M.F. that, steadily applied to a conductor whose resistance is one Inter- national ohm, will produce a current of one international ampere, and 1000 which is represented sufficiently well for practical use by — — of the E.M.P. between the poles or electrodes of the voltaic cell known as Clark's cell at a temperature of 15° C, and prepared in the manner described in the ac- companying specification (B). 4. As the unit of quantity, the International coulomb, which is the quan- tity of electricity transferred by a current of one international ampere in one second. 5. As the unit of capacity the international farad , which is the capacity 10 SYMBOLS, UNITS, INSTRUMENTS. of a conductor charged to a potential of one international volt by one inter- national coulomb of electricity. 6. As the unit of work, the joule, which is 10 7 units of work in the C.G.S. system, and which is represented sufficiently well for practical use by the energy expended in one second by an international ampere in an inter- national ohm. 7. As the unit of power, the watt, which is equal to 10 7 units of power in the C.G.S. system, and which is represented sufficiently well for practical use by'the work done at the rate of one joule per second. 8. As the unit of induction, the henry, which is the induction in the cir- cuit when the E.M.F. induced in this circuit is one international volt, while the inducing current varies at the rate of one international ampere per second. Specification A. In employing the silver voltameter to measure currents of about one ampere, the following arrangements shall be adopted: : The kathode on which the silver is to be deposited shall take the form of a platinum bowl not less than 10 cms. in diameter, and from 4 to 5 cms. in depth. The anode shall be a disk or plate of pure silver some 30 sq. cms. in area, and 2 or 3 cms. in thickness. This shall be supported horizontally in the liquid near the top of the solution by a silver rod riveted through its center. To prevent the disintegrated silver which is formed on the anode from falling upon the kathode, the anode shall be wrapped around with pure filter paper, secured at the back by suitable folding. The liquid shall consist of a neutral solution of pure silver nitrate, con- taining about 15 parts by weight of the nitrate to 85 parts of water. The resistance of the voltameter changes somewhat as the current passes. To prevent these changes having too great an effect on the current, some resistance, besides that of the voltameter, should be inserted in the circuit. The total metallic resistance of the circuit should not be less than 10 ohms. Method of making* a Measurement. — The platinum bowl is to be washed consecutively with nitric acid, distilled water, and absolute alcohol ; it is then to be dried at 160° 0., and left to cool in a desiccator. When cold it is to be weighed carefully. It is to be nearly filled with the solution, and connected to the rest of the circuit by being placed on a clean copper support to which a binding-screw is attached The anode is then to be immersed in the solution so as to be well covered by it, and supported in that position ; the connections to the rest of the circuit are then to be made. Contact is to be made at the key, noting the time. The current is to be allowed to pass for not less than half an hour, and the time of breaking contact observed. The solution is now to be removed from the bowl, and the deposit washed with distilled water, and left to soak for at least six hours. It is then to be nused successively with distilled water and absolute alcohol, and dried in a hot-air bath at a temperature of about 160° C. After cooling in a desiccator 11 m fc £ J* Y eighed a S am - The S ain in mass gives the silver' deposited. To find the time average of the current in amperes, this mass, expressed m grammes, must be divided bv the number of seconds during which th< current has passed and by 0.001118. In determining the constant of an instrument by this method the current should be kept as nearly uniform as possible, and the readings of the instru- ment observed at frequent intervals of time. These observations give a curve from which the reading corresponding to the mean current (time average of the current) can be found. The current is calculated from the voltameter results, corresponding to this reading. * & The current used in this experiment must be obtained from a battery and not from a dynamo, especially when the instrument to be calibrated is an electrodynamometer. Specification B. — The Volt. The cell has for its positive electrode, mercury, and for its negative elec- trode, amalgamated zinc ; the electrolyte consists of a saturated solution of SPECIFICATION B. 11 tine sulphate and mercurous sulphate. The electromotive force is 1.434 volts at 15° C, and, between 10° C. and 25° C, by the increase of 1° C. in tempera- ture, the electromotive force decreases by .00115 of a volt. 1. Preparation of the mercury. — To secure purity it should be first treated with acid in the usual manner, and subsequently distilled in vacuo. 2. Preparation of the Zinc Amalgam,- The zinc designated in commerce as " commercially pure" can be used without further prepara- tion. For the preparation of the amalgam one part by weight of zinc is to be added to nine (9) parts by weight of mercury, and both are to be heated in a porcelain dish at 100° C. with moderate stirring until the zinc has been fully dissolved in the mercury. 3. Preparation of the Mercurous Sulphate. — Take mercurous sulphate, purchased as pure, mix with it a small quantity of pure mercury, and wash the whole thoroughly with cold distilled water by agitation in a bottle ; drain off the water and repeat the process at least twice. After the last washing, drain off as much of the water as possible. (For further de- tails of purification, see Note A.) 4. Preparation of the Zinc Sulphate Solution. — Prepare a neutral saturated solution of pure re-crystallized zinc sulphate, free from iron, by mixing distilled water with nearly twice its weight of crystals of pure zinc sulphate and adding zinc oxide in the proportion of about 2 per cent by weight of the zinc sulphate crystals to neutralize any free acid. The crystals should be dissolved by the aid of gentle heat, but the temperature to which the solution is raised must not exceed 30° C. Mercurous sulphate, treated as described in 3, shall be added in the proportion of about 12 per cent by weight of the zinc sulphate crystals to neutralize the free zinc oxide remaining, and then the solution filtered, while still warm, into a stock bottle. Crystals should form as it cools. 5. Preparation of the Mercurous Sulphate and Zinc Sul- phate Paste. — For making the paste, two or three parts by weight of mercurous sulphate are to be added to one by weight of mercury. If the sulphate be dry, it is to be mixed with a paste consisting of zinc sulphate crystals and a concentrated zinc sulphate solution, so that the whole con- stitutes a stiff mass, which is permeated throughout by zinc sulphate crys- tals and globules of mercury. If the sulphate, however, be moist, only zinc sulphate crystals are to be added ; care must, however, be taken that these occur in excess, and are not dissolved after continued standing. The mercury must, in this case also, permeate the paste in little globules. It is advantageous to crush the zinc sulphate crystals before using, since the paste can then be better manipulated. To set up the Cell. —The containing glass vessel, represented in the accompanying figure, shall consist of two limbs closed at bottom, and joined above to a common neck fitted with a ground-glass stopper. The diameter of the limbs should be at least 2 cms. and their length at least 3 cms. The neck should be not less than 1.5 cms. in diameter. At the bottom of each limb a platinum wire of about 0.4 mm. in diameter is sealed through the glass. To set up the cell, place in one limb mercury, and in the other hot liquid amalgam, containing 90 parts mercury and 10 parts zinc. The platinum wires at the bottom must be completely covered by the mercury and the amalgam respectively. On the mercury, place a layer one cm. thick of the zinc and mercurous sulphate paste described in 5. Both this paste and the zinc amalgam must then be covered with a layer of the neutral zinc sul- phate crystals one cm. thick. The whole vessel must then be filled with the saturated zinc sulphate solu- tion, and the stopper inserted so that it shall just touch it, leaving, however, a small bubble to guard against breakage when the temperature rises. Before finally inserting the glass stopper, it is to be brushed round its upper edge with a strong alcoholic solution of shellac, and pressed firmly in place. (For details of filling the cell see Note B.) Fig 1. 12 SYMBOLS, UNITS, INSTRUMENTS. Pm -'C «o 1- co|^ © 3 ^ eo 00 co _r IS I § W <»r3~ si ss rH OO rH 00 s § - 8 ^2 CO j . CO M CO t- 1 t- - § 6SS Mug 5 2 rJ ^ rH rH ag-s » s CO l> g +9 rg * O IS <=> o t- 8 3 s * £ T* rH rH 5 cjdPh O 03 § S -d g Pn-d SIS 8 8 0S2 • p 58 boo rH rH <*> CO rH ^ CM 5 « § rH ro |rH © £ 10 S H *T a 2 . ago O-d § 3 ^ -r rt< ""' 8 © § g X X X so rH >r) 7 - O Ph o o Ph « w DESCRIPTION OF INSTRUMENTS. 13 Notes to the Specifications. (A). The Mercnrous Sulphate. — The treatment of the mercurous sulphate has for its object the removal of any mercuric sulphate which de- composes in the presence of water into an acid and a basic sulphate. The latter is a yellow substance — turpeth mineral — practically insoluble in water ; its presence, at any rate in moderate quantities, has no effect on the cell. If, however, it be formed, the acid sulphate is also formed. This is soluble in water, and the acid produced affects the electromotive force. The object of the washings is to dissolve and remove this acid sulphate, and for this purpose the three washings described in the specification will suffice in nearly all cases. If, however, much of the turpeth mineral be formed, it shows.that there is a great deal of the acid sulphate present ; and it will then be wiser to obtain a fresh sample of mercurous sulphate, rather than to try by repeated washings to get rid of all the acid. The free mercury helps in the process of removing the acid ; for the acid mercuric sulphate attacks it, forming mercurous sulphate. Pure mercurous sulphate, when quite free from acid, shows on repeated washing a faint yellow tinge, which is due to the formation of a basic mer- curous salt distinct from the turpeth mineral, or basic mercuric sulphate. The appearance of this primrose yellow tinge, which is due to the formation of a basic mercurous salt distinct from the turpeth mineral, or basic mer- curic sulphate, may be taken as an indication that all the acid has been removed ; the washing may with advantage be continued until this tint appears. (B). filling* the Cell. — After thoroughly cleaning and drying the glass vessel, place it in a hot-water bath. Then pass through the neck of the vessel a thin glass tube reaching to the bottom to serve for the intro- duction of the amalgam. This tube should be as large as the glass vessel will admit. It serves to protect the upper part of the cell from being soiled with the amalgam. To fill in the amalgam, a clean dropping-tube about 10 cms. long, drawn out to a fine point, should be used. Its lower end is brought under the surface of the amalgam heated in a porcelain dish, and some of the amalgam is drawn into the tube by means of the rubber bulb, The point is then quickly cleaned of dross with filter paper, and is passed through the wider tube to the bottom, and emptied by pressing the bulb. The point of the tube must be so fine that the amlagam will come out only on squeezing the bulb. This process is repeated until the limb contains the desired quantity of the amalgam. The vessel is then removed from the water-bath. After cooling, the amalgam must adhere to the glass, and must show a clean surface with a metallic luster. For insertion of the mercury, a dropping-tube with a long stem will be found convenient. The paste may be poured in through a wide tube reach- ing nearly down to the mercury and having a funnel-shaped top. If the paste does not run down freely it may be pushed down with a small glass rod. The paste and the amalgam are then both covered with the zinc sul- phate crystals before the concentrated zinc sulphate solution is poured in. This should be added through a small funnel, so as to leave the neck of the vessel clean and dry. For convenience and security in handling, the cell may be mounted in a suitable case so as to be at all times open to inspection. In using the cell, sudden variations of temperature should, as far as possible, be avoided, since the changes in electromotive force lag behind those of temperature. CHAPTER III. DESCRIPTION OW Il¥STIt¥Jlf«J]¥TS. Although no attempt will be made here to fully describe all the different instruments used in electrical testing, some of the more important will be named and the more common uses to which they may be put mentioned. The four essential instruments for all electrical testing of which all other instruments are but variations, are: the battery, the galvanometer, the resistance-box, and the condenser, and following will be found a concise description of the more important types of each. 14 SYMBOLS, UNITS, INSTRUMENTS. PRIMARY BATT£RI£§, A Voltaic Battery is a device for converting chemical energy directly into electrical energy. If a plate of chemically pure zinc and a plate of copper are immersed in dilute sulphuric acid no chemical action takes place. As soon, however, as the zinc and copper plates are connected by an electrical conductor outside of the liquid a vigorous chemical action is set up, the zinc dis- solves in the acid, and hydrogen is liberated on the copper plate. As long as this action takes place an electric current passes from the zinc plate through the acid to the copper plate and through the conductor back to the zinc plate. The chemical action in this simple voltaic cell soon becomes weaker, and at the same time the intensity of the electric current diminishes and finally becomes zero. The diminution of activity is chiefly due to the accumulation of hydrogen on the copper plate, causing what is known as "polarization." An agent introduced into a galvanic cell to prevent polarization is called a "depolarizer." The chemical reaction of a voltaic cell is directly proportional to the quantity of electricity passing through it. The quantity (in grammes) of an element liberated or brought into combination electrolytically by one coulomb of electricity, is called its electrochemical equivalent. (See table on second page of section on "Electrochemistry.") The theoretical con- sumption of material in a voltaic battery doing a certain amount of work can be calculated from the electrochemical equivalent of the material. For example, in a battery doing work equivalent to one horse-power hour 746 X 3600 X .003387 grammes of zinc will be dissolved; E being the E.M.F. of the battery. In practice the consumption of material in a galvanic cell is larger, due to local action. Commercial zinc always contains iron, carbon, or other impurities; as soon as these are exposed to the liquid, local closed circuits are formed resulting in the consumption of zinc. To prevent this wasteful action, the zinc must be amalgamated with mercury. The action of the mercury brings the pure zinc to the surface and in contact with the liquid. Amalgamated zinc is not attacked by diluted sulphuric acid. Zinc is amalgamated by immersing it in dilute sulphuric or hydrochloric acid for a few minutes to give it a clean surface, then mercury is rubbed on with a hard brush or cloth fixed on the end of a piece of wood. Primary Cells may be classified into two groups; closed circuit and open circuit. Closed Circuit Cells. — Cells of this group must be capable of work- ing on a closed circuit of moderate resistance for a long period without sen- sible polarization. They must, therefore, contain an effective depolarizer. The best depolarizers are copper sulphate CuS0 4 , strong nitric acid HN0 3 , chromic acid Cr0 3 , oxide of copper CuO, and chloride of silver AgCl. The following table contains data on the representative types of closed circuit cells. Name. +Plate. Electrolyte. Depolarizer. — Plate. E.M.F. R. Daniell Grove Bunsen Peggen- dbrff Lande Davy Zinc Sulphuric Acid ii ii Caustic Potash Ammonium Chloride Cop. sulphate Nitric Acid Bichromate of Potassium- Sulp. acid Copper Oxide Silver Chloride Copper Platinum Carbon Iron Silver 1.08 1.9 1.8 2. 1. 1.1 1. .15 .2 .2 .1 4.5 The values given as electromotive force and internal resistance of the different types of cells are approximate only. The E.M.F. depends upon the purity of the materials, the concentration of the solution; the internal resistance, furthermore, depends upon the dimensions and general arrangement of the cells. BATTERIES. 15 Open Circuit Cells. — Cells of this group are only suitable for use when the circuit is to be closed for a few seconds at a time, as for example for call bells, annunciators, etc. Such batteries do not need to contain a quick acting polarizer, as the effect of polarization can be taken care of during the intervals of rest, either by a slow acting depolarizer or even without any polarizer. It is, however, of the greatest importance that no local action takes place in these cells on open circuit. The following table contains data on the representative types of open circuit cells: Name. -f Plate. Electrolyte. Depolarizer — Plate. E.M.F. R. Leclauche Law Gassner Zinc Zinc Zinc Sol. of Sal-ammoniac ii ii ii Oxide of Zinc, sal-am- moniac, Chloride of zinc, plaster Binoxide of Manganese None Carbon Carbon Carbon 1.48 1.37 1.3 .5 .4 .2 Fig. 2. The Oravity Cell. The elements are copper and zinc; the solution is sulphate of copper, or "bluestone," dissolved in water. The usual form (see Fig. 2) is a glass jar, about 8 inches high and 6 inches diameter. The copper is made of two or more layers fastened in the middle, spread out, and set on edge in the bottom of the cell, the terminal being a piece of gutta-percha insulated copper wire extending up through the solution. The zinc is usually cast with fingers spread out, and a hook for suspending from the top of the jar as shown, the terminal being on top of the hook. This form of zinc is commonly called "crowfoot," and the battery often goes by that name. Some- times star-shaped zincs are suspended from a tri- pod across the top of the jar. The "bluestone" crystals are placed in the bottom of the jar about the copper, the jar then being filled with water to just above the "crowfoot" or zinc. A table- spoonful of sulphuric acid is added. A saturated solution of copper sulphate forms around the cop- per; and, after use, a zinc sulphate solution is formed around the zinc, and floats upon the cop- per sulphate solution. The line of separation between the two solutions is called the blue line. As the two solutions are kept separate because of their different specific gravities, the name "gravity cell" is employed. This cell does not polarize, and the E.M.F. is practically constant or uni- form at about 1 volt on a closed circuit. If the circuit is not closed, and the cell does not have work enough to prevent mixing of the two solutions, the copper sulphate coming in contact with the zinc will become decomposed; the oxygen forming oxide of zinc, and the copper depositing on the zinc hav- ing an appearance like black mud. Care of the Oravity Cell. — For ordinary "local work" about three pounds of "bluestone" per cell is usually found best. When this is gone it is better to clean out the cell and supply new solution than to try to re- plenish. "Bluestone" crystals should not be smaller than a pea nor as large as an egg. In good condition the solution at the bottom should be a bright blue, changing to water-color above. A brownish color in any part denotes deterioration. To prevent evaporation of the solution it is well to pour a layer of good mineral oil over the top when the cell is first set up. This oil should be odorless, free from naphtha or acid, and non-inflammable under 400° F. If oil is not used, dipping the top of the jar in melted paraffin for about an inch will prevent the salts of the solution from climbing over the edge. In starting a new battery it is best to short circuit the cells for twenty-four or forty-eight hours to form zinc sulphate and lower the internal resistance, 16 SYMBOLS, UNITS, INSTRUMENTS. The internal resistance of the ordinary gravity cell is 2 to 3 ohms, depending on a number of conditions, such as the size of plates, the nearness together, and the nature of the solution. Never let the temperature of gravity cells get below 65° or 70° F., as the internal resistance increases very rapidly with a decrease in temperature. The JLet lanche Cell. This cell is one of the most commonly used outside of telegraphy, and up to the advent of the so-called dry cell was practically the only one in use for house and telephone work. The elements are zinc and carbon, with per- oxide of manganese about the carbon plate for a depolarizing agent. As usually constructed — for there are many modifications of the type — the jar is of glass, about 7 inches high and 5 inches in diameter, or sometimes square. The zinc is in the form of a stick, about a half inch diameter by 7 inches long, and is placed in one corner of the jar in a solution of sal-ammoniac. The carbon plate is placed in a porous cup within the jar, and the space around the carbon in the cup is filled with small pieces of carbon and gran- ulated peroxide of manganese. The sal-ammoniac solution passes through the porous cup and moistens the contents. This cell will polarize if worked hard or short circuited, but recuperates quickly if left on open circuit for a while* The resistance of the Leclanche" cell varies with its size and con- dition, but is generally less than one ohm. The initial E.M.F. is about 1.5 volt. It is desirable not to use too strong a solution of sal-ammoniac, as crystals will be deposited on the zinc; and not to let the solution get too weak, as chloride of zinc will form on the zinc; both conditions will materially increase the internal resistance of the cell and impair its efficiency. Without knowing the dimensions of cells it is not possible to state the amount of sal-ammoniac to use; but perhaps as good a way as any is to add it to the water until no more will dissolve, then add a little water so that the solution will be weaker than saturation. Keep all parts clean, and add sal-ammoniac and water when necessary. Chloride of Silver Dry Cell Battery. This cell is extensively used for testing insulation of cables, etc., and its elements are a plate of chemi- cally pure zinc and a cast plate of chloride of silver in an electrolyte paste. As ordinarily constructed the jar is of glass about 2k" long by \" diameter with the zinc and silver plates set in as per Fig. 3. The paste is poured in and the cell is then hermetically sealed. The terminals are led through fiber tops to posts thereon. The small size of the cell renders it possible to con- struct a battery of from fifty to two hundred and twenty cells within a small compass. The containing box is provided with a pole-chang- ing switch in the cover and with selecting cords and tips so that the operator may select any number of cells desired. Fig. 4 shows a portable testing battery of fifty of these cells complete ready for use. The E.M.F. of the chloride of silver cell is .9 of a volt, the internal resistance being about 4 ohms. The current supplied is quite constant until within a few moments of its exhaustion ; they will not dry out in any climate, have a long life, and there is no local action developed when the cells are not in use. Fig. 3. Fuller Cell. The elements of this cell are zinc in a dilute solution of sulphuric acid and carbon in a solution of electropoin. Electropoin consists of three parts BATTERIES. 17 < Fig. 4. Chloride of Silver Cells. bichromate of potash, one part sulphuric acid, and nine parts water. Dis- solve the bichromate in the water at boiling, and when cool add the sul- phuric acid slowly. The zinc plate is in the form of a cone, and is placed in the bottom of a porous cup inside a glass jar. The carbon plate is out- side the porous cup. About two ounces of mercury are placed in the porous cup with the zinc, for amalgamation, and the cup is filled with a dilute solution of sulphuric acid. The outside jar is filled with the electropoin. In this the carbon plate is immersed. The E.M.F. is 2 volts, and the internal resistance is about half an ohm. The solution is originally of an orange color. When this becomes bluish in tint, add more crystals. Should the color be normal and the cell be weak, add fresh sulphuric acid. ESdison-I /2r To find the number of cells in series (?i«) and in parallel (n p ) required to give a current (/) through an external resistance (R) and to have an effi- ciency (F). „«, . „ External work Efficiency E = — — — Total work 1 2 R R I*(™ + X) ™+R \n P ) n P ' < n P ) n P The internal resistance of the whole battery is n 8 r R(l — F) , _ n,EF — = — ^—s — - and / = — — - n P F R IR Ir n ° = an? %> = E (1 — F) EIECTRICAL n^ASlBIYC; IWSTItUIflEMTS. The electrical measuring instruments most used in practice are galvanom- eters, resistance boxes, condensers, voltmeters, ammeters, and watt- meters, with variations of the same, such as millivoltmeters, milliammeters, etc. Gal vanometerg. These are instruments for measuring the magnitude or direction of electric currents. The term galvanometer can also be properly applied to the many types of indicating instruments, such as voltmeters and ammeters, where a needle or pointer is under the influence of some directive force, such as the earth's field, a spring, a weight, a permanent magnet, or other means, and is deflected from zero by the passing of an electric current through its coils. Nearly all galvanometers can be separated into two classes. The first is the moving-needle class. A magnetized needle of steel is suspended with its axis horizontal so as to move freely in a horizontal plane. The suspen- sion is by means of a pivot or fiber of silk, of quartz, or of other material. The needle normally points in a north and south direction under the influence of the earth's magnetic field, or in the direction of some other field due to auxiliary magnets. Near to the needle, and frequently surrounding it, is placed a coil of wire whose axis is at right angles to the normal direction of the needle. When a current is passed through the coil the needle tends to turn into a new position, which lies betw r een the direction of the original field and the axis of the coil. The second class is the moving coil or d'Arsonval class. A small coil is suspended by means of a fine wire between the poles of a magnet. Its axis is normally at right angles with the lines of the field. Current is led into the coil by means of the suspension wire, and leaves the coil by a flexible wire attached underneath it. The figure of merit of a galvanometer is (a) the current strength required to cause a deflection of one scale division ; or (6) it is the resistance that must be introduced into the circuit that one volt may cause a deflection of one scale division. This expression for the delicacy of a galvanometer is 22 SYMBOLS, UNITS, INSTRUMENTS. insufficient unless the following quantities are also given : the resistance of the galvanometer, the distance of the scale from the mirror, the size of the scale divisions, and the time of vibration of the needle. The sensitiveness of a galvanometer is the difference of potential neces- sary to be impressed between the galvanometer terminals in order to pro- duce a deflection of one scale division. Movingr-Needle Galvanometers. (a.) The Tangent Galvanometer. If the inside diameter of the coil which surrounds a needle, held at zero by the earth's held, be at least 12 times the length of the needle, then the deflections of the needle which correspond to different current strengths sent through the coils, will be such that the current strengths will vary directly as the tangents of the angles of deflec- tion. Such an instrument is called a tangent galvanometer. It was for- merly much used for the absolute measurement of current. It has, however, many correction factors, some of which are of uncertain magnitude ; and, furthermore, for accuracy in the results yielded by it one must have an exact knowledge of the value of the horizontal component of the earth's magnetism. This quantity is continually changing, and is affected much by the presence of large masses of iron and the existence of heavy currents in the vicinity. Let r = the radius of a tangent galvanometer coil, in centimeters n = the number of turns in the coil, H=. the horizontal intensity of the earth's magnetism, T= the current flowing in the coil in absolute units, and = the deflection of the needle, then Fig. 11. Tangent Galvanometers. GALVANOMETERS. 23 I=z L - Stan 9. 2ttU For convenience the term 2irn . i.e., the strength of the field produced at the center of the coil by the unit of current, is called the constant of the galvanometer, and is represented by G, whence 7— — tan 9 The current in amperes equals 10 I. (b.) Kelvin Galvanometers. The most sensitive galvanometers made are of a type due to Lord Kelvin. Fig. 12 shows one form of this instrument. The moving system consists of a slender quartz rod, to the center of which is fastened a small glass mirror. Parallel to the plane of the mirror, and at one end of the quartz tube, is fas- tened a complex of carefully se- lected minute magnetic needles. The north ends of those needles all point in the same direction. At the other end of the quartz tube is fastened a similar complex with the polarity reversed. Were the two complexes of exactly equal magnetic moment, then, when suspended in the earth's field, no directive action would be felt. In fact, this action is very small. The combination forms what is called an astatic system. Each magnetic complex is in- closed between two wire coils. The four coils are supplied with binding-posts, so as to permit of connection in series or in parallel. Current is sent through them in the proper direction, to produce in each case deflections the same way. Quartz fiber, which ex- hibits no elastic fatigue and which is very strong, is used as a suspension. An adjustable magnet is mounted on the top of the galvanometer. By means of it the directive action of the earth's field can be modified to any extent. Under weak direc- tive force the sensitiveness in- creases greatly, and the period of oscillation of the needle becomes long. The limit of sensitiveness is largely influenced by the pa- tience of the observer. For very precise work the de- flections of the needle are ob- served by means of a telescope and scale. Fig. 13 shows such an instrument. The moving mirror reflects an image of the scale into the objective of the telescope. Continuous work with the tele- scope is apt to injure the eyes, and is certainly tiresome. Where much gal- vanometer work is being done by the same person, a ray of light from a small electric, gas, or oil lamp is so directed as to be reflected from the mirror on the needle upon a divided scale. Such a lamp and scale is shown in Fig. 14. In order to bring the needle quickly to rest when under the in- Fig. 12 e — Kelvin Reflecting Astatic Galvanometer with Four Coils. 24 SYMBOLS, UNITS, INSTRUMENTS. FIG. 13. Fig. 14. GALVANOMETERS. 25 fluence of a current, some method of damping must be employed. One method is to attach a mica vane to the moving system, and allow it to swing in an inclosed chamber which contains air or oil. Sometimes the moving needle is inclosed in a hollow made in a block of copper. The eddy currents induced by the moving needle react upon it and stop its swinging. movingr-Coil Galvanometers. These galvanometers are to be preferred in all cases except where the utmost of delicacy is required. In the most sensitive form, with permanent magnetic field, they can be made to deflect one millimeter with a scale dis- tance of one meter, when one microvolt is impressed between the terminals of the coil. This is sufficient for nearly all purposes. The sensitiveness can be further increased by using an electromagnetic field. The moving-coil i Fig. 15. form of galvanometer has the following good points : its readings are but slightly affected by the presence of magnetic substances in the vicinity, and are practically independent of the earth's field ; the instrument can be easily made dead-beat; and many forms are not much affected by vibrations. Fig, 15 shows a form of D'Arsonval galvanometer of high sensibility. The coil (shown at the right) is inclosed in an aluminum tube. Eddy currents are induced in this tube when the coil swings. They cause damping, and, with a proper thickness of tube, the system may be made aperiodic. Ballistic Galvanometers. Ballistic galvanometers are used for measuring or comparing quantities of electricity such as flow in circuits when a condenser is discharged or mag- netic flux linkages are disturbed. The time of oscillation of the needle 26 SYMBOLS, UNITS, INSTRUMENTS. must in such cases be long as compared with the duration of the discharge. If there be no damping of the needle the quantities of electricity are pro- portional to the sines of half the angle of the first throws of the needle. All galvanometers have some damping. The comparison of quantities of electricity can easily be made with galvanometers of moderate or even Strong damping. Absolute determination of quantity by means of the ballistic galvanometer requires great experimental precautions. (See the Galvanometer, by E. L. Nichols.) Instrument for the Measurement of Alternating- Currents, by E. JF. Horthrup. (Abstract from Trans. A. I. E. E.) The instrument here described was developed to meet the frequent need of means for easily and accurately calibrating alternating-current instru- ments, ammeters and voltmeters, whatever their capacity. (a) It is used as a zero instrument, and does not depend upon any cali- bration or determination of any constant of the instrument; (6) it operates with extreme sensitiveness, and being perfectly "dead-beat" is adapted to work with fluctuating currents; (c) it may be used with or without low. resistance shunts; when used with them it has an unlimited upward range of current measurement; and when used without them its lower range is down to from two to five milliamperes; (d) as the operation of the instru- ment depends upon the heating effect of currents it is wholly independent of wave-form and frequency. Referring to Fig. 16, two small wires, AB, of No. 33 hard-drawn silver wire when shunts are used, lie parallel to each other at a distance of 0.158 in., being held near their extremities by ivory clamps, CC. Each of the ends of the two wires are connected to binding posts through the medium of heavy leads and soldered joints. One face of a small circular disk of ivory, D, rests against the two wires at their middle point, a 0.5-in. circular mirror being fastened to the other GALVANOMETERS. 27 face. Fastened at the center of the ivory disk and half way between the wires, when the disk is in position on the wires, is a small hook. To this, through the medium of a thread, is fastened a small adjustable spiral spring. The small ivory disk maintains its position by friction and the tension of the spring. The wires bend back under the tension of the spring about 0.875 in. from the vertical. The ivory disk does not rest directly upon the wires but bears upon each wire through the medium of a small agate stud shaped like the head of a screw, each wire being in the slot of the agate stud which rests upon it. Tne two ivory clamps holding the wires near their upper extremity are made separately adjustable in a vertical direction by means of thumbscrews which pass through the hard-rubber top of the instrument. Springs s s prevent lost motion when the ivory clamps are screwed up or down. The arrangement of parts above described is supported by a brass frame and a circular hard-rubber top. This frame drops into a circular nickel-plated brass case (Fig. 17). The case has a window in it directly in front of the mirror on the small ivory disk. Fig. 17 shows clearly the arrangement of parts and the appearance of the instrument. By means of the adjusting screws the tension of the two wires may be so adjusted that the plane of the mirror will be vertical to a line drawn in the direction of the spring which holds the mirror against the wires. Now if any elongation occurs in the wire on the right, that side of the mirror will be drawn down or back by the spring, or a deflection to the right is obtained. Like- wise, if an elongation takes place in the wire on the left, the mirror will deflect to the left. If, however, an exactly equal elongation occurs in both wires at the same time, the plane of the mirror will not tilt but simply move back keeping parallel to itself. If the mirror is observed with a telescope and scale, say at a distance of one meter, very minute angular deflections of the mirror will be easily observed, while a sinking back of the plane of the mirror away from the scale will not be observable. Now if an alternating current of unknown strength be sent through the wire A, the wire will elongate, deflecting the mirror toward the left. Pass an adjustable direct current, which can be measured, through the wire B, until the deflection is reversed and brought back to zero on the scale. If when the deflection is zero, and certain precautions to be stated later have been observed, the strength of the direct current is known, the strength of the alternating current will also be known; for it is exactly equal to the direct current. This, however, is on the assumption that equal currents through the wires A and B produce equal elongations of the wires. Pre- viously to comparing the currents, connect the wires A and B in series, and send a current through the circuit; if under these conditions the mirror be not deflected at all, or only slightly, it proves that the two wires are practically equally elongated by the same current strength. The limit of this possible small deflection may be taken as the true zero of the instru- ment. If this zero is maintained under working conditions, it means that the strength of the alternating current in the wire A, is equal to the strength of the adjustable and measured direct current in the wire B. The arrangement of the complete circuits for measuring a large alter- nating current for the purpose of calibrating an alternating current am- meter A is shown diagrammatically in Fig. 18. An important accessory to the instrument is a quick-acting double-throw switch, marked S in the diagram. Wa and Wd represent the two wires of the instrument and m the mirror. R is a low-resistance shunt, preferably of manganin, having 28 SYMBOLS, UNITS, INSTRUMENTS. a negligible temperature coefficient, furnished with tap-off points c and d, between which the resistance R has previously been determined. The ammeter indicated in the diagram will measure from one to two amperes of direct current; r 3 is a slide wire resistance along which a slider p may be moved, thereby varying the pressure difference at a-b from zero to the value of the electromotive force of the storage battery. The points a, b, on the direct-current side of the circuits have leads attached to them which go either to an accurately calibrated direct-current laboratory standard voltmeter, or to a potentiometer. AMMETER VOLTMETER OR POTENTIOMETER Fig 18. When the instrument is installed, a permanent adjustment of the re- sistances at any convenient temperature of the wires and leads must be made as follows: (see Fig. 18.) The resistances, 9 to 10 = 7 to 8, 10 to 1 -f- 9 to 5 = 8 to 4 + 7 to 2 and 2 to c + 4 to d — 3 to a + 6 to b. Thus while this gives the over-all resistance from a through the wire Wd to b equal to the over-all resistance from d through the wire Wa to c, the different portions of the circuit must be matched in resistance as stated above. When the switch S is closed on the alternating-current side the two wires Wa and Wd are thrown in parallel, and the two parallel-connected circuits have the same resistance, by construction, and that to these par- allel circuits at the points 2 and 4 is applied the same potential difference, this potential difference being the drop on the low resistance R carrying the alternating current. The drop over R, inasmuch as it is a low resis- tance, is only slightly lowered by being shunted by the two wires of the instrument and their leads, and this iowering of the potential is not appre- ciably greater when the two wires in parallel shunt the resistance R than when only one wire with its leads shunts the resistance. Disregarding the slight lowering of the potential, both wires will now have passing through them equal currents, each current being nearly the same as would pass through the one wire Wa if the switch S were open, and only this wire could receive current. . With the resistances of the parallel circuits correctly adjusted to equality, both wires will get equal currents, both will elongate equally or very nearly so, and the mirror m instead of rotating will move back, maintaining its plane parallel to the position which it has with no current passing. When the switch S is thrown to the direct-current side, the potential drop over the resistance R is now applied to the wire Wd only; and the direct potential difference between the points a and b is applied to the wire Wd. GALVANOMETERS. 29 This drop between a and b can be varied by the slider p and measured by a voltmeter or potentiometer applied at a, b. The ammeter gives the current taken by the wire Wd. The shunt resistance R may be designed to carry any current, however large. The same resistance R, or a combination of resistances, may be designed with several tap-off or potential points, so that the instrument may always have approximately the same potential applied to its alter- nating-current side, whatever the strength of the current to be measured. This potential drop is best made between 0.25 and 0.5 volt. The neces- sary drop of potential being so low, the energy dissipated in the shunts is small, and therefore they may be of very moderate size. It is also easy to make them practically non-inductive. { Galvanometer Shunt Boxes. It is often desirable to use a galvanometer ot high sensibility for work demanding a much lower sensibility. Again, it may be convenient to cali- brate a galvanometer of low sensibility, while it would be inconvenient to calibrate a more sensitive one It is therefore useful to be able to change the sensibility in a known ratio. Convenience dictates that sim- ple ratios be used, and those almost universally taken are 10, 100, and 1000; that is J,^,or ft § B , AAMAA- Fig. 19. ■ -f- 1 = the Multiplying power of the shunt. part of the current flowing is allowed to go through the galvanometer while the remainder is diverted through a shunt. In Fig. 19 let G == the resistance of the galvanometer, and S = the resistance of the shunt, then the joint resistance of the two is ■ • (jr -f- b If I = the total current flowing in the circuit, and if L = the part flowing through the galvanometer, then 1^ _ G + S _ G I, ~~ S ~ S The resistance of a shunt which will give a certain multiplying power, n, is equal to • Fig. 20 shows a form n — 1 of shunt used with a galvanometer, al- though it is perfectly feasible to use an ordinary resistance box for the purpose. Messrs. Ayrton & Mather have developed a new shunt, which can be used with any galvanometer irrespective of its resist- ance : following is a diagram of it. A and B are terminals for the galvano- meter connections. B and C are the in- going and outgoing terminals for battery circuit. To short circuit G, place plugs in j and f. To throw all the current through G, put a plug in f only. To use the shunts, place a plug in h, and leave it there until through using. In this method it is not necessary to know the resistance of either G or r. The shunt box can therefore be used with any galvanometer. Temperature variations make no differ- ence, provided they do not take place Fig. 20. during one set of tests. The resistance r may be any number of ohms, but in order not to decrease the sensibility too much r should be at least as large as G. The resistance r is divided for use as follows : permanent attachments to the various blocks are made at points in the coil corresponding with l I The method of mounting is seen in Fig. 48. The current coils are gener- ally made up of thin copper strip. At P P' are the pole-pieces, built of laminated iron, for concentrating the effect on the strip. There are two similar vibrating sets separated by an iron partition in the center, thus form- ing two different oscillographs which are quite independent of each other; even three sets can be mounted in this way. The oil tube T containing the mirror may be slid up and down by the lower screw v and may be turned hori- zontally by the endless screw V. On the left is seen the complete mounting; the front coil has an elliptical opening to allow the light to pass. At M is an adjustable mirror which gives a perma- nent spot of light to form the, base line of the curves. By using the vibrating band, a period of 50,000 vibrations per second has been reached, representing the oscilla- tion period proper to the band. In this case the instrument is sufficiently sensitive, although it may be made much more sensitive by using a band having 15,000 to 20,000 vibrations, which will answer in most cases where the wave forms are not too irregular. The sensitiveness in the latter case answers to a displacement of the spot of light of 100 millimeters per ampere on a screen one meter distant. The use of soft iron pole-pieces to concen- trate the effect gives a high magnetic intensity to the band, and in fact it is generally brought to saturation owing to its small volume. It is found that the band has an advantage in being saturated. The sensitiveness increases at first while the band is not yet sat- urated, then decreases when the mag- netization of the piece increases less rapidly than the field strength. The number of vibrations continues to in- crease, rapidly at first, then slowly, as the band becomes saturated. The re- sults depend in a great measure upon the quality of the iron used for the band. The mirrors must be very small and light when mounted on such a thin strip. They have now been reduced as low as 0.2 millimeter wide and 0.5 millimeter high, with a thickness of but 0.05 to 0.1 millimeter. Silvered glass or mica is used, and the mirrors are fastened to the bands with shellac before the latter are mounted. As the band is enclosed in an oil box it is free from rust and well protected. The sensitiveness of the instrument may be greatly varied by using an iron yoke which is placed against the poles of the perma- nent magnet and acts as a shunt to diminish the strength of the field at the poles. To the right of the box will be seen the arrangement of the oscillating mirror which gives the to-and-fro motion to the spot of light in order to form the wave. The device will be understood by the diagram, Fig. 49. *S is an arc lamp which throws a beam of light by means of the lens X and shutter F upon the mirror of the oscillograph n; this beam is then reflected and passes through the lens /, falling on the oscillating mirror m placed behind it. The latter is given a to-and-fro motion by a small synchronous motor. The beam of light thus far has two movements, one by the mirror n of the oscillograph and the other by the mirror m, and the resultant of the two gives the wave form which is projected above on the ground-glass screen P. The to-and-fro movement of the mirror is obtained by a cam fixed to the motor-shaft. During two complete periods of the wave the mirror must be moved at a continuous Fig. 47. Blondel Oscillograph, snowing Method of mounting Vibrating Band. DETERMINATION OF WAVE FORM. 54a « Fig. 48. Blondel Oscillograph, showing the Arrangement of the Magnet o ""tt^lria lU? .— ~---s Fig. 49. Diagram showing the Arrangement of the Apparatus in the Blondel Oscillograph. 54b symbols, units, instruments. rate from top to bottom, and during the next period it must be able to return so as to continue the movement (as will be ^ noticed on the photograph two complete r\jk /Xw waves are thrown on the screen). This is / X \ / A \ carried out by the profile of the cam which is / / \ \ / / \\ sucn tnat tne m i rror nas a uniform move- \~\/~7 \ \J ment during two cycles of the wave, and the \ /\f \ X next cycle is occupied by the return of the ^ ^ mirror (during this time an electrically oper- Fig. 50. ated shutter placed at F cuts off the light), so that the eye perceives only a continuous trace of the wave. To observe phenomena which are not periodic the motor is replaced by a pendulum device. MEASUREMENTS. Revised by W. N. Goodwin, Jr., and Prof. Samuel Sheldon. ELEXEXTARl' LAWS OF ELECTRICAL CIHCOI.I. Ohm'i law is the fundamental law of electrical circuits and is expressed in the following equations. R E = IR *=i where / = Current strength in amperes, R = Resistance in ohms, E = Electromotive Force in volts. The conductance of a conductor is the reciprocal of its resistance, and the unit is called a mho, so that Ohm's law may be stated as follows: I -EG where G =s conductance in mhos. multiple Circuits. — The conductance of any number of circuits in parallel is equal to the sum of the conductances of the individual circuits, which is, as stated above, the reciprocal of their resistances. The combined resistance then is the reciprocal of the conductance thus found. Thus in Fig. 1, if r and r\ be two resistances in parallel, the combined resistance = - : — — A— • r n The joint resistance of any number of resistances in parallel as a, b, c, and d is - - ■ a o c a Current in a multiple Circuit is divided among the separate circuits in direct proportion to their respective conductances, or inversely as their resistances. In Fig. 2, the total resistance of circuit ( r \ total current ' and i =■ r-\-n E (r + n) En ' Rr + Rr\ -{-rri Er Rr -\r Rr\ + rri Rr + Rri -f m Fig. 2. KIRCHOFF'§ LAWS. First Law, — If in any circuit a number of currents meet at a point, the sum of those flowing toward that point is equal to the sum of those flowirig away from it. Second law, — In any closed circuit, the algebraic • _ sum of the products formed by multiplying the re- sistance of each part by the current passing through it is equal to the sum of the electromotive forces in the circuit. By means of these, laws, the current in any part of an intricate system of conductors can be found if the resistances of the different parts and the electromotive forces are given. Thus in Fig. 3, according to the first law i = tj + i 2 and from the second law i=i x + i 2 and from the second law E = i^ and iir 2 = tirj. Fig. 3. From these three formulae, the three unknown currents can be deduced The same method can be applied to more complex circuits. 55 56 MEASUREMENTS. RESISTANCE MEASIJRIIIENIS. Substitution Method. — This is the simplest method of measuring resistance. The resistance to be measured is inserted in series with a galvanometer and some constant source of current, and the galvanometer deflection noted; then a known adjustable resistance is substituted for the unknown and adjusted until the same deflection is again obtained, then this value of the adjustable resistance is equal to that of the resistance to be measured. , Differential Galvanometer Ifletliod. — In galvanometers having two coils wound side by side, separate currents sent through them in opposite directions exert a differential action on the movable system. In a diner- ential galvanometer the two coils are equal in their magnetic action on the movable system for equal currents, so that equal currents sent through them in opposite directions will not deflect the needle. If tne currents are unequal, then the deflection is a measure of then difference. This form of galvanometer may be used to measure resistance by inserting the unknown resistance in circuit with one coil of the galvanometer and a known adjustable resistance with the other, both circuits being connected in multiple. Then when the resistance is adjusted until no deflection is produced the resistances in the two circuits are equal. The method is often used in the comparison of the conductivity of wire, and where rapid measurements not requiring great accuracy are desired. Wheat stone's Bridg-e. — For accurate measurements of resistance the Wheatstone Bridge method is almost universally used; Fig. 4 is a dia- gram of the connections in which a, o, and ti are known resistances and x the unknown resis- tance to be measured. G is the galvanometer, and B is a battery of several cells, the number . 4 of which may be varied according to the value k of the resistance x. R is adjusted until there is no deflection of the galvanometer needle when both keys are closed. The battery key should always be closed be- fore the galvanometer key is depressed or there will be a "kick" in the galvanometer due to the v A self inductance or capacity of the circuit under * IG * 4 * test. When a balance is established — =-, or x = R -• R a a The resistances a and b are, in practice, made even multiples of 10, so that x can be read directly from R, the proper number of figures being pointed off decimally. If a = b the value of x is the same as R. If x be greater than the ca- pacity of R, or low in comparison to it, then a and b must be so chosen that their ratio respectively multiplies or divides R. For example, let a = 10 ) 7, 1000 b = 1000 then a> = - , R = ^r X 243 = 24 ' 3( *°- ft = 243 ) a 10 The ratio of a to b being 100, any reading as R is multiplied by 100, or again let a — 1000 ) in b = 10 then x - -r^rz X 243 = 2.43. R — 243 J 100 ° The ratio of a to b being x fo, any reading as R should be divided by 100. A commercial form of Wheatstone Bridge of the Weston Model is shown diagrammatical ly in Fig. 5. This type, called the "plug in" type, or some modification of it, is most commonly used. It has the advantage over the 44 plug out " type in that fewer plugs are required, there being but one plug needed for each decade; this reduces the plug error to a minimum. RESISTANCE MEASUREMENTS. 57 Direct Reading* Ohmmeter. — Another form of instrument used for measuring resistances is known as the direct reading ohmmeter. Briefly described it is simply a slide wire bridge, the wire forming two of the arms of the bridge, a known resistance a third arm, and the unknown resistance Hllllh _ .<§) Ba (§>- THOUS. HUNDS. TENS UNITS R jW\AA (AVVWN ^VWWV) fVWWV| ill xi bo x ly* ih> |W/VW i > •? ? p* * 4 sQr 4 sOr 4 C) 6 4 G° {a? 3 K?' »Ep- »Q t %53 7 sS? 2 sOr a vOl? 8 2 sCv 8 2 vLv 8 s£-v ) 9 j < pAAVW Qio oQ Qio oQ Qio oQ Qio_ Qioo(^ idELSL.-! ^F fi..;n. j !Ba — - P Ga < X l AWVW — I L* Fig. 5. the fourth. The slide wire is graduated to read directly in ohms, and is printed with numbers in black and red. The black numbers refer to a low reading scale which is used when the single plug of the instrument is fitted into the hole marked black, and the red numbers refer to a higher scale Fig. 6. Fig. 7. Fig. 6 shows diagrammatically the connections of this Ohmmeter, and Fig. 7 gives the same ones expanded into the conventional Bridge Form. when the plug is inserted in the hole marked red. This instrument usually has four scales, although it is sometimes made with three and five. The slide wire is doubled back on itself by means of a heavy cross block of practically zero resistance. The detector circuit comprises a detecting instrument ordinarily a tele- phone receiver, and a stylus, which is touched at various points along the 58 MEASUREMENTS. slide wire until the detector by silence indicates a balance, when the result is read directly in ohms. In some of the instruments the battery is equipped with a small induction coil which provides alternating current. In this form the instrument is useful for measuring electrolytic resistance and other resistances containing electromotive forces that may be developed by the presence of current therein, and by the use of a suitable condenser in place of the known resistance, capacities can be compared. Directions for Use of j§ag*e Direct Reading* Ohmmeter. — To Measure Resistance. Connect the terminals of the circuit to be measured to the posts, A and D. Place the telephone receiver to the ear and close the battery key, K, located in the receiver. Hold the stylus, S, in the hand in the same manner as a pencil, and with it touch the straight wires along their entire length until a point is reached where gently tapping the stylus on the wire produces no sound in the telephone. The resistance sought is then that indicated by the scale under that point of the wire. During these readings the plug, P, must be in one of the sockets at the right-hand end of the rubber cross-bar. When in the socket marked ••red" the scale numerals printed in red should be used. When in the socket marked "blue" the blue numbers should be read, etc. Slide- wire Bridgre. — A very convenient form of bridge for ordinary use where extreme accuracy is not de- manded is the slide-wire bridge, shown in Fig. 8. It consists of a wire 1 meter long and about 1.5 mm. diameter stretched "2a ^~/x ~^ v ~\^ N ~\ parallel with a meter scale divided into ( r~\ 4/ ^ \ iv millimeters. A contact key is so arranged p / ip ii is — 2 — °i i° — z n \ < as to be moved along the wire so that <~A<\ | o ie 20 30 40 so Co 7o 8 o | 4^ P -=— contact with it can be made at any point. A known resistance R is connected as shown; x is the unknown resistance; the galvanometer and the battery are con- nected as shown in the figure; after closing the key k t the contact 3 is then moved along the wire until the galvanometer needle returns to zero; Fig. 8. then again; and a : b :: R : x, bR The Carey- Foster IfCetnod. — For the very precise comparison of nearly equal resistances of from 1 to 100 ohms this method yields exquisite results. In Fig. 9, Si and S 2 represent the two nearly equal resistances to be compared, and R t , R 2 represent nearly equal resistances, which, for best results, should not differ much in magnitude from Si and S 2 . Si and S 2 are connected by a slide wire whose resistance per unit length p is known. The battery and galvanometer are con- nected as in the diagram. A balance is obtained by moving the contact c along the stretched wire. Suppose the length of the wire on the left-hand side to the point of contact to be a units. Then exchange S t and S 2 for each other without alter- ing any other connections in the circuit. Upon producing a new balance, let a x be the length of wire to the left of the contact. Fio. 9. Carey-Foster Bridge. Then Si = S 2 + (« — ai) p. Special commutators are upon the market which have for their purpose the easy exchange of Si and S 2 . To avoid thermal effects, which are quite considerable with resistances made of some materials, the battery should be commutated for each position of the resistances to be compared. The readings for the two balances ac- companying the battery commutation should be averaged. RESISTANCE MEASUREMENTS. 59 Measurements of Low Resistances. Kelvin's Doable Bridge. — If a Wheatstone bridge be used to compare re- sistances having a value much less than one ohm, the terminal and contact resistances produce a considerable error in the re- sults. In conductors having such low resistance, the value of the resistance given or to he measured is considered as ly- ing between two definite points. In standard resistances these points are connected to two ter- minals called potential terminals. Kelvin has designed a modified form of Wheatstone bridge in which the above-mentioned errors are eliminated. The method is shown diagrammati- cally in Fig. 10, in which R and x, the resistances to be compared, lie between S and Si on one and between T and T\ of the other, and are connected together at y; n and o are auxiliary resistances also adjustable. A galva- nometer is connected through a key, as shown, to two points, one at the junction of n and o; the other at the junction of a and b. If n and o be so adjusted that n: o: : R: x, and o and b be adjusted so that the galvano- meter is balanced, then a :b : : R : x, bR or x = i Fig. 10. Kelvin's Double Bridge. In practice, n and o may be changed during the adjustment of a and b so as to maintain the ratio of n to o the same as that of a to b, either by changing n and o, on standard rheostats, or by opening the circuit at y and adjusting n and o, as in a regular bridge, for a balance after each trial value of a and b; then when a balance is obtained in the galvanometer with circuit at y both open and closed the above equation holds good. Another UKethod for Comparison of Ion Resistances. — For comparing the resistances of ammeter shunts, etc., with standard side terminal resistances of the Reichsanstalt form, the method of Sheldon yields very accurate results. The unknown resistance x, Fig. 11, which may be as- sumed to he supplied with branch po- tential points a 5, is connected by heavy conductors in series with a standard re- sistance R, having potential points c d. From the two free terminals T T 1 of these resistances are shunted two 10,000 ohm resistance boxes S P, adjusted to the same normal temperature, and wound with wire of the same or negli- gable temperature coefficient, and con- nected in series. From the point of connection e, between the two boxes, connection is made to one terminal of the galvanometer g, the other terminal being connected successively with the potential points a, 6, c, and d. At the outset all the plugs are removed from the box S, and all are in place in the box P. After connecting T and T 1 with a source of heavy current, plugs are transferred from one box to the corresponding holes in the other box (this keeps the total resistance in the two boxes constant) until no deflection is observed in the galvanometer. This operation is repeated for each of the potential points a, b, c, and d. Rep- resenting the resistances in the box S on the occasion of each of these bal- ances by Sa, Sb, Sc, and Sd respectively, we have the following expression for the value of the unknown resistance : Fig. 11. Precise Measurement. . Sa — Sb ' Se — Sd R. 60 MEASUREMENTS. Note. ■ — Mr. E. F. Northrup gives the following formula as handy in determining the percentage conductivity of metal wires. This conductivity is generally^ expressed as a certain per cent conductivity of Matthiessen's standard. To determine the conductivity, a resistance R of a sample is usually determined at a temperature 20° C and of a length I. From this measurement the per cent conductivity may be expressed as follows: Percentage conductivity = - where ' R20XW X 581,054* I — length in centimeters, _ W =z weight in grams, d = specific gravity. : resistance in ohms at 20° C, RE§I§TAIICE OF OALVA^OMETER§. When a second galvanometer is available, by far the most simple and sat- isfactory method is to measure the resistance of the galvanometer by any of the ordinary Wheatstone's bridge methods. Take the temperature at the same time, and, if the instrument has a delicate system, remove the needle and suspension. Hall Deflection method. — Connect the galvanometer in series with a resistance r and battery as in the following figure. r Note the deflection d ; then increase r so that the new deflection d x will be one-half the first, or - = d x ; call the new resistance r\ ; then Resistance of Galvanometer = 7\ — 2r. If the instrument be a tangent galvanometer, then d and d t should represent the tangents of the deflec- tions. Kelvin's Method. — Connect the galvano- meter, as a? in a Wheatstone's bridge, as in Fig. 13. Adjust r until the deflection of G is the same, whether the key is closed or open. G = r b -. a The result is independent of the resistance of the battery. The battery should be connected from the junction of the two highest resistances to that of the two lowest. Fig. 12. Fig. 13. RESISTANCE OF BAITERIi!§, I — vV&V-^A. Condenser Method. — For this test is needed a condenser C, a ballistic galvanometer G, a double contact key k x , a resistance R, of about the same magnitude as the supposed resistance of the battery B, and a single contact key k 2 . Connect as in the following figure. With the key k 2 open, press the key fc lf and observe the throw 9 1 in the galvanometer. Then, after the needle has come to rest, with key k 2 closed, repeat the operation observing the throw 2 . Then the resistance of the battery x = R d -±^- B, Reduced Reflection Method. — Connect the battery B in circuit with a galvanometer G and a resist- ance r as in Fig. 15. Note the deflection d. and then in- crease r to r 1 and note the smaller deflection d x ; then, if the deflections of the galvanometer be proportional to the currents, rydi — rd d — G. Fig. 15. If r x is such that d } = - then B ■(2r+G). RESISTANCE OF HOUSE CIRCUITS. 61 The E.M.F. of the battery is supposed to remain unaltered during the measurement. Mance's IfEetbod. — Connect the battery as x in Wheatstone's bridge as in Fig. 16. Adjust r until the deflection of G is the same whether the key be closed or open. Then B = r-' a The galvanometer should be placed between the junction of the two highest resistances and that of the two lowest. Resistance of Battery while Working-. — Connect the battery B with a resistance r, and also in parallel with a condenser C, galvanometer G, and key k ; shunt the battery through s with key k] , as in Fig. 17. Close the key k, and note the deflection d of the galvanometer, keeping k closed, close k x and note d u the deflection in the opposite direction. Then the battery resistance ( B = i d — dy- dis If r be large, the term dis is negligible, and Bz=i d — d^ » being the multiplying power of the shunt. Workshop Method, Applicable as well to Dynamos. — With dynamo or battery on open circuit, take the voltage across the terminals with a voltmeter, and call it d ; take another reading d± at the same points with the battery or dynamo working on a known resistance r : then the in- ternal resistance R z= — - — - r. In the case of storage batteries, if the current / be read from an inserted ammeter when charging, the resistance of the battery is and when discharging B = — j^-± . RESISTANCE OE AERIAL LIAES OR HOUSE CIRCUITS. Conductor Resistance. — When the circuit has metallic return, it is easily measured by any of the Wheatstone's bridge methods, or, if the circuit conductor can be supplied with current through an ammeter, then the fall of potential across the ends of the con- ductor will give a measure of the resistance by ohms law, viz., Resistance = drop in volts current "^Earth — Earth ~ Fig. 18. If the circuit has earth return as in tele- graph and some telephone circuits, then place far end of the line to earth, and con- nect with bridge as in Fig. 18. Then the total resistance x of the line and earth, is x — r — . If a second line be available, the resistance of the first line can be deter- mined separated from that of earth, as well as the resistance of earth. 62 MEASUREMENTS. Let r = resistance of first line, r\ = resistance of second line, r 2 = resistance of earth. First connect the far end of r and r x together, and get the total resistance R; connect r and r 2 , and measure the resistance R x , connect r x and r 2 , and get total resistance R 2 . Then if R±Rj_±Rt ~ 2 r = T — R 2 , n=T-R lt r 2 —T — R. This test is particularly applicable to rinding the resistance of trolley wires, feeders, and track. For other methods for resistance measurements see under "Tests with Voltmeter." Ro MEASUREMENT OF ELECTROMOTIVE FORCE. Of Batteries. — This can usually be measured closely enough for all practical purposes by a high class low-reading voltmeter (see Tests with a Voltmeter). Wheat»tone V Method. — Connect the cell or battery to be compared in circuit with a galvanometer and high resistance r, and note the deflec- tion d; t^ien add another high resistance r x (about equal to r), and note the de- flection d x . Next, connect the cell with which the first is to be compared in cir- cuit with the galvanometer, and connect in resistance until the galvanometer deflection is the same as d; then add further resistance R until the galvano- meter deflection is the same as d x ; then, if e equals the E.M.F. of the first cell, and E equals the E.M.F. of the cell with which it is compared, n : R : : e : E, _L and n Fig. 19. Or, the electromotive forces are pro- portional to the respective resistance which must be added to reduce the deflection the same amount. Lumsden's Method. — The two cells E x and E 2 to be compared are arranged as shown in Fig. 19. R x and R 2 are adjustable resistances which are large as compared with the resistances of the cells. R x and R 2 are changed until the deflection in the galvanometer is reduced to zero. Then Ei _ Ri E t R 2 If greater accuracy be required than that obtained by the above methods, some potentiometer method may be used, in which the cell to be measured is compared directly with a standard cell. Lord Rarleigrli'g Compensation Method. — In the following diagram let R and R x be two 10,000-ohm rheostats, B be the battery of larger E.M.F. than either of the cells to be compared, B x be one of the cells under test, G be a sensitive galvano- meter, HR be a high resistance to protect the standard cell, and k be a key. Obtain a balance, so that the galvanometer shows no deflection on closing the key k, by trans- Fig. 20. MEASURING CAPACITY. 63 f erring resistance from one box to the other, being careful to keep the sum of the resistances in the boxes equal to 10,000 ohms. Observe the resistance in R ana call it R x . Repeat with the other cell B 2 , and call the resistance R 2 . Then the E.M.F.'s of the two cells E\\ E% = R\\ R%. Note. — Special boxes are on the market which automatically change the resistances R and Ri, maintaining the sum of the resistances constant, the value of the resistance being read directly from the dials. Direct Reading* Potentiometer. — There are many forms of po- tentiometers available, which are used in connection with a standard cell, and on which the potential difference to be measured is read directly from the switch dials of the instrument when it is balanced as shown by a gal- vanometer. Such potentiometers generally read to 1.5 volts. To meas- ure higher voltages than this a volt box must be used, which is simply a high resistance, across which the voltage to be measured is connected. Connections are brought out from the resistance so as to include a known portion of it, having such a value that the potential difference across it will be less than 1.5 volts. This is then measured on the potentiometer, and the value found multiplied by the constant of the volt box. Measurement of Current Uy Potentiometer. — The current to be measured is passed through a standard low resistance, say, .01 or .001 ohm, and the difference of potential across its potential terminals meas- ured by means of a potentiometer. Then the current is by Ohm's law '-§ where E is the difference of potential as measured, and R the resistance of the standard. i HTEASURI1VG CAPACITY. Arrangement of Condensers. In Parallel. — Join like poles of the several condensers together as in the figure ; then, the joint capacity of the set is equal to the sum of the several capacities. Total capacity = c + c, + c n + c,,,. Condensers in Series. — Join the unlike poles as if connecting up battery cells in series as in Fig. 22, then the joint capacity of all is the Fig. 21. reciprocal of the sum of the reciprocals of the several capacities 1 Capacity C: i + I + A c c. C// +i Fig 22. Capacity l>y Direct Discharge. — Charge a standard condenser, Fig. 23, C« by a battery E for a certain time, say 30 sec- onds ; then discharge it through a ballistic galvanometer G ; note the throw d. Next charge the condenser to be measured, C lt by the same battery and for the same length of time, and discharge this through the same galvanometer noting the throw d x ; Then CkiC x iid: d x . and <\- For Kelvin's and Gott's methods see pages 326-327, " Cable Testing." 64 MEASUREMENTS. Bridg-e OTethod. — For comparing the capacities of two condensers, Ce and C, which are approximately the same, connect as in Fig. 24 through two rather high inductionless resistances i? t and JR 2 to the key k which makes and breaks contacts at each end. E is a bat- tery. A galvanometer is inserted between the ends of the condensers where they join the resistances. Adjust the resist- ances so that no deflection results when the key is manipulated. Then C=C.|l. Fla *• Mo. JjOS» of Potential Ifletnod. — The capacity of a condenser may be determined by the following formula: c = { — a 2.303 R log - e where C is the electrostatic capacity, in microfarads, of a condenser, the potential of whose charge falls from E to e when it is discharged during t seconds through a resistance of R megohms. If C is the known and R the unknown quantity, then R = * 2.303 k log - e In measuring the insulation resistance of a short cable by this method, the discharge deflection E, compared with the discharge deflection obtained with the same battery from a standard condenser, would give the value of k. For long cables, however, this does not give correct results, and the ca- pacity must be determined by other methods. JEJLECTJt0^i:ACJXJETJ.C iADlCTIOX, .Law of Induction. — When the magnetic induction or flux inter- linked with an electrical circuit is changed in any manner, an electro- motive force is induced in that circuit which is proportional in amount to the rate of change of the flux, and acts in a direction which would, by producing a current, tend to oppose that change. Symbolically expressed the induced electromotive force in volts is n d € ~ 10 8 dt 9 where is the magnetic flux through the circuit, n the number of turns of wire, and t the time. Self-induced electromotive forces are those induced in a circuit by change in the current in the circuit itself. Coefficient of ftelf-Ind uction. — The practical unit of self-induction is the henry, and is equal to 10 9 absolute units. The self-induction in henrys of any coil or circuit is equal numerically to the electromotive force in volts induced by a current in it changing at the rate of one ampere per second. Thus the electromotive force in volts pro- duced in a circuit by a varying current is T d i e =- L it' where L is the self-induction in henrys and i the current in amperes. If fa = n, represent the flux turns in the circuit, then fa = Li X 10 8 . Foi example, if a coil have 150 turns of wire, carrying a current of two MEASUREMENT OF COEFFICIENT INDUCTION. 65 amperes, producing 200,000 lines of force, or 200 kilogausses through it, the flux turns equal 200,000 X 150 = 30,000,000, and the self-induction is therefore i 30,000,000 L = Wi = 2X100,000,000 = - l0 henry ' If the current of 2 amperes die out uniformly in one second, then the electromotive force induced is i = L : .15 X2- .30 rolt. Coefficient of Self-induction of a Long 1 Solenoid. L ~ 10 9 when the permeability is unity. Where n =: total number of turns of wire, n 1 = number of turns per centimeter length, A = area of cross section of solenoid. For magnetic substances the above equation must be multiplied by fx, the permeability of the medium. measurements of The Coefficient of Induction. Comparison with Known Capacity. — The coefficient of self- Fig. 25. Fig. 26. induction may be determined by means of a Wheatstone bridge as follows: Let A and B, in Fig. 25, be the bridge ratio arms, Ri the adjustable rheostat. Connect the circuit to be measured as RL in series with a variable non-inductive resistance r and r t a portion of which r t is shunted by a standard condenser of capacity C. First balance the bridge for steady currents by adjusting R x , that is, when the key K is closed continuously. Then alter the proportion of non-inductive resistance r t , shunting the condenser until no deflection occurs in the galvanometer when the key K is open and closed. Then the self-inductance L = Crf. Comparison witn Know n Self-Inductance. — Arrange in form of bridge as shown in Fig. 26, L being the unknown and L t the standard self- inductance. Adjustable non-inductive resistances are connected in series with them. Call the resistances in each arm R and R lf A and B are non- inductive resistances. First adjust to a balance for steady currents by changing R and R lt then adjust A and B until no throw of the galvano- meter is observed when the galvanometer key is closed before closing the battery key. Then R and R t must be again adjusted for steady currents, 66 MEASUREMENTS. and so on until a balance is obtained for both steady and transient current!. L\ B R\ Then Fig. 27. Ayrton and Perry's Variable Standard of Self-induction. If L x be one of Ayrton and Perry's adjustable standards of self-induction (see Fig. 27), then the bridge can be balanced in the regular way for steady current, and for transient currents by varying the self-induction stand- ard. As shown in the illustration this instrument consists of two coils wound on sections of con- centric spherical surfaces, the in- side one of which can be rotated with reference to the outside one, and thus their coefficient of in- duction varied without changing their resistance. The scale is divided to read in millihenrys on one side and in degrees on the other. Its range is approximately from 3.5 to 42 millihenrys. Telephone Method. — A modification of the above, which is quicker and more practical, is by using a telephone in place of the galvanometer, and a source of alternating or rapidly interrupted direct current for the battery, _ shown in Fig. 28. The part ab is a slide wire with telephone contact at K; the self-inductances L and L\ are connected as in the previous m«tnod with adjustable non-inductive resistances. S is a source of alternating current. The return circuit should be run parallel and close to the slide wire to reduce inductive errors. The contact K is moved along the wire and placed in a position where the minimum sound is heard in the telephone and R and Ri are changed to reduce this sound to a lower minimum. These oper- ations are repeated until finally a point is reached where the minimum of sound is very sharply denned or silence occurs. Then £=£• Fig. 28. Measurement of Self-Inductance with an Alternating- Cur- rent of Known frequency. For this test is needed a high resistance or electrostatic alternating cur- rent voltmeter, a direct current ammeter, and a non-inductive resistance. Connect as in Fig. 29, where R x is an inductive resistance to be measured, and S a switch for short-circuiting the ammeter; the A. C. dynamo of fre- quency n is so arranged that its terminals may be disconnected, and a battery be substituted therefor. With the connections as in Fig. 29, close the switch S, and take the drop with the voltmeter from a to 6 and the drop from a to C; then disconnect the A. C. dynamo, and connect the battery B; open the switch s, and vary the continuous current until the drop from a to C is the same as with the alternating current, both measurements being made with the same volt- meter; then note the current shown by the ammeter, and measure t! e drop from a to 6 with the voltmeter. Call the drop across R x from a to \ with MEASUREMENT OF MUTUAL INDUCTANCE. 67 alternating current, E, and the same with continuous current, E lt and the reading of the ammeter with the latter, I. Then L = \/E 2 - E x 2 2-nnl If the resistance R t be known, and the ammeter be suitable for use with am £>* c r a ( rrrrrrrrrr* HI Load. Fig. 29. alternating currents, the switch an d non-indu ctive resistance may be dis- pensed with. We then have L = — }— — . where I x is the value of the 2-nnl alternating current. Note. — The resistance of the voltmeter must be high enough to render its current negligible as compared with that through the resistance R t . Measurement of Mutual Inductance. Connect the two coils whose mutual inductance is to be determined, first in series and then in opposition to each other. The self-induction of each combination is then measured by any suitable method. Let M = the mutual inductance between the two coils. L = the self-inductance of one coil. L y = the self-inductance of the other coil. L n = the self-inductance of both coils in series. Z//// = the self-inductance of both coils in opposition. Then since L u = L-\-L,-\-2M and L n , = L + L, - 2 M . Then the coefficient of mutual inductance desired is M = L„ - L„, Fig. 30. Comparison with a Known Ca- pacity. — Connect as shown in Fig. 30 where A and D are two coils whose mutual inductance M is required. R and R x are two adjustable non-inductive resistances and C a standard condenser placed in shunt to R and R lm Vary the resistances R and R* until no deflection is observed on the galvanometer when the key is opened or closed. Then the mutual inductance is M = CRR X . 68 MEASUREMENTS. Comparison with Known Self-1 nduction by Bridge. — In this method the mutual inductance of two coils is compared with the knowij self-inductance of one of them. The coil whose self-inductance is knowf is connected as R in Fig. 31. The othel coil is connected in the battery circuit witfc its magnetic circuit opposed to that of th< other coil. Then by adjusting the othel arms of the bridge to a balance for botli steady and transient currents, as in tht methods for self-inductance, the mutual inductance is r + ri Another Method. — In order that a balance may be obtained without the incon- Fig. 31. venience of trial and approximation as in the foregoing method, the battery circuit may be shunted by non-inductive resistance as S shown in Fig. 32. The other connections are similar to those of the previous test. The bridge is first balanced for steady currents in the regular way by adjusting the resistances R u r, and r x , and then S is changed until no deflection occurs when the key is opened or closed. Then the mutual inductance is M LRxS M ~ (R 1 + R)S + (R + r)R 1 Comparison of Mutual Inductance with Known Self-In- ductance of Another Coil. — Connections are made as shown in Fig. 33. One of the two coils whose mutual inductance is to be measured is con- Fig. 33. nected in the battery circuit, and the other in series with an adjustable non-inductive resistance as a shunt to the galvanometer. The known self-inductance L is connected in the bridge as R. The bridge is first balanced, as before, for steady current, then the resistance S is changed until no deflection occurs when the key is opened or closed. Then if S be the total resistance in the shunt circuit, the mutual inductance is M LR,S (R + Ri) 2 ' Telephone Method. — As in measurements of self-inductance, a tele- phone may be used in measurements of mutual inductance, as shown in Fig. 34. The coil of known self-inductance L is connected in one arm of the bridge, as shown at R. The other coil is connected in opposition to that coil in the main current circuit, the current supplied being either alternating or a rapidly interrupted direct current. The non-inductive resistance and the telephone circuit contact are varied until silence occurs in the telephone in a manner similar to that described for self-inductance. MEASUREMENT OF A.C. POWER. 69 Then if p is the resistance of the slide wire for unit length, and the position for a balance is a units from the right as shown, then the mutual inductance Secohmmeter. - In measurements of inductance, when balancing for transient currents the galvanometer deflects in one direction when the battery key is closed, and in the opposite direction when it is opened, lo increase the sensibility of such tests, Ayrton and Perry have devised the secohmmeter. The battery and galvanometer circuits are each commuted Fig. 34. Fig. 35. Ayrton and Perry's Secohmmeter. so as to produce a galvanometer deflection in one direction, and increased in amount. This apparatus may be used in connection with any of the above tests where galvanometers are used, the balance being obtained when the deflection is reduced to zero. Below is given a description of the apparatus as shown in Fig. 35. This instrument serves the purpose of making an alternating current to use in measurements of self-induction, and of commuting such portion of this current as flows in the galvanometer circuit to a direct current. The instrument consists of two rotating commutators mounted on one axis and a train of gears for rapidly driving them. The commutators are on the two sides of a cast metal case, one only being shown in the illustration. They are electrically insulated from each other. The brushes of one com- mutator are mounted on a disk, which can be rotated through an angle of 90° around the axis. The brushes can accordingly be set so that they will reverse the circuits in which they are connected at the same time, of so that one will reverse at any desired fraction of a period after the other. The driving handle may be attached at two places on the train of gears, thus giving two speeds. A pulley wheel is also provided, which may be used in place of the handle and the apparatus be driven by a motor. MEAiURE^EUTT OE POWER Il¥ ALTER^ATI^CJ CVRREIT CIRCUITS. In alternating current circuits having inductance in any part of the cir- cuit, such as motors, unloaded transformers, and the self-inductance of the line itself, the product of the values of the current and the E.M.F. as shown by an ammeter and voltmeter does not give the power in the circuit, since the current is not in phase with the E.M.F. t _ The power at any instant of time in any alternating current circuit is equal to the product of the instantaneous values of the current and E.M.F. This is shown graphically in (Cut A) Fig. 36. The mean power in the circuit is P = EI, where E is the effective E.M.F. and / the effective current. The effective values of E.M.F. and current are the square roots of the mean squares of their respective instantaneous values, or numerically, their maximum values divided by V^2 or 1.41. Alternating current measuring instruments of either the "hot wire' ' or dynamometer type indicate effective values. If the current is not in phase with the E.M.F., and the angular difference in phase is , then the power is P = EI cos 4>. 70 MEASUREMENTS. Fig. A. MEASUREMENT OF A.C. POWER. 71 Cos is called the power factor, since it is the factor by which the apparent power EI must be multiplied to obtain the true power. Suppose that curve No. 1 in Fig. B, page 70, represents the various values of the impressed voltage throughout a cycle, and that curve No. 2 represents the various values of the self-induced voltage. Curve No. 2, it will be noted, is not in phase with curve No. 1. Its highest value comes at a later time than that of curve No. 1, because the self-induced electromotive force is never in phase with the impressed electromotive force, as the self-induced electromotive force is obviously at its highest point when the lines of force induced by the coil are changing most rapidly. This occurs when the current is rapidly increasing or diminishing, and not when it is maintain- ing a momentarily steady value at its highest point. Current will flow in the circuit in proportion to, and in phase with, the resultant of the two curves, and the ordinates of this resultant will be the algebraical sum of the corresponding ordinate of the two curves. Curve No. 3 shows the resultant curve constructed in this way. It will be found to be similar to the other curves but of a different maximum value, also lagging behind the curve of impressed E.M.F., but occurring earlier than the curve of self-induced E.M.F. In Fig. C are shown the curves representing the impressed E.M.F. and the resulting current, and as will be seen the current lags behind. If the values of these curves be combined by multiplying them together, ordinate by ordinate, this curve representing power will result. This will be the true curve of power, as it obviously represents the power at every instant, the instantaneous voltage being multiplied by the instantaneous current, and consequently takes account of the fact that their maxima are shifted with reference to one another. If the current and voltage curves are arranged as shown in Fig. D, in which the maximum value of the voltage occurs at the same time as does the minimum value of the current, the result will be as shown, and no power will be produced. If the current is in phase with the electromotive force as shown in Fig. E, the power curve will appear above the zero line, and the true power will also be the apparent power. Three Voltmeter OTethod. Ayrton & lumpner. This method is good where the voltage can be regulated to suit the load. m In figure 37 let the non-inductive re- sistance R be placed in series with the load a b ; take the voltage V across the terminals of R ; J\ across the load a b, and Vo across both, or from a to c. Then the ( V2 V 2 V 2 True watts = -^ ^ . 2 R The best conditions are when V = V 1% and, if R = £ ohm, then W— V 2 2 — V? — F 2 . Combined Voltmeter and Ammeter Method. This method, devised also by Fleming, is quite accurate, and enables the accuracy of instruments in use to be checked. In Fig. 38 R is a non-inductive resistance connected in shunt to the induc- tive load a b, and the voltmeter V measures the p. d. across x y. A and A x are ammeters connected as shown ; then True watt! *=f(A' -A*- (f)> If the voltmeter stakes an appreciable amount of current, it may be tested as fol- lows : disconnect R and V at y, and see that A and A x are' alike ; then con- nect R and V at y again, and disconnect the load a b. Then A x =z current taken by R and V in multiple. 72 MEASUREMENTS. WATTMETER MIIHOI)^. (Contributed by W. N. Goodwin, Jr.) For measurement of power in electric circuits, the wattmeter gives the quickest and most accurate results. Since the instrument mechanically integrates the products of the instantaneous values of current and E.M.F., the power is indicated directly, regardless of the power factor. When a wattmeter is connected to a circuit, the instrument itself re- quires current and, therefore, some power is consumed in it. This error must be calculated and subtracted from the observed readings. Weston wattmeters are compensated for this error by means of a coil wound in opposition to the field coil and adjusted with it. The. following are a few of the important tests with a wattmeter used in power measurements. Fig. 39 shows the connections for measurement of power in either a direct or single phase alternating-current circuit. The power consumed by L is read directly from the instrument. Fig. 39. Fig. 40. In direct current measurements, to eliminate the effect of the earth's magnetic field, two readings must be taken; either the connections must be reversed for the second reading, or the instrument turned 180° from its first position; the mean of the two readings gives the true power. If the instrument have a multiplier, it should be connected as shown m Fig 40, so that the difference of potential between the stationary and mov- abTeopils shall be a minimum. Checking* IVattnieters. — In checking wattmeters either directly with other wattmeters, or by means of a voltmeter and ammeter, the wattmeter should be connected so as not to include its compensating coil. In a Wes- ton wattmeter the "independent" binding post should be used, shown in Fig 39, the pressure circuits being connected in parallel and the field or current coils in series. Three -Phase Power Measurements. — In unbalanced systems two wattmeters are required, connected as shown in Fig. 41. The total power transmitted is then the algebraic sum of the readings of the two watt- meters. If the power factor is greater than .50, the power is the arith- metical sum, and if it is less than .50, the power is the arithmetical differ- ence of the readings. Fig. 41. WATTMETER METHODS. 73 Balanced Three>Phase Systems. — One wattmeter may be used in three-phase eircuits in which the current lag is the same for all parts of the circuit and the load is uniformly distributed. The connections are shown in Fig. 42. The current coil of the wattmeter is connected in one < Fig. 42. of the leads as A; one end of the pressure circuit to the same lead, the other end is connected successively to each of the other leads as B and C, a read- ing being taken in each position. The power is then the sum of the sepa- rate readings. Second Method for Balanced Circuits. — Another method may be used by which the power may be obtained from a single reading of the instrument, as shown in Fig. 43. The current coil of the wattmeter is connected in one lead as A; one end of the pressure circuit is connected to the same lead. Fig. 43. The other end of the pressure circuit is connected to the junction of two resistances r and r, each equal in resistance to that of the wattmeter; the ends of these resistances are connected to the other two leads as shown at B and C. The power is then P = Sp where p is the instrument reading. If it be desired to use the instrument for higher voltages than that for which it was designed, then a resistance R must be added to the instru- R + r ment branch, of such a value that is equal to the multiplying con- r stant m desired. Each of the other two branches must be increased to R + r. Then the power is P = 3 mp. The Weston " Ybox" multiplier, which may be made for any multiplying constant, is constructed according to this principle. Any of the above methods can be used equally well for the delta as for the star connection. 74 MEASUREMENTS. > TESTS WITH A VOLTMETER. The following are a few of the more important tests for which voltmeters and ammeters are especially adapted. With some changes and additions they have mostly been condensed from an article by H. Maschke, Ph.D., of the Western Laboratory published in the Electrical World in April, 1892. The scales of the better known portable instruments read, in general, from to 150, or some even multiple or fraction of this value. Voltmeters are available having scales ranging from 1.5 volts to 750 volts for a full scale deflection, and when used with multipliers for any higher range. Two or more ranges may be had on the same instrument, so that by simply transferring connections from one binding post to another, voltages dif- fering greatly in amount may be measured on one instrument. Millivolt- meters may be had reading as low as 20 millivolts for a full scale deflection. Instruments with Permanent Iflag-nets should not be placed on or near the field magnets of motors or generators, nor should they be used for measurements in very strong magnetic fields, such as those produced in the vicinity of conductors carrying heavy currents. If the fields be not too strong, then the error produced in the instrument from this cause may be eliminated by taking the mean of two readings, one in position, and the other when the instrument is turned 180° from that position around its vertical axis. electromotive force of Batteries. -1» + Fig. 44. The positive post of voltmeters is usually at the right, and marked +• In a battery the zinc is commonly neg- ative, and should therefore be con- nected to the left or negative binding post. For single cells or a small number, a low-reading voltmeter, say one read- ing to 15 volts, will be used, the con- nections being as per diagrams. Electromotive Force of Dynamos. riliWiUJilHh For voltage within range of the instrument available for the purpose, it is only necessary to connect one terminal of the voltmeter to a brush of one polarity, and the other terminal to a brush of the opposite polarity, and read direct from the scale of the instrument. As continuous current volt- meters usually deflect forward or back according to which pole is connected, it is necessary sometimes to reverse the lead wires, in which case the polar- ity of the dynamo is also determined. Of course the voltage across any cir- cuit may be taken in the same way, or the dynamo voltage may be taken at the switchboard, in which case the drop in the leads sometimes enters into the calculations. Following are diagrams of the connections to bipolar and multipolar dynamos : Fig. 46. Fig. 47. TESTS WITH A VOLTMETER. 75 In the case of arc dynamos or other machines giving high voltage, it is necessary to provide a multiplier in order to make use of the ordinary in- strument ; and the following is the rule for determining the resistance which, when placed in series with the voltmeter, will provide the necessary multiplying power. Let e = upper limit of instrument scale, for example 150 volts, E == upper limit of scale required, for example 750 volts, R — resistance of the voltmeter, for example 18,000 ohms, r = additional resistance required, in ohms. Then r = R ^=^ or r = 18,000 75 °~ 15 ° = 72,000 ohms. e 150 The multiplying power = — or — - = 5. e 150 Should the exact resistance not be available, then with any available resistance r t the regular scale readings must be multiplied by ( — -j- 1 ) . Importance of II igr h Resistance for Voltmeters. It is highly important, as reducing the error in measurement, that the in- ternal resistance of a voltmeter be as high as practicable, as is shown in the following example : Let E in the figure be a dynamo, battery, or other source of electric energy, senaing current through the resistance r ; and vm. be a voltmeter indicating the pressure in volts between the terminals A and B. Be- fore the vm. is connected to the terminals A and R there will be a certain difference of potential, which will be less when the voltmeter is connected, owing to the les- sening of the total resistance between the two points ; if the resistance of the vm. be high, this difference will be very small, and the higher it is the less the error. Following are the formulas and computations for de- termining the error. In Fig. 48 let E be the E.M.F. of the generator, r the resistance of the circuit across A and B when the difference of potential is to be measured, r x the resistance of the leads, generator, etc., and R the resistance of the volt- meter. Before the vm, is connected the difference of potential between A and B is With the voltmeter connected the difference of potential indicated by the instrument is Fl = rM 1 rR-\-r t r-\- r x R The voltage across A and B is, therefore, reduced by the introduction of the voltmeter by the amount of V Vl {r + rJR The error is lOOn-! ( — .(^/-V + n)R The error r. inversely proportional to the resistance R of the voltmeter Example . Let E = 10 volts, r = 10 ohms, ri = 2 ohms, R = 500 ohms. Then the reading of the voltmeter is v 10 X 500 X 10 R . Vl (10 X 500) + (2X 10) -|- (2 X 500) ~ S,W °° V01t8, 7b' and the error ia V - v x MEASUREMENTS 10 X 2 X 8.3056 and the percentage error is (10 + 2) 500 .0277 volts. __ 100X10X2 _ P "(10 + 2) X 500" * 333%> If R be made 1000 ohms, then r = 10 X 1000 X 10 (10 X 1000) + (2 X 10) + (2 X 1000) = 8.32 volts and the error is v t -v and the percentage error is V = = 10 X 2 X 8.32 (10 + 2) 1000 100 X 10 X 2 .01387 - .166% is less than -— -• (10 + 2) X 1000 or just one-half the error with R = 500 ohms. If the error of measurement is not to exceed a stated per cent p, then r and n must be such that T\ r-\-r x If the circuit is closed by a resistance r lt and it be desired to measure the E.M.F. of the generator by connecting the voltmeter between any two points as A and B, then E =-- ( — ^— - ) Vi, where V± = reading on vm. The error between the true value of the E.M.F. of the generator and that shown by the voltmeter is E - V = and the percentage error p = 100 ( b ) ' riVi R If the error is not to exceed p per cent, then the resistance of the gen- erator, cables, etc., must not exceed — — • 100 For example, with a voltmeter having 15,000 ohms for 150 volts; if p must be less than |%, then r x may be as great as - — r^r- = 30 ohms. Comparison of E.^I.F. of Batteries. Wlieatstone's Method. — To compare E.M.F. of two batteries, A and X, with low-reading voltmeters, let E be the E.M.F. of A, and JSi the E.M.F. of X. -wwwww Fig. 49. First connect battery A in series with the voltmeter and a resistance r, switch B being closed, and note the deflection V\ then open the switch B, and throw in the resistance r lt and note the deflection Vy Now connect bat- tery X in place of A, and close the switch B, and vary the resistance r until the same deflection Voi voltmeter is obtained and call the new resistance r 2 ; next open the switch B, or otherwise add to the resistance r 2 until the deflec- tion V r of the voltmeter is produced ; call this added resistance r s , then If E be smaller than E u the voltmeter resistance R may be taken as r, and it is better to have r x about twice as large as the combined resistance of r and the resistance of A. It is not necessary that the internal resistance of the cells be small as compared with R. TESTS WITH A VOLTMETER. 77 Pog-gendorff's method Modified toy Clark. To Compare the E.M.F. of a battery cell or element with a standard cell. Let S he a standard cell, The a cell for comparison with the standard, ^bea battery of higher E.M.F. than either of the above elements. A resistance r is joined in series with the battery B and a slide wire A D. A millivoltmeter is connected as shown, both its terminals being connected to the like poles of the battery B and the Standard St Fig. 50. Move the contact C along the wire until the pointer of the instrument stands at zero, and let r t be the resistance of A C. Throw the switch b so as to cut out the standard S, and cut in the cell T ; now slide the contact C x along the wire until the pointer again stands at zero, and call the resistance of A C x r 2 , Then the E.M.Fs. of the two cells T\ S ::r 2 :r v If a meter bridge or other scaled wire be used in place of A Z>, the results may be read directly in volts by arranging the resistance r so that with the pointer at zero the contact C is at the point 144 on the wire scale, or at 100 times the E.M.F. of the standard S, which may be supposed to be a Clark cell. All other readings will in this case be in hundredths of volts ; and should the location of C, be at 175 on the scale when the pointer is at zero on the millivoltmeter then the E.M.F. of the cell, being compared, will be 1.75 volts. Measuring* Current Strength with a Voltmeter. If the resistance of a part of an electric circuit be known, taking the drop in potential around such resistance will determine the current flowing by J? ohms law viz., J= — • . M In the figure let r be a known resistance be- tween the points A and B of the circuit, and / the strength of current to be determined ; then il the voltmeter, connected as shown, gives a deflection of V volts, the current flowing in r will be I— — , r For the corrections to be applied in certain cases, see the section on Importance of High Resistance for Voltmeters, page 75. Always see that the resistance r has enough carrying capacity to avoid a rise of temperature which would change its resistance. If the reading is exact to — volt the meas- urement of current will be exact to am- p Xr Fig. 51. 78 MEASUREMENTS. peres. If r ■» .5 ohm, and the readings are taken on a low-reading volt- meter, say ranging from to 5 volts, and that can be read to 3 ^ volt, then the possible error will be * 1 300 X .5 If rbe made equal to 1 ohm, then the volts read also mean amperes. Measurement of Current with a ^tillivoltnteter. — This is the method generally used in practice for the measurements of currents, and is the same principle as the one outlined above with the substitution of a millivoltmeter for the voltmeter. As the drop is much lower, a comparatively low resistance shunt may be used, so that heavy currents may be measured without the shunt becom- ing disproportionately large. For portable instruments, detachable shunts are generally adjusted with the instrument so that the instrument scale reads directly in amperes. The snunts are constructed of resistance alloy having a negligible temperature coefficient. Switchboard instruments also have shunts with slotted terminals so that they may be connected directly to the bus-bars. In some cases where the currents to be measured are very large the in- struments are adjusted to the drop across a portion of the copper bus-bar through which the current passes. To compute the length of the copper bar ofa given cross section to give a certain drop for a given current, let A = the area of the cross section of bar in square inches, / = current in amperes, V = drop in millivolts desired for instrument for current I; a v y 119 then, length in feet = * at 20° C. measuring: Resistance with a Voltmeter. General Methods. — In the figure, let X = the unknown resistance that is to be measured, r = a known resistance, E, the dynamo or other steady source of E.M.F. When connected as shown in the figure, let the voltmeter reading be V ; then connect the voltmeter terminals to r in the same manner and let the reading be V x ; then X:r:: V: V, and X = — ~ — . "\ If, for instance, r = 2 ohms and V = 3 volts and V x = 4 volts then x _2x_3_ 15ohms 4 If readings can be made to — volt, the error of resistance measurement will then be P Fig. 52. 100 X p ( T + ~V\) P ercent ' and for the above example would be 1 (i + i) = 0.58%. Should there be a considerable difference between the magnitudes of the two resistances X and r, it might be better to read the drop across one of them from one scale, and to read the drop across the other on a lower scale. Resistance Measurement with Voltmeter and Ammeter. The most common modification of the above method is to insert an am- meter in place of the resistance r in the last figure, in which case X=z-^ where /is the current flowing in amperes as read from the ammeter. TESTS WITH A VOLTMETER. 79 If the readings of the voltmeter be correct to — and the ammeter read- ings be correct to the same degree, the possible error becomes : 100 X -f 1 r+i) per cent. ©. M.Vm, V L \ i X ( m < n Am y L r Measurement of very Small Resistances with a Millivolt- meter and Ammeter. By using a millivoltmeter in connection with an ammeter, very small re- sistances, such as that of bars of copper, armature resistance, etc., can be accurately measured. In order to have a reas- onable degree of accuracy in measuring resistance by the " drop" method, as this is called, it is necessary that as heavy currents as may be available be used. Then, if E be the dynamo or other source of steady E.M.F., X be the required resistance of a portion of the bar, V be the drop in potential between the points a and 6, and I be the current flowing in the circuit as indicated by the FIG. 53. v ammeter, then x=- r The applications of this method are endless, and but a few, to which it is especially adapted, need be mentioned here. They are the resistance of armatures, the drop being taken from opposite commutator bars and not from the brush-holders, as then the brush-contact resistance is taken in ; the resistance of station instruments and all switchboard appliances, such as the resistance of switch contacts ; the resistance of bonded joints on electric railway work, as described in the chapter on railway testing. Measurement of High Resistances. "With the ordinary voltmeter of high internal resistance, let R be the re- sistance of the voltmeter, X be the resistance to be measured. Connect them up in series with some source of electro- motive force as in the following figure. Close the switch b, and read the voltage V with the resistance of the voltmeter alone in circuit ; then open the switch, thus cutting in the resistance X, and take another reading of the voltmeter, V r Then Xz=r(£—i\. If the readings of the voltmeter be cor- rect to - of a volt the error of the above V \ i y j- y \ result will be 100 x — tt rr Tr per cent. pVi \ V— r,J Very Hig-li Resistance. — For the measurement of very high resis- tances a more sensitive voltmeter will give much better results for the reason that the reading Vi when the switch b is opened, becomes so small with the ordinary voltmeter that the error is relatively very great. Instruments are on the market having a sensibility of 1600 ohms per volt or about 250,000 ohms for 150 volts. FiCx. 54. 80 MEASUREMENTS. For example if x =» 1 megohm and an ordinary voltmeter be used R — 15,000 ohms for 150 volts, and E = 120 volts, T . . , , ER 120 X 15,000 , __. u V t would be =-—- = 1 nnn nnn , 1 _ - = 1.772 volts; while if ' X + R R were 250,000 ohms, . , . 120 X 250,000 Vi would be 1,000,000 + 15,000 = 24 volts, ' 1,000,000 + 250,000 that is with the high resistance instrument, with the same accuracy of the instrument scales, the percentage error is about ^ a s great as with the lower resistance instrument. Measuring- the Insulation Resistance of [Lighting* and Power Circuits with a Voltmeter. — For the measurement of in- sulation resistance, a high resistance sensitive voltmeter is needed. For rough measurements where the exact insulation resistance is not required but it is wished to determine if such resistance exceeds some stated figure, then a voltmeter of ordinary sensibility will answer. The methods in general are as follows : Let X = insulation resistance to ground as in Fig. 55, X t = insulation resistance to ground of opposite lead, R = resistance of voltmeter, V = potential of dynamo E, Vi = reading of voltmeter, as connected in figure, V2 — reading of voltmeter, when connected to opposite lead. Ground Fig. 55. Ground Then and X = R(f.-t), The above formula can be modified to give results more nearly correct by taking into account the fact that the path through the resistance R of the voltmeter is in parallel with the leak to ground on the side to which it is connected as shown in the following figure : ^=Ground -=r Fig. 5G. TESTS WITH A VOLTMETER. 81 In this case the voltage V of the circuit will not only send current through the lamps, but through the leaks e f to ground, and through the ground to d and c, thence through d to ft, and c to a, these two last paths being in par- allel, therefore having less resistance than if one alone was used ; thus if r be the resistance of the ground leak b d, and r x be the resistance of the leak e /, and R be the resistance of the voltmeter, then the total resistance by way of the ground, between the conductors, would be R X r , and if Then and R + r V= voltage of the circuit, v = reading of voltmeter from a to c, v, = reading of voltmeter from g to c. i r,= B ( r -% + v '>). The sum of the resistance r -\-r x will be = R f (V + v,){V—(v + v,))\ Insulation Resistance of Arc liig-ht Circuits. Arc lamps are to a great extent run in series, and the insulation resis- tance of their circuits is found in a manner similiar to that for multiple circuits, but the formula differs a little. Let the following figure be a typical arc circuit, with a partial ground at c. First find the total voltage V between a and b of the circuit. This can most handily be done with a voltmeter having a high resistance in a sepa- rate box and so calibrated with the voltmeter as to multiply its readings by r Ground FlJT57. some convenient number. For convenience in locating the ground, get the average volts per lamp by dividing the total volts V by the number of lamps on the circuit ; the writer has found 48 volts to be a good average for the ordinary 10 ampere lamp. With the 16 lamps shown in the above figure, V would probably be about 768 volts. Next take a voltmeter reading from each end of the circuit to ground. Call the reading from a to ground v, and from b to ground v n R being the resistance of voltmeter as before, and r the insulation resistance required. Then — b( v -( v + v ' ) \ and the location of the ground, provided there be but one and the general insulation of the circuit be good, will be found closely proportional to the leadings v and v, ; in the above figure say we find the voltmeter reading from a to ground to be 28, and from b to ground to be 36 ; then the distance of the ground c from the two ends of the circuit will be in proportion to the leadings 28 and 36 respectively. There being 16 lamps on the circuit, the number of lamps between a and c 82 MEASUREMEN TS. would be 28 — (28 -f- 36) = §| of 16 = 7, and from b to c would be 36 - (28 + 36) = |f of 16 = 9 ; that is, the ground would most likely be found be- tween the seventh and eighth lamps, counting from a. Insulation across a Double Pole fuse Block or Other Similar Revice where Both Terminals are on the lame Base. Let// be fuses in place on a base, V= potential of circuit, R = resistance of voltmeter, v =z reading of voltmeter, required the resistance r across the base a a, to ft 6 t . Then r = jR ZjZf. Fig. 58. MEASUREMENT OF THE Il¥SITEATIO]¥ RESIS- TANCE OF AN ELECTRIC WIROG SYSTEM WITH THE POWER ON. The following methods have been devised by Dr. Edwin F. Northrup for the measurement of insulation resistance of a circuit where it is im- practicable to shunt off the current. 1. — Voltmeter Method. Let A (Pig. 59) represent any wiring system in which X t and X 2 are the insulation resistances between the bus- bars, B± and B 2 and the earth (the gas or water pipes being taken as at the potential of the earth). In Fig. 59, J, 77, and 777 are equivalent diagrams in which y represents the unknown resistance of all the lamps, motors, etc., across the line. If direct current is supplied to the bus-bars, a direct-current voltmeter should be used. If the current is alternating, then an alternating-cur- rent voltmeter will be required. The resistances, X x and X 2y are determined by knowing g, the resistance of the voltmeter, and by taking three volt- meter readings. 1st. Measure the voltage, which we will call E, across the bus-bars (Fig. 59) 7. 2d. Connect the voltmeter be- tween the bus-bar, B lt and the earth and take its reading, which we will call V t (Fig. 59) 77. 3d. Connect the voltmeter be- tween the bus-bar, B 2 , and the earth and take its reading, which we will call V 2 (Fig. 59) 777. If the readings in either of the two latter cases are only a fraction of a scale division, then the insulation re- sistance is too high to be measured by this method and we maj" resort to Bi c <^f READsE II B, ^f REA DS Vj *i n£ reads Vo >AAA\AAV\AAAAAA/AA/1 C 2 1 1 Fig. 59. Voltmeter Method. MEASURING INSULATION RESISTANCE. 83 the second method to be described. Having taken the above three read- ings, it can be shown that X i ~ vt (1) x 2 - 'C*-r,-r,) . (2) The current /, which leaks to the ground will be, X x + X 2 For example, the insulation resistance of the wiring system of a large office building was determined by means of a Weston voltmeter, the fol- lowing readings and resistances were obtained: Q = E = fc = v 12,220 (113 12,220 ohms, 113 volts, 1 volt, 4 volts. -1-4) , Xl ~ 4 v 12,220 (113 A 2 = -1-4) ] = 329,940 ohms, = 1,319,760 ohms. The above example shows that where the sum of the resistances, X x and X 2 , are not over one or two million ohms, the voltmeter method is sufficiently accurate for the purpose. If one side of the line is grounded — that is, if X 2 = — then from (2) E = V x + V 2 = V lf as V 2 = 0, and the method fails to give X lm Expressions (1) and (2) above are obtained as follows: The meaning of the letters used are indicated in /, II, and 77/ (Fig. 59), C lt C 2 , etc., being currents and g the resistance of the voltmeter. C,= c 2 = -*2~r ■ — X t - E Cl " F+xl Cl = 1 ' or Cl = "~ xT^ C * ""F=ra C, ~7 ' or ° 2 -~x7d Hence, we have the two relations, E « Zij£±Zl) , and E F 2 (g-rl2) > xr I qXi X t g v _t oX 2 X 2 g from which the values for X x and X 2 are obtained as given above in equa- tions (1) and (2). Any instrument, as a galvanometer, in which the deflections are pro- portional to the currents, may be substituted for a voltmeter. In such a case, if D, d*. and d 2 are deflections corresponding to the readings E, V lt and V 2 , and G is the total resistance in series with the instrument, we have as before: X, = G (D ~f x " d *> (3) and X 2 = g (Z) "/* - <* 2 ? (4) do - d d x 84 MEASUREMENTS. If two or more electric lamps are connected in series, their resistances, while carrying current, can be determined by means of three readings, as above. If X 2 — oo, V t — 0, and X t = — — T , — i which is the ordinary ex- 2. pression used in measuring a resistance with a voltmeter by reading the voltmeter with the resistance in series with it and again with the resistance cut out. II. — Galvanometer Uletliocl. This method may be used when greater accuracy is required or when the insulation resistance to earth, of at least one side of the line, is over a megohm. The wiring system is represented in 1 of Fig. 60, and 2 of Fig. 60 gives equivalent circuits. The method consists in connecting across the bus-bars a moderately high resistance and rinding on this resistance a point, p, where the poten- tial due to the generator is the same as that of the earth, and then with Fig. 60. Galvanometer Method. the aid of a sensitive galvanometer and an external source of E.M.F., meas- uring the resistances, r t and r 2 , to earth in the following manner: & is a key and S an Ayrton universal shunt. This latter may be omitted if the source of E.M.F. can be varied in a known manner. It is evident from Fig. 60 that a balance will be had when - = — , the O To key, k, being in its upper position. If k is now depressed, the resistance, R, encountered by the current generated by the source, e, will be R - Ox + - 1 ■ + b-\-r 2 a-\-r t where g t is the resistance of the galvanometer; but in comparison with rj and r 2t g lt a an: 1 b can be neglected, and R = nr 2 By construction, — = 7 = TV, a known ratio. r 2 b tions we deduce m R(N+1) r 2 =- - From the last two rela- and N ri = R(N + 1). Taking d as the deflection of the galvanometer and K as the galvano- meter constant, the current through the galvanometer is e R d __ eK MEASURING INSULATION RESISTANCE. 85 K should be defined as the resistance which must be inserted in circuit with the galvanometer (including its own resistance), so that it will give, with one volt, a scale deflection of one scale division at the distance at which the scale is placed from the mirror during the test, usually taken as one meter. Then we will have: eK {N + l) r 2 =" Nd , eK(N + l) and n = - d Taking K = 10 8 as an average value for an ordinary D'Arsonval gal- vanometer and e = 100, n = 2, and d = 100, we have: 100X10 8 (2 + 1) tcn , r2 = 2^100 = 15 ° me * ohms ' 100 X 10 8 (2 + 1) on „ , r, = Yqq — — = 300 me go hms - This example shows that a galvanometer of very moderate sensibility will measure in this way a very high insulation resistance. If, on the other hand, the insulation is low, small battery power may be used or the deflec- tion of the galvanometer can be cut down to xV» ihs, tsVo» or Totyou by the Ayrton shunt. The only difficulty likely to be experienced in applying the above method is that, while making the test, the relative values of r t and r 2 will keep changing, due to motors or lights being thrown on or off the line. In this event it is only possible to obtain a sort of average value for the resistance to earth of each side of the line. Insulation Resistance of £ lee trie Circuits in Building's. In the United States it is quite common to specify that the entire installa- tion when connected up shall have an insulation resistance from earth of at least one megohm. The National Code gives the following : The wiring of any building must test free from grounds ; i.e., each main supply line and every branch circuit should have an insulation resistance of at least 100,000 ohms, and the whole installation should have an insulation resistance between conductors and between all conductors and the ground (not including attachments, sockets, receptacles, etc.) of not less than the following : Up to 5 amperes . . 4,000.000. Up to 200 amperes . . 100,000. Up to 10 amperes . . 2,000,000. Up to 400 amperes . . 50,000. Up to 25 amperes . . 800,000. Up to 800 amperes . . 25,000. Up to 50 amperes . . 400,000. Up to 1,600 amperes . . 12,500. Up to 100 amperes . . 200,000. All cut-outs and safety devices in place in the above. Where lamp-sockets, receptacles, and electroliers, etc., are connected, one-half of the above will be required. Professor Jamison's rule is : E.M.F. Resistance from earth = 100,000 x r ^ . number of lamps Kempe's rule is : — 75 Resistance in megohms == — . number of lamps A rule for use in the U. S. Navy is : E.M.F. ( Resistance = 300,000 x number of outlets $6 MEASUREMENTS. Institution of Electrical Engineers' rule is : 7900 X E.M.F. B= number of lamps Phoenix Fire Office rule for circuits of 200 volts is that ._ , . „ 12.5 megohms the least E= r- ^ . number of lamps Twenty-five English insurance companies have a rule that the leakage from a circuit shall not exceed ^-J^ part of the total working current. Below is a table giving the approximate insulation allowable for circuits having different loads of lamps. For a circuit having — 25 lamps, insulation should exceed . . 500,000 ohms. 50 lamps, insulation should exceed . . 250,000 ohms. 100 lamps, insulation should exceed . . 125,000 ohms. 500 lamps, insulation should exceed . . 25,000 ohms. 1000 lamps, insulation should exceed . . 12,000 ohms. All insulation tests of lighting circuits should be made with the working current. (See page 80, voltmeter test.) In the following table Uppenborn shows the importance of testing with tiie working voltage. Table I. shows the resistance between the terminals of a slate cut out. Table II. shows the resistance between two cotton-covered wires twisted. I. II. Volts. Megohms. Volts. Megohms. 5 10 13.6 27.2 68 53 45 24 5 10 16.9 27.2 281 188 184 121 Measuring* the Insulation of Dynamos. The same formula as that used for measuring high resistances (see Fig. 55) applies equally well to determining the insulation of dynamo conductors from the iron body of the machine. Fig. 61. Connect, as in Fig. No. 61, all symbols having the same meaning a3 before. Let r = insulation resistance of dynamo, then «*(£-*)• MEASURING INSULATION RESISTANCE. 87 JtEeasuring* the Insulation Resistance of motors. Where motors are connected to isolated plant circuits with known high insulation, the formula used for insulation of dynamos applies ; but where the motors are connected to public circuits of questionable insulation it is necessary to first determine the circuit insulation, which can be done by using the connections shown in Fig. 56. Fig. 62 shows the connections to motor for determining its insulation by current from an operating circuit. Fig. 62. Here, as before, the insulation r of the total connected devices zr If r = total resistance of circuit and motor in multiple to ground, and r t is the insulation of the circuit from ground, then X, the insulation of the motor will be X=' Measurement of the Internal Resistance of a Battery. In the following figure (No. 63), let E be the cell or battery whose resistance is to be measured, K be a switch, and r a suitable resistance. Let V = the reading of voltmeter with the key, K y open (this is the E.M.F. of the battery), and V t = the reading of voltmeter with key, K, closed (this is the drop across the re- sistance r), Then the battery resistance Z±It. r. The same method can be used to measure the internal resistance of dynamos. An ammeter may be connected in the r circuit, in which case V-Vi. Fig. 63. *V = r X - ohms. r t = ■ where / is the reading in amperes. Conductivity with a Millivoltmeter. This is a quick and convenient method of roughly comparing the conduc- tivity of a sample of metal with that of a standard piece. In Fig. 64, R is a standard bar of copper of 100% conductivity at 70° F.; this bar may be of convenient length for use in the clamps, but of known cross section. X is the piece of metal of unknown conductivity, but of the 88 MEASUREMENTS. I same cross section as the standard, j^isa source of steady current, and if a storage battery is available it is much the better for the purpose. M is a milhvoltmeter with the contact device d. The distance apart of the two points may be anything, so long as it remains unaltered and will go between the clamps on either of the bars. Now with the current howing through the two bars in series the fall of potential between two points the same distance apart and on the same flow- FlG, 64. line will, on either bar, be in proportion to the resistance, or in inverse pro- portion to the conductivity ; therefore by placing the points of d on the bars in succession, the readings of the millivoltmeter will give the ratio of the conductivities of the two pieces. For example : if the reading from R — 200 millivolts, and the reading from X=z 205 millivolts, then the percentage conductivity of X as compared with R is 205 : 200 : : 100 : conductivity of X, 200 X 100 or7 _~ or 2Q5 - = 97-50. MAGNETIC PROPERTIES OP IRON. Revised by Townsend Wolcott. With a given excitation the flux $ or flux-density (E of an electromagnet will depend upon the quality of the iron or steel of the core, and is usually rated as compared with air. If a solenoid of wire be traversed with a current, a certain number of magnetic lines of force, J£> will be developed per square centimetre of the core of air. Now, if a core of iron be thrust into the coil, taking the place of the air, many more lines of force will flow ; and at the centre of the solenoid these will be equal to (& lines per square centimetre. As iron or steel varies considerably as to the number of lines per square centimetre (ft which it will allow to traverse its body with a given excitation, its conductivity towards lines of force, which is called its permeability , is numerically represented by the ratio of the flux-density when the core is present, to the flux-density when air alone is present. This permeability is represented by /u.. The permeability /u. of soft wrought iron is greater than that of cast iron ; and that for mild or open-hearth annealed steel castings as now made for dynamos and motors is nearly, and in some cases quite, equal to the best soft wrought iron. The number of magnetic lines that can be forced through a given cross- section of iron depends, not only on its permeability, but upon its satura- tion. For instance, if but a small number of lines are flowing through the iron at a certain excitation, doubling the excitation will practically double the lines of force ; when the lines reach a certain number, increasing the excitation does not proportionally increase the lines of force, and an excita- tion may be reached after which there will be little if any increase of lines of force, no matter what may be the increase of excitation. Iron or steel for use in magnetic circuits must be tested by sample before any accurate calculations can be made. < Data for (ft-3C Curves. Average First Quality American Metal. (Sheldon.) A ,d Cast Iron. Cast Steel. Wrought Iron Sheet Metal. Ampere turns pe cent, leng Ampere turns pe inch leng JC i o ©33 Kilomax- wells per sq. in. GO 1 0) ©31 Kilomax- wells per sq. in. 03 tx Kilom ax- wells per sq. in. Kilomax- wells per sq. in. 10 7.95 20.2 4.3 27.7 11.5 74.2 13.0 83.8 14.3 92.2 20 15.90 40.4 5.7 36.8 13.8 89.0 14.7 94.8 15.6 100.7 30 23.S5 60.6 6.5 41.9 14.9 96.1 15.3 98.6 16.2 104.5 40 31.80 80.8 7.1 45.8 15.5 100.0 15.7 101.2 16.6 107.1 50 39.75 101.0 7.6 49.0 16.0 103.2 16.0 103.2 16.9 109.0 60 47.70 121.2 8.0 51.6 16.5 106.5 16.3 105.2 17.3 111.6 70 55.65 141.4 8.4 53.2 16.9 109.0 16.5 106.5 17.5 112.9 80 63.65 161.6 8.7 56.1 17.2 111.0 16.7 107.8 17.7 114.1 90 71.60 181.8 9.0 58.0 17.4 112.2 16.9 109.0 18.0 116.1 100 79.50 202.0 9.4 60.6 17.7 114.1 17.2 110.9 18.2 117.3 150 119.25 303.0 10.6 68.3 18.5 119.2 18.0 116.1 19.0 122.7 200 159.0 404.0 11.7 75.5 19.2 123.9 18.7 120.8 19.6 126.5 250 198.8 505.0 12.4 80.0 19.7 127.1 19.2 123.9 20.2 130.2 300 238.5 606.0 13.2 85.1 20.1 129.6 19.7 127.1 20.7 133.5 J€ = 1.257 ampere turns per cm. = .495 ampere turns per inch 90 MAGNETIC PROPERTIES OF IRON. i O T 1 \\ \\ Ice !\ ,1 ft "Pi - Id— o \ _JJ! \ f\\ \ H \ ! 1 I IV \ -M \ \ h-1 \ \ «\ \ \ \ \\ \ \ V \ \ \ \ \ \ \ \ \ © s ° a 3 r~ <£ IT -n M HEEL The methods of determining the magnetic value of iron or steel for elec- tro-magnetic purposes are divided by Prof. S. P. Thompson into the follow- ing classes : Magnetometric, Balance, Ballistic, and Traction. The first of these methods, now no longer used to any extent, consists in calculating the magnetization of a core from the deflection of a magneto- meter needle placed at a fixed distance. In the Balance class, the deflection of the magnetometer needle is bal- anced by known forces, or the deflection due to the difference in magnetiza- tion of a known bar and of a test bar is taken. The Ballistic method is most frequently used for laboratory tests, and for such cases as require considerable accuracy in the results. There are really two ballistic methods, the Ring method and the Divided-bar method. In either of these methods the ballistic galvanometer is used for measur- ing the currents induced in a test coil, by reversing the exciting current, or cutting the lines of force. Ring: method. — The following cut shows the arrangement of instru- ments for this test, as used by Prof. Rowland. The ring is made of the sample of iron which is to undergo test, and is uniformly wound with the BALLISTIC ... GALVANOMETER FIG. 2. Connections for the Ring Method. exciting coil or circuit, and a small exploring coil is wound over the excit- ing coil at one point, as shown. The terminals of the latter are connected to the ballistic galvanometer. 92 MAGNETIC PROPERTIES OF IKOX. The method of making a test is as follows : — The resistance, i?, is adjusted to give the highest amount of exciting cur- rent. The reversing switch is then commutated several times with the gal- vanometer disconnected. After connecting the galvanometer the switch is suddenly reversed, and the throw of the galvanometer, due to the reversal of the direction of magnetic lines, is recorded. The resistance, i?, is then adjusted for a somewhat smaller current, Avhich is again reversed, and the galvanometer throw again recorded. The test is carried on with various exciting currents of any desired magnitude. In every case the exciting cur- rent and the corresponding throw of the galvanometer are noted and recorded. If iz=i amperes flowing in the exciting coil, n l = number of turns of wire in exciting coil, / = length in centimetres of the mean circumference of the ring, then the magnetizing force nr . 47r ii x i . ___ nd 5C ~To x T~ or ^ 7x ~f' If I" — length of the ring in inches, then JC"=.495X %. If 9 =z the throw of the galvanometer, K= constant of the galvanometer, i? = resistance of the test coil and circuit, n 2 = number of turns in the test coil, a = area of cross-section of the ring in centimetres, then ^ 10 8 RKQ *** 2 an 2 To determine K, the constant of the galvanometer, discharge a condenser of known capacity, which has been charged to a known voltage, through it. and take the reading 1 , then If c z=z capacity of the condenser in microfarads, e = volts pressure to which the condenser is charged, coulombs is Q = then the quantity passing through the galvanometer upon discharge in c e 1,000,000' and the galvanometer constant c e K- 1,000,000 e 1 IMvided-Bar TOethod. obtain samples in the form of a ring, and still more incon- venient to wind the coils on it, Hopkinson devised the di- vided-bar method, in which the sample is a long rod \" diameter, inserted in closely fitting holes in a heavy wrought iron yoke, as shown in Fig. 3. In the cut the exciting coils are in two parts, and receive current from the battery and through the ammeter, resist- ance, and reversing switch, as shown. -As it is often inconvenient or impossible to AMMETER DIVISION IN V TEST PIECE V TO V-,MEAN_lENGT.H , L-OF TEST PIECgl BALLISTIC^ GALVANOMETER . ., Fig. 3. Arrangement for Hopkinson s di- vided-bar method of measuring permea- The test bar is divided near the centre at the point indicated in the cut, and a small light test coil fo placed over it, and so arranged with springs as MAGNETIC TEST METHODS. 93 to be thrown clear out of the yoke when released by pulling out the loos* end of the test bar by the handle shown. In operation, the exciting current is adjusted by the resistance B, the test bar suddenly pulled out by the handle, thus releasing the test coil and pro- ducing a throw of the galvanometer. As the current is not reversed, the induced pressure is due to N only, and the equation for (ft is (B = 1 and IP — ^ v n ^ ^-10 X L : 1.257 ^\ Where L = the mean length of the test rod as shown in the cut. In using the divided-bar method, a correction must be made, for the rea- son that the test coil is much larger than the test rod, and a number of lines of force pass through the coil that do not through the rod. This cor- rection can easily be determined by taking a reading with a wooden test rod in place of the metal one. An examination of the cut will show that the bar and yoke can also be used for the method of reversals. IVEag-iietic Square Method. — G. F. C. Searle (Journal I. E. E., December, 1904), has suggested another method of avoiding the use of the Rowland ring arrangement. The apparatus consists of a square, with strips laid overlapping at the edges. To obtain accurate results, the dimensions of the square must be large, as compared with the width of the strips. The same is true, but in a somewhat less degree, with the Rowland ring. According to A. Press, when the relative dimensions are correctly adjusted the ballistic galvanometer will give repeatable results, if the iron be first effectively demagnetized by means of an alternating current gradually reduced to zero, and then subjected to a series of reversals, from 50 to 200 with normal magnetizing current, before actual readings are taken. Traction Method. The following cut shows the method with sufficient clearness. A heavy yoke of wrought iron has a small hole in one end through which the test rod is pushed, through the exciting coil shown, and against the bottom of the yoke, which is surfaced true and smooth, as is the end of the test rod. In operation, the exciting current is ad- justed by the resistance R, and the spring balance is then pulled until the sample or test rod separates from the yoke, at which time the pull in pounds necessary to pull them apart is read. Then SPRING BALANCE (B = 1,317 X v/5 +JC - Where P = pull in pounds as shown on the balance, A z=z area of contact of the rod and yoke in square inches. 3C is found as in the Hopkinson method preceding this. Following is a description of a practical adaptation of the permeameter to shop-work as used in the factory of the Westinghouse Electric and Manu- facturing Co. at Pittsburgh, Pa. Fig. 4. S. P. Thompson's per- meameter. 94 MAGNETIC PROPERTIES OF IRON. Tlie Permeameter, as used by toe Westing-house Electric and Mfgr. Co» Design and Description prepared by Mr. C. E. Skinner. A method of measuring the permeability of iron and steel known as the 11 Permeameter Method " was devised by Prof. Silvanns P. Thompson, and is based on the law of traction as enunciated by Clerk Maxwell. According to this law the pull required to break any number of lines of force varies as the square of the number of lines broken. (A complete discussion of the theory of the permeameter, with the derivation of the proper formula for calculating the results from the measurements will be found in the " Electro Magnet,'* by Prof. S. P. Thompson.) A permeameter which has been in use for several years in the laboratory of the Westinghouse Electric and Manufacturing Company, and which has given excellent satisfaction, is shown in Figs. 5 and 6. The yoke, .4, consists of a piece of soft iron 7" x 8h" x 2J", with a rectangular open- ing in the center c l\" x 4". The sample, X, to be tested is §" in diam- eter and 7|" long, and is introduced into the opening through a § " hole in the yoke, as shown in the drawing. The test sample is finished very accurately to §" in diameter, so that it makes a very close tit in the hole in the yoke. The lower end of the opening in the yoke and the lower end of the sample are accurately faced so as to make a perfect joint. The upper end of the sam- ple is tapped to receive a \" screw §" long, twenty threads per inch, by means of which a spring balance is attached to it. The magnetizing coil, G\ is wound on a brass spool, S, 4" long, with the end flanges turned up so that it may be fastened to the yoke by means of the screws. The axis of the coil coincides with the axis of the yoke and opening. The coil has flexible leads, which allow it to be easily removed trom the opening for the inspection of the surface where contact is made between the yoke and the test sample. The spring balance, JF, is suspended from an angle iron fastened to the up- right rack, /, which engages with the pinion, J. The balance is suspended exactly over the centre of the yoke through which the sample passes, to avoid any side pull. A spring buffer, K. is provided, which allows perfectly free movement of the link holding the sample for a distance of about §", and then takes up the jar consequent upon the sudden release of the sample. The frame, B, which supports the pulling mechanism, is made of brass, and has feet cast at the bottom, by means of which the complete apparatus is fastened to the table. Two spring balances are provided, one reading to 30 lbs. and the other to 100 lbs. These spring balances are of special construc- tion, having comparatively long scales. (They were originally made self- registering ; but this was found unnecessary, as a reading could be taken with greater rapidity and with sufficient accuracy without the self-register- ing mechanism.) Any good spring balance may be used. The spring should be carefully calibrated from time to time over its whole range ; and if there is a correction it will be found convenient to use a calibration curve in cor- recting the readings. With a sample f" in diameter, or § of a square inch area cross-section, the maximum pull required for cast iron is about 25 lbs., and for mild cast steel about 70 lbs. With the number of turns on the coil given above, the current required for obtaining a magnetizing force of JC= 300, is about 12.5 amperes. This is as high a value as is ever necessary in ordinary work. For furnishing the current a storage battery is ordinarily used, and the variations made by means of a lamp board which has in addition a sliding resistance, so that variations of about .01 ampere may be obtained over the full range of cur- rent from 0.1 ampere to 12.5 amperes. The operation of the permeameter is as follows : — The sample to be tested is first demagnetized by introducing it into the field of an electro-magnet with a wire core, through which an alternating current is passing, and gradually removing it from the field of this electro- magnet. The sample is then introduced into the opening in the yoke, care being taken to see that it can move without friction. Measurements are taken with the smallest current to be used first, gradually increasing to the highest value desired. In no case should a reading be taken with a current of less value than has been reached with the sample in position, unless the sample is thoroughly demagnetized again before reading is taken. It is usually most convenient to make each successive adjustment of cur- THE PERMEAMETER. 95 rent with the sample out of position, then introduce the sample and give it a half turn, to insure perfect contact between the sample and the yoke. The lower end of the sample and the surface on which it rests should he care- fully inspected to see that no foreign matter of any kind is present which might introduce serious errors in the measurements. The pull is made by turning the pinion slowly by means of a handle, E, carefully noting each o fflj <§B= <§» h 4j LLs / J . r o 1 o x^ ) i Coil and Shell Fig. 5. position of the index of the spring balance as it advances over the scale, and noting the point of release. The mean of three or four readings is usually taken as the corrected value for pull, the current in the coil remain- ing constant. With practice the spring balance can be read to within less than 1%; and as the square root of the pull is taken, the final error become3 quite small, especially with high readings. 96 MAGNETIC PROPERTIES OF IRON. The evaluation of the results for the above permeameter is obtained by the use of the following formula : The magnetizing force JC = n . 10 I Where 7^ = number of turns in the magnetizing coil = 223, i = current in amperes, j = length of magnetic circuit in centimeters, estimated in this case as 11.74. Substituting the known values in the above formula we have 3C = 23.8i. Fig. 6. The number of lines of force per square centimeter i =r 1,317 V3 + 3C Where P = pull in lbs. y4 = area of the sample in square inches = 0.3068. JC = value of the magnetizing force for the given pull. THE PERMEAMETER. 97 Substituting the value of A in the above formula we have (B = 2,380 Vp + X There are several sources of error in measurements made by the permea- meter which should be carefully considered, and eliminated as far as possible. a. The unavoidable air gap between the sample and the yoke where it passes through the hole in the upper part of the yoke, together with the more or less imperfect contact at the lower end of the sample, increases the magnetic reluctance and introduces errors for which it is impossible to make due allowance. By careful manipulation, however, these can be reduced to a minimum, and be made practically constant. b. As the magnetization becomes greater the leakage at the lower end of the sample increases more rapidly; and there is considerable error at very high values from this source, as the leakage lines are not broken with the rest. c. Errors in the calibration and reading of the spring balance. None but the best quality of spring balance should be used, and the average of several readings taken with the current remaining perfectly constant for each point on the (B _ JC curve. As the square root of the pull is taken, the errors due to reading the spring balance make a larger and larger percent- age error in (gasP approaches zero, thus preventing accurate determina- tions being made at the beginning of the curve. From the above it will be seen that the permeameter is not well adapted for giving the absolute values of the quality of iron and steel, but is especially suitable for comparative values, such as are noted in ordinary work, where a large number of samples are to be quickly measured. A complete curve can be taken and plotted in ten minutes. By suitable comparison of known samples measured by more accurate methods, the permeameter readings may be evaluated to a sufficient degree for use in the calculations of dynamo electric machinery. Drysdale's Permeameter, This instrument is designed to enable one to test the magnetic quality of iron or steel magnet castings and forgings under commercial conditions, by drilling it with a special drill. A testing plug is inserted in the hole thus drilled and the magnetization or permeability is then directly meas- Fig. 7. ured on an instrument attaehed, without any calculations, by simply throwing over a reversing switch. Fig. 7 shows the special form of drill employed. It has four cutting edges at the lower end, which cut a cylindrical hole in the specimen. The drill is, however, made hollow, so that a thin rod or pin of the material is left standing in the center of the hole, as shown in Fig. 8, which shows a cast steel pole piece, and some small specimens of iron and steel actually drilled. In addition, cutting edges are provided at the top of the drill, which give a conical shape to the top of the hole drilled. The hole is about f in. deep and ^ in. in its largest diameter, while the pin is T V in. in diameter. Such a hole may be drilled in any position where a bolt hole is afterwards to be made in the back of a pole piece, or face of a joint, or otherwise in projections left specially for the purpose, which may be cut off the casting or forging on delivery and sent to the test room. 98 MAGNETIC PROPERTIES OP IRON. Fig. 8. Specimens Showing Holes and Pins. In this hole is inserted the testing plug, Fig. 9, which consists of a soft iron plug, accurately fitting the conical portion of the hole cut by the drill, and having a central hole fitting over the pin. The plug is also split length- wise, so that on forcing it into the conical hole the sides yield slightly and grip the pin, so making a very perfect magnetic joint. If the pin is magnetized the lines of force pass through the pin into the plug, and thence round the mass of the metal to the pin again, as shown in Fig. 9. The pin is magnetized by current in a coil wound round it, and the magnetization produced is tested by use of a second or search coil. On making or breaking or reversing the current in the first or magnetizing coil, the lines of force passing through the search coil are altered, and if this coil is connected to a galvanometer, kicks or throws of the galvanometer will be obtained proportional to the change in the magnetization of the pin. CORE BOSSES. These result from Hysteresis and Eddy Fig. 9. Section through Plug currents. > and Specimen. Professor Ewing has given the name Hysteresis to that quality in iron which causes the lagging of the induction behind the magnetic force. It causes a loss when the direction of the induction is reversed, and results in a heating of the iron. It increases in direct proportion to the number of reversals, and according to Steinmetz, as the 1.6th power of the maximum value of the induction in the iron core. The heat produced has to be dissi- pated either by radiation or conduction, or by both. Steinmetz gives the following formula for hysteresis loss in ergs per cubic centimeter, of iron per cycle; h = -q (fomax 1 - 6 , where i? = a constant depending upon the kind of iron. Taking -q at .002 and retaining (B in gausses, the loss in watts per cubic inch of material Ph will be, Ph -.338 (ftma* 1 - 6 / 10"*, in which /= cycles per second. CORE LOSS. 99 It is to be observed that, in practice, considerable variations in the mag- netic density take place in parts where the magnetomotive force is a con- stant, due to the differences in the lengths of the lines of flux. This will not only affect the measured hysteresis losses, but the eddy currents as well. For this reason, machines of geometrically different form will not obey quite the same law of losses. Considerable question has been raised regarding the constancy of the hysteresis index. According to A. Press, the experiments of Mordey and Hansard with transformer iron imply that the hysteresis index for the range taken should be at least 2. Lancelot Wild gave the index as 2.7 for densities varying from (ft = 200 to 05 = 400, W. E. Sumpner states that the index varies 1.475 to 2.7, depending upon the range of the density, and Prof. Ewing gives the index as varying from 1 .9 to 2 with densities (E = 200 to (ft = 500, depending upon the sample. Hysteretic Constants for Different Materials. Material. Hysteretic Constant. v. Best annealed transformer sheet metal . . Very soft iron wire .001 .002 Thin good sheet iron .003 Thick sheet iron .0033 Most ordinary sheet iron Transformer cores .004 .003 Soft annealed cast steel .008 Soft machine steel .0094 Cast steel .012 Cast iron .016 Hardened cast steel .025 Hysteresis I^oss JFactors. ®>maz ^max 1 ' 6 ^ma, 1 ' 6 in Gausses. r?= 0.002 V = 0.003 >? = 0.004 1,000 63,100 126 189 252 2,000 191,300 382 573 765 3,000 365,900 731 1,096 1,463 4,000 580,000 1,160 1,740 2,320 5,000 828,800 1,657 2,486 3,315 6,000 1,111,000 2,222 3,333 4,444 7,000 1,420,000 2,840 4,260 5,680 8,000 1,758,000 3,516 5,274 7,032 9,000 2,122,000 4,244 6,366 8,488 10,000 2,511,000 5,022 7,533 10,044 Eddy Currents are the local currents in the iron core caused by the E.M.F.'s generated by moving the cores in the field, and increase as the square of the number of revolutions per second. The cure is to divide or laminate the core so that currents cannot flow. These currents cause heating, and unless the core be laminated to a great degree are apt to heat the armature core so much as to char the insulation of its windings. Wiener gives tables showing the losses by Hysteresis and Eddy currents at one cycle per second, under different conditions. These are changed into any number of cycles by direct proportion. The formula for eddy current loss is: p e = 42 &" 2 fP 10~ 18 , in which Pe = watts per cu. in., Qfamax" — = maximum value of the magnetic density per sq. in., t = thickness of plate in mils, and/ = frequency. 100 MAGNETIC PROPERTIES OF IRON. Hysteresis factors for Different Core Densities. S? H -* ~ £ 2 * H «. w o w < 3 fe S3 Watts dissipated at a Frequency of One Complete Magnetic Cycle per Second. V = .002 10,000 15,000 20,000 25,000 30,000 31,000 32,000 33,000 34,000 35,000 36,000 37,000 38,000 39,000 40,000 41,000 42,000 43,000 44,000 45,000 46,000 47,000 48,000 49,000 50,000 51,000 52,000 53,000 54,000 55,000 56,000 57,000 58,000 59,000 60,000 61,000 62,000 63,000 64,000 65,000 Per cu. ft. 713 37 16 09 06 39 4.62 4.85 5.08 5.32 5.56 5.82 6.08 6.30 6.59 6.87 7.11 7.39 7.67 7.95 8.24 8.53 8.81 9.10 9.40 9.70 10.00 10.31 10.68 10.98 11.28 11.60 11.94 12.27 12.61 12.95 13.30 13.63 13.99 14.31 Per lb. .0015 .0027 .0045 .0064 .0086 .0091 .0096 .0101 .0106 .0111 .0116 .0121 .0126 .0131 .0136 .0142 .0148 .0154 .0160 .0166 .0172 .0178 .0184 .0190 .0196 .0202 .0208 .0214 .0221 .0228 .0235 .0242 .0249 .0256 .0263 .0270 .0277 .0284 .0291 .0298 V = .003 Per cu. ft. 1.069 2.055 3.24 4.64 6.09 6.59 6.93 7.28 7.62 7.98 8.34 8.73 9.12 9.45 9.89 10.31 10.67 11.09 11.51 11.93 12.36 12.80 13.22 13.65 14.10 14.55 15.00 15.46 15.97 16.47 16.92 17.40 17.91 18.41 18.91 19.42 19.95 20.45 20.98 21.47 Per lb. .0023 .0041 .0068 .0096 .0129 .0137 .0144 .0152 .0159 .0167 .0174 .0182 .0189 .0197 .0204 .0213 .0222 .0231 .0240 .0249 .0258 .0267 .0276 .0285 .0294 .0303 .0312 .0321 .0332 .0342 .0353 .0363 .0374 .0384 .0395 .0405 .0416 .0426 .0437 .0447 5 w ^ - & fc 2 H « L£J ID K £ £ o < S fc 66,000 67,000 68,000 69,000 70,000 71,000 72,000 73,000 74,000 75,000 76,000 77,000 78,000 79,000 80,000 81,000 82,000 83,000 84,000 85,000 86,000 87,000 88,000 89,000 90,000 91,000 92,000 93,000 94.000 95,000 96,000 97,000 98,000 99,000 100,000 105,000 110,000 115,000 120,000 150,000 Watts dissipated at a Frequency of One Complete Magnetic Cycle per Second. V = .002 Per cu. ft. 14.68 15.01 15.39 15.76 16.13 16.50 16.87 17.25 17.61 17.99 18.41 18.78 19.19 19.58 19.93 20.37 20.77 21.18 21.60 21.98 22.40 22.85 23.26 23.65 24.10 24.51 24.97 25.41 25.86 26.30 26.84 27.30 27.73 28.19 28.55 30.86 33.20 35.70 38.20 40.83 Per lb. .0305 .0313 .0321 .0329 .0337 .0345 .0352 .0360 .0368 .0376 .0384 .0392 .0400 .0408 .0416 .0424 .0432 .0440 .0448 .0456 .0465 .0474 .0483 .0492 .0501 .0510 .0519 .0528 .0538 .0548 .0558 .0568 ,0578 0588 0598 0643 0694 0746 0796 0850 V = .003 Per cu. ft. 22.02 22.52 23.05 23.64 24.19 24.75 25.31 25.88 26.41 26.99 27.62 28.17 28.78 29.37 29.90 30.55 31.15 31.77 32.40 32.98 33.60 34.27 34.87 35.47 36.15 36.76 37.44 38.11 38.79 39.45 40.26 40.95 41.59 42.28 42.85 46.29 49.80 53.55 57.30 60.25 CORE LOSS. IQl The Step-oy-Step method of Hysteresis Test. The samples for hysteresis tests, being generally of sheet iron, are made in the form of annular disks whose inner diameters are not less than § of their external diameter. A number of these disks are stacked on top of each other, and the composite ring is wound with one layer of wire form- ing the magnetizing coil of n x turns. This coil is connected through a re- versing switch to an ammeter in series with an adjustable resistance, and a storage battery. A secondary test coil of n 2 turns is connected with a bal- listic galvanometer, as shown in Fig. 10. BALLISTIC GALVANOMETER ( Fig. 10. To make the test, adjust the resistance for the maximum exciting current. Reverse the switch several times, the galvanometer being disconnected. Then connect the galvanometer, and reduce the current by moving the con- tact arm of the rheostat up one step. This rheostat must be so constructed that an alteration in resistance can be made without opening the circuit even for an instant. Note the throw in the galvanometer corresponding to the change in exciting current. Follow this method by changing resistance step-by-step until the current reaches zero. Reverse the direction, and in- crease step-by-step up to a maximum and then back again to zero. Reverse once more, and increase step-by-step to the original maximum. In every case note and record the value of the exciting current i, and the corre- sponding throw of the galvanometer, 0. Form a table having the following headings to its columns : — i, JC> 0> change of (B, (ft. Values of IT are obtained from the formula, JC = 10 I , when I = average circumference of the test ring. Change of 63 is obtained by the formula, W R KB a n 2 ' where all letters have the same significance as in the formula on page 92. Remember that we started in our test with a maximum unknown value of (U and that we gradually decreased this by steps measurable by the throw of the galvanometer, and that we aftenvards raised the 03 in an opposite direc- tion to the same maximum unknown value, and still further reduced this to zero, and after commutation produced the original maximum value. Ac- cording to this, if due consideration be paid to the sign of the (ft which is determined by the direction of the galvanometer throw, the algebraic sum of the changes in (ft should be equal to zero ; the algebraic sum of the first or second half of the changes in (ft should be equal to twice the value of the original maximum, (ft. Taking this maximum value as the first under the column of the table headed (ft, and applying algebraically to this the changes in (ft for successive values, we obtain the completed table. Plot a curve of JCand(ft. The area enclosed represents the energy lost in carry- ing the sample through one cycle of magnetization between the maximum limits + (ft and —(ft. Measure this area, and express it in the same units as is employed for the co-ordinate axes of the curve. This area divided by 4* 102 MAGNETIC PROPERTIES OF IRON. gives the number of ergs of work performed per cycle upon one cubic centi- meter of the iron, the induction being carried to the limits -f- (Sand — (ft. Tlie Wattmeter Method of Hysteresis Tests. Inasmuch as the iron, a sample of which is submitted for test, is generally to be employed in the manufacture of alternating-current apparatus, it is desirable to make the test as nearly as possible under working conditions. If the samples be disks, as in the previous method, and these be shellacked on both sides before being united into the composite test-ring in order to avoid as much as possible foucault current losses, the test can be quickly made according to the method outlined in the following diagram : Fig. 11. Wattmeter Test for Hysteretic Constant. Alternating current of / cycles per second is sent through the test-ring. Its voltage, E, and current strength, t, are measured by the alternating- current voltmeter, V, and ammeter, A. If r be the resistance of the test- ring coil of ?ii turns, then the watts lost in hysteresis W, is equal to the wattmeter reading W — i 2 r. If the volume of the iron be V cubic centi- meters, and the cross section of the iron ring be a square centimeters, then Steinmetz's hysteretic constant 10 7 W / V2tt ttt/a x 1 - 6 V E 10 8 )' Vf Foucault current losses are neglected in this formula, and the assumption is made that the current is sinusoidal. Ewing-'s Hysteresis Tester. — In this in- strument, Fig. 12, the test sample is made up of about seven pieces of sheet iron %" wide and 3" long. These are rotated betweec the poles of a permanent magnet mounted on knife edges. The magnet carries a pointei which moves over a scale. Two standards of known hyster- esis properties are used for reference. The de- flections corresponding to these samples are plotted as a function of their hysteresis losses, and a line joining the two points thus found is referred to in subsequent tests, this line show- ing the relation existing between deflection and hysteresis loss. The deflections are practically the same, with a great variation in the thick- ness of the pile of test-pieces, so that no cor- rection has to be made for such variation. This instrument has the advantage of using easily prepared test samples. Fig. 12. Hysteresis Ttteter, T *ed by General Electric Co. Designed and Described by Frank Holden. During the last few weeks of the year 1892 there was built at the works of the General Electric Company, in Lynn, Mass., under the writer's direction, an instrument, shown in Fig. 13, by which the losses in sheet iron were determined by measuring the torque produced on the iron, which was punched in rings, when placed between the poles of a rotating electro-mag- net. The rings were held by a fibre frame so as to be concentric with a CORE LOSS. 103 vertical shaft which worked freely on a pivot bearing at its lower end. They had a width of 1 centimeter, an outside diameter of 8.9 centimeters^ and enough were used to make a cylinder about 1.8 centimeters high. The top part of this in- strument, which rested on a thin brass cylin- der surrounding the rings, was movable. On the upper surface was marked a degree scale, over which passed a pointer, with which the upper end of a helical spring rotated. It was so constructed that when the vertical shaft with the rings and the upper part of the instru- ment with the spring was put in place, the lower end of the spring engaged with the shaft, and consequently rotated with the rings. A pointer moving with the lower end of the spring leached to the zero of the degree scale when the apparatus was ready for use. By this ar- rangement it was found what distortion it was necessary to give the spring in order to bal- ance the effect of the rotating magnet, and the spring having been calibrated, the ergs spent on the rings per cycle were determined by mul- tiplying the degrees distortion by a constant. A coil, so arranged that it surrounded but did not touch the rings, made contact at its ends with two fixed brushes that rested in diametrically oppo- site positions on a two-part commutator, which revolved with a magnet. The segments were connected each to a collector ring against which rubbed a brush, the latter two brushes being joined through a sensitive Weston voltmeter. If this were so arranged that the coil was at right angles to the Fig. 13. Hysteresis Meter. 1000 2000 3000 4000 5000 0000 7000 1 •[ 1 r -T_ 8000 l|NDL G FP'i' fp 16 -,4- -- " +.* 2* ±=- S000 s / 6U00 —l*\ t-70oa *-/-- : JZ - 2 t^eooo ~"V / ? 3 _- 5000 3000 - i -T _ ± __ i- 2:- 4000 - •*nm =£- O ^ 2 • Watts dissipated fa £ fa fa fa A H g A £ S fa O" Watts dissipated PER CUBIC FOOT OF IRON at a fre- quency OF 1 CYCLE PER CUBIC FOOT OF IRON AT A FRE- QUENCY OF 1 CYCLE = number of lines of force cut, NI W =l0' Rotation of Conductor (carrying- current) around a Itlagrnet role. If a conductor (carrying current) be so arranged that it can rotate about the pole of a magnet, the force producing the rotation, called torque, will be determined as follows : The whole number of lines of force radiating from the pole will be 4tt times the pole strength m. 4:tt ml . ___ T w — — 7q— = 1-257 ml. Dividing by the angle 2tt, the torque, T, is Every electric circuit tends to place itself so as to embrace the maximum flux. Two electric conductors carrying currents tend to place themselves in position such that their mutual flux may be maximum ; otherwise stated : if two cur- rents run parallel and in the same direction, each produces a field of its own, and each conductor tends to move across the other's field. In two coils or conductors lying parallel to each other, as in a tangent gal- vanometer, the mutual force varies directly in proportion to the product of their respective n/,and inversely as the axial distance they are apart. Principle of the Tlag-uetic Circuit. — The resistance that a mag- netic circuit offers to the passage or flow of magnetic lines of force or flux, has been given the name of reluctance, symbol (ft, and is analogous to resist- ance, to the flow of electric current in a conductor. The magnetic flux or lines of force are treated as current flowing in the magnetic circuit, and denoted by the symbol <£. The above two factors, together with the magneto-motive force described in the early part of this chapter, bear much the same reiation to each other as do resistance, current, and E.M.F. of electric circuits, and are expressed as follows : — , x .. „ Magneto-motive force Magnetic flux = - — . reluctance 1.257 nl Afx nl= +J& , 1.257 110 ELECTKOMAGNETS. If dimensions are in inches, and A is in square inches, then nl— -^- X .3132. and <£ = (&" A". TUe JLaw of Traction. — The formula for the pull or lifting-power of an electromagnet when the poles are in actual contact with the arma- ture or keeper is as follows : Pull (in dynes) = ^ — . Pull (in grammes) = 8 w x 981 Pull (iii pounds) = „™ nnn '• v ^ ' 11,183,000 fir\ 2 A" In inch measure: Pull (in pounds) = " • > Jd , 104,000 Traction. This proportionality to the square of the induction accounts for some anomalous peculiarities in the way that the keeper of a magnet holds fast to the poles. If the pole faces be perfectly true and flat and the face of the keeper the same, the keeper is actually held with less force than when the pole faces are very slightly convex. Or, again, if the keeper be slid to one side until only its sharp edge and that of the poles are in contact, it will be found to adhere more firmly than when placed squarely and centrally on the poles. In general, a magnet holds tighter to a slightly uneven surface than to one which perfectly fits the poles. The reason is that, when the area of contact is decreased, the intensity of the induction through the remaining contact is increased by the crowding together of the lines of induction; and, as the traction is proportional to the product of the area and the square of the intensity of the induction, so long as there is sufficient crowding of the lines so that the square of their intensity increases more than the area is diminished, the traction is increased by reducing the area of contact. ^. 2 The amount of the traction is usually determined by the formula, T z=z __ . in which T is the traction per square centimeter expressed in dynes: to express the traction in grammes, this figure is of course divided by 981, or for pounds avoirdupois per square inch it should be divided by 69090. This formula is correct for the force required to separate the halves of a straight bar magnet cut in the middle, if the winding be also in halves and these halves separate at the same time as their respective halves of the core and if, further, the winding fit the core closely. It is also correct for the separating force when the magnetism is residual; as in the case of a per- manent magnet. In other cases, for example, where an ordinary keeper is pulled away from a magnet, the formula is not strictly accurate on account of the keeper being attracted partly by the core of the magnet and partly by the current in the winding directly. However, the attraction exerted by the coil is usually small as compared to that exerted by the core; and the formula is not very much in error. The attraction between the two parts of the iron is always 2 71-3 2 dynes per square centimeter, 3 being the intensity of magnetization, that is the number of units of free magnetism per square centimeter. But (fc^z^nQ + JC so when J£ = 0, that is when there is no magnetizing force, 2 11-32 r= - — , which is evidently correct, as there is no attraction except between the two parts of the iron. When J£ is not equal to zero, that is, when the magnetism is not residual, there is a force between the coil and the part of the iron that is moved away from the coil equal to JC3 ( ^y nes per square centimeter, so that the whole force of separation is 2 tt^ 2 -f- J£3- When there is a coil on each part of the magnet and both parts of the magnet PROPERTIES OF ELECTROMAGNETS. Ill and both coils are just alike, there are two of these J£3 forces, because each coil attracts the other part of the iron; but as in this case J£ represents the intensity of the magnetizing force of the whole coil each half now attracts the other part of the iron with a force of ' and both forces 2 Hf>2 together equal 3C3- The two coils attract each other with a force of- 3C 2 8tt per square centimeter, so the whole force is 2 n 3 2 + 5C3 + s — > which 1 1 87r rt* may be written ^- (16 n* 3 2 + 8 n JC 3 + JC 2 ) = <±- (4 tt 3 + JC) 2 = -^- per square centimeter, so in this case also the traction is proportional to the square of the intensity of the induction. If the coils be loose upon the cores so that their areas are sensibly greater than those of the cores, the whole force of separation is greater than that given by the equation; but, in practical cases, the error is usually small. In all cases, the attrac- tion between the iron parts is 2 rr 3 2 per square centimeter. UEag-netization and Traction of Electromagmets*. (B (B" Dynes Grammes Kilogs Pounds Lines per Lines per per per per per sq. cm. sq. inch. sq. cm. sq. cm. sq. cm. sq. inch. 1,000 6,450 39,790 40.56 .04056 .577 2,000 12,900 159,200 162.3 .1623 2.308 3,000 19,350 358,100 365.1 .3651 5.190 4,000 25,800 636,600 648.9 .6489 9.228 5,000 32,250 994,700 1,014 1.014 14.39 6,000 38,700 1,432,000 1,460 1.460 20.75 7,000 45,150 1,950,000 1,987 1.987 28.26 8,000 51,600 2,547,000 2,596 2.596 36.95 9,000 58.050 3,223,000 3,286 3.286 46.72 10,000 64,500 3,979,000 4,056 4.056 57.68 11,000 70,950 4,815,000 4,907 4.907 69.77 12,000 77,400 5,730,000 5,841 5.841 83.07 13,000 83,850 6,725,000 6,855 6.855 97.47 14,000 90,300 7,800,000 7,550 7.550 113.1 15,000 96,750 8,953,000 9,124 9.124 129.7 16,000 103,200 10,170,000 10,390 10.390 147.7 17,000 109,650 11,500,000 11,720 11.720 166.6 18,000 116,100 12,890,000 13,140 13.140 186.8 19,000 122,550 14,360,000 14,630 14.630 208.1 20,000 129,000 15,920,000 16,230 16.230 230.8 Exciting- Power and Traction. — If we can assume that there is no magnetic leakage, the exciting power may be calculated from the follow- irig expression ; all dimensions being in inches, and the pull in pounds: nl=z^— x.3132, nXnl ^ — I" X .3132 ' also, (ft" = 8494 ^ Pull 7// 7i/=2661X — X If dimensions are in metric measure, \. Pull nl=3951 Area" . / Pull in kilos ▼ Area in sq. cms. ' ^=1316.6 / Pull in lbs. Area in sq. ins. ^ *~ n ~ l /Pull in kilos. (R = 4965 V -i w t Area sq. cm. 112 ELECTROMAGNETS. WINDING OF EUECTJRO^AOjIUTS. The method used by Cecil P. Poole for predetermining magnet windings is as follows: Temporary test coils, of wire much larger than will probably be required in the permanent coils, are wound to occupy the space that it is estimated the permanent coil will occupy. Current is passed through the temporary coils in series with a water rheostat or finely graduated resistance, by means of which the excitation may be closely adjusted. The exciting current is adjusted until the desired magnet performance is obtained ; the current producing this effect is represented by Ix. The current is then increased or decreased as may be required until the resist- ance per foot of the winding corresponds with the resistance per foot given by Table I herewith, after five hours. The current required to produce this result is indicated by In. The size of wire required to produce a given number of ampere-turns under given conditions of mean length and voltage is d°~ = KAtLm in which d 2 equals circular mils of the wire to be used, K is a coefficient de- pending upon the specific resistance of the wire, At equals the ampere-turns desired, Lm equals the mean length per turn of wire in inches, and V equals the volts at the terminals of the coil. With the best commercial grade of magnet wire, K becomes unity at a temperature of about 140° Fahr., since the resistance per mil-foot of the wire at that temperature is 12 ohms. The resistances of wires given by Table I are based on this temperature. Table II has been calculated from the foregoing formula for this temper- ature. From the first test made with the temporary winding the desired ampere- turns are obtained, and from Table II may be obtained the size of wire required to give the nearest number of ampere-turns per volt corresponding to this test and the proposed working voltage. Table I. - Resistance of 3Kagriet Wire at 140° Tempera- ture, Fahrenheit. Wire No. Resistance per Foot. Wire No. Resistance per Foot. 4 5 6 0.0002875 0.0003625 0.0004571 19 20 21 0.009316 0.01176 0.014814 7 8 9 0.00057662 0.0007268 0.0009168 22 23 24 0.018691 0.023575 0.0297 10 11 12 0.001156 0.0014575 0.001838 25 26 27 0.0375 0.04725 0.05956 13 14 15 0.0023175 0.002922 0.003684 28 29 30 0.0751 0.0947 0.1194 16 17 18 0.004646 0.00586 0.007389 31 32 33 . 1506 . 1899 0.2395 WINDING OF ELECTROMAGNETS. 113 The number of turns of wire in the test coil will, of course, be known, and the product of this number and the current, lx, is the required exciting force in ampere-turns. The mean length per turn of wire in the perma- nent winding will be the same as that in the test winding, subject to minor corrections that may prove necessary in rounding out the final results. Tentatively, at least, the mean length, Lm, will be equal to Gt + g 2 ' in which Gt is the girth of the test coil and g the girth of the bobbin or form in which it was wound. Having the ampere-turns required, the mean length per turn of wire and the voltage that will be applied to the terminals of the coil (or each coil, if there are more than one), the size of wire that must be used in the permanent winding is obtainable by the application of Table II. It may happen that none of the mean length values in the table will be found to correspond with that of the test winding; in that event, the nearest table value may be adopted and the mean length per turn of the permanent winding made to conform to this. In many cases it will be found that both the excitation per volt and the mean length per turn of the test winding will differ from all values in the table; in such a case, the nearest mean length value in the table should be adopted which gives the nearest excitation per volt in excess of the desired value. The table is worked out on the assumption that any two wires drawn to B. & S. gauge and differing in size by ten gauge numbers will have cross- sectional areas differing in the ratio of 1 to 10.163 or 10.163 to 1, according to which wire is considered first. As stated in the note at the foot of the table, the ampere-turns per volt in column a apply to the wire sizes in line A across the top of the table; the ampere-turns per volt in column b apply to the wire sizes in line B, and those in column c, to the wires in line C Thus, if a coil wound with No. 8 wire has a mean length of 45.11 inches per turn, its exciting force will be 366 ampere-turns for each volt at its terminals; a coil of the same mean length but wound with No. 18 wire will have 36 ampere-turns per volt, while a coil of No. 28 wire with the same mean length per turn will yield only 3.54 ampere-turns per volt of applied E.M.F. The table is calculated on the basis of the wire sizes in line B and the ampere-turns per volt in column b, hence the latter values are not numbers from which decimals have been dropped, but are exact. If the winding is to operate at constant potential, as most magnet wind- ings do, the watts dissipated will be exactly proportional to the current passing, and this will be inversely proportional to the length of the coil par- allel with the magnet core if the girth and temperature remain constant. The temperature will be unchanged, of course, the value Ih, of the current necessary to produce the working temperature having been ascertained by trial, as previously described. If the girth of the permanent winding cannot be made identical with that of the test winding, the correction in dimensions will be simple. First, the proper length on the hypothesis of unchanged girth must be determined. As the temperature of the coil is a function of the heat dissipated per unit of effective radiating surface, and the radiating surface is approximately proportional to the length of the coil parallel with the core (assuming the girth fixed), the heat dissi- pated per unit of surface will be approximately proportional inversely to the square of the coil length. Therefore, if the girth of the permanent winding were identical with that of the test winding, the proper length of the permanent coil would be given by the equation. LtX V^ = Lc (1) in which Lt is the length of the test coil and Lc the calculated length of the permanent coil on the basis of unchanged girth. Table III (divided into lour sections, Ilia, Illb, IIIc and Hid,) gives the corrected coil length, Lc, corresponding to a considerable practical range of test coil lengths, Lt, and ratios of Ik to Ih. If no correction in the mean length per turn 114 ELECTROMAGNETS. CO CO CM CO CO ooco^co Tf(NHH(N ©t^iOO-lO COiO00i-iiO © t^coos ©© © i— < co CMi-H©© I>tHIO©© lO-^COtNr-H CM CM CM CM CM ©©00 00 l> CM'-H'-ItHtH ©LO^^CO CMCMi-HrH© N CM CM CM CO CM©©*— 1 lO 00©1>CMCO ©©!>©r-H ©00 CO© Tt<©00 00 00 iO00©iOl> OCMiO©CO (NOOiSCO CO CO CM CM CM lO"^ CO CM CM CM CM CM CM CM i-H©00t^© CM CM i-h t-i i-i ©lO^COCO s CM CO I>CMCMiO iOiO00(MI> »o ioco ■^ t-i ©© i-i CM i-i CO 00 lO CO lOOOrH^-ll^ CMCMrt<©00 OOOCOiOCO "^ co co co co CMtH©00!> CO CO CO CM CM b^iOCOCMi-H CM CM CM CM CM ©©G0I>© CM T-I rH rH i-H o O CM o co CO lO CO 00 00 lO ^ 00 CO 00 lO CM ©co*ooo© ©©ocooo lOCOCMiHO lO CO CM CM CM i-iOOCO^CN lO ^ ^ ^ ^ ©ost^coio ^CO CO CO CO ^^©00© CO CO CO CM CM lO-^COCMv-t CM CM CM CM CM © © OS CM THiO CO -^COiO CO CM CO lOLOl^ ^ COiO00l>© ©CM 00 l> 00 ©i> co CM©CM 00 Tt©^ CMOI>iOCO tFtFCOCOCO CM©©00© CO CO CM CM CM 00 00 00 CM COCM CO CM CO 00 CO CO COCOrJH OSTfl*H COuO©i-iCO HNt>HN ©r-H CO ©©©CO 00 HNMON oot>i>t>co "tfCMOOO© ©©©iCuO ^©l>iOCM lO >o ^ "* ^ ©oo©uoco ^cocococo t> 1> 1—1 1> CM CM t^CO l>iO*OCM ©I^OOt-hcO © Tf © CM CM00© ©TjHCMt> CMl>«OiO© CMl>CO©iO ©©©00 00 i-HOOioco© 00t^l>l>I> oo^©©co ©©©lOiO t-i00©^CM lO ^ ^ ^ ^ CO CO CO CM rHOi^COCO •^COiO COCO©CM CMJ>iOt^ 00 lOCM »OLOt^i-i00 ©CMl>CMt^ NNHHO CO©LOCM© ©OS OS OS 00 ©©iOi-fl> 00 00l>t^© rf t-h0C©CO ©©iO»OiO iO LO LO OirMi-l©^ »o>o CMCO©COCO IOO0 t>i-H iOI>00t^ I> COtQCM © -#iO©00 00 CMiOOO'-iiO ©iO^rt 00 t^ l> l> © <* <* • CM o CO00l>COr-l iOO CO CM CMCSr-iCOCO to ©CO 00 t^oo-^oo© ©t^COCMiO iOOOCOOOt-h ©©00l>t^ CM HHHH Tt^l>CMCOi-i CO iO »0 ""*• ^t 1 ©0C©"#00 CO CM CM HO CMI>C0©»O CJ t^t^cOcOcO OS ©rH CM CO ©©©lO^O rtHiOOl>00 Cir-i COlO t"-» Tt< CO CO CO CM ©r-icoiot> r-i CM CM CM CM CM CM CM CM CM CM CO CO CO CO CO ^ ^ ^f ^ lO ©T-HCMCO0 lO CO h- 00 OS CM CM CM CM CM io»o©©© ©CMrt<©00 CO CO CO CO CO ix^ooooo © CM tJH © 00 WINDING OF ELECTROMAGNETS. 115 ooioooeot^ oocoooioio qicctfrHco c*oooiCMt> rH CM Tfi 00 CM I>COO>COCO OOCMOCO COrHCiOOCO O05 00l>l> COCOO*OiO iO^^C*CO CO CO CM CM CM T^t^ l>t> COCMiOCOiO CMCOiO^OO 00rHi>cO00 OOCOt^OOi-t 100*0'— U> "^OOCOCJiO NOMOW 0000I>l>CO CO iO »0 "tf "^ ■<* -^ CO CO CO CMCO CM CPIOCOOO COCO CO CM I> tO "^ O 00 rH 10 iO rH CO t> CM I> -<*< © 1> iO CM CO-tfCOCMrH ©00*000 OOt^COCOiO iOiO^t^tH ■*00CO(M CMt^CMiOiO CMOrHcO rHO0rHl>00 TjCO COI>OCOb- CMCMiOOOCO oocoococo ©OOt^iO"*" COCMCMrH© ©OOOI>I> CO CO CO iO »0 iOt-iio OOrHCOrH 00iOI>iO00 I> <* rt< 00 "# THrHrHCOiO 00 1> b- O CM iCOiOHN lOCOT-HOiOO t^cDiO^CO CMrH©©© 0000l>t^cO CMCMCMrHrH rHrHrHrHrH rH rH rH 00CMCO00 LO ^Oi '"tfCOCOOi (MiOiOfM^ "*iiO©CiCM COMHOO (NI>iO^CO OOi-nOOO (NOil>rtHCO rH©©00b- CO rt< CO CM ^ 00050500 COCMCMCMCM (M(NHHH rH rH rH rH rH rH rH COCOCOrHiO OJiOCO OOCMCOiOCO iO "tfl>00 OCMrHiOCM COCOOt>iO ^CON© CO00©C01> ON^HQ l>lO"t>00 00© 05 0^ CM CO ^i-OCOr^OO OOCJO'HiM (NC0^iO»O COOOOrHCO iOCOOOOh < 5^ T ^ C 5 rH COrHCOrHCO rH rH CM CM CM CM CM CM CO CO iOiOCOCOt* l>»0000OiOi ©rHCMCOH/i lOCOt^OOCl 5^ .aS H a> 3^.3 116 ELECTROMAGNETS. is necessary, this set of tables will, of course, give the proper length, L, of the permanent coil, which in such cases is identical with Lc. If a cor- rection in mean length is necessary and is such as to alter materially the girth of the coil, and, therefore, the radiating surface per unit of length, after making the correction in mean length as explained in a preceding paragraph, and ascertaining the calculated length of coil, Lc, by means of Table III, the final value for the length (L) of the permanent coil may be obtained by means of the formula Lc xGt —G—= L (2) being the girth that the permanent coil will have after correcting the mean length per turn, and Gt the girth of the test coil. For convenience in making corrections in the mean length per turn and the girth of the finished coil, Table IV (divided into IVa to IVe inclusive) has been prepared. This gives the depth of coil that will be obtained with different numbers of layers of the standard sizes of magnet wire, single and double cotton covered. The table is based on the insulation thicknesses used by the Roebling factory, and while the coil depths are given to the second and third decimal places, it will, of course, be understood that this is not intended as an in- timation that coils can be wound in practice to any such degree of accuracy, even if the insulation ran absolutely uniform always, which it does not do. The full figures are given in this, as in Tables I and II, merely in order that one may see what the exact theoretical values are. The table has not been made to include very small sizes of wire, for the reason that any approach to accuracy in calculations based on the insulated diameters of such wires is impossible. For coils wound around a continuously convex surface, such as that of a bobbin for a round magnet core or one of oval cross section, the mean length per turn of wire is readily obtained by means of the formula g + ,r d = Lm (3) in which g is the girth of the bobbin or former in which the coil is wound and d is the depth of the winding (in inch measure, or whatever unit of linear measurement may be used; not in layers). The girth of the coil will be obtainable by means of the formula g + 2-nd = G (4) The mean length per turn in a coil wound on a bobbin of substantially rectangular cross section will be greater than the value given this formula on account of the bulging of the wire away from the core in the parts of the winding which cover the straight surfaces of the bobbin or former. This is also true, and to a greater extent, of the girths of the finished coil. WINDING OF ELECTROMAGNETS. 117 Table Ilia. — Tor correcting* JLeng'tli of Jlagrnet Coil. Ix Length of Test Coil, Lt. Ih 1* H H U 2 2* k 2f 2i 2| 2f 2| .4 . . .95 1.03 1.11 1.19 1.27 1.35 1.43 1.5 1.58 1.66 1.74 1.82 .425 . .98 1.06 1.14 1.22 1.31 1.39 1.47 1.55 1.63 1.71 1.8 1.87 .45 . . 1.01 1.09 1.17 1.26 1.34 1.43 1.51 1.6 1.68 1.76 1.85 1.93 .475 . 1.03 1.12 1.21 1.29 1.38 1.47 1.55 1.64 1.72 1.81 1.9 1.98 .5 . . 1.06 1.15 1.24 1.33 1.42 1.5 1.59 1.68 1.77 1.86 1.95 2.03 .525 . 1.09 1.18 1.27 1.36 1.45 1.54 1.63 1.72 1.81 1.9 1.99 2.08 .55 . . 1.12 1.21 1.3 1.39 1.48 1.58 1.67 1.76 1.86 1.95 2.04 2.13 .575 . 1.14 1.23 1.33 1.42 1.52 1.61 1.71 1.8 1.9 1.99 2.09 2.18 .6 . . 1.16 1.26 1.36 1.45 1.55 1.65 1.74 1.84 1.94 2.03 2.13 2.23 .625 . 1.18 1.29 1.38 1.48 1.58 1.68 1.78 1.88 1.98 2.08 2.17 2.27 .65 . . 1.21 1.31 1.41 1.51 1.61 1.71 1.82 1.92 2.02 2.12 2.22 2.32 .675 . 1.23 1.34 1.44 1.54 1.64 1.75 1.85 1.95 2.05 2.16 2.26 2.36 .7 . . 1.26 1.36 1.47 1.57 1.67 1.78 1.88 1.99 2.09 2.2 2.3 2.41 .725 . 1.28 1.38 1.49 1.6 1.7 1.81 1.92 2.02 2.13 2.24 2.34 2.45 .75 . . 1.3 1.41 1.52 1.62 1.73 1.84 1.95 2.06 2.17 2.27 2.38 2.49 .8 . . 1.34 1.46 1.57 1.68 1.79 1.9 2.01 2.13 2.24 2.35 2.46 2.57 .85 . . 1.39 1.5 1.61 1.73 1.85 1.96 2.08 2.19 2.31 2.42 2.54 2.65 .9 . . 1.42 1.54 1.66 1.78 1.9 2.02 2.14 2.25 2.37 2.49 2.61 2.73 .95 . . 1.46 1.58 1.71 1.83 1.95 2.07 2.19 2.32 2.44 2.56 2.68 2.8 1. . . . 1.5 1.63 1.75 1.88 2. 2.13 2.25 2.38 2.5 2.63 2.75 2.88 1.05 . . 1.54 1.67 1.79 1.92 2.05 2.18 2.31 2.44 2.56 2.69 2.82 2.95 1.1 . . 1.57 1.71 1.84 1.97 2.1 2.23 2.36 2.49 2.62 2.75 2.88 3.02 1.2 . . 1.64 1.78 1.92 2.05 2.19 2.33 2.47 2.6 2.74 2.88 3.01 3.15 1.3 . . 1.71 1.85 1.99 2.14 2.28 2.42 2.57 2.71 2.85 3. 3.14 3.28 1.4 . . 1.78 1.92 2.07 2.22 2.37 2.51 2.66 2.81 2.96 3.11 3.25 3.4 1.5 . . 1.84 1.99 2.14 2.3 2.45 2.6 2.76 2.91 3.06 3.22 3.37 3.52 1.6 . . 1.9 2.06 2.21 2.37 2.53 2.69 2.85 3.01 3.16 3.32 3.48 3.64 1.7 . . 1.96 2.12 2.28 2.45 2.61 2.77 2.93 3.1 3.26 3.42 3.59 3.75 1.8 . . 2.01 2.18 2.35 2.52 2.68 2.85 3.02 3.19 3.35 3.52 3.69 3.86 1.9 . . 2.07 2.24 2.41 2.59 2.76 2.93 3.1 3.27 3.45 3.62 3.79 3.96 2. . . . 2.12 2.3 2.48 2.65 2.83 3 3.18 3.36 3.54 3.71 3.89 4.07 2.1 . . 2.17 2.36 2.54 2.72 2.9 3 '.08 3.26 3.44 3.62 3.81 3.99 4.17 2.2 . . 2.23 2.41 2.6 2.78 2.97 3.15 3.34 3.52 3.71 3.89 4.08 4.27 2.3 . . 2.28 2.47 2.65 2.84 3.03 3.22 3.41 3.6 3.79 3.98 4.17 4.36 2.4 . . 2.32 2.52 2.71 2.91 3.1 3.29 3.49 3.68 3.87 4.07 4.26 4.46 The above numbers (in the body of the table) are corrected lengths, Lc 118 ELECTROMAGNETS. Tabic 1 1 lb. — For correcting- Teng"tli of Mag-net Coil. Ix Length of Test Coil, Lt. Ih 3 3£ 3* 31 3i 31 3i 31 4 4* 4i 4f .4 . . 1.9 1.98 2.06 2.14 2.22 2.3 2.37 2.45 2.53 2.61 2.69 2.77 .425 . 1.96 2.04 2.12 2.2 2.28 2.36 2.45 2.53 2.61 2.69 2.77 2.85 .45 . . 2.01 2.1 2.18 2.26 2.35 2.43 2.52 2.6 2.68 2.77 2.85 2.94 .475 . 2.07 2.15 2.24 2.33 2.41 2.5 2.58 2.67 2.76 2.84 2.93 3.02 .5 . . 2.12 2.21 2.3 2.39 2.48 2.56 2.65 2.74 2.83 2.92 3.01 3.09 .525 . 2.18 2.26 2.36 2.45 2.54 2.63 2.72 2.81 2.9 2.99 3.08 3.17 .55 . . 2.23 2.32 2.41 2.5 2.60 2.69 2.78 2.87 2.97 3.06 3.15 3.23 .575 . 2.28 2.37 2.46 2.56 2.65 2.75 2.84 2.94 3.03 3.13 3.22 3.31 .6 . . 2.32 2 42 2.52 2.62 2.71 2.81 2.91 3. 3.1 3.2 3.29 3.39 .625 . 2.37 2.47 2.57 2.67 2.77 2.87 2.97 3.06 3.16 3.26 3.36 3.46 .65 . . 2.42 2.52 2.62 2.72 2.82 2.92 3.02 3.13 3.23 3.33 3.43 3.53 .675 . 2.46 2.57 2.67 2.77 2.88 2.98 3.08 3.19 3.29 3.39 3.49 3.59 .7 . . 2.51 2.62 2.72 2.82 2.93 3.03 3.14 3.24 3.35 3.45 3.56 3.66 .725 . 2.56 2.66 2.77 2.87 2.98 3.09 3.19 3.3 3.41 3.51 3.62 3.73 .75 . . 2.6 2.71 2.81 2.92 3.03 3.14 3.25 3.36 3.46 3.57 3.68 3.79 .8 . . 2.68 2.8 2.91 3.02 3.13 3.24 3.35 3.47 3.58 3.69 3.8 3.91 .85 . . 2.77 2.88 3. 3.11 3.23 3.34 3.46 3.57 3.69 3.81 3.92 4.03 .9 . . 2.84 2.97 3.09 3.2 3.32 3.44 3.56 3.68 3.8 3.91 4.03 4.15 .95 . . 2.92 3.05 3.17 3.29 3.41 3.53 3.66 3.78 3.9 4.02 4.14 4.26 1. . . . 3. 3.13 3.25 3.38 3.5 3.63 3.75 3.88 4. 4.13 4.25 4.38 1.05 . . 3.07 3.2 3.33 3.46 3.59 3.72 3.84 3.97 4.1 4.23 4.36 4.48 1.1 . . 3.14 3.28 3.41 3.54 3.67 3.8 3.93 4.06 4.2 4.33 4.46 4.59 1.15 . . 3.21 3.35 3.49 3.62 3.75 3.89 4.02 4.16 4.29 4.42 4.56 4.69 1.2 . . 3.28 3.44 3.58 3.72 3.85 3.99 4.13 4.27 4.4 4.54 4.68 4.82 1.25 . . 3.35 3.49 3.63 3.77 3.91 4.05 4.19 4.32 4.47 4.61 4.75 4.89 1.3 . . 3.42 3.56 3.71 3.85 3.99 4.13 4.28 4.42 4.56 4.7 4.85 4.99 1.35 . . 3.49 3.68 3.78 3.92 4.07 4.21 4.36 4.5 4.65 4.79 4.94 5.08 1.4 . . 3.55 3.7 3.85 3.99 4.14 4.29 4.44 4.59 4.73 4.88 5.03 5.18 1.45 . . 3.61 3.76 3.91 4.07 4.22 4.37 4.52 4.67 4.82 4.97 5.12 5.27 1.5 . . 3.67 3.83 3.98 4.13 4.29 4.44 4.59 4.75 4.9 5.05 5.21 5.36 1.6 . . 3.85 3.95 4.11 4.27 4.43 4.59 4.75 4.9 5.06 5.22 5.38 5.53 1.7 . . 3.91 4.08 4.24 4.4 4.56 4.73 4.89 5.05 5.22 5.38 5.54 5.71 1.8 . . 4.02 4.19 4.36 4.53 4.7 4.86 5.03 5.2 5.37 5.54 5.7 5.87 1.9 . . 4.14 4.31 4.48 4.65 4.8S 5. 5.17 5.34 5.51 5.69 5.86 6.03 2. . . . 4.25 4.42 4.6 4.77 4.95 5.13 5.31 5.48 5.66 5.83 6.01 6.19 The above numbers (in the body of the table) are corrected lengths, Lc. WINDING OF ELECTROMAGNETS. 119 Table IHc. — JFor correcting* T^ng-t Ii of Magpiet Coil. Ix Length of Test Coil, Lt. Ih 4i 4| 41 41 5 5i 5* 51 5* 51 51 51 .5 . . 3.18 3.27 3.36 3.45 3.54 3.62 3.71 3.8 3.89 3.98 4.07 4.16 .525 . 3.26 3.35 3.44 3.53 3.62 3.71 3.81 3.9 3.99 4.08 4.17 4.26 .55 . . 3.34 3.43 3.52 3.62 3.71 3.8 3.9 3.99 4.08 4.17 4.27 4.36 .575 . 3.41 3.51 3.6 3.7 3.79 3.89 3.98 4.08 4.17 4.27 4.36 4.46 .6 . . 3.49 3.58 3.68 3.78 3.87 3.97 4.07 4.16 4.26 4.36 4.46 4.55 .625 . 3.56 3.66 3.76 3.86 3.95 4.05 4.15 4.25 4.35 4.45 4.55 4.65 .65 . . 3.63 3.73 3.83 3.93 4.03 4.13 4.23 4.33 4.43 4.54 4.64 4.74 .675 . 3.7 3.8 3.9 4.01 4.11 4.21 4.31 4.42 4.52 4.62 4.72 4.83 .7 . . 3.77 3.87 3.97 4.08 4.78 4.29 4.39 4.5 4.6 4.71 4.81 4.92 .725 . 3.83 3.94 4.04 4.15 4.26 4.37 4.47 4.58 4.68 4.79 4.9 5.0 .75 . . 3.9 4.01 4.11 4.22 4.33 4.44 4.55 4.66 4.76 4.87 4.98 5.09 .775 . 4.01 4.07 4.18 4.29 4.4 4.51 4.62 4.73 4.84 4.95 5.06 5.17 .8 . . 4.03 4.14 4.25 4.36 4.47 4.58 4.7 4.81 4.92 5.03 5.14 5.25 .825 . 4.09 4.2 4.32 4.43 4.54 4.66 4.77 4.88 5. 5.11 5.22 5.34 .85 . . 4.15 4.27 4.38 4.5 4.61 4.73 4.84 4.96 5.07 5.19 5.3 5.42 .875 . 4.21 4.33 4.44 4.56 4.68 4.8 4.91 5.03 5.15 5.26 5.38 5.5 .9 . . 4.27 4.39 4.51 4.63 4.74 4.86 4.98 5.1 5.22 5.34 5.46 5.57 .925 . 4.33 4.45 4.57 4.69 4.81 4.93 5.05 5.17 5.29 5.41 5.53 5.65 .95 . . 4.39 4.51 4.63 4.75 4.87 5. 5.12 5.24 5.36 5.48 5.61 5.73 1. . . . 4.5 4.63 4.75 4.88 5. 5.13 5.25 5.38 5.5 5.63 5.75 5.88 1.05 . . 4.61 4.74 4.87 5. 5.12 5.25 5.38 5.51 5.64 5.76 5.89 6.02 1.1 . . 4.72 4.85 4.98 5.11 5.25 5.38 5.51 5.64 5.77 5.9 6.03 6.16 1.15 . . 4.83 4.96 5.09 5.23 5.36 5.5 5.63 5.76 5.9 6.03 6.17 6.3 1.2 . . 4.96 5.07 5.2 5.34 5.48 5.61 5.75 5.89 6.03 6.16 6.3. 6.44 1.25 . . 5.03 5.17 5.31 5.45 5.59 5.73 5.87 6.01 6.15 6.29 6.43 6.57 1.3 . . 5.13 5.27 5.42 5.56 5.7 5.84 5.99 6.13 6.27 6.41 6.56 6.7 1.35 . . 5.23 5.37 5.52 5.67 5.81 5.96 6.1 6.25 6.39 6.54 6.68 6.83 1.4 . . 5.33 5.47 5.62 5.77 5.92 6.07 6.21 6.36 6.51 6.66 6.81 6.95 1.45 . . 5.42 5.57 5.72 5.87 6.02 6.17 6.32 6.47 6.62 6.77 6.93 7.08 1.5 . . 5.51 5.67 5.82 5.97 6.12 6.28 6.43 6.58 6.74 6.89 7.04 7.2 1.55 . . 5.6 5.76 5.91 6.07 6.23 6.38 6.54 6.69 6.85 7. 7.16 7.32 1.6 . . 5.69 6.85 6.01 6.17 6.33 6.48 6.64 6.8 6.96 7.12 7.27 7.43 1.65 . . 5.78 5.94 6.1 6.26 6.42 6.58 6.74 6.91 7.07 7.23 7.39 7.55 1.7 . . 5.87 6.03 6.19 6.36 6.52 6.68 6.85 7.01 7.17 7.33 7.5 7.66 1.75 . . 5.96 6.12 6.28 6.45 6.61 6.78 6.95 7.11 7.28 7.44 7.61 7.77 1.8 . . 6.04 6.21 6.37 6.54 6.71 6.88 7.05 7.21 7.38 7.55 7.72 7.88 1.85 . . 6.12 6.29 6.46 6.63 6.8 6.97 7.14 7.31 7.48 7.65 7.82 7.99 1.9 . . 6.2 6.38 6.55 6.72 6.89 7.07 7.24 7.41 7.58 7.75 7.93 8.1 1.95 . . 6.28 6.46 6.63 6.81 6.98 7.16 7.33 7.51 7.68 7.86 8.03 8.21 2. . . . 6.37 6.54 6.72 6.9 7.07 7.25 7.42 7.6 7.78 7.96 8.13 8.31 The above numbers (in the body of the table) are corrected lengths, he. 120 ELECTKOM AGNETS . Table Hid. — Vor correcting* JLeiig-rf i of >1 ag-net Coil. lx Length of Test Coil, U Ih 6 6* 6i 6| 6i 6| 6f 61 7 7£ 7i 71 .5 . . 4.24 4.33 4.44 4.51 4.6 4.69 4.77 4.86 4.95 5.04 5.13 5.22 .525 . 4.35 4.44 4.53 4.62 4.71 4.8 4.89 4.98 5.07 5.16 5.25 5.34 .55 . . 4.45 4.54 4.64 4.73 4.82 4.91 5.01 5.1 5.19 5.29 5.38 5.47 .575 . 4.55 4.65 4.74 4.83 4.93 5.02 5.12 5.21 5.31 5.4 5.5 5.59 .6 . . 4.65 4.75 4.84 4.94 5.04 5.13 5.23 5.33 5.42 5.52 5.62 5.71 .625 . 4.75 4.84 4.94 5.04 5.14 5.24 5.34 5.44 5.53 5.63 5.73 5.83 .65 . . 4.84 4.94 5.04 5.14 5.24 5.34 5.44 5.54 5.64 5.75 5.85 5.95 .675 . 4.93 5.03 5.14 5.24 5.34 5.44 5.55 5.65 5.75 5.85 5.96 6.06 .7 . . 5.02 5.13 5.23 5.33 5.44 5.54 5.65 5.75 5.86 5.96 6.07 6.17 .725 . 5.11 5.22 5.32 5.43 5.53 5.64 5.75 5.86 5.96 6.07 6.17 6.28 .75 . . 5.2 5.3 5.41 5.52 5.63 5.74 5.85 5.95 6.06 6.17 6.28 6.39 .775 . 5.28 5.39 5.5 5.61 5.72 5.83 5.94 6.05 6.16 6.27 6.39 6.49 .8 . . 5.37 5.48 5.59 5.7 5.81 5.93 6.04 6.15 6.26 6.37 6.49 6.6 .825 . 5.45 5.56 5.68 5.79 5.91 6.02 6.13 6.25 6.36 6.47 6.59 6.7 .85 . . 5.53 5.65 5.76 5.88 5.99 6.11 6.22 6.34 6.45 6.57 6.69 6.8 .875 . 5.61 5.73 5.85 5.96 6.08 6.2 6.31 6.43 6.55 6.67 6.78 6.9 .9 . . 5.69 5.81 5.93 6.05 6.17 6.29 6.4 6.52 6.64 6.75 6.88 7. .925 . 5.77 5.89 6.01 6.13 6.25 6.37 6.49 6.61 6.73 6.85 6.97 7.09 .95 . . 5.85 5.97 6.09 6.21 6.34 6.46 6.58 6.7 6.82 6.95 7.07 7.19 1. . . . 6. 6.13 6.25 6.38 6.5 6.63 6.75 6.88 7. 7.13 7.25 7.38 1.05 . . 6.15 6.28 6.41 6.53 6.66 6.79 6.92 7.05 7.17 7.3 7.43 7.56 1.1 . . 6.29 6.43 4.56 6.69 6.82 6.95 7.08 7.21 7.34 7.47 7.61 7.74 1.15 . . 6.44 6.57 6.7 6.84 6.97 7.11 7.24 7.37 7.51 7.64 7.78 7.91 1.2 . . 6.57 6.71 6.85 6.99 7.12 7.26 7.39 7.53 7.67 7.81 7.94 8.08 1.25 . . 6.71 6.85 6.99 7.13 7.27 7.41 7.55 7.69 7.83 7.97 8.11 8.25 1.3 . . 6.84 6.98 7.13 7.27 7.41 7.55 7.7 7.84 7.98 8.13 8.27 8.41 1.35 . . 6.97 7.12 7.26 7.41 7.55 7.7 7.84 7.99 8.13 8.28 8.43 8.57 1.4 . . 7.1 7.25 7.4 7.54 7.69 7.84 7.99 8.13 8.28 8.43 8.58 8.73 1.45 . . 7.23 7.38 7.53 7.68 7.83 7.98 8.13 8.28 8.43 8.58 8.73 8.88 1.5 . . 7.35 7.5 7.66 7.81 7.96 8.11 8.27 8.42 8.57 8.73 8.88 9.03 1.55 . . 7.47 7.63 7.78 7.94 8.09 8.25 8.4 8.56 8.72 8.87 9.03 9.18 1.6 . . 7.59 7.75 7.91 8.07 8.22 8.38 8.54 8.7 8.86 9.01 9.17 9.33 1.65 . . 7.71 7.87 8.03 8.19 8.35 8.51 8.67 8.83 8.99 9.15 9.31 9.47 1.7 . . 7.82 7.99 8.15 8.31 8.48 8.64 8.8 8.96 9.13 9.29 9.45 9.62 1.75 . . 7.94 8.1 8.27 8.43 8.6 8.77 8.93 9.09 9.26 9.43 9.59 9.76 1.8 . . 8.05 8.22 8.39 8.55 8.72 8.89 9.06 9.22 9.39 9.56 9.73 9.9 1.85 . . 8.16 8.33 8.5 8.67 8.84 9.01 9.18 9.35 9.52 9.69 9.86 10.03 1.9 . . 8.27 8.44 8.62 8.79 8.96 9.13 9.3 9.48 9.65 9.82 9.99 10.17 1.95 . . 8.38 8.55 8.73 8.9 9.08 9.25 9.43 9.6 9.78 9.95 10.13 10.3 2. . . . 8.49 8.66 8.84 9.02 9.19 9.37 9.55 9.72 9.9 10.08 10.25 10.43 The above numbers (in the body of the table) are corrected lengths, Lc» WINDING OF ELECTROMAGNETS. 121 Table IVa.-Iinear Space occupied toy Single Cotton- Covered Wires. Turns or Layers. Wire Numbers, B. & S. Gauge : 6 .432 648 864 08 296 512 728 994 16 38 59 81 03 24 46 67 89 11 32 53 75 97 .388 ,582 ,776 ,97 164 358 552 746 94 ,14 33 52 ,72 ,91 .11 3 ,49 ,69 .88 07 27 46 65 .85 .05 344 516 688 86 032 204 376 548 72 89 07 24 41 58 75 93 1 27 44 61 78 95 125 3 47 64 81 98 16 .308 .462 .616 .77 .924 ,078 232 386 .54 ,69 .85 !l6 .31 .47 .62 .77 .93 .08 .24 .39 .54 .69 .85 .16 .31 .46 .62 .77 .93 .08 ,274 ,411 ,548 685 822 959 096 233 37 51 64 78 92 06 19 33 ,47 61 ,74 ,88 02 15 29 43 56 ,7 83 97 11 25 38 52 65 79 93 07 9 ~244 366 488 61 732 854 976 098 22 34 47 59 71 83 95 06 2 32 44 56 69 81 93 05 17 29 41 54 66 78 9 02 14 27 39 51 63 76 88 10 .216 .324 .432 .54 .648 .756 .864 .972 .08 .19 .3 .41 .51 .62 .73 .84 .95 .05 .16 .27 .38 .49 .59 .7 .81 .92 .03 .13 .24 .35 .45 .56 .67 .78 .89 .99 .1 .21 .32 .42 .53 .64 .75 .86 .97 11 7l94 .291 .388 .485 .582 .679 .776 .873 .97 ,07 ,17 ,26 .36 .46 .55 ,65 .75 .85 ,94 .04 .14 ,23 33 ,43 ,52 ,62 ,72 82 91 ,1 2 ,29 39 ,49 59 ,68 ,78 ,88 ,97 07 ,17 ,27 36 46 56 66 76 ,85 12 7l74 ,261 .348 ,435 ,522 ,609 .696 .783 ,87 .96 05 ,13 22 31 39 ,48 ,57 66 .74 83 92 !09 ,18 ,26 35 44 53 61 ,7 79 87 96 04 13 22 3 39 48 56 65 74 83 91 13 14 0.14 0.21 0.28 0.35 0.42 0.49 0.56 0.63 0.7 0.77 0.84 0.91 .98 1.05 1.12 1.19 1.26 1.33 1.4 1.4* 1.54 1.61 1.68 1.75 1.82 1.89 1.96 2.03 2.1 2.17 2.24 2.31 2.38 2.45 2.52 2.59 2.66 2.73 2.8 2.87 2.94 3.01 3.08 3.15 3.22 3.29 3.36 3.43 3.5 3.64 3.78 3.92 4.06 4.2 4.34 4.48 122 ELECTROMAGNETS. TA BLE IVl>.- Linear Space occupied by Single Cotton- Covered Wires. Turns or Layers. Wire Numbers, B. & S. Gauge : 15 0.38 0.44 0.5 0.57 0.63 0.75 0.82 0.88 0.95 1.01 1.07 1.13 1.2 1.26 1.32 1.39 1.45 1.51 1.57 1.64 1.7 1.76 1.83 1 1.95 2.02 2.08 2.14 2.2 2.27 2.33 2.39 2.46 2.52 2.58 2.65 2.71 2.77 2.83 2 2.96 3.02 3.09 3.15 3.27 3.4 3.53 3.65 3.78 3.91 4.03 4.16 4.28 4.41 16 17 0.34 0.4 0.46 0.51 0.57 0.63 0.68 0.74 0.8 0.85 0.91 0.97 1.03 1.08 1.14 1.2 1.25 1.31 1.37 1.42 1.48 1.54 1.6 1.65 1.71 1.77 1.82 1.88 1.94 2. 2.05 2.11 2.17 2.22 2.28 2.34 2.39 2.45 2.51 2.56 2.62 2 2.73 2.79 2.85 2.96 3.08 3.19 3.31 3.42 3.53 3.65 3.76 3 3.99 18 0.31 0.36 0.41 0.46 0.51 0.56 0.61 0.66 0.71 0.76 0.82 0.87 0.92 0.97 1.02 1.07 1.12 1.17 1.22 1.27 1.33 1.38 1.43 1.48 1.53 1.58 1.63 1.68 1.73 1.78 1.84 1. 1.94 1 2.04 2.09 2.14 2.19 2.24 2.29 2.35 2.4 2.45 2.5 2.55 2.65 2.75 2 2.96 3.06 3.16 3.26 3.37 3.47 3.57 19 0.28 0.32 0.37 0.41 0.46 0.51 0.55 0.6 0.64 0.69 0.74 0.78 0.83 0.87 0.92 0.97 1.01 1.06 1.1 1.15 1.2 1.24 1.29 1.33 1.38 1.43 1.47 1.52 1.56 1.61 1.66 1.7 1.75 1.79 1.84 1.89 1.93 1.98 2.02 2.07 2.12 2.16 2.21 2.25 2.3 2.39 2.48 2.58 2.67 2.76 2.85 2.94 3.04 3.13 3.22 20 0.25 0.29 0.34 0.38 0.42 0.46 0.5 0.55 0.59 0.63 0.67 0.72 0.76 0.84 0.88 0.92 0.97 1.01 1.05 1.09 1.13 1.18 1.22 1.26 1.3 1.34 1.39 1.43 1.47 1.51 1.55 1 1.64 1.68 1.72 1.76 1.81 1.85 1.89 1.93 1.97 2.02 2.06 2.1 2.18 2.27 2.35 2.44 2.52 2.6 2.69 2.77 2.86 2.94 21 0.23 0.27 0.3 0.34 0.38 0.42 0.46 0.49 0.53 0.57 0.61 0.65 0.68 0.72 0.76 0.8 0.84 0.87 0.91 0.95 0.99 1.03 1.06 1.1 1.14 1.18 1.22 1.25 1.29 1.33 1.37 1.41 1.44 1.48 1.52 1.56 1.6 1.63 1.67 1.71 1.75 1.79 1.82 1.8' 1.9 1.98 2.05 2.13 22 2 2.28 2.36 2.43 2.51 2.58 2.66 0.2 0.24 0.27 0.31 0.34 0.37 0.41 0.44 0.48 0.51 0.54 0.58 0.61 0.65 0.71 0.75 0.78 0.82 0.85 0.88 0.92 0.95 0.99 1.02 1.05 1.09 1.12 1.16 1.19 1.22 1.26 1.29 1.33 1.36 1.39 1.43 1.46 1.5 1.53 1.56 1 1.63 1.67 1.7 1.77 1.84 1.9 1.97 2.04 2.11 2.18 2.24 2.31 2.38 23 0.19 0.22 0.25 0.28 0.31 0.34 0.37 0.4 0.43 0.46 0.5 0.53 0.56 0.59 0.62 0.65 0.68 0.71 0.74 0.78 0.81 0.84 0.87 0.9 0.93 0.96 0.99 1.02 1.05 1.08 1.12 1.15 1.18 1.21 1.24 1.27 1.3 1.33 1.36 1.39 1.43 24 46 49 52 55 61 67 1.74 1.8 1.86 1.92 1. 2.05 2.11 2.17 0.17 0.2 0.22 0.25 0.28 0.31 0.34 0.36 0.39 0.42 0.45 0.48 0.5 0.53 0.56 0.59 0.62 0.64 0.67 0.7 0.73 0.76 0.78 0.81 0.84 0.87 0.9 0.93 0.95 0.98 1.01 1.04 1.06 1.09 1.12 1.15 1.18 1.2 1.23 1.26 1.29 1.32 1.34 1.37 1.4 1.46 1.51 1.57 1.62 1 1.74 1.79 1.85 1.9 1.96 0.15 0.17 0.2 0.22 0.25 0.27 0.3 0.33 0.35 0.38 0.4 0.42 0.45 0.47 0.5 0.52 0.55 0.57 0.6 0.62 0.65 0.67 0.7 0.72 0.75 0.77 0.8 0.82 0.85 0.87 0.9 0.92 0.95 0.97 1. 1.02 1.05 1.07 1.1 1.12 1.15 1.17 1.2 1.22 1.25 1 3 1 35 1 4 1 45 1 5 1 55 1 .6 1 65 1 « 1 .75 WINDING OF ELECTROMAGNETS. 123 Table IVc. — Linear Space occupied by Single Cotton- Covered Wires. Wire Numbers, B. &S. Gauge : Turns or Layers. 17 18 19 20 21 22 23 24 72 3.67 3.31 3.02 2.74 2.45 2.23 2.02 1.8 74 3.77 3.4 3.11 2.81 2.52 2.29 2.07 1.85 76 3.88 3.5 3.19 2.89 2.58 2.36 2.13 1.9 78 3.98 3.58 3.28 2.96 2.65 2.42 2.18 1.95 80 .... . 4.08 3.68 3.36 3.04 2.72 2.48 2.24 2. 82 4.18 3.77 3.44 3.12 2.79 2.54 2.3 2.05 84 4.28 3.86 3.53 3.19 2.86 2.6 2.35 2.1 86 4.39 3.96 3.61 3.27 2.92 2.67 2.41 2.15 88 4.49 4.05 3.7 3.34 2.99 2.73 2.46 2.2 90 4.59 4.14 3.78 3.42 3.06 2.79 2.52 2.25 92 4.23 3.86 3.5 3.13 2.85 2.58 2.3 94 4.32 3.95 3.57 3.2 2.91 2.63 2.35 96 4.42 4.03 3.65 3.26 2.98 2.69 2.4 98 4.51 4.12 3.72 3.33 3.04 2.74 2.45 100 4.6 4.2 3.8 3.4 3.1 2.8 2.5 102 4.28 3.88 3.47 3.16 2.86 2.55 104 4.37 3.95 3.54 3.22 2.91 2.6 106 4.45 4.03 3.6 3.29 2.97 2.65 108 4.54 4.1 3.67 3.35 3.02 2.7 110 4.18 3.74 3.41 3.08 2.75 112 4.26 3.81 3.47 3.14 2.8 114 4.33 3.88 3.53 3.19 2.85 116 4.41 3.94 3.6 3.25 2.9 118 4.48 4.01 3.66 3.3 2.95 120 4.56 4.08 3.72 3.36 3. Table IVd. — Linear Space occupied by Double Cotton- Covered Wires. Wire Numbers, B. & S. Gauge: Turns or Layers. 4 5 6 7 8 9 10 11 12 13 14 2. . . 0.444 0.4 0.356 0.32 0.284 0.252 0.224 0.202 0.182 0.16 0.15 3. . . 0.666 0.6 0.534 0.48 0.426 0.378 0.336 0.303 0.273 0.24 0.22 4. . . 0.888 0.8 0.712 0.64 0.568 0.504 0.448 0.404 0.364 0.32 0.29 5. . . 1.11 1. 0.89 0.8 0.71 0.63 0.56 0.505 0.455 0.4 0.36 6. . . 1.332 1.2 1.068 0.96 0.852 0.756 0.672 0.606 0.546 0.49 0.44 7. . . 1.554 1.4 1.246 1.12 0.994 0.882 0.784 0.707 0.637 0.57 0.51 8. . . 1.776 1.6 1.424 1.28 1.136 1.008 0.896 0.808 0.728 0.65 0.58 9. . . 1.998 1.8 1.602 1.44 1.278 1.134 1.008 0.909 0.819 0.73 0.66 10. . . 2.22 2. 1.78 1.6 1.42 1.26 1.12 1.01 0.91 0.81 0.73 11. . . 2.442 2.2 1.958 1.76 1.562 1.386 1.232 1.111 1.001 0.89 0.8 124 ELECTROMAGNETS. Table I Vd. — Linear Space occupied by Double Cotton- Covered "Wires. — Continued. Turns or Layers. Wire Numbers, B. & S. Gauge: 2.664 2 108 33 55 136 314 492 67 85 3.03 3.2 3.38 3.56 3.74 3.92 4.09 4.27 4.45 4.63 4.81 4.98 1.92 2.08 2.24 2.4 2.56 2.72 2.88 3.04 3.2 3.36 3.52 3.68 3.84 4. 4.16 4.32 4.48 4.64 4.8 4.96 1.704 1.846 1.988 2.13 2.27 2.41 2.56 2.77 2.84 2.98 3.12 3.27 3.41 3.55 3.69 3.83 3.98 4.12 4.2 4.4 4.54 4.6 4.8 4.97 1.512 1.638 1.764 1.89 2.01 2.14 2.27 2.39 2.52 2.65 2.77 2.9 3.02 3.15 3.28 3.4 3.53 3.65 3.78 3.91 4.03 4.16 4.28 4.41 4.54 4.66 4.79 4.91 5.04 10 1.344 1.456 1.568 1.68 1.79 1.9 2.02 2.13 2.24 2.35 2.46 2.58 2.69 2.8 2.91 3.02 3.14 3.25 3.36 3.47 3.58 3.7 3.81 3.92 4.03 4.14 4.26 4.37 4.48 4.59 4.7 4.82 4.93 5.04 11 12 13 14 1.212 1.313 1.414 1.51 1.62 1.72 1.82 1.92 2.02 2.12 2.22 2.32 2.42 2.53 2.63 2.73 2.83 2.93 3.03 3.13 3.23 3.33 3.43 3.54 3.64 3.74 3.84 3.94 4.04 4.14 4.24 4.34 4.44 4.55 4.65 4.75 4.85 1.092 1.183 1.274 1.36 1.46 1.55 1.64 1.73 1.82 1.91 2. 2.09 2.18 2.28 2.37 2.46 2.55 2.64 2.73 2.82 2.91 3. 3.09 3.19 3.28 3.37 3.46 3.55 3.64 3.73 4.28 4.37 4.46 4.55 4.73 4.91 WINDING OF ELECTROMAGNETS. 125 Table IVe. — linear Space occupied by Double Cotton- Covered Hires. Wire Numbers B. and S. Gauge: Turns or Layers. 15 16 17 18 19 20 21 22 23 24 7 . . . . 0.46 0.42 0.38 0.35 0.32 0.29 0.26 0.24 0.22 0.2 8 . . . . 0.53 0.48 0.43 0.4 0.36 0.33 0.3 0.27 0.25 0.23 9 . . . . 0.59 0.54 0.49 0.45 0.4 0.37 0.34 0.31 0.28 0.26 10 ... . 0.66 0.6 0.54 0.5 0.45 0.41 0.37 0.34 0.31 0.28 11 ... . 0.73 0.66 0.59 0.55 0.5 0.45 0.41 0.38 0.34 0.31 12 ... . 0.79 0.72 0.65 0.59 0.54 0.49 0.45 0.41 0.37 0.34 13 ... . 0.86 0.78 0.71 0.65 0.59 0.53 0.49 0.44 0.41 0.37 14 ... . 0.92 0.84 0.76 0.69 0.63 0.58 0.53 0.48 0.43 0.39 15 ... . 0.99 0.9 0.81 0.74 0.68 0.62 0.56 0.51 0-.47 0.42 16 ... . 1.06 0.96 0.86 0.79 0.72 0.66 0.6 0.54 0.5 0.45 17 ... . 1.12 1.02 0.92 0.84 0.77 0.7 0.64 0.58 0.53 0.48 18 ... . 1.19 1.08 0.97 0.89 0.81 0.74 0.86 0.61 0.56 0.51 19 ... . 1.25 1.14 1.03 0.94 0.86 0.78 0.71 0.65 0.59 0.53 20 ... . 1.32 1.2 1.08 0.99 0.9 0.82 0.75 0.68 0.62 0.56 21 ... . 1.39 1.26 1.13 1.04 0.95 0.86 0.79 0.72 0.65 0.59 22 ... . 1.45 1.32 1.19 1.09 0.99 0.9 0.83 0.75 0.68 0.62 23 ... . 1.52 1.38 1.24 1.14 1.04 0.94 0.86 0.78 0.72 0.65 24 ... . 1.58 1.44 1.3 1.19 1.08 0.98 0.9 0.82 0.75 0.67 25 ... . 1.65 1.5 1.35 1.24 1.13 1.03 0.94 0.85 0.78 0.7 26 ... . 1.72 1.56 1.4 1.29 1.17 1.07 0.98 0.88 0.81 0.73 27 ... . 1.78 1.62 1.46 1.34 1.22 1.11 1.01 0.92 0.84 0.76 28 ... . 1.85 1.68 1.51 1.39 1.26 1.15 1.05 0.95 0.87 0.79 29 ... . 1.91 1.74 1.57 1.44 1.31 1.19 1.09 0.99 0.9 0.81 30 ... . 1.98 1.8 1.62 1.49 1.35 1.23 1.13 1.02 0.93 0.84 31 ... . 2.05 1.86 1.68 1.54 1.4 1.27 1.16 1.06 0.96 0.87 32 ... . 2.11 1.92 1.73 1.58 1.44 1.31 1.2 1.09 0.99 0.9 33 ... . 2.18 1.98 1.78 1.63 1.49 1.35 1.24 1.12 1.02 0.92 34 ... . 2.25 2.04 1.84 1.68 1.53 1.4 1.28 1.16 1.05 0.95 35 ... . 2.31 2.1 1.89 1.73 1.58 1.44 1.31 1.19 1.09 0.98 36 ... . 2.38 2.16 1.95 1.78 1.62 1.48 1.35 1.23 1.12 1.01 37 ... . 2.44 2.22 2. 1.83 1.67 1.52 1.39 1.26 1.15 1.04 38 ... . 2.51 2.28 2.05 1.88 1.71 1.56 1.43 1.29 1.18 1.07 39 ... . 2.58 2.34 2.11 1.93 1.76 1.6 1.46 1.33 1.21 1.09 40 ... . 2.64 2.4 2.16 1.98 1.8 1.64 1.5 1.36 1.24 1.12 41 . . . . 2.71 2.46 2.22 2.03 1.85 1.68 1.54 1.4 1.27 1.15 42 ... . 2.77 2.52 2.27 2.08 1.89 1.72 1.58 1.43 1.3 1.18 43 ... . 2.84 2.58 2.32 2.13 1.94 1.76 1.61 1.46 1.33 1.21 44 ... . 2.91 2.64 2.38 2.18 1.98 1.81 1.65 1.5 1.37 1.23 45 ... . 2.97 2.7 2.43 2.23 2.03 1.85 1.69 1.53 1.4 1.26 46 ... . 3.04 2.76 2.49 2.28 2.07 1.89 1.73 1.57 1.43 1.29 47 ... . 3.1 2.82 2.54 2.33 2.12 1.93 1.76 1.6 1.46 1.32 48 ... . 3.17 2.88 2.59 2.38 2.16 1.97 1.8 1.63 1.49 1.34 49 ... . 3.23 2.94 2.65 2.43 2.21 2.01 1.84 1.67 1.52 1.37 50 ... . 3.3 3. 2.7 2.47 2.25 2.05 1.87 1.7 1.55 1.4 52 ... . 3.43 3.12 2.81 2.57 2.34 2.13 1.95 1.77 1.61 1.46 VM ELECTROMAGNETS. Table IVe. ■Linear Space occupied bv Double Cotton- Covered Wires. — Continued. Turns or Wire numbers, B and £ >. Gauge. Layers. 15 16 17 18 19 20 21 22 23 24 54 . . . 56 . . . 58 . . . 60 . . . 62 . . . 64 . . 66 . . . 68 . . . 70 . . . 72 . . . 74. . . . 76. . . . 78. . . . 80. . . . 82. . . . 84. . . . 86. . . . 88. . . . 90. . . . 92. . . . 94. . . . 96. . . . 98. . . . 100. . . . 102. . . . 104. . . . 106. . . . 108. . . . 110. . . . 112. . . . 114. . . . 116. . . . 118. . . . 120. . . . 122. .. . 3.56 3.7 3.83 3.96 4.09 4.23 4.36 4.49 4.62 4.75 3.24 3.36 3.48 3.6 3.72 3.84 3.96 4.08 4.2 4.32 2.92 3.03 3.13 3.24 3.35 3.46 3.57 3.67 3.78 3.89 4. 4.11 4.21 4.32 4.43 4.54 4.65 4.75 2.67 2.77 2.87 2.97 3.07 3.17 3.27 3 37 3.47 3.57 3.67 3.76 3.86 3.96 4.06 4.16 4.26 4.36 4.46 4.56 4.66 4.75 2.43 2.52 2.61 2.7 2.79 2.88 2.97 3.06 3.15 3.24 3.33 3.42 3.51 3.6 3.60 3.78 3.87 3.96 4.05 4.14 4.23 4.32 4.41 4.5 4.59 4.68 .... 2.22 2.3 2.38 2.46 2.54 2.63 2.71 2.79 2.87 2.95 3.04 3.12 3.2 3.28 3.36 3.45 3.53 3.61 3.69 3.77 3.86 3.94 4.02 4.1 4.18 4.27 4.35 4.43 4.51 4.59 2.03 2.1 2.18 2.25 2.33 2.4 2.48 2.55 2.63 2.7 2.78 2.85 2.93 3. 3.08 3.15 3.23 3.3 3.38 3.45 3.53 3.6 3.68 3.75 3.83 3.9 3.98 4.05 4.13 4.2 4.28 4.35 4.43 4.5 1.84 1.9 1.97 2.04 2.11 2.18 2.25 2.31 2.38 2.45 2.52 2.59 2.65 2.72 2.79 2.86 2.93 2.99 3.06 3.13 3.2 3.27 3.33 3.4 3.47 3.54 3.61 3.67 3.74 3.81 3.88 3.95 4.01 4.08 4.15 1.67 1.74 1.8 1.86 1.92 1.99 2.05 2.11 2.17 2.23 2.3 2.36 2.42 2.48 2.54 2.61 2.67 2.73 2.79 2.85 2.92 2.98 3.04 3.1 3.16 3.23 3.29 3.35 3.41 3.47 3.54 3.6 3.66 3.72 3.78 1.51 1.57 1.63 1.68 1.74 1.79 1.85 1.91 1.96 2.02 2.07 2.13 2.19 2.24 2.3 2.35 2.41 2.47 2.52 2.58 2.63 2.69 2.75 2.8 2.86 2.91 2.97 3.03 3.08 3.14 3.19 3.25 3.31 3.36 3.42 Note. — Because of the compression of the insulation on wires wound in layers, and the tendency of the wires of each layer to "bed" between those of the preceding layer, a given number of layers will occupy from 2% to 8% less space than the same number of turns side by side, according to the size of wire and thickness of the insulation. Most of the difference is due to the compression of insulation, the "bedding" effect being almost negligible. For wires of medium size with single cotton insulation, an allowance of 4% will usually be ample to cover the increase in number of layers within a given 8 pace. WINDING OF ELECTROMAGNETS. 127 Alternatingr-Current Xleetroinagrnets. The cores of electromagnets to be used with alternating currents must be laminated, and the laminations must run at right angles to the direc- tion in which eddy currents would be set up. Eddy currents tend to cir- culate parallel to the coils of the wire, and the laminations must, therefore, be longitudinal to or parallel with the axis of the cores. The coils of an alternating-current electromagnet offer more resistance to the passage of the alternating current than the mere resistance of the conductor in ohms. This extra resistance is called inductance, and this combined with the resistance of the conductor in ohms produces the quality called impedance. (See Index for Impedance, etc.) If L = coefficient of self-induction, N = cycles per second, R = resistance, Impedance = V#2 + 4 ir*WL?\ and, Maximum E.M.F. Maximum current = Mean current = Impedance Mean E.M.F. Impedance. Keating' of Hagn<»t Coil*. Professor Forbes. I = current permissible. r t = resistance of coil at permissible temperature. Permissible temperature = cold r X 1.2. t = rise in temperature C°. s = sq. cms. surface of coil exposed to air. / 0003 X t X s .24 X r x law of the Plunder Electromagnet. Charles R. Underhill gives the following formula as having been found by practise the most accurate and complete for the design of plunger electro- magnets. Let P = pull in pounds. B = flux density in the working air-gap. I = length of the air-gap. IN = ampere-turns in the winding. A = cross section of plunger in sq. in. Pe = pull at 10,000 ampere-turns and 1 sq. in. of plunger. n = ampere-turn factor. L = length of the winding in inches. Then, the pull due to an iron-clad solenoid is APc {IN - n) P = 10,000 - n and, at points along the uniform range of solenoids, the pull for the plunger electromagnet will be ZAT2 Pc(IN - nU ',075,600 I 2 "*" 10,000 - n )' Here I must include the extra length assumed due to the reluctance outside of the working air-gap. = A iu 128 ELECTROMAGNETS. Pull in Pound*, and Ampere-turn factor at Different Points along* an ^Electromagnet. L Pt n 1 2 3 4 - 5 6 33.0 28.3 23.4 19.2 16.0 13.8 12.2 11.0 10.0 9.2 8.4 7.8 7.2 6.8 6.4 6.0 5.7 5.3 5.0 4.7 3600 3150 2800 2500 2200 1970 7 8 9 10 11 12 - - 13 14 15 16 17 18 . . 19 20 1750 1580 1400 1230 1100 960 840 725 625 525 430 350 270 210 To approximate the curve of a plunger electromagnet at points between the center of the winding, and the end of the winding where the plunger enters, assume that the curve is a straight line for the last .4 of the dis- tance; then the pull at any point, la as measured in inches, back from the end of the winding, will be :dx IW laPc (IN - n) 075,600 I 2 ' .4 L (10,000 n)J (8) where L equals length of the winding. In this it is assumed that the winding is approximately as long as the inside of the frame. In cases where a low density in the core is used, the curve for the iron- clad solenoid effect cannot be calculated with so high a degree of accuracy. /fm7777777Zr77 7777777777//A ft//////////////////////////, Figs. 2, 3, 4 and 5. Shapes of Electromagnets. WINDING OF ELECTROMAGNETS. 129 40- "I — 38 > 5 ' r| \ J 5 z o 19 26 > z O O z 24 o Ifi ■0 T I § O \ * Z \ a' \ a" -f 1R i A If I is measured in centimeters and A in square centimeters, p is the resistance of a centimeter cube of the conductor. If I is measured in inches and A in square inches, p is the resistance of an inch cube of the conductor. In telegraph and telephone practice, specific resistance is sometimes expressed as the weight per mile-ohm, which is the weight in pounds of a conductor one mile long having a resistance of one ohm. Another common way of expressing specific resistance is in terms of ohms per mil-foot, i.e., the resistance of a round wire one foot long and 0.001 inch in diameter ; I is then measured in feet and A in circular mils. Microhms per inch cube = 0.3937 X microhms per centimeter cube. Pounds per mile-ohm = 57.07 X microhms per centimeter cube X specific gravity. Ohms per mil-foot == 6.015 X microhms per centimeter cube. 131 132 PROPERTIES OF CONDUCTORS. Specific Conductivity is the reciprocal of specific resistance. If c = specific conductivity cA = J_ C RA' 1 c = -■ P By Relative or Percent ag-e Conductivity of a sample is meant 100 times the ratio of the conductivity of the sample at standard tem- perature to the conductivity of a conductor of the same dimensions made of the standard material and at standard temperature. If p is the specific resistance of the sample at standard temperature and pa is the specific resist- ance of the standard at standard temperature, then Percentage conductivity = 100 — • Po In comparing different materials, the specific resistance should always be determined at the standard temperature, which is usually taken as 0° Centigrade. If it is inconvenient to measure the resistance of the sample at the standard temperature, this may be readily calculated if the tem- perature coefficient a of the sample is known, i.e., P ° = T+Vt where p t is the specific resistance at temperature t. MT atthiessen's Standard of Conductivity, which is the commercial standard, is a copper wire having the following properties at the standard temperature of 0° C. Specific gravity 8.89. Length 1 meter. Weight 1 gram. Resistance .141729 ohms. Specific Resistance 1.594 microhms per cubic centimeter. Relative Conductivity 100%. Specific Resistance, Relative Resistance, and Relative Conductivity of Conductors. Referred to Matthiessen's Standard. Resistance in Microhms at C ' C. Relative Relative Metals: Resis- Conduc- Centimeter Cube. Inch Cube. tance. % tivity. % Silver, annealed . . . 1.47 .579 92.5 108.2 Copper 1.55 .610 97.5 102.6 Copper (Matthiessen's 1.594 .6276 100 100.0 Standard). Gold (99.9% pure) . 2.20 .865 138 72.5 Aluminum (99% pure) 2.56 1.01 161 62.1 Zinc 5.75 2.26 362 27.6 Platinum, annealed . . 8.98 3.53 565 17.7 Iron 9.07 3.57 570 17.6 Nickel 12.3 4.85 778 12.9 Tin 13.1 5.16 828 12.1 Lead 20.4 8.04 1,280 7.82 Antimony 35.2 13.9 2,210 4.53 Mercury 94.3 37.1 5,930 1.69 Bismuth 130. 51.2 8,220 1.22 Carbon (graphitic) . . 2,400-42,000 950-16,700 Carbon (arc light) . . about 4,000 about 1,590 Selenium 6X10 10 2.38 X10 10 GENERAL. 133 Liquids at 18° C. Ohms per Centi- meter Cube. Ohms per Inch Cube. Pure water 2650 30 4.86 1.37 9.18 1.29 21.4 1050 Sea water 11 8 Sulphuric acid, 5% Sulphuric acid, 30% Sulphuric acid, 80% Nitric acid, 30% Zinc sulphate, 24% 1.93 .544 3.64 .512 8.54 Temperature Coefficient. The resistance of a conductor varies with the temperature of the con- ductor. Let R = Resistance at 0° Then R = Resistance at t°. R = Ro(l +'at). a is called the temperature coefficient of the conductor. 100 a is the per- centage change in resistance per degree change in temperature. The following values of the temperature coefficient have been found for temperatures measured in degrees Centigrade and in degrees Fahrenheit. It is to be noted that the coefficients vary considerably with the purity of the conductor. Pure Metals. Silver, annealed . , Copper, annealed Gold (99.9%) . . , Aluminium (99%) . Zinc Platinum, annealed Iron , Nickel Tin Lead Antimony . . . Mercury .... Bismuth .... Centigrade a Fahrenheit a 0.00400 0.00222 0.00428 0.00242 0.00377 0.00210 0.00423 0.00235 0.00406 0.00226 0.00247 0.00137 0.00625 0.00347 0.0062 0.00345 0.00440 0.00245 0.00411 0.00228 0.00389 0.00216 0.00072 0.00044 0.00354 0.00197 Matthiessen's formula for soft copper wire R = R (l + .00387* + .00000597* 2 ). The wire used by Matthiessen was as pure as could be obtained at the time (1860), but in reality contained considerable impurities; the above formula, therefore, is not generally applicable. Later experiments have shown that for all practical work the above equation for copper wire may be written R = R (1 + .0042*) for t in ° C. 134 PROPERTIES OF CONDUCTOKS. 3 *s ■$. * s $ z ft ~ 2 - t s 2 ^ .3.2 8 odod* >5a 00^ •+» O'- CD =2 ► « h a w 55 S fe -« CD.O "o3 S^ 2 e3 09 ofs P 1 CD CO O^P 73 cd oo w oo CD CD CD &8*J , 6.2 Jg S *3Q^2 ! I - V*- 1 oo.S, 73 d o d « 2 5 c3 £-3^ 03 oo*~ d « c3 O c3 £^-~ 2 >*}-& u> d I&sJ !*„•* 'a •£? .213 2 s ££'9 a d © ^-3.2 o bfl •spunoj 'qouj OS OSOS • oc D iq n O I J° ^Sia^ c • CN »o O CO COCO OS OS *>. •jtyiABJ*) onioadg N oi(N • oi d • l> •qouj a.renbg aad spunoj 'n^Suaj^g a I! SU9 X a^'Buii^IXl •Treaj^ •^eag otjioadg CN CN (N -8Q '^nioj Sm^aj^ O OO O OO CO coco "3 88JS8Q CO lOOi t-i 00 »o .iad aouB^sisa^j jo CN COCO OO ^ -*i-i CO CO CN i-H 8S^8J0UJ ^U80 J8J 'O oO V s 'sraqo 1°°J ~IIK jad 80HB^SIS8^J r^ O^ "^ 1^ 00 10 10 cOI> b- 1—1 T-il-i T-i 00 CN t^ T-i lO^ «# CO r-< CO CO •aqno O Oi-I *H CM CO 00 os TJOUJ J8d SUiqOJOIJ\[ ^ rH^ rH T-i T* •9qno CO l^OS O 10 cooo OS i-H CO T-i CO CO CO J8!J9Un;U8Q J8d CN CNCN — 1 d<° . CO CO , co .^: rt a «» ~ CO (S • - • d CD p lQ J ft . Cp CD ft . ft . CD . CD l> 0. 'OS 5 a ft ft O.O- CD CD > OS • m OS ^ s » ■ S-S-s .' if s ^ ""3 c c^ d d w uhausen Dmmerci leming aealed). per ce leming per ce Charpen per ce rpentier per ce leming . nze, Cu mh . . 5.S T3^ 73 2 2 « OS OS,c3 alsSaaS a "S a a§a° d-2 d l. d d u die d^. d^d umini and F umini Dewa umini umini Dewa umini (anne umini (hard umini Dewa umini cent). < o o o o o : •msaj\[ •^psajj oupadg OS O t^o • 1-ICO • HO • •« •O •c coco • • • 0505 • • • OO:';'; ; ; •3 saaaS -3Q ^UIOJ SUI^PI\[ o •O OO •O OO •q saaaSaQ jad aouB^sisay; jo 9SB9J0UI iu83 J8J •*© ■IOH •CO"* 00 oc 05 050i'-i i-h •aqno qonj JOd stuqojoij\[ io ^io»oo o iOO i-iG0t-h 1> CO *rHCO •aqnO ja^ampuao jad sinqoj'oiH *O o 0rB aom^sisa^i onioadg 1tH(N(N ,-h £% .* ©^" a © d 53 d D <3W S • ^ •a . o © g-0 s :*! O 03 ! £eo 12 © c3 ! ° c3 ° ins< att,] and tin Si -*S L d u & ft C * •OV< ► © V 1 a •TSX Tl (.040). ompressed pure). D< onze, cop er . . . o - © 0> N ©o d* . flO 08 ' 3 1 d o c3^^ "S o^ ^ S Q W 3*S d 02 1* 2 2 J3 £> ••> *» » CO a) BQ ss§a O Oh2 O— - c •J2 «j.d ^3-d A -c Woo OO O O , ^3 •TJ^ . ^ O ^ 2-S ^ rdn^« d ^ 0Q «2-^ 0) ^■g d-S «h .-sills, nil? . ."o 'o 'd t- b u £ C 2 • © ©fe aaaaaB rt« O O O O C WW OOOOO 136 PROPERTIES OF CONDUCTORS. •spnnoj -qDuj •OO •COCO •X^tabjq ogioads •coco •oso> spunoj *t[;Su9j;g • 10 C* ^00 c d-* CP o O CD 5 d •S-oSd> 2 d^ jj .» ^ IS '5 > *rt afc-d« a=2 NO &SS SS5«S Jg^' 3 1 |°i *J *.s ocoa)P 00 8 • (NOO^^r^ t>» I> rH i— 1 1— 1 00 OS 00 Oi 00 00 •X^iA'BJf) ogioadg •qouj aj^nbg jad spunoj 'q^Suaa^g •trBap^ '^38jj oijioadg CO CO rH •!-lT-l •00^00 •co^co © © CO o o o •0 O o VB smqo ^ooj fij^ jad aoirB^sisay; CO o o 50 TjH CO CO CO tFCNOCOCO CO OOOOrHCMCN CO CO rH^H > •tJ<"^ ft oJS39a We3 Lj i-a £ w ,9 S S w » ° e £* £ (D^ . a a a© ft.2 d AA ^ ft O o o 138 PROPERTIES OF CONDUCTORS. •spunoj 'qouj lOCO "^©^ co co co T}<«tf COlO HIN i-l •aqno qouj jad soiqojoij\[ 00 id Oi 00 CO 39.8 37.1 4.85 4.89 COOOO <* •aqno ja^anii^naQ jad smqojoiH 'Oo0^ aotn^sisa^j oijioads © © s CO CO CO CO oq oo cot^ OCOCOt* rH ■* •© -d • • -.. .**: •« -a • 9 MM M .« 'o o /w $ 'a -ooo -« 8 CP CJ PHYSICAL AND ELECTRICAL PROPERTIES OF METALS. 139 ^q^o I J° ^^Ai •X^iABjf) oijT09dg •0000 -iQiO iO .^-i^h -COO CO • coco nn- b- o>o> r>t> coco •0000 t ft e —< a V •g « K ■5 6 !■* I ijoiil ajBnbg J9d aiisuaj^ a^rai;{fx o o S 8 •ireaj\[ •^9jj oijioadg • • • n- »o*o . . . t>t^ i> a»a» CMi-h - -^ »o •9qno qoni J9d sranojoipf (NcONCOCOCO "* O CO 00 O CO CO OCO O N00 1-H lOiO 00 •9qno J9^9mi^n90 J9d sraqojoij\[ *OoO^ 90ire^STS9si ogioads CO to 1 CD C3 • 3 -2 fl^ § ."g *3§ .* • a 02 • o -fl • a C 'ft si g s s s-Sii " fe J &s • : ^if's^S?* * 3 3 . 9-dQ cl & c fl fl.So^g^ o flgoooflflfl^flSflg-flg a|flflc!flCflT3flgfl-2 Org lo'l'l'l'lli §'|q|3|^ C- — , G- jl< p- 2- Cu Sm 5 2- a a 5 • cu fl • fl c3 • '-g 3 .v3 . * •s_^ 0) . CD . t^ OO'-li-H qouj aj'Bnbg jad spunoj *q;3u8j;g — I * H C 2 -. 9 ■0 fi -: 1 •uraaj^ •;Bajj oijpadg •coco • r^ioio • O O • rH OS OS • © © ; i-J © © • oo • -»0*0 •COCO • -i-Hi-H *Q S89J3 -8Q ^moj Sm^a^ •q aajSaQ jad aoirB^sisay; jo •oooo •os^co •co^co •cOiO • oco •«*C0 3 O o ^ sraqo 1°°J -l!H J8d aon^sisa^j O CM CO Tt. S'§ g a o « § ° § 8 O J) © fl fid <-> CdpT) c3 o 8 a-**. U * U * g a oo t-© SS |i 16,510 13,090 10,380 8,232 6,528 us ^co efoT 1,624 1,288 1,021 ©CM 1$ 823 ©■># co il ©«oo OS CO t~© fed ©» © t~© usee i-i CO r* *$« rj< © © o©oo^cM^ co^T-^oot- ©i* CO CMOf fJVTlH ««*» t>-lO ©«* © ©■3< cc 00 CM WIN t^§£ ooo 00 t*- fed co t>rH©©i* oqo>-#©;* ooooco o^hnio r - ( © ©© © 222^2 £: J2 ^ ,■, ^<-h©©CO i-iOOCOOiCO HOtOOO) MOOhjC ©©cMa «J»(N00hO itOOrtO ©©CM©1> CO © 00 CO rf" t> t- i-f 00 lOO^riO OS CM © © t> i-H CO Tt< t-i 00 CM CM CM ^< © _ cooo©c~ © cm © "33 © # co os © © ©©.co '"J P^ ^ *": °. c£ J rf ! <:< l r ^ © © r* «# CO © © © CD © COi-H^odt^ r-< t-^ ** CM tH *H © © ©* © CM© © i-< © lO CM SCO 00 iHt-4lMH r-l t^CO©C^00 i-^t- "^ CM i-l r-l i-Ti>rTjr• © *£ CO © CM ' 8 88888 _©© CO © © i-H ^H C5 10N O © CM © t>- ©T£©©CM rHiHH C^^H^^O ©©©©© ©©© ©©©©© ©©©©© ©©©©© ©©©©© ©©©©© "is co o< co © © © t^ cm oo©^©oo i-it> : eoioi>- © i>- © « © © © cm cm © © © ^ © © CM © 00 © r-< O © 00 © © CM © ©CD ©©^O^i-^rJ^ r-©^©©^ © CM © 00 © cb^i-r^oTvo' CM~tH CnTuO~©~ ©"oTcToO© ©•^CO^oTcm' i-Ti-TrH" ©CO©00© © "V< © CM CM i-H i— IH 8©© ©©©©© ©©©©^ Q©^t~© 00^J©CMO © © t«ffi9oS C) h O 00 CO CO t~ © © © ^« CN 00 CM rH Dr-llO© CO ©t> i-l CM ©©©©CM lOHHCNW © © CM © CO £2 5 • M Ttl r-l © 00 t^ CM ©©CiCDCD JrfiCTlt^ ©O©©00 CM©00C5Tf 3^--© OHi«t-o laowH* , .Ji- 1 © © 00t>-©uO© ^HCM00 ©CM©©© c^^^ OAVV ^8 OH CM©^> - s 3 > fe fl U "3 3 « «4 EH u Tj * i- S -3 & s 'j CO id BQ - fl - M 0) a .4 e 4a a M N » N CD ^ *a ft « O fa M PS a a C a ? t s •; 2* pss HO - s A 2 fro fro" fro fed fro fro S3- *Ta 02 o '9 Ai T © OS rH T*i OS lO © lOCNOS t- rtHOO 3 os t>; os TijoqosOoq lC^TjJt^i-; ^ n 6 od to lOT* COCNCN rlrtH ^*j co co io ih Tfioioo^ CO CO CO CN © HlOt-ifl ^JCC^t>;i-J t^coooo id "«* CO CN CN r-!^H»-HO r>oo io i-i : © rs os © n? os io cn os t- lO ri OS IQ r* lOOS^JO ih X CO O ^J OS lO CN OS ©* <<* co co' cn hhho lOWOOOf OS t* t~ t~tf5 COCOOCOt* «00«0-mHOO OOOO CN OS ri 00 *H COHOH CN lO/O-^CN »OiO^ lQZQCC© co co co 3; t- ■^ OS t>- CO *T — lO COCN rH ooooo OOOO OOOOO OOOO ©©CO t~ 00"*i t- co os iG\a co co CO t* CO OS 00 r-t~ OOOO CM O CN rt< CN O00 ■ *■■ oo i-t t -" OlOt'OOO rHb-^cNi-H HOC 2M%£2 38888 881. . ©OOOO ©©©©'© ©*©'©'©© ©©*©'©* 888 ©©'©©"© CO rf< © rf t» CN © ^t< b-iOOSI>" © CN CO t> © CO lO O OS © © CO CN t^ © lOMb-t^ S"©t^rHOS CO tfj OS 00 CN OSOSi-HTjJoi* lOfNCfiV .OHiOO) lOCNOSt*-© ^COCOCNi-l r-ti-t *** ^ CO CN r» ( Hrt CNgsre©^ rnosco©os co©osr^o osiocnc ^COOCNIO ©iOCN©l>. ©1T5COCOCN Hrtrt © Y"i ^ CO CN NrtHH Ot-©©"«*l rftCOCOCN OOOcSS S C§ CO ^< £&^£8 S28£2£S ^PSS* 00 ^c£SS !3£SSt:^2 ^jcnt-iOqo t^ts'S'Q ^cococo §©!SoS ©©©©8 88888 1888 ©*©©©'© ©©*©©"© ©©©*©©* ©*©©*© ;^^5 148 PROPERTIES OF CONDUCTORS. - fe; So " 00 hi RSI ™ eg & lis ^ T3 * © S fl B = o B g *« * - s s * ^ .2 a - 5-3 S! I o feO feO feO s o fed fed fed 5 t- OS t> ooooo t> CO eN •O OJ OS W N COOOiHlOO hckcoosoo Tfiost-osc ?H eg « e ISlOMw 8iO^~CO «D tH CO t^oo t-co csooo OHrtMf OOOOO ooooo ooooo o © o' 6 d CO locoes CO t~ co eo «<5 t-C<> OS *H cs iiili ddddd iO «3o> ooooo ooo 83S ooooo ooooo ooo COOOi-f OS l^ 00 tH t>. OOiHt-OOCO «5coooc^^ 1»t5N lOOS^rHCS ddddd odd OS lO lO 99 25 os t- c5 cs ooo odd C3COrt00Ci COPPER WIRE TABLES. 149 IflN CO CO• © 4 »$ 55 © 00 i-h rfi 00 C} 00 © lO N (N rn lO »C CO © i-H CD 30 52 S 222^^ t-< cj co co ■<* io ■> ©. T-1 "*< oo co © © © © © ©©©©<© © © © © © ©©©,-< »H r-l CO CO CO "f © © © © © © © © © © <6<6<£d>Or-li-l i-hCOCOCOt* ©ddd© ©ddd© CO lOCO^? iO© 00© odd-' COCO©CO iO cooot^cot> 3 C-l © © © © © o fc t- t- t- 9 co JNt' t> i-i co co »0© ©C0©©C0 ©OOCOCOCO CO lO 00Tt© Tj5©'t>lo6i^ © £o ^ £( ;s ^co co© ; fH 00 lO-^ t>T* tH coco o fro OO© t-O ©©rcoi-roo o © l>©_r-^©_ © t^^co r-T © 00 © l> CO cn co © ©© •*^iO©©^eo e5"©«oeocN~ ©©©©© CO «H 00 t^^Tji^©^ ^©io^co © CN COCN r* ©©CO 00© ^tj< lO©00 COHO © co t^eo^ C0©iOiO© iO © r- uo^ >t>C0T-jr^ -t-iCO«Ot^ HHlfllOt* ©©©©*© gc€28^S CO^Ot^O^CO co"cfT*Too*"c^ © iOrt< CO CO CN CN CN CN °J ©'©©*©© COPPER WIRE TABLES. 151 2S££& 5 t- CO""f COCOC oiiOCOr-l i> coos cooic* «© ^ CO rH 1-3 i-l if rf CM Id coco^oooj^ CM cfi-TT-Ti-T CMiOi-jC -O lOOOli } CO OS *f Cl 5 C~lO"3< coco tfsos io ?E8S§©: t-1 if 11 Tfi CO OiO-f COCM t~lQ*«f CM i-i CM CM t-i CO 05 COCMOfr-Tr-T if i-t © 00 OO CM CO t-i-H 05 CO CM i-H rH CM CO lO CO Oi co^co -t> 10 cmoo co i-nf CO i-i t> if l& 2832 S8S CM CO CO l> CO CO l> t> 05 CM 05 CO CO co CM i-lr-ICMCO if cqcqo5 CM CM 00 CM CM t*HO CO 00© i-Tr-Tr-Jof CO 00 CI CM O Z t>- CO if i-i a © r- t>CM CI 5 2c3 coco 05 O CO r-i 05 c^iqcococo i*< t>- i-i !>• Tfi CO lO CO CM CM CO CM CO OS tA CO ^i t-coco C-CD CO U OirH © OO if r-<00O5 i-H CM CO t>- !>• io cm iQ t>Tf if 05 O CO CM r-iOOOO ododo i-|if COi-H 00 t-"f O CO J* o tg 00 CO ododo ■o SB JS 55 5 O OO OO "-©5CM CO lO CM n n CM i-l h i-H CM CO t~ 5 CO"* O CO CO t- -f CD O i-H 05 t- O5 1OCOC0CM cMr-idd© o t- t- ooooo jggSS t>. t>.co 88S88 COCO 05 Tf CO co th d d d O5 00O5 IXN'* t- CO 05 'JT' 00 CO O t>- i-i co 3 S hhooo OiOOO CO t~ i-H CM if rt»0»t- © © © © © o t co co lo cocm c©if co CO CM 05 05 C5 CO CO CO t~ ooooo dodo© ooc-too ooooo ooooo COCO CM t~CO<£3! »ococoiOoo •** a if t~ t** ooooo ^cocmSco i-l 00 iH CM CO CO O ^ O CO cft^iortTco CO 05 i-H CM \Q \d d d Tti tj5 HAUHIQ CO i*i CO CO CM rH 05 !>; »H iq i-H coc4 coco O iO»COi-Hl>. CO© CO 05 CM COiOCOtHth * OS ©i CO O^CO r-iC^ i-ToToioiji'' O O © CM CM CO^^t^C^O^ coofi-Tr-Ti-r oqqqq if io if d i*i 00 IQ C5 CO^ © qqqq i-H Tff 05 1OCO 00 CO if CN rl 50000 5 co -f co oi ooooo ooooo 088S8 00 05 O* ITS CM uo a it co co ooooo 00 S O0O5©T-i SSgc CNCNCNCOCO CO CO CO CO" CO 152 PROPERTIES OF CONDUCTORS. H S E M e . &£ s'l 3 § w £ " CO fa fl Z © V CO a? 82 ^^ gl W co" 2 £ * "S w i S e 1 « 1 - a H » a 2 Si M s h s o fed b» 00 fed fe*o ^3 o feo feO hOhhiO Tf-t<co cocot- Ol>«OJlft 00 T*l 00 rH CNlflOOHlfl 1-1 00 t- 00000 00000 000 00000 CO rl CN oi>co 00060 066 OS 1-HC0O52 •*CN feO OSt- COO TH 00 tH 00 00 §& ooo©o §§§§! III ooooo ooooo 000 gHH«T}» lOt-O^O CN«OtH odd do Sdidxddi odd ^ 58°" CNe0<*»O COPPER WIRE TABLES. 153 ©6060 iHCftt- CO CO CO ^> lO W^5 © o 10 w.n*o»o © *r co o CO © © © CO © © lO k~ CI 888SS 3S288 8 66060 ddooo 6 © © o © co co ©©i-» OOOHMH t^ CO CO iflO rH N lO H — -O5R1H OOOOO ^h t~©00© CO l£ CO CO © CO © 1-1 IQ fH 1-1 i— CO CO CO OOOOO 66666 oco t>© S38BS 66 6 6 6 00 00 COCO i-« t> 1-H T- 1*1 t- CO© l>- CO© CO CO CO t- CO COO ©Ot-^cotj* ©CO COCO lO i-Tt-T©'© lOCO t^coco CO'* O ©lOCO i-( HOO l>© © CO T*< «h CO ^ CO tj< i-i OO ©OO 1-1CO <6d> <6d>dd> CO©C0 Tf Tf co © co uo © 6 6 r-i co' co' 3© CO t c S © CO COG 100 <^" co 2? 2 8 o© l ^ , -J. c l co© t> CO © © co © CO CO©© CO © © © t>. © CO»£©C0 CO 6 6 rA co* co' ©lOQCOCO 10 t>© co 10 lO CO^i-* CO © 3< CO HOOOO CO rH CO © ©lOCO IO OO 1-1 © rf CO COCO |>© IO CC* T*< CO CO CO -* IO© CO CO -* UJ © t^ CO © © i~l CO CO ■<*! lO © COCOCOCOCO CO CO CO CO CO COCOCOCOCO 154 PROPERTIES OF CONDUCTORS. The following condensed copper wire tables for both solid and stranded conductors are more convenient for ordinary calculations. Solid Copper Wire-100% JtEatthiessen'rt Standard. Diam. Mils. Weight. Pounds. Resistance, 20° C. 68° F. No. B.&S. Area. Cir. Mils. Feet Bare. 1000'. Mile. per Pound. 1000'. Mile. oooo 000 00 460 409.6 364.8 324.9 211,600 167,800 133,100 105,500 640.5 508 402.8 319.5 3,381 2,682 2,127 1,687 1.561 1.969 2.482 3.130 .04893 .06170 .07780 .09811 .2583 .3258 .4108 .5180 1 289.3 83,690 253.3 1,337 3.947 .12370 .6531 2 3 4 257.6 229.4 204.3 66,370 52,630 41,740 200.9 159.3 126.4 1,062 841.1 667.4 4.977 6.276 7.914 .1560 .1967 .2480 .8237 1.0386 1.3094 5 g 181.9 162.0 33,100 26,250 100.2 79.46 529.0 419.5 9.980 12.580 .3128 .3944 1.6516 2.0824 7 144.3 20,820 63.02 332.7 15.87 .4973 2.6257 8 128.5 16,510 49.98 263.9 20.01 .6271 3.3111 9 114.4 13,090 39.63 209.2 25.23 .7908 4.1754 10 101.9 10,380 31 .43 166.0 31.82 .9972 5 . 2652 11 90.74 8,234 24.93 131.6 40.12 1.257 6.6370 12 80.81 6,530 19.77 104.4 50.59 1.586 8.374 13 71.96 5,178 15.68 82.79 63.79 2.000 10.560 14 64.08 4,107 12.43 65.63 SO. 44 2.521 13.311 15 57.07 3,257 9.858 52.05 101.4 3.179 16.785 16 50.82 2,583 7.818 41.28 127.9 4.009 21 . 168 17 45.26 2.04S 6.200 32.74 161.3 5.055 26.690 18 19 40.30 35.89 1,624 1,28$ 4.917 3.899 25.96 20.59 203.4 256.5 6.374 8.038 33.655 42.440 20 31.96 1,022 5 3.092 16.33 323.4 10.14 53,540 T>. 7« fe Or . STRANDED COPPER WIRE. 155 Atranded Copper Wire — 100% JUattliiessen's Standard. No. B.&S. Diam. Mils. Area Cir. Mils, Weight Pounds. Resistance 20° C. 68° F. Bare. 1,000'. Per Mile. Feet per lb. 1,000'. Mile. 1,152 1,125 2,000,000 1,500,000 1,250,000 1,000,000 950,000 6,100 4,575 3,813 3,050 2,898 32,208 24,156 20,132 16,104 15,299 .164 .219 .262 .328 .345 .005177 .006902 .008282 .010353 .010900 .02733 .03644 .04373 .05466 .05755 1,092 1,062 1,035 999 900,000 850,000 800,000 750,000 2,745 2,593 2,440 2,288 14,494 13,688 12,883 12,078 .364 .385 .409 .437 .01150 .01218 .01294 .01380 .06072 .06431 .06832 .07286 963 927 891 855 700,000 650,000 600,000 550,000 2,135 1,983 1,830 1,678 11,273 10,468 9,662 8,857 .468 .504 .546 .596 .01479 .01593 .01725 .01882 .07809 .08411 .09108 .09937 819 770 728 679 500,000 450,000 400,000 350,000 1,525 1,373 1,220 1,068 8,052 7,247 6,442 5,636 .655 .728 .819 .936 .02070 .02300 .02588 .02958 . 10930 .12144 . 13664 .15618 oooo 000 630 590 530 470 300,000 250,000 211,600 167,800 915 762 645 513 4,831 4,026 3,405 2,709 1.093 1.312 1.550 1.949 .03451 .04141 .04893 .06170 .18221 .21864 .2583 .3258 00 1 2 420 375 330 291 133,100 105,500 83,690 66,370 406 322 255 203 2,144 1,700 1,347 1,072 2.463 3.106 3.941 4.926 .07780 .09811 .12370 . 15600 .4108 .5180 .6531 .8237 3 4 261 231 52,630 41,740 160 127 845 671 6.250 7.874 .19670 .2480 1.0386 1.3094 This table is calculated for untwisted strands; if the strand is twisted the cross section of the copper at right angles to the length of the strand, the weight per unit length and the resistance per unit length will each increase from 1 to 3 per cent, and the length per unit weight will decrease from 1 to 3 per cent, depending on the number of twists per unit length and the number of wires in the strand. 156 PROPERTIES OF CONDUCTORS. Tensile Streng-th of Copper Wire. ROEBLING. Numbers, B.&S. Gauge. Breaking Weight, Lbs. Numbers, B.&S. Gauge. Breaking Weight, Lbs. Hard- drawn. Annealed. Hard- drawn. Annealed. 0000 000 00 1 2 3 4 5 6 7 8 8,310 6,580 5,226 4,558 3,746 3,127 2,480 1,967 1,559 1,237 980 778 5,650 4,480 3,553 2,818 2,234 1,772 1,405 1,114 883 700 555 440 9 10 11 12 13 14 15 16 17 18 19 20 617 489 388 307 244 193 153 133 97 77 61 48 349 277 219 174 138 109 87 69 55 43 34 27 The strength of soft copper wire varies from 32,000 to 36,000 pounds per square inch, and of hard copper wire from 45,000 to 68,000 pounds per square inch, according to the degree of hardness. The above table is calculated for 34,000 pounds for soft wire and 60,000 pounds for hard wire, except for some of the larger sizes, where the breaking weight per square inch is taken at 50,000 pounds for 0000, 000, and 00, 55,000 for 0, and 57,000 pounds for 1. Hard-Drawn Copper Telegraph Wire. ROEBLING. Size Resistance Breaking Weight Furnished in Coils as follows, Miles. Approx. Size B.&S. in Ohms Strength, per E.B.B. Iron Wire Gauge. per Mile. Pounds. Mile. Equal to Copper. 9 4.30 625 209 1 2 . JO 5.40 525 166 1.2 3 11 6.90 420 131 .52 4 12 8.70 330 104 .65 6 Iron-Wire 13 10.90 270 83 1.20 6* Gauge. 14 13.70 213 66 1.5U 8 15 17.40 170 52 2.00 9 16 22.10 130 41 1.20 10 In handling this wire the greatest care should be observed to avoid kinks, bends, scratches, or cuts. Joints should be made only with Mclntire Con- nectors. On account of its conductivity being about five times that of Ex. B. B. Iron Wire, and its breaking strength over three times its weight per mile, copper may be used of which the section is smaller and the weight less than an equivalent iron wire, allowing a greater number of wires to be strung on the poles. Besides this advantage, the reduction of section materially decreases the electrostatic capacity, while its non-magnetic character lessens the seV-induc- tion of the line, both of which features tend to increase the possible speed of signalling in telegraphing, and to give greater clearness of enunciation over telephone lines, especially those of great length. WEIGHT OF COPPER WIRES. 157 Weig-lit of Copper Wire. English System, per 1,000 Feet and per Mile, in Pounds. English Legal Standard. Birmingham. Brown & Sharpe. Weight. Weight. Eh Si Weight. 1 1000 Feet. Mile. 1000 Feet. Mile. 1000 Feet. Mile. 6-0 464 432 400 652 565 484 3,441 2,983 2,557 5-0 4-0 454 624" 3,294" 460 " 64i' ' 3,382"" 3-0 372 419 2,212 425 547 2,887 410 509 2,687 2-0 348 367 1,935 380 437 2,308 365 403 2,129 324 318 1,678 340 350 1,847 325 320 1,688 1 300 272 1,438 300 272 1,438 289 253 1,335 2 276 . 231 1,217 284 244 1,289 258 202 1,064 3 252 192 1,015 259 203 1,072 229 159 838 4 233 163 860 238 171 905 204 126 665 5 212 136 718 220 146 773 182 100 529 6 192 112 589 203 125 659 162 79 419 7 176 94 495 180 98 518 144 63 331 8 160 77 409 165 82 435 128 50 262 9 144 63 331 148 66 350 114 39 208 10 128 50 262 134 54 287 102 32 166 LI 116 41 215 120 44 230 91 25 132 12 104 33 173 109 36 190 81 20 105 13 92 25.6 135 95 27.3 144 72 14 80 19.4 102 83 20.8 110 64 12.4 65 15 72 15.7 83 72 15.7 83 57 9.8 52 16 64 12.4 65 65 12.8 68 51 7.9 42 17 56 9.5 50 58 10.2 54 45 6.1 32 18 48 7.0 36.8 49 7.3 38.4 40 4.8 25.6 19 40 4.8 25.6 42 5.3 28.2 36 3.9 20.7 20 36 3.9 20.7 35 3.7 19.6 32 3.1 16.4 21 32 3.1 16.4 32 3.1 16.4 28.5 2.5 13.0 22 28 2.4 12.5 28 2.4 12.5 25.3 1.9 10.2 23 24 1.7 9.2 25 1.9 10.0 22.6 1.5 8.2 24 22 1.5 7.7 22 1.5 7.7 20.1 1.2 6.5 25 20 1.2 6.4 20 1.2 6.4 17.9 .97 5.1 26 18 .98 5.2 18 .98 5.2 15.9 .77 4.0 27 16.4 .81 4.3 16 .77 4.1 14.2 .61 3.2 28 14.8 .66 3.5 14 .59 3.1 12.6 .48 2.5 29 13.6 .56 3.0 13 .51 2.7 11.3 .39 2.0 30 12.4 .47 2.5 12 .44 2.3 10.0 .30 1.6 31 11.6 .41 2.15 10 .30 1.6 8.9 .24 1.27 32 10.8 .35 1.86 9 .25 1.3 8.0 .19 1.02 33 10.0 .30 1.60 8 .19 1.02 7.1 .15 .81 34 9.2 .26 1.35 7 .15 .78 6.3 .12 .63 35 8.4 .21 1.13 5 .075 .40 5.6 .095 .50 36 7.6 .17 .92 4 .048 .256 5.0 .076 .40 The diameters given for the various sizes are those to which the wire is actually drawn. 158 PROPERTIES OF CONDUCTORS. Weight of Copper Wire. Metric System — Per Kilometer, in Kilograms. Number of Wire Gauge. Roebling. Brown & Sharpe. Birmingham or Stubs. English Legal Standard- 6-0 954.3 970.9 5-0 833.9 841.6 4-0 696.5 954! 3 929 !i 721.5 3-0 591.0 756.8 814.5 624.0 2-0 494.1 600.2 651.3 546.2 425.1 480.4 521.3 473.4 1 361.2 377.4 405.8 405.8 2 311.9 299.3 363.3 343.5 3 268.5 237.4 302.6 286.3 4 228.3 188.3 255.3 242.7 5 193.2 149.3 218.3 202.7 6 166.2 118.4 185.9 166.2 7 141.3 93.9 146.1 139.7 8 118.3 74.5 122.8 115.4 9 98.8 59.0 98.8 93.5 10 82.2 46.8 81.0 73.9 11 64.9 37.1 64.9 60.7 12 49.9 29.5 53.6 48.8 13 38.2 23.4 39.8 38.2 14 28.9 18.5 31.1 28.9 15 23.4 14.7 23.4 19.1 23.4 16 17.9 11.7 18.5 17 13.2 9.23 15.2 14.1 18 9.96 7.32 10.8 10.4 19 7.58 5.80 7.95 7.22 20 5.52 4.61 5.52 5.85 21 4.61 3.65 4.62 4.61 22 3.54 2.89 3.54 3.54 23 2.81 2.16 2.81 2.59 24 2.38 1.82 2.19 2.19 25 1.80 1.44 1.80 1.80 26 1.46 1.15 1.46 1.46 27 1.30 .908 1.16 1.21 28 1.15 .720 .884 .988 29 1.02 .572 .762 .833 30 .884 .452 .649 .694 31 .822 .359 .451 .607 32 .762 .284 .365 .525 33 .544 .226 .289 .451 34 .451 .179 .220 .381 35 .406 .141 .113 .319 36 .365 .113 .071 .260 STANDARD COPPER STRANDS. 159 Standard Copper Strands. ROEBLING. CM. Wires. Outside Diam. Weight lbs. per 1000 ft. No. Size. 2,000,000 1,950,000 1,900,000 127 127 127 .1255 .1239 .1223 1.632 1.611 1.590 6100 5948 5795 1,850,000 1,800,000 1,750,000 127 127 127 .1207 .1191 .1174 1.569 1.548 1.526 5643 5490 5338 1,700,000 1,650,000 1,600,000 91 91 91 .1367 .1347 .1326 1.504 1.482 1.459 5185 5033 4880 1,550,000 1,500,000 1,450,000 91 91 91 .1305 .1284 .1262 1.436 1.412 1.388 4728 4575 4423 1,400,000 1,350,000 1,300,000 91 91 91 .1240 .1218 .1195 1.364 1.340 1.315 4270 4118 3965 1,250,000 1,200,000 1,150,000 91 61 61 .1172 .1403 .1373 1.289 1.263 1.236 3813 3660 3508 1,100,000 1,050,000 1,000,000 61 61 61 .1343 .1312 .1280 1.209 1.181 1.152 3355 3203 3050 950,000 900,000 850,000 61 61 61 .1247 .1214 .1180 1.122 1.093 1.062 2898 2745 2593 800,000 750,000 700,000 61 61 61 .1145 .1108 .1071 • 1.031 .997 .964 2440 2288 2135 650,000 600,000 550,000 61 61 61 .1032 .0991 .0949 .929 .892 .854 1983 1830 1678 500,000 450,000 400,000 61 37 37 .0905 .1103 .1039 .815 .772 .727 1525 1373 1220 350,000 300,000 250,000 37 37 37 .0972 .0900 .0821 .680 .630 .575 1068 915 763 160 PROPERTIES OF CONDUCTORS. Standard Copper Strands. — (Continued) . ROEBLING. Size. Wires. Outside Diameter. Weight. Lbs. per 1000 ft. B. &. S. No. Size. 0000 000 00 19 19 19 .1055 .0941 .0837 .528 .471 .419 645 513 406 1 2 19 19 7 .0746 .0663 .0975 .373 .332 .293 322 255 203 3 4 5 7 7 7 .0866 .0771 .0688 .260 .231 .206 160 127 101 6 8 10 7 7 7 .0612 .0484 .0386 .184 .145 .116 80 50 32 12 14 16 7 7 7 .0306 .0242 .0193 .092 .073 .058 20 12 8 18 7 .0151 .045 5 0§€LATED COPPER WIRES AIJTD CABLE!. Weather-proof JLine and House Wire. Solid Conductor. Standard Underground Cable Co. Double Covered. Tripl e Covered. B. AS. Gauge. Lbs. per Lbs. per Diam. in Lbs. per Lbs. per Diam. in Mile. 1000 ft. Mils. Mile. 1000 ft. Mils. 0000 3690 699 725 3910 741 780 000 2970 562 655 3160 598 700 00 2390 452 585 2560 485 635 1860 352 545 2020 382 590 1 1500 284 505 1650 312 550 2 1225 232 470 1340 254 515 3 980 186 385 1050 199 450 4 800 151 360 860 163 430 5 640 121 335 700 132 400 6 520 98 300 575 109 360 7 420 79 270 465 88 335 8 345 65 245 390 74 265 9 275 52 225 320 60 255 10 235 45 195 265 50 220 11 190 36 180 225 42 205 12 145 27 165 180 34 185 14 105 20 140 130 24 160 16 80 15 130 100 19 150 18 55 10 125 80 15 145 20 42 8 122 68 12 135 PROPERTIES OP CONDUCTORS. ion A The following tables of weights of weatherproof wire are in general use by the manufacturers and are guaranteed to be correct within 3%. WEATHER-PROOF \\ I RE. Approximate Weights. — Solid. Double Braid. Triple Braid. Size. B. & S. Gauge. Lbs. per Lbs. p3r Lbs. per Lbs. per 1000 ft. Mile. 1000 ft. Mile. 0000 723 3,817 767 4,050 000 587 3,098 629 3,320 00 467 2,467 502 2,650 377 1,989 407 2,150 1 294 1,553 316 1,670 2 239 1,264 260 1,370 3 185 977 199 1,050 4 151 795 164 865 5 122 646 135 710 6 100 529 112 590 8 66 349 75 395 9 54 283 62 325 10 46 241 53 280 12 30 158 35 185 14 20 107 25 130 16 16 83 20 105 18 12 64 16 85 20 9 48 12 65 Approximate Weights. - — Stranded. Capacity. Circular Mills. 2,000,000 6,690 35,323 7,008 37,000 1,750,000 5,894 31,119 6,193 32,700 1,500,000 5,098 26,915 5,380 28,400 1,250,000 4,264 22,516 4,508 23,800 1,000,000 3,456 18,246 3,674 19,400 900,000 3,127 16,513 3,332 17,600 800,000 2,799 14,779 2,992 15,800 750,000 2,635 13,913 2,822 14,900 700,000 2,471 13,045 2,650 14,000 600,000 2,093 11,052 2,235 11,800 500,000 1,765 9,318 1,894 10,000 450,000 1,601 8,452 1,724 9,100 400,000 1,436 7,584 1,553 8,200 350,000 1,248 6,589 1,345 7,100 300,000 1,083 5,721 1,174 6,200 250,000 907 4,788 985 5,200 Size-B. & S. Gauge. 0000 745 3,935 800 4,220 000 604 3,190 653 3,450 00 482 2,544 522 2,760 388 2,051 424 2,240 1 303 1,599 328 1,735 2 246 1,301 270 1,425 3 190 1,004 206 1,090 4 155 820 170 900 5 126 668 140 740 6 103 544 115 610 8 68 359 78 410 160b WEATHER-PROOF WIRE. SLOW BURNING WEATHER-PROOF WIRE. Triple Braid — Black Outside. Capacity. Circular Mills. Stranded. Solid. Capacity. Circular Mills. Stranded. Solid. Lbs. per 1000 ft. Lbs. per Mile. Lbs. per 1000 ft. Lbs. per Mile. Lbs. per 1000 ft. Lbs. per Mile. Lbs. per 1000 ft. Lbs. per Mile. 1000000 3860 3520 3180 2820 2350 1990 1820 1650 1440 1270 1060 20400 18600 16800 14900 12400 10500 9600 8700 7600 6700 5600 Size— B.& S. Gauge. 0000 000 00 1 2 3 4 5 6 8 10 12 14 16 18 900 735 583 480 355 290 240 195 160 132 87 4750 3880 3080 2530 1870 1540 1270 1030 845 695 460 862 710 562 462 340 280 230 190 155 127 85 60 42 30 24 19 4550 900000 3750 800000 2970 700000 2440 600000 1800 500000 1480 450000 1220 400000 1000 350000 820 300000 670 250000 450 315 220 160 130 100 WEATHER-PROOF IRON WIRE. Approximate Weights Per Mile. Iron Wire Gauge. Double Braid. Triple Braid. Iron Wire Gauge. Double Braid. Triple Braid. No. 4 6 8 9 860 665 470 400 940 740 525 450 No. 10 12 14 350 225 145 400 260 175 slow mmno wire. Approximate Weights — Triple Braid. Capacity. Circular Mills. Stranded. Solid. Capacity. Circular Mills. Stranded. Solid. Lbs. per 1000 ft. Lbs. p-r Mile. Lbs. per 1000 ft. Lbs. per Mile. Lbs. per 1000 ft. Lbs. per Mile. Lbs. per 1000 ft. Lbs. per Mile. 1000000 3980 3640 3280 2920 2460 2080 1900 1700 1500 1310 1120 21000 19200 17300 15400 13000 11000 10000 9000 7900 6900 590C Size— B. & S. Gauge. 0000 000 00 1 2 3 4 5 6 8 10 12 14 16 18 960 785 625 510 380 335 280 230 195 165 105 5070 4150 3300 2700 2000 1770 1480 1220 1030 870 555 925 760 600 495 365 320 270 220 190 160 100 80 55 40 30 24 4890 900000 4020 800000 3170 700000 2610 600000 1930 500000 1690 450000 1425 400000 1160 350000 1000 300000 845 250000 530 420 290 210 160 130 RUBBER COVERED WIRES AND CABLES. 161 Underwriters' Test of Rubber Covered Wire. Adopted Dec. 6, 1904. The Electrieal Committee of the Underwriters National Association recommended the following, which was adopted. Each foot of the completed covering must show a dielectric strength sufficient to resist throughout five minutes the application of an electro- motive force proportionate to the thickness of insulation in accordance with the following table: Thickness Breakdown Test in 64ths inches. on 1 Foot. 1 3,000 Volts A. C. 2 6,000 " 3 9,000 '* 4 11,000 " 5 13,000 " 6 15,000 " 7 16,500 " 8 18,000 " 10 21,000 " 12 23,500 " 14 26,000 " 16 '. . . . .28,000 M Standard Rubber Covered Wire* and Cables. (Made by General Electric Company.) Rubber covered wires and cables are insulated with two or more coats of rubber, the inner coat in all cases being free from sulphur or other sub- stance liable to corrode the copper, the best grade of fine Para being em- ployed. All conductors are heavily and evenly tinned. Five distinct finishes can be furnished as follows: — White or black braid, plain lead jacket, lead jacket protected by a double wrap of asphalted jute, lead jacket armored with a special steel tape, white armored, for submarine use. For use in conduits the plain lead covering is recommended, or if corro- sion is especially to be feared, the lead and asphalt. For use where no con- duit is available, the band steel armored cable is best, as it combines moderate flexibility with great mechanical strength, enabling it to resist treatment which would destroy an unarmored cable. In addition to the ordinary galvanometer tests, wires and cables are tested with an alternating current (as specified in table) before shipping. Special rubber covered wire and cable with lead jackets will be covered with the following thicknesses of lead unless otherwise specified: Outside diameter of cable (inside diameter of lead pipe). Up to and including .500" - • eV .501" to .750", inclusive . . -is" .751" to 1.250", inclusive &" 1.251" to 1.5", inclusive &" Larger than 1.501" £" 162 PROPERTIES OF CONDUCTORS. Standard Conductor. National Electric Code, General Electric Company. I. Solid. Size. 3 a 03-73 ft Weight per 1000 ft. in lbs. o o '5 03 DQ $ o 03 - o a; Jl 03 1 BQ Test Pres- IJS per 1000 ft. o v_ 03 DO'S sure for 30 Size. in ibs. S3 S .fl- ° a min. B.&S. and CM. CM m 1 "™ 1 +3 - ■S-d 0QPQ 0> fcC V ea a .290 = t *6 o . 16 .196 28 43 210 A A 1000 1500 14 .212 35 50 228 .306 A A 1000 1500 12 .231 46 63 253 .325 A A 1000 1500 10 .255 63 81 288 .349 A A 1000 1500 8 .285 86 107 335 .379 3 64 A 1000 1500 6 .374 139 162 410 .433 A A 2000 2500 5 .396 165 189 455 .455 3 64 A 2000 2500 4 .422 197 221 507 .481 A A 2000 2500 3 .450 240 265 567 .509 A A 2000 2500 2 .512 289 316 639 .541 A A 2000 2500 1 .587 381 410 935 .647 1 T6~ A 2500 3500 100000 .616 447 476 1030 .676 A A 2500 3500 .626 464 493 1055 .686 1 T6~ A 2500 3500 125000 .656 513 544 1128 .716 A _5_ 64 2500 3500 00 .669 563 595 1202 .730 1 A 2500 3500 150000 .690 617 650 1275 .750 1 T6~ A 2500 3500 000 .721 683 716 1372 .781 A A 2500 3500 200000 .763 800 834 1532 .823 i A 2500 3500 0000 .779 835 869 1583 .839 A _5_ 64 2500 3500 250000 .873 1032 1095 2047 .948 A A 4000 5000 300000 .932 1218 1283 2303 1.008 A A 4000 5000 350000 .976 1381 1449 2527 1.056 A A 4000 5000 400000 1.027 1548 1617 2753 1.102 A A 4000 5000 500000 1.113 1888 1958 3202 1.189 A A 4000 5000 600000 1.222 2275 2354 3725 1.298 A A 5000 6000 700000 1.294 2619 2707 4148 1.370 A 7 64" 5000 6000 750000 1.328 2791 2880 4355 1.404 A A 5000 6000 800000 1.360 2959 3051 4912 1.436 A A 5000 6000 900000 1.423 3295 3390 5340 1.531 A A 5000 6000 1000000 1.482 3624 3721 5752 1.590 A A 5000 6000 1250000 1.650 4496 4600 7704 1.820 i 1 5000 6000 1500000 1.772 5319 5432 8754 1.942 J 1 t 5000 6000 2000000 1.992 6958 7075 10821 2.162 * i 1 5000 6000 Note. — Wire and cable No. 1 B. & S. and larger have tape over rub- ber in addition to braid. Add -fc" to single braid for diameter of double braid. RUBBER INSULATED WIRES AND CABLES. 163a Diameters and Weights oi Small Sizes of Cotton Covered Wire. Diameters Weighl > in Pounds per 1000 Feet. Size B. &S. s.c.c. D.C.C. S.* 3.C. D.S.C 5. S.C.C. D.C.C. s.s.c. D.S.C. 14 .0700 .0740 12.684 12.918 15 .0630 .0670 10.082 10.274 16 .0560 .0590 8.012 8.176 17 .0500 .0530 6.375 6.510 18 .0450 .0480 5.081 5.188 19 .0400 .0440 4.043 4.130 20 .0360 .0400 3.218 3.289 21 .0325 .0365 2.569 2.628 22 .0294 .0334 2.055 2.106 23 .0265 .0305 .( )260 .029( ) 1.630 1.676 1 573 1.604 24 .0241 .0280 .( )230 .026( ) 1.297 1.344 1 241 1.298 25 .0220 .0260 .( )210 .024( ) 1.036 1.082 991 1.040 26 .0200 .0240 .( )190 .022( ) .828 .873 791 .833 27 .0180 .0220 .( )170 .020( ) .661 .703 631 .666 28 .0166 .0206 .( )156 .018( 1 .524 .562 499 .521 29 .0153 .0193 .( )140 .017( ) .421 .457 397 .416 30 .0140 .0180 .( )125 .0151 ) .336 .372 315 .332 31 .0130 .0170 .( )114 .0131 ) .271 .307 254 .267 32 .0119 .0159 .( 1105 .013( ) .215 .248 203 .214 33 .0110 .0150 .( )095 .012( ) .174 .201 161 .172 34 .0103 .0143 .( )088 .0}K 1 .141 .161 130 .140 35 .0096 .0136 .( )076 .009( J .120 .137 110 .119 36 .0085 .0120 .( )070 .009( ) .099 .112 089 .096 38 ( )060 .008( ) 058 .065 40 ( )05C .( )07( ) 037 .040 164 PROPERTIES OF CONDUCTORS. General Electric Company Rubber Insulated Win* and Cable (fs" Rubber). Test Pressure. — Red Core, 2500 Volts; White Core, 3000 Volts, for 30 Minutes. Wire. Diameter, Single Braid, Inches. Weight per 1000 ft. in Lbs. 8a rH o5 S Si a? a a rt Xi O Hh3 Insulation Resistance in Megohms per Mile. Size. B. &.S. v-4 S u icffl Red Core. White Core. 16 14 12 10 8 .221 .234 .251 .272 .299 33 40 51 67 91 48 56 67 85 109 233 249 273 305 348 .315 .328 .345 .366 .393 1 64 A 350 350 350 350 350 600 600 600 600 600 Cable. 16 .227 39 56 242 .326 A 350 600 14 .243 43 61 260 .337 _3_ 64 350 600 12 .262 60 80 285 .356 A 350 600 10 .286 78 99 316 .380 A 350 600 8 .316 105 127 360 .395 & 350 600 Note. — Add t$" to single braid for diameter of double braid. RUBBER INSULATED WIRES AND CABLES. 165 General Electric Company Runner Insulated Wire and Cable (gV Runner). Test Pressure. — Red Core, 5000 Volts; White Core, 6000 Volts, for 30 Minutes. Wire. Size. B. & S. and Diameter, Single Braid, Inches. Weight per 1000 ft. in lbs. 8a Kg W)h3 tn -CI 14 .388 89 113 373 .447 A 600 1000 12 .407 103 127 401 .466 A 600 1000 10 .431 125 150 439 .490 A 600 1000 8 .461 156 182 491 .520 3 64 600 1000 6 .529 210 237 563 .558 A 550 900 5 .551 240 268 608 .580 A 550 900 4 .577 277 306 821 .636 A 550 900 3 .605 322 351 895 .665 1 550 900 2 .637 376 407 981 .697 a 550 900 1 .681 454 486 1104 .741 i T6~ 550 900 100000 .710 514 547 1192 .770 A 550 900 .720 530 564 1216 .780 A 500 800 125000 .750 582 616 1290 .810 A 500 800 00 .764 635 669 1364 .824 1 T6~ 500 800 150000 .784 689 723 1443 .844 A 500 800 000 .815 760 796 1545 .875 A 500 800 200000 .872 914 977 1929 .948 A 500 800 0000 .888 953 1018 1987 .964 A 500 800 250000 .955 1084 1149 2178 1.031 A 400 700 300000 .994 1278 1346 2444 1.070 A 400 700 350000 1.042 1445 1514 2672 1.118 A 400 700 400000 1.088 1617 1686 2901 1.164 A 400 700 500000 1.175 1958 2034 3350 1.251 A 350 600 600000 1.253 2308 2391 3790 1.329 A 350 600 700000 1.325 2657 2747 4222 1.401 A 350 600 750000 1.359 2831 2923 4781 1.466 A 300 500 800000 1.391 3031 3126 5012 1.498 A 300 500 900000 1.454 3343 3438 5432 1.561 A 300 500 1000000 1.513 3675 3773 5852 1.620 A 300 500 Note. — Add ^" to single braid for diameter of double braid. 168 PROPERTIES OF CONDUCTORS. General JElectric Company Rubber Insulated Wire and Cable ( 5 y Rubber). Test Pressure . — Red Core, 12,000 Volts ; White Core. 15,000 Volts, for 30 Minutes. L Solid. Size. fa 5« Weight per 1000 ft. in lbs. 11 1— 1 5 ° fctH u IT S*g °£ .§13 Insulation Resistance in Megohms per Mile. B. &S. ®T3 bird o S OQffl o * Red Core. White Core. 14 .534 156 184 512 .562 A 700 1200 12 .551 173 201 540 .580 3 64 700 1200 10 .572 196 224 735 .601 A 700 1200 8 .598 226 255 792 .658 A 700 1200 6 .632 272 303 872 .692 1 T6~ 700 1200 5 .652 302 333 924 .712 1 T6~ 600 1100 4 .674 340 372 982 .734 A 600 1100 3 .699 386 419 1053 .759 i 1 ? 600 1100 2 .728 441 474 1137 .788 A 600 1100 1 .759 509 543 1235 .819 A 600 1100 .795 592 638 1356 .855 A 550 1000 00 .850 696 732 1708 .926 A 550 1000 000 .895 851 926 1898 .971 5 64 550 1000 0000 .945 1011 1084 2109 1.021 5 64 550 1000 RUBBER INSULATED WIRES AND CABLES. 1C9 Cleneral Electric- Company Rubber Insulated Wire and Cable ( 5 y Rubber) — Continued. II. Stranded. Size. - X Weight per 1000 ft. in lbs. o • 1—1 II fc£H-} 'o - i ■a w ■ a c3 3 O 0> »-£ 02 O cj d M - ■s'S Insulation Resistance in Megohms per Mile. B. & S. and CM. 5-d Si 2 6 £ o 14 .543 162 190 524 .572 A 700 1200 12 .562 181 209 566 .591 3 64 700 1200 10 .586 205 233 758 .646 A 700 1200 8 .616 239 268 822 .676 1 700 1200 6 .654 290 320 912 .714 A 600 1100 5 .676 323 354 968 .736 1 T6~ 600 1100 4 .702 365 397 1034 .762 A 600 1100 3 .730 413 447 1112 .790 i IT 600 1100 2 .762 472 506 1201 .822 A 600 1100 1 .806 555 591 1332 .866 A 550 1000 100000 .850 619 656 1638 .926 5 64 550 1000 .860 637 675 1666 .936 5 64 550 1000 125000 .890 708 759 1750 .966 A 550 1000 00 .904 780 844 1834 .980 A 550 1000 150000 .924 838 903 1917 1.000 3 64 550 1000 000 .955 915 981 2032 1.031 A 550 1000 200000 .997 1042 1110 2212 1.073 3 64 500 900 0000 1.013 1083 1151 2271 1.089 A 500 900 250000 1.060 1225 1294 2473 1.136 A 500 900 300000 1.119 1424 1494 2745 1.195 A 500 900 350000 1.167 1600 1675 2980 1.243 A 450 800 400000 1.213 1781 1860 3218 1.289 A 450 800 500000 1.300 2138 2226 3679 1.376 A 450 800 600000 1.378 2497 2589 4474 1.485 A 400 700 700000 1.450 2854 2950 4938 1.557 A 400 700 750000 1.484 3030 3127 5161 1.591 A 350 600 800000 1.516 3205 3304 5384 1.623 A 350 600 900000 1.579 3557 3658 5829 1.687 A 350 600 1000000 1.638 3900 4004 7085 1.808 i 350 600 Note. — Add ^$" to single braid for diameter of double braid. For ^j" insulation the insulation resistance will be in proportion with &" and 5 y insulation. Test pressure for -^ ,f Red Core, 10,000 volts; White Core, 12,000 volts for 30 minutes. 170 PROPERTIES OF CONDUCTORS. Creneral Electric ( ompaiiv Three Conductor Cable, White Core Insulation, Test Pressure. — 3000 Volts for 30 Minutes. Leaded. Braided. Insula- tion Resist- ance in Megohms per Mile. Size. B. &S. and CM. ££2 Hi Pi t-" . CO «t r ^ 3 « .2 <8 5.g 3&S .SP .© 8 1192 A .805 A .740 449 600 6 1567 A .918 A .852 551 500 5 1728 tV .966 A .900 653 500 4 1889 1 T5 1.022 5 64 .956 756 500 3 2123 A 1.082 A 1.016 909 500 2 2358 A 1.152 A 1.077 1062 500 1 2847 A- 1.314 5 64 1.239 1352 500 100000 3032 A 1.376 A 1.301 1492 500 3217 A 1.398 5 64 1.327 1632 500 125000 3631 A 1.494 _5_ 64 1.386 1800 500 00 4045 A 1.524 A 1.427 1967 500 150000 4332 A 1.567 A 1.470 2175 500 000 4619 A 1.635 A 1.537 2381 500 200000 4968 A 1.724 JL 64 1.626 2638 500 0000 5318 A 1.759 _5_ 64 1.662 2895 500 Test Pressure. — 8000 Volts for 30 Minutes. Leaded. Braided. Insula- tion Resist- ance in Megohms per Mile. Size. B. &S. and CM. It <».S • -H 15 .a §13.3 JMJ 3.s 8 2452 1.376 A 1.310 913 1300 6 3077 1.490 A 1.392 1097 1200 5 3282 1.538 A 1.440 1224 1200 4 3488 1.594 A 1.496 1352 1200 3 3767 1.654 A 1.556 1536 1200 2 4046 1.723 A 1.626 1721 1200 1 4471 1.818 A 1.721 2020 1100 100000 5120 1.943 * 1.783 2160 1100 5769 1.965 * 1.800 2301 1100 125000 6055 2.030 * 1.865 2490 1100 00 6342 2.060 * 1.900 2679 1100 150000 6677 2.103 * 1.943 2907 1100 000 7013 2.170 * 2.010 3135 1100 200000 7418 2.261 ft 2.101 3421 1100 0000 7823 2.295 ft 2.135 3707 1100 Test Pressure. — 26,000 Volts for 30 Minutes. Leaded. Braided. Insulation Size. B. &S. and C. M. .sp a © t££2 H 02 n H.s §J rat— ( S.g bf7o '© 02 O Resistance in Megohms per Mile. 8 4103 1.878 A 1.781 1558 1600 6 4437 1.960 3 32 1.863 1770 1500 5 4661 2.008 _3_ 32 1.911 1919 1500 4 4885 2.064 & 1.967 2068 1500 3 5710 2.124 A 2.027 2281 1500 2 6535 2.256 1 8 2.096 2495 1500 1 6995 2.351 * 2.183 2792 1500 100000 7259 2.414 * 2.246 2968 1400 7523 2.436 4 2.271 3145 1400 125000 7828 2.500 1 t 2.335 3354 1400 00 8133 2.530 1 8 2.371 3563 1400 150000 8490 2.576 * 2.417 3813 1400 000 8848 2.641 * 2.481 4064 1400 200000 9292 2.731 * 2.571 4378 1300 0000 9736 2.766 1 2.606 4693 1300 172 PROPERTIES OF CONDUCTORS. General Electric- Company Extra Flexible Dynamo Cable. This is adapted for use as brush-holder leads, or to any use where great flexibility is required. The finish is black glazed linen braid. Each wire of the strand is No. 25 B. & S. Dimensions in Inches. Number Circular Mils. Wires in Strand. Diameter Thickness Diameter Bare. Rubber. Over All. 25 8,000 .108 .047 .275 50 16,000 .150 .047 .320 75 24,000 .205 .047 .375 100 32,000 .235 .047 .450 150 48,000 .285 .047 .500 200 64,000 .325 .047 .540 250 80,000 .350 .047 .600 300 96,000 .385 .065 .665 350 112,000 .425 .065 .705 400 128,000 .460 .065 .740 450 144,000 .485 .065 .765 500 160,000 .570 .065 .810 550 176,000 .530 .065 .830 600 192,000 .570 .065 .870 650 208,000 .605 .065 .935 700 224,000 .625 .065 .955 750 240,000 .640 .065 .970 800 256,000 .680 .065 1.010 900 288,000 .700 .065 1.030 1000 320,000 .725 .065 1.055 1250 400,000 .825 .065 1.165 1500 480,000 .880 .065 1.213 1750 560,000 .960 .093 1.360 2000 640,000 1.060 .093 1.410 2250 720,000 1.100 .093 1.500 2500 800,000 1.200 .093 1.600 2750 880,000 1.250 .093 1.650 3125 1,000,000 1.430 .093 1.830 SPECIAL CABLES. 173 Rubber Insulated Cable for Car Wiring-. Single Conductor, Weatherproof Finish. This class of cable is made with separator, standard code thickness of insu- lation tape and single-braid weatherproof finish. Standard Strands. Finished Weight Diameter in Inches. Size B. & S. Stranding. in Pounds per M Feet. 14 7/. 0243 37 .23 12 7/. 0306 48 .25 10 7/. 0386 64 .27 8 7/. 0485 90 .31 6 7/. 0613 139 .38 4 7/. 0773 197 .42 2 7/. 0974 289 .51 1 19/. 0664 381 .59 I/O 19/. 0746 464 .63 2/0 19/. 0838 563 .67 3/0 19/. 0940 683 .72 4/0 19/. 1056 835 .78 250,000 37/. 0823 1032 .87 For each additional braid, add approximately ^g inch to diameter. Single Conductor, Flameproof Finish. National Electric Code Standard. This class of cable is made with separator, standard code thickness of in- sulation and double braid finish — the first braid is cotton, well compounded, the second or finishing braid is filled asbestos. Standard Strands. Finished Weight Diameter in Inches. Size B. & S. Stranding. in Pounds per M Feet. 14 7/. 0243 65 .31 12 7/. 0306 78 .33 10 7/. 0386 99 .36 8 7/. 0485 128 .39 6 7/.0613 189 .47 4 7/. 0773 255 .52 2 7/. 0974 353 .58 1 19/. 0664 461 .68 I/O 19/. 0746 545 .72 2/0 19/. 0838 650 .77 3/0 19/. 0940 778 .82 4/0 19/. 1056 937 .88 250,000 37/. 0823 1138 .99 300,000 37/. 0906 1330 1.05 350,000 37/. 0974 1497 1.10 500,000 61/. 0906 2024 1.23 750,000 61/. 1110 2945 1.45 1,000,000 61/. 1281 3801 1.62 SPECIAL CABLES. 173a Rubber Insulated Cable for Type TO Control. For connecting contactors and controllers, 19/25 B. & S. single conductor s Vinch rubber insulation is used; double braid weatherproof finish. The nearest equivalent size is number 12 B. & S. The weight per 1000 feet is 52 pounds, and diameter .25 inches. Train Cables. Multiple conductors, each single conductor being composed of 19/25 B. & S. wires, rubber covered, single braid and a tape and braid finish overall. Number of Conductors^ Diameter Overall. Weight per 1000 Ft. 5 .7 255 6 .75 343 7 .75 373 9 .85 479 10 .93 503 12 1.03 613 20 1.28 893 Jumper Cables. Are similar in construction t* train cables with the exception that the group of conductors is surrounded by a rubber jacket and a double braid finish. Number of Conductors. Diameter Overall. Weight per 1000 Feet. 5 .88 371 6 .94 461 7 .94 491 9 1.00 632 10 1.07 687 12 1.30 846 20 1.54 1246 Both train and jumper cables have distinctive marking threads woven in the braid of each conductor. 174 PROPERTIES OF CONDUCTORS. WAVY STA\»AHJ) HIKES. In the following table are given sizes of Navy Standard Wires as per specifications issued by the Navy Department in March, 1897. d 3 4,107 9,016 11,368 14,336 18,081 22,799 30,856 38,912 49.077 60;088 75,776 99,064 124,928 157,563 198,677 250,527 296,387 373,737 413,639 8-T3 .J3 S3 *** ^QQ o . £.s N^ 1 m 1 14 7 19 7 18 7 17 7 16 7 15 19 18 19 17 19 16 37 18 37 17 61 18 61 17 61 16 61 15 61 14 91 15 91 14 127 15 Diameter Inches. Over copper. .06408 . 10767 . 12090 .13578 . 15225 .17121 .20150 .22630 . 25410 . 28210 .31682 .36270 .40734 .45738 .51363 .57672 . 62777 . 70488 .74191 Over Para rubber .0953 .1389 .1522 .1670 .1837 .2025 .2328 .2576 .2854 .3134 .3481 .3940 .4386 .4885 .5449 .6080 .6590 .7361 .7732 Diameter in 32ds of an inch. Over vulc. rubber . Over Over tape. braid. 7 9 11 10 12 14 10 12 14 10 12 14 11 13 15 12 14 16 12 14 16 13 15 17 14 16 18 15 17 19 16 18 20 18 20 22 19 21 23 20 22 24 22 24 26 24 26 28 26 28 30 29 31 33 30 32 34 at «1§ 56.9 103 108.5 115.5 140 1651 184 218 260i 314 371 463 557 647 794 970 1,138 1,420 1,553 Double Conductor, Plain, 2-7-22 B. & S. . . . Double Conductor, Silk, 2-7-25 B. & S Double Conductor, Diving Lamp, 2-7-20 B. & S. Bell Cord, 1-16 B. & S 181.5 28 218.3 29.7 PAPER nSILATED AXI> LEADED WIRES AND CABLES. GENERAL ELECTRIC CO. There will be found on the following pages data of a full line of paper insulated and lead covered wires and cables. All cables insulated with fibrous covering depend for their successful operation and maintenance upon the exclusion of moisture by the lead sheath; and this fact should be borne in mind constantly in handling this class of cables, consequently the lead on them is extra heavy. The use of jute and asphalt covering over the lead is strongly recommended on all this class of cables, inasmuch as their life is absolutely dependent upon that of the lead. Paper insulated cables cannot be furnished without the lead covering. PAPER INSULATED WIRES AND CABLES. 175 General Electric Company Paper Insulated and JLcad Covered Cable. I. Solid. $t" Insulation 3 y Insulation J, <» Test Pressure, 4000 Test Pressure, 6000 .2 S fe Volts for 30 Minutes. Volts for 30 Minutes. Size. B. & S. and CM. ft .9 t 08 08 C o>© £8 s A 6$£ II 5 A £ * » 10 413 •414 493 .477 300 8 461 .441 A 542 .503 1 300 6 530 .474 A 613 .537 A 300 5 574 .494 A 660 .557 A 300 4 626 .517 A 715 .579 A 300 II. Stranded. 6 558 .496 l 645 .559 A 250 5 605 .518 A 694 .581 A 250 4 662 .544 A 754 .607 l 250 2 814 .604 A 1,068 .698 A 250 1 1,072 .679 A 1,184 .742 A 250 100000 1,176 .708 A 1,289 .771 A 250 1,199 .718 A 1,315 .781 A 250 125000 1,276 .748 A 1,393 .811 A 200 00 1,354 .762 A 1,470 .825 A 200 150000 1,431 .782 A 1,547 .845 A 200 000 1,536 .813 A 1,655 .876 A 200 200000 1,703 .855 A 2,046 .949 A 150 0000 1,758 .871 A 2,106 .965 A 150 250000 2,165 .950 A 2,304 1.012 A 150 300000 2,435 1.009 A 2,574 1.071 A 150 350000 2,660 1.057 A 2,804 1.119 A 125 400000 2,890 1.103 A 3,041 1.165 A 125 500000 3,929 1.252 i 4,106 1.315 i 125 600000 4,409 1.330 i 4,598 1.393 1 8 125 700000 4,876 1.402 i 5,067 1.465 i 100 750000 5,106 1.436 i 5,298 1.499 8 100 800000 5,337 1.468 1 ¥ 5,523 1.531 i 100 900000 5,782 1.531 l 8 5,976 1.594 i 100 1000000 6,213 1.590 i 6,416 1.653 i 100 1250000 7,293 1.727 i 7,500 1.790 I 100 1500000 8,329 1.849 8 8,542 1.912 i 75 2000000 10,355 2.069 i 10,586 2.132 i 50 176 PROPERTIES OF CONDUCTORS. General Electric Company Paper Insulated and Lead Covered Cable. I. Solid. Size. & S. and CM. 3 y Insulation Test Pressure, 8000 Volts for 30 Minutes. £?" Insulation T«st Pressure, 10,000 Volts for 30 Minutes. 1 m .a S S3 B lid 11 Is t» _ 0) C Si la ^"0 *g S.S S3 00 S.S 3>S t-t w 10 576 .539 l T6~ 669 .602 A 400 8 632 .565 l T6~ 875 .659 A 400 6 707 .599 A 960 .693 A 400 5 753 .619 A 1,011 .713 _5_ 64 400 4 963 .672 5 64 1,075 .735 _5_ 64 400 II. Stranded. 6 737 .621 A 999 .715 A 400 5 943 .674 A 1,056 .737 A 400 4 1,012 .700 A 1,124 .763 A 400 2 1,182 .760 5 64 1,300 .823 5 64 350 1 1,300 .804 A 1,420 .867 5 64 350 100000 1,407 .833 A 1,529 .896 A 350 1,433 .843 5 64 1,555 .906 5 64 350 125000 1,513 .873 5 64 1,752 .967 3 32 350 00 1,593 .887 A 1,949 .981 A 300 150000 1,892 .939 A 2,029 1.001 A 300 000 2 006 .970 A 2,147 1.032 A 300 200000 2,187 1.012 A 2,330 1.074 A 250 0000 2,246 1.028 3 32 2,390 1.090 3 32 250 250000 2,451 1.075 A 2,597 1.137 A 250 300000 2,724 1.134 A 3,470 1.259 1 8 250 350000 2,958 1.182 A 3,715 1.307 ft 200 400000 3,795 1.290 A 3,980 1.353 ft 200 500000 4,298 1.377 i 4,488 1.440 J 200 600000 4,793 1.455 i 4,983 1.518 ft 200 700000 5,269 1.527 J 5,463 1.590 i 150 750000 5,500 1.561 i 5,702 1.624 i 150 800000 5,721 1.539 1 ¥ 5,931 1.656 J 150 900000 6,189 1.656 ft 6,390 1.719 ft 150 1000000 6,631 1.715 J 6,838 1.778 J 125 1250000 7,715 1.852 ft 7,943 1.915 ft 100 1500000 8,776 1.974 ft 9,001 2.037 ft . 100 2000000 10,834 2.194 ft 11,066 2.257 ft 100 PAPER INSULATED WIRES AND CABLES. 177 General Electric Company Paper Insulated and Lead Covered Caole. I. Solid. aV Insulation. Test Pressure, 16,000 Volts for 30 Minutes. §f" Insulation. Test Pressure, 22,000 Volts for 30 Minutes. J, DB .2 g t- 3 o:5- B. Size. &. S and CM. g3 £2 'So cj « a 5^~ Is •a o> o^ 0> o S § & 10 1,157 .820 A 1,770 1.039 'A 600 8 1,223 .846 A 1,846 1.065 A 600 6 1,313 .880 A 1,949 1.099 A 600 5 1,369 .899 A 2,013 1.119 A 550 4 1,659 .953 A 2,086 1.141 A 550 II. Stranded. 6 1,357 .902 A 2,001 1.121 A 500 5 1,639 .955 A 2,068 1.143 A 500 4 1,717 .981 A 2,155 1.169 A 500 2 1,917 1.041 A 2,959 1.292 i 500 1 2,052 1.085 A 3,121 1.336 i 500 100000 2,175 1.114 A 3,267 1.365 8 450 2,204 1.124 A 3,300 1.375 4 450 125000 2,293 1.154 A 3,404 1.405 J 450 00 2,382 1.168 4 3,508 1.419 1 8 450 150000 3,053 1.251 i 8 3,610 1.439 1 8 450 000 3,215 1.282 i 3,755 1.470 1 450 200000 3,400 1.323 i 3,970 1.512 i 400 0000 3,473 1.340 J 4,046 1.528 1 8 400 250000 3,706 1.387 * 4,293 1.575 1 8. 400 300000 4,023 1.446 1 8 4,611 1.634 i 400 350000 4,293 1.494 ft 4,888 1.682 i 350 400000 4,559 1.540 1 8 5,168 1.728 i 350 500000 5,088 1.627 1 8 5,707 1.815 I 300 600000 5,594 1.705 ft 6,228 1.893 i 300 700000 6,087 1.777 ft 6,740 1.965 i 300 750000 6,331 1.811 ft 6,983 1.999 i 300 800000 6,555 1.843 1 8 7,224 2.031 i 300 900000 7,040 1.908 ft 7,706 2.094 i 250 1000000 7,495 1.965 ft 8171 2.153 i 250 1250000 8,608 2.102 ft 9,324 2.290 i 200 1500000 9,702 2.224 1 8 10,424 2.412 i 150 2000000 11,810 2.443 i 12,579 2.631 i 150 178 PROPERTIES OF CONDUCTORS. General Electric Company Three Conductor Paper Insulated Cables. Test Pressure, 3000 Volts for 30 Minutes. o o . O n "8 *** 03 O o> Size. S .fl B. & S. and a.fl fel C^ C. M. ■5"S 2 3 .2^ ^ — v ©Pt4 Hh-3 £ 3 8 1388 .864 _5_ 64 6 1874 .979 _3_ 32 5 2072 1.027 3 3"2 4 2270 1.083 A 2 2837 1.213 A 1 3405 1.314 A 100000 3635 1.437 A 3864 1.459 A 125000 4142 1.524 A 00 4420 1.553 A 150000 4750 1.595 * 000 5081 1.663 A 200000 6300 1.815 * 0000 6700 1.852 1 ft? O • fl^^ •J flu 03 fl fl oS 150 125 125 125 100 100 100 100 100 100 100 100 100 100 Test Pressure, 8000 Volts for 30 Minutes. .fl « .SPrg tv)l-H 1,892 2,190 2,393 2,597 3.188 3,583 3,814 4,045 4,327 4,610 5,358 6,106 6,546 6,978 ■g" ■s O 4) fl) . 03 'A ^ 2 Sfl S fl fl^ •2li a J2 V 3 hJ s 1.029 * 1.114 A 1.162 A 1.218 A 1.345 A 1.441 A 1.504 A 1.525 A 1.591 A 1.622 A 1.663 A 1.795 * 1.876 I 1.919 1 8 .2 fl Si . flgS ■S-S8. Is 2 fl fl «3 200 175 175 175 150 150 150 150 150 125 125 125 125 125 Test Pressure, 15,000 Volts for 30 Minutes. Test Pressure, 26,000 Volts for 30 Minutes. 8 6 5 4 2 1 100000 125000 00 150000 000 200000 0000 2874 1.424 * 3199 1.508 _3_ 32 3422 1.557 A 3646 1.608 A 4274 1.740 A 4705 1.837 3 3~2 5407 1.962 1 8 6110 1.984 1 t 6433 2.049 * 6755 2.080 * 7134 2.122 1 8 7513 . 2.190 4 7980 2.298 1 8 8446 2.315 J 300 300 275 275 275 275 275 275 275 250 250 250 250 250 5,342 2.017 * 5,742 2 . 100 * 6,020 2.150 * 6,299 2.206 4 7,052 2.335 * 7,561 2.433 * 7,883 2.495 4 8,144 2.515 * 8,492 2.580 * 8,841 2.608 i 9,249 2.653 * 9,657 2.720 * 10,160 2.809 * 10,663 2.845 i 400 400 400 400 400 350 350 350 350 350 350 350 300 300 Thickness of insulation for 3000 volt class, sizes No. 2 and smaller, yV' paper on each conductor, T y paper over all ; sizes 0000 to No. 1 inclusive, fa" paper on each conductor, fa" paper over all. Thickness of insulation for 8000 volt class, all sizes, fa" paper on each conductor, fa" paper over all. Thickness of insulation for 15,000 volt class, all sizes, fa" paper on each conductor, fa" paper over all. Thickness of insulation for 26,000 volt class, all sizes, fa" paper on each conductor, fa" paper over alL PROPERTIES OF CONDUCTORS. 178a Varnished Cambric Cables. Special Finishes. The standard braided finish of varnished cambric cables is a weatherproof cotton braid. The following special finishes may be applied if desired: Asbestos Braid. — Generally applied over the regular cotton braid is filled with flameproof paint. It is especially recommended for interior wiring as a protection against the arcing of one cable affecting another. Asbestos braid adds about ^ to diameter of cable. Cotton Braid, Flameproof. — The standard braid may be treated with flameproof paint instead of being weatherproof ed, or one or more cotton braids may be applied, all being treated with flameproof paint. This style of finish is slightly more expensive than standard weatherproof but not as expensive as asbestos braided. Varnish Cambric cables leaded may have any of the special finishes described applied over the lead. Hoiking- and Test Pressures of Paper Insulated Lead Covered Cables. Factors by which to multiply working pressure to obtain proper test pres- sure for paper insulated lead covered cables. Test at Factory. For 5 mins., pressure = 2.5 X working pressure. For 30 mins., pressure = 2.0 X working pressure. For 60 mins., pressure = 1.6 X working pressure. Test After Installation by Manufacturer. For 5 mins., pressure = 2.0 X working pressure. For 30 mins., pressure = 1.6 X working pressure. For 60 mins., pressure = 1.3 X working pressure. CAMBRIC INSULATED WIRES AND CABLES. 179 Varnished Cambric Cables. Single Stranded Conductor — Leaded and Braided. For Working Pressures not Exceeding 1000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 A A .40 386 151 4 A A .45 490 202 2 A A .54 t 725 279 1 A A .61 1 880 362 A A .66 J 1015 448 00 A A .70 1120 534 000 s A .75 3 1301 642 0,000 A A .84 o 1690 778 250,000 A A .95 s 2267 1034 300,000 A A 1.01 s 2520 1220 350,000 3 3 2 A 1.06 >> 2780 1409 400,000 32 3 3 2 1.11 B 2994 1556 500,000 32 3 3 2 1.19 1 3473 1893 600,000 B 7 ¥ 3 S 2 1.30 3999 2281 700,000 & A 1.37 4388 2562 750,000 A A 1.41 4589 2731 800,000 A 1.44 ft 4794 2901 900,000 32 A 3 3 2 1.50 < 5241 3245 1,000,000 B5 B ? * 1.56 5656 3589 1,250,000 ^ 1,500,000 i 1,750,000 [ See 2000 volt class 2,000,000 ) For Working Pressures not Exceeding 2000 Volts. 6 4 2 1 00 000 0,000 250,000 300,000 350,000 400.000 500,000 600,000 700,000 750,000 800,000 900,000 1,000,000 1,250,000 1,500,000 1,750,000 2,000.000 32 3 3 2 3 3 2 A A A i A A I i i I i i i i 1 A B* 3 3 2 * .47 .52 .61 .65 .69 .74 .82 .87 .98 1.04 1.09 1.14 1.22 1.33 1.41 1.44 1.47 1.53 1.59 1.76 1.92 2.03 2.14 I 1 a a 468 180 570 248 870 352 963 426 1082 496 1,239 605 1,502 739 1,846 905 2,329 1063 2,590 1252 2,845 1440 3,045 1569 3,539 1924 4,068 2315 4,455 2597 4,658 2765 4,903 2938 5,311 3280 5,766 3632 7,185 4453 8,700 5297 9,793 6157 10,835 7010 Specifications, diameters and weights for solid conductors same as above. 180 PROPERTIES OP CONDUCTORS. Varnished Cambric Cables. Single Stranded Conductor — beaded and Braided. For Working Pressures not Exceeding 3,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 & & .56 565 228 4 & A .64 837 301 2 & 2 .70 961 416 1 & .74 1,128 495 & i 1 * .78 T3 1,417 570 00 £ & .86 T3 1,597 676 000 £ & .91 o3 1,786 818 0,000 £ £ 1.00 M 2,293 994 250,000 & & 1.05 s 2,495 1124 300,000 & & 1.11 a> 2,757 1319 350,000 £ & 1.15 a 3,019 1510 400,000 & & 1.20 03 3,176 1639 500,000 & & 1.29 >> 3,677 1996 600,000 & A 1.36 "3 4,145 2359 700,000 & A 1.44 03 S O 4,532 2639 750,000 & A 1.47 4,776 2811 800,000 & & 1.50 4,982 2986 900,000 & & 1.60 a 5,772 3330 1,000,000 £ & 1.65 Pi 6,237 3678 1,250,000 & i 8 1.82 7,717 4509 1,500,000 & l 1.95 8,802 5354 1,750,000 & 1 2.06 9,904 6222 2,000,000 & i 2.16 10,944 7072 For Working Pressures not Exceeding 5,000 Volts. 6 t 3 b | .69 871 285 4 t 3 b .74 999 365 2 i 3 b s .83 1,171 483 1 t 3 b A .87 1,509 568 t 3 b A .91 T3 1,608 638 00 i 3 * A .95 T3 2,023 757 000 x 5 * 3 3 2 1.04 s 2,242 1004 0,000 & A 1.09 H-3 2,525 1087 250,000 i 3 s A 1.14 s 2,695 1219 300,000 A A 1.20 o> 3,004 1442 350,000 & A 1.25 a 3,266 1619 400,000 j\ A 1.29 o3 QQ 3,427 1749 500,000 T 3 B A 1.38 >> 3,942 2116 600,000 1 3 B A 1.46 % 4,415 2485 700,000 & A 1.53 a 4,802 2771 750,000 I 3 B A 1.60 B 5,388 2946 800,000 IB A 1.63 *g 5,647 3123 900,000 IB i 1.72 6,499 3473 1,000,000 1 3 B I 1.78 p. <3 7,010 3823 1,250,000 T 3 B | 1.92 8,067 4664 1,500,000 TB | 2.04 9,168 5532 1,750,000 r 3 s * 2.15 10,282 6410 2,000,000 I 3 B i 2.26 11,318 7259 Specifications, diameters and weights for solid conductor approximately same s above. CAMBRIC INSULATED WIRES AND CABLES. 181 Varnished Camoric Cables. Single Stranded — JLeaded and Braided. Conductor For Working Pressures not Exceeding 7,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 i 4 & .81 1,112 405 4 1 s 5 * .89 1,252 497 2 | B 6 ? .95 1,653 622 1 1 B 5 i .99 1,802 714 1 & 1.06 1 2,183 812 00 I & 1.11 2,364 926 000 I 3 3 2 1.16 £ 2,594 1085 0,000 1 4 3 3 2 1.22 £ 2,898 1283 250,000 | 3?2 1.27 o3 3,144 1397 300,000 1 a 3 ? 1.32 © 3,363 1610 350,000 J T?2 1.37 i 3,642 1816 400,000 i i 1.42 i 3,888 1996 500,000 i 1.51 >> 4,327 2331 600,000 700,000 i 1 1.58 1.69 •a c3 4,816 5,633 2714 3022 750,000 1 ! 1.75 | 5,848 3197 800,000 1 I 1.78 °£ 6,546 3379 900,000 i I 1.85 8 p. 7,004 3746 1,000,000 | i 1.90 a 7,514 4111 1,250,000 i i 2.04 > 3985 1989 400,000 tb B 3 2 1.54 a> 4557 2115 500,000 5 32 1.63 a 5083 2514 600,000 R & 1.74 5598 2911 700,000 t\ I 1.84 8 6471 3213 750,000 i 5 b | 1.88 a 6756 3401 800,000 } 1.91 6984 3581 900,000 i* § 1.97 7460 3966 1,000,000 IB I 5 B i 2.03 7967 4331 Specifications, diameters and weights for solid conductor approximately same as above. 182 PROPERTIES OF CONDUCTORS. Varnished Cambric Cables. Single Stranded Conductor — Leaded and Braided. For Working Pressures not Exceeding 13,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 t A 1.09 1724 636 4 f A 1.14 | 1978 749 2 I 3 32 1.23 3 2493 878 1 f A 1.27 2668 986 t _3_ 32 1.31 o3 2766 1048 00 3 8 A 1.36 § 2997 1211 000 f A 1.41 i 3241 1383 0,000 1 A 1.47 3661 1596 250,000 f A 1.52 c3 3711 1715 300,000 f A 1.57 a 4042 1940 350,000 f A 1.62 o 4333 2159 400,000 f A 1.70 3 4981 2330 500,000 3 8 A 1.79 5469 2701 For Working Pressures not Exceeding 17,000 Volts. 6 A A 1.25 2193 755 4 A A 1.30 © 2488 873 2 A A 1.36 03 8 2803 1017 1 A A 1.40 h3 2981 1123 7 A 1.44 o3 3054 1161 00 A A 1.48 s 3316 1351 000 A A 1.53 o3 OS 3561 1530 0,000 A A 1.59 "© 3891 1757 250,000 A A 1.64 o3 J 4046 1872 300,000 A A 1.73 4793 2106 350,000 A A 1.78 g 5102 2334 400,000 A 1 8 1.86 5806 2548 500,000 7 1 8 1.94 6332 2884 Specifications, diameters and weights for solid conductor approximately same as above. CAMBRIC INSULATED WIRES AND CABLES. 183 Varnished Cambric Insulated Cables. — Single Conductor. Working Pressure, 10,000 Volts or Less. Test Pressure, 25,000 Volts. Size. B. & S. and CM. Thick. Ins. in Inches. Thick. Lead in Inches. Dia. in Inches. Braided. Weight in Lbs. per 1000 ft. Leaded. Weight in Lbs. per 1000 ft. 6 Sol. i A .80 424 1063 4 Sol. 1 A .84 498 1176 6 St. i A .82 441 1102 4 St. i A .87 521 1227 2 St. J A .96 712 1651 1 St. i A 1.04 793 1925 1/0 St. 1 A 1.08 891 2182 2/0 St. i A 1.12 1009 2365 3/0 St. 1 A 1,17 1150 2580 4/0 St. i A 1.23 1327 2839 250,000 1 A 1.28 1483 3058 300,000 i A 1.38 1707 3353 400,000 1 A 1.48 2087 4031 500,000 i A 1.57 2467 4709 750,000 tt A 1.80 3458 6470 1,000,000 * 1.96 4386 7688 Duplex cables larger than 250,000 Cm. are difficult to handle and there- fore are not recommended. The fourth column — Dia. in Inches — is the over -all diameter of the finished cable and is approximately the same for either braided or leaded. 184 PROPERTIES OF CONDUCTORS. Varnished Cambric Insulated Cables. — Single Conductor. Working Pressure, 15,000 Volts or Less. Test Pressure, 33,000 Volts. Braided. Leaded. Size. B. &. S. and Thick. Ins. in Thick. Lead in Dia. in Weight in Weight in C. M. Inches. Inches. Inches. Lbs. per 1000 ft. Lbs. per 1000 ft. 6 Sol. a A 1.05 660 1939 4 Sol. H A 1.10 767 2084 6 St. H A 1.08 705 1994 4 St. H A 1.12 797 2153 2 St. 1 1 3 2 A 1.18 927 2373 ISt. 1 A 1.29 1110 2693 1/0 St. 1 A 1.33 1225 2860 2/0 St. 1 A 1.37 1360 3051 3/0 St. t 3 32 1.42 1533 3288 4/0 St. t A 1.48 1732 3562 250,000 f A 1.53 1901 3795 300,000 t 7 FT 1.63 2130 4487 400,000 i J 1.73 2530 5246 500,000 i i 1.82 2930 6006 750,000 If I 2.05 3998 7468 1,000,000 a i 2.23 5005 8835 Duplex cables larger than 250,000 Cm. are difficult to handle and there- fore are not recommended. The fourth column — Dia. in Inches — is the over-all diameter of the finished cable and is approximately the same for either braided or leaded. CAMBRIC INSULATED WIRES AND CABLES. 185 Varnished Cambric Cables. Triple Stranded Conductor — Leaded and Braided. For Working Pressures not Exceeding 1,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 100 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 4 l l T6~ 64 i l A A 5 64 A .824 .959 i 1,245 1,820 538 760 2 1 16 64 A-A 64 64 64~64 5 1_ 64 64 5 1 64 64 64 64 1.085 1.279 2,290 3,066 1089 1384 00 000 0,000 250,000 A A A A 1.357 1.456 1.566 1.723 1.891 m o3 O 1 3,446 3,933 4,528 5,642 6,470 1660 2003 2416 2955 3537 300,000 350,000 400,000 500,000 _3 1_ 32 64 3 1 32 64 J- 1 W2 64 32 64 A l 8 1 8 1 8 2.023 2.150 2.253 2.438 a •a o u Q* a < 7,296 8,595 9,347 10,870 4155 4770 5288 6483 For ' Forking '. Pressures not Exci :eding 3,0 00 Volts. 6 4 2 5 1 5 1.016 1,803 692 64 16 A _ A 64 A 3 32 1.120 1.277 T3 -8 2,159 2,955 930 1273 1 64" 1~6" A 1.363 h5 3,290 1505 64~ — 16" A 1.451 cS 3,725 1795 00 64 — T6~ A 1.550 a 4,206 2139 000 64 "~ T6~ A 1.691 g 5,184 2573 0,000 64~~16 A 1.815 >> 5,928 3115 250,000 32~T6~ 7 64 1.984 6,805 3704 300,000 A~T6" 1 8 2.134 8,169 4344 350,000 3~2~1T 1 I 2.243 2 8,986 4973 400,000 3%~16~ 1 8 2.346 <5 9,692 5492 500,000 ft-A 1 8 2.531 11,288 6713 Note. — Under Thickness Insulation: The first fraction is thickness of insu- lation on each conductor; the second fraction is thickness of insulation over all. 186 PROPERTIES OF CONDUCTORS. Varnished Cambric Cables. Triple Stranded — Leaded and Braided. Conductor For Working Pressures not Exceeding 5,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 32"~3~2 A 1.15 2,092 835 4 32~ 3~2" A 1.28 o "d 2,765 1083 2 3lJ~"32 A 1.41 8 3,302 1444 1 32~32 A 1.50 h5 s 3,682 ■ 1686 A~A A 1.58 4,084 1982 00 32 - 32 A 1.71 a 4,989 2338 000 32"~ 32 A 1.82 3 5,640 2790 0,000 J- J~ 32 32 A 1.95 >> 6,356 3342 250,000 3 3 32 32 1 8 2.08 7,517 3835 300,000 32 32 1 8 2.20 a 8,398 4476 350,000 3~2~"~ 32 1 8 2.31 | 9,267 5113 400,000 32~32 1 8 2.41 9,978 5641 500,000 32~32 1 8 2.60 11,533 6866 For Working Pressures not Exceeding 7,000 Volts. 6 8 8 A 1.38 2,909 1083 4 i-i A 1.48 3,317 1352 2 hi A 1.61 t3 3,867 1733 1 l l 8 8 A 1.69 h3 4,268 1991 1 1 A 1.81 3 5,115 2302 00 1 1 8~8 A 1.91 a 5,651 2673 000 i-i 7 2.02 02 6,280 3139 0,000 i i 8 8 1 8 2.18 j>> 7,585 3713 250,000 1 1 8 8 1 8 2.28 1 8,259 4200 300,000 1 1 8 8 1 8 2.39 a •a o 9,183 4892 350,000 1 1 8 8 1 8 2.50 10,075 5550 400,000 H 1 8 2.61 a 10,800 6086 500,000 1 1 8 8 1 8 2.79 << 12,392 7348 Note. — Under Thickness Insulation: The first fraction is thickness of insula- tion on each conductor; the second fraction is thickness of insulation over all. CAMBRIC INSULATED WIRES AND CABLES. 187 Varnished Cambric Cables. Triple Stranded Conductor — Steaded and Braided. For Working Pressures not Exceeding 10,000 Volts. Size. Thick. Insulation in Inches. Thick. Lead in Inches. Diameter. Weight in Lbs. per 1000 Feet. Leaded. Braided. Leaded. Braided. 6 4 2 1 00 000 0,000 250,000 300,000 350,000 400,000 500,000 32"~32~ 5 5 32 32 3~2~~32" 3%~A 32 32 32~~32 _5 5_ 32 32 32 ~ 32 T2~ 32" 32~3"2 5 5 32 32 5 5 32 32 32~T2 i i 1 i i i i 1.57 1.68 1.80 1.92 2.01 2.11 2.25 2.37 2.47 2.59 2.70 2.81 2.99 T3 73 c3 O h3 03 C3 a o3 03 >> s o3 J 2 a p. 3,480 3,992 4,480 5,309 5,797 6,336 7,539 8,373 9,083 10,010 10,883 11,660 13,290 1378 1661 2065 2331 2656 3031 3526 4127 4654 5343 6023 6569 7868 187a cambric insulated wires and cables. Varnished Cambric Cables. Triple Stranded Conductor — Leaded and Braided. For Working Pressures not Exceeding 13,000 Volts. Thick. Thick. Diameter. Weight in Lbs. per 1000 Feet. Size. Insulation in Inches. Lead in Inches. Leaded. Braided. Leaded. Braided. 6 A-A 3 32 1.77 4,103 1720 4 A-A A 1.87 > 9,199 4582 250,000 _3 3_ 16 16 i 2.67 +3 9,933 5128 300,000 A-A i 2.79 1 10,884 5840 350,000 A-A i 2.90 O 11,779 6541 400,000 16 16 i 3.00 a a 12,511 7089 500,000 _3 3 16 TG i 3.18 . < 14,202 8402 For Working Pressures not Exceeding 17,000 Volts. 6 7 7 32 32 3 32 1.97 *6 0> 4,784 2123 4 JL_JL 32 32 A 2.10 03 5,724 2419 2 32 _ 32 7 64 2.23 Hi 6,381 2877 1 32~32 1 8 2.34 I 7,364 3164 32~32 1 8 2.43 7,906 3519 00 W2TTZ 1 8 2.53 c3 CO 8,459 3884 000 7 7 32 32 1 8 2.64 £ 9,218 4454 0,000 32~"~ 32" 1 8 2.77 -M d 10,091 5092 250,000 7 7 32 32 1 8 2.88 a 10,847 5664 300,000 32 32 1 8 2.99 11,802 6380 350,000 32~32 1 8 3.10 a a < 12,697 7086 Note. — Under Thickness Insulation: The first fraction is thickness of insulation on each conductor; the second fraction is thickness of insulation over all. ENAMELED WIRE. 187b Enameled IFire. Diameter in Inches. Comparative Weight per 1000 Feet in Pounds. Size B. &S. Bare. Over Enamel. Single Cotton- Covered. Enamel. 14 .0640 .0670 12.684 12.684 15 .0570 .0600 10.082 10.053 16 .0510 .0535 8.012 7.973 17 .0450 .0475 6.375 6.322 18 .0400 .0425 5.081 5.009 19 .0360 .0380 4.043 3.966 20 .0320 .0340 3.218 3.136 21 .0280 .0305 2.569 2.475 22 .0250 .0275 2.055 1.970 23 .0230 .0250 1.630 1.555 24 .0200 .0220 1.297 1.232 25 .0180 .0200 1.036 .980 26 .0160 .0175 .828 .777 27 .0140 .0155 .661 .616 28 .0126 .0140 .524 .485 29 .0110 .0123 .421 .384 30 .0100 .0113 .336 .303 31 .0090 .0102 .271 .242 32 .0080 .0092 .215 .192 33 .0070 .0082 .174 .152 34 .0063 .0075 .141 .121 35 .0056 .0063 .120 .101 36 .0050 .0056 .099 .081 37 .0045 .0051 .090 .061 38 .0040 .0046 .080 .0507 40 .0031 .0037 .060 .0304 188 PROPERTIES OF CONDUCTORS. TELEPHONE CABLE!. JLead Sheathed for Underground or Aerial Use. The insulation of these cables is dry paper. The following specifications have been adopted by the larger telephone companies and, therefore, may be considered standard. Cable Conductor. No. 19 B. and S. G., 98% conductivity, insulated with one or two paper tapes; conductor twisted in pairs; one of the pairs to have a distinctive colored paper for marker; length of twist not to exceed 3". Pairs to be laid up in reverse layers; insulation to be unsaturated ex- cept two feet from each end to prevent moisture from entering. The lead sheath to have an alloy of 2\ to 3^% of tin; thickness of sheath xV for fifty pair of cables, &" for one hundred pair of cables, and £" for larger sizes. Insulation resistance to be at least 100 megohms per mile after the cable is laid and spliced. Electrostatic capacity no greater than .054 with a maximum of .060 microfarads per mile. The aerial cables for telephone companies usually follow the same speci- fications as those for underground use, being purchased with the ultimate intention of being put underground. Cables that are to remain overhead indefinitely are usually made with a lighter sheathing of lead than that specified for underground work. Number Pairs. Outside Diameters. Inches. Weights 1000 feet. Pounds. 1 & 214 2 I 302 3 515 4 f 629 5 747 6 |1 877 7 T$ 912 10 t! 1,214 12 H 1,373 15 l 1,566 18 l& 1,758 20 25 li 6 * 1,940 2,332 30 l& 2,748 35 l* 2,985 40 U\ 3,176 45 if 3,365 50 if 3,678 55 m 3,867 60 u 4,055 65 m 4,241 70 2 4,430 80 2* 4,804 90 21 5,180 100 i 5,505 SUBMARINE CABLES. 189 TELEGRAPH < VBJLl]*. Lead Sheathed for Underg-round or Taped and Braided for Aerial Use. The insulation of these cables is made of a compound containing not less than thirty per cent pure Para rubber. These specifications may be considered standard, being used by the principal telegraph companies. Rubber Insulated Aerial Teleg-rapli Cable. Gauge B. & S. No. of Conductors. Outside Diameter. Weight per 1,000 ft. 14 14 14 7 10 19 .3." IF 425 lbs. 500 lbs. 890 lbs. Conductors No. 14 B. and S. insulated to diameter of 6-32", cabled together and covered with a rubber tape, one layer of tarred jute, a rubber tape, and a heavy cotton braid saturated with waterproof compound. SUBMARINE CABLE§. These cables are insulated with a rubber compound containing not less than thirty per cent (30% ) of pure Para rubber. These specifications have been adopted by the various telegraph com- panies and the United States Government for general use. No. of Conduc- tors. Gauge of Con- ductors. No. of Armor Wires. Gauge of Armor Wires. Outside Diameter. Weight per 1,000 feet. 1 2 3 4 5 6 7 10 14 B. & S. 14 B. & S. 14 B. & S. 14 B. & S. 14 B. & S. 14 B. & S. 14 B. & S. 14 B. & S. 12 16 14 16 19 21 21 22 8 B. W. G. 8 B. W. G. 6 B. W. G. 6 B. W. G. 6 B. W. G. 6 B. W. G. 6 B. W. G. 4 B. W. G. 1" w U" ItV if" H" 1J" 11" 1150 1675 2400 2750 3100 350C 3600 4600 Conductors built up of 7 No. 21 B. & S. copper wires, heavily tinned. Each conductor insulated with ^" Rubber and Taped. The above specifications refer only to river and harbor cables. Ocean cables are of an entirely different character, and consist of Shore End, In- termediate and Deep Sea Types. 190 PROPERTIES OF CONDUCTORS. •Joints in Rubber Insulated Cables. Preparation of Ivnds. — Remove the outside protecting braid or tape, and bare the conductor of its rubber insulation for two or three inches back from the end. Clean the metal carefully by scraping with a knife or with sandpaper. Metal Joint. — If solid conductor, scarf the ends with a file so as to give a good long contact surface for soldering. If conductor is stranded, carefully spread apart the strands, cutting out the centers so conductors can be butted together, the loose ends interlacing as in Fig. 1, and bind wires down tight as in Fig. 2, with gas or other- pliers. Solder carefully, Fig. 1. Fig. 2. using no acid; resin is the best, although jointers often use a spermaceti candle as being handy to use and easy to procure. Large cables are easiest soldered by dipping the joint into a pot of molten solder, or by pouring the molten metal over the joint. The insulation of all kinds of joints is done in the same manner, the only difference in the joint being the manner in which the conductors are joined together. Following are some of the styles of joining conductors, which are afterward insulated with rubber, and covered with lead when necessary. Fig. 3. Fig. 4. Figs. 5, 6, Seeley's Cable Connectors. — The cuts below show a style of cop- per connectors very handy in joining cables. They are copper tinned over, and after putting in place can be "sweated" on with solder; when dry can be insulated as previously described. Figs. 7, 8. JOINTS IN CABLES. 191 Insulating* the Joint. — Jointers must have absolutely dry and clean hands, and all tools must be kept in the best possible condition of cleanliness. Clean the joint carefully of all flux and solder ; scarf back the rubber insulation like a lead-pencil for an inch or more with a sharp knife. Carefully wind the joint with three layers of pure unvulcanized rubber, taking care not to touch the strip with the hands any more than neces- sary ; over this wind red rubber strip ready for vulcanizing. Lap the tape upon the taper ends of the insulation, and make the covering of the same diameter as the rubber insulation on the conductor, winding even and round. Cover the rubber strip with two or three layers of rubber-saturated tape. Lead covering-. — If the insulation is covered and protected by lead, a loose sleeve is slipped over one end before jointing, and slipped back over the joint when the insulation is finished, a plumber's wiped joint being made at the ends. Fig. 9. Joints in Waring- Cables. —This cable is covered with cotton, thoroughly impregnated with a composition of hydro-carbon oils applied at high temperature, the whole being covered with lead to protect the insula- tion. The insulating properties of this covering are very high if the lead is kept intact. Metal joints are made as usual, and a textile tape may be used for cover- ing the bare copper. A large lead-sleeve is then drawn over the joint, and wiped onto the lead covering at either end ; then the interior space is filled with a compound similar to that with which the insulation is im- pregnated. Joints in Paper Insulated Cables. — This cable is covered or insulated with narrow strips of thin manila paper wound on spirally, after which the whole is put into an oven and thoroughly dried, then plunged into a hot bath of resin oil, which thoroughly impregnates the paper. This insulation is not the highest in measurement, but the electrostatic capacity is low and the breakdown properties high. When used for telephone pur- poses the paper is left dry, and is wound on the conductor very loosely, thus leaving large air spaces and giving very low electrostatic capacity. Joints are made as in the Waring cable by covering the conductor with paper tape of the same kind as the insulation, then pulling over the lead sleeve, which is finally filled with paraffine wax. Dossert Joint. — Dossert & Company, New York City, make a mechanical joint for solid or stranded conductor which has great mechanical strength and an electrical conductance in excess of that of the cable. The joint illustrated in Fig. 10 consists of a nipple (A), two compression sleeves or bushings (B) and two compression nuts (C). As shown in Fig. 11, the compression sleeves are split lengthwise and tapered at both ends. The tapered ends of the sleeve fit into correspond- ingly tapered parts of the nipple and nut. When the nut is screwed upon the nipple the action of the taper causes the compression sleeve to decrease in diameter and grip the strands tightly together, thereby getting good elec- trical contact. To make a splice with this connector cut the insulation from the cables to a distance equal to half the length of the connector, slip the cable into the connector and screw the nuts up tightly on the nipple. 192 PROPERTIES OF CONDUCTORS. DETAIL TMO-W/iY TOIN~r m ■ e Fig. 10. LONGWUOlrfAL SECTION TO I NT TNO MY" SPL ICE COMPUTE Fig. 11. Lugs, 3-Ways, Y's, Reducers, Elbows and many other types of connectors are made with this principle for making the electrical connections and can be used for connections on switchboards, bus bars, transformers, meters, oil switches, storage batteries, electric smelting furnaces and the like. A special application of this joint is the cable tap as shown in Figs. 12 and 13. It consists of a hook (A), cover (B), jam nut C), compression sleeve (D) and compression nut (E). The hook is machined to fit the main cable while its shank is drilled and threaded to form the nipple of a standard Dossert joint for size of branch required. The branch is secured to the DETAIL CABLE TAP Brewh CONNECTION COn PLC TE Figs. 12 and 13. connector by inserting it in the sleeve (D) and screwing nut (E) up tight. Connection is made to main by placing the hook part of the connector over the main cable, inserting the cover (B) and screwing up the jam nut (C). For overhead work where the cables are subjected to considerable tensile strain the Company makes another type of joint. JOINTS IN COPPER WIRES. 193 Jointing- Cwutta-JPerclia Covered IFire. First remove the gutta-percha for about two inches from the ends of the wires which are to be jointed. Fig. 14. Fig. 14. Next cross the wires midway from the gutta-percha, and grasp with the pliers. Fig. 15. Fig. 15. Then twist the wires, the overlapping right-hand wire first, and then, reversing the grip of the pliers, twist the left-hand wire over the right. Cut off the superfluous ends of the wires and solder the twist, leaving it as shown in Fig. 16. Fig. 16. Next warm up the gutta-percha for about two inches on each side of the twist. Then, first draw down the insulation from one side, half way over Fig. 17. the twisted wires, Fig. 17, and then from the other side in the same way, Fig. 18. Fig. 18. Then tool the raised end down evenly over the under half with a heated Iron. Then warm up the whole and work the "drawdown" with the thumb and forefinger until it resembles Fig. 19. Now allow the joint to cool andfset. Fig. 19. Next roughen the drawdown with a knife, and place over it a thin coating of Chatterton's compound for one inch, in the center of the drawdown, which is also allowed to set. Next cut a thick strip of gutta-percha, about an inch wide and six inches 194 PROPERTIES OF CONDUCTORS. long, and wrap this, after it has been well warmed by the lamp, evenly over the center of the drawdown. Fig. 20. Fig. 20. The strip is then worked in each direction by the thumb and forefinger over the drawdown until it extends about 2 inches from center of draw- down. Then tool over carefully where the new insulation joins the old, after which the joint should be again warmed up and worked with the fore- finger and thumb as before. Then wet and soap the hand, and smooth and round out the joint as shown in Fig. 21. Fig. 21. Between, and at every operation, the utmost care must be exercised to remove every particle of foreign matter, resin, etc. Note. Chatterton's compound consists of 1 part by weight Stockholm tar; 1 part resin; 3 parts Gutta-percha. AMJUmTNUXK WIRE. Physical Constants of Commercially (99%) Pare Aluminum. Per cent Conductivity (Copper 100) Specific Gravity Pounds in 1 cubic foot Pounds in 1 cubic inch Pounds per mile per circular mil Ultimate strength, r— sq. m. Modulus of Elasticity, . '— r — in. X sq. in. Coefficient of Linear Expansion per ° C Coefficient of Linear Expansion per °F Melting Point in °C Melting Point in °F Specific Heat (watt-seconds to heat 1 lb. 1° C.) Thermal Conductivity (watts through cu. in. temperature grad- ient 1° C.) Resistance Microhms of centimeter cube at 0° C , Microhms of inch cube at 0° C , Ohms per mile-foot at 0° C , Ohms per mil-foot at 20° C Ohms per mile at 0° C Ohms per mile at 20° C Pounds per mile-ohm at 0° C Pounds per mile-ohm at 20° C Temperature coefficient per ° C Temperature coefficient per ° F 62 2.68 167 .0967 .00481 26.000 9,000,000 .0000231 .0000128 625 1157 402 36.5 2.571 1.012 15.47 16.70 81,700 cir. mils 88,200 cir. mils 393 424 .004 .0022 ALUMINUM WIRE. 195 Aluminum and Copper Compared. Aluminum wire of 62% conductivity is the generally accepted standard. Aluminum of 62% conductivity, bought at 2.13 times the price of cop- per per pound, will give the same length and conductivity for the same expenditure. Comparative Cost of Aluminum of G2% Conductivity and Copper for Equal JLeng-tli and Conductivity. Cost per Pound Cost per ] Pound of Copper of 100 % Conductivity. of Aluminum of 62% Conductivity. 14 cents 28.8 cents 15 * 32.0 " 16 ' 34.1 " 17 ' 36.2 •■ 18 4 38.4 " 19 ' 40.5 «• 20 * 42.6 •• 21 ' 44.7 " 22 • 46.8 «« 23 * 49.0 ■■ 24 ' 51.1 '• 25 - 53.2 i« Comparison of Copper and Aluminum of Various Conduc- tivities for Initial Length and Conductivity. Metal. Conduc- tivity. Cross Section. Weight. Breaking Weight.* Price per lb. Copper .... 100 100 100.0 100 100 Aluminum 54 180 54.0 85.1 185 55 176 53.0 83.5 189 56 173 52.0 82.0 192 57 170 51.1 80.6 196 58 167 50.2 79.2 199 59 164 49.4 77.9 203 60 162 48.6 76.6 206 61 159 47.8 75.3 210 62 157 47.0 74.1 213 63 154 46.3 72.9 216 * Breaking weights (pounds to break wire of equal conductivity) are cal- culated on the assumption of an ultimate strength of 55,000 pounds per square inch for copper and 26,000 pounds per square inch for aluminum. 196 PROPERTIES OF CONDUCTORS. Table of Iteciatances of Solid Aluminum Wire 62% Conductivity .* Pittsburg Reduction Co. Conductivity 62 in., the Matthiessen Standard Scale . Pure aluminum weighs 167.111 pounds per cubic foot. ¥ Resistances at 70° ] F. Log d 2 . R Ohms per 1000 Feet, Ohms per Mile. Feet per Ohm. Ohms per lb. Logfl. 0000 .07904 .41730 12652. .00040985 5.325516 3 .897847 000 .09966 .52623 10034. .00065102 5.224808 3 .998521 00 .12569 .66362 7956. .0010364 5.124102 T . 099301 .15849 .83684 6310. .0016479 5.023394 r . 200002 1 .19982 1.0552 5005. .0026194 4.922688 T . 300639 2 .25200 1.3305 3968. .0041656 4.821980 T . 401401 3 .31778 1.6779 3147. .0066250 4.721274 j. 502127 4 .40067 2.1156 2496. .010531 4.620566 1.602787 5 .50526 2.6679 1975. .016749 4.519860 T . 703515 6 .63720 3.3687 1569. .026628 4.419152 T . 804276 7 .80350 4.2425 1245. .042335 4.318446 T . 904986 8 1.0131 5.3498 987.0 .067318 4.217738 0.005652 9 1.2773 6.7442 783.0 .10710 4.117030 0.106293 10 1.6111 8.5065 620.8 .17028 4.016324 0.207122 11 2.0312 10.723 492.4 .27061 3.915616 0.307753 12 2.5615 13.525 390.5 .43040 3.814910 0.408494 13 3.2300 17.055 309.6 .68437 3.714202 0.509203 14 4.0724 21.502 245.6 1.0877 3.613496 0.609850 15 5.1354 27.114 194.8 1.7308 3.513788 0.710574 16 6.4755 34.190 154.4 2.7505 3.412082 0.811273 17 8.1670 43.124 122.50 4.3746 3.311374 0.912063 18 10.300 54.388 97.15 6.9590 3.210668 1.012837 19 12.985 68.564 77.06 11.070 3.109960 1.113442 20 16.381 86.500 61.03 17.595 3.009254 1.214340 21 20.649 109.02 48.44 27.971 2.908546 1.314899 22 26.025 137.42 38.4 44.450 2.807838 1.415391 23 32.830 173.35 30.45 70.700 2.707132 1.516271 24 41.400 218.60 24.16 112.43 2.606424 1.617000 25 52.200 275.61 19.16 178.78 2.505718 1.717671 26 65.856 347.70 15.19 284.36 2.405010 1.818595 27 83.010 438.32 12.05 452.62 2.304304 1.919130 28 104.67 552.64 9.55 718.95 2.203596 2.019822 29 132.00 697.01 7.58 1142.9 2.102890 2.120574 30 166.43 878.80 6.01 1817.2 2.002182 2.221232 31 209.85 1108.0 4.77 2888.0 1.901476 2.321909 32 264.68 1397.6 3.78 4595.5 1.800768 2.422721 33 333.68 1760.2 3.00 7302.0 1.700060 2.523330 34 420.87 2222.2 2.38 11627. 1.599354 2.624148 35 530.60 2801.8 1.88 18440. 1.498646 2.724767 36 669.00 3532.5 1.50 29352. 1.397940 2.825426 37 843.46 4453.0 1.19 46600. 1.297234 2.926064 38 1064.0 5618.0 .95 74240. 1 . 196526 3.026942 39 1341.2 7082.0 .75 118070. 1.095820 3.127494 40 1691.1 8930.0 .59 187700. 0.995112 3.228169 * Calculated on the basis of Dr. Matthiessen's standard, viz.: The re- sistance of a pure soft copper wire 1 meter long, having a weight of 1 gram = .141729 International Ohm at 0° C. The purest aluminum obtainable has a conductivity of over 63 per cent, but this gain in conductivity is at a greatly increased cost. STRANDED ALUMINUM WIRE. 197 Stranded Weatherproof Aluminum Wire. (Triple Braid.) Circular Mils and B. & S. Gauge. Diameter in Mils. Lbs. per 1000 ft. Circular Mils and B. & S. Gauge. Diameter in Mils. Lbs. per 1000 ft. 1,000,000 1.152 1408 400,000 .728 567 950,000 1.125 1340 350,000 .679 502 900,000 1.092 1270 300,000 .630 436 850,000 1.062 1202 250,000 .590 375 800,000 1.035 1135 0000 .530 280 750,000 .999 1067 000 .470 232 700,000 .963 1001 00 .420 192 650,000 .927 938 .375 155 600,000 .891 878 1 .330 132 550 000 .855 806 2 .291 108 500,000 .819 740 3 .261 88 450,000 .770 665 1 4 .231 72 Dimensions and Resistances of Stranded Aluminum Wire. H. W. Buck. Relative Conductivity 62%. Resistance per Mil-foot 16.95 ohms. Temperature 75° F. Elastic Limit 14,000 lbs. per square inch. Ultimate Strength 26,000 lbs. per square inch. -a Pounds per Ohms per 1 en 0; J Size C. M. 03 02 Area Sq. Feet 3jg and B. & S. Inch. per Lb. o^ s a 1 a* s 1000 Feet. Mile. 1000 Feet. Mil. 00 03 H pg & 1,000,000 1.15 .7870 920 4,858 1.087 .01695 .08950 10,995 20,420 950,000 1.12 .7470 874 4,617 1.144 .01784 .09420 10,440 19,400 900,000 1.09 .7075 828 4,374 1.208 .01883 .09942 9,900 18,380 850,000 1.06 .6680 782 4,131 1.279 .01994 .10529 9,350 17,360 800,000 1.03 .6290 736 3,888 1.359 .02119 .11188 8,800 16,340 750,000 1.00 .5890 690 3,645 1.449 .02260 .11933 8,230 15,320 700,000 .96 .5500 644 3,402 1.553 .02421 .12782 7,700 14,300 650,000 .93 .5120 598 3,159 1.672 .02608 .13770 7,150 13,270 600,000 .89 .4720 552 2,916 1.812 .02825 .14917 6,600 12,250 550,000 .85 .4330 506 2,673 1.977 .03082 . 16275 6,050 11,230 500,000 .81 .3930 460 2,430 2.041 .03300 . 17900 5,500 10,210 450,000 .77 .3540 414 2,187 2.415 .03766 . 19884 4,950 9,190 400,000 .73 .3141 368 1,944 2.718 .04237 .22370 4,400 8,170 350,000 .68 .2750 322 1,701 3.106 .04843 .25570 3,850 7,150 300,000 .63 .2360 276 1,458 3.623 .05652 .29830 3,300 6,130 250,000 .58 .1965 230 1,215 4.348 .06780 .35800 2,750 5,110 0000 .54 .1661 194.7 1,028 5.733 .08010 .42290 2,330 4,320 000 .47 .1317 154.4 816 6.477 . 10100 .53315 1,850 3,430 00 .42 .1045 122.4 647 8.165 .12740 .67270 1,460 2,720 .37 .0829 97.1 513 10.300 .16050 .84740 960 2,150 1 .33 .0657 77.0 407 12.990 .20250 1.0692 920 1,710 2 .30 .0521 61.0 323 16.400 .25540 1.3486 730 1,355 3 .26 .0413 48.5 256 20.620 .32200 1.7002 579 1,075 4 .23 .0327 38.5 203 25.970 .40600 2.1438 450 852 198 PROPERTIES OF CONDUCTORS. s a s g ST c (-1 t- ftft " 00 *§i o °-0 iO"tf <£> " "a B • • a .3 M o: , OP 6 ■5 5 .5 02 * s B ffl a o * S s W 9 SI? t o t o a ££§ :s:si g> S> a > > c3 tfPnPn S £ £a2 Co a 03 ."73 O So OS *2 jog Ph ^ p go o o«o pm 2 02 S 2 3^ O) fH TO . "god 00 P °r^ . ooooo ooooo ooooo ooooo OMOOO o qoco^ococni T-HOioot^io ^nhooo i>co io (MCOO^QO cNiOOCOi> riioOCOCO O00^COlo"'*C0 COcNt-To'csT 00NO<©iO ^"cOCNTcNtH i-Ti-T OrHOOOOiO 0"*. OCOOOON lOOiHMTjt lOO^iOO "tfcNOCOCO O CO Ci i>0000O>Ci rHiO!>iO'-i *0 OOOOO OOOOO tHt-ItHi-ith CN >-i iHl-li-H OS OS *0 OhhMcO ^ iO CO 1> 00 H«00»0 MOWOh OOi-iCOcOCO 00 r-H rH i-H ,-1 tH rH ,-( tH tH i-H C3 CN CN CO CO "^ iO CO 00 O N©0»0(N O OOOOO OOOOO OOOOO OOOOh t-it-icN(NCO ^ 00NOICO ©t^NOh CO O CO 00 OS "tf C0C0b-O O OOINCOOCO i-h CO »0 rtn lo t^CMrHCOO iCiOOOH H®t»N«3 00 «oi>i>oooo ooo'-ioq coowooq mnohio cmcoo'*^ "- 1 iO CO»OC5*OiO t^ OOrHCNcOT-i i-i»OOsr)HC0 O t> '-• t>- CO CHOOOcMO MOfOO^ '- | !>OrHCNTt< CONMOM CO*OI>000 »-iCN|00C0O5L0oOiO^O l>i0^01>C^ ^ TjH^HOiCOCO H00ONO5 r^r^T-iOO CO^-tfCOCo' i-HCOOOOO oo lOOOOCOO Cir-tiot^Ci M^OOOCO COO"*05iO 0^ CO ^COCOCN|t-i OO^00t> t>^0 OiO"^^CO COCOiOiO"* T^COCOCO MO5HO5C0 " : ^ o"r-To4"co^" i6<£i^cca$ O CO CO ^ CO COOOOCOiO ^ tOiO-^COCN hhOOOO t>l>COiO^ COCOCNCNr-i tH 1-1 OOOOO OOOOO OOOOO OOOOO r-icNC0Tt 3 .Q a 'S 3 E.B.B. B.B. Steel. E.B.B. B.B. Steel. * S £ pui 4 .225 730 | mile. 2,190 2,409 2,701 6.44 7.53 8.90 6 .192 540 £ mile. 1,620 1,782 1,998 8.70 10.19 12.04 8 .162 380 £ mile. 1,140 1,254 1,406 12.37 14.47 17.10 9 .148 320 J mile. 960 1,056 1,184 14.69 17.19 20.31 10 .135 260 £ mile. 780 858 962 18.08 21.15 25.00 11 .120 214 £ mile. * 642 706 792 21.96 25.70 30.37 12 .105 165 ■£■ mile. 495 545 611 28.48 33.33 39.39 14 .080 96 i mile. 288 317 355 48.96 57.29 67.71 The values given in this table are averages of a large number of tests, They are within the limits of the specifications of the Western Union Tele- graph Company. The average value of the mile-ohm is 4,700 for E. B. B. wire. The average value of the mile-ohm is 5,500 for B. B. wire. The average value of the mile-ohm is 6,500 for Steel wire. The average breaking strain is 3 times the weight per mile for E.B.B. wire. The average breaking strain is 3.3 times the weight per mile for B. B. wire. The average breaking strain is 3.7 times the weight per mile for Steel wire. The mile-ohm = weight per mile X resistance per mile. Galvanized Sig*nal Strand. Seven "Wires. Diameter, Weight per 1000'. Estimated Breaking Inches. Bare Strand. Double Braid W. P. Triple Braid W. P. Weight. 1-2 520 616 677 8,320 15-32 420 510 561 6,720 7-16 360 444 488 5,720 3-8 290 362 398 4,640 5-16 210 270 297 3,360 9-32 160 214 235 2,560 17-64 120 171 188 1,920 1-4 100 148 163 1,600 7-32 80 122 134 1,280 3-16 60 96 105 960 11-64 43 76 84 688 9-64 33 60 66 528 1-8 24 48 53 384 3-32 20 38 42 320 IRON AND STEEL WIRE. 201 Properties of Steel Wire. ROEBLING. Note. — The breaking weights given for steel wire are not those of Steel Telegraph wire. They apply to wire with a tensile strength of 100,000 pounds per square inch. This strength is higher than that of telegraph wire. No. t Diam- eter in Inches. Area in Square Inches. Breaking Strain 100,000 lbs. sq. inch. Weight in Pounds. Roeb- lingG Per 1,000 ft. Per Mile. Feet in 2,000 lbs. 6-0 .460 .166191 16,619 558.4 2,948 3,582 5-0 .430 .145221 14,522 487.9 2,576 4,099 4-0 .393 .121304 12,130 407.6 2,152 4,907 3-0 .362 .102922 10,292 345.8 1,826 5,783 2-0 .331 .086049 8,605 289.1 1,527 6,917 .307 .074023 7,402 248.7 1,313 8,041 1 .283 .062902 6,290 211.4 1,116 9,463 2 .263 .054325 5,433 182.5 964 10,957 3 .244 .046760 4,676 157.1 830 12,730 4 .225 .039761 3,976 133.6 705 14,970 5 .207 .033654 3,365 113.1 597 17,687 6 .192 .028953 2,895 97.3 514 20,559 7 .177 .024606 2,461 82.7 437 24,191 8 .162 .020612 2,061 69.3 366 28,878 9 .148 .017203 1,720 57.8 305 34,600 10 .135 .014314 1,431 48.1 254 41,584 11 .120 .011310 1,131 38.0 201 52,631 12 .105 .008659 866 29.1 154 68,752 13 .092 .006648 665 22.3 118 89,525 14 .080 .005027 503 16.9 89.2 118,413 15 .072 .004071 407 13 7 72.2 146,198 16 .063 .003117 312 10.5 55.3 191,022 17 .054 .002290 229 7.70 40.6 259,909 18 .047 .001735 174 5.83 30.8 343,112 19 .041 .001320 132 4.44 23.4 450,856 20 .035 .000962 96 3.23 17.1 618,620 21 .032 .000804 80 2.70 14.3 740,193 22 .028 .000616 62 2.07 10.9 966,651 23 .025 .000491 49 1.65 8.71 24 .023 .000415 42 1.40 7.37 25 .020 .000314 31 1.06 5.58 26 .018 .000254 25 .855 4.51 27 .017 .000227 23 .763 4.03 28 .016 .000201 20 .676 3.57 29 .015 .000177 18 .594 3.14 30 .014 .000154 15 .517 2.73 31 .0135 .000143 14 .481 2.54 32 .013 .000133 13 .446 2.36 33 .011 .000095 9.5 .319 1.69 34 .010 .000079 7.9 .264 1.39 35 .0095 .000071 7.1 .238 1.26 36 .009 .000064 6.4 .214 1.13 This table was calculated on a basis of 483.84 pounds per cubic foot for steel wire. Iron wire is a trifle lighter. The breaking strains are calculated for 100,000 pounds per square inch throughout, simply for convenience, so that the breaking strains of wires of any strength per square inch may be quickly determined by multiplying the values given in the tables by the ratio between the strength per square inch and 100,000. Thus, a No. 15 wire, with a strength per square inch of 150,000 pounds, has a breaking strain of 407 X \^qqq = 610.5 pounds. The "Roebling" or "Market wire Gauge" is now used as standard for steel wires in America. 202 PROPERTIES OF CONDUCTORS. II E*I*T A\C i: %V I It i:*. Specific Resistance and Temperature Coefficient. Substance. Microhms per Cubic Centimeter about 20° F. Temperature Coeffi- cient per ° C. Platinum silver (Pt 66, Ag 33) Patent-Nickel (Cu74.41,Zn0.23,Ni25.10,Fe0.42, Mn0.13) Platinoid (Cu 59 Zn 25.5, Ni 14, W 55) German Silver (Cu, Zn, Ni in various proportions) Manganin (Cu, Ni, and Fe-Mn in various propor- tions) Boker & Co.'s Iala, hard Boker & Co.'s Iala, soft Krupp's metal Driver-Harris Co.'s "S. B." Driver-Harris Co.'s "Advance" . . . . Driver-Harris Co.'s "Ferro-Nickel" . . Constantin 31.726 .000243 34.2 .00019 32.5 19 to 46 .00025 to .00044 42 to 74 .000011 to .00014 50.2 — .000011 47.1 + .000005 85.13 .0007007 55.8 Small 48.8 Very small 28.3 .00207 50 to 52 German Silver. German silver is an alloy of copper, nickel, and zinc. The electrical properties of the alloy naturally vary considerably with the proportions of the constituent metals. The proportion of nickel present is ordinarily used to distinguish the various alloys, as the amount of this metal present in the alloy fixes the proportions of the other constituents in order that the result- ing material may be easily worked. As made in the United States, com- mercial German silver is made with approximately the following propor- tions. (Dr. F. A. C. Perrine.) Designation. Constituents. Resistance at ° C. Per Cent. Alloy. Nickel. Copper. Zinc. Microhms Per Centi- meter. Ohms Per Mil Foot. 8 12.5 20 30 8 12.5 20 30 60 57 56 50 32 30.5 24 20 19 25 32 46 114 150 193 277 Specific gravity, 8.5. Temperature coefficient per ' C, .00025 to .00044. GERMAN SILVER WIRES. 203 Resistances of German Silver Wire at 20- I\ American Gauge. — (American Electrical Works). 18% Alloy. 30% Alloy. Resistance varies .03 of one Resistance varies .022 of one per cent for one degree per cent foi • one degree Centigrade. Centigrade. Size. Ohms per 1000 ft. Ohms per Ohms per Ohms per pound. 1000 ft. pound. No. 8 11.772 . 24702 17.658 .37054 9 14.83 .39249 22.22 . 58873 ' 10 18.72 .62443 28.08 .93666 ' 11 23.598 .99281 35.397 1.4927 ' 12 29.754 1.5785 44.631 2.3676 4 13 37.512 2.5101 56.268 3 . 7650 ' 14 47.304 3.9911 70.956 5.9862 ' 15 59.652 6.3462 89 . 478 9.5192 ' 16 75.222 10.090 112.833 15.135 * 17 94.842 16.045 142.263 24.066 ' 18 119.61 25.511 179.41 38.266 1 19 155.106 42.909 232 . 659 64 . 362 4 20 190.188 64.498 285.282 96.524 * 21 239.814 102.56 359.721 153.84 1 22 302.382 163.06 453.573 244.60 4 23 381.33 259.33 571.99 388.99 4 24 480.834 412.37 721.251 618.55 4 25 606.312 655.61 909 . 468 983 . 43 4 26 764.586 1042.7 1146.879 1563.8 4 27 964.134 1657.7 1446.201 2486.6 4 28 1215.756 2636.0 1823.634 3953.9 4 29 1533.06 4191.5 2299 . 59 6287.2 ' 30 1933.038 6666.5 2899 . 557 9999.6 ' 31 2437.236 10594. 3655 . 854 15890. 4 32 3073 . 77 16850. 4610.65 25275. 4 33 3875.616 26788. 5813.424 40181. 4 34 4888 . 494 42618. 7332.741 63927. 4 35 6163.974 67759. 9245.961 101640. 4 36 7770.816 107700. 11656.224 161540. 4 37 9797.166 171170. 14695.749 256770. 4 38 12357.198 269820 . 18535.797 404740. 4 39 15570.828 428720. 23356.242 643070. " 40 19653.57 682540. 29480.35 1023800. Specific Gravity 8.5 approx. Tffanganiii. Dr. F. A. C. Perrine. Perhaps the most remarkable resistance alloy which has been produced is manganin, invented by Edward Weston in 1889. It is composed of copper, nickel, and ferro-manganese in varying proportions. Prof. Nichols of Cornell, has shown that coils made of this material are apt to change their resistance when successively heated to 100° Cent, and cooled to 0° Cent., but Dr. Lindeck, working for the Reichsanstalt, states that when a completed coil is annealed at a temperature of 140° Cent, for five hours, no further difficulty is experienced from any aging change, whether produced by time or repeated heatings and coolings. A further advantage of manganin which has been noticed by Dr. Lindeck, when used for resistance coils, is its very feeble thermo-electric power when soldered to copper, as is almost always the case in standard coils. While for german silver the thermo-electric power is between 20 and 30 micro- volts per degree Centigrade, and for constantin, an alloy of copper 50 parts with nickel 50 parts, having a temperature coefficient between .00003 and .00004, a thermo-electric power of 40 micro-volts per degree Centigrade is found, the thermo-electric power of manganin is not above one or two micro- volts per degree. 204 PROPERTIES OF CONDUCTORS. Electrical Properties and Constitution of UEangpanin. Dr. F. A. C. Perrine. Mi- Composition. Ohms crohms Temper- ature Co- efficient. * Authority. per Mil- per Cubic Cu. Fe. Mn. Ni. Foot. Centi- meter. Nichols 78.28 14.07 7.65 0.000011 Nichols 51.52 31.27 16.22 0.000039 Perrine 70. 25. 392 65.15 Perrine 65. 30. 5 - Is 404 67.2 Perrine 65. 30. 5.JS3 443 73.6 Feussner and Lindeck 73. 24. 3. 287 47.7 0.00003 Lindeck 84. 12. 4. 253 42.0 0.00014 Dewar and Fleming . 84. 12. 4. 287 47.64 0.0000 Dimensions, Resistance, and Weights of Resistance Wires. Boker & Co.'s IaIa. Specific gravity ' 8.4 Microhms per centimeter cube, 0° C, hard 50.2 Microhms per centimeter cube, 0° C, soft 47 . 1 Microhms per mil-foot, 0° C, hard 310. Microhms per mil-foot, 0° C, soft 284. Temperature coefficient per 0° C, hard -.000011 Temperature coefficient per 0° C, soft + .000005 Carrying B. &. S. Gauge No. Diameter, Inch. Area, Circular, Mils. Ohms per 1000 Feet. Feet per Lb. Approxi- mately. Capacity with Free Radiation Amperes. 14 .0641 4107. 73.5 85. .... 16 .0508 2583. 116.9 135.3 17 .0453 2048. 147.4 170.6 18 .0403 1624. 185.9 215.5 15.8 19 .0359 1289. 234.3 271.0 13.6 20 .0320 1024. 295.6 342.3 11.5 21 .0285 812.3 374.4 433. 9.7 22 .0253 640.1 470.1 543.5 8.0 23 .0225 506.25 596.6 689.6 6.8 24 .0201 404. 747.6 870. 5.8 25 .0179 320.4 945.6 1098. 4.9 26 .0159 252.8 1192.9 1370. 4.1 27 .0142 201.6 1497.8 1724. 3.6 28 .0126 158.8 1890.1 2174. 3.1 29 .0113 127.7 2407.8 2777. 2.9 30 .0100 100. 3005.3 3448. 2.7 31 .0089 79.2 3789 . 2 4347. 32 .0080 64. 4779 . 1 5555. 2.5 33 .0071 50.4 6025 . 1 7142. 34 .0063 39.69 7600.4 9090. 2.2 35 .0056 31.56 9582.7 11100. 36 .005 25. 12081 . 14286. 2.6 37 .0044 19.83 15229. 17543. 38 .004 16. 19213. 22220. 39 .0035 12.25 24218. 27700. 40 .0031 9.61 30570. 35714. Supplied by Boker Co., 101-103 Duane St., New York. borer's resistance ribbon. 205 Resistance Ribbon. la la Quality. of Ohms per 1000 feet. iin. iin. I in. i in. f in. fin. iin. 1 in. .128 14.81 7.40 4.93 3.70 2.96 2.46 2.11 1.85 .114 16.69 8.34 5.56 4.17 3.34 2.78 2.38 2.08 .101 18.80 9.40 6.26 4.70 3.76 3.13 2.70 2.35 .0907 20.97 10.48 6.99 5.24 4.19 3.49 2.99 2.62 .0808 23.46 11.73 7.82 5.86 4.69 3.91 3.35 2.93 .0719 26.63 13.31 8.87 6.65 5.32 4.43 3.80 3.32 .0641 29.62 14.81 9.87 7.40 5.92 4.93 4.22 3.70 .0571 33.38 16.69 11.12 8.34 6.68 5.56 4.77 4.17 .0508 37.60 18.80 12.53 9.40 7.52 6.26 5.37 4.70 .0452 41.94 20.97 13.98 10.48 8.38 6.99 5.99 5.24 .0403 46.92 23.46 15.64 11.73 9.38 7.82 6.70 5.86 .0359 53.26 26.63 17.78 13.31 10.64 8.87 7.60 6.65 .0320 59.24 29.62 19.75 14.81 11.84 9.87 8.46 7.40 .0284 66.76 33.38 22.25 16.69 13.35 11.12 9.53 8.34 .0253 75.20 37.60 25.07 18.80 15.04 12.53 10.74 9.40 .0225 83.88 41.94 27.96 20.97 16.77 13.98 11.96 10.48 .0201 93.84 46.92 31.28 23.46 18.77 15.64 13.40 11.73 .0179 106.52 53.26 35.50 26.63 21.30 17.78 15.21 13.31 .0159 118.48 59.24 39.49 29.62 23.69 19.75 16.91 14.81 .0142 133.52 66.76 44.50 33.38 26.70 22.25 19.07 16.69 .0126 150.40 75.20 50.13 37.60 30.08 25.07 21.50 18.80 .0112 167.76 83.88 55.92 41.94 33.55 27.96 23.96 20.97 .0100 187.68 93.84 62.56 46.92 37.53 31.28 26.81 23.46 .0089 213.04 106.52 71.01 53.26 42.60 35.50 30.43 26.63 .0079 236.96 118.48 78.98 59.24 47.40 39.49 33.82 29.62 .0071 267.04 133 . 52 89.01 66.76 53.40 44.50 38.15 33.38 .0063 300.80 150.40 100.26 75.20 60.16 50.13 42.97 37.60 .0056 335.52 167.76 111.84 83.88 67.10 55.92 47.93 41.94 .005 375.36 187.68 125.12 93.84 75.07 62.56 53.62 46.92 .0044 426.08 213.04 142.02 106.52 85.21 71.01 60.87 53.26 .004 473 . 92 236.96 157.97 118.48 94.78 78.98 67.64 59.24 206 PROPERTIES OF CONDUCTORS. Krupp's Resistance Wires. Specific gravity 8.102. Specific resistance at 20° C. mean 85 . 13 microhms. Temperature coefficient, mean .0007007. Resistance per circular mil-foot 314. ohms. Resistance per 1000', 1 square inch area ... .8513 ohms. This metal can be permanently loaded with current sufficient to raise its temperature to 600° C. (1112° F.) without undergoing any structural change. It should never be put in contact with asbestos, however, as this material causes it to deteriorate rapidly. Diam. Diam. in inches. Near- est B. &S. Gauge Feet per lb. Resistance in ohms per foot. in m.m. at at at at No. 68° F. 176° F. 284° F. 428° F. 5 .1968 4 9 .0132 .0138 .0143 .0150 4* .1772 5 12 .0163 .0170 .0176 .0184 4 .1575 6 15 .0206 .0215 .0224 .0235 3* .1378 7 19 .0269 .0280 .0291 .0307 3 .1181 9+ 26 .0368 .0382 .0396 .0417 2| .1083 9— 31 .0437 .0455 .0472 .0497 2* .0984 10 37 .0528 .0550 .0570 .0601 H .0885 11 46 .0653 .0679 .0705 .0742 2 .0787 12 58 .0825 .0860 .0892 .0940 If .0689 13 76 .1078 .112 .116 .123 li .0590 15 104 .1468 .153 .159 .167 n .0492 16 150 .2115 .220 .229 .241 i .0393 18 234 .3305 .344 .356 .376 i .0295 21 415 .5870 .610 .633 .667 i .0196 24 937 1.324 1.38 1.43 1.51 American Agent, Thomas Prosser & Son, 15 Gold St., New York City. Resistance Wires Made by Driver-Harris Wire Co., Harrison, A . T. 11 S. B." — Resistance per mil-foot at 75° F. Low temperature coefficient and low thermo-electric effect against copper. Will not rust. 'Advance." — Resistance per mil-foot at 75° F. A copper-nickel alloy containing no zinc. Temperature coefficient practically nil. "Ferro-Nickel." — Resistance per mil-foot at 75° F. Temperature coefficient per ° F. About the same resistance as German Silver, but weighs about ten per cent less and is cheaper. 336 ohms 294 ohms 170 ohms .00115 DRIVER-HARRIS RESISTANCE WIRES. 207 Resistances of Driver-Harris Resistance Wires. "S. B." "Advance." "Ferro-Nickel." No. B. & S. Ohms per Ohms per Ohms per 1,000 ft. 1,000 ft. 1,000 ft. 10 32 28. 2.0 11 40 35.5 2.5 12 51 44.8 3.2 13 64 56.7 4.1 14 82 71.7 5.1 15 103 90.4 6.5 16 130 113 8.2 17 168 145 10.4 18 210 184 13.1 19 260 226 16.3 20 328 287 20.5 21 415 362 25.9 22 525 460 32.7 23 660 575 41.5 24 831 725 52.3 25 1,050 919 65.4 26 1,328 1,162 85 27 1,667 1,455 106 28 2,112 1,850 131 29 2,625 2,300 166 30 3,360 2,940 209 31 4,250 3,680 266 32 5,250 4,600 333 33 6,660 5,830 425 34 8,400 7,400 531 35 10,700 9,360 672 36 13,440 11,760 850 37 16,640 14,550 1,070 38 21,000 18,375 1,330 39 27,540 24,100 1,700 40 37,300 32,660 2,120 208 PROPERTIES OF CONDUCTORS. CURRENT CARRYING CAPACITY Or Willie A3TI> CABLES. Let D — diameter of wire or cable core in inches. T = temperature elevation of wire or cable core in ° Centigrade. / = current in wire in amperes. r — specific resistance of wire in ohms per mil-foot at final tem- perature. The following approximate formulae give results sufficiently accurate for practical purposes. Bare Overhead Wires Out of Doors. Stranded : Solid : /-nooy/lp. / = i250 V / ^ 3 - 3 Wires In Doors, Expose] ed: S y/™_ 3 . / = 660 Jl >r Rubber Covered Cable : ed :__ Solid : i/™ 3 . / - 530 yfe Bare Wires In Doors, Exposed. Stranded : Solid : 610 ■* * Single Conductor Rubber Covered Cable in Still Air. Stranded : Solid : j = 49Q a i/TD* Single Conductor Rubber Covered Lead Sheathed Cable in Underground Single Duct Conduit. Stranded : Solid : / = 490 l/M 7 - 530 \JIf. Single Conductor Paper Covered Lead Sheathed Cable in Underground Single Duct Conduit. Stranded : Solid : / = 430 yfe 1 = 470 y^ 3 . * Three-Conductor Rubber Covered Lead Sheathed Cable in Underground Single Duct Conduit. Stranded : Solid : / = 370 \/lf- I = 400 jlf. * Three-Conductor Paper Covered Lead Sheathed Cable in Underground Single Duct Conduit. Stranded : Solid : 7 - 320 V& / - 350 t/*l!. * / is here current per wire. CAPACITY OF WIRES AND CABLES. 209 Carrying* Capacity of Insulated Copper Wires for Interior Wiring*. National Electrical Code. B. &S. Co. 18 16 14 12 10 8 6 5 4 3 2 1 00 000 0000 Circular Mils. 1,624 2,583 4,107 6,530 10,380 16,510 26,250 33,100 41,740 52,630 66,370 83,690 105,500 133,100 167,800 211,600 Rubber Weather Covered proof Wires. Wires. Am- Am- peres. peres. 3 5 6 8 12 16 17 23 24 32 33 46 46 65 54 77 65 92 76 110 90 131 107 156 127 185 150 220 177 262 210 312 Circular Mils. 200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000 1,100,000 1,200,000 1,300,000 1,400,000 1,500,000 1,600,000 1,700,000 1,800,000 1,900,000 2,000,000 Rubber Covered Wires. Amperes. 200 270 330 390 450 500 550 600 650 690 730 770 810 850 890 930 970 1,010 1,050 Weather- proof Wires. Amperes. 300 400 500 590 680 760 840 920 1,000 1,080 1,150 1,220 1 290 1,360 1,430 1,490 1,550 1,610 1,670 Carrying* Capacity of Stranded Copper Conductors for Interior IViring*. National Electrical Code. B. d _ q n Area Actual z S * G * C. M. No. of Strands. Size of g B. &£ trand A i q Amperes. L9 1,288 L8 1,624 . L7 2,048 L6 2,583 6* ] L5 3,257 ] L4 4,107 i2 ] L2 6,530 17 9,016 *7 is ) 21 11,368 7 u I 25 14,336 7 Vt r 30 18,081 7 le > 35 22,799 7 u > 40 30,856 19 u \ 50 38,912 19 17 60 49,077 19 16 1 70 60,088 37 It 1 ' 85 75,776 37 17 100 99,064 61 IS 120 124,928 61 17 145 157,563 61 ie ► 170 198,677 61 15 200 250,527 61 14 235 296,387 91 IS 270 373,737 91 14 320 413,639 127 IS 340 For aluminum wire the carrying capacity of any given size is to be taken as 84 per cent of the value given in the above table. 210 PROPERTIES OF CONDUCTORS. Carrying* Capacity of Rubber Insulated Cables. (From technical letter of General Electric Company.) The following table of carrying capacity is based on tests of cables in still air. Insulation alone &" thick; lead j\" to I" thick; jute and asphalt jacket &" thick. Paper insulated cables heat 8% to 10% more than rubber insulated cables with same current and thickness of coverings. Cables require about four hours to reach final temperature. 60% of total increase in temperature in 1st hour. 30% of total increase in temperature in 2d hour. 8% of total increase in temperature in 3d hour. Cables immersed in water will carry 50% more current with same increase of temperature, and cables buried in moist earth about 15% more. Rubber cables should not be run above 70° C. Paper cables should not be run above 90° c. Amperes at 30° C. Amperes at 50° C Rise. Rise. Diameter Copper Core. Inches. Size. Leaded Leaded Braided. and Jute Braided. and Jute Covered. Covered. 6 B. & S. Solid .162 61 56 76 68 4 B. & S. Solid .204 85 78 104 94 2 B. & S. Stranded .300 133 121 162 146 1 B. & S. Stranded .325 155 141 189 170 B. & S. Stranded .390 191 174 231 210 00 B. & S. Stranded .420 218 199 268 241 000 B. & S. Stranded .475 266 242 325 293 0000 B. & S. Stranded .543 320 291 391 352 250000 CM. .570 355 324 435 392 300000 CM. .640 414 377 506 456 350000 CM. .680 460 419 563 507 400000 CM. .735 512 466 626 564 450000 CM. .787 562 511 687 618 500000 CM. .820 606 551 742 668 600000 CM. .900 694 631 848 763 750000 CM. 1.020 825 750 1016 915 900000 CM. 1.096 940 855 1149 1034 1000000 CM. 1.157 1017 925 1333 1200 1250000 CM. 1.298 1204 1095 1481 1328 1500000 CM. 1.413 1376 1251 1644 1480 2000000 CM. 1.760 1766 1606 2178 1960 Heating: of Cables in Multiple Duct Conduit. The mutual heating of cables in multiple duct conduit has been inves- tigated experimentally by H. W. Fisher, The following diagram and table shows the arrangement of the conduit system used by him and the size and kind of cable in each duct. Means were provided for connecting any or all the cables in series and observing the temperature of the con- ductor in each duct. CAPACITY OF WIRES AND CABLES. 211 © © © © © © © © © © © © Fig. 22. Number of Size B. & S. and Cable. Conductors. C. M. Insulation. A* .... . 000 fa" and fa" Paper B 1 500,000 fa" Paper C* 000 fa" and fa" Paper D 500,000 fa" Paper E 1,250,000 fa" Paper F 1,250,000 fa" Paper G 000 fa" Paper H 000 Rubber I 1,250,000 fa" Paper J 1,250,000 fa" Paper K ..... 000 Rubber L 000 ■fa" Paper * The three conductors of A and C in multiple. Fisher's results are summarized in the following table : Conductors 30° C. Rise. 50° C. Rise. Carrying Current. Conductor. Amperes. Conductor. Amperes. All A. &C. (l ( I ) E ' F f B D 130 155 180 600 590 560 535 355 400 A. &C. G L I E J F B D 180 G, H, K 5 L E, F, I, J A, B, C, D, E, F, I, J. ... 190 260 765 750 725 690 425 550 An inspection of this table will show that the current corresponding to a given temperature elevation is in each case less than that given by the formulae on page 208, the difference being from 4 to 25 per cent, depend- ing on the number of conductors in service and the location of the cable in question. It is to be noted that corner ducts radiate heat the best, and all outside ducts radiate heat much better than do the inside ducts. 212 PROPERTIES OF CONDUCTORS. Watt* per foot l,o*t in Single-Conductor Cables at Different Maximum Temperature with Different A mounts of Currents. (From Handbook No. XVII, 1906. Copyrighted by Standard Under- ground Cable Company.) Size B. & S. Current in Amperes. 6 66 81 93 104 114 123 5 74 91 105 117 128 138 4 84 102 117 131 144 153 3 93 114 132 148 161 175 2 105 128 148 166 181 196 1 118 148 166 186 203 220 132 162 187 209 228 247 00 149 181 210 235 256 277 000 166 204 235 263 288 311 0090 186 229 264 295 323 350 Area in 1000 C. M. 300 222 273 315 352 385 416 400 248 315 363 406 445 480 500 288 352 406 455 498 537 600 315 385 445 497 545 587 700 341 416 480 538 588 635 800 364 446 514 575 628 679 900 386 473 545 610 666 720 1000 407 498 575 642 703 758 1100 426 522 602 674 736 796 1200 446 546 630 705 772 833 1300 462 568 655 732 802 866 1400 480 590 681 761 834 900 1500 496 610 704 788 862 931 1600 512 629 726 812 889 960 1700 529 649 750 837 916 990 1800 543 667 770 862 943 1018 1900 557 686 792 886 970 1048 2000 573 705 813 910 995 1075 Temp. (100 of cond. 1 125 in°F. (150 Watts lost per ft. 1.81 1.91 2.00 2.71 2.87 3.00 3.62 3.82 4.00 4.52 4.78 5.00 5.43 5.73 6.00 6.33 6.69 7.00 The watts lost per foot means the amount of electric energy lost in heat- ing the conductor and is equal to the product of the resistance per foot of cable times the square of the current in amperes. The above table is useful in showing the watts lost in heating effect per foot of cable with different currents, and also in finding the size of con- ductor that must be used for a given current and watts per foot loss. For Two-Conductor Cables the watts corresponding to the dif- ferent currents must be multiplied by two, and to obtain the currents corresponding to the watts in the table multiply the currents given in the table by .707. For Threes Conductor Caliles the watts corresponding to the currents in the table, must be multiplied by 3, and to obtain the currents corresponding to the watts in the table multiply the currents given in the table by .577. CAPACITY OF WIRES AND CABLES. 213 Current Carrying* Capacity of Lead Covered Cables. (From Handbook No. XVII, 1906. Copyrighted by Standard Under- ground Cable Company.) The current carrying capacity of insulated copper cables sheathed with lead depends primarily upon (a) The size and number of conductors and their relative position. (6) The ability of the insulating material to withstand high tempera- tures and to conduct heat away from the copper conductor, — this latter being in turn dependent upon kind of insulation and its thickness. (c) The initial temperature of the medium surrounding the cable. (d) The ability of the medium surrounding the cable to dissipate heat with small temperature rise. (e) The number of operating cables in close proximity and their relative positions. Where a number of insulated conductors are under the same sheath, they are subject to an interchange of heat somewhat similar to that which takes place when a number of separate cables are laid closely together, and for that reason each conductor of a multi-conductor cable will have a smaller current carrying capacity than a single-conductor cable. If the various conductors are separately insulated and laid together in the form of flat or round duplex or triplex, their carrying capacity will be greater than if they are laid up in the form of two-conductor concentric or three- conductor concentric, since the enveloping conductors in the latter forma- tion seriously retard the dissipation of heat from the inner conductors. Assuming that unity (1.00) represents the carrying capacity of single- conductor cables, the capacity of multi-conductor cables would be given by the following: 2 cond. flat or round form, 3 cond. triplex form .87; concentric form, .75; concentric form, .79 .60 The following experiment on duplex concentric cable of 525,000 C. M. indicates clearly the danger in subjecting this type of cable to heavy over- loads of even short duration. The cable was first heated up by a current of 440 amperes for 5 hours. An overload of 50 per cent w T as then applied, the results in degrees Fahrenheit above the surrounding air being as follows: Time from Start. Min. 15 Min. 30 Min. 45 Min. 60 Min. 90 Min. Inner Conductor . Outer Conductor . Lead Cover . . . 70° 55 31 84° 65 35 98° 76 40 111° 85 45 123° 94 49 142° 108 57 In any cable the area over which dissipation of heat must take place is proportional to the circumference of the conductor or (since the circum- ference varies as the diameter), upon the diameter of the conductor, while the cross section of the conductor varies as the square of the diameter. Hence the size of conductor varies much more rapidly than its heat radiat- ing surface, and in consequence the amperage per square inch, or circular mil of copper section, must be less for large size conductors than for small, in order to have the same rise of temperature under the same conditions. The usual formula for carrying capacity, Current = 3 (diam. of Cond.) A constant takes account of this fact but not to a sufficient degree, and we find that for cables as ordinarily used in underground work, a more correct expression is the following: ~ (diam. of Cond.)^ Current = - — -~- A constant 214 PROPERTIES OF CONDUCTORS. Rubber insulation is a somewhat better heat conductor than dry or saturated paper, and therefore, when applied to the same size conductor in equal thickness, will permit of a larger current flowing in the conductor for the same rise of temperature above the surrounding air. On the other hand, rubber deteriorates much more rapidly at high temperatures than saturated paper, and while this disadvantage is apparently compensated for up to about 150° Fahrenheit by its superior heat dissipating qualities, at higher temperatures deterioration takes place and becomes so serious that its value as an insulating medium disappears in a comparatively short time. As the thickness of insulation is increased, the temperature of the con- ductor, with any given current flowing gradually, increases and therefore the current carrying capacity becomes reduced. The reduction in capacity however, is not very great, being in the ratio of about 93 for §f insulation to 100 for & insulation, so that the values in the table given below should be slightly decreased when greater thicknesses than ■& are used. As it is the final temperature reached which really affects the carrying capacity, the initial temperature of surrounding medium must be taken into account. If, for instance, the conduit system parallels steam or hot water mains, the temperature of 150° F. (which we have assumed in the table on page 215 to be the maximum for safe continuous work on cables) will be reached with lower values of current than would otherwise be the case; and as 70° is the actual temperature we have assumed to exist in the surrounding medium prior to loading the cables, any increase over 70° must be compensated for by reducing the current carried. For rough calculations it will be safe to use the following multipliers to reduce the current carrying capacity given in the table on page 215 to the proper value for the corresponding initial temperatures: Initial Temp. , 70 80 90 100 110 120 130 140 15C Multipliers . . 1.00 .93 .86 .78 .70 .60 .48 .34 .OC The ability of the surrounding medium to dissipate heat, directly affects the carrying capacity of the cables, as with the same current the cable might be comparatively cool if laid in good heat conducting material such as water, and dangerously hot if laid in poor heat conduct- ing material such as dry sand. Ordinary conduit systems of clay or terra cotta ducts laid in cement, dissipate heat fairly well, the outside ducts, however, being much more efficient in this function than the inner ones, so that an ideal system, from this point of view, would consist of a single horizontal layer of ducts. _ As this would require an enormous width of trench and considerable inconvenience in handling the cables in manholes when many cables are to be installed, we would suggest the form shown in Fig. 23 as being more practicable. Where more ducts are required, the vertical section shown could be easily duplicated, a considerable space, however, being left between them. With this arrangement, the carry- ing capacities given in the table on p. 215 could be somewhat increased. When a number of loaded cables are operating in close proximity to one another, the heat from one radiates, or is carried by conduction, to each of the others, and all raised in temperature beyond what would have resulted had only a single cable been in operation; and if the cables occupy adjacent ducts in a conduit system of approxi- mately square cross section laid in the usual way, the centrally located cable or the one just above the center in large installations (A in Fig. 24) will reach the highest temperature. This is equiv- alent to saying that its carrying capacity is reduced, and while this reduction does not amount to more than about 12 per cent (as compared with the cable most favorably located, — as at Z), Fig. 24) in the duct arrangement given, it may easily assume much greater proportions where large numbers of cables are massed together. Fig. 23. Fig. 24. CAPACITY OP WIRES AND CABLES. 215 Assuming that not more than twelve cables, arranged as shown in Fig. 24, can be used, the average carrying capacity may be taken as the crite- rion for proper size of conductor; and for cables of a given type and size the carrying capacities of all cables, even though placed in adjacent ducts, will be represented by the following figures, taking unity as the average carrying capacity of four cables: No. Cables 2 Multiplier 1.16 4 6 8 10 12 1.00 .88 .79 .71 .63 Recommended Current Carrying* Capacities for Cables and Watts JLost per foot. For each of four equally loaded single conductor paper insulated lead covered cables, installed in adjacent ducts in the usual type of conduit system where the initial temperature does not exceed 70° F., the maximum safe temperature for continuous operation being taken at 150° F. (From Handbook No. XVII, 1906. Copyrighted by Standard Under- ground Cable Company.) Size B. & S. G. Safe Cur- rent in Amperes. Watts * lost per ft. at 150° F. Size CM. Safe Cur- rent in Amperes. Watts * lost per ft. at 150° F. 14 18 .97 300,000 323 4.22 13 21 1.03 400,000 390 4.61 12 24 1.09 500,000 450 4.91 11 29 1.15 600,000 505 5.16 10 33 1.25 700,000 558 5.36 9 38 1.39 800,000 607 5.56 8 45 1.53 900,000 650 5.71 7 53 1.67 1,000,000 695 5.86 6 64 1.85 1,100,000 740 6.01 5 76 2.08 1,200,000 780 6.13 4 91 2.31 1,300.000 820 6.25 3 108 2.54 1,400,000 857 6.37 2 125 2.77 1,500,000 895 6.49 1 146 3.00 1,600,000 933 6.61 168 3.23 1,700,000 970 6.73 00 195 3.46 1,800,000 1010 6.85 000 225 3.69 1,900.000 1045 6.97 0000 260 3.92 2,000,000 1085 7.09 * This column represents the amount of energy which is transformed into heat and which must be dissipated. It is what is usually called the PR loss and it is figured by using for / the current values given ; and for R the resistance of the respective conductor at a temperature of 150° F. Note. — The table is compiled from a long series of tests made by us in conjunction with the Niagara Falls Power Company, the conduit system being of the type shown in Fig. 24. The ducts were of terra cotta with 3-inch openings. 216 PROPERTIES OF CONDUCTORS. Recommended Power Carrying* Capacity in Kilowatt* of .Delivered Energy , Three-Conductor, Three-Phase Cables. (From Handbook No. XVII, 1906. Copyrighted by Standard Under- ground Cable Company.) Volts Size in B. &S.G. 1100 2200 3300 4000 6600 j 11000 13200 | 22000 Kilowatts. 6 92 183 275 333 549 915 1098 1831 5 109 217 326 395 652 1087 1304 2174 4 130 260 390 473 781 1301 1562 2603 3 154 309 463 562 927 1544 1854 3089 2 179 358 536 650 1073 1788 2145 3575 1 209 418 626 759 1253 2088 2506 4176 240 481 721 874 1442 2402 2884 4805 00 279 558 836 1014 1674 2788 3347 5577 000 322 644 965 1172 1931 3217 3862 6435 0000 372 744 1115 1352 2231 3717 4462 7435 250000 413 827 1240 1503 24S0 4132 4960 8264 Single Conductor Cables , A. C . or D . C. Volts. Size in B.&S.G. 125 250 500 1100 2200 3300 6600 11000 Kilowatts. 6 8.0 16.0 32 70 141 211 422 704 5 9.5 19.0 38 84 167 251 502 836 4 11.4 22.8 45 100 200 300 601 1001 3 13.5 27.0 54 119 238 356 713 1188 2 15.6 31.2 62 138 275 413 825 1375 1 18.3 36.5 73 161 321 482 964 1606 21.0 42.0 84 185 370 554 1109 1848 00 24.4 48.8 97 215 429 644 1287 2145 000 28.1 56.3 113 248 495 743 1485 2475 0000 32.5 65.0 130 286 572 858 1716 2860 300000 40.4 80.8 162 355 711 1066 2132 3553 400000 48.8 97.5 195 429 858 1287 2574 4290 500000 56.3 112.5 225 495 990 1485 2970 4950 600000 63.1 126.3 253 556 1111 1667 3333 5555 700000 69.8 139.5 279 614 1228 1841 3683 6138 800000 75.9 151.8 304 668 1335 2003 4006 6677 900000 81.3 162.5 325 715 1430 2145 4290 7150 1000000 86.9 173.8 348 764 1529 2294 4587 7645 1100000 92.5 185.0 370 814 1628 2442 4884 8140 1200000 97.5 195.0 390 858 1716 2574 5148 8580 1400000 107.1 214.3 429 943 1885 2828 5656 9427 1500000 111.9 223 . 8 448 985 1969 2954 5907 9845 1600000 116.6 233 3 467 1026 2053 3079 6158 10263 1700000 121.3 242.5 485 1067 2134 3201 6402 10670 1800000 126.3 252.5 505 1111 2222 3333 6666 11110 2000000 135.6 271.3 543 1194 2387 3581 7161 11935 These tables are based on the recommended current carrying capacity of cables given on page 215. A power factor = 1, was used in the calcula- tion and hence the values found in the last table are correct for direct currents. For alternating currents the kilowatts given in both tables must be multiplied by the power factor of the delivered load. FUSING EFFECTS OF ELECTRIC CURRENTS. 217 ftjsiuto effects of electric currents. By W. H. Preece, F. R. S. See " Proc. Roy. Soc," vol. xliv., March 15, 1888. The Law — /= ad 2 , where /, current ; a, constant ; and d, diameter — is strictly followed; and the following are the linal values of the constant "a," for the different metals as determined by Mr. Preece : — Inches. Centimeters. Millimeters. Copper 10,244 2,530 80.0 Aluminum .... 7,585 1,873 59.2 Platinum 5,172 1,277 40.4 German Silver. . . . 5,230 1,292 40.8 Platinoid 4,750 1,173 37.1 Iron ....... 3,148 777.4 24.6 ^i n 1 (342 405.5 12.8 Alloy (lead and tin 2tol) l',318 325.5 10.3 Lead 1,379 340.6 10.8 Table Giving- tne Diameters of Wires of Various materi- als Wnicn Will Be Fused by a Current of Given Strength.— W. H. Preece, F. R. S. d= (LV /3 Diameter in Inches. GO - — .4 s . in 3 e *4 02^ 'do "SB 00 CO I rlo6 *£ 111 03 Q s rill 511 5 £ 111 ^ 8 1 0.0021 0.0026 0.0033 0.0033 0.0035 0.0047 0.0072 0.0083 0.0081 2 0.0034 0.0041 0.0053 0.0053 0.0056 0.0074 0.0113 0.0132 0.0128 3 0.0044 0.0054 0.0070 0.0069 0.0074 0.0097 0.0149 0.0173 0.0168 4 0.0053 0.0065 0.0084 0.0084 0.0089 0.0117 0.0181 0.0210 0.0203 5 0.0062 0.0076 0-0098 0.0097 0.0104 0.0136 0.0210 0.0243 0.0236 10 0.0098 0.0120 0.0155 0.0154 0.0164 0.0216 0.0334 0.0386 0.0375 15 0.0129 0.0158 0.0203 0.0202 0.0215 0.0283 0.0437 0.0506 0.0491 20 0.0156 0.0191 0.0246 0.0245 0.0261 0.0343 0.0529 0.0613 0.0595 25 0.0181 0.0222 0.0286 0.0284 0.0303 0.0398 0.0614 0.0711 0.0690 30 0.0205 0.0250 0.0323 0.0320 0.0342 0.0450 0.0694 0.0803 0.0779 35 0.0227 0.0277 0.0358 0.0356 0.0379 0.0498 0.0769 0.0890 0.0864 40 0.0248 0.0303 0.0391 0.0388 0.0414 0.0545 0.0840 0.0973 0.0944 45 0.0268 0.0328 0.0423 0.0420 0.0448 0.0589 0.0909 0.1052 0.1021 50 0.0288 0.0352 0.0454 0.0450 0.0480 0.0632 0.0975 0.1129 0.1095 60 0.0325 0.0397 0.0513 0.0509 0.0542 0.0714 0.1101 0.1275 0.1237 70 0.0360 0.CM40 0.0568 0.0564 0.0601 0.0791 0.1220 0.1413 0.1371 80 0.0394 0.0481 0.0621 0.0616 0.0657 0.0864 0.1334 0.1544 0.1499 90 0.0426 0.0520 0.0672 0.0667 0.0711 0.0935 0.1443 0.1671 0.1621 100. 0.0457 0.0558 0.0720 0.0715 0.0762 0.1003 0.1548 0.1792 0.1739 120 0.0516 0.0630 0.0814 0.0808 0.0861 0.1133 0.1748 0.2024 0.1964 140 0.0572 0.0698 0.0902 0.0895 0.0954 0.1255 0.1937 0.2243 0.2176 160 0.0625 0.0763 0.0986 0.0978 0.1043 0.1372 0.2118 0.2452 0.2379 180 0.0676 0.0826 0.1066 0.1058 0.1128 0.1484 0.2291 0.2652 0.2573 200 0.0725 0.0886 0.1144 0.1135 0.1210 0.1592 0.2457 0.2845 0.2760 225 0.0784 0.0958 0.1237 0.1228 0.1309 0.1722 0.2658 0.3077 0.2986 250 0.0841 0.1028 0.1327 0.1317 0.1404 0.1848 0.2851 0.3301 0.3203 275 0.0897 0.1095 0.1414 0.1404 0.1497 0.1969 0.3038 0.3518 0.3413 300 0.0950 0.1161 0.1498 0.1487 0.1586 0.2086 0.3220 0.3728 0.3617 218 PROPERTIES OF CONDUCTORS. Ti:\MO\ 4VD SAO IJ¥ WIHE SPAM. By Harold Pender, Ph.D. The accompanying charts* (No. 1 for long spans, No. 2 for short spans) enable one to determine without arithmetical computation the variation of the tension and sag in copper wire spans with the temperature and resul- tant load on the wire. Similar charts can be readily prepared for wires of any material. The symbols used in the discussion below are as follows: m = weight of wire per cubic inch in pounds. a = coefficient of linear expansion of wire per degree Faljr. M = modulus of elasticity of wire (pounds — square inch). p = ratio of resultant of the weight of wire, the weight of sleet and the wind pressure to the weight of wire. I = length of span in feet. t = rise in temperature in degrees Fahr. T — tension in thousands of pounds per square inch. D = deflection at center of span in feet in direction of resultant force when points of suspension are on the same level. S — vertical sag at center of span in feet when points of support are on the same level. The lines on the charts are plotted as follows: The hyperbolic curves on the right have the equation y = (~J where y is the ordinate and T the abscissa. A curve is plotted for p = 1.0, 1.2, 1.4 . . . 4.0. The value of p for each curve is indicated at the top of the chart. It is to be noted that the horizontal distance between these curves at any level is directly proportional to the increment in the value of p. These curves are independent of the material of the wire. 10° The inclined straight lines have the equation y = ^-r^ — — T. For a given material the equation of these lines depends only on the length of the span. The lines on the charts are drawn for copper wire for which m = 0.321 and M = 12 X 10 6 . The corresponding length of span is indicated on the right-hand margin of the charts. For any other material, the line for a given length of span will have a different slope. The temperature scale on the X axis to the right of the origin is laid off so that x = Ma t. The scale given on the chart is for copper, for which M = 12 X 10 6 and o = 9.6 X 10"*. This scale will be different for any other material. The parabolic curves on the left of the chart have the equation D = 0.0015 m I 2 Vy, where D is measured off from the left of the origin. For a given material these curves are fixed by the length of the span. The curves given on the chart are for copper, for which m = 0.321. The correspond- ing lengths of span are indicated on the curves. These curves will be different for any other material. Rules for the "Use of the Charts. Given: A span of length I and the points of support on the same level, tension I\; ratio of resultant force to weight of wire, pi; to find the tension T when the temperature rises t degrees and the ratio of resultant force to weight of wire changes to p (for example, sleet melts off). At the point 1 (Fig. 27) on the curve corresponding to pi and having the abscissa 7\ f lay off 12 = the ordinate of the point 3 on the line corre- sponding to I having the abscissa t on the temperature scale. * These charts were devised to obtain a graphical solution of the equa- tions deduced by the author in an article in the Electrical World for Jan. 12, 1907, Vol. 49, p. 99. The present article also appeared in the Electrical World for Sept. 28, 1907. WIRE SPANS. 219 Through 2 draw a line parallel to the line I : the abscissa of the point 4 where this line cuts the curve corresponding to p is the tension T at the temperature t when the ratio of resultant force to weight of wire is p. The abscissa of the point 5 where the horizontal line through 4 cuts the para- bolic curve corresponding to I gives the corresponding deflection D at the center of the span in feet. Instead of actually drawing the straight line 24, a pair of compasses may be used; i.e., lay off the distance 12, then open the compasses until the lower point touches the straight line Z; then keep- ing the compasses vertical, slide the lower point along I until the upper point intersects the curve corresponding to p. If t is negative, i.e., if the temperature decreases, lay off 12 in the opposite direction. To determine D with greater accuracy use the formula D = .0015 ml? 9 -' Fig. 25. Calculation of p. Let w = weight of wire in pounds per foot. The weight of sleet (and hemp core, if any) in pounds per foot of wire is Wt 0.314 (dj - d 2 ) + 0.25 do 2 , where d is the diameter of the wire, d\ the diameter over sleet and do the diameter of the core, all in inches. The wind pressure in pounds per foot of wire is * w 2 = 0.00021 V 2 d u where V is the actual wind velocity in miles per hour; d x = d in case of no sleet. The relation between indicated wind velocity (as given by U. S. Weather Reports) and actual velocity is as follows: Indicated Velocity. Actual Velocity. 0.00021 VK 10 9.6 0.0194 20 17.8 0.0667 30 25.7 0.139 40 33.3 0.233 50 40.8 0.350 60 48.0 0.485 70 55.2 0.640 80 62.2 0.812 90 69.2 1.01 100 76.2 1.22 • The ratio p is then P -/c 4- wA 2 /VA 2 0)/ \(0 / * H. W. Buck in Transactions International Electrical Congress, 1904. 220 PROPERTIES OF CONDUCTORS. WIRE SPANS. 221 222 PROPERTIES OF CONDUCTORS. Calculation of Vertical Sag 1 . In case of no wind the vertical sag S is the same as the deflection D. The wind pressure gives a horizontal component to the resultant force so that the vertical sag when wind is blowing is, S - •■ ♦ tw Example: A No. 00 stranded copper cable is to be strung in still air at 70° F. between two points on the same level 800 feet apart, so that at a temperature of zero degrees Fahrenheit, with a coating of sleet 0.41 inch thick all around and wind blowing perpendicularly to the cable at 69.5 miles an hour (actual velocity) the tension in the cable will be 30,000 lbs. per sq. in.; (1) at what tension must the cable be strung and (2) what will be the vertical sag at stringing temperature, i.e., 70°, also (3) what will be the sag at zero temperature when the cable is coated with i-in. of sleet and wind is blowing with a velocity of 65 miles an hour, and (4) what will be the sag at the temperature of 150°, in the still air? We have w= 0.406 w 1 =0.314 (1.2382 -0.418 2 ) =0.425 w 2 = .00021 + 69.52 + 1 . 238 = 1 . 26. Therefore, at zero degrees with wind and sleet, (1) Measure off with compasses, on chart No. 1, the vertical distance from t = 70 on X axis to the straight line corresponding to I = 800. Lay this distance off vertically above the point on the curve corresponding to p = 3.72 having the abscissa T = 30. Keep the upper point fixed, open the compasses until the lower point touches the line I = 800; then, keeping the compasses vertical, slide the lower point along the line I = 800 until the upper point intersects the curve p = 1 at T = 8.95; the cable must therefore be strung at a tension of 8950 lbs. per sq. in. (2) The abscissa of the point on the parabolic curve I = 800, having the same ordinate as the point corresponding to p = 1 and T = 8.95 is D = 34.4 feet, which is the vertical sag S, in still air at 70° F. (3) The deflection at zero degrees with sleet and wind is the abscissa of the point on the parabolic curve I = 800 having the same ordinate as the point corresponding to p = 3.72 and T = 30, i.e., D = 38.2 feet. The vertical sag is = 21.(Heet. ^ * git (4) To find the sag at 150° proceed as under (1) and (2) taking t = 150. The sag will be found to be S = 36.8 feet. WIRE SPANS. 223 Were Suspended from Points not on the Sanie Level. The charts also apply directly to the determination of the change in tension in spans when the points of support are at different heights. In this case, however, the vertical sag Si ( — deflection in case of no wind) below the highest point of support is given by the formula S ^ s ( l + Ts)" where h is the difference in height of the two points of support, and S is the vertical sag for a span of equal length, but points of support on the same level; S is calculated by the formula given above, i.e., D J, + (_a_y D being the deflection, taken directly from the chart, for a span of equal length but points of support on the same level; in case of no wind S = D. The distance of the point of maximum sag from the highest point of support is 2 V + 4 Sj When h is greater than 4 S the lowest point of support is the point of max- imum sag, i.e., the lowest point in the span. Example: In the example given above, suppose the difference in height of the points of support is 20 feet : Then (1) the tension at 70° will still be 8950 lbs. per sq. in. (2) The corresponding vertical sag at 70° in still air for points of support at same level is 34.4 ft., therefore, for the span under consideration the vertical sag from the highest point of support is (20 \ 2 (3) The vertical sag at zero degrees with sleet and wind for points of support on the same level is 21 ft.; therefore, for a 20-ft. difference in the height of points of support the vertical sag from the highest point of sup- port is 21 ( 1+ 4f2i) 2 = 321ft - (4) The vertical sag at a temperature of 150°, for points of support on the same level is 36.8 ft.; therefore, for a 20-ft. difference in height of the points of support the vertical sag from the highest point of support is 36 - 8 ( 1 + orW= 47 - 5ft - The accompanying table, giving the value of T and p for various values of y = f J;) will be found useful in plotting the hyperbolic curves in case one wishes to make charts on a larger scale than those given herein, or similar charts for wires having different constants. The other lines are readily plotted from the equations given above. 224 PROPERTIES OF CONDUCTORS. <^IN n **co cocon ©oco 00 ooo ox CMt- i— i i—i 00 CO iOCMiO (MOO OCOO OOH i-hn© 1O1-1N i-l CM-* OOrH (MO XO© i— i cm^o i-l coo CO CO •ocox ONOJ 00hO COO© OOiOO XXO t-hcooo CMCNCM ©CM© CO CO CO OOO OOiO iOOO CNOX© ONN o^co oOi-ith cm coco oo co oo CO^*'* COt- CM iO»OCO 00(DN tDNOO OLO CM CO COOt- T-Hl-00 NOO HHO nco TfHCOlO ©CM© NI>X OINO X(MTt< tHOCO NOOC0 C0^Tt( XCOt-i t^ic CO iOt- l-O O CM CM CM OOHCD (MCOCO >«*cM "*CO CM 00 i-i CO OOCO OSTt^O NiO© (MOO OrFCN -<*»0© i-iOO (N03N "<* rt< O OCOO OXO ot> CM CM CO co-*© COOOO ©©CM co*o© OOOrH T^CCrH t-co co^o O •<*co n (M^N CD00O HNO NO© lOCOO TtAG FOR ALUMINUM ST RANDED WIR E. 37 . i A N \ 1 by R.D.Johusdn t ^ r Mfc - '-CA V 3 X 34 33 _i_I.3P ^ < ' > \ _§£ 3 -A"' t A,- s°: L. _ 31 30 29 28 .:, ca T° \ \ -ivX \ r \\ \ T AA AA V • \ ■3 u V A~ _> 1000 Fe et Sp ai ^ UA \ On ^ k [■- .,, 27 \-AA \ s \ \ ~' 25 i > t-AV- \ ^ \ • lAA a A \ooo;f( ■et S] >a lS s ^3 1A Jl « A- sA s H i \ rv£ 5 \ A \A_a X) \ \i •*- „ \ r Av N -.; "- >. ; i \\ r A $ S; 11 10 _4 \ \ \. ^500 V * l\ \ .Sa - S * \ V too 'v\ v 8 7 * \\ s \ ^r * 1 \\ w ** - jr r ' 's. itiVA v ~. - 4 3 ra x *% !A\ - TO « \ 1 8 8 8 8 8 8 8 § § 1 1 8 § Circular Mils Fig. 28. — The top boundary of each span diagram is drawn for conditions at 100° Fahr.; the bottom boundary line for 0°. For other temperatures, interpolate or exterpolate proportionately. For mechanical reasons it is not recommended to string larger sizes of wire than appear in any closed span diagram, with any less sag than the minimum shown therein. The values of constants and assumptions of weather conditions are open to criticism. No responsibility is assumed beyond the correctness of the arithmetic. Assumptions. — Maximum stress at - 20° = 14,000 lbs. Ice coating one-half inch thick. Wind pressure 10 lbs. per sq.ft. proj. Diam. Diam. stranded conductor = 1.15 that of a solid wire of same section, Modulus of Elas- ticity = 7,500,000. 226 PROPERTIES OF CONDUCTORS. Deflections in Feet of Stranded Aluminum Wire in Still Air. H. W. Buck. Wire strung so that the maximum tension at minimum temperature of 0° F with wind blowing at 65 miles per hour (actual velocity) will be 14,000 lbs. per square inch. Span in Area of Wire Degrees Fahrenheit Rise above Minimum Temperature. Feet in Cir. Mils. 0° 20° 40° 60° 80° 100° 120° 140° 150° 200 553,150 265,400 132,300 .42 .45 .46 .51 .52 .55 .65 .65 .69 .83 .85 .92 1.07 1.13 1.30 1.57 1.65 1.82 2.20 2.27 2.45 2.75 2.80 2.95 2.97 3.03 3.10 400 553,150 265,400 132,300 1.80 1.95 2.20 2.20 2.42 2.75 2.70 2.90 3.40 3.35 3.70 4.20 4.15 4.50 5.10 5.05 5.45 6.00 6.00 6.40 7.00 6.90 7.35 7.85 7.20 7.78 8.50 600 553,150 265,400 132,300 4.3 5.1 6.2 5.1 6.1 7.2 6.0 7.1 8.4 7.0 8.2 9.7 8.2 9.5 11.0 9.5 10.8 12.2 10.8 12.0 13.3 11.9 13.1 14.4 12.5 13.6 15.7 800 553,150 265,400 132,300 8.4 10.3 14.0 9.5 11.7 15.4 10.8 13.2 16.9 12.3 14.7 18.3 13.8 16.4 19.6 15.4 17.7 29.0 16.9 19.1 22.2 18.3 20.4 23.4 19.0 21.5 25.5 1000 553,150 265,400 132,300 13.9 18.6 26.0 15.6 20.3 27.6 17.3 22.0 29.0 19.1 23.8 30.5 20.8 25.5 31.8 22.5 27.1 33.1 24.2 28.6 34.4 25.9 30.0 35.8 26.7 31.5 37.5 Deflections in Incites of Stranded Aluminum Wire in Still Air. H. W. Buck. Wire strung so that the maximum tension at minimum temperature of 0° F with wind blowing at 65 miles per hour (actual velocity) will be 14,000 lbs. per square inch. Calculations made for No. 2 B. and S. stranded conductor, but it is safe to follow this table for all sizes of cable, for the larger sizes will have slightly smaller deflections without exceeding their elastic limit on account of their greater relative strength. Degrees Fahren- Length of Span in Feet heit Rise above Minimum Temp. 200 180 160 140 120 100 6.3 5.3 4.2 3.1 2.2 1.7 10 7.0 5.7 4.5 3.4 2.4 1.8 20 7.8 6.4 5.1 3.8 2.8 1.9 30 8.8 7.3 5.8 4.5 3.2 2.2 40 10.2 8.4 6.7 5.2 3.8 2.7 50 12.0 9.8 7.8 6.4 4.6 3.3 60 14.0 11.5 9.4 7.5 5.6 4.0 70 16.5 14.0 11.5 9.2 7.0 5.2 80 19.8 17.0 14.3 11.4 8.9 6.8 90 23.1 20.0 16.8 13.8 10.3 8.8 100 26.6 23.3 20.0 16.6 13.1 10.8 110 29.8 26.6 23.0 19.5 16.5 13.1 120 33.5 29.8 25.8 22.2 18.7 15.2 130 36.8 32.8 28.7 24.5 20.8 17.2 140 40.0 35.8 31.5 26.8 22.8 18.8 150 43.0 38.4 33.6 29.1 24.8 20.3 DIELECTRICS. 227 PROPERTIES OF DIELECTRICS. Approximate Values of Specific Inductive Capacity of Various Dielectrics. Non-conducting materials or insulators are called dielectrics. The di- electric constant or specific inductive capacity of a dielectric is the ratio of the capacity of a condenser having the space between its plates filled with this substance to the capacity of the same condenser with this space filled with air. All gases and vacuum 1 . 00 Glass 3 to 8 Treated paper used in manufacture of power cables 2 to 4 Porcelain 4.4 Ebonite 2.5 Gutta-percha 2.5 Pure Para Rubber 2.2 Vulcanized Rubber 2.5 Paraffin 2.3 Rosin 1.8 Pitch 1.8 Wax 1.6 Mica 6 Water 80 Turpentine oil 2.2 Petroleum 2 Specific Resistance of Dielectrics at about 20° C. These are approximate values; the resistance of dielectrics varies greatly with their purity and method of preparation. Material. Benzine Ebonite Glass, flint Glass, ordinary Gutta-percha Mica Micanite Micanite cloth Micanite paper Oil asbestos Olive oil Ozokerite (crude) Paper, parchment Paper, ordinary Treated paper used in manufacture of power cables Paraffin Paraffin oil Shellac Vulcanized fiber, black Vulcanized fiber, red Vulcanized fiber, white . Wood, ordinary Wood, paraffined Wood, tar Wood, walnut . . Resistance in Resistance Millions of in Millions Megohms per of Meg- Cubic Centi- ohms per meter. Cubic Inch. 14 5.22 28,000 11,000 20,000 8,000 90 36 450 180 80 30 2,500 900 300 120 1,200 500 850 315 1 0.4 450 180 0.03 0.01 0.05 0.02 10 to 20 4 to 8 24,000 13,000 8 3 9,000 3,500 68 27 10 4 14 6 600 250 3,700 1,500 1,700 670 50 20 228 PROPERTIES OF CONDUCTORS. Variation of Resistance with Temperature. The variations in resistance of dielectrics with temperature is much more rapid than in the case of metals. The variation can be expressed by an exponential equation. Ro — Rji . Where Rq = resistance at standard temperature. R t — resistance at temperature differing t degrees from standard temperature. t = temperature. a = constant depending on the material. For gutta-percha, t in ° C a = 0.88 For pure rubber, t in ° C a = 0.95 For other substances, the processes of manufacture vary too widely to permit the establishment of temperature coefficients. Dielectric Strength of Insulating: Materials. C. KlNZBRUNNER. Let V = Voltage required to puncture a given thickness of material. v = Volts required to puncture a sheet of material .001 inch thick. t = Thickness of the material in thousandths of an inch. For all the materials given in table below, except pure para, For pure para, For all the materials given below, except ordinary paper and impreg- nated paper, the puncturing voltage is the same for a solid sheet of material as for a sheet built up of thin layers. In the case of ordinary paper and impregnated paper the puncturing voltage is proportional to the number of layers; i.e., V = nnVf, where n is the number of layers and V the thickness of each layer. Puncturing Voltages for Sheet .001 in. thick (v.) Presspahn 117 Manila paper 56 Ordinary paper 37 Fiber 57 Varnished paper 267 Red Rope paper 239 Impregnated paper 107 Varnished linen 256 Empire cloth 201 Leatheroid 73 Ebonite 682 Rubber 502 Gutta-percha 454 Para 370 DIELECTRICS. 229 The values in the preceding table are for tests made under the follow- ing conditions: 1. Electrodes, flat disks with round edges 1.5 inches in diameter. 2. Pressure on electrodes 0.5 pounds per square inch. 3. Voltage curve sinusoidal. 4. Frequency of the alternating current between 20 and 75 cycles per second. 5. Temperature 17° C, humidity of the air about 70 per Gent. 6. Pressure applied for 15 minutes. Rubber. Pure rubber is a liquid gum having a specific gravity of .915. The rubber of commerce is obtained by coagulating this gum by various means, the most approved method being by the hot vapor rising from a smudge made from oily nuts. Rubbers prepared in this way are called "Para" rubbers; Para is the name of a province of Brazil which supplies a large quantity of this kind of rubber. Vulcanized rubber is a mixture of this coagulated gum, thoroughly cleaned and dried, with sulphur. Pure rubber deteriorates rapidly, whereas vulcanized rubber is comparatively stable, and at the same time retains the properties which make it valuable as an insulating material. The amount of sulphur present varies from five to twenty per cent of the entire mass, the amount determining the hardness of the product. Rubber with a large admixture of sulphur is called vari- ously "hard rubber," "vulcanite" or "ebonite." Vulcanized rubber is used largely for insulating cables of all kinds. Specification* for 30% Rubber Insulating: Compound. Adopted 1906, by the following wire manufacturers: American Steel & Wire Co. Indiana Rubber & Ins. Wire Co. American Electrical Works. National India Rubber Co. Bishop Gutta Percha Co. New York Ins. Wire Co. Canadian Gen. Electric Co. John A. Roebling's Sons Co. Crescent Ins. Wire & Cable Co. Safety Ins. Wire & Cable Co. General Electric Co. Simplex Electrical Co. Hazard Mfg. Co. Standard Underground Cable Co. India Rubber & Gutta Percha Ins. Co. The compound shall contain not less than 30% by weight of fine dry Para rubber which has not previously been used in rubber compounds. The composition of the remaining 70% shall be left to the discretion of the manufacturer. Chemical. — The vulcanized rubber compound shall contain not more than 6% by weight of Acetone Extract. For this determination, the Acetone extraction shall be carried on for five hours in a Soxhlet extractor, as improved by Dr. C. O. Weber. mechanical. — The rubber insulation shall be homogeneous in char- acter, shall be placed concentrically about the conductor, and shall have a tensile strength of not less than 800 pounds per square inch. A sample of vulcanized rubber compound, not less than four inches in length shall be cut from the wire, with a sharp knife held tangent to the copper. Marks should be placed on the sample two inches apart. The sample shall be stretched until the marks are six inches apart and then immediately released; one minute after such release, the marks shall not be over 2| inches apart. The samples shall then be stretched until the marks are 9 inches apart before breaking. For the purpose of these tests, care must be used in cutting to obtain a proper sample, and the manufacturer shall not be responsible for results obtained from samples imperfectly cut. Electrical. — Each and every length of conductor shall comply with the requirements given in the following table. The tests shall be made at the Works of the Manufacturer when the conductor is covered with vulcan- ized rubber, and before the application of other coverings than tape or braid. 230 PROPERTIES OF CONDUCTORS. Tests shall be made after at least twelve hours' submersion in water and while still immersed. The voltage specified shall be applied for five minutes. The insulation test shall follow the voltage test, shall be made with a battery of not less than 100 nor more than 500 volts, and the reading shall be taken after one minute's electrification. Where tests for acceptance are made by the purchaser on his own premises, such tests shall be made within ten days of receipt of wire of cable by purchaser. Inspection. — The purchaser may send to the works of the manufacturer a representative, who shall be afforded all necessary facilities to make the above specified electrical and mechanical tests, and, also, to assure himself that the 30% of rubber above specified is actually put into the compound, but he shall not be privileged to inquire what ingredients are used to make up the remaining 70% of the compound. 30% Rubber Compound 'Voltage Test for 5 HEinutes. For 30 Minutes Test, Take 80% of These Figures. I. Size. Thickness of Insulation in Inches. A 2 32 5 64 3 32 A A 1,000,000 to 550,000 . 4,000 6,000 8,000 10,000 11,000 6,000 500,000 to 250,000 . 4,000 6,000 8,000 9,000 8,000 4/0 to 1 2 to 7 8 to 14 3,666 '4,666 5,000 4,000 6,000 7,000 10,000 12,000 13,000 II. Thickness of Insulation in Inches. A A A A A 10 "3"? 1,000,000 to 550,000 . 500,000 to 250,000 . 4/0 to 1 2 to 7 8 to 14 10,000 12,000 14,000 16,000 17,000 14,000 16,000 18,000 20,000 21,000 18,000 20 000 22,000 24,000 25,000 22,000 24,000 26 000 28,000 26,000 28,000 30,000 32,000 30,000 32,000 34,000 36 000 DIELECTRICS. 231 HEegpoIims per iflile GO Degrees F. One Minute Electrification. 1000000 C. M. 900000 C. M. 800000 C. M. 700000 C. M. 600000 C M. 500000 C. M. 400000 C. M. 300000 C. M. 250000 C. M. 4/0 Strd. 3/0 Strd. 2/0 Strd. 1/0 Strd. 1 Solid 2 Solid 3 Solid 4 Solid 5 Solid G Solid S Solid 9 Solid 10 Solid 12 Solid 14 Solid A A A A A 200 235 270 305 C40 350 375 390 420 430 470 455 500 ', L40 480 520 4 150 490 535 4 L60 500 545 t 190 540 590 i >20 530 635 . >66 \ >50 615 680 i >30 l >85 650 715 i K0 ( )20 690 750 i yjQ ( 355 720 790 \\ e J20 ( 390 760 840 610 \ no i 300 880 985 650 ' rso i 350 940 1050 690 ' 705 ! 305 1000 1120 750 \ 370 i )90 1110 1250 \ 300 < )30 1< 360 1200 1340 210 250 290 325 365 405 450 505 540 565 580 590 650 700 750 795 830 870 920 1060 1130 1200 1370 1470 A 235 280 325 370 420 465 530 590 630 660 675 690 760 830 900 940 990 1040 1100 1240 1310 1380 1540 1640 265 315 370 420 470 525 600 680 720 750 770 790 860 950 1040 1080 1130 1180 1230 1370 1440 1510 1680 1780 32 300 360 420 480 540 600 670 750 810 840 860 880 950 1060 1160 1210 1260 1300 1350 1490 1560 1620 1790 1890 Crutta-JPercha. A higher grade of insulating material is another gum, gutta-percha, which is used in its pure state. The use of this gum is confined almost entirely to the construction of the insulated core of submarine cables. Specific gravity, 0.9693 to 0.981. Weight per cubic foot, 60.56 to 61.32 pounds. Weight per cubic inch, 0.560 to 0.567 oz. Softens at 115 degrees F. Becomes plastic at 120 degrees F. Meltc at 212 degrees F. ' . Oxidizes and becomes brittle, shrinks and cracks when exposed to the air, especially at temperatures between 70 and 90 degrees F. Oxidation is hastened by exposure to light. Oxidation may be delayed by covering the gutta-percha insulation with a tape wnich has been soaked in prepared Stockholm tar. Where gutta-percha is kept continually under water there is no notice- able deterioration, and the same applies where gutta-percha leads are cov- ered with lead tubing. , Stretched gutta-percha, such as is used for insulating cables, will stand a strain of 1,000 pounds per square inch before any elongation. The breaking strain is about 3,500 pounds per square inch. The tenacity of gutta-percha is increased by stretching it. Resistance of Gatta-Percha under Pressure. — The resistance of gutta-percha under pressure increases according to the following formula, when R = the resistance at the pressure of the atmosphere, and r the resis- tance at p pounds per square inch. r = R (1 -1- 0.00023 p). 232 PROPERTIES OF CONDUCTORS. Let D = diameter in mils of over gutta-percha insulation. d — diameter of cable core. W = weight in pounds of gutta-percha per knot. w = weight in pounds of copper. Then for Solid Cable For Stranded Cables. D - V55W-J-49I W. 5-V w 1 -f- 8.93 - • w D = V70Aw+491 W i-V 1 + 6.97 W Approximate Electrostatic Capacity of a gutta-percha cable per knot is : ^r - : — : ■ microfarads. log D — log d The electrostatic capacity of a gutta-percha insulated cable compared with one of the same size insulated with india rubber is about as 120 is to 100. Dividing- Coefficients for Correcting- tite observed Resist- ance of Cirutta-PercUa at any Temperature to 95° Jk\ K. WlNNERTZ 1907. Degree F. Coefficient. Degree F. Coefficient. Degree F. Coefficient. 95 0.1415 74 1.089 53 6.015 94 0.1561 73 1.187 52 6.373 93 0.1721 72 1.293 51 6.722 92 0.1898 71 1.409 50 7.057 91 0.2105 70 1.535 49 7.377 90 0.2332 69 1.672 48 7.670 89 0.2574 68 1.821 47 7.943 88 0.2836 67 1.984 46 8.178 87 0.3125 66 2.161 45 8.383 86 0.3442 65 2.353 44 8.499 85 0.3833 64 2.562 43 8.585 84 0.4304 63 2.790 42 8.637 83 0.4801 62 3.035 41 8.678 82 0.5251 61 3.302 40 8.719 81 0.5848 60 3.588 39 8.757 80 0.6458 59 3.896 38 8.796 79 0.7066 58 4.223 37 8.834 78 0.7707 57 4.564 36 8.880 77 0.8406 56 4.919 35 8.932 76 0.9168 55 5.282 34 8.990 75 1.0000 54 5.650 33 9.053 DIELECTRICS. 233 Dielectric Strength of Air. The voltage required to break down the air between two terminals de- pends on the shape of the terminals, the distance between the terminals, and the constants of the circuit in series with the terminals. The following curves, published by Mr. S. M. Kintner in the proceedings of the American Institute of Electrical Engineers, give the voltage re- quired to break down air gaps of various lengths under various conditions. 75 70 65 GO 55 50 2 45 O | W 30 25 20 15 10 5 n. in ^X IV $ r NEEDLE POINT SPAKK GAP CURVE I A.I.E.E. Curve q II Water Uheostat in Gap Circuit .4.111 Small Condenser IV Two Small Condensers in Gap Circuit • V Gap Shielded with 8 7 Disc / 2 Inches 3 Fig. 29. With regard to the use of a spark gap for measuring high voltages, Mr. Kintner makes the following recommendations: "For the measurement of sudden pressure variations, such as those pro- duced on transmission lines by lightning, switching, grounds, short cir- cuits, etc., where the use of an oscillograph or similar device is not feasible, the spark-gap method is very useful. It is, in fact, the only method by which any satisfactory quantitative results can be obtained under such conditions. "When using a gap the writer prefers * round nose' (hemispherical shielded terminals); (slightly concave shields placed back of and coaxial with the terminals) ; the gap shouid be standardized over the range for which it is to be used just prior to taking measurements, and under as nearly the same surroundings, connections, etc., as possible. This preference is based on the convenience of operation and greater freedom from erratic behavior of this form of gap. "The spark gap, although apparently a very simple device, requires an expert operator to get results that are at all satisfactory." 234 PROPERTIES OF CONDUCTORS. 70 65 in ii 55 IV 50 O 40 O a 35 s CURVES OF JUMP DISTANCES Shielded Gaps. ^''Noses 6"Shields Placed %"Back of Terminals $ I Normal Gap O II Voltmeter Resistance in Gap Circuit e III Water Resistance >> >> " • IV Small Condenser »» " »« 30 $ 25 20 15 10 J 7 / I 1 5 3 4 ©ap Distances in Inches Fig. 30. .Puncturing- Voltage of Iflica in Transil Oil. W. S. Andrews. Thickness of Average Punc- Thickness of Average Punc- Mica. turing Voltage. Mica. turing Voltage. .00 1" 3,800 .006" 6,700 .0015" 4,500 .0065" 6.930 .002" 4,600 .007" 7,220 .0025" 4,750 .0075" 7,400 .003" 5,300 .008" 7,700 .004" 5,570 .0085" 8,550 .00475" 5,950 .01" 8,900 .005" 6,050 Specific Thermal Conductivity of Dielectrics. Watts Through Inch Cube. Temperature Gradient 1° C. Specific Specific Name of Substance. Conduc- Name of Substance. Conduc- tivity. .0006 tivity. Air Vulcanized Rubber . . .00105 Glass .0053 Beeswax .00093 Wood .032 Felt .00093 .00089 Caoutchouc Gutta-percha .... .0044 Vulcanite .0051 Cotton Wool .00046 Sandy Loam .... .085 Sawdust .00131 Bricks and Cement .032 Sand .00140 India Rubber .... .0043 Paraffin .00121 Sand with Air Spaces .96 DIELECTRICS. 235 minimum Size of Conductors for Hig-Ii Tension Transmission. The loss of energy in a high tension transmission line due to the brush discharge from the wires depends on the electric pressure, the size of the conductors and the atmospheric temperature and barometric pressure. For any given size of conductor a certain critical electric pressure exists for which there is a sudden rise in the curve of "loss between wires." Con- ductors should never be used in practice so small that the operating pres- sure is greater than this critical pressure. Mr. H. J. Ryan has deduced the following table, giving the minimum size of conductor which should be used for pressures from 50,000 to 250,000 volts for a distance between con- ductors of 48 inches: Operating Pressure; Minimum Diameter 90 per cent of Critical Effective Volts. of Conductor in Inches. 50,000 0.058 75,000 0.106 100,000 0.192 150,000 0.430 200,000 0.710 250,000 0.990 The equation showing the relation between the maximum value of the pressure wave, the atmospheric temperature and barometric pressure, the distance between the line conductors and the radius of the conductors for conductors larger than No. 4 B. and S. gauge is as follows: where E= E 17.946 459 + t X 350,000 log*) (f) (r + .07) critical pressure at which the sudden increase in the brush discharge takes place, r = radius of conductors in inches. s = distance between conductors from center to center in inches. t = atmospheric temperature in degrees Fahrenheit. b = barometric pressure in inches of mercury. PROPERTIES OP CONDUCTORS CARRYING ALTERNATING- CURRENTS. Ke vised by Harold Pender, Ph.D. * Besides the ohmic resistance of a wire, the following phenomena affect the flow of an alternating current: Skin effect, a retardation of the current due to the property of alter- nating currents apparently flowing along the outer surface or shell of the conductor, thus not making use of the full area. Inductive effects, (a) self induction of the current due to its alternations, inducing a counter E.M.F. in the conductor; and (6) mutual inductance, or the effect of other alternating current circuits. Capacity effects, due to the fact that all lines or conductors act as elec- trical condensers, which are alternately charged and discharged with the fluctuations of the E.M.F. ErFECTIVE RE§I§TAl¥€E~§KI]y EITECT. The effective resistance of a circuit to an alternating current depends on the shape of the circuit, the specific resistance, permeability, cross section and shape of the conductor, and the frequency of the current. The current density over the cross section of the conductor is a minimum at the center, increasing to a maximum at the periphery; in a solid conductor of large cross section the current is confined almost entirely to an outer shell or •'skin." The "Skin Effect Factor" is the number by which the re- sistance of the circuit to a continuous current must be multiplied to give the effective resistance to an alternating current. The following curve, formulae and table give the "Skin Effect Factor" for a straight wire of circular cross section, the return wire of the circuit being assumed suffi- ciently remote to be without effect, which is practically the case in an aerial transmission line. Let R = Resistance of wire in ohms to a continuous current. R' = Effective resistance of wire in ohms to an alternating current. / = Cycles per second. A = Cross section of wire in circular mil3. ix == Permeability of wire in C.G.S. units. t = Temperature in °C. a = Temperature coefficient per °C. C = Percentage conductivity of wire referred to Matthiessen's copper standard at 0° C. R Then §-' = function of f^-V This function is a complex one, and can be represented best by the accompanying curve; however, for ^>3 X1 0.o. the approximate formula - = 10~ 5 i/ ;\ +0.28 tc t 1 ~x~ at is sufficiently accurate for all practicable purposes. 236 SKIN EFFECT FACTORS. 237 Skin effect Factor* at 20° C. for Straig-ht Wire* Having* Circular Cross Section. Product of Cir- cular Mils by Cycles per Sec- ond. fX A. Factor * for Product of Cir- cular Mils by Cycles per Sec- ond. fXA. Factor for Iron Wire. C= 17 n = 150. Copper W r ire C = 100 Aluminum Wire C = 62 /*= 1. 500,000 1.000 5,000,000 1.000 1.000 1,000,000 1.015 10,000,000 1.000 1.000 2,000,000 1.068 20,000,000 1.008 1.000 3,000,000 1.144 30,000,000 1.025 1.006 4,000,000 1.234 40,000,000 1.045 1.015 5,000,000 1.332 50,000,000 1.070 1.026 6,000,000 1.435 60,000,000 1.096 1.040 7,000,000 1.535 70,000,000 1.126 1.053 8,000,000 1.628 80,000,000 1.158 1.069 9,000,000 1.714 90,000,000 1.195 1.085 10,000,000 1.795 100,000,000 1.230 1.104 12,500,000 1.974 125,000,000 1.332 1.151 15,000,000 2.14 150,000,000 1.433 1.206 17,500,000 2.29 175,000,000 1.530 1.266 20,000,000 2.42 200,000,000 1.622 1.330 25,000,000 2.68 250,000,000 1.790 1.455 30,000,000 2.90 300,000,000 1.937 1.575 35,000,000 3.11 350,000,000 2.07 1.686 40,000,000 3.31 400,000,000 2.20 1.787 45,000,000 3.49 450,000,000 2.31 1.879 50,000,000 3.67 500,000,000 2.42 1.965 55,000,000 3.83 550,000,000 2.53 2.05 60,000,000 3.99 600,000,000 2.63 2.13 sj.u / / L. 1 * Curve for Determining / / S kin Effect Facte r / l.b H.Pender / / U ■ t / Lfl / ps| i / Lb / / f L4 / / / 1 / t / 7 ffiC A xl"- 10 i.O _-*= ■/ 1+Ht .2 .4 .0 .3 LU UJ.4 1.6 1.3 2.0 2.2 2.4 2.6 2.9 \SL Fio. 1. * This corresponds to E.B.B. telegraph wire. 238 CONDUCTORS. The approximate formula For Iron (E.B.B. telegraph wire), reduces to |-' = 479 X 1(T 6 V 'JA + 0.28 for jA > 12.5 X 10 6 and t = 20° C. For Copper, reduces to ^ = 96 X HT 6 \/fA +0.28 ti for jA > 300 X 10 6 and t = 20° C. For Aluminum, reduces to jL' = 76 x io-sZ/Z+o.28 for /A > 500 X 10 6 and t = 20° C. Examples: To find the effective resistance of a round-wire .5 inch in diameter, permeability 500, conductivity 10 per cent, at 15 cycles per second and 0° C: fuCA 15 X 500 X 10 X .25 X 10 6 , cc v 1ni0 1 + at ~ 1 - 1.88 X 10 . /?' From the curve ■=— = 1.63 ti or effective resistance R' = 1.63 R. To find the effective resistance of the same wire at 60 cycles per second: ffiCA 1+at therefore, from formula R_ R or effective resistance R' = 3.01 R. = 7.5 X lO" 10 , - 2.73 + 0.28 = 3.01 SELJF i;]¥I>UCTJM>]¥ A»D E¥DCCTIVE REACTANCE OJP TltJLWSJfllSSJLO^r CIRCUITS FORMED BY PARALLEL WIRES. The Coefficient of Self Induction (L) of an elementary circuit is defined as the ratio of the number of lines of induction produced by a current flowing in the circuit divided by the current in the circuit. When the conductor has a finite croc- section the exact definition of the coefficient of self induction is the ratio of twice the energy of the magnetic field produced by the cur- rent flowing to the square of the current. The practical unit of self induction is the henry; sometimes the milli- henry is used, which is equal to x^s of a henry. The coefficient of self induction of a circuit depends on the size and shape of the circuit, the cross section and shape of the conductor, the per- meabilities of the conductor and the surrounding medium, also, when the skin effect is large, upon the frequency of the current and the specific re- sistance of the conductor. The instantaneous E.M.F. induced in a cir- cuit by any change of the current flowing in the circuit is e = — -z- (Li), or, if L is constant, which is strictly true when there is no iron in the circuit, and approximately so in any case, e = — L -=r • at When a constant E.M.F. is impressed on a circuit or coil containing inductance, the current does not reach its full value instantly, as it is SELF INDUCTION AND INDUCTIVE REACTANCE. 239 opposed at first by a counter-electromotive force due to the inductance. This counter-electromotive force gradual^ grows less until the current reaches its full strength, which theoretically takes an infinite time, and in practice it is usual to determine the time taken for the current to attain 63.2% of its full value and this period is called the time-constant. Time-constant in seconds — -= : ohms resistance _ henrys X final amperes applied volts If the impressed E.M.F. varies according to the sine law and L is con- stant, the effective value of the counter inductive E.M.F. is E = 2 vfLI where f = cycles per second or frequency and J = the effective value of the current. 2 nfL is called the inductive reactance or simply the inductance of the circuit. The induced E.M.F. lags 90° behind the current. The E.M.F. required to overcome the induced E.M.F. leads the current by 90°. Form u Up for Self Induction and Inductive Reactance. Let r = radius of wire in inches. n = number of wire on B. and S. gauge.* D = distance between wires in inches. I = distance of transmission (length of one wire) in 1000 feet. L = coefficient of self induction of 1000 feet of wire in millihenrys. / = frequency of current in cycles per second. X = 2 nfL X 10~ 3 = inductive reactance of 1000 feet of wire in ohms. Single-phase Circuit — 2 Wires. ID Fig. 2. Total self induction of circuit = 2 IL. Total inductive reactance of circuit = 2 IX. Three-phase Circuit — 3 Wires. Fig. 3. Total self induction per phase (circuit formed by any two wires) =\/3 IL. Total inductive reactance per phase = \/3 IX. For XOHr-JfEAOHTKTlC WIRES, L = 0.01524 + 0.14 log 10 (^\ - 0.00705 n + A. where A = 0.14 log 10 + 0.1258. * See table on next page for values of n for wires larger than No. 0. 240 CONDUCTORS. For IllOX WIRE, L-- . 0.01524 n + 0.14 log, •(f) where u = permeability of the iron. /u. varies with the quality of the iron and also with the strength of the current. The above formula is true only in case /a is constant over the cross section of the wire, which in any practical case is only approximately true. The tables on p. 248 are calcu- lated for /u. = 150, corresponding to good quality telegraph wire, and, therefore, L - 2.286 + 0.14 log 10 j - Values of A (=©.14 logr 10 I> + . 1858) for Various Interaxial Distances. D. A. I in. .0662 .0837 £ .1083 1 .1258 2 .1679 3 .1925 6 .2347 12 .2768 18 .3015 24 .3190 36 .3436 48 .3611 60 .3747 72 .3857 Values of n for Wires Larger than ~No. O, B. and S. Size. 00 B. and S. 000 0000 250,000 C. M. 800,000 350,000 400,000 450,000 500,000 550,000 600,000 650,000 700,000 750,000 800,000 850,000 900,000 950,000 1,000,000 -1 - 2 -3 - 3.719 - 4.505 -5.170 - 5.746 -6.254 - 6.708 -7.119 -7.495 -7.840 -8.159 -8.457 - 8.735 - 8.997 -9.243 - 9.476 - 9.697 n = 49.8812 - 9.92978 log CM. SELF INDUCTION. 241 Self Induction in ItEillihenrys per lOOO Feet of Solid Non-magnetic Wire. Note. — The self induction of a stranded wire is slightly less than that of a solid wire of the same cross section, and slightly greater than that of a solid wire having the same diameter, but more nearly equal to that of a solid wire with equal cross section. The exact value of the self induction of a strand is a complex expression involving both the size and number of the individual wires. (See L'Eclairage Electrique, Vol. Ill, p. 20.) For all practical purposes the self induction of a strand may be taken equal to that of a solid conductor having the same cross section. L = .00705 n + A. B. and S. Gauge. Interaxial Distances. 1" h" r 1" 2" 3" 0000 !ioi3 .1084 .1225 .1367 .1507 .1647 .0871 .0943 .1012 .1083 .1154 .1223 .1364 .1436 .1506 .1647 .1788 .1928 .2068 .1046 .1116 .1187 .1258 .1329 .1398 .1539 .1610 .1681 .1822 .1963 .2103 .2243 .1467 .1538 .1608 .1679 .1750 .1820 .1961 .2032 .2102 .2243 .2384 .2525 .2665 .1714 000 .1778 00 .1855 .1926 1 2 4 5 6 8 10 12 14 0977 1117 1189 1259 1401 1541 1682 1822 .1946 .2066 .2207 .2278 .2349 .2490 .2631 .2772 .2911 Interaxial Distances. Cir. Mils and B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .1659 .2080 .2327 .2502 .2748 .2923 .3059 .3169 900,000 .1691 .2112 .2359 .2534 .2780 .2955 .3091 .3201 800000 .1727 .2148 .2395 .2570 .2816 .2991 .3127 .3237 700 000 .1768 .2189 .2436 .2611 .2857 .3032 .3168 .3278 600,000 .1815 .2236 .2483 .2658 .2904 .3079 .3215 .3325 500 000 .1871 .2292 .2539 .2714 .2960 .3135 .3271 .3381 450,000 .1903 .2324 .2571 .2746 .2992 .3167 .3303 .3413 400,000 .1939 .2360 .2607 .2782 .3028 .3203 .3339 .3449 350,000 .1980 .2401 .2648 .2823 .3069 .3244 .3380 .3490 300,000 .2027 .2448 .2695 .2870 .3116 .3291 .3427 .3537 250,000 .2083 .2504 .2751 .2926 .3172 .3347 .3483 .3593 0000 .2135 .2556 .2803 .2978 .3224 .3399 .3535 .3645 000 .2206 .2627 .2874 .3049 .3295 .3470 .3606 .3716 00 .2276 .2648 .2945 .3120 .3366 .3541 .3677 .3787 .2347 .2768 .3015 .3190 .3436 .3611 .3747 .3857 1 .2418 .2839 .3086 .3261 .3507 .3682 .3818 .3928 2 .2488 .2909 .3156 .3331 .3577 .3752 .3888 .3998 4 .2629 .3050 .3297 .3472 .3718 .3893 .4029 .4139 6 .2770 .3191 .3438 .3613 .3859 .4034 .4170 .4280 8 .2911 .3332 .3579 .3754 .4000 .4175 .4311 .4421 10 .3052 .3473 .3720 .3895 .4141 .4316 .4452 .4562 242 CONDUCTORS. Inductive Reactance in Ohms Per lOOO feet of Solid Aon* Jflagrnetic TFire. 100 Cycles per Second. X = 0.6283 L. Note. — Inductive reactance at other frequencies proportional to values given in this table. Interaxial Distances. Gauge. ¥ ¥ 3// 4 1" 2" 3" 0000 .0547 .0657 .0922 .1076 000 .0592 .0701 .0966 .1116 00 . .0635 .0745 .1010 .1165 .0680 .0790 .1055 .1209 1 .0725 .0834 .1099 .1254 2 .0613 .0768 .0878 .1143 .1298 4 .0702 .0857 .0966 .1231 .1386 5 0636 .0747 .0902 .1011 .1276 .1431 6 0681 .0791 .0946 .1056 .1320 .1475 8 0770 .0879 .1034 .1144 .1409 .1564 10 0858 .0968 .1123 .1233 .1497 .1652 12 0946 .1056 .1211 .1321 .1586 .1741 14 1034 .1144 .1299 .1409 .1674 .1828 Cir. Mils and Interaxial Distances. B. and S. Gauge. 6" 12" 18* 24" 36" 48" 60" 72" 1,000,000 .1042 .1307 .1462 .1572 .1727 .1837 .1922 .1991 900,000 .1062 .1327 .1481 .1592 .1747 .1857 .1942 .2011 800 000 .1085 .1350 .1505 .1615 .1769 .1879 .1965 .2034 700,000 .1111 .1375 .1531 .1640 .1795 .1905 .1990 .2060 600,000 .1140 .1405 .1560 .1670 .1825 .1954 .2020 .2089 500,000 .1176 .1440 .1595 .1705 .1860 .1970 .2055 .2124 450,000 .1196 .1460 .1615 .1725 .1880 .1990 .2075 .2144 400,000 .1218 .1483 .1638 .1748 .1902 .2012 .2098 .2167 350,000 .1244 .1509 .1664 .1774 .1928 .2038 .2124 .2193 300,000 .1274 .1538 .1693 .1803 .1958 .2068 .2153 .2222 250,000 .1309 .1573 .1728 .1838 .1993 .2103 .2188 .2257 0000 .1341 .1606 .1761 .1871 .2026 .2136 .2221 .2290 000 .1386 .1651 .1806 .1916 .2070 .2180 .2266 .2335 00 .1430 .1695 .1850 .1960 .2115 .2225 .2310 .2379 .1475 .1739 .1894 .2004 .2159 .2269 .2354 .2423 1 .1519 .1784 .1939 .2049 .2203 .2313 .2399 .2468 2 .1563 .1828 .1983 .2093 .2247 .2357 .2443 .2512 4 .1652 .1916 .2072 .2181 .2336 .2446 .2531 .2601 6 .1740 .2005 .2160 .2270 .2425 .2535 .2620 .2689 8 .1829 .2093 .2249 .2359 .2513 .2623 .2709 .2778 10 .1918 .2182 .2337 .2447 .2602 .2712 .2797 .2866 INDUCTIVE KEACTANCE. 243 Inductive Reactance in Ohmi Per lOOO feet of Solid \ on- HEag-netic IFire. 25 Cycles Per Second. X = .1571 L. Interaxial Distances. B. and S. Gauge. 3 » f V 1" r 3" 0000 .0137 0169 .0230 .0269 000 .0148 0175 .0242 .0279 00 .0159 0186 .0253 .0291 .0170 0198 .0264 .0302 1 .0181 0209 .0275 .0313 2 !6i53 .0192 0220 .0286 .0325 4 .0176 .0214 0242 .0308 .0347 5 0159 .0187 .0225 0253 .0319 .0358 6 0170 .0198 .0236 0264 .0330 .0369 8 0192 .0220 .0259 0286 .0352 .0391 10 0215 .0242 .0281 0308 .0374 .0413 12 0237 .0264 .0303 0330 .0396 .0435 14 l0259 .0286 .0325 0352 .0418 .0457 Cir. Mils and Interaxial Distances B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .0261 .0327 .0366 .0393 .0432 .0460 .0481 .0498 900,000 .0266 .0332 .0371 .0398 .0437 .0465 .0486 .0503 800,000 .0272 .0338 .0377 .0404 .0443 .0471 .0492 .0509 700,000 .0278 .0344 .0383 .0410 .0449 .0477 .0498 .0515 600,000 .0285 .0351 .0390 .0417 .0456 .0484 .0505 .0522 500,000 .0294 .0360 .0399 .0426 .0465 .0493 .0514 .0531 450,000 .0299 .0365 .0404 .0431 .0470 .0498 .0519 .0536 400,000 .0305 .0371 .0410 .0437 .0476 .0503 .0525 .0542 360,000 .0311 .0377 .0416 .0444 .0482 .0510 .0531 .0548 300,000 .0319 .0385 .0423 .0451 .0490 .0517 .0538 .0556 250,000 .0327 .0393 .0432 .0460 .0498 .0526 .0547 .0564 0000 .0335 .0402 .0440 .0468 .0505 .0534 .0555 .0573 000 .0347 .0413 .0452 .0479 .0518 .0545 .0567 .0584 00 .0358 .0424 .0463 .0490 .0529 .0556 .0578 .0595 .0369 .0435 .0474 .0501 .0540 .0567 .0589 .0606 1 .0380 .0446 .0485 .0512 .0551 .0578 .0600 .0617 2 .0391 .0457 .0496 .0523 .0562 .0589 .0611 .0628 4 .0413 .0479 .0518 .0545 .0584 .0612 .0633 .0650 6 .0435 .0501 .0540 .0568 .0606 .0634 .0655 .0672 8 .0457 .0523 .0562 .0590 .0628 .0656 .0677 .0695 10 .0480 .0546 .0584 .0612 .0651 .0678 .0699 .0717 244 CONDUCTORS. Inductive* Reactance in Ohm* per lOOO Feet of Solid Non-HEagrnetic Wire. 60 Cycles Per Second. 0.3770 L. Interaxial Distances. B. and S. Gauge. r ¥ r 1" 2" 3" 0000 .0328 .0394 .0553 .0646 000 .0355 .0421 .0580 .0670 00 .0381 .0447 .0606 .0699 .0408 .0474 .0633 .0726 1 .0435 .0501 .0659 .0752 2 0368 .0461 .0527 .0686 .0779 4 0421 .0514 .0580 .0739 .0832 5 0382 .0448 .0541 .0607 .0766 .0859 6 0409 .0474 .0567 ,0633 .0792 .0885 8 0462 .0528 .0621 .0687 .0845 .0938 10 0515 .0581 .0674 .0740 .0898 .0991 12 0568 .0634 .0727 .0793 .0951 .1044 14 0621 .0687 .0779 .0845 .1004 .1097 + Interaxial Distances. Cir. Mils and B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .0026 .0784 .0877 .0943 .1036 .1102 .1153 .1194 900,000 .0638 .0796 .0880 .0955 .1048 .1114 .1166 .120G 800,000 .0652 .0810 .0903 .0fo9 .1062 .1128 .1170 .1220 700,000 .0667 .0825 .0918 .0^84 .1077 .1143 .1194 .1235 600,000 .0685 .0843 .0936 .1002 . 1005 .1101 .1212 .1253 500,000 .0706 .0864 .0957 .1023 .1116 .1183 .1233 .1274 450,000 .0718 .0876 .0969 .1035 .1128 .1194 .1245 .1286 400,000 .0731 .0890 .0983 .1049 .1141 .1207 .1259 .1300 350,000 .0746 .0905 .0998 .1064 .1157 .1225 .1274 .1316 300,000 .0764 .0923 .1016 .1082 .1175 .1241 .1292 .1333 250,000 .0785 .0944 .1037 .1103 .1196 .1262 .1313 .1354 0000 .0805 .0964 .1057 .1123 .1216 .1282 .1333 .1374 000 .0832 .0991 .1084 .1150 .1242 .1308 .1360 .1401 00 .0858 .1017 .1110 .1176 .1269 .1335 .1386 .1427 .0885 .1043 . 3 36 .1202 .1295 .1361 .1412 .1454 1 .0911 .1070 .1163 .1229 .1322 .1388 .1439 .1481 2 .0938 .1097 .1190 .1256 .1348 .1414 .1466 . 1507 4 .0991 .1150 .1243 .1309 .1402 .1468 .1510 .1561 6 .1044 .1203 .1296 .1362 .1455 .1521 .1572 .1613 8 .1097 .12o6 .1349 .1415 .1508 .1574 .1625 .1667 10 .1151 .1309 .1402 .1468 .1561 .1627 .1678 .1720 INDUCTIVE REACTANCE. 245 Inductive Reactance of Loop Formed by Two Wires of a Three-Phase Transmission JLine. Ohms per 1000 Feet of Line* (Conductor Non-Magnetic) 100 Cycles per Second. *i oop vSx f( or single wire. Note. — Inductive reactance at other frequencies proportional to values given in this table. Interaxial Distances B. and S. Gauge F ¥ 3// 4 1" 2" 3" 0000 .0947 .1138 .1596 .1864 000 .1025 .1214 .1673 .1933 00 .1100 .1291 .1749 .2018 .1178 .1368 .1827 .2094 1 .1255 .1445 .1903 .2171 2 1062 .1331 .1521 .1980 .2248 4 1215 .1484 .1674 .2133 .2401 5 L102 1293 .1563 .1758 .2210 .2478 6 L179 1369 .1638 .1828 .2286 .2554 8 L333 1523 .1791 .1982 .2440 .2708 10 . 1487 1677 .1945 .2135 .2593 .2862 12 .639 1830 .2097 .2288 .2746 .3014 14 • 1791 1982 .2250 .2440 .2898 .3167 Interaxial Distances. Cir. Mils and B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .1807 .2265 .2533 .2724 .2992 .3183 .3330 .3450 900,000 .1842 .2300 .2568 .2759 .3027 .3218 .3305 .3485 800,000 .1881 .2339 .2607 .2798 .3066 .3257 .3404 .3524 700,000 .1926 .2384 .2652 .2843 .3111 .3302 .3449 .3569 600,000 .1977 .2435 .2703 .2894 .3102 .3353 .3500 .3620 500,000 .2038 .2496 .2764 .2955 .3223 .3414 .3561 .3681 450,000 .2073 .2530 .2799 .2989 .3258 .3449 .3596 .3716 400,000 .2111 .2570 .2839 .3029 .3296 .3487 .3636 .3755 350,000 .2156 .2615 .2884 .3074 .3341 .3532 .3681 .3800 300,000 .2208 .2665 .2934 .3125 .3393 .3584 .3731 .3851 250 000 .2268 .2726 .2995 .3185 .3454 .3644 .3792 .3911 0000 .2324 .2783 .3052 .3242 .3511 .3702 .3840 .3969 000 .2402 .2861 .3130 .3320 .3587 .3778 .3927 .4047 00 .2478 .2937 .3206 .3397 .3665 , .3856 .4003 .4123 .2556 .3014 .3282 .3473 .3742 .3932 .4079 .4199 1 .2632 .3092 .3360 .3551 .3818 .4008 .4157 .4277 2 .2709 .3168 .3437 .3627 .3894 .4085 .4234 .4353 4 .2863 .3320 .3591 .3780 .4048 .4239 .4386 .4508 6 .3015 .3475 .3743 .3934 .4203 .4393 .4540 .4660 8 .3170 .3627 .3898 .4088 .4355 .4546 .4695 .4814 10 .3324 .3781 .4050 .4241 .4509 .4700 .4847 .4967 * Length of line equals one half the total length of wire in the loop. 246 CONDUCTORS. Inductive Reactance of Loop Formed oj Two Wires of a Three-Phase Transmission Line. Ohms Per 1000 Feet of Line.* (Conductor Non-Magnetic.) 25 Cycles per Second. Xloop = \/3 X for single wire. Interaxial Distances. B. and S. Gauge. 1" 2 3.// 4 1" 2" 3" 0000 .0237 .0285 .0399 .0466 000 .0256 .0304 .0418 .0483 00 .0275 .0323 .0437 .0504 .0294 .0342 .0457 .0524 1 .0314 .0361 .0476 .0543 2 0266 .0333 .0380 .0495 .0562 4 0304 .0371 .0418 .0533 .0600 5 .0 276 0323 .0391 .0438 .0552 .0620 6 .0 295 0342 .0409 .0457 .0572 .0639 8 .0 333 0381 .0448 .0495 .0610 .0677 10 .0 372 0419 .0486 .0534 .0648 .0715 12 .0 410 0457 .0524 .0572 .0687 .0754 14 .0 448 0495 .0562 .0610 .0725 .0792 Interaxial Distances. Cir. Mils and B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .0452 .0566 .0633 .0681 .0748 .0796 .0832 .0862 900,000 .0461 .0575 .0642 .0690 .0757 .0805 .0841 .0871 800,000 .0471 .0585 .0652 .0700 .0767 .0815 .0851 .0881 700,000 .0482 .0596 .0663 .0711 .0778 .0826 .0862 .0892 600,000 .0495 .0609 .0676 .0724 .0791 .0839 .0875 .0905 500,000 .0510 .0624 .0691 .0739 .0806 .0854 .0890 .0920 450,000 .0518 .0633 .0700 .0747 .0815 .0862 .0899 .0929 400,000 .0528 .0643 .0710 .0757 .0824 .0872 .0909 .0939 350,000 .0539 .0654 .0721 .0769 .0835 .0883 .0920 .0950 300,000 .0552 .0666 .0734 .0781 .0848 .0896 .0933 .0963 250,000 .0567 .0682 .0749 .0796 .0864 .0911 .0948 .0978 0000 .0581 .0696 .0763 .0811 .0878 .0926 .0962 .0992 000 .0601 .0715 .0783 .0830 .0897 .0945 .0982 .1012 00 .0620, .0734 .0802 .0849 .0916 .0964 .1001 .1031 .0639 .0754 .0821 .0868 .0936 .0983 .1020 .1050 1 .0658 .0773 .0840 .0888 .0955 .1002 .1039 .1069 2 .0677 .0792 .0859 .0907 .0974 .1021 .1059 .1088 4 .0716 .0830 .0898 .0945 .1012 .1060 .1097 .1127 6 .0754 .0869 .0936 .0984 .1051 .1098 .1135 .1165 8 .0793 .0907 .0975 .1022 .1089 .1137 .1174 .1204 10 .0831 .0945 .1013 .1060 .1127 .1175 .1212 .1242 * Length of line equals half the total length of wire in the loop. INDUCTIVE REACTANCE. 247 Inductive Reactance of JLoop Formed or Two Wires of a Xhree-Phase Transmission ILine. Ohms Per 1000 Feet of Line.* (Conductor Non-Magnetic.) 60 Cycles per Second. Xloop = V3 X for single wire. Interaxial Distances. B. and S. Gauge. t" ¥ 3.// 4 1" 2" 3" 0000 .0568 .0615 .0660 .0707 .0753 .0683 .0728 .0774 .0821 .0867 .0958 .1004 .1049 .1096 .1142 .1118 000 .1160 00 .1211 .1257 1 .1303 2 0637 .0798 .0912 .1188 .1349 4 0729 .0890 .1004 .1280 .1440 5 .0 R61 0776 .0938 .1051 .1326 .1487 6 .0 708 0821 .0983 .1097 .1372 .1533 8 .0 300 0914 .1075 .1189 .1464 .1625 10 .0 392 1006 .1167 .1281 .1556 .1717 12 .0 983 1098 .1258 .1373 .1648 .1809 14 .1 375 1189 .1350 .1464 .1739 .1900 Cir. Mils and Interaxial Distances. B. and S. Gauge. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .1084 .1359 .1519 .1634 .1795 .1909 .1998 .2070 900,000 .1105 .1380 .1540 .1655 .1816 .1930 .2019 .2091 800,000 .1129 .1404 .1564 .1679 .1840 .1954 .2043 .2115 700,000 .1156 .1431 .1591 .1706 .1867 .1981 .2070 .2142 600,000 .1187 .1462 .1622 .1737 .1898 .2012 .2101 .2173 500,000 .1223 .1498 .1658 .1773 .1934 .2048 .2137 .2209 450,000 .1244 .1518 .1679 .1793 .1955 .2069 .2158 .2230 400,000 .1267 .1542 .1703 .1817 .1978 .2092 .2182 .2253 350 000 .1294 .1569 .1730 .1844 .2005 .2119 .2209 .2280 300,000 .1325 .1599 .1760 .1875 .2036 .2150 .2239 .2311 250,000 .1361 .1636 .1797 .1911 .2072 .2186 .2275 .2347 0000 .1394 .1670 .1831 .1945 .2107 .2221 . 2309 .2381 000 .1441 .1717 .1878 .1992 .2152 .2267 .2356 .2428 00 .1487 .1762 .1924 .2038 .2199 .2314 .2402 .2474 .1534 .1808 .1969 .2084 .2245 .2359 .2447 .2519 1 .1579 .1855 .2016 .2131 .2291 .2405 .2494 .2566 2 .1625 .1901 .2062 .2176 .2336 .2451 .2540 .2612 4 .1718 .1992 .2155 .2268 .2429 .2543 .2632 .2705 6 .1809 .2085 .2246 .2360 .2522 .2636 .2724 .2796 8 .1902 .2176 .2339 .2453 .2613 .2728 .2817 .2888 10 .1994 .2269 .2430 .2545 .2705 .2820 .2908 .2980 * Length of line equals one half the total length of wire in the loop. 248 CONDUCTORS. Self Induction in Jlillihenrys per lOOO Feet of Solid Iron Wire. Permeability 15© C. O. S. Units. L - 2.286+ .14 1og 10 (j'y Interaxial Distances. Roebling Dia. In. Gauge. 1" 2" 3" 6" 9" 12" 18" 24" 4 .225 2.4189 2.4610 2.4857 2.5278 2.5525 2.5699 2.5946 2.6121 6 .192 2.4285 2.4706 2.4953 2.5374 2.5621 2.5796 2.6042 2.6217 8 .162 2.4389 2.4809 2.5056 2.5478 2.5724 2.5899 2.6146 2.6321 9 .178 2.4443 2.4865 2.5111 2.5533 2.5779 2.5954 2.6201 2.6376 10 .135 2.4499 2.4921 2.5167 2.5589 2.5835 2.6010 2.6257 2.6432 11 .120 2.4571 2.4992 2.5239 2.5660 2.5907 2.6082 2.6328 2.6503 12 .105 2.4652 2.507412.5319 2.5742 2.5988 2.6163 2.6409 2.6584 14 .080 2.4817 2.5239 2.5485 2.5907 2.6153 2.6328 2.6575 2.6749 Inductive Reactance in Ohms per lOOO feet of Solid Iron l^ire. 100 Cycles Per Second. X = 0.6283 L. Note. — Inductive reactance at other frequencies proportional to values given in this table. Interaxial Distances. Roebling Dia. In. Gauge. 1" 2" 3" 6" 9" 12" 18" 24" / 4 .225 1.5191 1.5455 1.5610 1.5875 1.6029 1.6139 1.6294 1.6404 6 192 1.5251 1.5516 1.5671 1.5935 1.6090 1.6199 1.6355 1.6465 8 162 1.5316 1.5581 1.5735 1.6000 1.6155 1.6265 1.6419 1.6529 9 148 1.5350 1.5615 1.5769 1.6035 1.6189 1.6299 1.6454 1.6564 10 135 1.5386 1.5650 1.5805 1.6069 1.6225 1.6335 1 . 6489 1.6599 11 i?n 1.5431 1.5695 1.5850 1.6115 1 . 6269 1.6379 1.6534 1 . 6644 12 105 1.5482 1.5746 1.5901 1.6166 1.6320 1.6430 1.6585 1.6695 14 .080 1.5585 1.5850 1.6005 1.6269 1.6424 1.6534 1.6689 1.6799 CAPACITY. CAPACITY REACTANCE, Al¥» CHAItO- K¥« CURRENT OF TMAXSIfllSSlOHr CIRCUITS FOIl.lI E1» BY PARALLEL WIRES. Whenever a difference of potential is established between two or more conductors a static charge manifests itself on each conductor. If there are but two conductors present these static charges are equal and opposite. Two conductors thus carrying equal and opposite charges are said to form a condenser. The ratio of the charge (q) on one of the conductors to the difference of potential (e) between the two conductors is called the capa- city (C) of the condenser, i.e., + 215. For values of B see p. 251. For stranded wires neither formula is strictly accurate; the logarithmic formula gives results practically correct; values calculated by the second formula are about 3 per cent too small. TRANSMISSION CIRCUITS. 251 Concentric Cable in Grounded Metallic Sheath, Single -Phase. Let C ' = capacity in microfarads per 1000 feet of condenser formed by the two conductors. C" = capacity in microfarads per 1000 feet of condenser formed by outer conductor and sheath. Then C ' = C" = ♦007354 K x log l0 — .007354 K 2 logio -7 ^3 Fig. 9. Total charging current = I b' V + I b" Vo. Three Overhead Wires, Three-Phaie. .003677 SO' o c = logio - B + 13.7 n Total capacity per wire = 2 I C. Total capacitance per wire = 2 I b. m , !_ • • 2Z6V Total charging current per wire = — — = \/3 2lbV . Fig. 10. Three Wires in Metallic Sheath, Three-Phase. C - .007354 K T3 a 2 (ft 2 - a 2 )H logw L"F" ft 6 - a* J Total capacity per wire = 2 Z C. Total capacitance per wire = 2 I b. Total charging current per wire = — - = 2 V3 bVo- Fig. 11. Sheath Grounded. Values of B= £»£ log- l0 I* + «15. D. B. 1 99 i 133 i 181 l 215 2 297 3 344 6 426 12 508 18 556 24 590 36 638 48 672 60 698 72 720 * B = 272 log^o D + 215. For values see table. For stranded wires neither formula is strictly accurate; the logarithmic formula gives results practically correct; values calculated by the second formula are about 3 per cent too small. 252 CONDUCTORS. Capacity in .flicro farad* per lOOO feet of Circuit (2000 Feet of Wire) .Formed l>y Two Parallel Aerial Wires. C = 0.003677 D ' B+ 13.7 n 6 . .13 Interaxial Distances. pq Dia. over Insul. r ¥ r 1" 2" 3* 6* 12" 18" D000 .00748 .00723 .00696 .00669 .00643 .00678 .Q0626 .00601 .00576 .00591 .00541 .00499 .00459 .00716 .00575 .00534 .00497 .00465 .00437 .00413 .00371 .00353 .00336 .00308 .00284 .00264 .00246 .00391 .00371 .00353 .00337 .00522 .0030S .00284 .00274 .00264 .00246 .00230 .00217 .00205 .00329 .00315 .00302 .00290 .00279 .00269 .00250 .00242 .00234 .00220 .00207 .00196 .00180 .00259 .00250 .00242 .00234 .00227 .00220 .0020G .00202 .0019G .00186 .00177 .00109 .00162 .00214 .00208 .00202 .00197 .00191 .00183 .00177 .00173 .001G9 .00162 .00155 .00148 .00143 .00194 000 .00652 .0018& 00 .00598 .00184 o .00553 .00180 1 .00514 00175 2 .00624 .00480 .00424 .00401 .00380 .00344 .00314 .00288 .00268 o00171 4 5 6 8 10 12 14 ".00597 .00552 .00479 .00423 .00380 .00344 .00533 .004b6 .00465 .00412 .00370 .00336 .00308 .00163 .0016C .00156 .0015C .00144 .00139 .00134 Interaxial Distances. Size Cir. Mils 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .00361 .00279 .00246 .00227 .00204 .00191 .00182 .00173 900,000 .00353 .00274 .00242 .00223 .00202 .00189 .00180 .00173 800,000 .00345 .00269 .00238 .00220 .00199 .00186 .00178 .00171 750,000 .00340 .00266 .00236 .00218 .00198 .00185 .00177 .00170 700,000 .00335 .00263 .00233 .00216 .00196 .00184 .00175 .00169 600,000 .00325 .00257 .00229 .00212 .00193 .00181 .00172 .00166 500,000 .00314 .00250 .00223 .00207 .00189 .00177 .00169 .00163 450,000 .00308 .00246 .00220 .00205 .00186 .00175 .00168 .00162 400,000 .00302 .00242 .00216 .00202 .00184 .00173 .00166 .00160 350,000 .00295 .00237 .00213 .00199 .00181 .00171 .00163 .00158 300,000 .00287 .00232 .00209 .00195 .00178 .00168 .00161 .00156 250,000 .00278 .00226 .00207 .00191 .00175 .00165 .00158 .00153 0000 .00271 .00222 .00250 .00188 .00172 .00163 .00156 .00151 000 .00261 .00215 .00195 .00183 .00168 .00159 .00153 .00148 00 .00252 .00209 .00190 .00178 .00164 .00156 .00149 .00145 .00244 .00203 .00185 .00174 .00161 .00152 .00147 .00142 1 .00235 .00197 .00180 .00170 .00157 .00149 .00143 .00139 2 .00227 .00192 .00175 .00165 .00153 .00146 .00140 .00136 4 .00214 .00182 .00167 .00158 .00147 .00140 .00135 .00131 Solid 6 .00196 .00169 .00156 .00148 .00139 .00132 .00128 .0012? Solid 8 .00186 .00162 .00150 .00143 .00133 .00128 .00124 .0012C Solid 10 .00177 .00155 .00144 .00137 .00129 .00123 .00120 .0011? * For stranded wires the last formula gives values about 3% too small. TRANSMISSION CIRCUITS. 253 Charging* Current in Amperei per lOOO JFeet of Single- Phase Circuit (2000 feet of Wire) Formed by Two Parallel Aerial Wires. Pressure, E = 10,000 Volts. Frequency, / = 100 Cycles per Second. Charging Current = 6.283 C. Note. — Values of charging current at other pressures and frequencies are proportional to those given in this table. oooo ooo 00 1 2 4 5 Interaxial Distances. Dia. over Insul, .04699 .04542 .04373 04203 .04040 .04260 . 03933 .03776 . 03619 .03513 .03399 .03135 .02834 .03751 . 03468 .03009 .02658 .02387 02161 P 03920 03348 03116 02921 02588 . 02325 02111 01935 .04498 .04096 .03757 .03474 .03229 .03016 .02664 .02519 .02387 .02161 .01973 .01809 .01684 .03613 .03355 .03123 .02921 .02745 .02595 .02331 .02218 .02111 .01935 .01784 .01658 .01545 .02456 .02331 .02218 .02117 .02023 01935 .01784 .01721 .01658 .01545 .01445 .01363 ,01288 .02067 .0^979 .01897 .01822 .01753 .01690 .01571 .01520 .01470 .01382 .01300 .01231 .01168 .01627 .01571 . 01520 .01470 .01426 .01382 .01307 .01269 .01231 ,01168 ,01112 ,01061 ,01017 12* .01344 .01306 .01269 .01237 .01200 .01168 .01112 .01087 .01062 .01018 00973 ,00929 ,00898 18* .01218 .01187 .01156 =01130 .01099 .01074 .01024 .01005 .00980 .00942 .00905 .00873 .00842 Size Cir. Mils Interaxial Distances. Stranded. 6" 12" 18" 24" .01426 36" 48" 60" 72" 1,000,000 .02268 .01753 .01545 .01281 .01200 .01143 .01099 900,000 .02271 .01721 .01520 .01401 .01269 .01187 .01131 .01087 800,000 .02167 .01690 .01495 .01382 .01250 .01168 .01118 .01074 750,000 .02136 .01671 .01483 .01369 .01244 .01162 .01112 .01068 700,000 .02105 .01654 .01404 .01357 .01231 .01156 .01099 .01062 600,000 .02042 .01615 .01430 .01332 .01213 .01137 .01081 .01043 500,000 .01972 .01571 .01401 .01300 .01187 .01112 .01062 .01024 450,000 .01935 .01545 .013S2 .01288 .01168 .01099 .01055 .01018 400,000 .01897 .01520 .01363 .01269 .01156 .01086 .01043 .01005 350,000 .01853 .01489 .01338 .01250 .01137 .01074 .01024 .00993 300,000 .01803 .01457 .01313 .01225 .01118 .01055 .01011 .00980 250,000 .01746 .01426 .01300 .01200 .01099 .01036 .00993 .00961 0000 .01702 .01395 .01256 .01181 .01080 .01024 .00980 .00949 000 .01640 .01351 .01225 .01149 .01055 .00999 .00961 .00930 00 .01583 .01313 .01194 .01118 .01030 .00980 .00936 .00911 .01533 .01275 .01162 .01093 .01011 .00955 .00923 .00892 1 .01476 .01238 .01131 .01068 .00986 .00936 .00898 .00873 2 .01426 .01206 .01099 .01043 .00961 .00917 .00879 .00854 4 .01344 .01143 .01049 .00993 .00923 .00879 .00848 .00823 Solid 6 .01231 .01062 .00980 .00936 .00873 .00829 .00804 .00785 Solid 8 .01168 .01011 .00942 .00898 .00835 .00804 .00779 .00754 Solid 10 .01112 .00973 .00905 .00861 .00810 .00773 .00754 .00735 10 254 CONDUCTORS. Charging- Current in Amperes per lOOO Feet of Single* Phase Circuit (2000 feet of Wire) formed i>j Two Parallel Aerial Wires. Pressure, E = 10,000 Volts. Frequency, / =» 25 Cycles per Second. Charging Current = 1.571 C. Note. — Values of charging current at other pressures are proportional to those given in this table. e . Interaxial Distances. GO ^ #£ Dia. /A over i" *" r 1" 2" 3" 6" 12" 18" PQ Insul. 0000 .01175 .01124 . 00903 .00614 .0051 7 .0040 7 .00336 .00305 000 .01135 «... . 01024 .00839 . 00583 .0049 5 .0039 3 .00326 .00297 00 . 01093 .00939 . 00781 .00554 .0047 4 .0038 ) .00317 .00289 .01051 .00868 . 00730 .00529 .0045 5 .0036 7 .00309 .00282 1 . 01010 .00807 . 00686 .00506 .0043 8 .0035 3 .00300 .00275 2 . 01065 .... .00980 .00754 .00649 .00484 .0042 2 .0034, 5 .00292 .00268 4 .00983 .00837 . 00666 .00583 .00446 .0039 3 .0032 7 .00278 .00256 5 .00944 .00938 .00779 .00630 .00554 . 00430 .0038 .0031 J .00272 .00251 6 . 00905 . 00867 . 00730 . 00597 . 00528 . 00414 .0036 7 .0030* 1 .00265 . 00245 8 . 00928 . 00752 .00647 .00540 . 00484 . 00386 .0034 5 .0029i I .00254 .00235 10 .00850 . 00664 . 00581 .00493 .00446 .00361 .0032, 5 .0027* * .00243 .00226 12 . 00784 .00597 . 00528 .00452 .00414 .00341 .0030 3 .0026* ) .00232 .00218 14 .00721 .00540 .00484 .00421 .00386 .00322 .0029 2 .002# [ .00224 . 00210 Interaxial Distances. Size Cir. Mils Stranded. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .0056 7 .00438 .00386 .00356 .00320 .00300 .00286 .00275 900,000 .0055 4 .00430 .00380 .00350 .00317 .00297 .00283 .00272 800,000 .0054 2 .00422 .00374 .00345 .00312 .00292 .00279 .00268 750,000 .0053 4 .00418 .00371 .00342 .00311 .00290 .00278 .00267 700,000 .0052 6 .00414 .00366 .00339 .00308 .00289 .00275 .00265 600,000 .0051 3 .00404 .00360 .00335 .00303 .00284 .00270 .00261 500,000 .0049 3 .00393 .00350 .00325 .00297 .00278 .00265 .00256 450,000 .0048 4 .00386 . 00345 .00322 .00292 .00275 .00264 .00254 400,000 .0047 1 .00380 .00341 .00317 .00289 .00271 .00261 .00251 350,000 .0046 3 .00372 .00334 .00312 .00284 .00268 .00256 .00248 300,000 .0045 L .00364 .00328 .00306 .00279 .00264 .00253 .00245 250,000 .00431 3 .00356 .00325 .00300 .00275 00259 .00248 .00240 0000 .0042^ 5 .00349 .00314 .00295 .00270 00256 .00245 .00237 000 .004K ) .00338 .00306 .00287 .00264 00250 .00240 .00232 00 .0039( 3 .00328 .00298 .00279 .00257 00245 .00234 .00228 .0038: J .00319 .00290 .00273 .00253 00239 .00231 .00223 1 .0036< ) .00309 .00283 .00267 .00246 00234 .00224 .00218 2 .0035( ) .00301 .00275 .00261 .00240 . 00229 .00220 .00213 4 .0033( 5 .00286 .00262 .00248 .00231 . 00220 .00212 .00206 Bolid 6 .0030£ I .00265 .00245 .00234 .00218 . 00207 .00201 .00196 Solid 8 .0029S 5 .00253 .00235 .00224 .00209 . 00201 00195 .00188 Solid 10 .0027£ 1 .00243 .00226 .00215 .00202 . 00193 .00188 .00184 TRANSMISSION CIRCUITS. 255 Charging* Current in Amperes per lOOO feet of Single^ Phase Circuit (SOOO JFeet of Wire) Formed by Iwo Parallel Aerial Wires. Pressure, E = 10,000 Volts. Frequency, / = 60 Cycles per Second. Charging Current = 3.77 C. Note. — Values of charging current at other pressures are proportional to those given in this table. . ad 3 Interaxial Distances. Dia. over Insul. 3// 8 2 3// 4 1" 2" 3" 6" 12" 18* noon .02819 .02725 .02623 .02521 .02424 .02556 .02359 .02265 .02171 .02227 .02039 .01881 .01730 .02698 .02167 .02013 .01873 .01752 .01647 .01557 .01398 .01331 .01266 .01161 .01070 .00995 .00927 .01473 .01398 .01330 .01270 .01213 .01161 .01070 .01032 .00994 .00927 .00867 .00817 .00772 .01240 .01187 .01138 .01093 .01051 .01014 .00942 .00912 .00882 .00829 .00780 .00738 .00700 .00976 .00942 .00912 .00882 .00855 .00829 .00784 .00761 .00738 .00700 .00667 .00636 .00610 .00806 .00783 .00761 .00742 .00720 .00700 .00667 .00652 .00637 .00611 .00583 .00557 .00538 .00731 .00712 .00693 .00678 .00659 .00644 .00614 .00603 .00588 .00565 .00543 .00523 .00505 000 .02457 00 .02254 .02084 i .01937 ? .02352 .01809 .01598 .01511 .01432 .01296 .01183 .01085 .01010 4 .02008 5 6 8 10 12 14 .02251 .02080 .01805 .01595 .01432 .01296 .01869 .01752 .01552 .01395 .01266 .01161 Size Cir. Mils Interaxial Distances. Stranded. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .01360 .01052 .00927 .00855 .00768 .00720 .00686 .00659 900,000 .01330 .01032 .00912 .00840 .00761 .00712 .00678 .00652 800,000 .01300 .01014 .00897 .00829 .00750 .00700 .00670 .00644 750,000 .01281 .01002 .00889 .00821 .00746 .00697 .00667 .00640 700,000 .01263 .00993 .00878 .00814 .00738 .00693 .00659 .00636 600,000 .01224 .00969 .00863 .00800 .00727 .00682 .00648 .00625 500,000 .01183 .00942 .00840 .00780 .00712 .00667 .00637 .00614 450,000 .01161 .00927 .00829 .00772 .00700 .00659 .00633 .00610 400,000 .01138 .00912 .00817 .00761 .00693 .00651 .00625 .00603 350,000 .01111 .00893 .00802 .00750 .00682 .00644 .00614 .00596 300,000 .01081 .00874 .00787 .00735 .00670 .00633 .00606 .00588 250,000 .01047 .00855 .00780 .00720 .00659 .00621 .00596 .00576 0000 .01021 .00837 .00753 .00708 .00648 .00614 .00588 .00569 000 .00984 .00810 .00735 .00689 .00633 .00599 .00576 .00558 00 .00949 .00787 .00716 .00670 .00618 .00588 .00561 .00546 .00919 .00765 .00697 .00655 .00606 .00573 .00553 .00535 1 .00885 .00742 .00678 .00640 .00591 .00561 .00538 .0052; 2 .00855 .00723 .00659 .00626 .00576 .00550 .00527 .00512 4 .00806 .00685 .00629 .00595 .00553 .00527 .00508 .00493 Bolid 6 .00738 .00636 .00588 .00561 .00523 .00497 .00482 .00471 Solid 8 .00700 .00606 .00565 .00538 .00501 .00482 .00467 .00452 Solid 10 .00667 .00583 .00543 .00516 .00486 .00464 .00452 .00441 256 CONDUCTORS. Charging* Current in Ampere.* per Wire per lOOO feet of Three-Phase Circuit Formi'il Uy Three Parallel Aerial Wires. Pressure between Wires, E = 10,000 Volts. Frequency, / = 100 Cycles per Second. Charging Current per Wire = 7.26 C. Note. — Values of charging current at other pressures and frequencies are proportional to those given in this table. Interaxial Distances. Dia. over Insul. r ¥ 3." 4 1" 2" 3" 6" 12" 18" 0000 .05430 .05249 .05053 .04857 .04668 .04922 .04545 .04363 .04182 .04291 .03928 .03623 .03332 .05198 .04174 .03877 .03608 .03376 .03173 .02998 .02693 .02563 .02439 .02236 .02062 .01917 .01786 .02839 .02693 .02563 .02447 .02338 .02236 .02062 .01989 .01917 .01786 .01670 .01575 .01488 .02388 .02287 .02192 .02105 .02025 .01953 .01815 .01757 .01699 .01597 .01503 .01423 .01350 .01880 .01815 .01757 .01699 .01648 .01597 .01510 .01466 .01423 .01350 .01285 .01227 .01176 .01554 .01510 .01466 .01430 .01387 .01350 .01285 .01256 .01227 .01176 .01125 .01074 .01038 .01408 000 .04733 .01372 00 .04341 .01336 o .04015 .01307 1 .03732 .01270 2 .04530 .03485 .03078 .02911 .02759 .02497 .02280 .02091 .01946 .01241 4 .03869 .01183 5 6 8 10 12 14 .04334 .04007 .03477 .03071 .02759 .02497 .03601 .03376 .02991 .02686 .02439 .02236 .01162 .01132 .01089 .01045 .01009 .00973 Interaxial Distances. Size Cir. Mils Stranded. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .02621 .02025 .01786 .01648 .01481 .01387 .01321 .01270 900,000 .02563 .01989 .01757 .01619 .01466 .01372 .01307 .01256 800,000 .02505 .01953 .01728 .01597 .01445 .01350 .01292 .01241 750,000 .02468 .01931 .01713 .01583 .01437 .01343 .01285 .01234 700,000 .02432 .01911 .01691 .01568 .01423 .01336 .01270 .01227 600,000 .02359 .01866 .01662 .01539 .01401 .01314 .01249 .01205 500,000 .02280 .01815 .01619 .01503 .01372 .01285 .01227 .01183 450,000 .02236 .01786 .01597 .01488 .01350 .01270 .01220 .01176 400,000 .02192 .01757 .01568 .01466 .01336 .01256 .01205 .01161 350,000 .02142 .01721 .01546 .01445 .01314 .01241 .01183 .01147 300,000 .02084 .01684 .01517 .01416 .01292 .01220 .01169 .01132 250,000 .02018 .01641 .01503 .01387 .01270 .01198 .01147 .01111 0000 .01967 .01612 .01452 .01365 .01249 .01183 .01132 .01096 000 .01895 .01561 .01416 .01328 .01220 .01154 .01111 .01074 00 .01829 .01517 .01379 .01292 .01191 .01132 .01082 .01053 .01771 .01474 .01343 .01263 .01169 .01103 .01067 .01031 1 .01706 .01430 .01307 .01234 .01140 .01082 .01038 .01009 2 .01648 .01394 .01270 .01198 .01111 .01060 .01016 .00987 4 .01554 .01321 .01212 .01147 .01067 .01016 .00980 .00951 Solid 6 .01423 .01227 .01132 .01074 .01009 .00958 .00929 .00907 Solid 8 .01350 .01176 .01089 .01038 .00965 .00929 .00900 .00871 Solid 10 .01285 .01125 .01045 .00995 .00936 .00893 .00871 .00849 TRANSMISSION CIRCUITS. 257 Charging* Current in Amperes per \\ ire per lOOO Feet of Three-Phase Circuit Formed by Three Parallel Aerial Wire*. Pressure between Wires, E= 10,000 Volts. Frequency, / = 25 Cycles per Second. Charging Current per Wire = 1.815 C. Note. — Values of charging current at other pressures are proportional to those given in this table. ml *5 Interaxial Distances. 1000 000 00 1 2 4 5 6 8 10 12 14 Dia. over Insul. .01358 .01312 .01263 .01214 .01167 .01230 .01136 .01091 .01045 .01073 .00982 .00906 .00833 .01083 .01002 .00869 .00768 .00690 .00624 .01132 .00967 .00900 .00844 .00748 .00671 .00610 .00559 .01299 .01183 .01085 .01004 .00933 .00871 .00769 .00728 .00690 .00624 .00570 .00523 .00486 .01044 .00969 .00902 .00844 .00793 .00749 .00673 .00641 .00610 .00559 .00515 .00479 .00446 .00710 .00673 .00641 .00612 .00584 .00559 .00515 .00497 .00479 .00446 .00417 .00394 .00372 .00597 .00572 .00548 .00526 .00506 .00488 .00454 .00439 .00425 .00399 .00376 .00356 .00337 6" .00470 .00454 .00439 .00425 .00412 .00399 .00377 .00367 .00356 .00337 .00321 .00307 .00294 12" .00388 .00377 .00367 .00357 .00347 .00337 .00321 .00314 .00307 .00294 .00281 .00269 .00259 18" .00352 .00343 .00334 .00327 .0031S .00310 .00296 .0029C .00282 .0027S .00261 .00255 .0024c Interaxial Distances. Size Cir. Mils Stranded. 6" 12" 18" 24" 36" 48" 60" 72" 1,000,000 .00655 .00506 .00446 .00412 .00370 .00347 .00330 .00318 900,000 .00641 .00497 .00439 .00405 .00367 .00343 .00327 .00314 800,000 .00626 .00488 .00432 .00399 .00361 .00337 .00323 .00310 750 000 .00617 .00483 .00428 .00396 .00359 .00336 .00321 .00308 700,000 .00608 .00478 .00423 .00392 .00356 .00334 .00318 .00307 600,000 .00590 .00466 .00416 .00385 .00350 .00328 .00312 .00301 500,000 .00570 .00454 .00405 .00376 .00343 .00321 .00307 .00296 450,000 .00559 . 00446 . 00399 .00372 .00337 .00318 .00305 .00294 400,000 .00548 .00439 .00392 .00367 .00334 .00314 .00301 .00290 350,000 .00535 .00430 .00386 .00361 .00328 .00310 .00296 .00287 300,000 .00521 .00421 .00379 .00354 .00323 .00305 .00292 .00283 250,000 .00504 .00410 .00376 .00347 .00318 .00299 .00287 .00278 0000 .00492 .00403 .00363 .00341 .00312 .00296 .00283 .00274 000 .00474 .00390 .00354 .00332 .00305 .00288 .00278 .00269 00 .00457 .00379 .00345 .00323 .00298 .00283 .00270 .00263 .00443 .00368 .00336 .00316 .00292 .00276 .00267 .00258 1 .00426 .00357 .00327 .00308 .00285 .00270 .00259 .00252 2 .00412 .00348 .00318 .00299 .00278 .00265 .00254 .00247 4 .00388 .00330 .00303 .00287 .00267 .00254 .00245 .00238 Solid 6 .00356 .00307 .00283 .00269 .00252 .00239 .00232 .00227 Solid 8 .00337 .00294 .00272 .00259 .00241 .00232 .00225 .00218 Solid 10 .00321 .00281 .00261 .00249 .00234 .00223 .00218 .00212 258 CONDUCTORS. Charging- Current in Amperes per Wire per lOOO Feet of Three-Phase Circuit formed hv Three Parallel Aerial Wires. Pressure between Wires, E = 10,000 Volts Frequency, /= 60 Cycles per Second. Charging Current per Wire = 4.356 C. Note. — Values of charging current at other pressures are propor- tional to those given in this table. 4d t/5 "5 Interaxial Distances. ^ o Dia. over Insul. 1" 2 4 1" 2" 3" 6" 12" 18* 3000 .03258 .03149 .03032 .02914 .02801 .02953 .02727 .02618 .02509 .02574 .02356 .02174 .01999 .03119 .02505 .02326 .02165 .02025 .01903 .01799 .01616 .01538 .01464 .01342 .01237 .01150 .01071 .01703 .01616 .01538 .01468 .01403 .01342 .01237 .01193 .01150 .01071 .01002 .00945 .00893 .01433 .01372 .01315 .01263 .01215 .01172 .01089 01054 .01019 .00958 .00902 .00854 .00810 .01128 .01089 .01054 .01019 .00989 .00958 .00906 .00880 .00854 .00810 .00771 .00736 .00706 .00932 .00845 000 .02840 .00906) .00823 00 .02605 .00880 .00801 o .02409 .00858 .00832 .00810 .00771 .00753 .00736 .00706 .00675 .00645 .00623 .00784 1 .02239 .00762 2 .02718 .02091 .01847 .01747 .01655 .01498 .01368 .01254 .01167 .00745 4 . 02322 .00710 5 6 8 10 12 14 .02600 .02404 .02086 .01842 .01655 .01498 .02160 .02025 .01795 .01612 .01464 .01342 .00697 .00679 .00653 .00627 .00605 .00584 Interaxial Distances. Size Cir. Mils Stranded. 6" 12" 18" 24" 36" 48" 60" 72" 2,000,000 .01847 .01372 .01189 .01089 .00971 .00906 .00858 .00823 1,500,000 .01721 .01298 .01137 .01045 .00936 .00871 .00828 .00793 1,250,000 .01651 .01259 .01106 .01019 .00915 .00854 .00810 .00780 1,000,000 .01572 .01215 .01071 .00989 . 00889 .00832 .00793 .00762 900,000 .01538 .01193 .01054 .00971 .00880 .00823 .00784 .00753 800,000 .01503 .01172 .01037 .00958 .00867 .00810 .00775 .00745 750,000 .01481 .01159 .01028 .00950 . 00862 .00806 .00771 .00740 700,000 .01459 .01147 .01015 .00941 .00854 .00801 .00762 .00736 600,000 .01416 .01119 .00997 .00923 .00841 .00788 .00749 .00723 500,000 .01368 .01089 .00971 .00902 .00823 .00771 .00736 .00710 450,000 .01342 .01071 .00958 .00893 .00810 .00762 .00732 .00706 400,000 .01315 .01054 .00941 .00880 .00801 .00753 .00723 .00697 350,000 .01285 .01032 .00928 .00867 .00788 .00745 .00710 .00688 300,000 .01250 .01010 .00910 .00849 .00775 .00732 .00701 .00679 250,000 .01211 .00984 .00902 .00832 .00762 .00719 . 00688 .00666 0000 .01180 .00967 .00871 .00819 .00749 .00710 .00679 .00658 000 .01137 .00936 .00849 .00797 .00732 .00693 .00666 .00645 00 .01098 .00910 .00828 .00775 .00714 .00679 .00649 .00632 .01063 .00884 .00806 .00758 .00701 .00662 .00640 .00618 1 .01024 .00858 .00784 .00740 .00684 .00649 .00623 .00605 2 .00989 .00836 .00762 .00719 .00666 .00636 .00610 .00592 4 .00932 .00793 .00727 .00688 .00640 .00610 .00588 .00571 Solid 6 .00854 .00736 .00679 .00645 .00605 .00575 .00557 .00544 Solid 8 .00810 .00706 .00653 .00623 .00579 .00557 .00540 .06523 Solid 10 .00771 .00675 .00627 .00597 .00562 .00536 .00523 .00510 ALTERNATING CURRENT CIRCUITS. 259 SI UIMLV ALTERNATING CVRREX1 CIRCUITS. The impedance (z) of a circuit is defined as the ratio of the difference in pressure (effective) between the two ends of the conductor to the current (effective) flowing through the conductor. The E.M.F. required to overcome impedance is E = Iz. In the case of direct currents z = r. The following are typical alternating current circuits: Let R = resistance in ohms. Z = impedance. w = 2 *7. L = coefficient of self induction. C = capacity. Resistance and Inductance, in Series. Z = Vfl2 + L2«2, or diagrammatically Fig. 12. Resistance and Capacity in Series. Z = or diagrammatically, Fig. 13. Resistance, Inductance, and Capacity in Series. z - y/jp + (x-- 5J. or diagrammatically, Note. — In transmission lines the capacity is in parallel with the resist- ance and inductance; the above formulae involving capacity do not there* fore apply. For the discussion of capacity of transmission lines see p. 264. 260 CONDUCTORS. THE DIMENSIONS OF CONDUCTORS FOR DISTRIBUTION SYSTEMS. By Harold Pender, Ph.D. To proportion properly the size of the conductors for a distribution system, the following data with regard to each circuit is necessary: 1. The maximum power to be transmitted, or the maximum load on the line. 2. The load factor, or the variation of the power delivered with time. 3. The length of the line. 4. The distribution of the load along the line. 5. The pressure at which the power is to be transmitted. 6. The loss of power which may be allowed in the line. These six conditions will determine a conductor of a definite cross sec- tion, but no conductor should ever be used which is not of sufficient size both to insure the proper mechanical strength and also to prevent a dan- gerous temperature elevation; the first condition is of particular impor- tance in overhead lines, the second in underground and interior wiring. Assuming that the amount and distribution of the load and the trans- mission distance are known, the engineer has next to determine what line pressure to employ and what power loss to allow. To do this, he must keep in mind two fundamental facts, namely, that the transmission system is but part of the entire plant, and that the object of the plant as a whole is to gain the maximum net revenue for the least expenditure of money; also, that there is usually a limit to the capital available for the enter- prise, which the first cost of the entire plant must not exceed, even though a further increase of the capital outlay might gain a desirable revenue. Consequently, in the selection of the pressure and efficiency for a distribu- tion system, many complex factors enter, such as the nature of the supply of energy, the nature of the load supplied, the probability of increase in the demand for power, etc., as well as the relative costs of the various parts of the plant. Space does not permit of a detailed discus- sion of all these factors here; it will suffice to state briefly the general Amer- ican practice under the most common conditions. lilME PRE§§€RE. — To transmit a given amount of power a given distance at a fixed efficiency, the amount of copper required will vary inversely as the square of the pressure. High pressure then means de- crease in the cost of the conducting material, but an increase in the cost of insulating the line and the rest of the system. As a general rule, espe- cially in longdistance transmission, the saving in copper as the pressure is increased more than offsets the increased cost of insulation, up to about 60,000 volts, but in many cases other factors fix a much lower economical limit to the line pressure. Recent improvements in the design of insula- tors accompanied by a decrease cost of manufacture have raised the economic limit of line pressure to 100,000 volts. Direct Current Distribution. — On direct current systems supply- ing directly incandescent lamps and small motors, the maximum pressure allowable is 125 volts for two-wire distribution, 250 volts for three-wire distribution; in certain cases where cheap power may be had, these figures may be increased to 250 and 500 respectively. For large direct current motor systems the corresponding figures are 500 to 600 volts for two-wire and 1000 to 1200 volts for three-wire systems. The limiting transmission pressure is fixed by the maximum pressure which can be employed on the various translating devices, motors, lamps, and the like. Future devel- opments in the latter may set a new limit to the allowable pressure; in fact, the compensating pole direct current motors now being placed on the market will permit the use of pressure as high as 1200 volts for two-wire and 2400 volts for three-wire systems. On circuits supplying direct cur- rent series arc lamps, pressures as high as 5000 volts are used. DIMENSIONS OF CONDUCTORS. 261 Alternating: Current Distribution. — The line pressure on that part of an alternating current distribution system connected directly to the various translating devices, motors, lamps, and the like, is fixed by the practicable pressure that may be used on these devices. For direct distribution for incandescent lighting, the line pressure between wires should not exceed 125 volts, or possibly 250 volts if power is cheap and 220 to 250 volt incandescent lamps can be advantageously employed. Distribution in Cities. — In the larger cities the tendency of modern practice (1907) is to generate three-phase alternating current at 11,000 or 13,000 volts (delta), and to transmit the power at this pressure either to static transformer or rotary converter sub-stations. For the dis- tribution of direct current from rotary converter sub-stations see above under "Line Pressure for Direct Current Distribution." At the static transformer sub-stations the pressure is reduced to 2200 volts, and the power transmitted at this pressure to the centers of distribution, where another reduction in pressure to about 125 or 250 volts takes place, and from here the energy is distributed directly to the lamps, motors, or other translating device. In smaller cities, or when it is desired to employ overhead lines entirely (since 11,000 volts overhead in cities is not advis- able), the sub-stations may be omitted and generators for 2200 volts be used. Large induction motors may be supplied directly with 2200 volt current, the very largest sometimes with current at 11,000 or 13,000 volts. POWER LOi§ I]¥ THE MWJE. — To transmit a given amount of power a given distance at a given pressure, the amount of copper required will vary inversely as the amount of power lost in transmission. Low efficiency, therefore, means decrease in the cost of the conducting material, but an increase in the central station output. Kelvin's law, — In general, if two quantities A and B are both func- tions of the same variable x, then the sum of A + B is a minimum when the rate of change of A with respect to that variable is equal and opposite to the rate of change of B with respect to that variable, i.e., when dA = _dB t dx dx Numerous attempts have been made to apply this law to the determi- nation of the most economical efficiency for a transmission line. At first sight it would seem logical to proportion the costs of the central station and transmission line so that the annual cost of delivering an additional kilowatt of power by increasing the central station capacity will equal the annual cost of delivering an additional kilowatt of power by adding more copper to the line. On this basis a very simple law is found to hold, namely, that the most economical current density per million circular mils is * /Kc Kp* where Kc = increase in annual charges on transmission line, resulting from increasing the weight of copper one ton (2000 lbs.), and Kp = increase in annual operating and capital charges on the central station, resulting from increasing the output one kilowatt. This law, however, is true only for a given current; when the power sup- plied by any plant, and therefore the current, varies over wide limits during the year, as is almost invariably the case, the current density as determined by the above law refers to the square root of the mean square current for the year, a quantity which can be determined only to the roughest approximation. Further, the whole discussion of economical cross section is based on two assumptions, usually unwarranted, namely, that the amount of capital available is unlimited, and that a market can be found for the maximum output of the plant; it will evidently not be economical to install copper to save power which cannot be sold. In short, neither Kelvin's law nor 380 y/^ The formula for aluminum is 165 V K, 262 CONDUCTORS. any modification of it is a safe general guide in determining the proper allowance for loss of power in the line. Each plant has to be considered on its individual merits, and various conditions are likely to determine the pressure and loss in different cases. Distribution Direct to translating: Devices. — The power loss in a transmission line also fixes the pressure loss or volts drop. In direct current systems the per cent power loss equals the per cent pressure loss; in an alternating current line there is also a fixed relation between the two, see page 264. In that part of a distribution system connected directly to the translating devices, lamps, motors, etc., the regulation of the line, or the percentage pressure loss, must not exceed a certain amount con- sistent with reasonably" efficient operation of these translating devices. For example, the maximum variation in pressure on incandescent lamps should not be more than 2 per cent; distribution lines which supply incan- descent lamps and on which the pressure at the sending end is fixed, should therefore be of sufficient size to insure a pressure loss of not over 2 per cent at maximum load. When a line supplies a large number of lamps, all of which are not likely to be burning simultaneously, the per cent drop in pressure for the connected load may be taken considerably greater. For example, if the probable maximum load be figured at one third of the connected load, a drop of 6 per cent for all lamps burning may be allowed. Distribution in General. — The following discussion of the proper power loss to allow in transmission lines is taken from Bell, "Electric Power Transmission." "The commonest cases which arise are as follows, arranged in order of their frequency as occurring in American practice. Each case requires a somewhat different treatment in the matter of line loss, and the whole classification is the result not of a priori reasoning but of the study of a very large number of concrete cases. Case I. General distribution of power and light from water-power. This includes something like two thirds of all the power transmission enter- prises. The cases which have been investigated by the author have ranged from 100 to 20,000 H.P., to be transmitted all the way from one to one hundred and fifty miles. The market for power and light is usually uncer- tain, the proposition of power to light unknown within wide limits, and the total amount required only to be determined by future conditions. The average load defies even approximate estimation, and as a rule even when the general character of the market is most carefully investigated little certainty is gained. For one without the gift of prophecy the attempt to figure the line for such a transmission by following any canonical rules for maximum econ- omy is merely the wildest sort of guesswork. The safest process is as fol- lows: Assume an amount of power to be transmitted which can certainly be disposed of. Figure the line for an assumed loss of energy at full load small enough to insure good and easy regulation, which determines the quality of the service, and hence, in large measure, its growth. Arrange both power station and line w T ith reference to subsequent increase if needed. The exact line loss assumed is more a result of trained judgment than of formal calculation. It will be in general between 5 and 15 per cent, for which losses generators can be conveniently regulated. If raising and lowering transformers are used the losses of energy in them should be in- cluded in the estimate for total loss in the line. In this case the loss in the line proper should seldom exceed 10 per cent. A loss of less than 5 per cent is seldom advisable. It should not be forgotten that in an alternating circuit two small con- ductors are generally better than one large one, so that the labor of installa- tion often will not be increased by waiting for developments before adding to the line. It frequently happens, too, that it is very necessary to keep down the first cost of installation, to lessen the financial burden during the early stages of a plant's development. Case II. Delivery of a known amount of power from ample water- power. This condition frequently arises in connection with manufactur- ing establishments. A water-power is bought or leased in toto, and the problem consists of transmitting sufficient power for the comparatively fixed needs of the works. The total amount is generally not laree seJdorq DIMENSIONS OF CONDUCTORS. 263 more than a few hundred horse-power. Under these circumstances the plant should be designed for minimum first cost, and any loss in the line is permissible that does not lower the efficiency enough to force the use of larger sizes of dynamos and water-wheels. These sizes almost invari- ably are near enough together to involve no trouble in regulation if the line be thus designed. The operating expense becomes practically a fixed charge so that the first cost only need be considered. Such plants are increasingly common. A brief trial calculation will show at once the conditions of economy and the way to meet them. Case III. Delivery of a known power from a closely limited source. This case resembles the last, except that there is a definite limit set for the losses in the system. Instead, then, of fixing a loss in the line based on regu- lation and first cost alone, the first necessity is to deliver the required power. This may call for a line more expensive than would be indicated by any of the formulae for maximum economy, since it is far more impor- tant to avoid a supplementary steam plant entirely than to escape a con- siderable increase in cost of line. The data to be seriously considered are the cost of maintaining such a supplementary plant properly capitalized, and the price of the additional copper that render it unnecessary. Maxi- mum efficiency is here the governing factor. In cases where the motive- power is rented or derived from steam, formulae like Kelvin's may some- times be convenient. Losses in the line will often be as low as 5 per cent, sometimes only 2 or 3. Case IV. Distribution of power in known amount and units, with or without long distance transmission, with motive-power which, like steam or rented water-power costs a certain amount per horse-power. Here the desideratum is minimum cost per H.P., and design for this purpose may be carried out with fair accuracy. Small line loss is generally desirable unless the system is complicated by a long transmission. Such problems usually or often appear as distributions only. Where electric motors are in competition with distribution by shafting, rope transmission, and the like, 2 to 5 per cent line loss may advantageously be used in a trial com- putation. The problem of power transmisson may arise in still other forms than those just mentioned. Those are, however, the commonest types, and are instanced to show how completely the point of view has to change when designing plants under various circumstances. The controlling element may be minimum first cost, maximum efficiency, minimum cost of trans- mission, or combinations of any one of these, with locally fixed require- ments as to one or more of the others, or as to special conditions quite apart from any of them. In very many cases it is absolutely necessary to keep down the initial cost, even at a considerable sacrifice in other respects. Or economy in a certain direction must be sought, even at a considerable expense in some other direction. For these reasons no rigid system can be followed, and there is constant necessity for individual skill and judgment. It is no uncommon thing to find two plants for transmitting equal powers over the same distance under very similar conditions, which must, however, be installed on totally different plans in order to best meet the requirements." 264 CONDUCTORS. CALCULATION OF TRANSMISSION LINUS. Harold Pender, Ph.D. Let E = pressure between adjacent wires at receiving end in volts. W = power delivered in kilowatts. £ = power factor of the load expressed as a decimal fraction. A = cross section of each wire in millions of circular mils. w = total weight of conductors in pounds. I = length of circuit (length of each wire) in feet. R = resistance of each wire in ohms. U = reactance factor of line = ratio of line reactance to line resistance (Table II). Q = per cent power loss in terms of delivered power. p = per cent pressure drop in terms of delivered pressure. Put F = M- * (kE) 2 In Table I are given formulae for calculating the cross section, weight, and power loss for any kind of conductor. The per cent pressure drop, P, can be readily calculated when the per cent power loss is known by means of the formula P = MQ + NQ 2 . Where M and N are constants depending on the pov/or factor (/;) and the ratio t\ of the line reactance to the line resistance, tine ratio is called the "reactance-factor"; Tables III and IV cive the values of the constants II and N for various values of k and t x . To a close approximation, except when the power factor is nearly unity, or the receiver current is leading, the term NQ 2 may be neglected, i.e., in most practical cases P = MQ. The complete expression P = MQ + NQ 2 is exact in all cases for a 10 per cent power loss; it is in error less than 3 per cent for any value of P less than 30; in any case likely to arise in practice the discrepancy is less than 1 per cent in the value of P. The exact expression for P in terms of is P = Vi()4 4- 200 (1 4- th) k 2 Q + (1 + h 2 ) k 2 Q 2 - 100 where t is the tangent corresponding to the cosine k. (See p. 276.) Effect of Line Capacity. The effect of the capacity of the line is to reduce the pressure drop, i.e., improve the regulation, and to decrease or increase the power loss depend- ing on the load and power factor of the receiver. Let b = 2 nfC X 10 -6 . Where C is the capacity of the condenser in microfarads formed by any pair of wires of the line, / is the frequency; b is called the capacity susceptance of the line (for a single-phase line, the charging current is bE; for a three- phase line the charging current per wire is 1.155 bE. Table V gives the values of the capacity susceptance per 1000 feet of circuit for various sizes of wire spaced various distances apart for a frequency of 100 cycles per second; the values for other frequencies are directly pro- portional. (Continued on p. 270.) CALCULATION OF TRANSMISSION LINES. 265 - u \ i 8 « t S c « c fa V fa 5 e h I M I — ; x . .2 S S ^ 03hW 73.5 9'S ^ • >»?ftll § c II *o ft, a CO 10 0- ca. CM QP 3 CO ft, Q. CO 10 ^ CO 00 CM -a <3 o- i fen K 3 ft, fc Ph o 9§>^g» l> CO Of 00 QP l> CO ^ ^ -^ ^ Tji CM <-* O H CO o a j^ fc"§ ^ '"§ » fa ft, K 3 ft, Ph O ."tf CO 12 ° O o» 10 O? ^ © CM o o 1— 1 _. „ "* 9 • ea w* ft, s 1 3 ft, a. «o 1 «© Q. co CO O o> 00 |Q» OS 00 CO O ^ ■4 Ph >> 7 2 n s C II P. " o CO 10 1 f* CO CO ** u CO CO 5 co QP 3 8 1— 1 ft, CO CO ^ X C 3 <% N ° w o CO 1 0) s . 3 .4, Qhq-mCOTSo & §, £c>gg g 8 ft, 00 a> co CM C* CO CO ft, 00 CM ^ •H 1, || II II II Ttj § 9 O^ Q> CO CO ,Q jO la ' CO §S .9 5S ja t ^3 a CO • *S c?l O O -^ • A T2 O T3 s 3 T3 ft* • &'•£ CT3 a"2 G „ 03^ 03 fl 03 co ^h to ~ CO ^9 . O * e3 O o3 > O ► "S >^ > fe > - "S in "2 © .U O • -H CO •- CO 1 00 OH OH OPh • OPh 266 CONDUCTORS. 0> $ nO £ T^o g fa 09 00 ooo co oo oo OS CD CO fr- t- t- «-< CD © © © Co 00 00 00 o©© © © 00 i-i CO © 00 t- ©©00 t-COlO Tj«COCO C0 o io co © ©00 CO i-© CD CD CO t-' CO O 8 §8 2 "<*' co CO 838 CO CO th" COt- t» © CO © 00 t- © ©C<1 t~ CO CO CO 00 00 © cd t- © SS© 1 888 s 0J c fa O OS 00 t>© O ■«* co co COCOi-l s IO coco — OOh 00"* 00 t-i CM CO O CD Ci O'tN rf< © O © CO CO © CO ©CO © ©©00 t-© O t CO CO COr-lr-t *3 s -# Ci Ci 00 CO Ci CD t- CO© CO O 00 ,-H t- ^H CO 00 O CM © © Tf CO © CO © < 0) c e 05 00 t> CO LOO rf CO 0^ «HH so +a 05 s CO © t-O © ~ co CO t- -* OO 00 rf O CO © 00 ** © -*< © ~* rtfllb. © 00 00 t- CDUO'* Tf CO co CO i-h i-i w " i co CO lOTf CO CO CM (Ni-ih •"' - OO CO © CO t- ©CD© 00 00 00 CO ©^ Ot-C0 •<-> © t> © O ©©O o-* co COCO CO ,-<^^ aO .3 c 3 3 5 COt^T-H OOO r-cooo r-oco Tt^iOCO OOO oo CiCiO r^OicN OOrH 00OCM OOO oooo ^oo CO-CNO CO BD u a> a a o O NOOl CO COIN OOO COO-* OiTf^ aJ a> fr^9 gtxico ro *« 5^ s o o t^COiO o© oo_ ooo ^©- 50 _ HO oooo o© oocoto CO 000000 00 ooo o ooo o ooo o ooo o ooo o ooo o ooo o ooo o Q* A r-^T-KN MNO O0»-0 iCOSOO 05^-<* (M Ot^CO 05^0 CO i-l CO p — drop ) Decrease in per) ^ = cent power loss ) 50 bX at ~2WQ 100 6* where a = 100 bR and t is the tangent corresponding to the cosine k. (See p. 276.) The true regulation of the line is then P — p, and the true per cent power loss is Q — q, P and Q being calculated by the formulae given on pages 264 and 265. These formulae are approximate, being deduced on the assumption that the line capacity can be represented by a condenser of half the capacity of the line shunted across the line at each end, but they are sufficiently accurate for any case likely to arise in practice. It is to be noted that the change in regulation is independent of the load and the power factor, and is independent of the line resistance; the change in the per cent power loss varies with both the load and the power factor. Direct Current, Three-Wire System. — Figure the weight and cross section of the outer conductors as if the middle or neutral wire was not present, putting E = volts between outside wires. The neutral wire is usually taken from one- third to full size of each outer conductor. The total weight of copper required will therefore be one-sixth to one-half greater than the weight determined by the above formula. Two-Phase, Tour- IFire System. — Treat each phase separately, remembering that half the power is delivered by each phase, and E =■ volts between diametrically opposite wires. Two-Phase, Three-Wire System. — Let E V = pressure between each outer and middle wire at receiving end in volts. = pressure between each outer and middle wire at generating end in volts. Other symbols as above. Then for equal rise of temperature in the three conductors the following formulae hold. (The total weight of conductor required for this condition is only a fraction of one per cent greater than for the condition of maximum economy.) Cross section of each ) outer wire in million \ A x = CM. ) Cross section of middle ) a __ wire in million CM. \ A0 Total weight in pounds w = Total weight in pounds w = Copper. 100 % conduc- tivity. 20° Centigrade or 68° F. 0.93F Q 1.26^! 9.85/^! 9.15ZF Aluminum. 62 % conduc- tivity. 20° Centigrade or 68° F. 1.50F Q 1.26^! 2.97L4, 4A51F Any Material, p = microhms per cu. in. 5 = lbs. per cu. in. 1.37pF Q 1.26^! 30. 7 81 A t 42.1pSlF On the B. & S. gauge the middle wire is larger than each outer by one number (see p. 145). NUMERICAL EXAMPLES OF CALCULATIONS. 271 Two or more Circuit* in iSeries. The above formulae and tables are also applicable to the case of two or more circuits in series, i.e., a transmission line and transformer, if we put R = Ri + R2 4- . . . . rp = Rlt\ + ^2*3 , R I----- where R\, R 2 , etc., are the resistances of the separate circuits and t\, £2. etc. : are the reactance factors of the separate circuits. NUMERICAL EXAIftEPJLKS OF CAICITATIO^, OJP WEIGHT, CROSS IECTIOKT, ETC. Direct Current, Two-Wire System. Copper Wirhs. Given W = 40 kilowatts. E = 200 volts. I = 500 feet. Q = 5 per cent. „n. rr 500 X 40 Then F~ -7200)^ = ' 5 - 2 08 X 5 Cross section A = — — = 0.208 million CM. o The nearest commercial size is No. 0000 B. & S. (see Table II) which hat an area of 0.212 million CM. Total weight of copper w = 6.06 X 500 X 0.212 = 641 pounds. r> 1 r> 208 X0.5 Power loss Q = — — =4.92 per cent. Pressure drop P = Q = 4.92 per cent. Pressure at generating end = 1 . 0492 X 200 = 209 . 84 volts. IMrect Current, Three- Wire System. Take the same constants as in the preceding case, considering E = 200 volts as the pressure between outer wires. If the neutral wire is to be half the size of each outer, the total weight of copper required will be 641 + ?|i = 801 pounds. When the system is balanced there will be no current in the neutral wire and the regulation and efficiency will be the same as above. If one side of the system is fully loaded, and the other side not loaded at all, the volts drop in the loaded outer will be the same as if the system was balanced, since the same current flows, and the volts drop in the neutral will be twice the drop in the outer (same current and double resistance); hence total drop will be 14.8 volts in 100 volts or 14.8 per cent. The power loss will also be 14.8 per cent or 2.96 kilowatts. 272 CONDUCTORS. Alternating* Current, Single Phase. Copper Wires Spaced 3 Feet Apabt. Given f = 25 cycles per second. W — 500 kilowatts. E = 10,000 volts. I = 45,000 feet. k = 0.9, i.e., 90 per cent power factor. Q = 10 per cent. ™-„ *» - 45,000 X 500 _ Q Then F ~ (0.9 X lO.OOO)* = °- 278 * 2.08X0.278 n A __ .... _„ Cross section ^4 = r^r = . 0578 million CM. The nearest commercial size is No. 2 B. & S. (Table II), which has an area of 0.0664 million CM. Total weight of copper w = 6.06 X 45,000 X 0.0664 = 18,100 lbs. tt i ^ 2.08 X 0.278 Exact power loss Q = — ~ ~ aaA — =8.71 per cent. U . Uoo4 1 44 Reactance factor t x = -~— = 0.36. (Table II). Therefore M = 0.95 (Table III). N = 0.000. (Table IV). Then, neglecting the capacity of the line, Pressure drop P = 0.95 X 8. 71 = 8.27 per cent. Pressure at generating end = 1.0827 X 10,000 = 10,827 volts. Two-Phase, Three-Wire System. Copper V/ires Spaced 3 Feet Apart. Given / = 25 cycles per second. W = 500 kilowatts. E = 10,000 volts. I = 45,000 feet, k =0.9, i.e., 90 per cent power factor. Q = 10 per cent. Then F 45,000 X 500 (0.9X10,000) 2 u -^'°- 93 X 278 Cross section of outers A t = — -^ = 0.0259 million CM. The nearest commercial size is No. 6 B. & S. (Table II) which has an area of 0.0263 million CM. The middle wire must therefore be No. 5 B. & S. Total weight of copper w = 9.85 X 45,000 X 0.0263 = 11,600 lbs. tt * i n °- 93 X0.278 _ _ Exact power loss Q = — n = 9.87 per cent. The pressure loss will depend upon how the wires are arranged on the poles. As a first approximation for any ordinary arrangement, the reac- tance of each phase can be considered the same as in a single phase system with wires of the same cross section as the outer, spaced a distance apart equal to that between each outer and the middle wire. From Table II the reactance factor of a No. 6 wire corresponding to a three-foot spacing and 25 cycles is d- ^1=0.15. 4 Whence M = 0.87. NUMERAL EXAMPLES OF CALCULATIONS. 273 Then neglecting the capacity of the line, and using the approximate formula P = MQ, Pressure drop P = 9.87 X 0.87 = 8.59 per cent. Pressure at generating end = 1.0859 X 10,000 = 10,859 volts. Alternating: Current, Three-Phase. Copper Wires Spaced 6 Feet Apart. Given / =60 cycles per second. W = 10,000 kilowatts. E = 60,000 volts. I = 400,000 feet. k = 0.85, i.e., 85 per cent power factor. Q = 12 per cent. „ 400,000 X 10,000 , . . rhen F = (0.85X60,000)' = 1M - Cross section A = — — — : — = 0. 133 million CM. The nearest commercial size is No. 00 (see Table II), which has an area of 0.133 million CM. Total weight of copper w = 9 .09 X 400,000 X 0. 133 = 484,000 lb. Neglecting line capacity, T7 * i n 1-54 X 1.04 10 , Exact power loss Q = — _ „ = 12 per cent. Reactance factor t t = 3 . 06 X . 6 = 1 . 84. Therefore M = 1 . 55. N = 0.003. Pressure drop P = 1.55 X 12 + [0.003 X (12) 2 ] = 19.0. Effect of line capacity (see p. 264). b = .00000089 X 0.6 X 400 = 0.000214. (Table V). R =0.0778X400 = 31.1 (Table II). X = 1.84 X 31.1 = 57.2. Then Decrease in per cent pressure drop = p = 100 X 0.000214 X 57.2 = 1.2. a =100X0.000214X31.1=0.67. t = 0.62. Decrease in per cent power loss = q =2x0. 67X0. 62- ( Q ' N9 ' =0.8. (U. oo)^ Xl^ Whence True pressure drop = 19.0 — 1.2 = 17.8 per cent. True power loss = 12.0 — 0.8 = 11.2 per cent. Pressure at generating end = 1 . 178 X 60,000 = 70,680 volts. 274 CONDUCTORS. tranSiIKissioh mi: ojf known constants. The following formulae and tables give an exact method of calculating the efficiency and regulation of a transmission line of known constants, in terms of the pressure between adjacent wires at the generating end of line. Given: The kind of system, direct or alternating, n = number of phases, for the " single phase " system n — 2. f = frequency in cycles per second. V — pressure between adjacent wires at generating end, in volts. W = power delivered in watts. • cos a. = power factor of load at receiving end. R — resistance of each wire in ohms. X = inductive reactance of each wire in ohms. Z = \/R 2 + X 2 = impedance of each wire. Required: E = pressure between adjacent wires at receiving end in volts. / = current per wire in amperes. H = total power lost in watts. The values of E, I, and H are given in the table on p. 275. For approx- imate calculations J can be taken equal to unity; the exact value of J is given in the table below. Values of .1. .00 .01 .02 .03 .04 .05 000 1.0000 1.0001 1.0004 1.0009 1.0016 1.0025 .001 1.0000 1.0001 1.0004 1.0010 1.0017 1.0026 .002 1.0000 1.0001 1.0005 1.0010 1.0017 1.0027 .003 1.0000 1.0002 1.0005 1.0011 1.0018 1.0028 .004 1.0000 1.0002 1.0006 1.0012 1.0019 1.0029 .005 1.0000 1.0002 1.0006 1.0012 1.0020 1.0030 .006 1.0000 1.0003 1.0007 1.0013 1.0021 1.0031 .007 1.0000 1.0003 1.0007 1.0014 1.0022 1.0032 .008 1.0001 1.0003 1.0008 1.0014 1.0023 1.0034 .009 1.0001 1.0004 1.0008 1.0015 1.0024 1.0035 e .000 .002 .004 .006 .008 e .000 .002 .004 .006 .008 .06 1.004 1.004 1.004 1.004 1.005 .29 1.102 1.104 1.106 1.108 1.110 .07 1.005 1.005 1.005 1.006 1.006 .30 1.111 1.113 1.115 1.117 1.119 .08 1.006 1.007 1.007 1.007 1.008 .31 1.121 1.123 1.125 1.127 1.129 .09 1.008 1.008 1.009 1.009 1.010 .32 1.131 1.133 1.135 1.137 1.139 .10 1.010 1.010 1.011 1.011 1.011 .33 1 . 141 1.143 1.146 1.149 1.151 .11 1.012 1.012 1.013 1.013 1.014 .34 1.154 1.156 1.158 1.161 1.163 .12 1.014 1.015 1.015 1.016 1.017 .35 1.167 1.169 1.171 1.174 1.177 .13 1.018 1.018 1.019 1.019 1.020 .36 1.180 1.183 1.186 1.189 1.192 .14 1.021 1.021 1.022 1.022 1.023 .37 1.195 1.199 1.202 1.206 1.209 .15 1.024 1.024 1.025 1.025 1.026 .38 1.213 1.216 1.220 1.224 1.227 .16 1.027 1.027 1.028 1.029 1.030 .39 1.231 1.234 1.238 1.242 1.246 .17 1.031 1.032 1.032 1.033 1.034 .40 1 . 250 1.254 1.258 1.263 1.267 .18 1.034 1.035 1.03G 1.037 1.038 .41 1.272 1.276 1.280 1.285 1.290 .19 1.039 1.040 1.041 1.042 1.043 .42 1.296 1.301 1.307 1.312 1.318 .20 1.044 1.045 1.046 1.046 1.047 .43 1.324 1.330 1.336 1.342 1.349 .21 1.048 1.049 1.050 1.051 1.052 .44 1.356 1.363 1.370 1.377 1.385 .22 1.053 1.054 1.05G 1.057 1.058 .45 1.393 1.401 1.410 1.409 1.428 .23 1.059 1.0G1 1.062 1.063 1.065 .46 1.437 1.447 1.457 1.468 1.479 .24 1.066 1.067 1.068 1.070 1.071 .47 1.491 1.504 1.518 1.532 1.547 .25 1.072 1.074 1.075 1.076 1.078 .48 1.563 1.580 1.599 1.620 1.643 .26 1.079 1.081 1.082 1.083 1.084 .49 1.668 1.697 1.733 1.778 1.835 .27 1.086 1.087 1.089 1.090 1.092 .50 2.000 .28 1.094 1.096 1.098 1.099 1.100 TRANSMISSION LINE OF KNOWN CONSTANTS. 275 I Q Z I ^_^ >TN >"V >^s b 1 8 ti <3 a B j3 a tq >^k CO w fc 93 c fe aj . co 8k 3 "3 CO O fefa K5 3! K3 Bq &3 V—-' v^^.^ "*»•— "^ ^»—^' s — — '"' « &3 A3 ft^ ft^J(N CM CM 0$| 8 $ ti * fe CO 8 •*H fe|K|fc C5 B C Q fc DO C =r CO O co o Ol > K3 X b 1 8 ^ ICO > g 8 CM fcq |05l^|Qq|S|cq|^!Qq|S|QqiS lOQI^s ^^^* ""^J ->- ^> tsi « X U S3 a DQ a CO O © a) (D 0) 0) CD o 1 1 - t- 1 *o CM m CD CO CO O CO 03 H 9 Ph rP -3 J§ J5 ftn ftH Ph PL. Q rH CM CO "* 8 P. 50 >> a> D1 a a; OQ 5 a -i X -r u d 9) *s Cv ^ d u 0) XJ i— ^' ,s § a W B3 +3 a> . rr, •"" a [> h * ^ 276 CONDUCTORS. 0.000 100 50.0 33.3 25.0 20.0 16.6 14.3 12.5 11.1 9.95 9.03 8.27 7.63 7.07 6.59 6.17 5.80 5.47 5.17 4.90 4.66 4.43 4.23 4.05 3.87 3.71 3.57 3.43 3.30 3.18 3.07 2.96 33 2 86 0.34 2.77 0.35 2 2.59 2.51 2.43 2.36 2.29 2.22 2.16 2.10 2.04 1.98 1.93 1.88 1.83 1.78 0.002 500 83.3 45.4 31.2 23.8 19.2 16.1 13.8 12.2 10.8 9.75 8.87 8.14 7.51 6.97 6.50 6.09 5.73 5 40 5.11 4.85 4.61 4.39 4.19 4.01 3.84 3 3.54 3.40 3.28 3 16 3.05 2.04 2.84 2.75 2 2.58 2.50 2.42 2 35 2.28 2.21 2.15 2.09 2 03 1 1 1.87 1.82 1 77 0.004 250 71.4 41.6 29 4 22.7 18.5 15.6 13.5 11.9 10.6 9.56 8.71 8.00 7.40 6.87 6.42 6.02 5.66 5.34 5.06 4.80 4.56 4.35 4.15 3.97 3.81 3.65 3.51 3.38 3.25 3.13 3.02 2.92 2.82 2.73 2.64 2 56 2.48 2.40 2.33 2.26 2.20 2.14 2.08 2.02 1.96 1.91 1.86 1.81 1 76 0.006 167 62 5 38 4 27 7 21 7 17.8 15 1 13.1 11.6 10.4 9 38 8.56 7.87 7 28 6.78 6.33 5.94 5.59 5.28 5 00 4 75 4 52 4 31 4.12 3.94 3.78 3.62 3.48 3.35 3.23 3.11 3.00 2.90 2 2 71 2.63 2.54 2 46 2 39 2 32 2.25 2.19 2.12 2.06 2.01 1.9! 1.90 1.85 1. 1 75 0.008 125 55 5 35 7 26 3 20 8 17 2 14 7 12 8 11 3 10 2 9 21 8.41 7 75 7 18 6 68 6 25 5 5.53 5.22 4 95 4 70 4 4 4. 3 91 3 75 3.59 3 46 3 33 3.20 3.09 2 2.88 2 78 2 2 61 2.53 2.45 2.38 2.30 2.24 2.17 2.11 2.05 2.00 1.94 1. 1.84 K 50 51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.60 0.61 0.62 0.63 0.64 0.65 0.66 0.67 0.68 0.69 0.70 0.71 0.72 0.73 0.74 0.75 0.76 0.77 0.78 0.79 0.80 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 90 0.91 0.92 93 0.94 0.95 96 97 1.732 1.687 1.643 1.600 559 1.519 479 1.442 1.404 .368 .333 1.299 265 1.233 1.201 1.169 138 1.108 0.000 1.078 1.049 1.020 0.992 0,964 0.936 0.909 0.882 0.855 0.829 0.802 0.776 0.750 0.724 0.698 0.672 646 0.620 0.593 0.567 0.540 0.512 0.489 0.456 426 0.395 0.363 0.329 0.292 0.251 0.002 1.723 1 678 1.634 1 592 550 1.511 1.471 1 434 1.397 1.361 1 . 326 1 . 292 1.259 1.226 1.194 1.163 1.132 1.102 1.072 1.043 1.015 0.986 958 0.931 0.904 0.877 0.850 0.823 0.797 0.771 0.745 0.719 0.693 0.667 0.641 0.614 0.588 0.5G1 0.534 507 0.479 0.450 0.420 0.389 0.356 0.321 0.284 0.242 1 714 1.669 1.626 1.79 98 0.203 0.192 181 169 156 1.74 99 0.143 0.127 0.110 0.090 0.063 0.004 583 542 503 464 427 390 354 319 "286 1 252 1 220 1.188 1.157 1.126 1.096 1 067 1.037 1.009 981 0.953 0.925 0.898 0.871 0.845 0.818 0.792 0.766 0.662 0.635 0.609 0.583 0.556 0.529 0.501 473 0.444 0.414 0.383 0.350 0.314 0.276 0.232 0.006 1.705 1.660 1.617 1.575 1.534 1 495 1.457 1 419 1.383 1.347 1 313 1.279 1.246 1.213 1.181 1.151 1.120 1.090 1 061 1 032 1.003 0.975 947 0.920 0.893 0.866 0.839 0.813 0.787 0.760 0.734 0.708 0.682 0.656 0. 630 0.604 0.577 0.551 0.523 0.496 0.467 0.438 408 0.376 0.343 0.307 0. 268 0.223 Note: This table is to be used like a table of logarithms, e. g. % the reac- tance factor corresponding to the power-factor k = 0.816 is 2 = 708. TRANSMISSION LINE CALCULATIONS. 277 PARALLEL DISTRIBUTION. When the translating devices, whether lamps or motors, are scattered over a considerable area, the ucuai method of supplying them with power is to run a single feeder to some point near the "center of gravity" of the load, and from this center run out branches to feed groups of lamps or motors in parallel. The center of gravity of the load can be readily deter- mined as follows: Let itfi, v)2, 103, etc. represent the individual loads, and x it X2, £3. etc and y u 2/2, 2/3. etc., represent the distances of these loads from any two fixed lines OX and OY at right angles to each other. Then the center of gravity is that point which is the distance = XlWl+X2W2 + X3W , + ... from0X w x + w 2 + w z + and y = 2/^+2/2^2+2/3^3 + ... from 0Y Wl + W2 + W 3 + . . . The center of gravity of the load is by no means always the most economi- cal location for the center of distribution, as considerations of the relative cost of establishing the center at this point in comparison with the cost at other points, the probable change in the distribution of the load with the growth of the system, etc., have all to be taken into account. The general scheme of feeders, centers of distribution, and branches can be developed still further, and sub-centers, sub-feeders, etc., estab- lished, until a point is reached where the saving in the cost of copper is balanced by the increase in the cost of the centers of distribution. Calculation of Cross Section, Weight, &c. When a transmission line is loaded at more than one point, the conductor should have such dimensions that the pressure drop at the end of the line, when the line is supplying the maximum load at each point, shall not exceed a given amount. Whether the conductor shall be made of uniform section throughout the length of the line, or be reduced in size as the current carried diminishes, will depend on the relative amounts of energy sup- plied at, and the distances between, the various points at which the line is loaded. Below will be found formulae for determining the weight and cross section of a line of uniform cross section, and having no reactance, supplying a distributed load. When the line has no inductive reactance the weight and cross section of the conductor for a given pressure drop are to a close approximation independent of the power factor of the loads at the various points. When the line has reactance, the formulae will give only a first approximation to the correct weight and cross section. The error involved can be determined by considering each section of the line separately, and calculating the drop in each section, assuming the dimen- sions given by the approximate formulae. (See page 264.) If the pressure drop at the end of the line thus calculated differs considerably from the permissible drop given, choose a larger size wire and make another trial calculation, etc., until the proper size is found. 278 CONDUCTORS. 1} <— It- Fig. 15. In the figure let G be the generating end of the line ; J the far end of line Given: E = k, h P = Required: A = w = Put = pressure between adjacent wires at far end of line in volts. W 2 , Wz, etc., the loads in kilowatts at the points 1, 2, 3, etc. , lz, etc., the distances of these points from the generating end in feet. per cent pressure drop at far end of line in terms of delivered pressure. cross section of each wire in million CM. total weight of conductors in pounds. W = w t + W2 + Wz + . . . total power delivered in kilowatts. I = l L + l 2 + lz + . . . total length of circuit (length of each wire) in feet. F - hW t + hW 2 + hWz + . . . E* Then, for a line having no reactance : Copper. 100% conduc- tivity. 20° Centigrade. Aluminum. 62% conduc- tivity. 20°Centigrade. 3.34F Any Material. p = microhms per cu. in. 5 — lbs. per cu. in. Single Phase. Cross section in million C M A = w = w = A = w = w = 2.0SF 3.06pF Total weight of conduc- tors Or total weight of con- ductors Three Phase. Cross section in million CM Total weight of conduc- tors Or total weight of con- ductors P 6.06L4. 12. 6FI P 1 . 83L4 6.11FZ P 18.9MA 57. 881F P 1.04F P 1.67F P 1.53 P F P 9.09L4 9.48FZ P 2.741A 4.58*7 P 28.3SL4 43. 2 P 81F P P P TRANSMISSION LINE CALCULATIONS. 279 Economical Tapering* of Conductor. When the distances between the points at which the line is loaded are considerable, it is usually advantageous to taper the conductor; the most economical pressure drop per section must be determined, and each section of the line calculated independently. The following formulae give the most economical division of the drop, taking into account the cost both of conductor and insulation. For short runs the saving in cost of con- ductor and insulation may be more than offset by the extra cost of handling two or more sizes of wire. The same notation as in the preceding paragraph is used. In addition, let U t = W x -f- W 2 -h Wz + • • • = total load in kilowatts at and beyond point 1 . JJ 2 = W 2 + TF3 + . . . = total load in kilowatts at and beyond point 2. U3 = W% -+-.•• = total load in kilowatts at and beyond point 3. 3tC. A x = Zj = distance in feet from generating end to point 1. *2 =■= h — h = distance in feet between points 1 and 2. ^3 = h ~ h = distance in feet between points 2 and 3. etc. Thea the most economical per cent pressure drop for the ith section is Pi== XiV uj x P *i ^Ut + A 2 ^U 2 + A3 ^U 3 House Wiring*. y* As a rule, the size of wire used in wiring ordinary buildings for light &nd power as fixed by the permissible heating of the wire (see p. £65) is of sufficient size to keep the pressure drop within the prescribed limit, since the distances the wires are run are comparatively short. It is always well, however, to calculate the drop in the heaviest and longest circuits, to be sure that one is on the safe side as regards regulation. Chart and Table for calculating* Alternating-Current JLi ne.«. Ralph D. Mershon, in American Electrician. The accompanying table, and chart on page 282 include everything neces- sary for calculating the copper of alternating-current lines. The terms, resistance volts, resistance E.M.F., reactance volts, and react- ance E.M.F., refer to the voltages for overcoming the back E.M.F.'s due to resistance and reactance respectively. The following examples illustrate the use of the chart and table. Problem. — Power to be delivered, 250 k.w.; E.M.F. to be delivered, 2000 volts; distance of transmission, 10,000 ft.; size of wire, No. ; distance be- tween wires, 18 inches ; power factor of load, .8 ; alternations, 7200 per min- ute. Find the line loss and drop. The power factor is that fraction by which the apparent power or volt-am- peres must be multiplied to give the true power or watts. Therefore the 250 k w apparent power to be delivered is ^ — - = 312.5 apparent k.w., or 312,500 volt-amperes, or apparent watts. The current, therefore, at 2000 volts will be 312 500 -xqVvq- = 156.25 amperes. From the table of reactances, under the heading M 18 inches," and corresponding to No. wire, is obtained the constant, .228. Bearing the instructions of the table in mind, the reactance volts of this 280 CONDUCTORS. line are 156.25 (amperes) X 10 (thousands of feet) x .228 = 356.3 volts, whicli are 17.8 per cent of the 2000 volts to be delivered. From the column headed " Resistance Volts," and corresponding to No. wire, is obtained the constant .197. The resistance volts of the line are, therefore, 156.25 (amperes) x 10 (thousands of feet) x .197 = 307.8 volts, which are 15.4 per cent of the 2000 volts to be delivered. Starting, in accordance with the instructions of the sheet, from the point where the vertical line, which at the bottom of the sheet is marked " Load Power Factor .8," intersects the inner or smallest circle, lay off horizontally and to the right the resistance E.M.F. in per cent (15.4), and "from the point thus obtained," lay off vertically the reactance E.M.F. in per cent (17.8). The last point falls at about 23 per cent, as given by the circular arcs. This, then, is the drop in per cent of the E.M.F. delivered. The drop in per 23 cent of the generator E.M.F. is, of course, o = 18.7 per cent. J.UU —j" — o The resistance volts in this case being 307.8, and the current 156.25 am- peres, the energy loss is 307.8 x 156.25 = 48.1 k.w. The percentage loss is 48.1 250 4- 4g .. = 16.1. Therefore, for the problem taken, the drop is 18.7 per cent, and the energy loss is 16.1 per cent. If the problem be to find the size of wire for a given drop, it must be solved by trial. Assume a size of wire, and calculate the drop in the manner above indicated ; the result in connection with the table will show the direction and extent of the change necessary in the size of wire to give the required drop. The table is made out for 7200 alternations per minute, but will answer for any other number. For instance, for 16,000 alternations, multiply the reactances by 16000 -f- 7200 = 2.22. As an illustration of the method of calculating the drop in a line and trans- former, and also of the use of the table and chart in calculating low-voltage mains, the following example is given : — Problem. —A single-phase, induction motor is to be supplied with 20 am- peres at 200 volts ; alternations, 7200 per minute ; power factor, .78. The distance from transformer to motor is 150 ft., and the line is No. 5 wire, 6 inches between centres of conductors. The transformer reduces in the ratio 2000 : 200, and has a capacity of 25 amperes at 200 volts ; when delivering this current and voltage, its resistance E.M.F. is as 2.5 per cent, and its reactance E.M.F. 5 per cent, both of these constants being furnished by the makers. Find the drop. The reactance of 1000 ft. of circuit, consisting of two No. 5 wires, 6 inches 150 apart, is .204. The reactance-volts, therefore, are .204 x tt^ X 20= .61 volts. 150 The Tesistance-volts are .627 x ^^^ X 20 = 1.88 volts. At 25 amperes, the re- j 1000 sistance-volts of the transformers are 2.5 per cent of 200, or 5 volts. At 20 20 amperes they are — of this, or 4 volts. Similarly, the transformer reactance volts at 25 amperes are 10, and at 20 amperes are 8 volts. The combined re- actance-volts of transformer and line are 8-f- .61 = 8.61, which is 4.3 per cent of the 200 volts to be delivered. The combined resistance- volts are 1.88-}- 4, or 5.88, which is 2.94 per cent of the E.M.F. to be delivered. Combining these quantities on the chart with a power factor of .78, the drop is 6 per cent of the delivered E.M.F., or -— = 4.8 per cent of the impressed E.M.F. The 105 transformer must therefore be supplied with 2000 -f- .952 = 2100 volts, in order that 200 volts shall be delivered to the motor. To calculate a four-wire, two-phased transmission circuit, compute, as above, the single-phased circuit required to transmit one-half the power at the same voltage. The two-phase transmission will require two such circuits. To calculate a three-phase transmission, compute, as above, a single-phase circuit to carry one-half the load at the same voltage. The three-phase transmission will require three wires of the size obtained for the single-phase circuit, and with the same distance (triangular) between centres. By means of the table calculate the Resistance- Volts and the Reactance- TRANSMISSION LINE CALCULATIONS. 281 Volts in the line, and find what per cent each is of the E.M.F. delivered at the end of the line. Starting from the point on the chart where the vertical line corresponding with power factor of the load intersects the smallest circle, lay off in per cent the resistance E.M.F. horizontally and to the right; from the point thus obtained lay off upward in per cent the reactance E.M.F. The circle on which the last point falls gives the drop in per cent of the E.M.F. delivered at the end of the line. Every tenth circle-arc is marked with the per cent drop to which it corresponds. Size of Wire B.&S. o o o -§£ §M £h.S fcj) o a §^ o • io ss .22 © '- d C" r; o ;> u £ Reactance- Volts in 1000 ft. of Line ( = 2000 ft. of Wire) for One Ampere ( V Mean Square) at 7200 Alternations per Minute for the Distance given between Centers of Conductors. V 1" 2" 3" 6" 9" 12* 18" 24" 30" 36" 3000 639 .098 .046 .079 .111 .130 .161 .180 .193 .212 .225 .235 .244 000 507 .124 .052 .085 .116 .135 .167 .185 .199 .217 .230 .241 .249 00 402 .156 .057 .090 .121 .140 .172 .190 .204 .222 .236 .246 .254 319 .197 .063 .095 .127 .145 .177 .196 .209 .228 .241 .251 .259 1 253 .248 .068 .101 .132 .151 .183 .201 .214 .233 .246 .256 .265 2 201 .313 .074 .106 .138 .156 .188 .206 .220 .238 .252 .262 .270 3 159 .394 .079 .112 .143 .162 .193 .212 .225 .244 .257 .267 .275 4 126 .497 .085 .117 .149 .167 .199 .217 .230 .249 .262 .272 .281 5 100 .627 .090 .121 .154 .172 .204 .223 .236 .254 .268 .278 .286 6 79 .791 .095 .127 .158 .178 .209 .228 .241 .260 .272 .283 .291 7 63 .997 .101 .132 .164 .183 .214 .233 .246 .265 .278 .288 .296 8 50 1.260 .106 .138 .169 .188 .220 .238 252 .270 .284 .293 .302 282 CONDUCTORS. ili£ II :q 1 ! r' ; | ±r : igfeffiife t±w± : =f :~ hzr feffi - -H- - h jjjj f " ■{ J-;- -i-hb }:i±ii: j+j^ rite±E ffT r tcb Small Division quals one Percent ^44- ^k- i — | — i — j — I — . — 1 — 4: •^■fcrl'.H m "HsTf: *Kp- -"- - Mi -H-P N J '~ KX^p ;~- - > ■ fL^frp i+trH^ -fl^fl+f tttttttt ^^^j^1Ikx> s- ~:- T ' ; tjfH I--S- i : l SB ^vtfNsTm'T^ M-" 'i^- — A >SIl§ fe^jT HtipE m tm l^sBi". v ' r^~ ^ v f^|P #1 , ^ rt fjsff +ff- lllllllllllllf^ ^fflfflPffife ;n.\ ll^vS^ cfitji)} t* r-n'T^S^^t ^— -<•"! -.\ — ,J1 \ mt? ggrb-r ^r^fe H ^-rjy; liSS ill s| HS+^-f-V Vy V \ Sill!? vT^JtIt -- - j 1 j 1 i i\>. .: - IP 11 i# : IffiS it xVf.t wM ss A, iNu - . . ^ : .TTX w : ffl+ £|$£p|| i^ALia iffi i^^tO: ;^H ©sii ^-|'['i\xi 11 r?w ||§s ft^ w if(+j+ lUi ^V SS||K fffistt isffifi fe§s H 11 "X"- r- wv^ Vfjr tjfff5 W'hm 111 .- - 1 . \ il Wffn sill 4ffi# II n Iflllillllllll f- 1 j.J . It HT 111 III 3u$*fe+ gg .6 .7 .8 Load Power Factors 10 20 30 Drop in Percent of E.M.F. Delivered Fig. 1(5. TRANSMISSION LINE CALCULATIONS. 283 The following curves published by the General Electric Company give the pounds of copper per kilowatt delivered for various percentages of power loss and various pressure gradients (volts per mile). It is to be noted that these curves are correct only for unity power factor. Line Loss in per cent of Power Delivered. rirn v fflt PP^t I L A wtitt , T i V A flrtttm 3 3 x fflt+t+^h- f V v nitui. t X X a \ \ \ \ \ \ \ EUUJ i_3-- A. t a pjnj^t A- a v A t ttx t 3 \r v , £ ^443 A V > £ Ii \\\\ \ X \ X. ftttJJ I C V v \ _nrf l_l_3 a A S -u \ ( IE ±3 tlL L \ A \ ' EJ5 t-5\. v\ s: jk \ D±ti_t CL_C A \ V rntri ii 3 v 3^ V UJII3 U V i ^ X rmi l. n t ^ _l\, s ^ fttri v 3 1 K-tX \ s. flttCC 3 V V !^-PX ^ % TpTTl \ ^ ^4K *> n ll\ kj V v >. <|L \ ^ V u\ 1 \ V \ > ^^ x x "^ ^^ DIC33A £ 35^ N \ Ns \\\ ) \ \ > ]\ ^\ \ ^"v ^^ "^"^ \\ \ \ \ jJ wv \ V, ""^^ "^-v, \\\ \'^*\ S W "^^ "^^ ^^^ v\\rfev ^s ^^ ^^ ^^-^ *^^- \\\\^k v^>^ -^ ^-_ ^ ^^ i ^ SV ^ S ^^ u ^ > "'^^ """""*--*. ^ "^— - !fc ." < ■ - ^^^£===5==-====:== 002 S6T 061 S8I 081 Sit Oil S9T 091 SSI 0ST src on S2T SST g osi CD( 15 dczxizx: Type Derivation No. of Type =0 IXZDCDC XZDCZXIZJCZI^CZXZ -^"?- DCDC DCDC DCZ>CDC :3 = 1 + 2 A :5 . - 1 + 4 -6 = 2+4 :: =3 + 4 18 19 = 1 + 8 110= 2 + 8 :ii= 3 + 8 112= 4+8 113= 5+8 114=6 + 8 115= 7 + 8 -J Fig. 27 its center, has almost no beneficial effect. The exposure of 1 to 5 is 4; of 2 to 6 and 3 to 7, £; of 2 to 8 and 2 to 9, Induction Induction Sec tion Seel ;ion Section Section Fig. 29. Niagara Lin«*, — The conductors on this line are bare cables of 19 strands, equivalent to 350,000 circuit mils, and are arranged as shown in the following diagram. The first arrangement was with two three-wire cir~ A -4-1-8^- ± : Fig. 30. Niagara-Buffalo Line. 11000 to 22000 Volts. cuits on the upper cross-arm, the wires being 18 inches apart. So much trouble was experienced from short circuits by wires and other material being thrown across the conductors, that the middle wire was lowered to the bottom cross-arm as shown, since which time no trouble has been experienced. With porcelain insulators tested to 40,000 volts there is no appreciable leakage. These circuits are interchanged at a number of points to avoid inductive effects. TRANSPOSITION OF LINES. 291 Three-Phase Circuits. — The diagram (Fig. 31) shows another ar- rangement now seldom used although it makes lines conveniently accessible for repairs. Under the ordinary loads usual in the smaller plants the unbal- ancing effect is so small # as to be inappreciable. i. Fiq. 31. Convenient Arrangement of Three-Phase Lines for, 600010,000 Volts. h— «*-h •&, -L.6, Fig. 32. Arrangement of Two-Phase Circuit, of Phases necessary. No Reversal Two-Phase, Four-wire Circuits. — The arrangement of conductors shown in Fig. 32 is probably the best for two-phase work; as no reversals of wires are needed, the inductive effects of the wires of one circuit on those of the other are neutralized. 292 CONDUCTORS. Two-Phase Circuit* in lame Plane. — If the phases are treated as separate circuits, and carried well apart, as shown in Fig. 33, the interfer- Fio. 33. ence is trifling; and should the loads carried be heavy enough to cause notice- able effect, the reversal of one of the phases in the middle of its length will obviate it. The following diagram illustrates the meaning. PH AS E B. y C Fig. 34. Arrangement of Two-Phase, Four-Wire Circuit with Wires on same Plane. Wires of One Phase should be interchanged at the Middle Point of the Distance between Branches, and between its Origin and First Branch. Messrs. Scott and Mershon of the Westinghouse Electric and Manufactur- ing Co. have made special studies of the question of mutual induction of circuits, both in theory and practice; and their papers can be found in the files of the technical journals, and supply full detail information. Mutual Neutralization of Capacity and Inductance. — In order to completely neutralize phase displacement due to distributed in- ductance a distributed capacity is essential. Localized capacity can, how- ever, produce a partial neutralization. Excessive distributed capacity can also be partially neutralized by inserting inductances at proper inter- vals. In treating of local neutralization of capacity by inductance, the assumption is frequently made that the capacity is constant irrespective of the voltage, and that the inductance is constant irrespective of the current. Under these conditions neutralization can be obtained. As, however, inductance is dependent upon the permeability of the associated magnetic circuit, and permeability varies with the saturation of the iron, — that is, with the current, — complete neutralization cannot be obtained with iron inductances. Over-excited synchronous motors, or synchronous converters, take cur- rents which lead the electromotive force impressed upon them, and they therefore operate as condensers, and they may be utilized advantageously in neutralizing the line inductance. The power factor of the transmission system can therefore be varied by varying their excitation. BELL WIRING, 293 10§§ IX SHEATH OF THREE-CONDUCTOR COVERED CABLEI. IEAD John T. Morris (Electrician, London) gives the following formula, con- firmed by experiments, for the loss of power in the lead sheath of a three- conductor cable. Let / = current in amperes. / = frequency. I = length of cable in 1000 ft. t = thickness of sheath in mils. Then: Watts loss = 123 X 10^° PfW' 1 . If the cable is placed in an iron pipe the loss is increased about 75%. BELL WIRING. The following diagrams show various methods of connecting up-call bells for different purposes, and will indicate ways in which incandescent lamps may also be connected to accomplish different results. 5=6= Fio. 35. One Bell, operated by One Push. Fig. 36. One Bell, operated by Two Pushes. t_t J- Fig. 37. Two Bells, operated by One Push. Fig. 38. Two Bells, operated by Two Pushes. When two or more bells are required to ring from one push, the common practice is to connect them in series, i.e., wire from one directly to the next, and to make all but one single-stroke ends. Bells connected in multiple arc, as in diagram No. 24, give better satisfaction, although requiring more wire. Fig. 39. Three-Line Factory Call. A number of Bells operated by any number of pushes. All bells rung by each push. Fig. 40. Simple Button, Three- Line Return Call. One set of battery. ug gCM'I- j Fig. 41. Simple Button, Two-Line and Ground Return Call. One set of battery. 6 T*1 Fig. 42. Two-Line Return Call. Illustrating use of Return Call Button. Bells ring separately. 294 CONDUCTORS. Fig. 43. One-Line and Ground Return Call. Illustrating use of Return Call Button. Bells ring separately. Fig 44. Simple Button, Two- Line Return Call. Bells ring together. Q Fig. 45. Simple Button, One-Line and Ground Return Call. Bells ring together. The use of com- plete metallic circuit in place of ground connection is advised in all cases where expense of wire is not considerable. Fig. 46. Four Indication Annuncia- tor. Connections drawn for two buttons only. A burglar alarm cir- cuit is similar to the above, but with one extra wire running from door or window-spring side of bat- tery to burglar alarm in order to operate continuous ringing attach- ment. (j> (j> Annunciator ; Numeral indicates number of Points. — ^ Speaking Tube. — © Watchman Clock Outlet. — J Watchman Station Outlet. — © Master Time Clock Outlet. — J) Secondary Time Clock Outlet. ffl Door Opener. |Xl Special Outlet ; for Signal Systems, as described in Specifications. j|||||l||| | Battery Outlet. {Circuit for Clock, Telephone, Bell or other Service, run under floor, concealed. Kind of Service wanted ascertained by Symbol to which line connects. {Circuit for Clock, Telephone, Bell or other Service, run under floor above, concealed. Kind of Service wanted ascertained by Symbol to which line connects. In writing circular mill sizes a quick and handy method is to draw a circle and place in it the size in hundred thousands of circular mills, as, (?) = 300,000, (§) = 750,000, @ = 250,000. This is unhandy to print. UNDERGROUND CONDUITS AND CONSTRUCTION. With the establishment of the first commercial Morse telegraph line probably commences the history of the "underground wire" when a gutta-percha covered cable was laid in a trench made by an ox-drawn plough. Stages in the evolution of the present " monolithic " conduit are promi- nently marked by the system of grouping wires permanently installed and separated by the pouring about them in the trench of various insulating compounds; by the "built up system" made of creosoted boards so placed as to form square channels or ducts; by the "pump log" system or squared timber bored to required size and creosoted; by the cement lined iron pipe system; by the use of paper moulded and treated with dielectric compounds; and by the now very largely used vitrified clay. Clay conduits should be manufactured from a clay which will vitrify to a highly homogeneous and non-absorbent condition and be free from chemical elements (iron, sulphur, etc.) which under the action of heat in the kilns result in nodes or blisters in the ware. There are two established styles of clay conduit commonly designated as "single duct" and "multiple duct." The standard unit of the single duct is of square cross section measuring 4$ " by 4$" with corners chamfered, is 18 inches in length, and has a 3£ inch round bore or hole. The standard multiple duct units are of two, three, four, or six duct sections, the bore of each duct of any section being square and measuring 3£ by 3|, the interior and exterior wall being f" thick; the lengths of units are, for two and three duct, 24 inches, and for four or six duct 36 inches. The demand for 3£ inch and 4 inch bores or even larger is constantly increasing. Multiple duct conduit of nine duct and twelve duct sections have been offered to the trade but so far have not come into extensive use. Single duct conduits being more flexible are better adapted to use where service pipes, curves, or obstacles are frequently encountered. Laid with broken joints the possibility of the heat from a burning cable , being com- municated to a neighboring cable, is precluded. Where high construction on a small base (two ducts wide by more than five ducts high) is required, singles are not used to advantage. A mason should, under fair working conditions, average in a day of eight hours from twelve hundred to sixteen hundred duct feet of single duct conduit. Multiples have in their favor fewer joints, greater weight per unit, and the fact that their installation requires only unskilled labor. Two men selected from a gang of laborers will lay from eighteen hundred to twenty-four hun- dred duct feet per day of ten hours. Through town or city streets the conduit should have a foundation of concrete at least 3 inches thick. Where frequent excavations for other works are probable a complete encasement of 3 inches to 4 inches of concrete should be placed on both sides and on top of the ducts. The side protec- tion is, however, sometimes omitted and creosoted boards substituted for concrete on top. The top covering over ducts should be not less than 24 inches below the surface of the street. The several conduit terms are generally defined as follows: The word "Conduit" means the aggregation of a number of hollow tubes of duct material and includes all of the ducts in a cross section of the subway. In general a conduit will consist of four ducts or more. The word "Duct" means a single continuous passageway between man- holes or through any portion of the conduit or laterals. The word "Manhole" means an underground chamber built to receive electrical equipment and suitable to give access to the conduit. The word "Service Box" means an underground chamber similar to a manhole but of smaller size, and designed primarily to give access to dis- tributing conductors. 301 302 UNDERGROUND CONDUITS. The word "Lateral" means one or more ducts extending from a manhole or service box or from one or more of the main conduit ducts to some dis- tributing point. In general laterals will consist of one or two ducts for the same service connections. One or more laterals may be installed in the same trench. Manholes vary so much according to the ideas of the different engineers that it is difficult to give data that would suit all of them. However, the average size of manhole is 5' X 6' X 6' in the clear with a 12" wall. The covers for same vary from 800 to 1400 lbs. The general practice is to have ventilated covers and sewer connections with automatic back-water traps. The Service Boxes are made generally of concrete with an 8" wall, either 2' X 2' or 2' X 3' in length and width, and extending in most cases to the top layer of the conduit system, which would make the depth of the service boxes vary according to the depth of the conduit system proper, the upper tier of ducts being used for distribution. Covers for service boxes, including inside pan, weigh from 400 to 600 lbs. ; Usual Practice of Conduit Work. Manhole walls, where built of concrete are generally 8 to 12 inches thick, made of Portland Cement concrete, using, \\ inch stone, mixed in the pro- portion of an 1-2-5 and in some instances as high as 1-3-8. While in some cases the conduits proper are surrounded with Portland Cement concrete, the usual practice throughout the country is with casing of hydraulic cement concrete in a 1-2-5 mixture, stone \ inches to 1 inch. The Cost of Conduits. (A. V. Abbot in Electrical World and Engineer.) The items of cost of conduit construction are: 1. Duct material. 2. Pavement per square yard. 3. Street excava- tion per cubic foot, including the removal of paving, the refillment of the excavation after the ducts are laid, and the temporary replacement of the paving. 4. Concrete deposited in place. 5. Labor of placing duct ma- terial 6. Engineering expenses. 7. Manholes. 8. Removal of obstacles. TABLE No. 1. Cost of Manholes in Dollars. A. Brick with Brick Roof. Item. Amount. Rate (Dollars). Min. Amt. $ Per Ct. Av. Am. $ Per Ct. Max. Amt. Per Min. Ave. Max. Ct. Excavation Concrete . Brick . . Cover . . Iron . . . Repaving . Cleaning . 375 cu. ft. .7 yard 2200 1 500 lbs. 6 yards 10 loads .02 5.00 12.00 5.00 .015 .75 .50 .03 7.00 15.00 10.00 .03 2.00 .75 .04 9.00 18.00 15.00 .05 4.00 1.00 7.50 3.50 26.40 5.00 7.50 4.50 5.00 12.6 5.9 44.5 8.4 12.6 7.6 8.2 11.25 4.90 33.00 10.00 14.00 15.00 7.50 11.8 5.3 35.3 10.6 16.1 12.8 8.1 15.00 6.00 39.60 15.00 25.00 24.00 10.00 11.2 4.4 29.4 11.2 18.6 17.8 7.4 Totals . . .... 59.40 100.0 93.65 100.0 134.00 100.0 COST OF UNDERGROUND CONDUITS. 303 B. Brick ivith Concrete Roof. Rate (Dollars) Item. Amount. Per Unit. Min. Amt. % Per Ct. Av. Am. $ Per Ct. Max. Amt. Per Min. Ave. Max. ct. Excavation 375 cu. ft. .02 .03 .04 7.50 14.8 11.25 14.4 15.00 13.8 Concrete . 1.9 yards 5.00 7.00 9.00 9.50 18.7 13.30 17.0 17.10 15.7 Brick . . 1600 12.00 15.00 18.00 19.20 37.8 24.00 30.9 28.80 25.7 Cover . . 1 5.00 10.00 15.00 5.00 9.0 10.00 12.8 15.00 13,8 Repaving . 6 yards .75 2.00 4.00 4.50 8.9 12.00 15.4 24.C0 21.9 Cleaning . 10 loads .50 .75 1.00 5.00 9.9 7.50 9.5 10.00 9.1 Totals . . • • • • 50.70 100.0 78.05 100.0 109.90 100.0 C. Concrete Manhole. Item. Amount. Bate (Dollars) Per Unit. Min. Amt. Per Ct. Av. Am. Per Ct. Max. Amt. Per Min. Ave. Max. Ct. Excavation Concrete . Cover . . Repaving . Cleaning . 375 cu. ft. 4.5 yards 1 6 yards 10 loads .02 5.00 5.00 .75 .50 .03 7.00 10.00 2.00 .75 .04 9.00 15.00 4.00 1.00 7.50 22.50 5.00 4.50 5.00 16.8 50.5 11.2 10.2 11.2 11.25 31.50 10.00 12.00 7.50 15.5 43.6 13.9 16.6 10.4 15.00 40.50 15.00 24.00 10.00 14.3 38.8 14.4 23.0 9.5 Totals . . .... • • • ■ • • 44.50 100.0 72.25 100.0 104.50 100.0 Whenever practicable, a sewer connection to each manhole is desirable to provide exit for street drainage. Such sewer connections are essential in all cases where manholes are equipped with ventilating covers, otherwise the manholes will fill during every storm. TABLE Wo. «. Cost of Sewer Connections in Dollars. Rate (Dollars) Per Unit. Min. Ave. Max. Item. Amount. Amt. Per Am. Per Amt. Per Min. Ave. .03 Max. .04 $ Ct. $ Ct. $ Ct. Excavation 225 cu. ft. .02 4.50 35 1 6.75 26 9.00 21.4 Concrete . 5 yards .75 2.00 4.00 3.75 29.2 10.00 38.8 20.00 47.0 Brick 1 1.00 2.50 4.00 1.00 7 6 2.50 19.6 4 00 9 3 Cover 16 feet .04 .07 .10 .64 5 1.12 4 4 1.60 3 6 Repaving . 2 loads .50 .75 1.00 1.00 7 6 1 50 5.8 2 00 4 7 Cleaning . 1 2.00 4.00 6.00 2.00 15.5 4.00 15.4 6 00 14 Totals . . . . . 12.89 100.0 25.87 100 42.60 100 304 UNDERGROUND CONDUITS. Manholes will occur at intervals of from 250 to 500 feet, consequently the constant cost per conduit foot for this item is obtained by dividing the various manhole costs by the distances between them. TABLE No. 3. Constant Cost per Conduit foot for Manholes in Dollars. Distance between Manholes in Feet. 250 300 350 400 500 Brick manhole with brick roof . . . (Min. ] Ave. ( Max. .238 .372 .536 .196 .310 .427 .170 .248 .384 .148 .236 .335 .118 .186 .268 Brick manhole with concrete roof . . ( Min. < Ave. ( Max. .203 .300 .440 .169 .260 .363 .145 .223 .314 .127 .195 .272 .102 .156 .218 Concrete manhole (Min. I Ave. (Max. .176 .278 .416 .148 .242 .347 .127 .209 .298 .111 .180 .260 .089 .144 .208 Sewer connection (Min. jAve. (Max. .051 .104 .170 .043 .086 .142 .038 .074 .121 .032 .064 .105 .025 .051 .084 Engineering expense will vary from a minimum of 5 cents per conduit foot to a maximum of 12 cents, depending chiefly upon the difficulty of the work. The cost of the removal of obstacles is an item impracticable to estimate a priori with any degree of certainty, as it is impossible to foresee, and usually impracticable to ascertain, even with the greatest care, the impedi- ments to be encountered beneath street surface. Experience indicates that this expense will vary for small subways from 10 cents to 62 cents per foot of conduit; for medium-sized ones from 12 cents to $1.10, and for large conduits from 15 cents to $2.25. The cost of paving is partially dependent upon the number of ducts. It is impracticable for workmen to perform their avocations in a trench less than 18 inches wide, and, therefore, a strip of pavement of this width must be opened irrespective of the number of ducts to be installed. The cost of repaving will further vary with the kind of paving. In Table No. 4, the usual kinds of pavement encountered, the minimum, average, and maximum prices per square yard, and cost per conduit foot are given. Allowing a disturbance of paving for six inches on each side of the trench, the cost per lineal foot for small conduits will vary from 2.3 to 26.3 cents; for medium-sized ones from 4.6 to 29.2 cents, and for large conduits from 6.9 to 35.0 cents. Similarly the cost of excavation is only partially dependent upon the number of ducts. COST OP PAVING. 305 a m o S s 9 •e a c © e e h a el H 8 s * ft e o g I 8 b- >* d CQ o CO fr- CO CO CO 8 CO OS CM 8 8 h,OS co 2 fi d »o lO CM CO 8 8 ss g to 3^2 d CO CO CO CM CM © 1-1 2 8 •d d CO CO •CO CM ^ CO © o CM I- CM a 8 CO CM E 8 . © d g.S^d 8 o o § 8 S » S£^ CO CO CM " CM rt ? t- d 3 cm © g CO CO t» co § ti 1 d *■. CM ". °. R uj CO 2 d ft^3«+-l 8 ^ o CM CO CM CI £ 8 55 "£ 3 °5 d CO CM CM o © OT3 O 2 8 o •d d g CO CM I o CO CO CM 8 CO 1 CO Ave. Price per Sq. Yd. 8 CN CM U5 t- 8 8 B c~ d d CO 8 CO CO CM co o © 8 8 © CO 2 o d S) 00 CO CO CO SI CO CN d CM © © © 0-3 g CO 3° C6 CM 8 d d CO S 8 CM CO CO CO 8. lO s CO 8. . o> d G.s^d £ CM § s c^ 8 c^ aai* - 1 ' cm' ^ rH ' +3 o o «H b£ -JJ s d g8 •d Ah O o o u > +3 o o o .id s 'el IS • .a CO JO •d •a .4 ft s c3 T! o «4-( o3 O OS 3 X W Pi 0) , o> ^ <4 O " H 2 i 306 UNDERGROUND CONDUITS. Experience shows that 3 feet 6 inches is a minimum permissible dep^ for the bottom of subway construction, and that the cost of street excava- tion will vary from two to four cents per cubic foot of material excavated, including the removal of the pavement, the refillment of the trench, and the replacement of temporary paving. The cost of excavation will, there- fore, stand as in Table No. 1. TABLE Wo. 5. Cost of Street Excavation per Conduit foot in Dollars, 1 to 9 ducts 10 to 16 ducts 17 to 25 ducts Minimum .02 per Cu. Ft. .105 .160 .225 Average .03 per Cu. Ft. .1575 .240 .3375 Maximum .04 per Cu. Ft. .210 .320 .450 Table No. 5 summarizes these constant items; for conduits of from one to nine ducts, ten to sixteen ducts, and seventeen to twenty-five ducts, giving the minimum, average, and maximum prices of all, together with the percentage that each bears to the total. Table No. 6 enumerates the probable prices for the various forms of duct material laid into place, calculated in a manner similar to the preced- ing tables, including a percentage column showing the effect of each item upon the total expense. TABIE Wo. O. Constant Cost per Conduit Foot in Dollars. Minimum. Average. Maximum. Item. Cost. Per Cent. Cost. Per Cent. Cost. Per Cent. 1 to 9 ducts. Excavation .... Paving Engineering .... Removal of obstacles . .105 .0695 .05 .10 32.6 21.2 15.2 32.0 .1575 .185 .08 .25 23.4 27.5 11.9 37.2 .210 .279 .12 1.00 13.0 17.4 7.5 62.1 Total .3245 100.0 38.6 20.2 12.1 29.1 .6725 .24 .222 .08 .28 100.0 29.1 27.0 9.8 34.1 1.609 .32 .3315 .12 1.10 100.0 10 to 16 ducts. Excavation .... Paving . .16 .0845 17.0 17.7 Engineering .... Removal of obstacles . .05 .12 6.5 58.8 Total .4145 100.0 43.0 18.6 9.6 28.8 .822 .3375 .26 .08 .35 100.0 32.8 25.3 7.8 34.1 1.8715 .45 .52 .12 1.25 100.0 17 to 25 ducts. Excavation .... Paving Engineering .... Removal of obstacles . .225 .0970 .05 .15 19.2 22.2 5.1 53.5 Total .522 100.0 1.0275 100.0 2.34 100.0 COST OF UNDERGROUND CONDUITS. 307 From the data thus collected, the total cost of a conduit of any size is readily determined by taking first the cost per foot of street for manholes and sewer connections; second, the cost of the constant street items as given in Table No. 6 depending upon the number of ducts, and third, the cost per duct foot determined from Table No. 5 multiplied by the number of ducts to be laid, and adding these three items together, giving immediately the total cost per conduit foot. TABLE Wo. 9. Cost of Duet Material in Place in Dollars, Minimum. Average. Maximum. Item. Cost. Per Cent. Cost. Per Cent. Cost. Per Cent. Hollow brick. Duct material . . . Placing Encasement .... .02 .005 .02 44.4 11.2 44.4 .035 .01 .05 36.8 10.5 52.7 .05 .015 .08 34.5 10.3 55.2 Total .045 100.0 67.5 2.2 30.3 .095 .05 .0025 .0475 100.0 50.0 2.5 47.5 looTo - 53.6 3.4 43.0 .145 .065 .004 .07 100 Multiple duct. Duct material . . . Placing Encasement .... .035 .011 .015 .061 46.7 2.9 50.4 Total 100.0 62.5 3.2 34.3 .10 .06 .004 .05 .139 .08 .006 .088 100 Cement-lined pipe. Cement pipe. Wood pulp. Duct material . . . Placing Encasement .... .04 .002 .022 48.2 3.6 48.2 Total .064 100.0 98.04 1.96 0.00 .114 .05 .0015 .00 100.0 98.0 3.0 0.0 .174 .06 .003 .00 100 Creosoted wood. Duct material . . . Placing Encasement .... .04 .0008 .00 95.0 5.0 0.0 Total .0408 100.00 0515 100 .063 100 Cost per Conduit foot in Cities. Cost per Number of Ducts. Trench Foot. 2 4 6 12 16 20 24 Atlanta . . . Louisville . . Cincinnati . Boston . . . Springfield . . Brooklyn . . $.88 .89 .92 1.06 .90 .95 $1.14 1.12 1.18 1.34 1.16 1.21 $1.43 1.40 1.48 1.65 1.45 1.51 $2.31 2.29 2.36 2.66 2.34 2.45 $2.76 2.76 2.82 3.13 2.78 2.91 $3.22 3.19 3.26 3.66 3.24 3.39 $3.53 3.63 3.72 4.10 3.68 3.84 308 UNDERGROUND CONDUITS. J»d $sbo a « * .2 ^ ft © t> © © N <* ft 8 §3 SSlglgislk! . g£2 83 "* CM rH NO CO CO OOOlOOOt- H s? .a T* ft CO lO 00 'i S°2 COC5fOuOO"^t-rfJt-OTH(N © O C NO CM© (N00O^OOOrtOt»O^ MOt t- 1-1© © f- i-H S M 3 t~ CS NO CO t> © CO CN ST 8 SB h *^ « ft O COO rj*ift(N»Oi5(N6(NWOO^ "** CN NO © © OCO i-i i-i CM CN 00 O CO O t-t^CN L- rf CO t~ NO 00 O © "# ^< CM >N0— <©©C i-iOO © t*- 1-H © >; co i-» © co t~ NO "5 « t- Tf T-( >©©©©co8co 88 a. 5 • 08 ftg CO ^ft S.a 00* a > a a c$ CD ft 0h« CD y o o OH xj 3 «H •** X . o o X ' X' ^0*-'OS- * 6^?* | »-,'■ f ] i : < ; jr| \4\ i ' I ( -t-H [E^SS^ rSttrEjfa'mk in 1 vj—M c«»iat;rS«fig®5;5? n ?a.^e^ss^2^#^ SECTION ELEVATION SECTION* B-B Figs. 3 and 4. Manholes. 310 UNDERGROUND CONDUITS. Section A-A Fig. 5. Plan and Sectional View of Manholes. mmm 9 DUCTS 48 DUCTS Fig. 6. Feeder Ducts in Position. MANHOLES. 31] Fig. 7. Transformer Manhole. 312 UNDERGROUND CONDUITS. Fie. 8. Fig. 9. Gest's Patent Manhole Designs. Part of Frame showing Roughening Fig. 10. Sectional View of Manhole Covers. 314 UNDERGROUND CONDUITS. INNER COVER w Upper Side SUDSQ0 <:aosoa '/ i 1 n n i 1 oasoosaflsaosaasaos aasQDSQDSDQsoasQa aas[ra=aas!oasDDSQQs QQSDD aasoos 1 STREET COVER Fig. 11a. Manhole ^vers. TIGHT COVER FOR MANHOLE. 315 H-,r*4-,-rff ^^^^^^^^^^^^^g « IP-so"— I 4-1 42 PLAN OF FRAME Plan of Lock Bar Fig. 11b. Manhole Covers. 316 UNDERGROUND CONDUITS. Itemized Cost of Conduit. W. P. Hancock, Boston Edison Company. Material and Labor. Material. Lumber at $15.00 per M., or .015 cents per square foot, B. M Concrete at $4.85 per cubic yard, or 18 cents per cubic foot Mortar at $3.98 per cubic yard, or 14 cents per cubic foot . Ducts laid down beside the trench at $.0502 per duct foot Labor. Excavate and backfill at 15 cents per hour or $.0278 per cubic foot Cut and place lumber at 20 cents per hour, or $.0006 per square foot B. M. . . . . Mix and place concrete at 15 cents per hour, or $.0222 per cubic foot .... Mix and place mortar at 25 cents per hour, or $.0925 per cubic foot Lay the ducts at 60 cents per hour, or $.0040 per duct foot Haul away the dirt at 50 cents per hour, or $.0142 per cubic foot Pave the trench at $1.44 per square yard, or $.16 per square foot Cost of manholes per duct foot __ Total cost of manholes _ 490.28 ~ Total number of duct feet "~ 22,200 Inspection at 50 cents per hour, or $.0033 per duct foot Engineering expenses at $.0214 per duct foot Incidental expense at 5 per cent of total Cost per Duct Foot. .0105 .0231 .0026 .0502 .0004 .0029 .0016 .0040 .0047 .0500 .0221 .0033 .0214 .0116 $.2350 Cost per Conduit Foot. Total Expense. .1575 .3465 .0390 .7530 .3990 .0060 .0435 .0240 .0600 .0705 .7500 .3315 .0495 .3210 .1740 $3.5250 Total Cost for each Item for the Total Line. 233.10 514.15 58.90 1114.44 592.06 9.32 63.48 37.00 88.00 104.72 1109.92 490.28 73.26 475.08 248.22 $5212.73 Cost of 5' X »' X »' manhole. W. P. Hancock, Boston Edison Company. 23.76 cubic feet concrete, cost in place $.202 per foot $4.78 2,500 hard sewer bricks, cost $9.00 per M 22.50 If S. 6" trap and connections cost 5 . 65 30' 6" Akron sewer pipe, cost 30 cents per foot 9.00 R. R. steel (60 lbs. to the yard), 8 pieces 6' 4" long (1013 lbs.) cost $.0125 per lb 12.67 \\ yards mortar, cost per yard $3.98 4.47 1 manhole frame and cover, 962 lbs., cost $.015 lb 14.43 $73.50 COST OF MANHOLES. 317 We shall need labor that will cost as follows: Excavate and backfill part of same, including that for sewer connections, 785 cubic feet, cost $.0278 per Foot Remove from street 304 cubic feet of dirt, cost 50 cents double load or $.0142 per foot Pave 11.08 yards (including manhole and sewer connection), cost $1.44 per yard 1 mason, 10 hours, cost 40 cents per hour 2 mason helpers, 10 hours each = 20 hours, cost 15 cents per hour, $21.82 4.3^ 15.95 4.00 3.00 $49.07 Total cost 1 manhole, complete $122.o7 Cost of Underground Conduits in Cliicag-o. G. B. Springer, civil engineer of Chicago Edison Co., says: The difference in local conditions, variations in cost of material and labor, make it very difficult to give a set of figures which will hold good in many places or in fact in the same place under different circumstances. The following table, however, is submitted as a guide in approximating the cost of work of this character as a result of conduit construction cov- ering ten years in Chicago. The cost of manholes is not included in this table, but is given in the one following. Table for Estimating* Cost of Conduit, Per Duct Foot, in Different Groups, in Various Pavements: Kinds of Pavement Number of Ducts. 2 4 6 9 12 16 20 25 30 No pavement $.18 $.18 $.18 $.18 $.18 $.18 $.18 $.18 $.18 Macadam .24 .21 .20 .20 .19 .19 .19 .19 .19 Cedar .26 .22 .21 .20 .20 .19 .19 .19 .19 Cedar reserve and granite . .31 .24 .23 .21 .21 .20 .20 .20 .19 Granite reserve .43 .31 .28 .24 .24 .23 .22 .21 .21 Asphalt and brick reserve . .68 .43 .37 .31 .29 .26 .24 .24 .23 The following table contains approximate figures based on conditions prevailing in Chicago, and may be used as a guide in estimating the cost of conduit construction in connection with the table preceding. Table for .Estimating* Total Cost of Manholes in Different Kinds of Pavements : Kinds of Pavement. Size of Manholes in Feet. 3X3 $41 42 43 44 46 50 3X4 $47 48 49 50 53 58 4X4 4X5 5X5 6X6 6X7 7X7 8X8 9X9 No pavement .... Macadam Cedar Cedar reserve and granite .... Granite reserve . . . Asphalt and brick reserve $53 55 56 57 60 67 $64 66 67 68 72 80 $109 111 112 113 117 126 $133 135 136 138 144 156 $142 146 146 149 155 168 $160 163 164 167 174 188 $189 193 194 198 207 224 $222 2C6 227 231 ,°43 264 The above figures are based on the same prices for repaving, labor, brick- layers, cement and sand, as given in the table for conduit, and upon the following unit prices: Brick work including labor and material .... $12.50 per cu. yd. Concrete tops and bottoms $7 50 per cu' yd Back water gates '. | 6 .50 each! Sewer grates 30 cents each. Sewer connections $12.50 each. Sewer permits m $5.50 each. Manhole frames and covers * $15 *00 each' 12 318 UNDERGROUND CONDUITS. Grouping* of Ducts in Iflanholes. H. W. Buck in Electric Club Journal, April, 1904. Attention is called to the grouping of ducts and construction of manholes. Ordinarily ducts are bunched together and brought out at the center of the manhole, as shown in Fig. 12. Here the cables divide, half passing on one side and half on the other side of the manhole, being racked on the manhole walls. This design is objectionable for a number of reasons. » -N>^>\1 ft^fa E2S Plan Section Fig. 12. Ordinary Type of Manhole. First, it exposes every cable in the conduit to damage from short-circuit at the points A- A, where they are in close proximity to each other. Secondly, it necessitates bending every cable sharply at points A and B in every manhole, which tends to crack the insulation and cause trouble. Most break-downs in underground work do occur at these points. Another objection to this form of conduit construction is from the standpoint of heating. The cables in the inner ducts, if heavily loaded, will rise to a Section 1ULJ Plan Fig. 13. Improved Form of Manhole Construction. high temperature, for there is no way for the heat to get away by con- duction. The inner ducts are surrounded by chambers containing still air, which constitute the best possible insulator of heat. Ducts should never be grouped more than two in width, so that every duct will have an outlet for heat conduction through the surrounding earth. A much better form of construction is shown in Fig. 13. Here the ducts are grouped only two in width, and the conduit enters the manhole at the side, so that the cables can pass straight through on the side UNDERGROUND CABLES. 319 without bending. A further step in design leads to the arrangement shown in Fig. 14. Here the ducts are still laid in one trench but the ducts are placed in four separate groups, spaced apart by concrete, as shown. The Section Plan Fig. 14. Manhole Construction Adopted by Niagara Falls Power Company. manholes are built with a vertical division wall through the center and two entrance holes. Removable soapstone shelves divide the groups of cables horizontally, so that not more than one-quarter' of the number of Section Plan Fig. 15. Manhole Construction of Shallow Trenches. cables in the conduit can be damaged by short-circuit at any time. In this design the cables also run straight through the manhole without bend- ing. In places where rock is near the surface of the ground the construction shown in Fig. 15 is adopted. uarDEitoROuxn cjabmjs. Cables are placed underground in several different ways, chief among which are the " solid " and " drawing in" systems, as noted on page 301. One type of the solid systems is that in which the conductors, properly insulated, lead covered and protected by armor, are laid directly in the earth, a plan that has been widely adopted in Europe. The "Drawinff In " Plan is the one now most generally adopted in this country. This plan utilizes .the manholes and conduits just described. The cables are drawn into the ducts from manhole to manhole by means of a rope that has been previously drawn through the duct by a process termed "rodding." Rodding consists of screwing one rod on to another in the manhole and pushing them through the duct until the further end is reached. The rope is attached to the last rod and the rods are withdrawn from the ducts bringing the rope with them. Sometimes in place of rods a stiff steel wire is pushed through the ducts. The rope is 320 UNDERGROUND CONDUITS. attached to the cable by a mechanical device which securely grips the end of the cable. Various means of drawing the cables into the ducts are availed of, depending somewhat on the size of the cable and the length of the run; haud power, man power with windlass, horses, electric motors and gasoline engines being thus employed. Xjpen of" Underground Cables. — The type of cable employed for underground service varies largely with the requirements. Virtually all underground cables are lead covered to prevent injury to the insulation by moisture, gases, etc. For telephone purposes, lead covered, dry paper, insulated cables are universally used, to obtain low static capacity. (See pages 180 and 188.) For telegraph purposes rubber insulation (see page 229) and oil saturated cotton or paper are utilized, as in the telegraph service ; static capacity is not of so much importance, but still cannot always be disregarded, especially in high speed telegraph signaling. The conductor commonly used in underground telegraph cable is No. 14 B. & S. copper, having a conductivity of 98 per cent. In the case of cotton fiber or paper cable, each conductor is insulated to six thirty-seconds (&) of an inch outside diameter. The insulating material is thoroughly dried and then saturated with an insulating oil or compound. JFor Electric JLig-ht and Power purposes rubber, paper and varnished cambric insulation are largely used. (See pp. 174 and 180.) Owing to its high cost, rubber cables are not now in as high demand as formerly, especially as oil saturated paper cables appear to be quite as durable, efficient and reliable as rubber insulation for high potential work. It was formerly the practice to place as many as six lead covered electric light cables in one duct, but experience demonstrated that this was not advisable owing to the difficulty in withdrawing when necessary one or more cables from the duct without injury to the remaining cables. A burn- out in one cable also frequently injured adjoining cables in the duct. Present practice favors having only one cable in each duct, although there may be several conductors within the lead covering. (See page 185.) To prevent burning of light and power cables due to short circuits in the manholes and other places where the cables are bunched, the cables are frequently covered with asbestos strips about 3 inches wide and T % inch thick, well impregnated with a solution of silicate of soda which soon hardens over the lead. The lead covers of cables carrying alternating currents of high amperage and low E. M. F. should be bonded or carefully insulated in the manholes to prevent sparking and possible consequent damage, due to induced currents in the lead cover of the cables. All lead covered cables used on high potential circuits should be pro- tected from damage by static discharge by flared ends or bells, that is, by enlargement of the lead sheath to fully twice the diameter of the lead over the cable, for a distance of about a foot. The bell should then be filled with some good insulating material like Chatterton Compound, the con- ductor ends, in case of multiple conductor cable, being carefully separated. Cable Heads. — To prevent the entrance of moisture to the ends of telegraph and telephone paper cables the conductors of a short length (about two feet) of rubber covered cable are spliced to those of the paper cable. These splices are then insulated. A lead sleeve is passed over the rubber insulated conductors and the lead casing of the paper cable to which it is then soldered. The outer terminal of the rubber cable is led into a metal box or head to which the lead sleeve is soldered. The free conductors are solidly connected to insulated binding posts on the inside of the box, which binding posts extend to the outside of the head, thus giving access to the conductors externally. The sleeve and box are then filled with a melted rubber compound, the temperature of which must be below that at which the rubber insulation will soften ; otherwise the rubber will be seriously damaged. Bells for Cable End*. — AH lead-covered cable ends should be pro- tected from damage by static discharge by flared ends or bells, that is, by enlargment of the lead sheath to fully twice the diameter of the lead over the cable, for a distance of about a foot. Lead or brass cable heads or bells are much used on the ends of high potential underground cables. This bell should then be filled with some good insulating material like Chatterton Compound, the conductor ends, in case of multiple conductor cables, being carefully separated and arranged CABLE TESTING. Revised by Wm. Mayer, Jr. Cables — Underground and Submarine. The majority of the methods of tests and measurements given herein are applicable to aerial, underground, and submarine cables. Insulation Resistance. Direct Deflection Method, with Mirror Galvanometer. — This method, Fig. 1, is generally used in this country in underground and submarine work. CABLE Fig. 1. a and b z=. leads. G = galvanometer, Thomson or D'Arsonval, mirror type. S = shunts for G, usually X, T fo, T ^ . B =z battery, 20, 50, or 100 chloride silver cells. R z=. resistance box of megohm or more. BK = battery reversing key. SK=: short-circuit key for G. First connect a to lower contact point of SK, and take constant of G, using T fo v shunt, and small number of cells, say 5 (depending upon the sen- sitiveness of G), with standard resistance B only in circuit, b being discon- nected as shown. If 5 cells are used in taking constant, and 100 cells are to be used for test, G deflec. X shunt X R X 20 , Constant _-= 1>0 oo,000 = me g° hms ' After obtaining the constant, measure insulation resistance of lead b, by joining it instead of SK to a, disconnecting the far end of b from the cable. The result should be infinity ; but if not, deduct this deflection from the deflection to be obtained in testing the cable proper. Now connect the far end of b to the conductor of the cable, the far terminal of latter being free. Then open SK carefully, and observe if there are any earth currents from the cable. If any, note deflection due to the same, and deduct from bat- tery reading if in the same direction, or add to it if in opposite direction. Short-circuit G with SK, and close one knob of BK, using, say, the T £ — c,; -+ Fig. 6. condensers are now constructed so that these two methods of arranging the plates of a condenser may conveniently be combined in one condenser, thereby obtaining a much wider range of capacities. CABLES. 325 Vesting- Capacity l*y IMrec* Dischargee. — It is frequently de- sirable to know the capacity of a condenser, a wire, or a cable. This may be ascertained by the aid of a standard condenser, a trigger key, and an astatic or ballistic galvanometer. First, obtain a constant. This is done by noting the deflection d, due to the discharge of the standard condenser after a charge of, say, 10 seconds from a given E.M.F. Then discharge the other condenser, wire, or cable through the galvanometer after 10 seconds charge, and note the deflection d'. The capacity c' of the latter is then c being the capacity of the standard condenser. Capacity l>y Thomson's method. — This method is used with accurate results in testing the capacity of long cables. In the figure (Fig. 7) £jp mth|||| — l^QyJ^* CABLE C, Fig. 7. B = battery, say 10 chloride silver cells. R — adjustable resistance. Rj— fixed resistance. G=z galvanometer. C =. standard condenser. 1,2, 3, 4,5, keys. To test, close key 1, thus connecting the battery B, through the resist- ances R, R n to earth. Then V: V,r. R-.R, where Fand V, = the potentials at the junctions of the battery withi? R r Next close keys 2 and 3 simultaneously for, say 5 minutes, thereby char- ging the condenser to potential V, and the cable to potential V Let Cbe the capacity in microfarads of the condenser, and C, capacity of cable, and let Q and Q, be their respective charges when the keys were closed. Then Q : Q, :: VC : V,C r Open keys 2 and 3, keeping key 1 closed for say 10 seconds, to allow the charges of cable and condenser to mix or neutralize, in which case, if the charges are equal, there will be no deflection of the galvanometer when key 5 is closed. If there is a deflection, it is due to a preponderance of charge in Cor C r Change the ratio of R to R„ until no deflection occurs. Then, VC— V, C, or V : V :: C : C . But we found V. : V : : r' : & or Rs'.R.iC.C,. and Cv= w C microfarads. 326 CABLE TESTING. Capacity t>y Oott's Method. — Fig. 8 shows the connections for testing the insulation of a cable by this method, which is considered some- what better than Kelvin's, as it does not necessarily require a well insulated battery. First adjust the resistances R and R t to the proportions of C x to C, as nearly as may be, by moving the slider S. Depress K for five seconds, which will charge both cable and condenser. At the end of the time, de- press k and observe if there is any deflection of the galvanometer G. If there be any such deflection, open A; again, let up the key K, and short- Cable c Fig. 8. Gott's Method of Cable Testing with Condenser. circuit the condenser C x with its plug for a short time, then readjust R and R x and repeat the operation until there is no deflection of the galvano- meter G\ then C :C 1 ::R 1 :R and C = ^ C t . The best conditions for this test are when R and R x are as high as pos- sible, say 10,000 ohms, and C x and C are as nearly equal as possible. Testing* Capacities Uy Lord Kelvin's Dead-Beat, Multi- cellular Voltmeter. — Suitable for short lengths of cable (See Fig. 9.) M V = multicellular voltmeter. AC = air condenser. B = battery. & = switch. Q = total charge in condenser and M V, due to battery. Ca = capacity of AC. C6 = capacity of cable. First close switch S on upper point 1 and charge M V and AC to a desired potential, V. Next move switch S from point 1 to lower point 2, and note the potential V, and M V. Then Q = V (C + Ca) = V,(C + Ca + Cb), where r is the capacity of volt- meter. Ordinarily C can be neglected, as comparea with the capacities of AC and the cable, in which case, by transposition, Cb = (V-V,)Ca = F,. CABLES. 327 Conductors of telephone cables are measured for capacity with the lead sheathing of armor and all conductors but the one under test grounded. Fig. 9. Locating* Breaks in Cables or Overland Wire* by Capa- city Tests. —When the capacity per mile or knot of the conductor of a cable is known its total capacity up to the break is measured by comparison with a standard condenser. Then x = — , , x being distance to fault in miles, m m' capacity of conductor per mile and m total capacity of conductor from the testing station to break. A clear break in the caole or conductor is assumed. Locating- Crosses in Cables or Aerial Wires. — IProf. Ayr- ton Method. — To locate the cross at d (Fig. 10) arrange the connections -tvrf • Fig. 10. as shown. This is virtually a Wheatstone bridge, in which one of the wires, n, is one of the arms of same. Adjust r until a (x -\- y) =. br y when r will be equal to x + y> if a = b. d Fig. 11. 328 CABLE TESTING. Next connect the battery to line m instead of to earth, as in Fig- 11, and adjust a until ax — by. r^ x b Then = — = , , ■ and as x -f- y = r in the first arrangement, v b X r hence, x = =— - -. — • ' b-\- a This test may be varied by transposing G and the battery, in Fig. 9, which is the old method of making this test. Locating* Fault** in Aerial W ires or Cables l>r the loop Test. — Two conductors are necessary for this test, or both ends of a cable must be available at the testing-point. Also it is assumed there is but one defect in the conductor. The resistance of the fault itself is negligible in this test. Measure the resistance L of the loop by the ordinary Wheatstone bridge. Murray Method. — Connect as in Fig. 12, in which a and b are the arms of a Wheatstone bridge, and y x are resistances to fault, the conduc- tors being joined at J" (in the case of aerial wire, for instance). Close key and note the deflection of needle due to E.M.F. of chemical action at fault if any. This is called the false zero. Fig. 12. Now apply the positive or negative pole of the battery, by depressing one of the knobs of reversing key K, and balance to the false zero previously obtained by varying the resistance in arms a or b. Then, by Wheatstone bridge formula, ax =z by, and l — x-j-y y= I —x a 4- b To ascertain distance in knots or miles from 2 to F, divide x by resistance per knot or mile ; to ascertain distance from 1 to F, divide y by resistance per knot or mile. The foregoing test is varied in the case of comparatively short lengths of cable, in the manner shown in Fig. 13, in which the positions of the battery and galvanometer are transposed. Otherwise the test and formula are the same. It is advisable to reverse the connections of cable or conductors at 2 and 1, and take the average of results obtained in the different positions. In this latter method, battery B should be of low resistance, and well insu- lated. Best conditions for making test, according to Kempe.— Resistance of b should be as high as necessary to give required range of adjustment in a. CABLES. 329 Resistance of galvanometer should not be more than about five times the resistance of the loop. CABLE Fig. 13. Varley Loop Test. — Measure resistance of looped cable or conduc- tors as before. Then connect, as shown in Fig. 14, in which r is an adjustable resistance. If currents due to fault be present, obtain false zero as before. Then close key K, and adjust r for balance. In testing, when earth current is present, the best results are obtained when the fault is cleared by the negative pole, and just before it begins to polarize. Fig. 14. Then where x is the distance of fault, in ohms, from point 2 of cable proper. Then x + by the resistance of the cable or conductor per knot or mile gives the distance of fault in knots or miles. When the resistance of the "good" wire used to form a loop with the defective wire, together with that portion of the defective wire from ./ to F, is less than the resistance of the defective wire from the testing station to fault, the resistance r must be inserted between point 1 and the good con- ductor, the defective wire being connected directly to point 1. The formula in this case is x = — - — , x, as before, being the distance to fault in ohms. To Localize Fault when Resistance of Conductor is Known and a Parallel Good Wire is not Available. — Measure by Wheatstone bridge resistance (r) from A to earth through fault F, and resistance (r') from A' to earth through fault, Fig. 15. Let R be resistance of conductor from A to A', x the actual resistance of conductor from A to F and y actual resistance of conductor from A' to F. The* x = R + r- r ' 330 CABLE TESTING. and 2/ = in ohms, from which the distance in feet or miles may be calculated. » R_ £ < ~ *< —, > Fig. 15. Locating- Fault* in Insulated Wires. — The following, so to speak, "rule of thumb," or point to point electro-mechanical methods of locating faults in unarmored cables, in which the defect is not a pronounced one, have been found successful. Warren's Method. — The cable should be coiled on two insulated drums, one-half on each drum. The surface of the cable between the drums is carefully dried. One end of the conductor is connected to a battery which is grounded. The other terminal is connected to the insulated quadrants of an electrometer, the other pairs of quadrants of which are connected to the earth. Both drums being well insulated, no loss of potential is observed after three or four minutes. An earth wire is now connected first to one and then another of the drums, and the fault will be found on the drum which shows the greater fall on the electrometer. The coil is now uncoiled from the defective drum to the other drum, and tests are made at intervals until the defect is found. JF. JTacoo coils the core from a tank to a drum. The battery is con- nected between the tank and the conductor, one end of which is free. A galvanometer is joined between the tank and drum, which need only be partially insulated. The needle shows when the fault has passed to the drum, and it can be localized by running the galvanometer lead along the insulated wire. Copper Resistance, or Conductivity of Cables. The copper resistance of the submarine and underground cables used in telephony and telegraphy is always tested at the factory, usually by the Wheatstone bridge method. In such a case both ends of the cable are ac- cessible. When the cable is laid, if the far end is well grounded, the cop- per resistance may be measured, either by the Wheatstone bridge method, or by a substitution method, as follows: First, note the deflection due to copper resistance of conductor. Then substitute an adjustable resistance box and vary the resistance in the box until the deflection equals that due to cable. This latter resistance is the resistance of the cable. If there are earth currents on the cable, take readings of cable resistance with each pole of battery. Should there be any difference between the results obtained with the respective poles of the battery, the actual resistance will, according to F. Jacob, be equal to the harmonic mean of the two results, i.e., r 4- r where R is the actual resistance, r is the resistance with + pole, / is the resistance with — pole. To measure copper resistance of conductors by the voltmeter, first measure the E.M.F., V of testing battery. Then place the voltmeter in series with the battery and conductor or instrument to be tested, exactly . as a galvanometer would be placed, and note the deflection V in volts. It will be less than in the first instance. Unknown resistance x will be found by the formula: where r is the resistance of the voltmeter coil. CABLES. 331 Tenting 1 Submarine Cable During* manufacture and liajing*. The Core of the cable, that is, the insulated copper conductor, is made, as a rule, in lengths of 2 knots, which are coiled upon wooden drums, and are then immersed in water at a temperature of 75° F. for about 24 hours. The coils are then tested for copper resistance, insulation resis- tance, and capacity ; the results of which tests, together with data as to length of coils, weight, etc., are entered on suitably prepared blanks. After the tests of some of the coils have been made, the jointing up of the cable begins, which is followed by the sheathing or armoring. The joints are tested after 24 hours immersion in water. During the sheathing process, continuous galvanometer or electrometer tests are made of the core, to see that no injury befalls the cable during this process. In fact, practically continuous tests of the cable for insulation resistance, copper resistance, and capacity should be made until the laying of the cable begins. During laying, the cable should be tested continuously, and communica- tion should be* practically constant between the ship and the shore. An arrangement to permit such tests and communication is shown in Fig. 14. SHIP 8H0RE K Fig. 16. In this figure, G x is a marine galvanometer, B is a battery of about 100 cells on ship-board. In the shore station, L is a lever of key K, C is a con- denser, G 2 is a galvanometer. Normally key K is open and the cable is charged by battery B. If, while the cable is being paid out a defect occurs in the insulation, or if the conductor breaks, a noticeable throw of the galva- nometer follows, and the ship should be stopped and the cause ascertained. By pre-arrangement the lever of shore key K is closed, say every 5 minutes, thereby charging the condenser C, which causes a throw of the galvano- meters' needles. If the ship or shore fails to get these periodic signals, or if they vary as to their strength, it indicates the occurrence of a defect. At the end of every hour the ship reverses the battery, which reverses the direction of the deflection of the galvanometers. If the ship desires to communicate with the shore, the battery is not reversed at the hour, or is reversed before the hour. If the shore wishes to speak with the ship, the key K is opened and closed several times in succession. In either event both connect in their regular telegraphing apparatus for conversation. Compound Cables, that is, cables of more than one conductor, have their conductors connected in series for these tests. If there is an even number of conductors, two of them must be connected in parallel. Locating* Faults in Underground Cables. To localize a fault in a conductor of a cable, form a loop consisting of the defective conductor and a good conductor of equal resis- tance and length, with battery E as shown, Fig. 17. Place an ammeter in each leg of loop L. If current in leg A to fault F is /, and current in leg A' to fault is /' ; D being length of loop L and x the distance from A' to fault F t Fig. 17. 332 CABLE TESTING. then / x IL 77 = 7T and x = /' D — x I +/' The compass method of locating faults in underground cables consists, briefly, in sending a constant continuous current of about 10 amperes into the cable through the ground, the current first passing into an automatic reverser which reverses the direction of the current flow every ten seconds. A manhole is then opened near the center of the cable length and a pocket compass laid on the lead sheathing of the faulty cable and observed for say half a minute. If the ground is further from the source of reversed current the compass needle will swing around approximately 180° upon every reversal at the end of each ten seconds interval. The manhole is immediately closed and another opened, say a mile further away from the source of test current, and if no motion of the compass needle occurs, then the fault has been passed and another manhole is opened between the two first positions, and so on until the fault is finally located in a section be- tween two manholes. H. G. Stott, in Trans, A. I. E. E. Hig-ti Voltage or Dielectric Vests of Cables or Other Apparatus. Cables intended for high pressure circuits ranging from 500 to 60,000 volts or more are usually tested at the factory to ascertain their ability to withstand specified voltages. For the lower voltages the cables are generally tested for three or four times the contemplated working pressure. For higher voltages the cables are usually tested for one and a half to twice the working electro- motive force. See standardization rides of A.I. E. E. The present limit for underground power cables is about 30,000 volts. The alternat- ing electromotive force for these tests is supplied by specially de- signed step-up transformers, which must be of sufficient kw. capacity to supply the charging current called for by the cable to be tested. The charging current varies directly as the frequency, directly as the E.M.F., and directly as the static capacity, and as apparent energy (Skinner, Electrical Age, July, 1905) is equal to current multiplied by E.M.F., the apparent output of the transformers required must vary directly as the frequency, directly as the square of the E.M.F., and directly as the static capacity in microfarads of the cable or apparatus under test. For example, an under- ground cable having a static capacity of one microfarad, and tested at 20,000 volts, 60 cycles, requires a testing transformer of 150 kilowatt capac- ity; tested at 40,000 volts the same cable would require a testing trans- former of 600 kilowatt capacity. The testing electromotive force is regulated in several ways, for instance, by means of a rheostat in the field of the generator, as in Fig. 18, or by employing a number of small transformers capable of being connected up, as indicated in Fig. 19, in which the range is from 10,000 to 40,000 volts in steps of 10,000 volts. The voltmeter or vol- CABLES. 333 Supply Mains or Spark Gap Fig. 19. tage indicator may be placed in the primary circuit of the transformer, in which case the E.M.F. in testing circuit is calculated by the ratio of primary to secondary of the transformer, or the voltmeter may be placed directly in the testing circuit. A spark gap in the testing circuit is frequently employed across the cable or apparatus under test (Fig. 19), the E.M.F. in this case being obtained from a table of voltages of spark lengths in air. (See p. 233.) In applying high voltage, say anything above 5000 volts, to a cable or to a piece of apparatus for the purpose of testing its insulation, care should be taken to build it up gradually to the point required; and for this it is best to place a voltmeter across the primary of the testing transformer and place needles for a spark gap across the secondary, gauging their points at the distance given in the rules of the committee of standards of the A.I.E.E. Run the voltage up gradually, reading the voltmeter as the pres- sure is built up. until the current jumps the gap, when the indication of the voltmeter should be carefully taken. When the test is being made the needles should be set about 10*% farther apart, and the pressure obtained can be read on the voltmeter direct. Choke coils of many turns, and other high resistances should be placed in series with each side of the spark gap so as not to cause damage when the gap closes. Water rheostats consisting of glass tubes about 3 feet long, J" diameter, and filled with water, make good high resistance for this purpose. DIRECT-CURRENT DYNAMOS AND MOTORS. Revised by Cecil P. Poole. KfOTAIIOX. Except where other definitions are given, the definitions of the symboli used throughout this section are as follows : — A =: Area in square inches. Ab =: Aggregate area of all brush faces. B« =: Magnetic density in armature core body at full load. &m = Magnetic density in field magnet core at full load. Bj> = Average magnetic density over pole-face at full load. Bt = Magnetic density in armature tooth tops at full load. Bt'=: Approximate magnetic density in armature tooth tops at full load. B< = Magnetic density in armature tooth roots at full load. B*' = Approximate magnetic density in armature tooth roots at full load. Bt = Magnetic density in armature teeth at a specified point. Bt' = Approximate density in armature teeth at a specified point. b = Brush-face dimension crosswise of commutator bars. y = Average distance between interpolar edges of adjacent pole-faces. Da = Diameter of armature core over teeth. Dk =: Diameter of commutator barrel. Do = Diameter of central hole in armature core. Dp m Diameter of pole-face bore. Dt = Diameter of circle drawn through narrowest parts of armature core teeth. d — Diameter of bare round wire, in mils. A = Depth or thickness of winding in a magnet coil. 8 = Air-gap length from pole-face to tops of armature teeth. E =. Total E.M.F. generated in an armature. Ew == E.M.F. delivered by a dynamo or applied to a motor. e r= E.M.F. at terminals of one magnet coil. F = Ampere-turns per pole required by complete magnetic circuit at full load. F = Ampere-turns per pole required by complete magnetic circuit at no load. Fa = Ampere-turns per pole required by armature core at full load. F g = Ampere-turns per pole required by air-gap at full load. Fm = Ampere-turns per pole required by magnet core at full load. F P = Ampere-turns per pole required by pole-piece or shoe at full load. Fr =. Ampere-turns per pole required to balance full-load armature reaction. F$ = Ampere-turns per pole in series field-winding at full load. Fth =. Ampere-turns per pole in shunt field-winding at full load. Ft = Ampere-turns per pole required by armature teeth at full load. F v =z Ampere-turns per pole required by field-magnet yoke at full load. / = Ampere-turns per inch length of magnetic path at full load : Subscripts a, m, p, t and y apply to armature core, magnet core, pole-shoe, armature teeth and magnet yoke, respectively. G — Girth or perimeter of a complete magnet coil. g = Girth or perimeter of form or bobbin on which a magnet coil is wound. h =z Depth of armature coil slot. la — Total armature current. I»h ~ Shunt field current. Iu> — Current delivered from a dynamo, i r= Current in a specified conductor, or coil. i a == Current in each armature conductor. kch = sin (180 t+p); kch Dp = chord of polar arc. kg = a coefficient ; kg 8 = increase of air-gap span due to flux spread. kg 2 = a coefficient ; k ff2 8 = increase of air-gap width due to flux spread. k* z=. Number of commutator bars between the two to which the termi. nals of each armature coil are connected. 334 NOTATION. 335 X« = Length of magnetic path in armature core beneath slots. Lf — Length of a specified field-magnet coil parallel to flux path. Lm = Length of magnetic path in one field-magnet core. L P = Length of magnetic path in one magnet pole-piece or shoe. Ly = Length of magnetic path in field-magnet yoke between adjacent poles. le = Total length of each armature conductor. m =2 Number of windings in a multiplex armature winding. Ke =. Total number of armature conductors around armature periphery. Nk = Number of commutator bars and armature coils Nt = Number of armature teeth (and slots). rik = Maximum number of commutator bars simultaneously in contact with one brush at any instant. v — Coefficient of magnetic leakage. P A = Total watts lost in armature. P' A = Total watts lost in armature exclusive of projecting parts of the winding. Pb z=. Watts lost at all brush faces. Pe = Watts lost by eddy currents. Ph z=: Watts lost by hysteresis. Pr = Watts lost in entire armature winding alone. Pr' r= Watts lost in armature winding exclusive of projecting parts. Pa = Watts lost in series field-magnet winding. Psh — Watts lost in shunt field-magnet winding. P w = Watts of dynamo armature output or motor armature intake. p = Number of field-magnet poles. q = Number of parallel paths through an armature winding ; Note: — In a multiplex winding, q = total paths in all the windings. R =p Resistance of armature, commutator and brushes, warm. Pa = Resistance of armature winding, warm. R a ' = Resistance of embedded part of armature winding, warm. Rb z=z Effective resistance of all brush-face contacts ; I*Rb = Volts drop at brush faces. r rr Resistance of a specified conductor or coil in ohms, r.p.m. == Revolutions per minute. r.p.s. == Revolutions per second. 5 = Width of one armature coil slot. T = Torque in pound-feet. T =z Width of one armature tooth at the top. t ■=. Width of one armature tooth at the narrowest part, except in equation 32 and Table V. t =z Number of turns per armature coil ; only in equation 32 and Table V. A = Temperature rise of armature, Fahrenheit degrees. 6k rr Temperature rise of commutator, Fahrenheit degrees. 0/ = Temperature rise of field winding, Fahrenheit degrees. t = Width of one armature tooth at a specified point. $ z=z Magnetic flux passing from one pole-face to armature at full load. $m = Magnetic flux in magnet core at full load.

3000 / / 2500 / 2000 / 1500 / CHARACTERISTIC CURVE SPEED-B00 REV. PER MIN. 1000 / 500 / 4 5 7 8 9 10 11 12 13 14 AMPERES Fig. 3. Characteristic curve of Brush 125-Light Arc Dynamo without Regulator, 338 DYNAMOS AND MOTORS. machine without any regulator. The readings were all taken at the spark- less position of commutation. This curve is remarkable from the fact that after we get over the bend, the curve is almost perpendicular, and is prob- ably the nearest approach to a constant current machine ever attained. By winding more wire on the armature the machine could have been made to deliver a constant current of 9.6 amperes at all loads, without shunting 100 m . so \n / t "'GO / ELECTRICAL EFFICIENCY AMPERES 9.6 8PEEO 600 R.P.M. / 50 40 8S 2000 3000 4000 5000 6000 VOLTS Fig. 4. Electrical Efficiency Curve of Brush 125-Light Arc Dynamo. J t / COMMERCIAL IFFICIENCY 5 9.6 / AMPERE SPEED 600 R.P.M. 3000 4000 VOLT8 Pro. 5. Commercial Efficiency Curve of Brush 125-Light Arc Dynamo. any of the current from the field ; but this would have increased the internal resistance, and also have made the machine much less efficient at light loads. By the present method of regulation the 1*R loss at one-quarter load is reduced from 4,018 to 3,367 watts, the gain being almost one electrical horse-power. Fig. 4 is a curve of the electrical efficiency. It will be noticed that this at full load reaches 94 per cent, which is accounted for by the liberal allow- ance of iron in the armature, thus reducing the reluctance of the magnetic circuit, and by the large size of the wire used on both field and armature. Fig. 5 is a curve of the commercial efficiency. At full load this is Dver 90 per cent, and approaches very closely the efficiency of incandescent dynamos of equal capacity, but the most noteworthy point is the high effi- ciency shown at one-quarter load. Fig. 6 is a curve of the machine separately excited, with no current in the armature. The ordinates are the volts at the armature terminals, and the abscissae the amperes in the field. This is in reality a permeability curve of the magnetic circuit. By a comparison of the voltage shown here when DYNAMO CHARACTERISTICS. 339 there are nine amperes in the field, -with that of the machine when deliver- ing current, can be seen the enormous armature reaction. The curve also 9000 1000 6000 5000 ] 4000 8000 E. M. F. 1000 500 1 Fig. 1&345678 9 10 11 12 13 14 M AMPEBE8 6. Permeability Curve of Magnetic Circuit of Brush 125-Light Arc Dynamo. indicates a new departure in arc dynamo design, namely, that the magnetic circuit is not worked at nearly as high a point of saturation as in the old types. Shunt dynamo. The shunt dynamo has, besides an external characteristic, shown below, an internal characteristic. The first is developed from the volts read while the load in amperes is being added, the armature revolu- tions being kept constant (See Fig. 7.) Adding load to a shunt dynamo means simply reducing the resistance of the external circuit. With all shunt machines there is a point of external resistance, as at n, beyond which, if the resistance is further reduced, the volts will drop away abruptly, and finally reach zero at a short circuit. /ft Fig. 7. External Characteristic Fig. 8. Internal Character- of Shunt-wound Dynamo. istic of Shunt Dynamo. The internal characteristic, Fig. 8, or, more correctly, curve of magnetiza- tion, of a shunt dynamo, is plotted on the same scale as those previously described, from the volts at the field terminals and the amperes flowing in the field winding. 340 DYNAMOS AND MOTORS. The resistance line o a only applies to the point a on the curve, and the resistance value a b for that point is determined by ohms law, or as fol- lows : As the curve of magnetization is determined from the reading of volts plotted vertically and amperes horizontally, and asr=yOrr= — r and ^-r = tang aob, therefore the resistance at any point on the curve will o b be the tangent of the angle made by joining that point to the origin o. Compound dynamo. As the compound dynamo is a combination of the series and shunt machines, the characteristics of both may be obtained from it. The external characteristic is of con- siderable importance where more than one dynamo is to be connected to the same circuit, or when close regulation is necessary. Fig. 9 is a sample curve from a com- pound-wound dynamo, where the in- crease of magnetization of the fields due to the series coils and load causes amperes the terminal voltage to rise as the load ._, _ ~, . . 4. « A „„„ is increased. This is commonly done Fig. 9. Characteristic of Over- tQ make for d in feeder / to the compounded Compound - wound cen tre of distribution. It is impossi- Dynamo. kl e m ordinary commercial dynamos to make this curve closely approach a straight line, and the author has found it difficult for good makes to approach a straight line of regulation nearer than 1£ per cent either side of it for the extreme variation. Curve of IWtagrnetic ^Distribution. — This curve is constructed from existing dynamos to show the distribution of the field about the pole- pieces ; it can be plotted on the regular rectangular co-ordinate plan, or on the polar co-ordinate. The following cuts illustrate the commonest methods of getting the data for the curve. With the dynamo running at the speed and load desired, the Fig. 11. pilot brush, a, in Fig. 10, or the two brushes, a and b, in Fig. 11, is started at the brush x, and "moving a distance of one segment at a time, the difference in volts between the brush x and the location of the pilot brush, a, is read on the voltmeter. Where the one pilot brush is used, the total difference between that and the origin is read ; while with two brushes, as a and 6, which are commonly fastened to a handle in such a manner as to be the width of a segment apart, just the difference between the two adjacent segments is read, and the total difference is determined by adding the individual differences together. ARMATURES. 341 ARMATURE!. Direct-current armatures are divided into two general forms, — drum arma- tures, in which the conductors are placed wholly on the surface or ends of a cylindrical core of iron ; and ring armatures, in which the conductors are wound on an iron core of ring form, the conductors being wound on the out- side of the ring and threaded through its interior. Another form used somewhat abroad is the disk armature, in which the conductors are arranged in disk form, the plane of which is perpendicular to the shaft, and without iron core, as the disk revolves in a narrow slot be- tween the pole-pieces. Armatures of the slotted or toothed core type are almost exclusively em- ployed now. The coils are set into the slots, with the results that eddy cur- rents in the conductors are prevented and the conductors are positively driven by the core teeth. The cores are built up of sheet steel disks in small sizes, annular sheets in medium sizes, and staggered circular segments in large sizes ; the steel is from 15 to 25 mils thick and the sheets are clamped firmly together by end-plates. In order to prevent eddy currents in the core, the disks or sheets are either coated with an insulating varnish or separated by tissue paper pasted over the entire surface of one side of each disk or sheet. The toothed armature has the following advantages and disadvantages as compared with the smooth body: Advantages. 1. The reluctance of air-gap is minimum. 2. The conductors are protected from injury. 3. The conductors cannot slip along the core by action of the electrody- namic force. • 4. Eddy currents in the conductors are almost entirely obviated. 5. If the teeth are practically saturated by the fiefd magnetism, the> oppose the shifting of the lines by armature reaction. Disadvantages. 1. More expensive. 2. The teeth tend to generate eddy currents in the pole-pieces. 3. Self-induction of the armature is increased. If the slots can be made less in width than twice the air-gap, so that the lines spread and become nearly uniform over the pole-faces, but little effect will be felt from eddy currents induced in the pole-faces. When it is not possible to make such narrow slots, pole-pieces must be laminated in the same plane as the disks of the armature core, or the gap must be con- siderably increased. Hysteresis in the armature core can be avoided to a great extent by using the best soft sheet iron or mild steel, which must be annealed to the softest point by heating to a red heat and cooling very slowly. Disks are always punched, and are somewhat hardened in the process ; annealing will entirely remove the hardness, and any burrs that may have been raised. Disks should be punched to size so carefully as to need no filing or trueing up after being assembled. Turning down the surface of a smooth-body armature core burrs the disks together, and is apt to cause dangerous heating in the core when finished. Light filing is all that is permissible for truing up such a surface. Slotted cores should be filed as little as possible, and can sometimes be driven true with a suitable mandrel. Armature shafts must be very strong and stiff, to avoid trouble from the magnetic pull should the core be out of center. They are made of machin- ery steel, and have shoulders to prevent too much endwise play. Core Insulation. — A great variety of material is used for insulating the core, including asbestos, which is usually put next to the core to prevent damage from heating of that part, oiled or varnished paper, linen, and silk ; press board ; mica and micanite. For the slots of slotted cores the insula- tion is frequently made into tubes that will slide into the slots, and the con- ductors are then threaded through. Special care must be taken at corners and at turns, for the insulation is often cut at such points. 342 DYNAMOS AND MOTORS. Armature Winding-*. For all small dynamos, and in many of considerable 6ize, the winding is of double cotton-covered wire. Where the required carrying capacity is more than that of a No. 8 wire, B. & S. gauge, the conductor should be stranded for smooth-core armatures. In large dynamos, rectangular cop- per bars, cables of twisted copper, and in some cases large cable compressed into rectangular shape, are more commonly used. If the copper bars are too wide, or wide enough so that one edge of the bar enters the field percep- tibly before the remaining parts of the bar, eddy currents are induced in it ; such bars are therefore made quite narrow, and it is common to slope the pole-face a trifle, so that the bars may enter the field gradually. Method* or arrangement of windings are of a most complex nature, and only the most general in use will be described here, and these only in theory ; Parshall & Hobart have described about all the possible combinations ; S. P. Thompson, Hawkins &Wallis, and others have also written quite fully on the subject. Ring- or Gramme Winding**. There are two fundamental types of armature winding : ring and drum. In a ring-wound armature, the core is necessarily annular, the wire being wound through the core as well as along the exterior, as indicated in Figs. 12 to 15. This form of winding is now used only in arc-light dynamos and very small motors. The simplest form of ring winding is the two-circuit single winding, where a continuous conductor is wound about the ring, and taps taken off to the commutator at regular intervals. Fig. 12. Fig. 13. Fig. 14. The first variation on this will be the multi-circuit single winding, used where there are more than one pair of poles. Fig. 13 shows the four-circuit single winding. Where it is advisable to reduce the number of brushes in use, the multi- circuit winding can be cross-connected ; that is, those parts of the winding occupying similar positions in the various fields are connected in parallel to the same commutator bar. Fig. 14 shows one of the simplest forms of cross- connected armatures. Where, from the shape of the frame, the magnetic circuits are somewhat unequal, the winding shown in Fig. 15 will average up the unequal induc- tion values, and prevent sparking to some extent. It also halves the number of commutator segments ; that is, there are two coils connected ARMATURES. 343 to each segment instead of one, as in the previously mentioned windings. If Nk = number of coils, and p = number of poles, each coil is connected (?*»)* in advance of it. across to a coil \ p i Two-Circuit Winding's for Multipolar Field's. —This is an important class of windings, and, as it has but two circuits irrespective of the number of poles, has the advantage over the multiple-circuit windings 2 that it needs but — as many conductors as are necessary in that class. But two sets of brushes are necessary for the two-circuit windings, unless the current is heavy enough to require a long commutator, in which case other sets of brushes can be added, up to the number of poles. Fig. 15. Ring Winding Cross-connected to Reduce Unequal Induction. In the short-connection type of this class, conductors under adjacent field poles are connected together so that the circuits from brush to brush are influenced by all the poles and are therefore equal. In the long-connection type the conductors under every other pole are con- nected, so that the conductors from brush to brush are influenced by but one-half the number of poles. The number of coils in a two-circuit long-connection multipolar winding is determined by the formula **=? y ± i, where Nk = the number of coils, p = the number of poles, and y = the pitch. The number of commutator segments is equal to the number of coils and must be a number not divisible without a remainder by the num- ber of pairs of poles. The pitch, y, is the number of coils advanced over for the connections, as, for instance, m an armature with a pitch of 7 the end of coil number 1 is connected to the beginning of coil 1+7 = 8, and from 8 to 8 + 7 = 15, and so on. In multipolar ring long-connection windings y may be any integer. Mr. Kapp gives in the following table the best practice as to angular dis- tance between brushes for this class of windings. 344 DYNAMOS AND MOTORS. Number of poles. Angular distance between brushes. Degrees. Degrees. Degrees. Degrees. Degrees. 2 180 4 90 6 60 180 8 45 135 10 36 108 180 12 30 90 150 14 25.7 77 128 180 16 22.5 67.5 112 158 18 60 100 140 180 20 54 90 126 162 Fig. 16 shows a simple form of two-circuit multipolar single winding, and Fig. 17 another sample as used with a greater number of poles. WmMm Fig. 16. Two-path Multipolar Windings. Fig. 17. Both of the above samples are of the long -connection type. In the short* connection type the formula for determining the number of the coil is m = py ± 2, and Fig. 18 is a sample diagram of this type. ARMATURES. 345 Fig. 18. Short-connection Two-path Ring Winding. Dram Winding's. In order that the E.M.F.'s generated in the coils of a drum armature may be in the same direction, it is necessary that the two sides of each coil be in fields of opposite polarity, and therefore the sides of the coils are connected across the ends of the core ; directly across, for bipolar machines, and part way so for those of the multipolar type. Fig. 19. Bipolar Drum Winding. The drum winding is wholly on the exterior of the core. Fig. 19 is a dia- gram of a bipolar drum winding on a smooth core ; the dotted lines indicate the crossings of the wires over the rear head of the core. Drum windings are mostly of the two-layer type, of which Fig. 20 is a diagram; with a slotted core, the numbered conductors would lie within the slots. In this diagram each pair of conductors having numbers differing by 15 compose the two " sides " of one coil, and are therefore integral with each other. 346 DYNAMOS AND MOTORS. There are two general types of drum winding : lap and wave. If each coil has more than one turn, " lap-connected" and "wave-connected" are more appropriate distinguishing terms. Bipolar machines necessarily Fig. 20. Bipolar two-layer drum winding. FIG. 21. Two-path single four-pole winding have lap-connected windings. In multipolar machines the two " sides " of each coil are located a distance apart approximately equal to the pol« pitch instead of on opposite sides of the core (see Fig. 21). The proportion of armature circumference spanned by each coil is preferably a trifle less ARMATURES. 347 than the pole pitch ; for a toothed armature the number of teeth embraced by each coil should be equal to Nt-^p — xt . If Kt — p is a whole number, xt = 1 ; if it is a mixed number, xt = the fractional part or 1 -\- that part ; it should seldom exceed 2 in any case. All lap windings have p m parallel paths. A multiplex winding consists of two or more distinct windings, the conductors of which are arranged in regular sequence around the core ; the windings are connected to m sets of commutator segments assembled in a single commutator, as indicated Fig. 22. Six-path single drum winding. by Fig. 23. The terminals of each coil of any lap winding must be con- nected to two commutator segments between which there are m — 1 other segments. Wave-connected windings may have any even number of parallel paths regardless of the number of magnet poles, within practical limits. The number depends on the number of coils and method of connecting them. The relation between the number of coils (and commutator segments), num- ber of paths, number of magnet poles and method of connection is as follows : — Nh __ (l+l' 8 )p± q ~~ 2 and h= 8J ft*g-l (2) (3) The smaller value of ka is preferable, but choice between the two is usu- ally determined by the choice between the resulting classes of winding. If k» + 1 and JVk have a common factor, the winding will be of the plural or multiplex type ; if not, a simple wave-connected winding will result, pro- vided qT^p. In slotted armatures the number of conductors must be a multiple of the number of conductors per slot. 348 DYNAMOS AND MOTORS. Fig. 23 is a diagram of a two-path triplex winding, i.e., three two-path windings connected in parallel by the brushes. It is mathematically the equivalent of a single six-path winding. Fig. 23. Fig. 24 shows diagrammatically the characteristics of the usual two-path armature winding used on street railway motors, in which there are three times as many coils as there are slots. In this case xt ~ 0.25 and k» == 48. Fig. 24. ARMATURES. 349 Balancing* the Magnetic Circuits in Dynamo*. Difficulty has been experienced in the operation of large multipolar direct- current machines with parallel wound armatures, owing to differing mag- netic strengths in the poles. The potential generated in conductors under one pole differed from that generated in conductors similarly situated under another pole of the same polarity, the result being a slight difference of potential between brushes of similar polarity. This caused currents to flow from one brush to another, and from one section of the armature winding to another, attended by wasteful heating of conductors and sparking at the brushes. This difficulty is obviated by the Westinghouse Electric & Manu- facturing Company by the following method of balancing : A number of points in the armature winding corresponding to the num- ber of pairs of poles, which are normally of equal potential, are connected by leads through which currents may pass from one section to the others with which it is connected in parallel. The currents are alternating in character 'and lead or lag with reference to their respective E.M.E.'s. They thus magnetize or demagnetize the field magnets and automatically produce the necessary balance. This method of balancing is also of advan- tage in eliminating the sparking at the brushes and the wasteful heating, which occur when an armature becomes decentralized, owing to wear of the bearings, or to other causes. When an armature gets out of center the air-gap on one side is greater than the air-gap on the opposite side. The potential generated in the coils — if the armature has the ordinary multiple winding — will be much greater on the side having the smaller air-gap than that generated under poles of the same polarity on the opposite side. Con- sequently, a current corresponding to this difference of potential flows through the brushes from one section of the winding to another. This flow of current will act the same as if two generators were coupled rigidly on one shaft and the potential of the one raised above that of the other. The machine having the higher potential would act as a generator, and the other would run as a motor. This, of course, would result in bad sparking and the burning of the brushes. By the use of the above balancing method, however, the armature could be considerably out of center and no injurious results occur, as the balanc- ing currents flow, not through the brushes, but, as explained above, through specially provided connections. In addition, the currents in these conduc- tors are alternating currents — " leading" in some coils and " lagging " in others — a fact which enables a relatively small current to balance the cir- cuits effectively. Heating- of Armatures. The temperature an armature will attain during a long run depends on its peripheral speed, the means adopted for ventilation, the heating of the conductors by eddy currents, the heating of the iron core by hysteresis and eddy currents, the ratio of the diameter of the insulated conductor to that of its copper core, the current density in the conductor, the radial depth of winding, whether the armature is of cylinder or drum type, and the amount and character of the cooling surface of the wound armature. The higher the peripheral speed of the armature the less is the rise of temperature in it. Mr. Esson gives, as the result of some experiments on armatures with smooth cooling surfaces, the following approximate rule : 55 P A 350 P x "" £(1+0.00018 V) T S'{ 1 + 0.00059 V) ' where A = difference of temperature between the hottest part of the arma- ture and the surrounding air in degrees, Centigrade, P A = watts wasted in armature, S = .ictiVe cooling surface in square inches, S' = active cooling surface in square centimeters, V = peripheral speed of armature in feet per minute, V =. peripheral speed in meters per minute. 350 DYNAMOS AND MOTORS. The more efficient the means adopted for ventilating the armature by currents of air, the smaller is the temperature rise. Some makers leave spaces between the winding at intervals, thus allowing the air free access to the core and between the conductors. A draught of air through the in- terior of the armature assists cooling and should be arranged for whenever possible. For heavy currents it is sometimes necessary to subdivide the conductors to prevent eddy currents; stranded conductors, rolled or pressed hydraulic- ally, of rectangular or wedge-shaped section, have been used. Such sub- division should be parallel to the axis of the conductor, and preferably effected by the use of stranded wires rather than laminae. Few armature conductors of American dynamos of to-day are divided or laminated in any degree whatsoever. Solid copper bars of approximately rectangular cross- section are often used, and little trouble is found from Foucault currents. Mr. Kapp considers 1.5 square inches (9.7 square centimeters) of cooling surface per watt wasted in the armature a fair allowance. Esson gives the following for armatures revolving at 3000 feet per minute : P x = watts wasted in heat in winding and core, S = cooling surface, exterior, interior, and ends, in square inches, S' = cooling surface, exterior, interior, and ends, in square centi- meters, A = temperature difference between hottest part of armature and surrounding air in C°. Then _ 35 P A w 225 P A Specifications for standard electrical apparatus for IT. S. Navy say, " No part of the dynamo, field, or armature windings shall heat more than 50° F. above the temperature of the surrounding air after a run of four hours at maximum rated output." According to the British Admiralty specification for dynamos, the tem- perature of the armature one minute after stopping, after a six hours' run, must not exceed 30° F. above that of the atmosphere. In this test the ther- mometer is raised to a temperature of 30° F. above that of the atmosphere before it is placed in contact with the armature, and the dynamo complies (or does not comply) with the specification according as the thermometer does not (or does) indicate a further rise of temperature. The best dynamo makers to-day specify 40° and 45° C. as the maximum rise in temperature of the hottest part of a dynamo, or 55° if the tempera- ture of the commutator surface is to be measured. Armature Reactions. In many direct-current dynamos having no special devices for reversing the current in each armature coil as it passes through the " commutating zone," it is necessary to give the brushes a forward lead so that the mag- netic fringe from the pole-tip toward which the coil is moving may induce an E.M.F. in the coil and reverse the current. In motors the brushes are shifted rearward instead of forward, the polarity of the approaching pole- tip being of the wrong sign. With the forward lead given to the brushes the effect of the armature cur- rent is to weaken and distort the magnetic field set up by the field mag- nets ; a certain number — depending on the lead of the brushes — of the ar- mature ampere-turns directly oppose those on the field-magnets and render a somewhat larger number of these ineffective, except as regards wasting power ; the remaining armature ampere-turns tend to set up a magnetic field at right angles to the main field, with the result that the resultant field is rotated forward in the direction of motion of the armature, and that the field strength is reduced in the neighborhood of every trailing pole-piece horn, and is increased in that of every leading pole-piece horn. When, therefore, the brushes have a forward lead each armature section as it comes under a brush enters a part of the field of which the strength is reduced by ARMATURES. 351 the armature cross-induction ; and, if this reduction is great, the field strength necessary for reversing the current in the section (in the short time that the section is short-circuited under the brush) may not be ob- tained, and sparkless collection may thus be rendered impossible. Various devices for reversing the currents in the armature sections, as they pass successively under the brushes, without giving a forward lead to the brushes, have been proposed ; a number of these were described in a Saper by Mr. Swinburne ; an improvement by Mr. W. B. Sayers consists in iterposing auxiliary coils between the joints of adjacent armature sections and the corresponding commutator bars. Each auxiliary coil is wound on the armature with a lead relative to the two main armature sections and the commutator bar which it connects together. The result of this arrange- ment is that the difference between the E.M.F.'s in the two auxiliary coils connecting any given armature section to the two corresponding commutator bars may be made sufficient to reverse the current in the armature section when short-circuited under a brush, even if the brush has a backward in- stead of a forward lead. In the Thompson-Ryan dynamo the effects of armature reaction are neu- tralized by a special winding through slots across the faces of the pole- pieces, parallel with the axis of the armature ; this winding is in series with the armature, and the same current flowing in both, but in such direction that all effects on the field magnets are neutralized, the ampere-turns of the shunt are therefore much less than in other dynamos, there is no sparking , under any ordinary conditions of load, the brushes are placed permanently when the machine is set up, and the efficiency is high through a wide range. The method which is most widely employed is to put small auxiliary field-magnet poles between the main poles and connect their windings in series with the armature. This method is applied chiefly to constant-poten- tial motors designed to run at several speeds. Drag- on Armature Conductors. _ i n dynamos, each armature conductor has to be driven in opposition to an effort or drag proportional at every instant to the product of the current carried by the conductor into the strength of the magnetic field. This drag on a conductor, varies, there- fore, with the position of the conductor relative to the field-magnet poles, and is a maximum when the conductor passes through that part of the air- gap at which the magnetic induction is greatest. The arrangements for driving the armature conductors must, of course, be adapted to the greatest value of the drag to which a conductor is exposed, and this is given for smooth core armatures by the formula below. Let %• = current in amperes carried by each conductor, (ft = maximum induction in air-gap per square centimeter, B = maximum induction in air-gap per square inch, Wa = length of armature core in inches. mv ($>Waia B Wa la -, . „ . _. Then 1^300 = 13,302,360 = Maximum P ul1 m lbs - on each conductor. In slotted armatures the core teeth take the drag. COJ1 yi\ T II OH* AID BRUSHES, Commutators are built up of the best grade of copper, preferably hard drawn. The insulation between segments should invariably be of the best quality of amber mica ; white mica is usually too hard and brittle, and does not wear down as rapidly as the copper segments, so that eventually the mica strips project above the surface and cause the brushes to chatter and spark. The insulation at the ends is usually of micanite, and should be as hard as possible. Brushes are invariably of carbon except on machines built for very low voltages. The high resistance of the carbon reduces the "short-circuit" current in a coil undergoing commutation and also reduces the inductive opposition to the reversal of current in the coil, thereby facilitating com- mutation. The current density under brush faces should not exceed 60 amperes per square inch for carbon, 200 for woven wire, or 250 amps, per sq. in. for soft 352 DYNAMOS AND MOTORS. leaf copper brushes. The proper density to be used in any given case de- pends upon other features of commutator and brush proportions. See 44 Practical Design," page 361. FIEJLO Mie]¥ETS. Field magnets are bipolar in small sizes and multipolar in large sizes | the dividing line between bipolar and multipolar construction varies from 1 kilowatt to 10 kilowatts ; it is quite common practice to make machines of 5 kilowatts and over multipolar. Magnet cores are made either of wrought iron or steel, except in very small machines in which cast iron is used. Pole-pieces or shoes are of either cast iron or steel, according to their shape and disposition. Cast-iron shoes are attached to the sides of the pole and merely extend the pole-face surface ; steel shoes are bolted against the free ends of the poles so that the entire air-gap flux passes through the shoe. In many cases no shoes are used, the poles being carried to the air-gap without change in cross-section, or else provided with integral polar extensions at the free ends. Field magnet yokes are either of cast steel or cast iron. The latter is preferable on every score except weight, for the reason that steel castings are seldom perfectly sound throughout and rarely within & inch of calcu- lated dimensions. Magnet cores are generally bolted to the yoke, but a few builders still " cast-weld " them in. Coil Surface Necessary for Safe Temperature. Esson gives the following method of determining the surface necessary for a magnet coil to keep its heat within assigned limits. Let P == watts wasted in heating, S = cooling surface in square inches of coil, not including end flanges and interior, ^rr same as above in square centimeters, = temperature of hottest part above surrounding air, then F.°= 99 ^ or C.°= 335 ^. Maximum current -/ degs.F. x sq. ins. 99 X hot r Hot r = cold r -}- 1% for each additional 4.5° F. Table of Cooling- Surfaces. Excess temperature above sur- rounding air. Cooling surface per watt in F.° C.° square inches. sq. centimeters. 15 3.67 23.7 30 — 3.30 21.3 20 2.75 17.8 40 — 2.48 16.0 — 25 2.20 14.2 50 — 1.98 12.8 30 1.83 11.8 60 — 1.65 10.7 — 35 1.57 10.1 70 — 1.41 9.1 ~~ 40 1.38 8.9 DIRECT-CURRENT MOTORS. 353 Ctyrostatic Action on Dynamos in Ships. L — (Lord Kelvin.) and P = where g* L = moment of couple on axis, P=z pressure on each bearing, Wz=z weight of armature, k = radius of gyration about axis, n=r -= A ■=. maximum angular velocity of dynamo in radians per second due to rolling of ship, A == zrr^ = amplitude in radians per second, loO (Radian is unit angle in circular measure.) d = degrees of roll from mean position, Tz=. periodic time in seconds, F A1TIO DESIGN.* It is safe to follow the rule of using bipolar field-magnets for machines of 4 kilowatts or less and multipolar magnets for larger machines. For commutation reasons the current passing any one set of brushes should not exceed 250 amperes ; this gives a criterion of the number of poles for machines of 250 amperes output or more. Lap windings should be used on such machines. Then P — 0.008 la (6) The number of poles on machines having wave-connected armatures is determined by commutation considerations chiefly ; more than six poles are seldom used. The best construction is a laminated magnet pole with extensions at the air-gap end, bolted to a cast-steel yoke. Fairly good results are obtained, however, with cast-steel poles. Laminated cores, cast-welded into either iron or steel yoke and provided with cast-iron shoes embracing the ends at the air-gap, give excellent results if the cast-welding is properly done. When the ratio of air-gap length to the width of each armature core-slot opening is much less than 0.5, the pole-face should be laminated in order to prevent excessive eddy curents in it ; otherwise it may be solid. A cast-iron pole-shoe must not cover the end of the magnet core, but should surround it and serve merely as lateral extensions ; the cross section of the core should be slightly reduced where it is surrounded by the pole-shoe. * Cecil P. Poole. 356 DYNAMOS AND MOTORS. The E.M.F. generated in the direct-current armature is, from eq. (X), p op q 60 which reduces to E = 0.05236 Dp Wp \f/ Bp & r.p.m. 10 -8 -J- q. The output in watts is Pw =. Ew Iw, which for preliminary purposes may be considered the equal to E ia q ; whence Pu> — 0.05236 Dp Wp^i Bp i« Nc r.p.m. 10 ~ 8 .... (7) For economical use of material, the projected outline of a pole-face should be square, so that the width parallel to the armature shaft should approxi- mately equal the chord of the average polar arc; whence Wp should be— Dp sin (180 \p -f- p). For moderately high-speed machines, ^ may be taken at 0.7 ; for slightly lower speeds, at 0.72, and for slow-speed machines, at 0.75. For reversing motors it is best put at 0.6666, except series-wound reversing motors ; for these, let $ =: 0.7. Representing sin (180 i// — p) by kch, page 371, results. The average magnetic density over the pole-face ranges from 25,000 to 60,000 lines per square inch, according to the designer's method and the size of the machine. It is rational to make Bp = cX Dp 0,15 ,c being a coefficient varying according to the type of machine. For constant-potential dynamos and motors for general service, 28,120 is a suitable value for c ; for shunt or compound-wound reversing motors, 33,850 is appropriate, and for series re- versing motors, 36,620. The permissible number of ampere-conductors around the armature peri- phery ranges from 1200 to 2200 per inch of armature diameter. For ma- chines designed according to the method outlined herein, it is good practice to apply the formula: . ._ he Dp*-™ taNeZ=Z p The values of kc are as follows : Dynamos and motors for general service, kc = 679. Shunt and compound reversing motors, kc = 564. Series-wound reversing motors, kc = 678. From the foregoing equation an equivalent is obviously obtainable for iaNc «//, and substituting this and the equivalents for Bp a ^d Wp previously obtained, equation (7) reduces to the following two : For all machines except series-wound reversing motors : _ fc*ZV».s r .p. m . p "~ ioo — (8) For series-wound reversing motors : Pw = 0.013 kch Dp™ r.p.m (9) For belted machines which need not have any particular rate of speed, an economical rate is 8500 Considering Da and Dp equal, which is allowable in preliminary " rough- ing out," and substituting in equation (8) the above equivalent for r.p.m.: Pw = 85 kch Dp*- 6 (10) Armature Details. — Core disks 25 mils thick may be used in most armatures ; only those in which the core is subjected to high rates of mag- netic reversal need have thinner disks. When p x r.p.m. exceeds 3000. it is advisable to use disks 20 mils thick, or less ; wnenp x r.p.m. exceeds 4000, 15 mils should be the limiting thickness. The final criterion, however, is the eddy current loss in the core and teeth. PRACTICAL DYNAMO DESIGN. 357 Having a means of determining the pole-face width parallel to the arma- ture shaft, the length of the armature core follows within close limits. The armature core should extend beyond the edges of the pole-face at each end by a small amount — not less than the air-gap length, and preferably 1.5 times the air-gap. Armature cores more than 5 inches long should have ventilating ducts not less than § inch wide at intervals of 2J to 3^ inches. The exact duct width is usually determined by the amount of steel required parallel to the shaft in order to keep the magnetic density in the teeth within suitable limits. The" nominal " magnetic density at the narrowest part of the teeth should be between 140,000 and 155,000 lines per square inch of net cross section. The " nominal " density is that which would exist if the flux did not spread beyond the geometrical contour of the pole-face in passing from the latter to the armature, and if all of the flux passed through the teeth ; that is, --. ,-*. — = nominal density at tooth roots, Nt^tWa wa = 0.9 (Wa— ventilating ducts). In order to obtain dimensions that will result in a " nominal " density at the roots of the teeth that will be within the specified range, the number of teeth (and slots) may be approximated by means of the formula -nDt- Nt = - kt Wa (11) The number of teeth must, of course, be an integer ; if the result of eq. (11) should be a mixed number, therefore, the fractional part should be discarded if it is 0.8 or less ; if it be more than 0.8, the next higher integer is to be taken as the number of teeth. The net measurement of the armature iron parallel with the shaft must then be corrected to satisfy the equation, kt ' it Dt — sNt (12) The value of kt for all cases is __ 0.053 Dp* W P When the armature conductors are round wires, the size of the coil slot is determined chiefly by the size and arrangement of the wires. Form-wound WW/////. Fig. 29. and separately-insulated coils are generally used, so that the coil slot is ordinarily of one of the shapes shown in Fig. 29, the slots a or b being used when binding wires are employed to keep the wires in their slots, and one of the others when the coils are held in by wedges. Two-layer windings are almost invariably used in this country. Fig. 30 shows two half-coils *• abreast " in each layer, each coil having three turns of wire ; this makes 358 DYNAMOS AND MOTORS. the total number of coils twice the number of slots. Fig. 31 shows three half coils " nested," with two turns per coil ; this gives three times as many coils as there are slots, " three coils per slot." It is extremely objectionable to " nest " the coils, but sometimes unavoidable when round wires are used. Table II, p. 372, gives slot widths and depths suitable for various arrange- ments of round conductors drawn to B. & S. gauge, based on two-layer wind- ings and the insulation indicated in Fig. 32. The individual coils are wrapped Slot trough Fig. 30. Fig. 31. each in a single fold of 0.015-inch mica-treated press-board, each group of coils is wrapped with a single covering of 0.01-inch oiled tape, half lapped, and the slot is lined with a trough of 0.02-inch mica-treated press-board. If the press-board is well varnished with insulating compound, and the coils are dipped and baked before being assembled in the slots, this insula- tion will be adequate for 550-volt armatures. The Avidth of a coil slot should not be less than $ of its depth nor more than £ the depth. The depth of the coil slot, for armature of 16 inches diameter or over, may be estimated for preliminary purposes by means of the formula a=i+5 < 13 > Appropriate trial depths for the coil slots of smaller cores are given by Table III, page 373. Table IV, page 373, gives empirical but practical trial values for the mini- mum allowable number of armature coils, and Table V, page 374, gives values for the maximum allowable number of turns per coil, for use in preliminary 11 roughing-out." The former are somewhat elastic, but the latter can sel- dom be exceeded without risk of sparking at the brushes. Table VI, page 375, gives trial values for armature conductor sizes ; the actual allowable current density in the conductors, however, is determined by the heating of the armature. Armature bosses. —The total losses in the armature should not exceed the value which will give a temperature rise, under full load, of 70°F. The relation between lost watts, radiating surface, peripheral velocity and temperature rise is, for fairly well ventilated armatures in non-enclosed field magnet frames, approximately as follows : 35P' A DaWa[l + (• y(7),r.p.m.)» 420,000 and allowing a rise of 70° this transposes to 210 )' r.p.m.) 8 \ ^000 / (H) PRACTICAL DYNAMO DESIGN. 359 The reason for taking P' A instead of P A as the criterion of heating is that the projecting parts of the winding do not act effectively in radiating the heat produced by the core and teeth losses, although their radiating surface is always ample for the i^r loss in them. Since they are not included in the radiating surface, the loss in them is not included in considering the heating. With round conductors, the watts lost in the embedded part of the wind- ing will be, with sufficient accuracy, Pr' = WaNc ~\ a 2 if the conductors are rectangular in cross section, — — must be substituted for -^ in this equation. a 2 The losses in the armature teeth must be estimated separately from those in the body of the core, the densities being widely different in the two parts. The general formula for hysteresis loss in either part of the core is Ph — 48 kh vp r.p.m. 10 — 7 and the formula for eddy current loss is Pe = 4 k e vp 2 (r.p.m. )2 10 -■ in which kh is the loss per cubic foot of iron due to hysteresis, as given in the table on page 100 and ke the corresponding eddy current loss as given in the table on page 106. It should be borne in mind that although the con- stants taken from the tables mentioned are based on losses per cubic foot of iron or steel, the volume of iron or steel represented by v in the equations is in cubic inches. Combining the three equations just given, the total loss to be considered in estimating the heating of the armature is P A ' —WaNc l -£+p r.p.m. JO" » [48 (v a kha + vt kht) -f 0.4 p r.p.m. (Vakea + Vt ketj\ (15) In order to allow for the crowding of the magnetic flux toward the slots the cross section of the armature core body may be taken at 0.8 of the actual cross section, making the effective volume Va = 0.2 7T (2)«2 — Do 2 ) Wa (16) and the effective density will be, accordingly, # Ba= 0.8(Z)< — J) )w a ' • (17) For computing the probable losses in the teeth the following relations may be assumed without appreciable error : active teeth ) (2k ff S stive teeth ) _ /2_M £ \ # per pole \~ [nVp+p)"*' average width of each tooth =. (t -|- 2 t) -±- 3 ; and since (t -f 2 1) -^- 3 = [n (D a — 1.33 h) —Nt s] -=- Nt, and the average density in the teeth, for the present purpose, is equal to the flux per pole -J- active teeth per pole x average cross section per tooth, the average density will be Avg.B T = /0 ^ ,,„ * • • • (18) (v25 + |) ^( Da ~ 1 ' 3 ^ + Nt ^ Wa The volume of iron in the teeth is Vt— V-Da 2 — ™-D#—hsNA Wa (19) 360 DYNAMOS AND MOTORS. The value of kg in eq. (18) depends upon the relation between the inter- polar space (distance between neighboring pole-tips of opposite polarity) and the air-gap length, and also upon the slope of the pole from the tip toward Armature Center Fig. 33. the main part of the core. Table VII, page 376, gives practical values of k g within ordinary limits, and Fig. 33 indicates the angle represented in the table heading by a. Fig. 34 affords a simple method of estimating roughly the armature core losses which is favored by Messrs. Parshall and Hobart ; the curves here 450 t* ^ A ■D Jd M S ,^ 400 s > y y s y S y y 350 y y* / s s t y n/\n s 300 y y / / s 250 / > y * y y / y 200 / / 150 A <-^\i „A 1 • *A-~ ces 100 / "D. 01 ' 4.4. A ~ ~L A :esr // O L ^U itu * // V 7 50 // // // u 1 f ) .2 .4 .6 .8 L 1 .2 1 .4 l 6 l .8 2 Watts per cubic inch; core and teeth. Fig. 34- PRACTICAL DYNAMO DESIGN. 361 shown were plotted by Messrs. Esterlein and Reid from tests made on a large number of actual machines. In estimating before hand the efficiency of a machine, the loss in the pro- jecting parts of the armature winding must, of course, be considered. The actual total losses in the armature winding and core will be approximately P A = lc Nc ^+P r.p.m. 10~ 7 [48 (Vakha -J- vtkht) -hOApr.p.m. (vtket + vak*a)] (20) In a barrel winding, the length of each conductor (l e ) will be practically that given by the formula *e= W. + kw (Da — h) -f 0.8 (1 + h), if the conductors are bent around £-inch pins, as indicated in Fig. 35, and Fig. 35. afterward pulled out to span the proper number of teeth. Table VIII, page 376, gives values of kw for different numbers of poles. Each coil will project beyond the armature core at each end about -^ (Da — h) -\ — inches, and the distance from center to center of the winding pins must be equal to Wa + kw (Da — h) inches. Commutator and Brushes. — The number of commutator bars = number of armature coils or elements, in practically all modern windings. The diameter of the commutator barrel must be kept as small as possible in order to reduce the friction loss at the brush faces as well as to keep down the cost of the commutator and to favor good commutation. From purely mechanical considerations, Dk > 0.06 x Number of segments (21) For commutation reasons and to keep down friction, Dk < 10,000 -f r.p.m (22) In finally rounding out the dimensions, the following relation should be ob- served, if possible, t^ -W* b ___,. Dk— (23) 3 rik v ' and n* should preferably be an integer. The current density in each commutator segment should not much exceed 2000 amperes per square inch in the horizontal part and 2500 amperes per square inch in the connecting lugs or risers. The brush faces should be of such area and number that the current den- sity at the faces will not exceed 40 amperes per square inch for carbon brushes, 150 amperes per square inch for woven wire or gauze brushes, or 362 DYNAMOS AND MOTORS. 200 amperes per square inch for leaf copper brushes. Good average face densities are 30, 120, and 160 amperes per square inch, respectively. With pressures of 11 to 1\ lbs. per square inch of brush face, the effective resistance of the brushes will usually be Carbon brushes : Copper brushes : Ab 0.0125 — -a — = Rb - Ab The total drop in volts at the brush faces, therefore, will be Carbon brushes : Copper brushes : ^~ = volts drop • . . (24a) The loss in watts due to the friction of the brush contacts with the com- mutator is Ab Dk r.p.m. kb ' kb varying according to the brush pressure, condition of commutator and quality of brush. The total losses at the brush faces, therefore, are Carbon brushes : ^^ + 0^" m Copper brushes : AbDk r.p.m. /a 2 __ Pjk , OK . U- - + 8Q-Ab- Pb (25a) reasonably good condition, With ordinary grades of copper and carbon brushes and a commutator in d c 560 kb = : brush pressure in lbs. per sq. inch The maximum efficiency is obtained when the two terms of eqs. (25) and (25a) are equal, i. c, when the friction loss equals the J 2 R loss. The temperature rise of the commutator will usually be 85 X total lost watts /0 _. — uk (^b) ■»(* + ^ If the lugs of the commutator segments are of considerable length, the rise of temperature will be somewhat less than calculated; on the other hand, if the commutator and brushes are not in good condition, the losses will be considerably more than given by eq. (25) or (25a) and the tempera- ture rise will be correspondingly greater. The temperature rise should in no case exceed 75° Fahrenheit, and it is preferable to keep it down to 65° or 70°. The dimension of the brush face transverse to the commutator segments, is determined almost solely by commutation requirements, and these in- volve so many widely varying factors that no hard-and-fast general rule can be laid down. For machines of ordinary types and fairly large sizes — 100 kilowatts and over, say — the span of a carbon brush may be roughly estimated by means of the formula . Dk r ,. iaNc 1 /rt _. 6 " T L * (1 ~ *' ~ MSTBbJ (27) PRACTICAL DYNAMO DESIGN. 363 This formula will apply with sufficient closeness for all practical work by determining the value, for a given type of design, of the coefficient in the denominator of the bracketed fraction. For reversing motors of a certain type, for example, it is 1600, and for small, shunt-wound motors of conven- tional design, it ranges from 800 to 1000. Air-Crap. — The mechanical air-gap, from the pole-face to the tops of the armature teeth, should be made the nearest commercial dimension to that given by the formula General Service Machines. 18p All Other Machines. 20p (28) Thus, if the formula gives 0.188 inch as the proper air-gap length and the machine is to be built by English measure, the actual value to be used would be T ^ inch. In such cases a revision of the pole-face density should be made in order that the ampere-turns devoted to the air-gap shall conform tc Machine Axis Machin e Fig. 36. Fig. 37. the plan of design which is the basis of this section. See " Checking up Preliminary Dimensions below." role-Face. — The dimensions of the pole-face are determinable as previously described, the average chord being equal to kch Dp and the width parallel to the shaft being preferably equal to the chord. If solid pole-faces are used, the interpolar edges should not be strictly parallel to the armature slots. A common expedient for avoiding this par- allelism is to round the interpolar edges as in Fig. 36, or to make them slightly oblique with respect to the axis of the machine, as in Fig. 37. If laminated poles without shoes are used, the corners of alternate sheets of Fig. 38. Fig. 39. steel should be cut away as in Fig. 38 for straight poles, or the tips cut off, as in Fig. 39, for polar extensions. The length of the pole-face span should never exceed 2.5 Dp 4- p ; practi- cal values are given in the beginning of this section (page 356.) Checking* up Preliminary Dimensions. — Before passing on to the field-magnet proportions, and preferably before taking up the probable armature losses, the preliminary dimensions should be checked up in order to make sure that the desired E.M.F. is obtainable at the desired speed without entailing the use of excessive magnetic densities. 364 DYNAMOS AND MOTORS. Having ascertained by means of eq. (11) the maximum number of coil slots allowable and adjusted the net armature iron dimension axially by eq. (12) the E.M.F. or counter E.M.F. of the armature should be tested by the formula: _ frzy-"yr P *iy«r.p.m.io-» pq8 ' ' and if the E.M.F. is not what is desired, the armature diameter should be changed to correct it rather than change the value of either W p or ty or both. On the basis of the author's method, the E.M.F. is proportional to ZV 15 » if it be assumed that the number of wires will increase or diminish in proportion to small variations in the diameter ; therefore, if the preliminary dimen- sions do not give the proper E.M.F., the correct dimensions may be closely approximated by Tri al ZV-15 x E Trials = Correct B *> > the word " trial" referring to the diameter and E.M.F. first obtained. If the air-gap length actually adopted is not precisely the value given by eq. (28), the pole-face density should be adjusted to satisfy the equation, _ kdD P ^ Bp ~ p & < 30) The values of kv and kd are as follows : Type of Machine: General Service. Co^^g. **<&££* k v — 81 88 95 kd— 1562 1692 1831 The tendency to field distortion and sparking at the brushes should also be checked (after correcting the armature dimensions and pole-face density as just explained) before taking up the field magnet proportions. Armature Reaction and Commutation. — In order to guard against excessive field distortion the relation between the air-gap ampere- turns and armature ampere-turns should be as indicated by the following formula, for operation with fixed brushes at all loads : & p p8>kriaNc*l/ (31) The value of kr varies as follows : In general service machines, kr = 2.3. In shunt and compound reversing motors, kr 2j 3. In series-wound motors, kr = 2.7. The formula is based on the facts that Bp5 is approximately proportiona! to the ampere-turns required by the air-gap, and ia Nc \jt -f- p — armature ampere-turns tending to distort the field under each pole-face. The tendency to sparking at the brushes is proportional to the inductance of each coil, the number of coils simultaneously short-circuited by one brush, the number of coils in series between one positive and one negative brush and the current in the coil being commutated, and inversely propor- tional to the length of time the coil is short-circuited by the brush. The induotance of the coil is proportional to the length of the conductor and the square of the number of turns per coil. The following formula, based on these considerations, is an excellent criterion as to the sparklessness of a machine : (JVa + 0.llc)tHank^ ^r.p.m. 10~ 6 = Kk (32) The value of Kk varies as below : Kilowatts of machine : Up to 15 30 60 100 500 1000 or over. Kk— 80 70 60 50 40 35 field Magrnet. — Cores of circular cross section are most economical of wire in the field windings, and a square cross section is next best in this respect. The temperature rise is greater, however, in a round coil of given PRACTICAL DYNAMO DESIGN. 365 magnetizing power than in a square one, the cross section of the core and length of coil along the core being the same in both cases. Round coils are easier to wind, and are usually preferred. The length of a magnet core from the yoke to the pole-shoe or beginning of polar extensions, i.e., the space available for windings, parallel to the flux path in the core, may be roughly estimated for preliminary laying-out as follows : im= 9ow^iff) (33) The trial core length obtained by means of this formula will usually require revision in order to obtain the proper radiating surface for the coils. The magnetic density in field-magnet cores ranges from 90,000 to 100,000 lines per square inch for cast steel, and from 100,000 to 110,000 for sheet steel. The density in magnet yokes ranges from 35,000 to 45,000 maxwells per square inch in cast iron, and 85,000 to 95,000 for cast steel. In railway motors and others of extraordinarily light weight, the yoke density is con- siderably higher than in stationary machines ; the core density is also somewhat higher, but the diiference is not so great as in the yoke. The density is not uniform throughout the length of path in the core, nor is it so in the yoke, but for convenience the maximum density is assumed to exist throughout the length of each path. Leakage of magnetic lines between adjacent poles and between each pole and the yoke surfaces makes the flux in the field magnet considerably Fig. 40. greater than that in the air-gap. The relation between the magnet -core flux and the air-gap flux is The value of v varies widely with different types of machines and different sizes of a given type. For well-designed machines of conventional tvpes it may be assumed tentatively to have the values given in Table X. It is con- siderably higher for poor designs. In the absence of data from existing machines of the type being designed, the field magnet may be proportioned on the basis of the values in Table X, page 376, tentatively, and the leakage roughly checked up as follows: Lay out to a rather large scale two poles of the machine and the corre- sponding portion of the yoke, as shown in Fig. 40 for a circular yoke. The average length of the leakage path between the upper surface of the polar extension and the inner surface of the yoke will be about as indicated by the dotted line Z, and the length of the leakage path between the neighbor- ing polar extensions will be about as shown by the line Z x . The mean length of the leakage path between the flanks of neighboring pole-ends is practically equal to the distance between the centers of the two measured along a circular arc concentric with the armature ; represent it by Z 2 . The mean length of the leakage path between each pole-piece flank and the yoke surface lying between x and z may be called equal to Z. The maximum flux in the magnet core will be approximately as given by the equation, ♦.= . + 3.2* ( *A+A+A 3 + A i + ±A_ l + S _A,j . . m Field-Jftag-net Excitation. — In order to estimate beforehand the excitation required by the machine, the quality of the iron and steel to be 366 DYNAMOS AND MOTORS. used in its construction should be known. In the absence of such data however, the curves in Fig. 41 will serve for estimates. 140 130 120 110 100 90 £ 80 Ampere-turns per inch of length 500 1000 1500 o o 70 .9 6° «50 50 100 150 Ampere^turns per inch of length Fig. 41. The flux in the air-gap of a dynamo at no load is 60 q Ew 10 8 The flux at full load is *n = * = pNcT.p.m. 2000 rr s : • ' / / / / ! • J I I / / ' 1 - L* « fjgz X y } / / / / / J / . / / 1 1 i GOg^^ + ^/q)^ p No r.p.m. 200 (35) (36) PRACTICAL DYNAMO DESIGN. 367 For a motor the flux is the same at full load as at no load, except in special cases where a series winding is used in order to start a heavy load, and ex- cepting series- wound motors. The maximum air-gap flux for a motor haying to start under a load is no 11 ** = 117 pia#. ' < 3?) The full-load ampere-turns per pole for a dynamo or motor are F-\- Fr. F ' = Fa + Ft + F g + F P + Fm -f F y ; Fa =f a £a -J- 2 ; Ft = /t h ; F P =f p L P ; Fm '^zjm Lrn ) -by ZZZJy ±jy — 2. The ampere-turns per inch for the armature teeth will be the mean be- tween the ampere-turns per inch required to produce the density at the tops and those required to produce the density at the roots — not the ampere- turns required to produce the average density in the teeth. The approxi- mate density at th© roots of the armature teeth will be, at full load, and the approximate density at the tops of the teeth will be b t , = — — ^ o.mm "\ (39) Wa (nDa — S Nt) ( - + sr^ ) \P T Dp ] As some of the flux passes to the armature core body through the slots and ventilating spaces, the actual densities in the roots and tops of the teeth are less than the approximate densities given by the above formulas. The actual densities cannot be computed directly, but may be derived from the relation between the actual and approximate densities, which is as follows: B/=B T + 3.192/ T [^(l+f)-.l] (40) Since the formula cannot be transposed to solve for B T because B T and/ T are interdependent and vary at different rates, a table should be prepared showing values of B/ corresponding to different values of f T at different ratios of s -f- t and Wa -~ wa. The preparation of such a table is greatly facilitated by first preparing a table of values for representing this expression by k T , and thereby reducing eq. (40) to B/=B T -r-* T /r ( 41 > Table XI, page 377, gives values for k T for practical ranges of values for the two ratios mentioned. From eq. (41) and curves such as those in Fig. 41, a table of corresponding values for B T ' and/ T is easily prepared. From such a table the value of / T should be ascertained for the root and top of the tooth and also for two or three equidistant intermediate points between the root and top ; the average of these will be the working value. The ampere-turns per pole required by the air-gap will be f.'~ °' 3133 * (42) <^+*<^>0§f + *') 368 DYNAMOS AND MOTORS. Table IX, page 376, gives values of k g2 and Fig. 42 gives those of k h within ordi- nary ranges. The constant k g2 is merely the number which, multiplied by the air-gap length, gives the extent to which the air-gap dimension parallel to the shaft is increased by the bowing outward of the magnetic flux in pass- Slot Air-gap Pig. 42. ing from the pole-face edges to the armature core teeth. The constant k$ is the proportion of the physical air-gap length, 6, by which the gap is increased effectively by the passage of flux into the sides of the armature core teeth. This has been taken from Mr. F. W. Carter's article in the Electrical World and Engineer for Nov. 30, 1901. The value of F r cannot be predetermined with any approach to accuracy unless one has data from existing machines of corresponding type and out- put. The following empirical formula will serve to estimate roughly the value of F + Fr for modern American dynamos and non-reversing motors : PRACTICAL DYNAMO DESIGN. 369 j. + jR . = (0-g-0-»»>fcM + J F2+ ^UN. y , _ (43) For reversing motors, 0£4rUNc\2 (43a ) F+Fr= Jf* + / 0-6«M*JVc \ The no-load excitation of a shunt-wound dynamo need not be predeter- mined. The no-load excitation of a compound-wound dynamo is Fg ~ Fdo -f- Fto -f- Fgo -\- Fpo -j- Fmo -\- Fyo* The ampere-turns of the several parts of the magnetic circuit are deter- mined as in the case at full load, taking into account the differences in magnetic density in each part. After the first machine of a given type has been constructed, with the exception of the field-magnet coils, it should be tested with temporary exciting coils ; the results of these tests should be taken as the foundation of the magnet coil calculations. JF ield-UIagmet Winding's. — The field winding of a series or shunt- wound dynamo must be capable of giving the excitation required at full load. The field winding of a shunt-wound motor must give the excitation re- quired at the proper full-load speed. The field winding of a series-wound motor must give the excitation re- quired to produce the starting flux, 4> T . The shunt winding of a compound-wound dynamo must give the excita- tion required at no load ; the series winding must give the difference be- tween this and the excitation required at full load. The shunt winding of a compound-wound motor must give the excitation required at normal no-load speed ; the series winding must give the differ- ence between this and the excitation required to produce the starting flux, *T. The surface of any field magnet coil on a dynamo or motor of open con- struction (non-enclosed frame giving the external air free access to the windings), should be L fG— —jj- (44) r being the resistance of the coil when warm. For enclosed or poorly ven- tilated frames, the coil surface per watt per degree of temperature rise must be determined by trial ; no general rule will apply. In all cases 9/ should not exceed 70°. The proper size of wire to be used in a shunt field coil is approximately given by d 2 - ^fr + TA) (45) e Should the calculated value of d 2 not correspond with any standard size, the nearest standard size should be adopted and the depth of the winding ad- justed to suit it by transposing eq. (45) and solving for A, thus : d 2 e Fsh -9 A = — (46) 7T See also Magnet Windings, page 112. The minimum number of turns per pole for the series coils of a com- pound-wound machine is CF-\- Fr — Fah Turns : \ I F -\- Fr — Fah (short shunt) (long shunt) (47) 370 DYNAMOS AND MOTORS. The cross section of the series conductor need not exceed 0015 sauare inches per ampere actually carried by the coil, and should not be less tian ?hTheft q ing 1 : e mCh Pei ' amP<3re ordi * aril y J " will be finally determined by ti™K b ° th f the Seiie . S and f hu ?i 3 eM ma g net coils, the maximum possible number of ampere-turns should be made from 10% to 15% greater tEanth« calculated maximum in order to provide a margin fol -differenced Tin the ?r U epanciesf C ° PPer and ^ USed ' aS Wel1 M othw ^?onteol2b£ at* 1H.P. Fig. 43, Efficiency.-— Efficiencies. range from 60% to 95#i SLOonrdi^tr +r» fv,<* «,v« ~o the machine and the character of service. Table XII plgefr'^ves aver l&V„V UeS / 0r or ? i,lar y constant-potential dynamos, and *?ig. 43 gives simil wPSStiSSS^KXttSS!"- Traction and "ASK^SSi Procedure in raying- out a Desig-n.— The following will be Receding pages :° US p, ' ocedure in H-Hjl^ft. method described in the Determine 1. A trial polar bore, eq. 8 or 9 or 10. Type of armature winding ; number of paths. Number of poles ; eq. 6, for lap-wound machines. Ratio of pole-face span : pole pitch (i//). Maximum pole-face width ( W P < lech Dp). . Air-gap, eq. 28 ; the armature diameter follows. 7. Turns per armature coil, Table V. 8 Trial size of conductor, Table VI. 2. 3. 4. 5. PRACTICAL DYNAMO DESIGN. 371 9. Size of coil slot, based on number of conductors per slot, either Table III or eq. 13, and rules s < 25 and s = — to — . 10. Possible number of coil slots, eq. 11 ; hence, total number of arma- ture conductors, keeping in view type of winding, eq. 2. 11 Corrected pole-face density, eq. 30. 12! Field-distorting armature reaction, eq. 31 ; if kr comes out too small ' the polar bore must be increased, thereby increasing the pole-face density and air-gap ; then solve eq. 31 for Nc, taking the nearest smaller value that will fit the winding. . l. ^ 13. Corrected pole-face width, by solving eq. 29 for W P ; if the result ^ kch D P , accept it ; if not, take a still larger polar bore, with the corre- sponding air-gap, and start over from Determination No. 11. 14. Net axial iron measurement in armature, eq. 12. 15. Gross length of armature core (= W P + 2bto W P + 4 5) ; the differ- ence between this and the net iron to be occupied by ventilating ducts. 16. Number of armature coils ; check by Table I\ roughly ; a discrep- ancy of 25% is not prohibitive. , _ . _ , 17. Diameter of commutator barrel, eqs. 21 and 22 ; Dk should never ex- ceed 0.9 Da, and 0.7 Dais an excellent limit ; if the diameter comes out too polar revising the air-gap by eq. 28. 18. Complete commutator and brush dimensions, eqs. 25, 26, and 27. 19. Probable tendency to sparking, eq. 32 ; if Kh is excessive, and the turns per coil cannot be reduced without entailing an unwieldy number of coils, the polar bore must be increased in order to permit reducing the length of the armature core, the determinations being revised from No. 11 after finding the new air-gap, eq. 28. 20. Armature losses with respect to heating, eq. 15 et seq. ; if PjJ ex- ceeds the limit set by eq. 14, and cannot be brought within the limit by re- ducing the hole in' the center of the core, the ventilating ducts may be reduced sufficiently to accomplish the result ; if not, and if Wa cannot be sufficiently increased on account of eq. 32, the polar bore must be increased, the corresponding air-gap adopted, and the determinations revised, begin- ning with No. 11. Having progressed this far, the remainder of the design is straight work, only a slight revision of the trial magnet core length being probably neces- sary to obtain the minimum quantity of field copper within the heating limit. TABLE I. Values of Uch. Poles. yfj = 0.666. v// = 0.7. xjj = 0.72. 1// == 0.75. 2 4 6 0.866 0.5 0.342 0.891 0.5225 0.3584 0.9048 0.5358 0.3681 0.9239 0.5556 0.3827 8 10 12 0.2588 0.2079 0.1736 0.2714 0.2181 0.1822 0.279 0.2244 0.1874 0.2903 0.2334 0.1951 14 16 18 0.149 0.1305 0.1161 0.1564 0.137 0.1219 0.1609 0.1409 0.1253 0.1676 0.1467 0.1305 20 22 24 0.1045 0.0949 0.0872 0.1097 0.0998 0.0915 0.1129 0.1026 0.0941 0.1175 0.1069 0.0979 372 DYNAMOS AND MOTORS. it* © g 2 © «s is •^ 2 - cj eS ^ ©,d +3 1«§ .2^ © « o C fi © *a" S§1 s ® 53 flt~ tJOccO . •*? (ID *•£ • © © u. >• rt 5* ® -fl CO* 3 O.S « 8 Scg © §* * a^ o °£© .2 - © so +j a & ©3 ^ H.S'o l-« ° 1 «8*3 it © c« •8ZIS QJl^i ■<*-aoo©»H?©t^ooa>© £ ^ 5 ■<*< co cn *-i © © 00 00 'S^SS CO"* t-i-H CO tH O Ci CO 00 l> mcoc^H005 0)(»t-i>?i lOt- lO »-* 00 t~ CO in CO t> 50HNM05 05t>OTt<0«DiOiO'* lOcocsrHocioot-t^ocoioiiOiaTti'^i'^ •QZTg 911^ CO o X 5 ,0 OOCOCOt>'CDlOnOiO"*'5'^COCOCOC r*io«ob-oooiOi-HCNJco^io©t^ooo><; 5. CO CN1 <34 t> -* CO rH OS -* rJCOOt^-^CNi-HC75t^^)lLArfO»HtNJ0T*<»0«0l>00C7>Cl PRACTICAL DYNAMO DESIGN. 373 TABLE III. Trial Armature Coil Slot Depth 0. Core Diameter. Slot Depth. Core diameter. Slot Depth. 6 7 i 10* 11 11* If 8 8* I B 12 12* 13 it 9 10 1 It 1 * 13* 14 15 1A ii 5 * if TABLE IV. Trial Values for Minimum lumber of Armature Coils. Formula : N, = 0.8 p°' 8 X \A#X ^/KW.* The numbers in the table are values of \JE x \J KW.* KW* 125 volts. 250 volts. 600 volts. 1 11.2 15.8 24.5 2 14.1 19.9 30.9 3 16.1 22.8 35.3 4 17.8 25.1 38.9 5 19.1 26.9 41.9 6 20.3 28.7 44.5 8 22.4 31.6 49. 10 24.1 34.1 52.8 15 27.6 39. 60.4 20 30.4 42.9 66.5 25 32.7 46.2 71.6 30 34.7 49.1 76.1 40 38.2 54.1 83.8 50 41.2 58.2 90.2 60 43.7 61.9 95.9 75 47.1 66.7 103.3 100 51.9 73.4 113.7 125 55.9 79. 122.5 150 59.4 84. 130. 200 65.4 92.5 143. 250 70.4 99.6 154. 300 74.8 105.8 164. 400 82.4 116.5 180. 500 88.7 125. 194. 600 94.3 133. 207. 700 99.3 140. 218. 800 103.8 147. 227. 1000 112. 158. 245. *KW. = Kilowatts output of dynamo or intake of motor. For;? = 2 4 6 8 10 12 14 0.8p°- 8 = 1.4 2.4 3.35 4.2 5 6.8 6.6 1.6 73 374 DYNAMOS AND MOTORS. TABLE V. Trial Values for Maximum Allowable Number of Turns per Armature Coil. Formula : t 2 :5 240 q ~ iap. Lap Winding. Two-path Windings. Turns per Coil. p — q. p =z 4. p = 6. p = 8. t ia Id ia id 240 120 80 60 1 60 30 20 15 2 26 13 9 6.6 3 15 7.5 5 3.75 4 9.6 4.8 3.2 2.4 5 6.6 3.3 2.2 1.66 6 4.9 2.4 1.6 1.22 7 3.75 1.87 1.25 0.93 8 3 1.5 1 0.75 9 2.4 1.2 0.8 0.6 10 1.8 0.9 0.6 0.45 11 1.66 0.83 0.55 0.42 12 1.42 0.71 0.47 0.35 13 1.22 0.61 0.41 0.3 14 1.06 0.53 0.35 0.26 15 PRACTICAL DYNAMO DESIGN. 375 TABLE VI. Trial Value* for Carrying- Capacity of Armature Conductors. 2 or 4 Wires in Parallel Considered a Single Conductor. Round Wires, Drawn to B. & S. Gauge. Rectangular Conductors. 2 in par- allel. 4 in par- allel. Da X r.p.m. = Da X r.p.m. = Single. 4000 to 8000 to 10,000 to 15,000 to 20,000 to 6000. 10,000. 12,000. 17,000. 22,000. No. No. No. Amperes. A A 20 2 2* ,d 19 . . 2i 3i q cd .9 18 • • • • 3 4 CD U CD 17 16 20 19 • • 4 5 5 6 B c3 c« S3 15 18 • • 6 n CO u CD 00 U CD ■ CD 14 17 20 7i 9i ft Pi ft 13 16 19 9 111 CD CD CD 12 15 18 11 14 cd ft u CD ft CD ft 11 14 17 13* 17* ! i a oB 10 13 16 17 21* 9 12 15 21 26* I > £> >> 6 9 12 40 50 '35 1 8 11 52 66 CD CD T3 CD 7 10 65 80 43 43 6 9 80 100 a CD U U CD CD M N 8 104 132 3 S3 7 130 160 O o O 6 160 200 376 DYNAMOS AND MOTORS. TABLE VII. From " The Dynamo," by Hawkins & Wallis. Values of U 9 . a *•■* | =1 0. I= 12 - *-» I *>*» 0° 10° 20° 30° 40° 50° 1.95 1.85 1.75 1.66 1.58 1.52 2.18 2.05 1.95 1.84 1.75 1.666 2.38 2.23 2.10 1.98 1.S9 1.80 2.55 2.38 2.25 2.12 2.00 1.90 2.7 252 2.38 2.24 2.12 2.00 TAJBI.E VIII. Barrel Armature Winding* Constants. Poles = 4 6 8 10 12 14 16 18 20 24 hw = 0.8 0.56 0.42 0.36 0.3 0.256 0.225 0.2 0.18 0.15 TABLE IX. From " The Dynamo," by Hawkins & Wallis. Values of U gt . JVa—W P _ 1 1.5 2 2.5 3 3.5 0.74 1.0 1.2 1.38 1.54 1.68 4 1.8 T4BIE X. Averag-e Mag-netic L^aka-je Coefficients. Kilowatts = 10 25 40 50 75 100 200 300 500 1000 v — 1.35 1.3 1.27 1.25 1.23 1.2 1.18 1.15 1.13 1.12 PRACTICAL DYNAMO DESIGN. 577 TABLE VI. Values of k r . s »s=ua 1.17 1.18 1.19 1.20 1.22 1.24 T Wd 0.70 3.10 3.16 3.21 3.26 3.32 3.43 3.54 0.75 3.29 3.34 3.40 3.45 3.51 3.62 3.73 0.80 3.47 3.53 3.59 3.64 3.70 3.82 3.93 0.85 3.66 3.72 3.78 3.84 3.89 4.01 4.13 0.90 3.84 3.90 3.96 4.02 4.09 4.21 4.33 0.95 4.03 4.09 4.15 4.21 4.28 4.40 4.53 1.00 4.21 4.28 4.34 4.40 4.47 4.60 4.72 1.05 4.40 4.46 4.53 4.59 4.66 4.79 4.92 1.10 4.58 4.65 4.72 4.78 4.85 4.98 5.12 1.15 4.77 4.84 4.91 4.97 5.04 5.18 5.32 1.20 4.95 5.02 5.09 5.16 5.23 5.37 5.52 1.25 5.14 5.21 5.28 5.35 5.43 5.57 5.71 1.30 5.32 5.40 5.47 5.54 5.62 5.76 5.91 1.35 5.51 5.58 5.66 5.73 5.81 5.96 6.11 1.40 5.69 5.77 5.85 5.92 6.00 6.15 6.31 1.45 5.88 5.96 6.04 6.11 6.19 6.35 6.51 1.50 , 6.06 6.14 6.22 6.30 6.38 6.54 6.70 1.55 6.25 6.33 6.41 6.49 6.57 6.74 6.90 1.60 6.43 6.52 6.60 6.68 6.77 6.93 7.10 1.65 6.62 6.70 6.77 6.87 6.96 7.13 7.30 1.70 6.80 6.89 6.98 7.06 7.15 7.32 7.49 1.75 6.99 7.08 7.17 7.25 7.34 7.52 7.69 1.80 7.18 7.26 7.35 7.44 7.53 7.71 7.89 1.85 7.36 7.45 7.54 7.63 7.72 7.91 8.09 1.90 7.55 7.64 7.73 7.82 7.92 8.10 8.29 2.00 7.92 8.01 8.11 8.20 8.30 8.49 8.68 tibli: XII. Average Dynamo Efficiencies. Appropriate Distribution of Losses BO in Per Cent. . « «a 'd . PJ o 68 © P © © Armature Losses. .2 w> ri %* O fe 3 fa a o ©T3 ■4-» © « 2 M fao Copper. Iron. fa h} 30 90 4.0 3.0 2.5 0.5 10 40 90.5 3.8 2.8 2.4 0.5 9.5 50 91 3.6 2.7 2.3 0.4 9 75 91.5 3.4 2.5 2.2 0.4 8.5 100 92 3.2 2.4 2.0 0.4 8 200 93 2.7 2.15 1.8 0.35 7 300 93.5 2.5 2.0 1.65 0.35 6.5 500 94 2.3 1.8 1.55 0.35 6 750 94.5 2.0 1.7 1.5 0.3 5.5 1000 95 1.8 1.5 1.4 0.3 I 5 378 TESTS OF DYNAMOS AND MOTORS. TESTS OF DYNAMOS AND MOTORS. Revised by Cecil P. Poole and E. B= Raymond All reliable manufacturers of electrical machinery and apparatus are now provided with the necessary facilities for testing the efficiency and other properties of their output, and where the purchaser desires to confirm the tests and guaranties of the maker, he should endeavor to have nearly, and in some cases all such tests carried out in his presence at the factory, unless he may be equipped with sufficient facilities to enable him to carry out like tests in his own shops after the apparatus is in place. Some tests, such as full load and overload, temperature, and insulation (except dielectric) tests are best made after the machinery has been installed and is in full running order. Owing to the ease and accuracy with wmich electrical measurements can be made, it is always more convenient to make use of electrical driving power for dynamos, and electrical load for the dynamo output, and in the case of motors, a direct-current dynamo with electrical load makes the best load for belting the motor to. No really accurate tests of dynamo efficiencies can be made with water- wheels, and only slightly better are those made by steam-engines, owing to unreliability of friction cards for the engine itself and the change of fric- tion with load. Where it is necessary to use a steam-engine for dynamo testing, all fric- tion and low load cards should be taken with the steam throttled so low as to cut off at more than half stroke, and to run the engine at the same speed as when under load. The tests of the engine as separated from the dynamo are as follows : — a. Friction of engine alone. b. Friction of engine and any belts and countershaft between it and the dynamo under test. Consult works on indicators and steam-engines for instructions for deter- mining power of engines under various conditions. The important practical tests for acceptance by the purchaser, or to deter- mine the full value of all the properties of dynamos and motors, are to learn the value of the following items : — Rise of temperature under full load. Insulation resistance. Dielectric strength of insulation. Regulation. Overload capacity. Efficiency, core loss. Bearing friction, windage and brush friction. I 2 R loss in field and field rheostat, I 2 B loss in armature and brushes. Note. — If a separate exciter goes with the dynamo, its losses will be determined separately as for a dynamo. Methods of determining each of the above-named items will be described, and then the combinations of them necessary for any test will be outlined. Temperature.— -The rise of temperature in a dynamo, motor, or transformer, is one of the most important factors in determining the life of such piece of apparatus; and tests for its determination should be carried out according to the highest standards that can be specified, and yet be within reasonable range of economy. The A. I. E. E. standards state the allowable rise of temperature above surrounding air for most conditions, but special conditions must be met by special standards. For instance, no ordinary insulation ought to be subjected to a degree of heat exceeding 212° F., or 100° C. And yet in the dynamo-room of our naval vessels the temperature is said to at times reach 130° F., or even higher, which leaves a small margin for safety. It is obvious that specifications for dynamos in such locations should call for a much lower temperature rise in order to be safe under full load. For all practical temperature tests it is sufficient to run a machine under its normal full-load conditions until it has developed its highest temperature, although at times a curve of rise of temperature may be desired at various loads. TEMPERATURE. 379 Most small dynamos, motors, and transformers, up to, say, 50 K.W., will reach maximum temperature in five hours run under full load, if the tem- Eerature rise is normal ; but larger machines sometimes require from 6 to 18 ours, although this depends quite as much on the design and construction of the apparatus as on size, as, for instance, the 5,000 h.p. Niagara Falls Gen- erators reach full temperature in fiv r e hours. Temperature tests can be shortened by overloading the apparatus for a time, thus reaching full heat in a shorter period. On dynamos and motors the temperatures of all iron or frame parts, com- mutators, and pole-pieces, have to be taken by thermometer laid on the surface and covered by waste. Note that when temperatures are taken with the machine running, care must be taken not to use enough waste to influence the machine's radiation. Where there are spaces, as air spaces, in armature cores or in the field laminations, that will permit the insertion of a thermometer, it should be placed there. Temperature of field coils should be taken by thermometer laid on the surface and covered with waste, and by taking the resistance of the coils first at the room temperature and again while hot immediately after the heat run. Temperature rise of arma- ture windings can be taken by surface measurement and by the resistance method also ; although being nearly always of low resistance, very careful tests by fine galvanometer and very steady current are required in order to get anything like accurate results. The formula for determining the rise of temperature from the rise of resistance is as follows : Temperature by rise in resistance; for copper. — The in- crease in resistance due to increase in temperature is approximately 0.4% for each degree Cent, above zero, the resistance at zero being taken as the base. If then t x = temperature of copper when cold resistance is measured (Cent.), R x = resistance at temperature t lt t 2 — temperature of copper when hot resistance is taken, i?2 = resistance at temperature t 2 , Then first reducing to zero degrees, we have p — A n \ M ° - 1 + 0.0042 t t ' ' ' ■ (1) The increase in resistance from to t t degrees is i? 2 — i? , and hence we have for final temperature, t 2 = B2 ~ B ° + 0.0042 (2) Substituting (1) _ i? 2 (1 -f 0.0042 t x ) — i? t 2_ 0.0042^ " (3) It is often convenient to correct all cold resistances to a temperature of 20° C, in which case we first reduce to zero and then raise to 20°. The general formula for obtaining the resistance at t degrees is Rt = (1 + 0.0042 t) i? . Hence i? 20 = 1.084 K and in terms of the cold resistance at temperature t. _ (1.084 i? ) ^2 - ( i _j_ 0.0042 t) (4) Formula (3) then becomes, when the cold resistance is at 20°, ^ = ^f 2 X ^-6i42 = 258 5- 238 < 5 > As the first formula requires but oue setting of the slide rule, and the sub- traction of the constant 238 can usually be done mentally, the advantage of the temperature equation in this form is very great as regards both speed aud accuracy. The temperature coefficients most generally used are For copper 0042 For iron 0045 For German silver 00028 to .00044 380 TESTS OF DYNAMOS AND MOTORS. The following parts should be tested by the resistance method and the surface method also : Field coils series, and shunt. Armature coils. In 3-phase machines, take resistance between all three rinss. The following parts should be tested by thermometer on the surface : — Room, on side opposite from steam-engine, if direct connected, and always in two or more parts of the room, within six feet of machine. Bearings, each bearing, thermometer held against inner shell, unless oil from the well is found to be of same temperature as the bearing. Commutators and collector rings. Brush-holders and brushes, if thought hotter than the commutator. Pole-tips, leading and following. Armature teeth, windings, and spider. Terminal blocks, for leads to switch-board, and those for leads from the brushes. Series shunt, if in a compound-wound machine. Shunt field rheostat. On transformers which are enclosed in a tank filled with oil/temperatures by thermometer should be taken on — Outside case, in several places. Oil, on top, and deeper by letting down thermometer. Windings, by placing thermometer against same, even if under oil. Laminations, by placing thermometer against same, even if under oil. Terminals. Room, as with dynamos and motors. Also resistance measurements of primary and secondary windings, from which the temperature by resistance can be calculated as shown. On transformers cooled by air forced through spaces between windings %nd spaces in laminations, temperatures by thermometer should be taken on — Outside frame. Air, outgoing from coils. Air, outgoing from iron laminations. Windings. Terminals. Room, in two or more places. Also resistance measurements, hot and cold, should be taken, from which rise of temperature, by resistance can be calculated. Finally, the cubic feet of air, and pressure to force same through spaces (easily measured by " U " tube of water), should be measured. When other fluids are used for cooling, such as water passing through piping submerged in oil, in which also the windings and core are submerged, or through windings of transformers themselves (made hollow for the pur- pose), the temperature of incoming and outgoing fluid should be measured, the quantity used and the pressure necessary to force it through the path arranged, besides the other points mentioned above. Careful watch of thermometers is necessary in all cases, as they will rise for a time and then begin to fall ; and the maximum point is what is wanted. British authorities state a definite time to read the thermometers after stopping the machine. Care must also be taken to stop the machine rotating as soon as possible, so that it will not fan itself cool. A handy method of constructing a curve showing the rise of temperature in the stationary parts of a machine at full load is to insert a small coil of fine iron wire in some crevice in the machine in the part of which the tem- perature is desired. Connect the coil with a mirror galvanometer and battery. The temperature coefficient of iron is high, and the gradual increase in resistance of the coil will cause the readings on the galvanometer to grow gradually less ; and readings taken at regular intervals of time can be plotted on cross-section paper to form a curve showing the changes in temperature. TEMPERATURE. 381 Records of temperature test. —During all heat runs readings should be taken every fifteen (15) minutes of the following items: On direct and alternating current motors and generators — Armature, Volts (between the various rings where machine is more than single-phase, in the case of alternators, and between brushes, in the case of a D. C. machine). Amperes (in each line). Speed. Field, Volts. Amperes. On synchronous converters : — Armature, Volts (between all rings on A. C. end, and between brushes on D. C. end). Amperes, per line A. C. end, also D. C. end. Speed. Field, Volts. Amperes. On transformers, compensators, potential regulators : — Volts, primary. Volts, secondary. Amperes, primary. Amperes, secondary. Cycles. Amount and pressure of cooling-fluid (if any is used). On induction motors : — Volts, between lines. Amperes, in line. Speed. Cycles. Overload. —The A. I. E. E. standards contain suggestions for overload capacity (see page 303). The writer has uniformly specified a standard overload of 25% for 3 hours, and there seems to be no especial difficulty in getting machines for this Standard that do not heat dangerously under such conditions. Insulation test. — Insulation resistance in ohms is of much less im- portance than resistance against breakdown of the insulation under a strain test, with alternating current of high pressure. Make all insulation tests with a voltage as high, at least, as that at which the machine is to be worked. The following diagram shows the connections to be made with E some external source of E.M.F. The formula used is R =z resistance of voltmeter. E =: E.M.F. of the external source. mmror, e = reading of voltmeter connected as in wg. diagram. t/C\ i W n x=z insulation resistance in ohms. X--^ «BB— ■ x— insulation resistance in ohms. armature Then x : »(!-')• a^ y , — 4- FRAME According to the A. I. E. E. standards, the insulation resistance must be such that Fig. 1. Connections for volt- the rated voltage of the machine will not meter test of insulation re- send more than y^^s of the full-load cur- sistance of a dynamo. rent through the insulation. One megohm is usually considered sufficient, if found by such a test. Where one megohm is specified as sufficient, the maximum deflection that will produce that value, and that must not be exceeded in the test, may be found by the fol- lowing variation of the above formula : RXE ~ 1,000,000 -f R Strain test. — The dielectric strength of insulation should be deter- mined by a continued application of an alternating E.M.F. for at least one (1) minute. The transformer from which the alternating E.M.F. ht takax* should have a current capacity at least four (4) times the amount oi current 382 TESTS OF DYNAMOS AND MOTORS. necessary to charge the apparatus under test as a condenser. Strain tests should only be made with the apparatus fully assembled. Connect on a D.C. machine as in the following diagram. Strain tests should be made with a sine wave of E.M.F., or with an E.M.F. having /■Khini m w VM - tne same striking distance between needle rooaojjooo .so points in air. I >T> I Jpy See article 219 A. I. E. E. standards for X-^Tfuja J jL proper voltages. Regulation. — The test for regula- tion in a dynamo consists in determining its change in voltage under different FRAME loads, or output of current, the speed be- Fig. 2. Connections for strain ing maintained constant, test of dynamo or motor or The test for regulation in a motor transformer insulation. consists in determining its change of speedy under different applied loads, when the voltage is kept constant. Standards. — For full details of standards of regulation of different machines, see report of the Committee on Standardization of the A. I. E. E. at the beginning of this chapter. Regulation Tests, Dynamos, Shunt or Compound, and Alternators. The dynamo must be run for a sufficient length of time at a heavy load to raise its temperature to its highest limit ; the field rheostat is then adjusted, starting with voltage a little low, and bringing up to proper value to obtain the standard voltage at the machine terminals, and since a constant temper- ature condition has been reached, must not again be adjusted during the test. Adjust the brushes, in the case of a D. C. machine, for full-load con- ditions, and they should not receive other adjustment during the test. This is a severe condition, and not all machines will stand it ; but all good dy- namos, with carbon brushes, will stand the test very well, provided tne brushes are adjusted at just the non-sparking point at no load. Load is now decreased by regular steps, and when the current has settled the following readings are taken : — Speed of dynamo (adjusted at proper amount). Current in output (a non-inductive load should be used). If alternator, current in each line if more than single-phase. Volts at machine terminals. Amperes, field. Volts, field. Note sparking at the brushes (they should not spark any with carbon brushes) . Readings should be taken for at least ten intervals, from full load to open circuit (no load) ; and load should then be put on gradually and by the same steps as it was brought down ; and the same records should be made back to full-load point, and beyond to 25% overload. If the readings are to be plotted in curves, as they always should be, it will make little difference if the intervals or steps are not all alike ; and should the steps be overreached in adjusting the load, the load must not, in any circumstances, be backed up or readjusted back to get regular inter- vals or a stated value, as the conditions of magnetization change, and throw the test all out. In case the current is broken, or the test has to be slowed down in speed or stopped, it must be commenced all over again. Finally, when the curves are plotted, draw, in the case of a compound-wound ma- chine, a straight line joining the no-load voltage and the full-load voltage ; and the ratio of the point of maximum departure of the voltage from this line to the voltage indicated by the line at the point will be the regulation of the machine. The readings as obtained give what is called a field compounding curve. In the case of a shunt or separately excited machine, the procedure for the test is the same ; but when the curve is plotted, the regulation is figured as equal to the difference between the no-load voltage and full-load voltage, divided by the full-load voltage. The curve is called a characteristic in this case. DYNAMO EFFICIENCY. 383 For alternators that are too large to apply actual load as suggested above, another "no-load" method commonly used with satisfactory results upon alternators designed upon the usual lines is to short-circuit the alternator ar- mature upon itself and determine the amperes in the field required to produce normal current in the armature so short-circuited, the speed of the machine being normal at the time ; call this current F. Take another reading of the field current required to produce normal voltage at the machine ter- minals, with the armature on open circuit and the speed normal ; call this current C. Then the current required in the field winding for full non- inductive load will be /= ViT2_j_ c*. Having calculated the value of this current, pass it through the field windings of the alternator with the armature on open circuit and running at normal speed, and read the volts V. Let E =. normal voltage, then the regulation = — = — >• The current F is called the " Synchronous impedance " field current, being so named by Mr. C. P. Steinmetz, who proposed and has used the above- described method. When regulation is desired for a power factor other than unity the field currents F and C must be combined at the proper angle corresponding to the power factor. For instance, for a power factor of (i.e., 90° lag) the field currents would be directly added. This method is used extensively and gives results agreeing very well with those of actual tests. Reg-illation Tests, Motors, Shunt. Compound, and Induction. After driving the motor under heavy load for a length of time sufiicient to develop its full heat, full-rated load should be applied, the field rheostat, if any is used, and brushes adjusted for the standard conditions ; then the load should be gradually removed by regular steps, and the following read- ings be made at each such step : — Amperes, input. Volts at machine terminals (kept constant). Watts, if induction motor. Speed of armature. Note sparking at brushes. Amperes, field (in D. C. machines). At least ten steps of load should be taken from full-rated load to no load. The ratio of the maximum drop in speed between no-load and full-load, which will be at full-load, to the speed at full-load, is the regulation of the motor. Efficiency Tests. Dynamos. The term efficiency has two meanings as applied to dynamos ; viz., electrical and commercial. The electrical efficiency of a dynamo is the ratio of elec- trical energy delivered to the line at the dynamo terminals to the total electri- cal energy produced in the machine. The commercial efficiency of a dynamo is the ratio of the energy delivered at the terminals of the machine to the total energy supplied at the pulley. Otherwise the electrical efficiency takes into account only electrical losses, while the commercial efficiency includes all losses, electrical, magnetic, and frictional. Core-lioss Test, and Test for friction and Windage. These losses are treated together for the reason that all are obtained at the same time, and the first can only be determined after separating out the others. A core-loss test is ordinarily run only on new types of dynamos and motors, but is handy to know of any machine, and if time and the facilities are available, should be run on acceptance tests by the consulting engineer. It consists in running the armature at open circuit in an excited field, driv- ing it by belt from a motor the input to which, after making proper deduc- tions, is the measure of the power necessary to turn the iron core in a field of the same strength as that in which it will work when in actual use. 384 TESTS OF DYNAMOS AND MOTORS. Connect as in the following diagram, in wMcli A is the dynamo or motor under test, and B is the motor driving the arma- f ■ ^ ture of A by means of gk PjJrvJ • the belt. The field of A (M, field J^rAJ'* 1 motor field must, of necessity, be T^ / ~ ~ =1 .exciter separately excited, as LrV_J f^S I /SC\ rS? rfiJ *-^ its own armature circuit ~^£Z Y^Sy) ^X^^hM-^-^y must be open so that excite* V^X belt, \ y generator eor there may be no current uSHS?^- DR,VJNB W0T0R C0J5REN T generated in its conduc- mDER TES ' W0T0R - tors. Fig. 3. Connections for a test of core loss. The motor field is sep- arately excited and kept constant, so that its losses and the core loss of the motor itself, being constant for all conditions of the test, may be cancelled in the calculations. The motor B should be thoroughly heated ; and bear- ings should be run long enough to have reached a constant friction condi- tion before starting this test, so that as little change as possible will take place in the different "constant" values. It is necessary to know accu- rately the resistance of the armature, B, in order to determine its I 2 R loss at different loads, and to use copper brushes to practically eliminate the I i R of brushes. It is well to make a test run with the belt on in order to learn at what speed it is necessary to run the motor in order to drive the armature A at its proper and standard speed. Friction, core loss, and windage of motor. — The speed having been determined, the belt is removed, and the motor field kept at its final adjustment, and enough voltage is supplied to the motor armature to drive it free at the standard speed. The watts input to the armature is then the measure of the loss (I 2 R) in the motor armature plus the friction of its bear- ings, plus its windage, plus core loss, or the total loss in the motor at no load. This is called the " running light " reading. Friction and windage of dynamo. — After learning the losses in the driving motor, the belt is put on and the dynamo is driven at its standard speed without excitation, and in order to be sure of this a volt- meter may be connected across the armature terminals ; if the slightest indication of pressure is found, the dynamo field can be reversely excited, to be demagnetized, by touching its terminals momentarily to a source of E.M.F. Take a number of readings of the input to the motor in order to obtain a good mean, and the friction and windage of dynamo is then the input to the motor, less the " running light" reading previously obtained, the I 2 R of motor armature having been taken out in each case. Let P = watts input to motor, P,= PR loss in motor armature when driving dynamo, f = " running light " reading of motor, h = friction and windage of dynamo armature, P 2 = 1 2 R loss of motor armature when " running light," then A = p_ ( p 1+ /_p 2 ). II rush friction. — The friction of brushes is ordinarily a small portion of the losses ; but when it is desirable that it should be separated from other losses, it can be done at the same time and in the same manner as the test for bearing friction. The brushes can be lifted free from the commutator or collector rings when the readings of input to the driving motor for bearing friction are taken ; dropping the brushes again onto the commutator and taking other readings, the difference between these last readings and those taken with brushes off will be the value of brush friction. Note, that allow- ance must be made as before for increase of I 2 R loss in the motor armature. Test for core loss. — Having determined the friction and other losses that are to be deducted from the total loss, a current as heavy as will ever be used is put on the dynamo field, the motor is supplied with current enough to drive the dynamo at its standard speed, and the reading of watts and current input to the motor armature is taken. The dynamo field current is now gradually decreased in approximately regular steps, readings of the input to the motor being taken at each such step until zero exciting current is reached, when the exciting current is reversed and the current increased in like steps until the highest current DYNAMO EFFICIENCY. 385 reading is again reached. This may now be again decreased by intervals back to zero, reversed and increased back to the starting-point, which will thus complete a cycle of magnetization ; ordinarily this refinement is not, however, necessary. This test must always be carried through without stop ; and although it is desirable to make the step changes in field excitation alike, if the excitation be changed in excess of the regular step it must not be changed back for the purpose of making the interval regular, as it will change the conditions of the residual field. When the readings are plotted on a curve, regularity in intervals of magnetization is not entirely necessary. The following ruling makes a convenient method of tabulation : — Dynamo. Motor. Speed amperes in field Speed amperes in field amperes in armature i volts in armature e Constant Constant. Constant. Computations. Watts in armature, belt on P„ = ie Running light reading / PR in arm, belt on Px ' PR in arm, belt off P% Core loss P t s-{Pi+f-PJ Plot on curve with exciting-current values on the horizontal scale, and the core loss on the vertical, and the usual core-loss curve is obtained. Separation of Core ILoss into Hysterenis and Eddr Current Lo*v Losses due to hysteresis and friction vary directly with the speed ; losses due to eddy currents vary as the square of the speed. Current and voltage must now be applied to the dynamo armature to drive it as a motor at proper speed, with the current in the separately excited field kept constant at proper value. Drive the motor (dynamo) at say two different speeds, one of which may be K times the other ; let P = total loss in watts, f x = loss in friction, H = loss by hysteresis, D rr loss by eddy currents, or P = /v, -f- H-\- D at the first speed, P. z=.'Kf y -f KH-\-K*D at second speed, KP — JtA -f kH+ KD, P 1 — KP=K*D — KD, P. — KPz^KDiK—l) n -P.-KP If K=2, then £>-. K(K-1) . Pi — 2P_ 2 (2 — 1) 2 Kapp and Housman separately devised the above method of separating the losses, but stated them somewhat differently. With the field separately excited at a constant value, different values of current are supplied to the armature at different voltages to drive it as a motor. The results are plotted in a ^urve which is a straight line, rising as the volts are increased. 386 TESTS OF DYNAMOS AND MOTORS. The following diagram shows how the losses are plotted in curves. The test as a separately excited motor is run at a number of different values of voltage and current in the armature, and the results are plotted in a curve as shown in the following diagram. The line a, 6, is plotted from the results of the current and volt readings. The line a, c, is then drawn parallel to the base, and represents the sum of all the other losses, as shown by previous tests, and they may be further separated and laid off on the chart. The triangle a, b, c, represents one-half of the value of the foucault cur- rent loss. . If another run be made with a different value of excitation, a curve, at, b t . or one below the original a, b, will be gotten, according to whether the total losses have been increased or decreased. If the higher values of current tend to demagnetize, by reason of the eddy currents in the armature, the curve a, 6, will curve upward somewhat at the upper end. .. , . Knowing the core-loss, friction and windage of a dynamo and the resis- tance of the various parts, the efficiency is quickly calculated, thus Let P = core-loss 4- friction (obtained as shown), V = voltage of armature, / = current of dynamo armature, /i = current of dynamo field, R = resistance of armature and brushes, Rx = resistance of field. Then, considering the above as the only losses (i.e., neglecting rheostats, etc.), y Efficiency = Vi + pr + 1 J r^T ' This is a satisfactory method of getting the efficiency, but does not take in M load losses " if any should exist. The simplest method of determining the effi- ciency of a direct-current machine is to run it light as a motor, without load or belting or gearing, at its proper field strength and its proper speed and measure the input to the armature. From this value subtract the PR loss in the armature and the remainder is the core and friction loss. Know- ing this and the resis- tance of the remaining circuits, all the losses are known, and hence the efficiency can be cal- culated. This method is an accurate one and is easy to carry out. Another teat for efficiency. — If the dy- namo under test is not of too large capacity, and a load for its full output is available, either in the form of a lamp bank, water rheostat, or other adjustable resistance, then one form of test is to belt it to a motor. By separately exciting the motor fields, and running the motor free with belt off, its friction can be determined, and with the resistance of the arma- ture known, the input to the motor in watts, less the friction and the PR loss in its armature at the given load, is a direct measure of the power ap- plied at the pulley of the dynamo. The output in watts, measured at the dynamos terminals, then measures the efficiency of the machine. BRUSH FRICTION BERBtflCl ERICTION AND WINDAGE WETS IN ARMATURE' 6? Fig. 4. Diagram showing separation of losses in dynamos. DYNAMO EFFICIENCY. 38T Let P = watts input to motor, Pi— losses in motor, friction, Pit, and core-loss, P x == watts output at dynamo terminals. P % of efficiency = 100 X -p — ~- ;= commercial efficiency. Knowing the current flowing in the armature and in the fields, and also knowing the resistance of the same, the I 2 R losses in each may be calcu- lated, which, added to the output at the dynamo terminals, shows the total electrical energy generated in the machine. If a = the I 2 R loss in the armature, / =the I 2 R loss in the fields. The electrical efficiency in per- centage will be 100 x — The adjoining diagram shows the connections for this form of test. It must be obvious that a steam- engine, or other motive power that can be accurately measured, may be used in place of the electric motor ; but measurements of mechanical power are so much more liable to error that they should be avoided where possible. The only objection to this method is that the friction of the driving-motor varies with the load, and the loss in the belt is not considered. SENERATOfi WATER -^RHEOSTAT. =3 FOR LOAD Fig. 5. Connections for efficiency test of a generator, driven by an electric motor. Kapp'i Test with Two Similar Direct-Current Djnamoi, Where two similar dynamos are to be tested, and especially where their capacity is so great as to make it difficult to supply load for them, it is com- mon to test them by a sort of opposition method ; that is, their shafts are either coupled or belted together, the armature leads are connected in series, the field of one is weakened enough to make a motor of it ; this motor drives the other machine as a generator, and its current is delivered to the motor. The difference in currents between the two machines, and for exciting the fields of each, is supplied by a separate generator. The following diagram shows the method of connecting two similar SWITCH Fig. 6. Connections for Kapp's method of efficiency test of two similar dynamos. 388 TESTS OF DYNAMOS AND MOTORS. dynamos for Kapp's test. D is the dynamo ; M the machine with field weakened by the resistance R, that acts as a motor, and G is the generator that supplies the energy necessary to make up the losses, excitation and differences. Start the combination and get them to standard voltage, as shown by the voltmeter ; then take a reading of the current with the switch on b, and another with the switch on a. Let the first reading be m, and the second rf, and let x be the efficiency of either machine, then Per cent efficiency of the combination = 100 X -j-» and :V( 10 0X^). In using this formula the efficiency of the dynamo at its load is assumed the same as the motor at its simultaneous load, which is usually true above the | load point. The loss in motor-field rheostat should also be allowed for. Another similar method, called "pumping back ," is to connect the shafts of the two machines as before, by clutch or belt ; arrange the electrical connections and instruments as in the following diagram : A.M. A.M. V.M. n v.m, Fig. 7. Efficiency test of two similar dynamos. D is the dynamo under test ; M is the similar machine used as a motor ; and G is the generator for supplying current for the losses and differences between M and D. The speed of the combination, as well as the load on D, can be adjusted by varying the field of M. The motor, M, drives D by means of the shaft or belt connection. M gets its current for power from two sources, viz., G and D. In order to determine the amount of mechanical power developed by M, and also to be able to separate the magnetic and frictional losses in the two machines, a, core-loss test should have been made on the machine M at the same speed, current, and E.M.F. as it is to have in the efficiency test. The loss in the cable con- nections between M and D must also be taken into account, and is equal to the difference in volts between voltmeters c and 6, X the current flowing in ammeter n. Let V— E.M.F. of D, shown on c, V, — E.M.F. of M by vm. 6, V„ — E.M.F. of G by vm. a, I =z amperes current from D by am. n, 7 / =c amperes current from G by am. I, I n = amperes current in M = I-\- /,, t = drop in connections between D and M = V — V m L = loss in connections between D and M = e X I t r = D's internal resistance, r x =: M's internal resistance, w r= core loss + armature loss -f field loss + friction of M in watts -f- L (loss in connections). ELECTRICAL METHOD OP SUPPLYING LOSSES. 389 Then W = the useful output of D= V X /, Wj = energy supplied by G = V„ X T n W + W, = total energy supplied to M, 7P" -f" FT/ — «* = energy required to drive D, % commercial efficiency of D == - I 2 r =. electrical loss in D, % electrical efficiency r= - W+ W, — w x 100. w ^ + /v xioo. The other way of calculating the efficiency with this arrangement is to measure the output — W x from G, with full load on D. W x then is the losses of both machines under load ; and knowing the I 2 R loss in the arma- ture and field of each, the efficiency is quicklv and accurately calculated. This method is best, as no core loss is required, and includes the " load Electrical Method of Supplying* the losses at Constant Potential. Modification of <4 Kapp Method" by Prof. Wm. L. Puffer, from notes privately printed for the students of the Massachusetts Institute of Technology. Specification. Two similar shunt dynamos under full load, one as a motor driving the other as a loaded dynamo through a mechanical coupling. Mains at same voltage as dynamos, and only large enough to supply the full-load losses of both dynamos. Line up the two dynamos carefully, and mechanically connect them by a good form of mechanical coupling, strong enough to transmit the full load to the dynamo. Connect the field magnet windings of each machine to the supply mains, putting a suitable field rheostat in each. If desirable for any reason, the field of the dynamo may be left connected as designed ; but the field of the motor, which does not in any way enter as a quantity to be measured during the test, should be connected to the supply mains. Fig. 8. Diagram of Connections for Professor Puffer's Modifi- cation of Kapp's Dynamo Test. Method of Starting*. Close the field circuit of the motor, and by the motor starting rheostat gradually bring the motor up to full speed. The dynamo armature will be also at proper speed and on open circuit. Now close the dynamo field and adjust the field rheostat until the dynamo is at about normal voltage. Adjust the speed roughly at first by the use of the field rheostat of the motor, remembering that an added resistance will cause the speed to rise. Next see that the voltage of the dynamo is equal to that of the motor, or, in other words, that there is no difference of potential between -opposite sides of the main switch on the dynamo. Close this switch and there may, or may not, be a small current in the dynamo armature. Now carefully 390 TESTS OF DYNAMOS AND MOTORS. increase the armature voltage of the dynamo, watching the ammeter, and weaken that of the motor ; a current will now from the dynamo to the motor, and the motor will transmit power mechanically to the dynamo. The current which was first taken from the supply wires to run the motor and dynamo armatures will increase somewhat. By a careful adjustment of the two rheostats and the lead on each machine, the conditions of full load of the dynamo may be produced. The motor is overloaded and its arm- ature will carry the sum of the dynamo and supply currents. Great care must be taken in adjusting the brushes of the macbines, because of great changes in the armature reactions which take place as the brushes are moved. It is well to remember that a backward lead to the motor brushes will increase the speed, as the armature reactions will considerably weaken the effective field strength. Cautions. The increase of speed will raise the dynamo voltage, and cause the cur- rent flowing in the armatures to greatly increase. A forward movement of the motor brushes will reduce both speed and current. A forward move- ment of the dynamo brushes will increase the armature reaction, and cut down the current through the armatures, while a backward movement will cause it greatly to increase. Very great care must be taken in adjusting the brush lead, as a movement of the brushes of either machine, which would be of little importance usually, will produce sometimes a change in current value equal to the full-load current. It is quite possible but poor practice to produce the load adjustment by use of the brushes alone. It is best to have ammeters of proper size in all circuits, but those actually required are in the dynamo leads and in the supply mains. A single volt- meter is all that is required. The field magnet circuits ought to be connected as shown, and the am- meters placed so that the energy in the fields does not come into the test of the losses in the armatures. The magnet of the machine under test, a dynamo in this case, should be under the proper electrical conditions for the load, yet not in the armature test, because the object of the test can best be made the determination of the stray power loss under the conditions of full load ; then having found this, assume the exact values of E, I, and speed, and so build up the data for the required efficiency under a desired set of conditions which might not have been exactly produced during the test. Immediately after the run, all hot resistances should be measured as rapidly and carefully as possible, to avoid any error due to a change in temperature. The energy given to the two armatures less the I 2 R in each armature, will be the sum of all the armature losses of the two dynamos under the conditions of the test, so that we measure directly the armature losses of the dynamos while fully loaded. It is evident that the two armatures are not under exactly the same con- ditions, except as to speed, for the dynamo armature w r ill have an intensity of magnetic field that will give an armature voltage of Vf + 7^ 2? ^4, while the motor will be weaker as Vf is the same for both armatures, and the motor armature voltage will be Vf — I a Bj t All the iron core losses will be made much greater in the dynamo than in the motor. The motor armature must carry a current equal to the sum of the dynamo and supply currents, and will get much hotter ; its reaction will also be greater, and there will be a tendency for greater sparking at the brushes. The total stray power thus obtained may be divided between the two armatures equally, but preferably in proportion to the armature voltages, unless the true law for the armatures is known. All resistances of wires, etc., must be noted and corrections applied, unless entirely negligible. Two 15-H.P. dynamos were tested by the class of '93, using this method. One of the full-load tests is here given as a sample of calculation. The exact rating of the dynamos is not known, but is nearly 45 amperes at 220 volts, with the dynamo at a speed of 1600 r.p.m. ELECTRICAL METHOD OF SUPPLYING LOSSES. 391 The averages of the observed readings taken during the test, and after a run of about five hours to become heated, was as below. Example of Calculation. (Connections as shown in Fig. 8.) Volts at supply point 220.3 Amperes of 15.71 Output of dynamo, amperes 45.80 Dynamo field current . . . 1.945 Speed 1594. To Measure Armature Resistance. Motor V— 1.952 J— 10.18 Dynamo V — 2.406 7=10.08 The motor field is out of the test while the dynamo field is in the test. Calculation. Watts supplied 220.3 X 15.71 = 3461. Dynamo armature R. = Motor armature R. = PadRad I 2 am Ram la = 45.80 -f 1.94 = 47.74 la = 45.80 + 15.71 =. 61.51 47.742 x .2387 = 554 = Pa Rod 61.51 2 X .1918 =. 725.4 = Pa Ram Dynamo Field = 1.945 X 220.3 = 428.4 Watts supplied = 3461 Dynamo field = 428.4 PR M = 725.4 PR D — 554.0 Total heat lost = 1697.8 1698 Total stray power = 1763 watts, for both machines. Vad Vam Vt+IaRa Vt — IaRa 47.74 X .2387 = 11.4 + 220.3 61.51 X .1918 = 11.8 -f 220.3 = 231.7 = Vad. = 208.5 = Vam. Divide the total stray power between the two armatures as their arma- ture voltages. 231.7 Stray power of dynamo, 23 17 ■' 208.5 X 1763 = 928, Stray power of motor = 1763 — 928.0 r= 835.0. The quantity 928.0 is the object of our test, i.e., the stray power when as nearly as may be under actual running conditions. Calculation of Efficiencies. As run. Output of dynamo = 220.3 X 45.80 = 10090 Watts output 554 I^Rad 10090 428 Field 544 928 Stray power 428 11990 Watts input to the dynamo. U062 == Work done by current- 392 TESTS OF DYNAMOS AND MOTORS. Efficiency of Conversion: 11062 x 100 11990 = 92.2 per cent. Commercial efficiency: 10090 X 100 119.90 Power required to run dynamo: 11990 = 84.1 per cent. 746 = 16.1 H.P. In this test, carbon brushes were used, and the lead adjusted as carefully as possible. If the exact rating of this dynamo had been 45 amperes and 220 volts at a speed of 1600, and we wished to find the efficiencies corresponding, we should proceed in this way. The test was made under conditions as nearly as possible to the rating, and the stray power as found will not be perceptibly different from what it would be under the exact conditions. When the load has been as carefully adjusted as in this test, it is seldom worth while to make these corrections, as they are smaller than changes pro- duced by accidental changes of oiling, temperature, brush pressure, etc., of two separate tests. Advantages of the Method. Small amount of energy used in making the test, namely, only the losses. No wire or water rheostat required. Test made under full load, and yet the losses are directly measured. All quantities are expressed in terms de- pending on the same standards, and therefore the efficiency will be but little affected by any error in the standards. No mechanical power measure- ments are made, and all measurements are electrical. Disadvantages. Requires two similar machines. Armature reactions are not alike in both machines. Leads are not alike. The iron losses are not the same. No belt pull on bearings. Must line up machines and use a good form of mechanical coupling. Sometimes difficult to set the brushes on the motor. The motor armature is much overloaded. u&$&m FOR FIELDS OF MOTORS Fig. 9. Diagram of Connections for Test of Street Motors, Prof. Puffer. Car ELECTRICAL METHOD OF SUPPLYING LOSSES. 393 Fig. 10. Diagram of Connections of Modification of the Previous Diagram, by Prof. Puffer. This method is of advantage in the test of railway series motors, if slightly modified by the separate excitation of the motor fields. If the series field windings be not separately excited there will be a great deal of unneces- sary difficulty from great changes of speed as the load is varied. However, one field may be kept in circuit on the machine used as a motor, as the test can then be made with the motor under its exact conditions. There will be a very great change of speed during adjustment of load, but there will be no danger of injuring anything, as the separate excitation of the dynamo field is an aid to steadiness. Railway motors, as generally made, will not stand their full rated load continuously, and the motor is likely to get too hot if not watched ; the machine used as a dynamo will run cold, as it will not have a large current in it. The friction of brushes is very large in these motors, and in general there is a want of accuracy in the division of the total stray power between the two armatures. It can only be very approxi- mately done by the aid of curves showing the relation between speed and stray power, and armature voltage and stray power, Hopkiaioni Vest of two Similar Direct-Current Dynamos. In the original Hopkinson method, the two dynamos to be tested were placed on a common foundation with their shafts in line, and coupled to- gether. The combination was then driven by a belt from an engine, or other source of power, to a pulley on the dynamo shafts. The leads of both ma- chines were then joined in series, and the fields adjusted so that one acted as a motor driven by current from the other. The outside power in that case supplied, and was a measure of the total losses in the combination, the efficiency of either machine being taken as the square root of the efficiency of the combination. Many modifications of this test have been used, especially in the substitu- tion of some method of electrically driving the combination, as the driving- power is so much easier measured if electrical. This test is somewhat like that last given, but the two machines are con- nected in series through the source of supply for the difference in power, such as a storage battery or generator. The following diagram shows the connections for the Hopkinson test, with a generator for supplying the dif- ference in power. In this test the output of G plus energy taken by M x (motor driving the system), gives losses of motor and dynamo (the losses of Mj being taken out). These losses being known, the efficiency can be calculated. If the two machines D and M are alike, G supplies the I 2 It losses of arma- tures, and M the friction, core losses, and I 2 R of fields. Another method useful where load and current are both available, is to drive one of two similar dynamos as a motor, and belt the second dynamo to it. Put the proper load on the dynamo, and the efficiency of the com- bination is the ratio of the watts taken out of the dynamo to the watts supplied to the motor. The efficiency of either machine, neglecting smalJ differences, is then the square root of the efficiency of both. 394 TESTS OF DYNAMOS AND MOTORS. Fig. 11. Diagram of connections for Hopkinson's test of two similar dynamos. If P = watts put into the motor, P x z=. watts taken from the dynamo, x = per cent efficiency of the combination, y =z efficiency of either machine, A x loo *- P ' The above test is especially applicable to rotary converters, the belt being discarded, and the a c sides being connected by wires ; thus the first ma- chine supplies alternating current to the second, which acts as a motor generator with an output of direct current. The only error (usually small) is due to the fact that both machines are not running same load, since that one supplies the losses of both. Fleming*'* Modification ofHopkimon Test. — In this case the two dynamos under test are connected together by belt or shafts,- and are Fig. 12. driven electrically by an external source of current, say a storage battery or another dynamo, which is connected in series with the circuit of the two machines. Figure 12 shows the connections for this test, which will be found carried out in full in Fleming's " Electrical Laboratory Notes and Forms." Motor Tests. Probably the most common method of testing the efficiency and capa- city of motors is with the prony brake, although in factories where spare dynamos are to be had, with load available for them 5 there can be no MOTOR TESTS. 395 question that belting the motor to the dynamo with an electrical load is by far the most accurate, and _i * the easiest to carry out. Piony brake test. — In this test a pulley of suitable dimensions is applied to the motor-shaft, and some form of friction brake is applied to the pulley to absorb the power. prony brakeT 55 ^^ ^ ^ The following diagram shows ■p IG 13 one of the simplest forms of prony brake ; but ropes, straps, and other appliances are also often used in place of the wooden brake shoes as shown. Note. — See Flather, " Dynamometers and the Measurement of power." As the friction of the brake creates a great amount of heat, some method of keeping the pulley cool is necessary if the test is to continue any length of time. A pulley with deep inside flanges is often used ; water is poured into the pulley after it has reached its full speed, and will stay there by reason of the centrifugal force until it is evaporated by the heat, or the speed is lowered enough to let it drop out. Rope brakes with spring bal- ances are quite handy forms. The work done on the brake per m inute is the product of the following items : I = the distance from the centre of the brake pulley to the point of bearing on the scales, in feet, n = number of revolutions of the pulley per second, w •=. weight in lbs. of brake bearing on scales. Power = 2 it I n w = foot-pounds per second, and H.P. =?^?= 0.011424 In w. The input to the motor is measured in watts, and can be reduced to horse- power by dividing the watts by 746; or the power absorbed by the brake can be reduced to watts as follows: Brake watts = 8.52 I n w = P. If the length, I, be given in centimeters, and the weight, iv, be taken in kilograms, the horsepower absorbed by the brake is given by the formula H.P. = S26lnwl0~\ Again taking the length in centimeters and the weight in kilograms, the watts absorbed by the brake are Brake watts = 0.616 In w. p The watts input = Pi and efficiency in percentage = pT X 100. Using feet and pounds in the measurements, the efficiency in percentage will be _ 852 I n w Eff * = Pi ' Using centimeters and kilograms the efficiency will be 61.6 In w Eff. = ■ Pi If it is desired to know the friction and other losses in the motor, after the brake test has been made, the brake can be removed, and the watts neces- sary to drive the motor at the same speed as when loaded, can be ascertained. Electrical load test (including loss in belting, and extra loss in bear- ings due to pull of belt). — This test consists in belting a generator to the motor and measuring the electrical output of the generator, which added to the friction and other losses in the generator, makes up the load on the motor. The efficiency is then measured as before, by the ratio of output to input. The great advantage of this form of test is, that it can be carried on for any length of time without trouble from heat, and the extra loss in bearings due to pull of belt is included, which is therefore an actual com- mercial condition. 396 TESTS OF DYNAMOS AND MOTORS. In this form of test the losses in the generator are termed counter torque, and the method of determining them is given following this. Counter torque. — In tests of some motors, especially induction mo- tors, the load is supplied by belting the motor under test to a direct current generator having a capacity of output sufficient to supply all load, including overload. In determining the load applied to the motor and the counter torque, it is necessary to know, besides the /. E. or watts output of the generator, the following : — 1*11 of generator armature, Core loss of generator armature, Bearing and brush friction and windage of generator, Extra bearing friction due to belt tension. It is necessary to know the above items for all speeds at which the com- bination may have been run during the testing. This is especially useful in determining the breakdown point on induction and synchronous motors, both of which can be loaded to such a point that they " fall out of step." While the motor is under test especial note should be made of the speeds at which the motor armature and generator armature rotate, and of the watts necessary to drive the motor at the various speeds without load. The counter torque will then be the sum of the following three items : — P =r PR of generator armature, Pe = core loss of generator armature, F = bearing and brush friction and windage of the generator armature. The field of the dynamo must be separately excited and kept at the same value during the load tests and the tests for " stray power " and does not enter into any of these calculations. Belt-on test. — After disconnecting current from the motor under test, and with the belt or other connection still in place, supply sufficient voltage to the dynamo armature to drive it as a motor at the speeds run during the motor test, holding the field excitation to the same value as before, but adjusting the voltage supplied to the armature for changing the speed. Take readings of Speed, i.e., number of revolutions of dynamo armature. Volts at dynamo armature. Amperes at dynamo armature. Construct a curve of the power required to drive the combination at the various speeds shown during the motor test. Belt-off test. — Throw the belt or other connection off, and take read- ings similar to those mentioned above, which will show the power necessary to drive the dynamo without belt. Then for any speed of the combination the " stray power" will be found as follows : — P, = watts from belt-off curve, required to drive the dynamo as a motor. p n ■=. watts from belt-on curve, required to drive the combination. Pc ■=. core loss in dynamo armature. F = friction of dynamo belt-off. F, •=. friction of motor under test, running light and without belt. f— increase in bearing friction of dynamo, due to belt tension. /,= increase in bearing friction of motor, due to belt tension. From the belt-off curve, P, = Pc + F (1) From the belt-on curve, P„ = Pe + F+F s +f+f 4 (2) INDUCTION MOTORS. 397 Subtract (1) from (2) ■ />, = *■,+/+/, (3) The values of /and f, cannot be determined accurately ; but if the ma- chines are of about the same size as to bearings and weights of moving parts, it is very close to call them of equal value, when, /or/, _ (P„ - P, - F t ) (4) The friction F t of the motor under test has been previously found by noting the watts necessary to drive it at the various speeds. If it is an in- duction motor, the impressed voltage is reduced very low in determining the friction in order that the core loss may be approximately zero. As all the values of the quantities on the right-hand side of the equation (4) are now known,/ is determined, and may be added to P, to give the total 11 strew power." A curve is then plotted from the values of " stray power' at different speeds. Counter torque = (P, -f- f). Total load z=IE + PR + (P, +/), where IE = watts load on the D. C. machine when it is being driven by the motor. If S = Pi +/ = " stray power" then Total load = IE + l*R + 8. The value of/ is so small when compared with the total load, that any ordinary error in its determination will be unimportant. BOOSTER SUPPLYING IP. Test of Street-Railway Motors. The " pumping-back " test, as described before, with some little modifica- tion serves for testing street-railway motors. The following diagram shows the arrangement and electrical connections. The motors are driven mechanically by another motor, the input to which is a measure of the losses, frictional, core Eg g shaft losses, gears, bearings, t^^TfT^wT H H supplying core etc., in the two motors; VMr -§f D l_l &_J M ^g 5 losses and friction the two motors are connected in series, through a booster, B, care being taken to make the connections in such a manner as to have the direction of rotation the same ; and their voltages op- posing. Readings are taken and the efficiencies are calculated as in the u pumping- back " test. In eliminating the friction of bearings, etc., and of the driving-motor, it is run first without belts, the input being recorded as taken, at the speed necessary. The belt is then put on and a reading taken at proper speed, with both the motors under load. The load being adjusted by varying the field of booster B, the total losses of the system are then IE from booster plus the difference between belt-on reading with full load through the motors, and belt-off reading as noted (allowance being made for change of I 2 R of driving-motor). If the two motors are similar, half this value is the loss in one motor, from which the efficiency can be calculated as previously shown. Induction motors. — In addition to the tests to which the D. C. motor Fig. 14. Diagram of connections and arrange- ment of street-railway motors. 398 TESTS OF DYNAMOS AND MOTORS. is ordinarily submitted, there are several others usually applied to the in- duction motor, as follows : — Excitation ; Stationary impedance ; Maximum output ; and some variations on the usual heat and efficiency tests. Excitation: This is also the test for core loss-f friction, allowance being made for I 2 E of field ; with no belt on the pulley the motor is run at full impressed voltage. Read the amperes of current in each leg, and total watts input. The amperes give the excitation or " running-light" current, and the watts give core loss -}- friction + I 2 R of excitation current. Stationary impedance : Block the rotor so it cannot move, and read volts and amperes in each leg, and total watts input. This is usually done at half voltage or less, and the current at full voltage is then computed by proportion. This then gives the current at instant of starting, and a meas- ure of impedance from which, knowing the resistance and core loss, other data can be calculated, such as maximum output, efficiency, etc. Maximum output : This might be called & break-down test; as it merely consists in loading the motor to a point where the maximum torque point is passed and thus the motor comes to rest. Keep the impressed voltage constant and apply load, reading volts, am- peres in each leg, the total watts input, and revolutions ; also record the load applied at the time of taking the input. Then take counter torque as explained before, from which the efficiency, the apparent efficiency, the power factor, and maximum output are immediately calculated. Heat test. — Run motor at full load for a sufficient length of time to develop full temperature, then take temperatures by thermometer at the following points : — 1. Room, not nearer to the motor than three feet and on each side of motor. 2. Surface of field laminations. 3. Ducts (field). 4. Field or stator conductors, through hole in shield. 5. Surface of rotor. 6. Rotor spider and laminations. 7. Bearings, in oil. During heat run, read amperes and volts in each line. Efficiency test. — Apply load to the motor, starting with nothing but friction ; make readings at twelve or more intervals, from no load to break- down point. Keep the speed of A. C. generator constant, also the impressed voltage at the motor. Read, Speed of motor. Speed of A. C. dynamo. Amperes input to motor, in each leg. Volts impressed at motor terminals. Watts input to motor, by wattmeter. Current and volts output from D. C. machine belted to motor, Counter torque as explained above, and excitation reading watts. From the above the efficiency, apparent efficiency, power factor -^- — ^ — 7s— ; ) » & n N c p r.p.s. 10— 8 1.11 k Np r.p.s. 10~ 8 jjj — _ _ == . 2 vlrnq mq In three-phase alternators the E.M.F. between terminals will depend upon the method of connecting the armature conductors. The two most common methods are called the delta connection and the Y or star connection, both shown in the following diagrams. DELTA CONNECTION Y OR STAR CONNECTION Figs. 1 and 2. Values of E.M.F. in three-phase connections when x = y=zz. In the delta- connected armature the E.M.F.'s between terminals are those generated in each coil, as shown in the diagram. In the Y-connected armature the E.M.F. between any two terminals is the E.M.F. generated by one of the coils in that phase multiplied by the V3 or 1.732. Two-phase circuits are sometimes connected as a three-phase circuit ; that is, both phases have a common return wire. In this case the pressure be- tween the two outgoing wires is ^2 x E, and the current in the common return will be / V2, both conditions are on the assumption that E and / in each phase is the same. 404 ENERGY IN THREE-PHASE CIRCUITS. 405 The current from an alternator depends upon inductance and resistance. The coefficient of inductance is represented by the letter L. The E.M.F. of an alternator follows approximately a sine curve, and the current from it is represented by the same kind of curve. Since in a circuit, lines of force exist in proportion to the current flowing, at each of its different cur- rent values there is a new value of lines in force. Thus, in a circuit of varying current there is a continuously varying flux, and hence there is in- duced a back E.M.F. This back E.M.F. is called the back E.M.F. of self- induction, and it retards the current flow just as does resistance. This back E.M.F. of self-induction combines with the resistance, but at right angles thereto, the result being called impedance. The coefficient of self-induction = max. flux x turns L = w .,_. — = henrys. amperes X 10 8 J Henrys multiplied by 2 *■/ =. reactance ohms (/ = cycles per second). In a circuit where R = resistance ohms, and 2 tt fL = reactance ohms, these combine at right angles to produce impedance ohms, or the total opposing force of the current, thus: Impedance = V R* -f (2 n fL)*. Hence in an alternator circuit if the coefficient of self-induction of the alternator be L, and that of the external circuit be L x \ and if the resistance of the alternator armature be R, and that of the external circuit be i? t , and the effective E.M.F. generated in the alternator armature =: E f then the current flowing will be E V ( 22 + R % + ( 2 nfL + 2 nfLJ* Energy in an Entirely Non-Inductive and Balanced Three-Phase Circuit. In the following diagram of a Y-connected multiphase generator and cir- cuits, let e ± = E.M.F. of any phase in the armature, i x = current of any phase in the armature, E = E.M.F. between mains, / = current in any main, 1 Fig. 3. P t = power of one phase of the armature, P =. total power, P 1 = e 1 i l ; but hence P = 3 w x = 5-5f= 1.732 EI, V3 and /- I. _ ' 1.732 E 406 ALTERNATING- CURRENT MACHINES. In the following diagram of a delta-connected polyphase generator and circuits, let € 2 =zE f I=i 2 V3, P 2 — e 3 i 21 P=z3P 2 = ^= = 1.732 EI, V3 P I — 1.732 E Fig. 4. Where the circuit is inductive, in order to determine the real power tin above result must be multiplied by the " power factor," or the cosine of the " angle " bv which the current lags behind or leads the E.M.F. Thus the power in a circuit in which the current lags 9 degrees behind the E.M.F. = IE cos 9. If the current lags 90° behind the E.M.F. there will be no energy developed as cos 90° = 0. The cosine of the angle of lag 0, or the power factor, is equal to the ratio of the true watts to the apparent watts. In ordinary lighting distribution, the power factor is high so that rough calculations are made without ita value being exactly known. Angle of Las': To determine with a watt meter in three* phase circuit* (Fig. 5) : Connect the current coil in one lead ; connec* Wm Fig. 5. one end of the potential coil to x on the same lead ; now connect the re maining end first to one of the remaining leads y, then to z, calling the firs'* reading P t and the second, P n ; then if 9 = angle of lag, When 9 is greater than 60 degrees, one reading will be negative, so tha> the difference of readings will be greater than their sum. If i? = resistance per leg of Y-connected armature, r= resistance per phase of A-connected armature, then, PR loss in Y-connected armature = 3 PR PR loss in A-connected armature =»©; = Pr. Energy in UTon-Inductive Three-Phase Circuits. / — current in any one of the three wires of external circuit, i = current in one phase of the armature for delta connection, p- watts output of a balanced three-phase generator, 1.732 = V3, .577 = 1 — V3, E = volts between terminals (or lines) on either delta or Y system, v — volts of one phase of the armature it connected in " Y," R= resistance per leg, of Y-connected armature, r = resistance per phase of A-connected armature, P=3 I,v = 3 ' —F E 1.732 (either with Y or A armature). V3 COPPER LOSS IN ARMATURES. 407 For A P = 3r y i = 3v, V3 v.— E v p — r , 7 = 1.732 i£ /, which shows statement in brackets to be true. V 3 w '— E X 1.732 /, = 1.732 i in delta system. I 2 R loss in Y connected armature = 3 7/72. 7 2 i? loss in A connected armature = 3 ( -~ \ r =. Ifr. — x y \*\ E %*& 1" E E !i L. fC< r E E Fig. 6. Fig. 7. E=^E,z= 1.1Z2E,. E=E, / AMPERES = 1 .732 X 2 OT X I AMPERES ~Z \ I AMPERES = 1.732X Z or y AMPERES ■ 1.782.X V or X \Z AMP8. )J£™£: I AMPERES =y 3CAMP8. / AMPERES =0J Delta Connection. Star or Y Connection. ftGS. 8 and 9. Values of current in three-phase connections, where x=zy=zz. Copper LoMi in the Armatures of Alternators. A. Ruckgaber. In the armature of any alternating-current dynamo or motor of either single or polyphase the copper loss is always equivalent to -~— , in which /= total amperes and R = the measure of resistance between leads of a phase, usually taken as an average of the measurements of the armature resistance of each phase. Let R = Resistance as measured (average). r — Resistance per phase. / z= Total amperes = watts -J- volts. I x — Amperes per lead. i — Amperes per phase, in winding. Single-Phase. - Here /= I x = i ; and R : l^R loss = I*R. 408 ALTERNATING-CURRENT MACHINES. Two-Phase Independent Wind- ing* (Fig. 10). R is measured from 1 to 3 and 2 to 4. '=S _ watts volts '1=4 Then I*R loss = 2 X PB _JLI 2 B 4 r-R 2 Fig. 10. Two-Phase Winding-* Connected in Series (Fig. 11). E X 2 The I X 2 R loss = A i*r = R is measured from 1 to 3 and 2 to 4, the average of these two being taken for the value of R. B= (r + r)(r + r) _ r h _ V2~ I : 2V2 4 l^r __ 8 ~~ /2 r 2 Then 4 r /2/2 : 2 The IJR loss Three-Phase Star Connection (Fig. 12). E V 3 Then the /^ loss = 3 i*r = 3 T^r = i 2 r. J? is measured from 1 to 2, 2 to 3, and 3 to 1, the average of the three being the value used for R. Then R as measured — 2r. PR .-. The IJR loss = ■ Fig. 12. Three-Phase Delta Connection (Fig. 13) E l V3 *V3 3* Then Vi? loss = 3 iV = I*r 3 * -ft is measured from 1 to 2, 2 to 3, and 3 to 1, the average of these being ^ 3 taken as the value of R. r ( r _L r ) 9 T2J? Then R as measured = ^ ; = - r and the IJR loss = ^. r -+• r -\- r 6 * 2 -AA/WV\\WvWWWV-- r Fig. 13. 2\ REGULATORS LOR GENEKATORS. 409 Compensated Revolving* field Alternators. The General Electric Company in October, 1899, placed on the market a new type of polyphase alternator, which is claimed to overcome many of the faults common to the old style of machine, especially when used on combined lighting and motor loads. While it has been found a compara- tively easy matter to compound and over-compound for non-inductive loads, it has been heretofore quite difficult to add excitation enough to compound for inductive loads which require considerably more held current than do loads of a non-inductive nature. The following description is taken from the bulletin issued by the makers describing the machine, which is of the revolving field type : — " The means by which this result is accomplished are as follows : The shaft of the alternator which carries the revolving field carries also the armature of the exciter, which has the same number of poles as the alter- nator, so that the two operate in synchronous relation. In addition to the commutator, which delivers current to the fields of both the exciter and the alternator, the exciter has three collector rings through which it receives current from one or several series transformers inserted in the lines leading from the alternator. This alternating current, passing through the exciter armature, reacts magnetically upon the exciter field in proportion to the strength and phase relation of the alternating current. Consequently the magnetic field and hence the voltage of the exciter, are due to the combined effect of the shunt field current and the magnetic reaction of the alternating current. This alternating current passes through the exciter armature in such a manner as to give the necessary rise of exciter voltage as the non- inductive load increases, and without other adjustment, to give a greater rise of exciter voltage with additions of inductive load." REGULATORS FOR ilTERXATOG C1IBREWT GENERATORS. General Electric Company. This regulator automatically maintains the voltage of the generator at the desired value by varying the exciter voltage. This is done by rapidly opening and closing a shunt circuit across the exciter field rheostat. Fig. 14 shows the elementary connections of the regulator. The rheostat shunt circuit is opened and closed by a deferentially wound relay. The current for operating this relay is taken from the exciter bus bars and is controlled by the floating main contacts. The current for operating the direct-current con- trol magnet is also taken from the exciter bus bars. The relay and the direct- current control magnet constitute the direct-current portion of the regulator, and maintain not a constant but a steady exciter voltage. The alternating- current portion of the regulator consists of a magnet having a potential winding connected, by means of a potential transformer, to the bus bars or the circuit to be regulated. This magnet also has an adjustable compen- sating winding which is connected in series with the secondary of a current transformer usually inserted in the principal lighting circuit. The core of this magnet is attached to a pivoted lever carrying a counterweight which is balanced by the attraction of the magnet. If a load is thrown on the genera- tor the voltage will tend to drop, the alternating-current magnet will weaken and destroy the balance of the core and lever and cause the main contacts to close ; this in turn will close the relay contacts and entirely short-circuit the exciter field rheostat, thus increasing the exciter voltage until the origi- nal balance of the alternating-current magnet core and lever is restored and the alternating-current voltage maintained at the required value. In some cases the exciter voltage will vary from 70 to 125 volts from no load to full load. This is especially true if the load is partly inductive and the regulator is adjusted to compensate for the line loss. In order to get the full range of regulation within the scope of the regulator in such cases, the alternating field rheostat should be turned entirely out and the exciter field rheostat adjusted to lower the alternating-current voltage about 65 per cent below normal. When the regulator is switched in, it will close the rheostat shunt circuit and instantly build the voltage up to nor- mal, and maintain normal voltage by rapidly opening and closing the rheostat shunt circuit. 410 ALTERNATING-CURRENT MACHINES. MAIN CONTACTS POTENTIAL INDING POTENTIAL GENERATOR TRANSFORMER ^^ RHEOSTAT PCT 5 CURRENT TRANSFORMER A.C. GENERATOR Fig. 14. Diagram of Tirrell regulator and connections for a single genera- tor and exciter. Alternating-Current Armatures. Almost any continuous current armature winding may in a general way be used for alternating currents, but they are not well suited for such work, and special windings better adapted for the purpose are designed. Alternating current armature windings are open-circuit windings, except- ing in the rotary converter, where the rings are tapped directly on to the direct current armature windings. Early forms of armature windings of this type, as first used in the United States, had pan-cake or flat coils bound on the periphery of the core. In the next type the coils were made in a bunched form, and secured in large slots across the face of the core. Both these types were used for single- phase machines. After the introduction of the multiphase dynamo, arma- ture windings began to be distributed in subdivided coils laid in slots of the core ; and this is the preferred method of to-day, especially so in the case of revolving field machines. The single coil per pole type of winding gives the larger E.M.F., as the coils are thus best distributed for influence by the magnetic field. This type also produces the highest self-induction with its attendant disadvantages. The pan-cake and distributed-coil windings are much freer from self-induc- tion, but do not generate as high E.M.F. as does the single-coil windings. In well-considered multiphase windings the E.M.F. is but little less for distributed coils than for single coils, and has other advantages, especially where the use of step-up transformers permits the use of low voltages, and consequently light insulation for the coils. The distributed-coil winding offers better chance for getting rid of heat from the armature core, and the conductor can in such case be made of less cross-section than would be required for the single-coil windings. The greater number of coils into which a winding is divided, the less will ARMATURE WINDINGS. 411 >e the terminal voltage at no load. Parshall & Hobart give the following •atio for terminal voltage under no-load conditions : Single-coil winding = 1. for the same total number of conductors, the spacing of conductors being uniform over the whole circumference. Two-coil winding =: .707. Three-coil winding — .667. Four-coil winding = .654. When the armature is loaded, the current in it reacts to change the termi- nal E.M.F., and this may be maintained constant by manipulation of the exciting current. With a given number of armature conductors this reac- tion is greatest with the single coil per pole winding, and the ratios just given are not correct for full-load conditions. Single -phase Winding's. — The following diagram shows one of the simplest forms of single-phase winding, and is a single coil per pole winding. Fig. 15. Another similar winding, but with bars in place of coils, is shown in the following figure. It can be used for machines of large output. Fig. 16. 412 ALTERNATING-CURRENT MACHINES. The following figure shows a good type of three bars per pole winding, which is simple in construction. Fig. 17. Two-phase Winding's. — The following diagram shows a good type of winding for quarter-phase machines. It utilizes the winding space to good advantage, and is applicable to any number of coils per pole per phase. Fig. 18. Fig. 19 is a diagram of a bar winding for a quarter-phase machine, with four conductors per pole per phase. Tnre«-pliase Winding's. — Fig. 20 is a diagram of a three-phase ARMATURE WINDINGS. 413 srinding connected in Y, in which one end of each of the three windings is connected to a common terminal, the other ends being connected to three collector rings. Fig. 19. Fig. 31 is a sample of a three-phase delta winding, in which all the con- ductors on the armature are connected in series, a lead being taken off to a collector ring at every third of the total length. Fig. 20. Fig. 21. In the Y windings the proper ends to connect to the common terminal and to the rings may be selected as follows : Assume that the conductor in the middle of the pole-piece is carrying the maximum current, and mark its direc- tion by an arrow ; then the current in the conductors on either side of and ad- jacent to it will be in the same direction. As the maximum current must be coming from the common terminal, the end toward which the arrow points must be connected to one of the rings, while the other end is connected to the common terminal. It is quite as evident that the currents in the two adjacent conductors must be flowing into the common terminal, and there- fore the ends toward which the arrows point must be connected to the com- mon terminal, while their other ends are connected to the remaining two rings. In a delta winding, starting with the conductors of one phase in the mid- dle of pole-piece, assume the maximum current to be induced at the moment in this conductor ; then but one-half the same value of current will be included at the same moment in the other two phases, and its path 414 ALTERNATING-CURRENT MACHINES. and value will best be shown in the following diagram, in which x may be taken as the middle collector-ring, and the maximum current to be flowing from x toward z. It will be seen that no current is coming in over the line y, but part of the current at z will have been induced in branches b and c. Most three-phase windings can be connected either in Y or delta ; but it must be borne in mind that with the same windings the delta-connection will stand 1.732 times as much current as the Y- connection, but gives only — — as much voltage. Fig. 22. Path and Value of Current in Delta- eonnected Armature. Armature Reaction of an Alternator. Since the armature core is a part of the magnetic circuit, and since the armature winding surrounds this core and also carries current, it must be expected that this current influences the total magnetism of the machine and hence its voltage. This effect, combined with the natural inductance of the winding, itself constitutes what is called armature reaction. Fig. 23 a* ]a A r ,mmmm "" l"^' A it\ V * < } » / N b'\ ^ffj' /b ' S ( I ,,B* c 'Xe A m \ m Fig. 23. ■hows an alternator in its elements. The armature winding is tapped in two places and connected to the collector rings d and e, from which the current flows to the external circuit. This current passing through the winding on the armature creates a magneto-motive force, which tends to produce the flow of magnetism as shown by the dotted lines a—b—c; a' — b' — c', or in a general direction, m — n. The field current proper entering at A and coming out at B tends to pro- duct magnetism in the direction x — y, at right angles to m — n. Under such conditions, therefore, the ampere-turns of the armature are acting at right angles to the ampere-turns of the field. This is the condition under non-inductive load, the maximum current of the armature occurring in time and space simultaneously with the maximum E.M.F. If the maximum of the current of the armature occurs later than the maximum of the E.M.F. , or in other words, if the current lags behind the E.M.F., the ampere-turns of the armature are no longer acting in a direc- tion m — n when the current is a maximum, but in a direction m' — n\ partially opposing the main flux x — y. If the lag of current becomes 90° the armature reaction would turn still more around, becoming, in fact, just opposite to x — y. Thus, on non-inductive load, the armature ampere-turns combine with the field ampere-turns at right angles, and with increasing lag show a higher and higher resultant until at 90° lag the two combine by direct addition. Just similar to all this is the self-induction component of the armature inductance. As has been pointed out, self-induction lags in its opposing ARMATURE REACTION. 415 effects behind the current, thus on non-inductive load, the opposing effect of self-induction is shown by Fig. 24. Fig. 24. Let a — c =zlz=. the current, a — dz= E = the E.M.F. generated by the revolutions of the arma- ture, a — b-=z the resistance drop r= IR in phase always with the current, a—g = IX '= the inductive drop 90° away from the current. The resultant of these = a — e = E =: the total E.M.F. necessary to pro- duce to give the value E under the conditions. If the current lags these values are as shown in Fig. 25, the current lag- FlG. 25. ging behind and E.M.F. by the angle 0. At 90° lag the E.M.F. of self- induction is just in line with E, hence is added directly to give the total E.M.F. E necessary to generate to product E. Thus a similarity exists between the armature reactive effect shown in Fig. 23 and the armature self-inductive effect shown in Figs. 24 and 25. On this account it has been suggested by Mr. C. P. Steinmetz that the two values be combined into one and the combined value be given the term " synchronous impedance." This value is obtained in an actual alternator by short-circuiting the armature upon itself and reading the ampere-turns in the field coils necessary to give full armature current, which is then expressed in terms of ampere-turns. Since on short-circuit the armature ampere-turns are exactly opposing the field ampere-turns, this reading gives a direct measure of the armature opposing forces, but conveniently converted into ampere-turns. To calculate from this value the amount of ampere-turns necessary in a given alternator to give a certain voltage, pro- ceed as follows : Let A equal the ampere-turns necessary to produce the terminal voltage E of the alternator when running on open circuit : let B equal the syn- chronous impedance ampere-turns obtained as above. Then the total ampere-turns required to produce the voltage E on non-inductive load = V J2 -|_ B 2 If the current is not non-inductive the two values must be combined with proper phase relation, as shown in Figs. 24 and 25. The 416 ALTERNATING-CURRENT MACHINES. method has been extensively used and for ordinary designs seems a very useful one to follow. A designer can calculate this value to a very close approximation, thus predetermining the regulation. It can be seen from this that a single-phase alternator gives a pulsating armature reaction. A polyphase armature gives a constant armature reaction since it can be shown that at any instant the magnetic resultant of the current is the same. For this reason, among others, a polyphase alternator is more efficient than a single-phase machine since the pulsating armature reaction sets up eddy currents from its variable nature, which increases the losses. §YICHRONIZEH§. There are numerous methods of determining when alternators are in step, some acoustic, but mostly using incandescent lamps as an indicator. In the United States it is most common to so connect up the synchronizer that the lamp stays dark at synchronism ; in England it is more usual to have the lamp at full brilliancy at synchromism, and on some accounts the latter is, in the writer's opinion, the* better of the two, as, if darkness indi- cates synchronism, the lamp breaking its filament might cause the machines to be thrown together when clear out of step ; on the other hand, it is some- times difficult to determine the full brilliancy. The two following cuts show theory and practice in connecting synchro- nizers. IO.TERNA.TOR HO. 1 /& /& —.moo, 1 |— vQj&fls K\ ^© "^ Fig. 26. Synchronizer Connections. When connected as shoivn, the lamp will show full c.p. at synchronism. If a and b are reversed, darkness of lamp will show synchronism. Fig. 27. Synchronizer Connections. Lamp lights to full c.p. when dyna- mos are in synchronism. Two transformers having their primaries connected, one to the loaded and the other to the idle dynamo, have their secondaries connected in series through a lamp ; if in straight series the lamp is dark at synchronism ; if the secondaries are cross-connected the lamp lights in full brilliance at synchronism. The Lincoln Synchronizer is so made as to move a hand around a dial so that the angle between the hand and the vertical is always the phase angle between the two sources of electro-motive force to which the synchronizer is connected. If the incoming alternator is running too fast the hand deflects in one direction, and if too slow, in the opposite direction. Coincidence in phase occurs when the moving hand stands vertically. A complete revolution of the hand indicates a gain or loss of one cycle in the frequency of the incoming alternator as compared with bus-bars. SYNCHRONIZING GENERATORS. 417 Suppose a stationary coil F, Fig. 28, has suspended within it a coil A, free to move about an axis in the planes of both coils and including a diameter of each. If an alternating current be passed through both coils, A will take up a position with its plane parallel to F. If now the currents in A and F be reversed with respect to each other, coil A will take up a position 180° from its former position. Reversal of the relative directions of currents in A and F is equivalent to changing their phase relations by 180°, and therefore this change of 180° in phase relations is followed by a correspond- ing change of 180° in their mechanical relations. Suppose now, that instead of reversing the relative direction of currents in A and F, the change in phase relations between them be made gradually and without disturbing the current sjbrength in either coil. It is evident that when the phase difference between A and F reaches 90° the force between A and F will become reduced to zero, and a movable system, of which A may be made a part, is in condition to take up any position demanded by any other force. Let a second member of this movable system consist of coil B, which may be fastened rigidly to coil A, with its plane 90° from that of coil A, and the axis of A passing through a diameter of B. Further, suppose a current to circulate through B, whose difference in phase rela- tive to that in A, is always 90°. It is evident under these conditions that when the differ- ence in phase between A and F is 90°, the movable system will take up a position such that B is parallel to F, because the force between^ and F is zero, and the force between B and F is a maximum ; similarly when the difference in phase between B and F is 90°, A will be parallel to F. That is, beginning with a phase difference be- tween A and F of 0, a phase change of 90° will be followed by a mechanical change on the movable system of 90°, and each suc- cessive change of 90° in phase will be followed by a corresponding mechanical Fig. 28. Lincoln Synchronizer, change of 90°. For intermediate phase relations it can be proved that under certain conditions the position of equilibrium assumed by the movable element will exactly represent the phase relations. That is, with proper design, the mechanical angle between the plane of F and that of A and also between the plane of F and that of B is always equal to the phase angle between the current flowing in F and those in A and B respectively. As commercially constructed coil F consists of a small laminated iron field-magnet with a winding whose terminals are connected with binding posts. The coils A and B are windings practically 90° apart on a laminated iron armature pivoted between the poles of the magnet. These two windings are joined, and a tap from the junction is brought out through a slip-ring to one of two other binding posts. The two remaining ends are brought out through two more slip-rings, one of which is connected to the remaining binding post, through a non-inductive resistance, and the other to the same binding post through an inductive resistance. A light aluminum hand attached to the armature shaft marks the position assumed by the armature. < lUTDUCTOIt tym: synchroscope. From The Electric Journal. This type is especially applicable where voltage transformers are already installed for use with other meters. As it requires only about ten apparent watts it may be used on the same transformers with other meters. There are three stationary coils, A r , M and C, Fig. 29, and a moving system com- prising an iron armature, A, rigidly attached to a shaft, S, suitably pivoted and mounted in bearings. A pointer, B, is also attached to the shaft S. The moving system is balanced and is not subjected to any restraining 418 ALTERNATING-CURRENT MACHINES. force, such as a spring or gravity control. The axes of the coils N and M are in the same vertical plane, but 90 degrees apart, while the axis of C is in a horizontal plane. The coils A x and ikf are connected in " split phase " rela- tion through an inductive resistance P and non-inductive resistance Q, and these two circuits are paralleled across the bus-bar terminals 3 and 4 of the synchroscope. Coil C is connected through a non-inductive resistance across the upper or machine terminals 1 and 2 of the synchroscope. In operation, current in the coil C magnetizes the iron core carried by the shaft and the two projections, marked A and " Iron Armature" in Fig. 29. There is, however, no tendency to rotate the shaft. If current be passed through one of the other coils, say M, a magnetic field will be pro- duced parallel with its axis. This will act on the projections of the iron armature, causing it to turn so that the positive and negative projections assume their appropriate position in the field of the coil M. A reversal of Pointer -B n ' ' ' I Fig. 29. the direction of the current in both coils will obviously not affect the posi- tion of the armature ; hence alternating current of the same frequency and phase in the coils C and M cause the same directional effect upon the armature as if direct current were passed through the coils. If current lagging 90 degrees behind that in the coils M and C be passed through the coil jV, it will cause no rotative effect upon the armature because the maximum value of the field which it produces will occur at the instant when the pole strength of the armature is zero. The two currents in the coils M and N produce a shifting magnetic field which rotates about the shaft as an axis. As all currents are assumed to be of the same frequency, the rate of rotation of this field is such that its direction corresponds with that of the armature projections at the instants when the poles induced in them by the current in the coil C are at maximum value and the field shifts through 180 degrees in the same interval as is required for reversal of the poles. This is the essential feature of the instrument, namely, that the armature projections take a position in the rotating magnetic field which corresponds to the direction of the field at the instant when the projections are magnetized to their maximum strength by current in the coil C. If the frequency of the currents in the coils which produce the shifting field is less than that in the coil which magnetizes the armature, then the arma- ture must turn in order that it may be parallel with the field when its poles PARALLEL OPERATION. 419 are at maximum strength, consequently rotation of its armature indicates a difference in frequency, and the direction and rate of rotation show, respectively, which current has the higher frequency and the amount of the difference. Note on the Parallel Running* of Alternators. — There is little if any trouble in running alternators that are driven by water-wheels, owing to the uniform motion of rotation. With steam-engine driven ma- chines it is somewhat different, owing to more or less pulsation during a stroke of the engines, caused by periodic variations in the cut-off, which cause oscillations in the relative motion of the two or more machines, accompanied by periodic cross currents. Experiments have proved that a sluggish governor for engines driving alternators in parallel is more desi- rable than one that acts too quickly ; and it is sometimes an advantage to apply a dashpot to a quick-acting governor, one that will allow of adjust- ment while running. It is quite desirable also that the governors of engines designed to drive alternators in parallel shall be so planned as to allow of adjustment of speed while the engine is running, so that engines as well as dynamos may be synchronized, and load may be transferred from one machine to the others in shutting down. Foreign builders apply a bell con- tact to the same part of all engines that are to be used in this way, and throw machines together when the bells ring at the same time. These bells would also serve to determine any variation, if not too small, in the speed of the machines, and assist in close adjustment. Manufacturers do not entirely agree as to the exact allowance permissible for variation in angular speed of engines, some preferring to design their dynamos for large synchronizing power, and relatively wide variation in angular speed, while others call for very close regulation in angular varia- tion of engine speed, and construct their dynamos with relatively little syn- chronizing power. Dynamos of low armature reaction have large synchronizing power, but if accidentally thrown out of step are liable to heavy cross-currents. On the contrary, machines with high armature reaction have relatively little syn- chronizing power, and are less liable to trouble if accidentally thrown out of step. The smaller the number of poles the greater may be the angular variation between two machines without causing trouble, thus low frequencies are more favorable to parallel operation than high ; and this is especially so where the dynamos are used to deliver current to synchronous motors or rotary converters. Specifications for engines should read in such a manner as to require not more than a certain stated angular variation of speed during any stroke of the machine, and this variation is usually stated in degrees departure from a mean speed. The General Electric Company states it as follows : — "We have . . . fixed upon two and one-half degrees of phase departure from a mean as the limit allowable in ordinary cases. It will, in certain cases, be possible to operate satisfactorily in parallel, or to run synchronous apparatus from machines whose angular variation exceeds this amount, and in other cases it will be easy and desirable to obtain a better speed con- trol. The two and one-half degree limit is intended to imply that the max- imum departure from the mean position during any revolution shall not 2i exceed ^ of an angle corresponding to two poles of a machine. The angle ooO of circumference which corresponds to the two and one-half degrees of phase variation can be ascertained by dividing two and one-half by one-half the number of poles ; thus, in a twenty-pole machine, the allowable angular 2i variation from the mean would be -^ = .25 of one degree." Some foreign builders of engines state the conditions as follows : Calling N the number of revolutions per minute, the weight of all the rotary parts of the engine should be such that under normal load the variation in speed dur- . ,. Nmax. — Nmin. ... , 1 „ 1 mg one revolution — will not exceed — - • Some state — - • N average 250 200 Oudin says : " The regulation of an engine can be expressed as a percent- age of variation from that of an absolutely uniform rotative speed . A close solution of the general problem shows that 1£° of phase displacement cor- I 420 ALTERNATING-CUKRENT MACHINES. responds to a speed variation, or *' pulsation," with an alternator of two n poles, as follows : — 2 75% In the case of a single cylinder or tandem compound engine — ■ A cross compound . 5.5% A working out of the problem also shows . . . that no better results are obtained from a three-crank engine than a two-crank. The Westinghouse Company designs its machines with larger synchro- nizing effect by special construction between poles, and allows somewhat larger angular variation, stating it as follows: The variation of the fly- wheel through the revolution at any load not exceeding 25% overload, shall not exceed one-sixtieth of the pitch angle between two consecutive poles from the position it would have if the motion were absolutely uniform at the same mean velocity. The maximum allowable variation, which is the amount which the armature forges ahead plus the amount which it lags behind the position of absolute uniform motion is therefore one-thirtieth of the pitch angle between two poles. The number of degrees of the circumference equal to one-thirtieth of the pitch angle is the quotient of 12 divided by the number of poles. The cross currents of alternators can be shown by reference to Fig. 30, Fig. 30. which represents the E.M.F. vectors of two alternators which have swung apart in phase due to any cause, such as variation in speed of their prime movers or fluctuations of speed during a revolu- tion. Let O—A = E.M.F. vector of alternator A. O—B — E.M.F. vector of alternator B. As drawn, the vectors are displaced in phase by the angle 0. When these alternators are con- nected in multiple there will be acting between them the E.M.F. A — B, or drawn to the center point O, the E.M.F. O — B. This E.M.F. acts through the two armatures in series, the circuit being a — b — c — d, (Fig. 31); the current result- ing is equal to the volts O — D divided by the im- pedance of the two armatures in series, which is equal to V(/? a + J? 6 )2 + (2 7r/Za + 2 nfLb)* where Ra and Bb = the resistance of the two al- ternator armatures respectively and La and Lb their inductances. Since in such a circuit the proportion of inductance is greater than tht, resistance, the current flowing from the E.M.F. O — D is lagging a large amount as shown by the line O — C. Hence the E.M.F. 's O — A and — B Fig. 31. Two Alterna- tors Connected in Multiple. ALTERNATING-CURRENT MOTORS. 421 of the alternators proper are in phase approximately with this cross current and hence under such conditions as the figure indicates there will be an ex- change of energy (since E.M.F. and current are in phase) which is what actually happens, thus tending to bring the two alternators together in phase. Fig. 32 shows the vectors of two alternators A and B in phase but the C ->A Fig. 32. E.M.F. O — A smaller than the other, — B, due, for instance, to the field of one being weaker than that of the other. In this case there is a difference of — D volts to act through the armatures of the two alternators in series, as in Fig. 31. As shown in Fig. 32 the current from this E.M.F. O — D lags 90° and is indicated by the vector O — C. This current is, how- ever, 90° away from the E.M.F.'s O — A and O — B of the machine proper and hence does not represent an exchange of energy ; therefore, it has no tendency to bring the machines together or increasing the dephasing. Synchronizing*. It is plain from the foregoing that to connect an idle alternator in parallel with one or more already in use : Excite the fields of the idle machine until at full speed the indicator shows bus-bar pressure, or the pressure that may have been determined on as the best for connecting the particular design of alternator in circuit. Connect in the synchronizer to show when the machines are in step, at which point the idle machine may be connected to the bus bars. The load will now be unequally divided, and must be equalized by increasing the driv- ing-power of the idle dynamo until it takes on its proper part of the load. Very little control over the load can be had from the field rheostats. To disconnect an alternator from the bus-bars : Decrease its driving power slowly until the other machines have taken all the load from it, when its xiain switch may be opened and the dynamo stopped and laid off . ALTERWATIMG-C1JRREKT MOTORS. The single-phase alternating-current motor has been quite well developed luring the last few years, but it has as yet come into rather limited use. The polyphase motor has come into very general use, its relative simplicity oeing a strong feature. Only the most elementary formulas will be given here, and the reader is referred to the numerous books treating on the subject ; among others, S. P. Thompson, Steinmetz, Jackson, Kapp and Oudin. Following is a statement of the theory of the polyphase motor, condensed from a pamphlet of the Westinghouse Electric and Manufacturing Com- pany. i 422 ALTERNATING-CURRENT MACHINES. Elementary Theory of the Polyphase Induction Motor. If a horse-shoe magnet be held over a compass the needle will take a posi- tion parallel to the lines of force which flow from one pole to the other. It is perfectly obvious that if the magnet be rotated the needle will follow. If a four-pole electromagnet be substituted for the horse-shoe, and current be made to flow about either one of the sets of poles separately, the needle will take its position parallel with the lines of force that may be flowing, as will be seen by the following figures. Fig. 33. Fig. 34 If the two sets of poles are excited at the same time by currents of equal strength, then the needle will take its position diagonally, half way be- tween the two sets of poles, as will be seen by the following diagram. It is now easily conceivable that if one of these currents is growing stronger while the other is at the same time becoming weaker, the needle will be at- tracted toward the former until it reaches its maximum value, when if the currents are alternating, the strong current having reached its maximum begins to weaken, and the other current having not only re- versed its direction but begun to grow strong, attracts the needle aAvay from the first current and in the same direction of rotation. If this process be continually repeated, the needle will continue to re- volve, and its direction of rotation will be determined by the phase relation of the two currents, and the direction of rotation can be reversed by reversing the leads of one phase. If the compass needle be replaced by an iron core wound with copper conductors, secondary currents will be induced in these windings, which will react on the field windings, and rotation will be produced in the core just as it was in the compass needle. Two cranks at right angles on an engine shaft are analogous with the quarter-phase motor, and three to the three-phase motor, which depends on the same principle for its working. Fig. 35. Theory of the Polyphase Induction motor. Condensed from C. P. Steinmetz. The following names and symbols are used for designating the parts and properties of the induction motor : — THE INDUCTION MOTOR. 423 Statorzz stationary part, nearly always corresponding to the field. Rotor = rotating part, corresponding to the armature of the direct-current motor. Analytical Theory of Polyphase Induction Motor. Let r = resistance per circuit of primary, r t = resistance per circuit of secondary , being reduced to primary system by square of the ratio of turns. Let p = number of poles, x = reactance of primary, per circuit, x x z=z reactance of secondary, per circuit, reduced to primary system by square of the ratio of turns. Let s = per cent of slip, / = current per circuit of primary, E z= applied E.M.F. per circuit, Z=. impedance of whole motor per circuit, T=z torque between the stator and rotor, f = frequency of applied E.M.F. Let the primary and secondary consist of m circuits on an m phase system. n = primary turns per circuit, *&i= secondary turns per circuit. Let a = — ratio of transformation. «i Then sE /(neglecting ex. current) =• V / , Na , 9/ , — r, » B v (r t -f s r) 2 + s 2 (x -f- x t )* Toraue T- mpr^s ^ ""4 7r/[(r 1 + sr)2 + s 2 (^i+^) 2 ] n mr x E 2 s (1 — s) Power = ■ — " (r x -\- s r)2 + s 2 (x x + x)* mpE 2 Max. torque = g -^^=== , Max. power = 2[r " r f +Jg] at the slip . = v ^< E Starting current = i =2 — » JO , ,. mpE 2 n Starting torque = ~^y X ~ Note that the maximum torque is independent of secondary resistance r t . and thus the speed at maximum torque depends on the secondary resistance Current at maximum torque is also independent of secondary resistance. The maximum torque occurs at a lower speed than the maximum output. A resistance can be chosen that when inserted in the secondary, the maximum 424 ALTERNATING-CURRENT MACHINES. torque will be obtained at starting ; that is, the speed at which maximum torque occurs can be regulated by the resistance in the rotor. Ro. 36. Torque curves for Polyphase Induction Motor. Curves 1, 2, and 3 show the effect of successive increases of rotor resist- ance, rotor run on part of curve a—b ; for here a decrease of speed due to load increases the torque. Speed of Induction Motor. — The speed or rotating velocity of the magnetic field of an induction motor depends upon the frequency (cycles per second) of the alternating current in the field, and the number of poles in the field frame, and may be expressed as follows : — r.p.m. = revolutions per minute of the magnetic field, p = number of poles, /= frequency ; then r.p.m. ~ 120 - P The actual revolutions of the rotor will be less than shown by the formula, owing to the slip which is expressed in a percentage of the actual revolu- tions ; therefore the actual revolutions at any portion of the load on a motor will be r.p.m. x slip due to the part of the load actually in use. actual speed = r.p.m. (1 — % of slip.) The following table by Wiener, in the American Electrician^ shows the speeds due to different numbers of poles at various frequencies. Speed of Rotary field for Different Number* of Poles and for Various frequencies. O Speed of Revolving Magnetism, in Revolutions per Minute, when Frequency is : a ° 25 30 33£ 40 50 60 66§ 80 100 120 125 133$ 2 1500 1870 2000 2400 3000 3600 4000 4800 6000 7200 7500 8000 4 750 900 1000 1200 1500 1800 2000 2400 3000 3600 3750 4000 6 500 600 667 800 1000 1200 1333 1600 2000 2400 2500 2667 8 375 450 500 600 750 900 1000 1200 1500 1800 1875 2000 10 300 360 400 480 600 720 800 960 1200 1440 1500 1600 12 250 300 333 400 500 600 667 800 1000 1200 1250 1333 14 214 257 286 343 428 514 571 686 857 1029 1071 1143 16 188 225 250 300 375 450 500 600 750 900 938 1000 18 167 200 222 267 333 400 444 533 667 800 833 889 20 150 180 200 240 300 360 400 480 600 720 750 800 22 136 164 182 217 273 327 364 436 545 655 682 720 24 125 150 167 200 250 300 333 400 500 600 625 667 THE INDUCTION MOTOR. 425 Slip. —The slip, or difference in rate of rotation between rotating field and rotor, is due to the resistance opposed to rotor current. Slip varies from 1 per cent in a motor designed for very close regulation to 40 per cent in one badly designed, or designed for some special purpose. Weiner gives the following table as embodying the usual variations : Slip of Induction Motors. Capacity of Motor, H.P. Slip, at full load, per cent. Usual limits. Average. , 20 to 40 30 1 10 ■ 1 30 20 1 10 ■ « 20 15 1 8 ■ « 20 14 2 8 « * 18 13 3 8 * * 16 12 6 7 ■ 1 15 11 n 6 ■ 1 14 10 10 6 « ■ 12 9 15 6 « ♦ 11 8 20 4 ■ « 10 7 30 3 ■ * 9 6 50 2 « " 8 5 75 1 7 4 100 « 6 3.5 150 • 5 3 200 ♦ 4 2.5 300 • 3 2 Core of Stator and Rotor. — Both the field-frame core, or Stator > and the armature core, or Itotor, are built up of laminated iron punchings in much the same manner as are the armature cores of ordinary dynamos. The windings in both cases are laid in slots across the face of either part, and for this reason both parts are punched in a series of slots or holes for the reception of the windings. The following cuts, taken from the " Ameri- can Electrician," show the usual form of slots used. Figs. 37 and 38. Forms of Punchings of Induction Motors. The number of slots in the stator must be a multiple of the number of poles and number of phases, and Weiner gives the following table, in the " Ameri- can Electrician," as showing the proper number to be used in various cases, both for two- and three-phase machines. In practice the number of poles is determined by the speed required and the available frequency ; then the number of slots is so designed as to be equally spaced about the whole innei periphery of the stator. 426 ALTERNATING-CURRENT MACHINES. Somber of Slots in Field-Frame of Induction Motors. Capacity of Motor. Number of Poles. Slots per Pole. Slots per Pole per Phase. Two-Phase. Three-Phase. £ H.P. to 1 H.P. 4 to 8 3 4 ? 1 1 H.P. to 1 H.P. 4 to 6 5 6 ? 2 4 to 10 5 6 9 2 2 HJP. to 5 H.P. 4 to 6 7 8 9 ? 3 6 H .P. to 50 H.P. 6 to 12 7 8 9 4J 3 4 to 8 10 11 12 5 4 10 to 20 7 8 9 3 50H.P.to200H.P. 8 to 12 10 11 12 13 5 4 6 to 10 14 15 16 7 ? 5 The number of slots per pole per phase in the rotor must be prime to that of the stator in order to avoid dead points in starting, and to insure smooth running, and commonly range from 7 to 9 times the number of poles, or any integer not divisible by the number of poles, in the squirrel cage or single conductor per slot windings. The proper number of slots may be taken from the following table by Wiener ; THE INDUCTION MOTOR 427 Number of Rotor Slots for Squirrel-Cagre Induction Motors up to 5 H.I*. Capacity. Number of Poles, p. Limits of Slots, Number 7 p. to 9 p. Number of Rotor Slots. 4 6 8 28 to 36 42 " 54 56 " 72 29, 30, 31, 33, 34, 35, 37. 43, 44, 45, 46, 47, 49, 50, 51, 52, 53. 57,58,59,60,61,62,63,65,66,67, 68, 69,70,71. In large machines, where there is more than one conductor in each slot and in which the winding is connected in parallel, the number of slots in the rotor must be a multiple of both the number of phases and the number of pairs of poles. The following table gives numbers of slots for various field-slots : Number of Rotor-Slots for Induction Motors of Capacities over 5 H.P. Number of Field-Slots per Pole. Number of Rotor-Slots. (n« = Field-Slots.) = number of 8 fn,. or § n*. 9 § n 8 . 10 fn«. » fn*. 12 |iu. " fn«. 14 fn«. " §n«. 15 f n«. " fn 8 . 16 |n 8 . " |n«. Flux Density. — This must be settled for each particular case, as it will be governed much by the quality of iron and the particular design of the motor. Hysteresis loss increases as the 1.6 power of the flux density ; and eddy current losses are proportional to the square of the density and also to the square of the frequency. The following table shows practical values : flux-Densities for Induction Motors. (Wiener.) Flux-Density, in Lines of Force per Square Inch. Capacity of Motor, H.P. For Frequencies from 25 to 40. For Frequencies from 60 to 100. For Frequencies from 120 to 180. Practical Values. Aver- age. Practical Values. Aver- age. Practical Values. Aver- Age. i 12000 to 18000 15000 " 25000 18000 " 32000 15000 20000 25000 10000 to 15000 12000 " 18000 15000 " 25000 12500 15000 20000 7000 to 11000 7500 " 12500 8000 " 17000 9000 10000 12500 428 ALTERNATING-CURRENT MACHINES Flux-Densities for Induction Motors —(Continued). Flux-Density in Lines of i orce per Square Inch. Capacity of For Frequencies For Frequencies For Frequencies Motor, from 25 to 40. from 60 to 100. from 120 to 180. H.P. Practical Aver- Practical Aver- Practical Aver- Values. age. Values. age. Values. age. 1 20000 to 40000 30000 18000 to 32000 25000 9000 to 21000 15000 2 25000 " 45000 35000 20000 " 40000 30000 10000 " 25000 17500 5 30000 " 50000 40000 25000 " 45000 35000 11000 " 29000 20000 10 40000 " 60000 50000 30000 " 50000 40000 12500 " 32500 22500 20 50000 " 70000 60000 35000 " 55000 45000 15000 " 35000 25000 50 60000 " 80000 70000 40000 " 60000 50000 17500 " 37500 27500 100 70000 " 90000 80000 45000 " 65000 55000 20000 " 40000 30000 150 80000 " 100000 90000 50000 " 70000 60000 25000 " 45000 35000 200t 90000 " 110000 100000 60000 " 80000 70000 30000 " 50000 40000 t And over. In the earlier induction motors it was considered the most efficient method to connect the driving current to the revolving part or rotor ; and as it is highly important that the number of windings on the rotor be prime to that of the stator, Fig. 39 shows a winding with an odd combination of con- ductors, being 51, or three times 17. The stator windings would then be bars, con- nected at either end to a heavy copper ring, this forming a sort of " squir- rel-cage." In the modern ma- chines the winding shown would be in coils on the stator, the three ends being carried to terminal blocks on the outside of the machine instead of to rings as shown, and the " squirrel- cage " would then be placed on the rotor and be made of bars as men- tioned. Starting* and It em- ulating- Devices. — Small induction motors, up to about 5 h. p. capa- city, are started by closing the circuit directly to the motor. In large ma- chines this would not be safe, as the rotor is standing, and would act in a lesser degree as the short-circuited secondary of a static transformer, and cause a heavy rush of current. Resistance in Rotor. —This is a favorite method with the General Electric Company. A set of strongly constructed resistances is secured inside the rotor ring, and so arranged with a lever that they may be closed orshoU-circuited after the motor has reached its full speed. These resist- Fig. 39. THE INDUCTION MOTOR. 429 ances are in the armature circuits. In order to give maximum starting torque total armature resistance should be r, = V r2 + (a?/ + y)l Where r x — rotor resistance per circuit reduced to field system, x x := rotor reactance per circuit reduced to field system. r = resistance per field circuit. y =z reactance per field circuit. This method serves the double purpose of keeping down the starting cur- rent and increasing the starting torque. Resistances in Stator. — Resistance boxes may be connected in the circuits supplying induction motors ; three separate resistances in three- phase circuits, and two separate resistances in two-phase circuits. They must be all connected in such a manner as to be operated in unison. Under these conditions the pressure at the field terminals is reduced, as is of course the starting current and the starting torque. In order to start a heavy load, under this arrangement, a heavy starting-current is necessary. Compensators or Auto-Transformers. —This method is greatly favored by the Westinghouse Electric Manufacturing Company, and is used extensively by the General Electric Company. It consists of connecting an impedance coil across the line terminals, the motor being fed, in starting, from some point on the winding where the pressure is considerably less than line pressure. This avoids heavy drafts of current from the line, thus not disturbing other appliances attached thereto, but as regards starting current and torque has the same effect as resistances directly in the line ; that is, greatly reduces both. Rotor Winding's Commutated.— In this arrangement all or a part of the rotor windings are designed to be connected in series when starting, and are thrown in parallel after standard speed is attained. Another design has part of the conductors arranged in opposition to the remainder in starting, but all are thrown in parallel in regular order when running at standard speed. These commutated arrangements have not been much used in the United States. The single-phase alternating-current motor brought out by the Wagner Electric Manufacturing Company of St. Louis, is, in mechanical construc- tion, similar in many respects to the two and z three-phase motors on the market. A field is built up of iron plates very much like A of Fig. 40, and an armature core is also built up from iron plates very much like B. The field is wound with so-called pan-cake coils threading through the slots of the punchings, as shown at C, thus producing a magnetic pole of intensity, varying from a maximum along the radius x — y to zero along the radius x — z. The armature core is wound with an ordinary direct- current progressive winding, connected up to a commutator in exactly the same fashion as is the direct-current motor winding. Fig. 40. -^ ne commutator of this armature is so designed that it may be completely short-circuited by intro- ducing into it a short-circuiting circle of copper segments. When so short-circuited, the winding affords a substitute for the squirrel-cage form of winding, above described, differing from the squirrel cage, in that instead of currents being able to select paths for themselves, they are restricted to flowing in paths afforded by the individual coils. The operation of this motor, as stated, is based wholly upon the principle that an induction motor with a completely short-circuited armature will, when up to the running speed, operate on single-phase current supply in exactly the same manner as does a two or three-phase motor with two or three- phase current supply. The armature winding is short-circuited through carbon brushes bearing upon the commutator surface, and the currents flowing in it are generated by induction from the field. These currents flow out through the carbon rushes either into an outside resistance box, or where a direct short cir- 430 ALTERNATING-CURRENT MACHINES. cuit of the brushes is provided, out through one brush and back into the armature through the other. By the shifting of the brushes on the com- mutator surface, they are forced to take such position relative to the mag- netic poles of the field, that repellant action between them and the poles of the fields is eifected, and rotation results. When running speed is attained, the brushes are no longer required and the armature winding is completely short-circuited, as stated. The short-circuiting ring is made up of small copper links, which links, being in turn mounted upon a short- circuiting band, are thrown into the annular opening in the commutator and by making close contact with the individual segments, produce a very effec- tive short-circuiting of the entire armature winding. In the operation of the motor, it is very advantageous to have this short-circuiting operation performed either at or slightly below the running speed, so these motors are built with an automatic device for performing this operation. This device consists of a set of governor weights acting against a spiral spring. The centrifugal action of the weight will, at the proper speed, force the short- Fig. 41. Cross Section of Wagner Motor. circuiting links into the commutator, against the action of the spring. At the same instant and by the same means, the brushes bearing upon the commutator are thrown off. Fig. 41 shows a view in cross-section of the Wagner motor, and the dia- gram, Fig. 42, shows the elementary connections of the same ; the first diagrammatic motor being shown as in the starting condition, and the diagram at the right showing the condition of the armature after it has attained full running speed and the commutator is short-circuited. S¥ICHR0^01J§ MOTORS. Alternators are convertible into motors ; and one alternator will run in synchronism with another similar machine after it is brought to the same speed, or, if of unlike number of poles, to some multiple of the speed of the driven dynamo, provided the number of pairs of poles on the motor is SYNCHRONOUS MOTORS. 431 ORDIN**a»J Rheostat .'/Sometimes i I // : pop STARTING COMMUTATOR CONNECTIONS •REPULSION MOTOR Rotot Winding Stator Winding Commutator. >rt Circuiting Links and ng automatically introduced mo Cjra mutator: Carbou SruthM mu!uneou»_'y removed from. iCommuutpr.. RUNNING COMMUTATOR CONNECTIONS •INDUCTION MOTOR Fig. 42. Connections of Wagner Single-Phase Motor. divisible into the multiple. Such motors will run as if geared to the driven dynamo even up to two or three times its normal full torque or capacity. Single-phase synchronous motors have no starting-torque, hut synchronous motors for multiphase circuits will come up to synchronism without much load, giving about 25 % starting-torque, starting as induction motors, with the d. c. field open. When connected to lines on which are connected induction motors that tend to cause lagging currents and low-power factor of the line, over excita- tion of the synchronous motor fields acts in the same manner as a condenser introduced in the line, and tends to restore the current to phase with the impressed E.M.F., and therefore to do away with inductive disturbances. It is necessary to provide some source from which may be obtained con- tinuous current for exciting the fields of the synchronous motor ; and this is oftenest done by the use of a small d. c. dynamo belted from the motor- shaft, the exciting current not being put into use until the motor armature reaches synchronism. In starting a synchronous motor the field is open-circuited, and current is turned on the armature. In practice, field coils are connected in various ways to obviate the dangers of induced voltage, and a low resistance coil similar to the series winding of the d. c. machine is sometimes so arranged on the field poles as to give the necessary reaction for starting. Another way is to use a low-pressure excitation, and therefore few turns on the field coils ; also the field coils are " split up " by a switch at starting. The field excitation is thrown on after the rotating part approaches synchronism, which may be indicated by a lamp or other suitable device at the operating switchboard. Considerable care must be exercised in the use of synchronous motors, and their best condition is where the load is quite steady, otherwise they intro- duce inductive effects on the line that are quite troublesome. The field of such a motor can be adjusted for a particular load, so there will be neither leading nor lagging current, but unity power factor. If the load changes, then the power factor also changes, until the field is readjusted ; if the load ALTERNATING-CURRENT MACHINES. has been lessened the current will lead, and if it increases the current will lag. If induction motors are connected to the same line, with a synchro- nous motor that has a steady load, then the field of the synchronous motor can he over-excited to produce a leading current, which will conteract the effect of the lagging currents induced by the induction motors. If two or more synchronous motors are connected to the same circuit, and the load on one of them is quite variable, and its field is not changed to meet such changing conditions, a pumping effect is liable to take place in the other motors, unless especial provision has been made in the design of the motors to prevent it. It is only necessary to arrange one of the motors of the number for preventing this trouble, but better to make all alike. A copper shield between pole- pieces, and covering a portion of the pole-tip, will prevent the trouble ; and the Westinghouse Electric and Manufacturing Company use a heavy copper strap around each pole-piece, with a shoe covering part of the pole-tip in the air-gap. Theory of the Synchronous motor. Let R = resistance of whole circuit, L = inductance of whole circuit, E x = generator E.M.F., E 2 = motor E.M.F. E<> resultant Fig. 43. Take the origin at 0. Let E represent maximum value, e = instantaneous value, e x = E x sin (o> t -f- <£), e 2 = E 2 sin (), where w = 2 tt/, and/ number of complete cycles per second e = E sin (w t — \p), where i// = angle of lag of E with respect to the origin. E Q 2 = EJ + E 2 2 -f- 2 E X E 2 cos 2 , For Eo > E x , E leads, E 2 , r» j? tan rjf = 2 * tan , (E. + E 2 ) sin »// = V *1, — - cos are known, Energy shifts the origin by the angle \]/. e x = E x sin (a> t + -\- \f/), e 2 = E 2 sin (ut t — -f" #)• Now THE SYNCHRONOUS MOTOR. E n 433 / = " V,R2 + ta ,2£2 and / lags behind E by the angle 5 where LP tan 5 z=i — =■ • K By introducing the angle \f/ we are referring the E.M.F.'s of both machines to the zero point of the resultant wave as origin. In general 1 f T > M E1 = ji f « idt=— cos®, P = -7* where Let Then P = the power in watts, and $ =. lag or lead of / with respect to E, E and / are maximum values, T= - » or the periodic time. n P x as power given to the circuit by the generator, P 2 =. power absorbed from the circuit by the motor, XT e, idt— — l , E ° cos ( + i/r + 8) [i = /sin (P 1 t — «)], 2 Vm + i^L* * = ^ E 2 Vi22_|_ w 2£2 sin 6 as - [cos (4> + \p) cos 6 — bin ($ + \fj) sin 8], cos 5 = - " V.R2 _J_ JJl ^2 ^R2 4. 0,2 2,2 ''' Pl = 2(R^+!iL*) {^ co s(<^ + ^)-Z wS in ($ + *)}, and substituting — for -f- <£ we get { R cos (0 — «J0 + 2/ w sin ( ig} P 2 = An angle <£i is introduced such that R sin 2^ — Vi22 _|. w 2 £2 - » and cos 2 <£i : ' Vtf2 + w 2 £2 434 ALTERNATING-CURRENT MACHINES. Substitute in P 2 1 P* = and [e x Vi22 + w 2i2 S in {2 + 2 + <*>': that is, the 4i sine term " = unity. P 2 is positive provided E l R — > - i J£ 2 Vtf2 + W 2£2 FIELD Fig. 44. which shows that it is possible to have E 2 greater than E x if there is the proper ratio of resistance and reactance in" the circuit. Now, if we plot from an actual motor the armature current and the field excitation we get a curve shown in Fig. 44. This shows that the armature current varies with the excitation for a given load. The flatter curves are for increase of load. Point a shows under excitation, b shows over excitation, c shows the excitation which makes the power factor unity ; it is well from the point of stability of operation to slightly over excite, and this makes E 2 >E, , and also counteracts the inductive drop m the line, thus showing that the action of an over excited synchronous motor is similar to a condenser. Graphical treatment. Eg = generator E.M.F. Em = motor E.M.F. Eo — resultant E.M.F. Io = resultant current. O I g z=z projection of Io on Eg. O Im — projection of Io on O Em. O Ig . OEg — ug = energy given up by the generator. O Im . Em = wm = energy absorbed by the motor from the cir- cuit, com is negative, which shows that com is the motor, because it is taking energy from the circuit ; and similarly 4UOTORM. These are of two styles, one for changing direct current of one voltage into direct current of a different voltage, and usually called in America motor-generators; the second class changes alternating current into direct current or vice versa, the voltage not being changed excepting from alter- nating Vmean 2 values to direct-current values equal to the top of the alternating wave ; these latter machines are now called rotary converters, and are largely used. DIRECT-CURRENT BOOSTERS. 435 Dynamotors are now largely used in telegraph offices for reducing the pressure of the supply current to voltages suitable for use in telegraphy and for ringing and charging generators in telephone offices. Theory. Let E = voltage at motor terminals. e = voltage at generator terminals. / zz current in motor armature. B zz resistance of motor armature. Nc = number of conductors in motor armature. L zz current in generator armature part. ijL zz resistance of generator armature part. iv c ,= number of conductors in generator armature part. Nc — — zz k =. coefficient of transformation. E zz induced E.M.F. in motor part. E x zz induced E.M.F. in generator part. E = r.p.s. X Nc X . ^i = r.p.s. X Nc Y X <*>. E —E — RI. E 1 = e-\- EJ V ke = E = EI — krJ v If it be assumed that losses by hysteresis and eddy currents be negligible, or that EI zz E X I Y whence I x zz kl, then E k ' «,+!)/, Such machines run without sparking at the commutator, as all armature reactions are neutralized. !>fi 11 ■;< T-C I ItJREUT IftOOfcTXIlS. This is a tvpe of motor generator much in use for raising or lowering the pressure on long feeders on the low-pressure system of distribution, and is to be found in most of the larger stations of the Edison companies. It is also much used in connection with storage-battery systems in charging cells. The " booster " consists of a series generator driven by a motor direct con- nected to its armature shaft. The terminals of the generator are connected in series with one leg of the feeder ; and it is obvious that the current in the feeder will excite the series field just in proportion to the current flowing, provided the design of the iron magnetic circuit is liberal enough so that the field is way below saturation (on the straight part of the iron curve way below the knee). As the armature is being independently rotated in this field, it will produce an E.M.F. approximately in proportion to such excitation, which E.M.F. will be added to that of the feeder or will oppose that E.M.F., ac- cording as the terminal connections are made. On three-wire systems two generators are direct connected to one motor, and for convenience on one bed-plate. . . Such a booster can be so adjusted as to make up for line loss as it in- creases with the load. One danger of a booster that is not always taken into account is. that if the shunt of the driving-motor should happen to open, or, in fact, anything should happen to the driving-motor that would result in its losing its power, the generator would immediately become a series motor, taking current from the line to which it is connected, and by its nature would reverse in direction of rotation, and increase in speed enormously, and if not discon- nected from its circuits in time would result in a complete wreck of the machine. It is always safest to have the generator terminals connected to their line through some automatic cut-out, so arranged that should the shunt break, as suggested, it would actuate the device, and automati- cally detach the booster from the circuit before harm could be done, 436 THE ROTARY CONVERTER. ROTARY CONVERTERS. A rotary converter is the name given to a machine designed for changing alternating currents into direct currents. If the same machine be used inverted, i. e., for changing direct currents into alternating, it is some- times known as an inverted converter. Again, if the same machine be driven by outside mechanical power, both alternating and direct currents may be taken from it, and it then becomes known as a double current generator. Theoretically the rotary converter is a continuous current dynamo with collector rings added, which are connected by leads to certain parts of the armature windings, sometimes at the commutator segments. In the following figure, which represents in diagram the single-phase rotary converter, the collector rings r and r± are connected by leads to dia- metrically opposite segments or coils of the armature at c and c v It is obvious that as the armature revolves the greatest difference of potential between the rings, or maximum E.M.F., will be at the instant the segments c and Cj pass under and coincide with the brushes B and B 1 ; and this E.M.F. will decrease as the rotation continues, until the lowest E.M.F. will occur when the segments c and c 1 are directly opposite the centre of the pole-pieces P and P v SINGLE PHASE ROTARY CONVERTER Fig. 46. The maximum alternating E.M.F. will be equal to the direct-current voltage at the brushes B and B lt and if the machine be designed to produce a sinuso idal curve of E.M.F., then the alternating E.M.F., that is, the Vmean 2 or effective E.M.F., will be, e = JL=.707 J£, V2 where e = Vmean 2 value of the alternating E.M.F. , and E = direct-current voltage between brushes. In a bipolar machine the frequency = r.p.s., and in a machine with p poles the frequency will be — r.p.s. Neglecting losses and phase displacement the supply of alternating cur- rent to the rings must be / V2 = 1.414 / where / is the direct-current output. If, as shown in Fig. 47, another pair of rings be added, and connected to points on the winding at right angles to the first, then another and similar THE ROTARY CONVERTER. 43? E.M.F. will be produced, but in quadrature to the first. The E.M.F. will be the same for each phase as in the single-phase connection previously shown, and still neglecting phase displacement and losses the current will be for each of the two phases -£= = .707 /. V 2 TWO PHA8E, OR QUARTER PHASE ROTARY CONVERTER Fig. 47. If three equidistant points on the armature windings be connected to three rings, as shown in the following diagram, a three-phase converter is produced. THREE PHASE ROTARY CONVERTER Fig. 4& As the connections of a three-phase rotary are always delta, the E.M.F.'s as compared with the continuous current E.M.F. E have the following value : E Voltage between collector ring and neutral point e = — — = .354 E. 2 V 2 Voltage between collector rings e 1 = E^3 Alternating current input : 2V^ IE _ 2/V~2 : 3e" 3 .612 E. : .943 1. Steinmetz, in the Electrical World of Dec. 17, 1898, gives the following tables of values of the alternating E.M.F. and current in units of direct current. 438 THE ROTARY CONVERTER. © fl 1:1* 8 ii CO 1 |CN| > « CN ! d *« ICN > g s 6 > cp CO a ii 1 CO 1 | £ A ilCN H Pa 1 CN cp "* 8 S CN CN 5 jjjj 55 ,d Pi II ii ii |CN| ii |CN| j t- o 6 (0 CO ii to t>; «3 Is H e3 ,d ii 1H ICN ii -,|i 11 r-l ICN & ® •«*< i CN CO 3 CP CO c3 ii ii N || ,d ip ICO|| > rfc > r > > 1 CN 1 CN CN 1 CN 1 CO S 5 "<* "*< cp 6 t^ ^ rf| CO II ||CN - 1 > 1 CN 'bio d 33 03 ,d II ~\> II |CN > II > 03 3 o d «a S3 cp c (-1 3 ' H r-t *"• ^ o Q o O d J. ^ * 03 o oJ cp d d * cp CO CP^ § bo d * • H St; a 4a CP d CP OJ2 co <».S CO CP u 00 «P CP c« co *? =3 CP cp'd Sgg ^3 O M a ftoS O-H fl O 8 a k |> -5 4 THE ROTARY CONVERTER. 439 The values of E.M.F. and of current stated above are theoretical, and are varied in practice by reason of drop in armature conductors and phase displacement. In converting from a.c. to d.c, if the current in the rotary is in phase with the impressed E.M.F. , armature self-induction has little effect ; but with a lagging current, which may be due to under-excitation, the induced d.c. E.M.F. is somewhat reduced ; and if the machine be over- excited, thus producing a Leading current, the induced d.c. E.M.F. will be raised. The same is the case in converting from d.c. to a.c, the a.c. volts being down on a lagging circuit. The corrections for the theoretical ratios of voltages as shown are, first for drop in the armature ; and second, they have to be multiplied by the factors shown above. Steinmetz says that the current flowing in the armature conductors of a rotary is the difference between the alternating current input and the con- tinuous current output. The armature heating is therefore relatively small, and the practical limit of overload is limited by the commutator, and is usually far higher than in the continuous current generator. In six-phase rotaries the I 2 R losses of the armature are but 29 % of the regular I 2 R loss in the armature as used for d.c. dynamo. Kapp shows that width of pole-face has a bearing on the increase in out- put of a rotary converter over the same machine used as a continuous cur- rent dynamo. He compares the output of two converters, one in which the pole-face is two-thirds the pole distance, and another in which it is one- half the pole distance. In single-phase converters the output is not equal to that of the d.c. dynamo, and two- and three-phase machines are much different. He gives, in the following table, the percentage of d.c. output of what would be the output of the same machine used as a d.c. dynamo. Pole-width. § i rcos — 1 88% 81 73 63 95% Single-phase \ gj = ;§ ; ; ; ; ; ; ; ; ; ; LCos= .7 88 80 70 ( Cos — 1 138 128 117 144 Three-phase \ Cos = .9 (Cos= .8 137 126 »n ( COS — 1 167 160 144 170 T wo-or Cos- .9 167 four-phase { ^ = % ; v ; ; ; ; ; ; ; ; ; 153 To find the voltage required between collector rings on rotary con- verters, when T— number of turns in series between collector rings, * = flux from one pole-piece into the armature, /■=. cycles per second, E = required E.M.F. Then For single-phase and two-phase machines E = 2.83 T /*$ 10- 8 , For three-phase machines E — 3.69 T f S> lO- 8 . 440 THE ROTARY CONVERTER. The single-phase rotary lias to be turned up to synchronous speed by some external power, as it will not start itself. The polyphase rotary will start itself from the a.c. end, but takes a tre- mendous lagging current, and therefore, where possible, it should be started from its d.c. side. The starting of rotaries that are connected to lines having lights also con- nected, should always be done from the d.c. side, as the large starting cur- rent taken at the moment of closing the switch will surely show in the lamps. Polyphase rotaries are sometimes started, as are induction motors, by use of a " compensator." In starting a rotary , the field circuit must be opened until synchronism is reached, after which it is closed. The d.c. side must also be disconnected from its circuit, as it is obvious that the current produced is alternating until synchronism is reached. Care must be taken to keep the field circuit closed when the d.c. side is connected in parallel with other machines, and the a.c. side open, or the armature will run away and destroy itself. As the change in excitation of the field of a rotary changes the d.c. voltage but little, and on the other hand produces wattless currents, the regulation of E.M.F. must be accomplished by some other method. This can be done by changing the ratio of the static transformer by cutting in and out turns as its primary, or by the introduction of self-induction coils in the a.c. leads to the rotary. The first introduces a complicated set of connections and contacts, but is unlimited in range. The second method seems especially suited for the purpose, but is some- what limited in range. Theoretically the action is as follows : Suppose the excitation to be low enough so that the current lags 90° behind the impressed E.M.F., the E.M.F. of self-induction lags 90° behind the current, and is therefore 180° behind the impressed E.M.F., and therefore in opposition to it. On the other hand, if the excitation is large, and produces a leading current of 90°, the E.M.F. of self-induction is in phase with the impressed E.M.F. and adds itself to it. Therefore, with self-induction introduced in the a.c lines, it is only necessary to vary the excitation in order to change the con tinuous current E.M.F. A rotary can thus be compounded by using shunt and series field, to maintain a constant E.M.F. under changes of load, the compounding taking place, of course, in the a.c. lines and not in the field of the machine, as usual in d.c. dynamos. In handling the inverted converter care must be exercised in starting it under load, as it is apt to run away if not connected in parallel with other alternators. If they are started from the d.c. side, and have lagging cur- rents flowing from a.c. side, this current will tend to demagnetize or weaker the fields, and the speed of the armature is liable to accelerate to the dan- ger limit. A lagging current taken from an inverted rotary, even after having reached synchronism, will cause an immediate increase in speed, and if enough lag- ging will cause an approach to the danger point. Running as a rotary, and converting from a.c. to d.c, the phase of the en- tering current has no effect on the speed, this being determined by the cycles of the driving generator, nor upon the commutation, simply influen- cing the heat in the armature and ratio of voltages slightly. Double-current generators are useful in situations where continuous cur rent can be used for a portion of the day and the current transferred througl the a.c. side to some other district for use in another portion of the day, thus keeping the machine under practically constant load. The size of double-current generators is limited by the size of the d.c. gen- erator that can be built with the same number of poles as a good alternator. The heating of the armature depends upon the sum and not the difference of the currents, as in the rotary, and the capacity is therefore no greater than a d.c. machine of the same total output. Automatic compounding of double- current generators is scarcely feasible in practice, and the field must be very stable, as the demagnetizing effect of the lagging a.c. currents tends to drop the excitation entirely. Such machine? run better separately excited. ROTARY CONVERTER WINDINGS. 441 €O^T£RTER U4H4TIKE WINDINGS. Two-Circuit Winding* for Two-Phase Horary Transforms rs. The following diagram shows the connections of the four rings to the dif- ferent sections of the armature. The connections are made at the commu- tator segments at four points, although there are six poles. Fig. 49. Two-Circuit Winding- for Three -Phase Rotary Transform ers. The following diagram shows the connections of the three collector rings to the continuous current winding of a six-pole dynamo. As in the last fig- ure, the rings are connected to points on the commutator at nearly equi< distant points. Fig. 50. 442 ROTARY CONVERTER CONNECTIONS. Note, — Connection of Transformer* and Rotary Converters. In the use of rotary converters, two or more of these machines are some- times connected in multiple to the secondary of the transformers, and their direct current leads then conducted in multiple to a common bus-bar circuit, as shown in Fig. 51. UfcSHfcNAr.OK I — ^ — I rQQOmOQbOQQW QfflOOOQOOOO.OWi TRANSFORMER 00006] fffiW] KWlSt fOi rOi mm mm 4 ROTARY ROTARY Fig. 51. Fig. 52. With the'above connections, currents are often formed in the rotaries that . disturb the point of commutation, and it becomes practically impossible to adjust the brushes so they will not spark. Rather than connect across in the above manner, it is better that each rotary have its own transformer, or at least its own secondary on the transformer, as shown in Fig. 52. Current Densities. Current leads from brushes to binding-posts, must be ample to produce no appreciable drop in voltage. The following table gives current densities, etc., for brush-holders, conductors, bolted joints, and switches. Average Current Densities for Cross Section and Contact Surface of Various materials. Material. Square Mils, per Ampere. Amperes per Square Inch. Cross section i Copper wire . . . Copper rod . . . Copper-wire cable . Copper casting . . Brass casting . . 500 to 800 800 " 1,200 600 " 1,000 1,400 " 2,000 2,500 " 3,300 1,200 to 2,000 800 " 1,200 1,000 " 1,600 500 " 700 300 " 400 Brush contact < Copper brush . . . Carbon brush . . . 5,700 " 6,700 28,500 " 33,500 150 " 175 30 " 35 Switch jaws Copper — copper . . Brass ; 10,000 " 15,000 J 20,000 " 25,090 67 " 100 40 " 50 Screwed contact J Copper — copper B '«* 2 3 4 5 6 7 8 S TIME IN HOURS CURVES 8H0WING RESULTS Duf TO USE OF OIL IN TRANSFORMERS Fig. 8. thermometer. The difference of temperature of transformers operated with and without oil as shown in these curves is greatly exaggerated in larger sizes. When the transformers are of such size that sufficient radiating surface cannot be had in the tank to dissipate the heat, it becomes necessary to provide artificial means for cooling the same. The principal methods employed are the use of a forced blast of air and by the circulation of water through the coils immersed in oil-cooled transformers. The former are known as air-blast transformers and the latter as water- cooled. Some special forms of water-cooled transformers have been built, wherein water has been circulated through the conductor itself. Transformers have been constructed in sizes up to about 4000 K.W., using water circulation for cooling. An Air-Blast Transformer, or one in which ventilation and radi- ation of heat is, by means of a blast or current of air, forced through the transformer coils and core, is shown in Figs. 14 and 15. In this transformer, the coils are built up high and thin, and assembled with spaces between them, the air being forced through these spaces. The iron core is also built with numerous openings through which the air is forced for cooling pur- poses. This style of transformer has been constructed in si^es up to about 450 THE STATIC TRANSFORMER. m 1 ~r~ A" ii ii fc- : /QOT^ r - £ " i i Fw. 9. 300 K.W. Oil-Insulated Water-Cooled Transformer. Fig. 10. Water-Cooled Oil- Insulated Transformer. Fig. 11. 200 K.W. 22,000- Volt Oil-Insulated, Self-Cooling Transformer. WATER-COOLED TRANSFORMERS. 451 Fia. 12. Water-Cooled Transformer out of Tank. Fig. 13. Water-Cooled Transformer in Tank with Switch for Voltage Regulation. Fig. 14. 250 K.W. Single-Phase Air- Blast Transformer. Fig.. 15. Section of Air-Blast Transformer. EFFICIENCIES. 453 EFFICIENCIES. The efficiency of a transformer is the ratio of the output watts to the input watts. Thus Efficiency = Output Output watts _ Input watts — Output + Core loss + Copper loss The core loss, which is made up of the hysteresis loss and eddy current loss, remains constant in a constant potential transformer at all loads 28000 1 1 1 1 1 1 II II 1 1 1 1 1 1 1 = -B -C TRANSFORMER IRON.AGEING TESTS. BYH.F. PARSHALL HYSTERESIS IN THE IRON AS RECEIVED HYSTERESIS TRANSFORMER AFTER , SHORT PERIOD OF LIGHT WORK > HYSTERESIS TRANSFORMER AFTER THREE YEARS OF HEAVY WORK u / / / tol / / °. / => / O / UJ i c / co / tc 12000 C v * i ' / CO / / / y m gQQQ / F * / m / t s K / s £ / s * I <, 2 4000 / S / t s S / r* y y 1 i \ t 1 i 2 14 LINES PFR SQUARE CENTIMETER Fig. 16. 100 o 00 A - 80 m. z \ 70 / \ 5- / MANUFACTURE BY A.H. FORD,AT UNIVERSITY OF WISCONSIN. JAN. FEB. MAR. 1897. B- TEST ON WAGNER TRANSFORMER , FEB. MAR. APR. 1897. / *m j / / 60 t B "*"? 40 V DAY8 RESULTS OF AG~.ING TESTS Fig. 17. 454 THE STATIC TRANSFORMER. while the copper loss, or I 2 R loss, varies as the square of the current in the primary and secondary. Methods for determining all the losses are fully described in the chapter on transformer testing. In a service where a transformer is generally worked at full load while connected to the circuit, as in power work, the average or "all-day" effi- ciency will be about the same as its full-load efficiency. By " all-day " effi- ciency is meant the percentage which the energy used by the customer is of the total energy sent into the transformer during twenty-four hours. In lighting work the transformers are usually connected to the mains or ------- 7 __ _ __ ^:_ :?: : _s_ 7 ■_ :_^ 7 __::_ : S __ _ _^__ . -_ -; _+ ::i-^: "1 Z- --I ---I :__^ _ j_:::::i:::ii: _ — ^ -_.-._. _^_ _ z _:::::::::>-2 / . , . <■' . . : : :z~~ ~ - ::;zi:!:::: . _ / „,jl . / ~ " ? .:::::::: _/ _-Q0'i^-' ----------- Q ^OU - - ^r* 7 \9V". P H J o4f) - JSZ' __ - - --F- ^- ZiU -+q , _ -^^ -,,."-- II"""III p ,^_u ^ - -Eg;*-. - - _ r _^-_j_„ _ __ HO - tt- -- - l ™~ r/~^ ~ ~ : : ; 100 -- 32 ^ Jlt. 80 2 g2 - _- - 60--3fc^___ __ _. J£f_ _2 " : — 20 * ■ ■ - ii i i_. 0123 45 6 7.5 10 15 20 2i 30 36 40 4^ 6t K. W. CAPACITY Fig. 18. Comparative Curves of Core Losses and Regulation, Showing the Improvement made in Transformers from 1897 to 1902. are excited the full twenty-four hours per day, while the customer draws current from them during from three to five hours in the twenty-four. Assuming on an average five hours full load, the losses will be 5 hours I 2 R and 24 hours core loss. The calculation of the "all-day" efficiency can, therefore, be made by the following formula: All-day efficiency = Full load X 5 Core loss X 24 + I 2 R X 5 + Full load X 5 From this it is evident that while for power work or continuous fulljoad the relative amount of the core and copper losses will not affect the " all- day " efficiency seriously, yet in the design of transformers which are worked at full load only a short time, but are always kept excited, a large core loss means a very low " all-day " efficiency. MAGNETIC FATIGUE. 455 450 "TT't" -— — — — — - till 425 1899 ' ^r'" 400 1902 375 INDICATE/? DIFFERENCE BETWE TRANSFORMERS OF 1899 & 190 / , EN . gi4:S.5_ " z £250 .zzfcp-^r ^ P CO « • iO • -lOCO CM -i— 1 i-H • • 665 1,680 V,670 1,540 1,515 1,445 1440 *o d o O CO .2 *Sj CU cu 43 01 >> a In- crease per cent. Or-itM «S0 • cMOO •© »OI>. '(50 ♦ O500t> • «rt* i-Ht-H -i-H • i-H,-Hi-« • »*H 02 >> w In- crease per cent. ONN 00 «b» «CM • 00 •© •© IQCO b- '''iH -CO • CO •"* •«* •rH .»-H • rH • ,-H • i~t ^d 610 960 1,020 1,090 1,325 1,450 1,465 1,465 CO d a 00 .2 *oo GO 02 >> n In- crease per cent. ©CO© «000 CO .CO 0000 02 GO ^d OiOO -too 10 «o © *»o o»o 005l> • 00 rH i-i -CO CO •!> O5C0 COCOt^- -00O5 05 -OS © •© Oi-h CM d CO .2 "S cu (h CU 18 >> n In- crease per cent. ©CMCM "*t •© i-iCO rH .f-lrH CMCM • CO "0 • 10 -IS '*"<* COCOCO -t>b- t^00 • 00 -00 0505 ■H d o O 02 *02 CU (h CU 43 02 In- crease per cent. . . CO -co ... -co • O ' ' rH -iH ... -rH . ^ .'01 .GO 02 • <6 10 • • CO .10 ... .10 • •© CO . . ^ .^ ... -rf • CO -05 CO • • CO -CO • • • >CO • CO *CO 6 O cu H E if O^CM cO'tfCO 1^00© CMcOiO OiOI>. T-irHr-t CMCNCN 458 THE STATIC TRANSFORMER. Regulation. The most important factor in the life of incandescent lamps is a steady voltage, and a system of distribution in which the regulation of pressure is not maintained to within 2% is liable to considerable reduction in the Hfe and candle-power of its lamps. For this reason it is highly important that the regulation, i.e., the change of voltage due wholly to change of load on the secondary of a transformer, be maintained within as close limits as possible. In the design of a transformer, good regulation and low core loss are in direct opposition to one another when both are desired in the highest de- gree. For instance, assuming the densities will not be changed in the iron or in the copper, if we cut the section of the core down one-half we decrease the core loss one-half. The turns of wire, however, are doubled, and the reactance of the coils quadrupled, because the resistance changes with the square of the turns in series. A well-designed transformer, however, should give good results, both as regards core loss and regulation, the relative values depending upon the class of work it is to do, and the size of the transformer. Comparative Expense of Operating- JL.i rg*e and Small Transformers. It is obvious that the design of the distributing system has quite as much to do with the maintenance of a steady voltage as does the regulation of the transformers, and the proper selection of the size of transformers to be used requires skilled judgment. When transformers were first used it was the custom to supply one for each house, and sometimes two or three where the load was heavy. Expe- rience and tests soon made it evident that the installation of one large transformer in place of several small ones was very much more economical in first cost, running expenses (cost of power to supply loss), and regulation. Where transformers are supplied one for each house, it is necessary to provide a capacity for 80 % of the lamps wired, and allowing an overload of 25% at times. Where one large transformer is installed for a group of houses, capacity for only 50% of the total wired lamps need be provided. For resi- dence lighting, where the load factor is always very low, it is often best to run a line of secondaries over the region to be served, and connect a few large transformers to them in multiple. A study of the following curves will show in a measure the results to be expected by careful selection and placing of the transformers. The first curve, Fig. 22, shows the relative cost per lamp or unit of transformers of different capacity, showing how much cheaper large ones are than small ones. S a I id > 1 V -->— — LU I I 100 200 300 400 500 600 FIG. 22. Relative Cost of Transformers of Different Capacities. The second set of curves (Fig. 23), shows the power saved at different loads, and using different sizes of transformers. Power factor is the ratio of the actual watts in a line to the volt amperes or apparent watts in that line. It is also defined as the cosine of the angle of phase displacement of the current from the voltage in the circuit. TESTING TRANSFORMER. 459 The power factor of most commercial transformers is low at no load varying from 50% to 70%, while at high loads the power factor is very Fig. 23. Relative Efficiency of Large and Small Transformers. nearly 100%. For this reason it is better to distribute the transformers on tne line so that they will carry load enough most of the time to keen the power factor reasonably high. TESTIJ¥« Tftl^FOIUIEH. The term testing transformer is a commercial one for describing a trans- former used in testing the insulation of cables, transformers and other ap- Fig. 24. Shop Testing Set. to 12,000 Volts by 200 Volt Steps. paratus. Such apparatus is generally tested at a voltage from 2 to 1U times the working pressure. It is necessary, therefore, to build such 460 THE STATIC TRANSFORMER. transformers for very high voltages, some having been made for pressures as high as 500,000. Because of the severe nature of the service to which they are subjected it is essential that more than ordinary attention be paid to the insulation Fig. 25. of the windings so that a minimum potential strain results between adja- cent portions, and that sufficient insulation be provided between the two windings. These transformers are generally of the core type of design, because the construction of this type of transformer lends itself more readily to the TESTING TRANSFORMER. 461 sub-division of the high voltage coils into separate and independent parts of few turns, thus reducing the potential strain within such coils to a very low figure. Such transformers are almost invariably oil-insulated and the best prac- tice is to place them in metal cases which are connected to the ground to protect the operator against accident from the static induced by the high voltage winding. Figs. 24 and 26 show two types of this appliance, Figs. 24 and 25 show- ing a handy shop testing set with diagram of connections, and Fig. 26 showing a set for moderately high voltage. The only practical way of measuring the high potential generated by these transformers is by spark-gap shunted across the terminals of the TEST LINE 50 OR Flo. 26. S. K. C. High Voltage Testing Set. transformer. Ordinarily the spark-gap is set for the desired voltage by use of a calibration curve or by preliminary calibration by means of a voltmeter connected to the low potential side, the ratio of the transformer being known. A high resistance should be connected in series with the spark-gap to prevent the flow of an appreciable amount of current should the potential jump across the needle points: this will prevent the accumulation of high frequency voltage which might otherwise result. 462 THE STATIC TRANSFORMER. . - '. . ' ^-,Z J 2 ""• ;/6 — 2.5 ,Z ^ ^~Z 7 -^ / 2 » -R° -+- / • 7 ^ ^ ,* ■ T 1 !^ -7*'- V 4- ; 14 — 1'5 ^ r y ^ ^ it /•* _J /U £ -*£^ J M-. 5 -/ y^ ■ 5 ± Z 3M: . 4 / ff\ 3 J^- -X?- -? it -.22 *£ 7 *s5- ^^r ^ac JO 20 30 40, ,50 60 -70 00 90 100 HO IZQ 130 MO /50 160 Kilo Vqlt3 Fig. 27. Sparking Distances Across Needle Points. Transformer for Constant Secondary Current. Several methods have been tried with more or less success to obtain con< stant current at the secondaries of transformers. The simplest and earliest system for obtaining a constant current in the secondary is by means of transformers whose primaries are connected in CONSTANT CURRENT LINE SERIES TRANSFORMERS ooooq_r CIRCUIT POINTS \JJJL^JLT REACTIVE COIL s-arc lamps -* Fig. 28. series, and a constant current maintained in the primary. This is shown in diagram in Fig. 28. Series transformers for this purpose have never been very successful, due to the trouble caused by the rise of potential in the TYPES OF TRANSFORMERS. 463 secondary when opened for any cause. Various devices (Fig. 28), such as short-circuiting points separated by a paraffined paper, or a reactive or choking-coil connected across the secondary terminals, have been intro- duced to prevent any complete opening of the secondary by reason of any defect in the lamp or other device connected in the circuit. Reactive coils used as shunt devices have been used under different names ; as compensators, choking coils, and economy coils. A device of this kind has been introduced by the Westinghouse Electric and Mfg. Company, and others, for use in street-lighting by series incan- descent lamps. It is shown diagrammatically in Fig. 29. The lamp is m r 3_ ir o i__r > i — 1 .tJUUL,y ljte.roi Lc2-£j£J- CONSXA'NT 13 — CT Fig. 29. placed in shunt to the coil ; when the filament breaks, the total current passes' through the coil, maintaining a slightly higher pressure between its terminals than when the lamp is burning. It is thus evident that the regu- lation of the circuit is limited, due to the excessive reactance of the coils when several lamps are taken out of circuit. Economy Coils or Compensators. A modification of the above is built by several companies for use on ordi* nary low potential circuits, where it is desired to run two or three arc lamps. It is a single coil transformer, and is shown in Fig. 30, and diagram- matically in Fig. 31, same page. If any lamp is cut out or open-circuited, D. P. SWITCH D. P. FUSE BOX COMPENSATOR S.P.SWITCH Fig. 30. Arrangement of Apparatus for use of Economy Coil or Compensator. Fig. 31. Westinghouse Econ- omy Coil, for A. C. Arc Lamps. the current in the main line decreases slightly. As more lamps are cut out the remaining lamps receive less current, and it is necessary to replace the bad lamps in order to obtain normal current through the circuit. 464 THE STATIC TRANSFORMER. Transformers for Constant Current from Constant Potential. The transformers represented in Fig. 32 show a design that will give out an approximately constant current when connected to constant potential circuits. The transformer has its core jy .secondary so designed that there is a leakage >£\ I ln c ° RE n x X X XX . P atn for tne flux between the primary Co) 1| I Ep • " arc lamp^ ^ I and secondary. This is shown in the V^l +3LI b L^- — * — x X ti ' diagram at a and 6. At open second- primary J _ 1 ar y circuit there is little or no ten- dency for the flux to leak across the Fig. 32. Constant-Current or Series S a P- When current flows through the Transformer. secondary, thus creating a counter magneto-motive force, there is then a leakage across this path, and if properly proportioned, this leakage will act to regulate the current in the secondary, so that it will be approximately constant. General Electric Constant Current Transformers. The transformer thus described has the disadvantage that its regulation is fixed for any transformer and may vary in transformers of the same design without any ready means of adjustment. The transformer also regulates for constant current over but a limited range in the secondary loads. The General Electric Company constant-current transformer shown in Figs. 35 and 36 is constructed with movable secondary coils, and fixed pri- mary coils. primary circuit J FUSE BOX f TUBULAR .PLUG SWITCH Fig. 33. Constant-Current Trans- former showing Counterweight and Primary and Secondary Leads from Winding. Fig. 34. Connections for Alter nating Series Enclosed Arc Lighting System, with 50, 75, or 100 Light Transformer. The weight of the movable coil is partially counterbalanced, so that at normal full-load current the movable coil or coils lie in contact (see Fig. 35) with the stationary coil, notwithstanding the magnetic repulsion between them. When, however, one or more lamps are out of the circuit, the in- creasing current increases the repulsion between the coils, and separates them, reducing the current to normal. (See Figs. 35 and 36.) At mini- mum load, the distance between the coils is maximum. The regulation is thus entirely automatic, and is found to maintain practically constant current, or a departure from constant current if desired. The transformer can be adjustea for practically constant current for positive regulation; TYPES OF TRANSFORMERS. 465 i.e., increasing current from full load to light loads, or for a negative regu- lation, i.e., decreasing current from full load t light loads. This adjust- Fig. 35. Diagram of Connections. Fig. 36. Mechanism of Oil-Cooled Con- stant Current Transformer— 100 Lamps. ment is obtained by changing the position of a cam from which the counter- weights are suspended. The curves shown in Fig. 37 show the range obtained in a 100-light transformer. The transformers are enclosed in cast iron or sheet iron tanks filled with transil oil. The oil, in addition to being an insulating and cooling medium, serves to dampen any sudden movement of the secondary coils. These transformers are connected to the regular constant potential mains, 1 RE 5ULATI0N TE8T 10 )L. I G. E.CO iSTA NT e.8 «; e. e> £6.4 X < 6.2 iJ66 TJVg g*2ULA7 JON cu RREf ,TTF AN8 ■ORM ER.. , RE.GULA TION ive3 EGO} *f\0 N-,_ Nl isfi LOAD Vt LOAD •A LOAD fUJ-l LOAl Fig. 37. and the larger siaes are arranged for multiple circuits in the secondary. After having been started on a run the transformers need no attention, as they are entirely automatic in their action. In the Westinghouse constant-current transformer the movable coil is partially counter-balanced by a weight or another movable coil, depending upon the size of the regulator. A dashpot is arranged to permit free sep- aration of the coils, but slow approach. This device is important at starting and overcomes the tendency to pump, common to such transformers. The full-load efficiency of this type is practically the same as that of a constant-potential transformer of the same capacity. The power factor of the system at full load is about 85 per cent, due to the reactance of alternat- ing arc lamps. At fractional loads the power factors necessarily are much tower, and it is therefore not desirable to operate such a system at light load, 466 THE STATIC TRANSFORMER. RE4CTAIVCE FOR 4LTEH\4TOOJ CIJRREIII ARC CIRCUITS. For low voltage circuits required on transformers, a modification of the constant-current transformer has been devised in the regulating reactance connected in series with the line. Fig. 38 shows a typical construction Fro. 38. Regulating Reactance Coil by Manhattan General Construction Co. adopted by one of the leading manufacturers. It consists of a single coil of insulated wire arranged to inclose more or less of one leg of a " W "-shaped magnet as shown in the following cut. The coil is suspended from one end of a lever and counterbalanced by a weight on the other, and so arranged that at all points of its travel it just balances the varying magnetic pull of the coil. The arc circuit is connected in series with this coil with a switch to open the circuit. Without cur- rent flowing, the normal position of the coil is at the top or off the leg of the magnet. When the switch is closed, current flows in the circuit (and coil), and draws the coil down on the leg to a point where the reactance of the coil holds the current strength at a pre- determined point; as, say, 6.6 am- peres. It is said that this device will maintain a current constant within one-tenth of an ampere. The losses are the iron losses and PR losses in the coil, which, with constant current, are the same un- der all conditions of load. As it is not always, or even often, that it is necessary to pro- vide for regulation of an arc cir- cuit to the extent of its full load, the makers have adopted the pol- icy of supplying instruments to care for but that part of the load , , . that is expected to vary, in some eases 10% of the circuit and in others 75%, thus avoiding the need for Arger apparatus, or for insulation for the total voltage of the circuits. Fro. 39. "G. I." Series A. C. Regulator. POTENTIAL REGULATORS. 467 They claim another advantage in being able to connect the device in one leg of the series circuit, and allowing the other end of the circuit to be con* nected to the mains at any such point as may be the nearest at hand. Fio. 40. Potential Regulator** An alternating current potential regulator is essentially a transformer hav- ing its primary connected across the mains, and its secondary in series with the mains. The secondary is arranged so that the voltage at its terminals can be varied over any particular range. Fio, 41, Diagram of Connections for Single-Phase Potential Regulator, Westinghouse Elec. and Mfg. Co. 468 THE STATIC TRANSFORMER. The several different styles of feeder regulators have been devised, differ- ing in principle of operation, but all of them have the primary coil con- nected across the mains, and the secondary coils in series with the mains. The " Stillwell " regulator, which was designed by Mr. L. B. Stillwell, has the usual primary and secondary coils, and effects the regulation of the cir- cuit by inserting more or less of the secondary coil in series with the line. This secondary coil has several taps brought out to a commutating switch, as shown in Fig. 40. The apparatus is arranged so that the primary can be reversed, and therefore be used to reduce as well as to raise the voltage of the line. It is evident from an observation of the diagram that if two of the segments connected to parts of the coils were to be short-circuited, it would be almost certain to cause a burn-out. To prevent this, the movable arm or switch-blade is split, and the two parts connected by a reactance, KAPPS MODIFICATION Fig. 42. this reactance preventing any abnormal local flow of current during the time that the two parts of the switch-blade are connected to adjacent seg- ments. The width of each half of the switch-arm must of necessity be less than that of the space or division between the contacts or segments. As the whole current of the feeder flows through the secondary of the booster, the style of regulator which effects regulation by commutating the secondary cannot well be designed for very heavy currents because of the destructive arcs which will be formed at the switch-blades. To overcome this difficulty, Mr. Kapp has designed the modification which is shown in Fig. 42. In this regulator the primary is so designed that sections of it can be commutated, thus avoiding an excessive current at the switch. This regulator, however, has a limited range, as the secondary always has an E.M.F. induced in it while the primary is excited ; and care must be taken to see that there are sufficient turns between the line and the first contact in order to avoid excessive magnetizing current on short circuit. FROM TO LOAD SWITCHBOARD SECONDARY -f.| L PRIMARY Fig. 43. Connections for M. R. Feeder Regulator of G. E. Co. RAISE V0LTAaE_^^2DI» *ER VOLTAGE •CONTMCL'INa HAND WHEEL Fig. 44. Diagram of Con- nections of Feeder Po- tential Regulator. The General Electric Company have brought out a feeder regulator, in which there are no moving contacts in either the primary or secondary, and which can be adapted for very heavy currents. This appliance is plainly shown in Figs. 43 and 44. The two coils, primary and secondary, are set at right angles in an annular body of laminated iron, and the central lami- THREE-PHASE REGULATORS, 469 nated core is arranged so as to be rotated by means of a worm wheel and shaft as shown. The change in the secondary voltage, while boosting or lowering the line voltage, is continuous, as is also the change from boosting or lowering, or vice versa. In this regulator, the change of the secondary voltage is effected by the change in flux through the secondary coil, as the position of the movable core is changed by the turning of the hand wheel and shaft. There are, therefore, no interruptions to the flow of current through either the primary or secondary coils, and the regulator is admirably adapted for in- candescent lighting service, where interruptions in the flow of current, how- ever instantaneous, are objectionable. Separate Circuit Regulators, Where a number of circuits are run out from the same set ol bus bars, regulation of each circuit is provided for by the use of a single coil trans- former from various points, on the winding of which leads are brought out to a regulator head, from which any part or all of the transformer may be thrown into service to increase the pressure on the line. Three-Phase Jleg-ulators* The regulator described above is suitable only for operation on single- phase circuits. The primary is connected in a shunt and the secondary in series with the circuits to be controlled. Two or three-phase regulators of similar design, but having either primary or secondary on the moving H3 Fig. 45. Three-Phase Induction Potential Regulator. core, are commonly used. The voltage in such a design is constant in each phase of the secondary winding, but by varying the relative positions of primary and secondary the effective voltage of any phase of the secondary in its circuit is varied from maximum boosting to maximum lowering. Referring to the diagram which represents graphically the voltage of a single phase of the regulator, e o = Generator voltage or the E.M.F. im- pressed on the primary; a o =» E.M.F. generated in the secondary coils, and is constant with constant generator E.M.F.; 6' a* = Secondary E.M.F. in phase with the generator E.M.F.; e' a' = Line E.M.F. or resultant of the generator E.M.F. and the secondary E.M.F. The construction of the regulator is such that the secondary voltage o a is made to assume any desired phase position relative to the primary E.M.F., as o /, o b, o c, etc. When its phase relation is as represented by o /, which is the position when the north poles and the south poles of the primary and secondary windings are opposite, the secondary voltage is in phase with the primary voltage and is added directly to that of the generator. The regulator is then said to be in the position of maximum "boost," and by rotating the armature with reference to the fields, the phase relation can be changed to any extent between this and directly opposed voltages. When the voltage of the secondary is directly opposed to that of the primary or gen- erator, its phase relation is as represented by o d in the diagram, while o b represents the phase relation of the secondary when in the neutral position. 470 THE STATIC TRANSFORMER. TIIREE-I»1I.4*E TRAMIFORHEB§. This type of transformer has been commonly used " abroad " for a long time and has recently been introduced into American practice. Such transformers differ little from the single-phase designs and may be built in either core and shell type. The three-phase shell type transformer consists simply of the single- phase units so united that considerable of the iron in the core becomes unnecessary. This is illustrated by the following cuts. Fig. 46. Three-Phase Core Type Transformer. Fig. 47. Core of Three- Phase Core Type Transformer. A three-phase core type transformer consists of three legs of single-phase core transformer placed side by side and united at either end by a yoke of the same cross section as each single-phase leg. Fia. 48, Cross Section of the Cores and Coils of Three Single-Phase Air-Blast Transformers. liiiiii I Hi Fio. 49. Cross Section of the Same Coils Combined in One Three-Phase Air- Blast Transformer of a Capasity Equal to the Total Capacity of Those Above. RATIO OP TRANSFORMATION, 471 Fig. 50. Three-Phase Air-Blast Transformer in Process of Building. Fig. 51. A Typical Three-Phase Air-Blast Transformer. RATIO OIF Tit AX&JFOmiATIOtf T'X IIIREE.PHASfi Transformers are usually built with both their primary and secondary coils wound in two or more sections in order to facilitate changes of trans- formation ratio. This is especially useful where three transformers are used in a three-phase system. Let n = ratio of transformation from one section of high-tension side to one section of low-tensionside, expressed as an integer; Y = total number of sections in series in each arm of the star, high- tension side; D = total number of sections in series in each arm of the delta, high- tension side; y, and d, being the corresponding quantities for the low-tension side. Then, H.T. line volts L.T. line volts ' yVs + d This formula is applicable to combination stars and deltas as well as to simple stars and deltas. Example. ~ Fig. 52 shows a combined star and delta for the H.T. side and a simple star for the L.T. side. Ratio =» 10 2\/3 4-3 3\/3 + n = 10, Y = 2, D = 3, -• — •- L. T. Side Fig. 52. 472 THE STATIC TRANSFORMER. TRAJfiFORMEB COWtfECTIOHTS. Some of the advantages claimed for alternating current systems of dis- tribution over the direct current systems is the facility with which the potential, current, and phases can be changed by different connections of transformers. On single-phase circuits, transformers can be connected up to change from any potential and current to any other potential and current; but in a multi-phase system, in addition to the changes of potential and current, the phases can be changed to almost any form that may be desired. Single -Phase. The connections of the single-phase step-down and step-up transformers, having parallel connections, need no explanation. For residence lighting, a favorite method of supply is through single-phase transformers with three-wire secondaries. A tap is brought out from the middle of the sec- BALANCING TRANSFORMER Fio. 53. Arrangement of Balancing Transformer for Three- Wire Secondaries. ondary winding, this tap connecting to the middle or neutral of the three- wire system. In this way a few large transformers can be connected by three-wire secondaries in a residence or other district, and will take care of a large number of connected lamps. 4 1000 » Fig. 55. Single-Phase, Fig. 54. Single- with Three- Wire Sec- Fig. 56. Two- Fig. 57. Three- Phase. ondary, Useful for Phase, Four Wire, Two- Residence Circuits. Wires. Phase. TRANSFORMER CONNECTIONS. 473 Kapp shows a modification of the three-wire circuits, in which the out- side wires are fed by a single transformer, and the neutral wire is taken care of by a balancing transformer, connected up at or near the center of distribution. The capacity of the balancing transformer need be but half the greatest variation in load between the two sides. Some makers of transformers have the connection board in their trans- formers so arranged that the two primary coils may be connected either in series or parallel by mere changes of small copper connecting links, so that the same transformer can be connected up for either 1000- or 2000-volt circuits, and the secondary for either 50 or 100 volts. Two-Phase. The plain two-phase or quarter-phase connection (Fig. 56) is simply two single transformers connected to their respective phases, the phases being kept entirely separate. In the three-wire quarter-phase circuit, one of the leads can be used as a common return, as shown in ' g. 57. Three-Phase. The three-phase connections shown in diagram 58 are known as the delta connections, and are of great advantage where continuity of service is very important. The removal of any one transformer does not interrupt Wj nm nrain Fig. 58. Three-Phase Delta Connection. Fig. 59. Three-Phase Star Connection. the three-phase distribution, and the removal of two transformers still admits of power transmission on a single phase of the circuit. The Y or star connection, as shown in diagram 59, has one of the terminals of each primary and secondary brought to a common connec- tion, the remaining three terminals being brought to the main line and the distributing lines. The advantage of the star connection over the delta con- nection is, that for the same transmission voltage each transformer is wound ior only 58% of the line voltage. In high-voltage transmission this admits of much smaller transformers being built for high potentials than is possi- ble with the delta connection. 474 THE STATIC TRANSFORMER. Arrangement of Transformers for Stepping* Up and Down for .Long- Distance Transmission. Figures GO, 61, and 62 show diagrammatically the connections for adapting three-phase transmission to quarter-phase generators, with interchangeable and non-interchangeable transformers. GENERATOR STEP UP TRANSFORMER 8TEP DOWN TRANSFORMEP Fig. 60. Changing Quarter-Phase to Three-Phase, Non-Interchangeable Step-up Transformers. Fig. 61. Changing Quarter- Phase to Three-Phase, and back to Quarter-Phase. All Transformers Inter- changeable. -«2000-V> SLSLMJULT 2000Vt* tfiJUUlSJiai STEP UP nrjprafftfri transformers 0000->:<10000V7> MV&1*-*-\sxjjxI*-L*SM*Jm) ^T^^^ -110V— -110-V— Fig. 62. Changing Quarter- Phase to Three-Phase. All Step-up Transformers Inter- changeable. TRANSFORMER CONNECTIONS. 475 Three-Phase to Six-Phase Connections. A rotary converter wound for six-phase has a greater capacity for work than the same machine wound for three-phase. Three-phase transmission, however, is very economical, and in Figs. 63 and 64 is shown a diagram by which six phases can be obtained from three phases by the use of only three transformers. Each transformer has two secondary coils. One secondary of each trans- former is first connected into a delta, then the remaining secondary coils are UwwwJ Iwvw w J UvwwvJ SMA/VV^^ n/VvWn fVV\AAAa nAA/W\n ( AAAAA ) /W\A^ 'PI f Six- Phase A Figs. 63 and 64. Three-Phase to Six-Phase Connection. connected up into a delta, but in the reverse order of the first delta. This is an equivalent of two deltas, one of which is turned 180° from the other. In the diagram ABC represents one delta, and DEF the other. Kvwwf ! ' .J W\AAA/J s.j W WW L n A-WWW^ pAMA/W-^ Wvww' Fig. ( Diagrams of Connections for Changing from Three-Phase to Six-Phase. In the same way the two secondaries can be connected up Y, and one Y turned 180° to obtain six phases. The disadvantage of Y connec- tion, however, is that in case one transferrer is burned out, it is not possi- ble to continue running, as can be done with delta connections 476 THE STATIC TRANSFORMER. Methods of Connecting: Transformers to Rotary Converters. vwwwvwJ VwwwvwJ A/VWA Fig. 66. Two-Phase. /VwWa\A/\ WyVWA FIG. 68. Three-Phase J. WwAwJ Wm\W^ IwWWwW ^^^^^^ ,, a/vw\a 2, /wyvvv^ Fig. 70. Six-Phase Diametrical. ■ WMNVWWAMM/W VVWWVWWVVW M/VWNAAA K/WWW\ iL -H 1 2' c «fcjv|3- si. J?2 FIG. 72. Six-Phase T. Wwwwwv wvwvvAW Vwwvww/ i3 ^wwv t 7^A^^A-ysAA^J Fig. 67. Three-Phase A. VwwW IvwwW IwwwJ kvw\ kA/vw,tww\ 2| 3 Fig. 69. Three-Phase y. •WwwwJ UwvJ Uvwww l5_ Iwwi^A/wvyww] |WW\ /WW^ j 2'f L y^) Fig. 71. Six-Phase A. Wwwww/ WW. tvvwwvv! JVWVy twwy IvwVvil WW^ /VV\AA M/V\A L Fig. 73. Six-Phase y. CONVERTER AND TRANSFORMER CONNECTIONS. 477 A/N/N, ^0 rj J™^*^™* Fig. 74. Three Transformers Arranged in Inter-connected Star, Operating a Three-Phase Rotary Converter on a D. C. Three-Wire System. Converter and Transformer Connections. The ,, Scott" connection is used a great deal in transmissions and distri- butions (See Fig 75.) One transformer is designated the main, and the other the teaser. Two transformers are required. They are made exactly alike, so that with proper connections either may be used as mam or teaser. The winding is provided with a 50% tap and with taps so that 86.6% of the winding may be used. 1-2-3 are three-phase voltage, A~A one-phase, B-B' the other of the two-phase circuit. Reference to the small diagram shows the reason for using 86 6% of winding of one transformer; also the necessity for the 50% tap. Teaser Hvwwwv^ Main 100% /oo% Fw. 75. Fig. 76. WE1SIBOO POWER O 8IX-PHASE CIBCVIII. Use two pairs of wattmeters, each pair connected to one of the three- phase circuits as shown in Fig. 76. If power factor is less than 60% one meter of each pair will read negative. The algebraical sum of the read- ings of each of the two pairs will be the result required. 478 THE STATIC TRANSFORMER. Y OH A COOTVECTIOl? OF TRANSFORMERS. (F. 0. Blackwell. Trans. A. I. E. E M 1903.) Transformers. Assuming that three transformers are to be used for a three-phase power transmission and that the potential of the line is settled, each of the trans- formers, if connected in Y, must be wound for — — or about 58 per cent of V3 the line potential, and for the full line current. If connected in A, each transformer must be wound for the line potential and for 58 per cent of the line current. The number of turns in the transformer winding for Y connection is, therefore, but 58 per cent of that required for A connection, to avoid eddy current losses that occur when the cross section of the con- ductor is too large. The Y connection requires the use of three transformers, and if any- thing goes wrong with one of them the whole bank is disabled. With the A connection, one of the transformers can be cut out and the other two still deliver three-phase power up to 86.6 per cent of their aggregate capacity, or 66.6 per cent of the capacity of the entire bank. Fio. 77. Step-down Transformer for 4000 Volt Y Distribution. Combined three-phase transformers are now made of any size and are prefer- ably Y connected on the high potential side. Grounding* the Neutral. If the common connection of transformers joined in Y is grounded, the potential between windings and the core is limited to 58 per cent of that of the line. Under normal conditions, the potential between any conductor of a three-phase transmission circuit and the ground is 58 per cent of the line potential, with either Y or A connection, but the neutral may drift so as to increase the potential with an ungrounded system. If one branch is Fig. 78. Step-down Transformer for 200 Volt Y Distribution. partly or completely grounded, the potential between the other two branches and the ground is, of course, increased and may be the full line potential. With a grounded neutral Y system, a ground is a short circuit of the trans- formers on the grounded branch, and the transmission becomes inoperative. CONNECTION OF TRANSFORMERS. 479 From the point of view of safety to life and prevention of fires this is a desirable condition, especially if the low tension distribution is also grounded. If the high tension circuit makes contact with the ground or low potential system, it can be immediately cut out by fuses or automatic circuit breakers. The difficulty is that a power transmission with grounded neutral is likely to be frequently shut down by temporary grounds, such as would be caused by a tree blowing against one of the wires. Even if the circuit is not opened, the drop in the pressure due to the sudden "short" on the line will cause synchronous apparatus to fall out of step. Unstable Neutral. If two transformers are connected in series, there is no certainty that they will divide the potential equally between them. A system in which all the electrical apparatus is connected in Y has somewhat the same char- acteristics. The neutral may drift out of its proper place and there will be unequal potentials between it and the three conductors of the circuit, due to unequal loading and differences in the transformers or transmission cir- cuits. Such unbalancing would cause unequal heating of the transformers, and if a four-wire three-phase system of distribution were employed, would seriously interfere with the regulation of the voltage. If transformers, therefore, have Y secondaries, it is desirable that the primary should be A connected. Two systems in common use with which A primary wind- ings should be used, are shown in Figs. 77 and 78. Rise of Potential. The high potential windings of transformers are necessarily of high reactance, and if left in series with a circuit of large capacity, as shown in Figs. 79, 80, 81, and 82, the leading charging current flowing over the react- ance may set up extraordinarily high pressures. Figs. 79 and 80 represent Y-connected banks of three transformers each connected so as to cause such rOi -mw- rWWW- Fiq. 79. Fig. 80. a rise of potential. In Fig. 79 the primary of one transformer is excited by a generator, the primary of the other two transformers being open-circuited. In Fig. 80 the primary of one transformer is open-circuited, the other two being connected to the generator. Figs. 81 and 82 show T-connected banks of two transformers, which might be used to transform from either two- phase or three-phase to three-phase or vice versa, and are similar in action to Fig. 79. If in anyone of Figs. 79, 80, 81 and 82 the secondaries are con- nected to a long distance transmission circuit, a pressure of many times the normal potential will beset up between A and B, and between B and C, that between A and C not being affected. It is theoretically possible for a potential 100 times that for which a trans- former is wound, to be caused by opening the primary switches of one or more of the transformers of a bank connected in Y before the secondary switches are used. Actually, the current jumps across the insulation at some point in the system before there car. be any such increase in pressure. If there are a number of banks of transformers in parallel, this phenomena cannot occur except when all but one bank are disconnected. This source of trouble could be obviated by employing oil switches on the high poten- 480 THE STATIC TRANSFORMER. tial side which disconnect the line before the low tension switches are used, or by triple pole switches on the primary which open all three branches of vhe bank of transformers at once. The selection of Y or A connection of transformers for long distance r-^ V^ I 1 — 'waaw B Fig. 8i. Fig. 82. transmissions should only be determined after a careful consideration of the conditions in each case. There is little choice between Y or A without a grounded neutral. Note. — For further information on this subject [see discussion on this paper in Proceedings of A. I. E. E. for 1903. CcJtLXEIl AJL ELECTRIC COMPANY ^EBCIH V ARC RECTIFIER!. H GH © -WWWWAMr- TRANSFORMER -A/WW\AAA- A.C. SUPPLY (By P. D. Wagoner.) A detailed idea of the operation of the mercury arc rectifier circuit may be obtained from Fig. 83. Assume an instant when the terminal H of the supply transformer is positive, the anode A is then positive and the arc is free to flow between A and B, B being the mercury cathode. Following the direction of the arrows still further the I current passes through the load J, through the reactance coil E and back to the nega- tive terminal G on the transformer. A little later, when the impressed electromotive force falls below a value sufficient to. main- tain the arc against the counter electro- motive force of the arc and load, the reactance E, which heretofore has been charging, now discharges, the discharge current being in the same direction as formerly. This serves to maintain the arc in the rectifier until the electromotive force of the supply has passed through zero, reverses and builds up to such a value as to cause A' to have a sufficiently positive value to start an arc between it and the mercury cathode B. The discharge circuit of the reactance coil E is now through the arc A'B, instead of through its former circuit. Consequently the arc A'B is now supplied with current, partly from the trans- former and partly from the reactance coil E. The new circuit from the transformer is indicated by the arrows inclosed in circles. The amount of reactance inserted in the circuit reduces the pulsations of the direct current sufficiently for all ordinary com- mercial purposes. Where it is advisable to still further reduce the ampli- tude of the pulsations, as, for instance, in telephone work, this is done with very slight reduction in efficiency by means of reactances. t© F E Fio.83. Rectifier Connections Shown Diagrammatically. WESTINGHOUSE MERCURY ARC RECTIFIER OUTFITS. 481 wssraveuaoijss: merccry arc rectifieh OUTFITS. For Arc Lamps* These outfits are a development of the constant current transformei adapted for use with the mercury rectifier, receiving alternating current at a constant potential, and delivering a constant direct current. By a special J^flQO.PAQ.Q.Qfl.ri"-- .QQ.Q.O.QQ.QQQQft. Fig. 84 and Pig. 85. Diagrams of Westinghouse Mercury Arc Rectifier. arrangement of coils the usual sustaining reactance is omitted, resulting in reduced floor space and an improved efficiency. A boiler iron tank with cast iron cover, two alternating currents and two direct currents leads, describes the simple and rugged appearance of an outfit. (See Fig. 84.) The connections (Fig. 85) explain the operation. P-P and S-S are respec- tively the primary and secondary; ST the starting transformer, R the rectifier, and A the auxiliary coil for exciting the starting transformer. «| — -mm i , . » mmmmm .smj-. < ' - 5 Fio. 86. The outfit is started by tipping the bulb, causing a spark between the terminals of the starting transformer as the current path through the mercury is interrupted. This breaks down the high resistance of the nega- tive electrode and permits the establishment of the direct current. The bulb is carried in a box which is easily slid in or out between guides to the bottom of the containing tank, thus making the bulb replacement a matter of but a few moments. Simple variable weights permit of adjusting the transformer so as to deliver its exact rated direct current (Fig. 86), at all loads. The power factor at full load averages over 70 per cent and the efficiency well over 90 per cent for all sizes of rectifier outfits. These are regularly built in 25, 35, 50, 75 and 100 light capacities, either 25 or 60 cycles, for 2200 V., 6600 V., 11,000 V., and 13,200 V. circuits. 482 THE STATIC TRANSFORMER. For Battery Charging:* Fig. 87. Westinghouse Mercury Arc Kectifier for Battery Charging. These outfits are intended to operate from low constant potential circuits and deliver a constant D. C. voltage, varying from 5 to 125 volts, according to design. Fig. 87 indicates a method of connec- tion which is essentially the same as for the arc lighting outfits. SR is a starting resistance, for the rectifier; MN, the auto- transformer, BB f the D. C. terminals, and A A ' the A. C. terminals. These outfits are started by tipping the bulb. A spark due to interrupting the current in the starting resistance breaks down the high negative electrode resist- ance, permitting the direct current to be established. In this outfit, like the arc outfit, a special arrangement of coils per- mits the omission of the usual sustaining coil. The D. C. voltage is varied by changes in the connection to the auto- transformer, or by changes in the A. C. impressed voltage made by an adjustable series reactance. Control panels carrying instruments, control dial, circuit breaker, etc., are furnished. Thirty amperes, 110 volts, is at present the maximum capacity for which these outfits are built, for either 25 or 60 cycle service. TRANSFORMER TESTING. Although the standard types of transformers of to-day are made on lines found by long experience to be the best for all purposes, and are subject to careful inspection and test at the factory in most cases, yet the various makers have such different ideas as to the value of the different points, that in order to obtain fair bids on such appliances when purchased, it is always best to prepare specifications, and the buyer should be prepared to conduct or check tests to determine whether the specifications have been fulfilled. Large stations should have a full outfit of apparatus for conduct- ing such tests ; but smaller purchasers can do quite well by having a compe- tent superintendent, or by hiring an outside engineer to witness the tests at the factory. It is not always necessary to put each individual transformer through all the tests, but the break-down test for insulation should be ap- plied to all. Prof. Jackson gives the following requirements for guaranties of trans- formers. Iron loss for 1000-volt transformers and for frequencies over 100 as follows : Capacity. Iron Loss. Exciting Current. 1000 watts 1500 watts 2000 watts 2500 watts 4000 watts 6500 watts 17500 watts 30 watts 40 watts 50 watts 60 watts 80 watts 100 watts 150 watts • .055 amperes. .080 amperes. . .150 amperes. .200 amperes. For frequencies less than 100 it may be advisable to allow 10 to avoid excessive cost. Note. year. higher loss Guaranties for iron loss should cover ageing for at least one TRANSFORMER TESTING. 483 Drop in secondary pressure not to exceed 3 % between no load and full load. Rise of temperature after 10 hours' run under full load, 70° F. (about 40° C). Note. — This measurement was probably meant by Professor Jackson to be made by thermometer. It is better to take the rise by resistance meas- urement, in which case the allowable temperature is 50° C. Disruptive strength of insulation after full-load run, between coils and between primary coil and iron, at least 10 times the primary volt- age. Insulation resistance to be not less than 10 megohms, and guaranteed not to deteriorate with reasonable service. Note. — See previous matter as to test voltage. Exciting- current for 1000-volt transformers not to exceed values given in the above table, when the frequency is above 100. The exciting current increases as the frequency decreases, and varies inversely as the voltage. For intermediate capacities proportional values may be expected. He further says : " Transformers which do not meet the insulation and heat- ing guaranties are unsafe to use upon commercial electric lighting and motor circuits, while those which do not meet the iron loss, regulation, and exciting current guaranties icaste the company's money." The characteristics of a transformer, to be determined by tests, are as follows : (1) Insulation strength between different parts. (2) Core loss and exciting current. (3) Resistances of primary and secondary and PR. (4) Impedance and copper loss, direct measurement. (5) Heating and temperature rise. (6) Ratio of voltages. (7) Regulation and efficiency, which may be calculated from the results of tests (2), (3), and (4), or may be determined directly by test. (8) Polarity. The instruments required to make these tests should be selected for each particular case, and consist of ammeters, voltmeters, and indicating watt- meters. For central station work, the following instruments will suffice for nearly any case which may come up in ordinary practice. A. C. Voltmeter, reading to 150 volts, and with multiplier to say 2500 volts. A. C. Ammeter, reading to 150 amperes, with shunt multiplier if necessary to carry the greatest output. Indicating wattmeter, reading to 150 or 200 watts. Note. — For full data and examples of transformer testing, see pamphlet No. 8126, " Transformer Testing for Central Station Managers," by Gen- eral Electric Company, and Westinghouse Pamphlet No. 7035. Insulation Test. This is the simplest and most important test to be made, for the reason that one of the principal functions of a transformer is its ability to thor- oughly and effectually insulate the secondary circuit from the primary circuit. Tests of the insulation of practically all high-potential apparatus are now carried out by high pressure, rather tnan by test of the insulation resistance by galvanometer. Some insulations will show a very high test by galva- nometer, but will fail entirely under test with a voltage much exceeding that at which it is to be used. On the other hand, it is not uncommon to find insulation such that, while the galvanometer tests show low resistance, it will not break down at all under the ordinary voltages. For this reason, it is common practice among manufacturers of transformers to apply a mod- erately high voltage, from two to three times the working voltage, for a short period, usually about one minute. The Committee on Standardization of the A. I. E. E. has given certain voltages which they recommend to be used in the testing of all electrical ap- paratus, and the tables and methods of application for the testing of trans- formers will be found in paragraphs Nos. 217 to 221, both inclusive in tiie 484 THE STATIC TRANSFORMER. latest revision of the rules of that Committee which will be found elsewhere in the book. In standard transformers these insulation tests should be (1) between pri- mary and secondary, and between primary and core and frame ; (2) between secondary and core and case. To obviate any induced potential strain, the secondary should be grounded while making the test between the primary and secondary, and between primary and core and case. In testing between the primary and secondary, or between the primary and core and frame, the secondary must be connected to the core and It is also important that all primary leads should be connected together as well as all secondary leads, in order to secure throughout the winding a uniform potential strain during the test. Note. — See index for sparking-gap curve, and use new needles after every discharge. From one point of view, the factor of safety of the secondary need not be greater than that of the primary, and if 10,000 volts is considered a sufficient test for a 2000-volt primary, 1000 voits might be sufficient for a 200-volt sec- ondary. But a thin film of insulation may easily withstand a test of 1000 volts, although it is so weak mechani- O CONNECT CALIBRATING VOLTMETER BETWEEN A AND B 104 OR 52 VOLT MAINS Fig. 88. cally as to be dangerous. A 200-volt secondary should therefore be tested for at least 2500 volts in order to guar- antee it against breakdown due to mechanical weakness. The duration of the insulation test may vary somewhat with the magni- tude of the voltage applied to the transformer. If the test is a severe one, it should not be long continued; for while the insulation may readily withstand the momentary applica- tion of a voltage five or ten times the normal strain, yet continued applica- tion of the voltage may injure the in- sulation and permanently reduce its strength. Attention has been called to the fact that in testing between the primary and the core or the secondary, the sec- ondary should be grounded. In test- ing between one winding and the core, for example, an induced potential 3train is obtained between the core and the other winding which may be much greater than the strain to which the insulation is subjected under normal working conditions, and greater therefore than it is designed to withstand. In testing between the primary and the core, the induced po- tential between the secondary and the core may be several thousand volts, ind the secondary may thus be broken down by an insulation test applied to the primary under conditions which do not exist in the natural use of the transformer. Attention is further called to the fact that during the test all primary leads as well as all secondary leads should be connected together. If only one terminal of the transformer winding is connected to the high potential transformer, the potential strain to which it is subjected may vary through- out the winding, and may even be very much greater at some point than at the terminals to which the voltage is applied. Under such conditions the reading of the static voltmeter affords no indication of the strain to which the winding is subjected. Indications which are best learned by experience reveal to the operator the character of the insulation under test. The transformer in test requires a charging current varying in magnitude with its size and design. From the reading of the ammeter, placed in the low potential circuit of the test- ing transformer, the charging current may be ascertained. It will increase as the voltage applied to the insulation is increased. If the insulation under test be good there will be no difficulty in bringing the potential up tc the desired point by varying the rheostat. If the insula- TRANSFORMER TESTING. 485 tion be weak or defective, it will be impossible to obtain a high voltage across it, and an excessive charging current will be indicated by the am- meter. Inability to obtain the desired potential across the insulation may be- the result merely of large electrostatic capacity of the insulation and the conse- quent high charging current required, so that the high potential trans- former may not be large enougn to supply this current at the voltage desired. A breakdown in the insulation will result in a drop in voltage indicated by the electrostatic voltmeter, an excessive charging current, and the burn- ing of the insulation if the discharge be continued for any length of time. Core JLoss and Exciting Current. In taking measurements of core loss and exciting current, the instruments required are a wattmeter, voltmeter, and ammeter. One of the two following described methods for connecting up the instru- ments is usually employed, although several others might be shown. These methods differ only in the way of connecting up the instruments, and are as follows : UEethocl 1. — The voltmeter and pressure coil of the wattmeter are con- nected directly to the terminals of the test transformer. When the pressure of the voltmeter is at the standard voltage the reading of the wattmeter will be the core loss in watts. It is evident from an inspection of diagram 89 that the wattmeter will indicate, in addition to the watts consumed by the test transformer, the I 2 R or copper loss in both the pressure coil of the wattmeter and voltmeter. This error, however, being constant for any pressure, is easily corrected. This method is very good for accurate results, and where the quantities to be measured are small it is most desirable. §P 111 — wvw ^J> VARIABLE RESISTANCE TEST TRAN8 Fig. 89. Core Loss (Method 1). Method 2. —The current coils of the wattmeter are inserted between a terminal of the test transformer and the terminal of the voltmeter and pressure coil of the wattmeter (see diagram 90). In this method the error introduced is the I 2 R loss in the current coil of the voltmeter. This is a very much smaller error than in Method 1, but does not allow of an easy or accurate correction, and the results obtained by it must, therefore.be taken without correction. For this reason Method 2 is more convenient, and for the measurement of large core losses, and for commercial purposes, it is sufficiently accurate. VARIABLE SWITCH RESISTANCE -NAAA/W- WATTMETER TEST TRANS Fig. 90. Core Loss (Method 2). Core losses and exciting current should be measured from the low-poten- tial side of the transformer to avoid the introduction of high voltage in the test. Notes on Core Loss and Excitation Current. In an ordinary commercial transformer, a given core loss at 60 cycles may consist of 70 per cent hysteresis and 30 per cent eddy current loss, while at 125 cycles the same transformer may have 55 per cent hysteresis loss and 45 per cent eddy current loss. 486 THE STATIC TRANSFORMER. The core loss is also dependent upon the wave form of the impressed E.M.F., a peaked wave giving somewhat lower core losses than a flat wave. It is not uncommon to find alternators having such a peaked wave form that the core loss obtained, if the transformer is tested with current from them, Avill be 5 per cent to 10 per cent less than that obtained if the trans- former is tested from a generator giving a sine wave. On the other hand, generators are sometimes obtained which have a very flat wave form, so that the core loss obtained will be greater than that obtained from the use of a sine wave. The magnitude of the core loss depends also upon the temperature of the iron. Both the hysteresis and eddy current losses decrease slightly as the temperature of the iron increases. It is well known that if the tempera- ture be increased sufficiently, the hysteresis loss disappears almost entirely, and since the resistance of iron increases with the temperature the eddy current losses necessarily decrease. In commercial transformers, an in- crease in temperature of 40° C. will cause a decrease in core loss of from 5 per cent to 10 per cent. An accurate statement of core loss thus necessi- tates that the temperature and wave form be specified. If, in the measurement of core loss, the product of impressed volts and excitation current exceeds twice the measured watts, there is reason to suspect poorly constructed magnetic joints or higher iron densities than are allowable in a well-designed transformer. Mea§urement of Resistance. Resistance of the coils can be measured by either the Wheatstone Bridge or Fall of Potential Method. For resistances below one or two ohms it is generally more accurate to use the Fall of Potential Method. Resistances should always be corrected for temperature, common prac- tice being to correct to 20° centigrade. For pure soft-drawn copper this cor- rection is .4 % per degree centigrade. Readings should be taken at several different current values, and the average value of all the readings will be the one to use. (See Index for correction for rise of temperature.) • Having obtained the resistance of the primary and secondary coils, the PR of both primary and secondary can be calculated ; the sum of the two being (very nearly) equal to the copper loss of the transformer. If it is preferred to measure the copper loss directly by wattmeter, then we must make test No. 4. The fall of potential method is subject to the following sources of error : (1) Witeh the connections as ordinarily made the ammeter reading includes the current in the voltmeter, and in order to prevent appreciable error the resistance of the voltmeter must be much greater than that of the resistance to be measured. If the resistance of the voltmeter be 1000 times greater, an error of ^ of 1 per cent will be introduced, while a voltmeter resistance 100 times the coil's resistance will mean the introduction of an error of 1 per cent. Correction of the ammeter reading obtained in (3) may thus become necessary, but whether or not it be essential will depend upon the accuracy desired. (See example below.) (2) The resistance of the voltmeter leads must not be sufficient to affect the reading of the voltmeter. (3) Since the resistance of copper changes rapidly with the temperature, the current used in the measurement should be small compared with the carrying capacity of the resistance, in order that the temperature may not change appreciably during the test. If a large current is necessary, read- ings must be taken quickly in order to obtain satisfactory results. If a gradual increase in drop across the resistance can be detected within the length of time taken for the test, it is evident that the current flowing through the resistance is heating it rapidly, and is too large to enable accu- rate measurement of resistance to be secured. It is quite possible to use a current of sufficient strength to heat the wind- ing so rapidly as to cause it to reach a constant hot resistance before the measurement is taken, thus introducing a large error in the results. Great care should be taken, therefore, in measuring resistance to avoid the use of more current than the resistance will carry without appreciable heating. (4) Considerable care is necessary to determine the temperature of the winding of the transformer. A thermometer placed on the outside of the winding indicates only the temperature of the exterior. The transformer TRANSFORMER TESTING. 487 should be kept in a room of constant temperature for many hours in order that the windings may reach a uniform temperature throughout. The surface temperature may then be taken as indicative of that of the interior. Impedance and Copper-IoM Test. method 1. — In this method, which was first described by Dr. Sumpner, the secondary coil is short-circuited through an ammeter. A wattmeter and a voltmeter are connected up in the primary circuit in a manner similar to either of the two methods described for the core-loss test. An adjustable resistance or other means for varying the impressed voltage is placed in series with the primary circuit. To make the test, the voltage is raised gradually until the ammeter shows that normal full-load current is flowing through the secondary circuit. Readings are then taken on the wattmeter and voltmeter. This method of measuring the impedance and copper loss of a transformer is now seldom used, on account of the liability to error due to the insertion of the ammeter in the secondary. In addition to being inaccurate, it usu- ally requires an ammeter capable of measuring a very heavy current. IfEetliod 2. — This method differs from Method 1 only in that the sec- ondary is short-circuited directly on itself, an ammeter being inserted in the primary circuit. The diagram of connections is shown in Fig. 91. In con- necting up the voltmeter and the potential coil of the wattmeter, the same corrections hold as in the measurement of core loss and exciting current, and connections made according to whether accuracy of results or simplicity of test is the more imporant. "I p| vwwv H^) WATTMETER Fig. 91. Impedance Test with Wattmeter. Having the readings of amperes, volts, and watts, we obtain from the first two the impedance of the transformer. This impedance is the geo- metrical sum of the resistance and reactance, and is expressed algebraically as follows : z = V IP + (2wnL)*> where z = Impedance, i?= Resistance, L = Coefficient of self-induction, /= Current in amperes, n = Frequency in cycles per second, 2?r n L = reactance of the circuit. In a test on a transformer with secondary short-circuited as in Fig. 91 above, and primary connected to 2000 volts, the impedance volts were 97 at full-load primary current of 2.5 amperes, then 97 Impedance = — = 38.8 ohms, and 97 X 100* Impedance drop = = 4.85 per cent. The reading on the wattmeter indicates the combined T 2 R of the primary and secondary coils, and in addition includes a very small core loss, which can be neglected, and an eddy current loss in the conductors. In standard lighting transformers, the impedance voltage varies from 2 per cent to 8 per cent. In making this test, careful record of the fre- quency should be made, as the impedance voltage will vary very nearly with the frequency. 488 THE STATIC TRANSFORMER. nMM/V 1 Ua/VW-' ^VWvV "-mam, A/S rAAAAAA- | rAM/W^i SECONDARY TRANSFORMER NO. 1 PRIMARY TO THREE-PHASE ALTERNATOR B PRIMARY TRANSFORMER NO. 2 C p/VWW- single-pha8e alternator Fig. 92. PRIMARY L > /^^ r l-VV\/VVV-' WVW\A/-' SECONDARY r V\/W"| r J WVWS f MWA 1 rAAAMAi ^ Law-T L^yvJ Uaaa-J i|ii lift lift ^^^p^p^ TO THREE-PHASE ALTERNATOR B TO THREE-PHASE " ALTERNATOR A Fig. 93. Figures 92 and 93 show a method of loading three-phase transformer* for heat test. TRANSFORMER TESTING. 489 Heat Vests. To test the transformer for its temperature rise, it is necessary to run it at full excitation and full-load current for a certain length of time. An eight-hour run at full load will usually raise the temperature to its highest point, and in the case of lighting transformers a full-load run very seldom continues longer than eight hours in practice. If it is desired to find just what is the final temperature rise under full load (as is often the case with transformers for power work) the transformer can be operated for two or three hours at an overload of about 25 %, after which the load should be reduced to normal, and the run continued as long as may be necessary. There are several methods for making heat runs of transformers, and all of them approximate the condition of the transformer in actual service. Heat Test, Uletliod 1. — The primary is connected to a circuit of the proper voltage and frequency, and the secondary loaded with lamps or resistance until full-load current is obtained. The temperature of all acces- sible parts should be obtained by thermometer, and the temperature rise of the coils determined by increase of resistance. Frequent readings should be taken during the run to see to what extent the transformer is heating. Heat Test, Tlethod 3. — Where the transformer is of large size, or sufficient load is not obtainable, the motor generator method of heat test is preferable. Two transformers of the same voltage, capacity and frequency are required, and are connected up as shown in Fig. 94. AMMETER -O-i A SWITCH TO CIRCUIT THIS VOLTAGE TO BE APPROX. TWICE THE IMPEDANCE VOLTAGE OF EACH TRANSFORMER, IT MUST BE ADJUSTED UNTIL FULL LOAD CURRENT .FLOWS IN. TRANSFORMERS. Fig. 94. NOTE: THI8 VOLTAGE TO BE THAT OF THE SECONDARY OF EACH TRANSFORMER The two secondaries are connected in parallel, and excited from circuit A at the proper voltage and frequency. The two primaries are connected in series in such a way as to oppose each other. The resultant voltage at B will be zero, however, because the voltage of the two primaries is equal and opposite. Any voltage impressed at B will thus cause a current to flow independent of the exciting voltages at the transformer terminals, and approximately twice the impedance voltage of one transformer will cause full-load current to flow through the primaries and secondaries of both transformers. The total energy thus required to run two transformers at full load is merely the losses in the iron and copper. Circuit A supplies the exciting current and core losses, and circuit B the full-load current and copper Heat Test, Ifletliod 3. —When only one transformer is to be tested, and this transformer is of large capacity, a modification of the motor gen- erator method can be used as described below : This method was first used in testing an 830 k.w. 25-cycle transformer made for the Carborundum Company of Niagara Falls. The connections are shown in Fig. 95. Both primary and secondary windings are divided into two parts, the pri- mary coils x and y being connected in multiple to the dynamo circuit, but an auxiliary transformer capable of adding a few per cent E.M.F. to that half of the primary is connected as shown in the y half. 490 THE STATIC TRANSFORMER. By this means the primary coils are properly magnetized, and full-load currents can be passed through them by varying the auxiliary E.M.F. The two halves of the secondary coils are connected in series in opposi- tion to each other, and are subject to an auxiliary E.M.F. from the same generator, but reduced to the proper voltage by the auxiliary trans- former B. The currents were measured in all three transformer circuits, and the E.M.F. of one-half the secondary was measured. The method is accurate enough for large units, and is quite handy where no large dynamo can be gotten for supplying full-load currents, as in this case current is required only for the transformer losses and for supplying the auxiliary transformers. 1 DYNAMO y X b a .OCOOQ i II! I ! onrai Wfr oooo.ou PRIMARY in , 1 SECONDARY Fig. 95. General Electric Method of Testing One Large Transformer. Fig. 96 shows connections for heat-run on three single-phase trans- formers, or one three-phase transformer. The primaries and secondaries are connected in delta, and in one corner of the primary impedance vol- TO ALTERNATOR SUPPLYING COPPER LOSSES TO THREE-PHASE ALTERNATOR SUPPLYING CORE LOSS Fig. 96. tage for the three transformers connected in series is impressed. The current circulates in the delta connections and is entirely independent of the secondary voltage. The method outlined above requires only power enough to supply the losses. TRANSFORMER TESTING. 491 Temperature Rise. To ascertain the temperature rise of the different parts of a transformer, thermometers are placed on the various parts, and readings taken at fre- quent intervals. These readings, however, indicate only. the surface tem- perature oi a body and not the actual internal temperature. The average rise of temperature of the windings can be more accurately determined by means of the increase of resistance of the conductor, and is determined by knowing the resistances hot and cold. Let Re =r resistance of one coil, cold. Rh — resistance of one coil, hot. Tc =. temperature of one coil in cent, degrees, cold. Th = temperature of one coil in cent, degrees, hot. K=z temperature of coefficient of copper .004. _ i?ft(l + .0047 T c ) — Re h — .OOiRc This equation is based on the assumption that the resistance of pure cop- per increases .4 % of its value at zero for every degree centigrade rise in temperature. If it be desired to know the temperature rise of both primary and second- ary coils, their hot and cold resistances must be determined separately ; but it is customary to determine the temperature rise by resistance of only one coil, usually the primary, and comparing the secondary temperatures by the thermometer measurements. The method for taking these measurements is described in the paragraph in this section on measurement of resistance. Ratio. As a check against possible mistakes in winding the coils and connecting up. a test should be made for ratio of voltages. The ratio test is made at a fractional part of the full voltage at no-load current, and should not be substituted for a regulation test. An error of one or two per cent is quite admissible in making this test, because of its being taken at partial voltages. Reg- ulation. The regulation of a transformer can be determined either by direct meas- urement or by calculation from the measurements of resistance and reac- tance in the transformer. Since the regulation of any commercial trans- former is at the most but a few per cent of the impressed voltage, and as errors of observation are very liable to be fully one per cent, the direct method of measuring regulation is not at all reliable. Regulation by Direct measurements. Connect up the transformer with a fully loaded secondary, as in Fig. 97. If the primary voltage is very steady, voltmeter No. 2 only will be neces- sary, but it is better to use one on the primary circuit also as shown. A w.tt WM. 3 2 5 o *M* LAMP LOAD Fig. 97. Test for Regulation of Transformer. reading of voltmeter No. 2 is taken with no load, and again with load, the difference in the two readings being the drop in voltage on the secondary. We, therefore, have, ~ T:> . ,. ,__ /100 X Reading at full load"* % Regulation = 100 — ( =: r . ~ : > \ Reading at no loaa / 492 THE STATIC TRANSFORMER. Reg-iilation Uy Calculation. Several methods of calculating the regulation of transformers from the measurements of resistance and reactive drop have been devised. Below is a method of Mr. A. R. Everest, which has been found to answer the requirements of daily use. Let 1R = Total resistance drop in transformer expressed as per cent of rated voltage. IX = Reactive drop, similarly expressed. P = Proportion of energy current in load or power factor of load. For non-inductive load P = 1. W = Wattless factor of primary current. (With non-inductive load, W = Magnetizing current expressed as a fraction of full-load current. With inductive load, W = Watt- less component of load, plus magnetizing current.) Then if volts at secondary terminals = 100%, Primary voltage — For Hon-Inductive Load: E = V(ioo + PIR + fT/X) a + (ZX) 2 . Cor Inductive Load : E = V (100 + PIR + WIXy + (PIX - WIR) 2 . In each of these equations the last expression within parentheses repre- sents the drop " in quadrature." ™ . . i/7^ 3 T^ /Core loss\2 The magnetizing current = y (Exciting current J — ( - J . For frequencies of 60 cycles or higher, magnetizing current may be taken as 75 per cent of the exciting current. Extracting the square root in the expression for regulation may be avoided in the use of the following table : Quadrature Drop. Increase in Primary Voltage. 2.5 per 3 3.5 " cent. .025 .04 .06 per ii cent. 4 4.5 " " .08 .10 ti it ii ii 5 5.5 " .13 .15 ft 6 6.5 " « .18 .21 ii u ii 7 7.5 « .24 .27 it ii II ii 8 8.5 •« 11 .31 .35 14 il II 9 9.5 " << it .39 .45 ii II ii 10 " .50 ii " EFFICIENCY. 493 As an example, take a 2 k.w. transformer having the following losses: IR drop = 2%. IX drop = 3.5%. Exciting current = 4% or .04 ; then magnetizing current = 75% of this, or .03. 1. 3f on- Inductive Load. — Secoudary voltage = 100%. Primary voltage in phase = 100 + 2% + (.03 X 3.5%) = 102.1%. Quadrature drop = 3.5% ; this from table adds .06% of total primary volt- age - 102.16%. 2 16 The drop is 2.16% of secondary voltage, or ' = 2.11% of primary volt- 102.1o age, which is the true regulation drop. 2. Inductive Load. — With a power factor of .86, wattless factor of load = .5, and adding magnetizing current (which in most cases might be neglected on inductive load), W becomes .52. The primary voltage in phase is now 100% -f- (2% X .86) + (3.5 X .52) = 103.54%. The quadrature drop is (.86X3.5%) -(.52X2%) = 1.97. From the table 1.97% adds .02% to primary voltage or 103.54 + .02%= 103.56. Primary voltage = 103.56 3.56 Regulation drop = ' = 3.43% of primary voltage. Regulation drop should always be expressed finally in terms of primary voltage. The above-described methods of traasformer testing are in use by one of the large manufacturers, and present average American shop practice. The following matter is largely from the important paper by Mr. Ford and presents the commonest theoretical test methods. EFFICIENCY. The efficiency of a transformer is the ratio of its net power output to its gross power input, the output being measured with non-inductive load. The power input includes the output together with the losses which are as follows : (1) The core loss, which is determined by test at the rated frequency and voltage. (2) The P R loss of the primary and the secondary calculated from their resistances. Example. Transformer, Type H, 60 Cycles, 5 k.w., 1000-2000 Volts Prim., 100-200 Volts Sec. Amperes. Primary, at 2000 volts 2.5 Secondary, at 200 volts . 25 Resistance. Ohms at 20° C. Primary 10.1 Secondary 0.067 At Full Load. Losses. Watts. Primary PR 63 Secondary I 2 R 42 Total PR 105 Core Loss 70 Total Loss 175 Output at Full Load 5000 Input " " " 5175 Efficiency 5000/5175 or 96.6% 494 THE STATIC TRANSFORMER. At Half Load. Losses. Watts. Total PR 26 Core Loss 70 Total Loss 96 Output 2500 Input 2596 Efficiency 2500/2596 or 96.2% The all-day efficiency of a transformer is the ratio of the output to the input during 24 hours. The usual conditions of practice will be met if the calculation is based on 5 hours at full load, and 19 hours at no load. Output. Watt Hrs. 5 Hours at Full Load 25000 19 Hours at No Load Total, 24 Hours 25000 Input. 5 Hours at Full Load 25875 19 Hours at No Load (Neglecting PR Loss due to Excitation Current) 1330 Total, 24 Hours 27205 All-day Efficiency 25000/27205 or 91.9% In calculating the efficiencies in both of the above examples, the copper loss due to excitation current of the transformer has been neglected. This current, in the example given above, is less than 3%, and its effect on the loss of the transformer is thus negligible. Even at no load the total P R loss introduced by it is less than one watt. It is quite necessary, however, that the loss introduced by the excitation current should be checked in all cases. In some transformers, for example, the excitation current may reach 30 % of the full-load current, and thus its effect is noticeable at large loads, while at \ load the loss in the primary winding due to excitation current is greater than the loss due to the load current. Inasmuch as the losses in the transformer are affected by the tempera- ture and the wave form of the E.M.F., the efficiency can be accurately specified only by reference to some definite temperature, such as 25° C, and by stating whether the E.M.F. is sine or otherwise. The foregoing method of calculating the efficiency neglects what are known as " load losses," i.e., the eddy current losses in the iron and the conductors caused by the current in the transformer windings. The watts measured in the impedance test include " load losses " and I 2 R losses to- gether with a small core loss. Considering the core loss as negligible, the " load losses " are obtained by subtracting from the measured watts the PR loss calculated from the resistance of the transformer. It is sometimes assumed that the " load losses" in a transformer when it is working under full-load conditions are the same as those obtained with short-circuited secondary, and it is stated that these losses should enter into the calcula- tion of efficiency. Many tests have been made to determine whether or not the above assumption is correct, and while the results cannot be considered as conclusive, they indicate in every case that, under full-load conditions, the "load losses" are considerably less than those measured with short- circuited secondary. Inasmuch as these losses, in general, form a small percentage of the total loss in a transformer, and in view of the difficulty in determining them with accuracy, they may be neglected in the calcula- tion of efficiency for commercial purposes. The measurement of watts in the impedance test is, however, useful as a check on excessive eddy current losses in a poorly designed transformer. DATA TO BE DETERMINED BY TESTS 495 JPOJLAMTY. For lighting and other small uses, transformers are generally designed so that the instantaneous direction of flow of the current in certain selected leads is the same in all transformers of the same type. For example, re« ferring to Fig. 98, the transformer there shown is de- signed so that the current at any instant flows into the primary at " A " and into the secondary at " C." This is the system adopted for small transformers by the ma- jority of manufacturers. The polarity test should be unnecessary when bank- ing transformers of the same type and design. When, however, transformers manufactured by different com- panies are to be run in parallel, it is necessary to test them in order to avoid the possibility of connecting them in such a way as to short circuit the one on the other. Their polarity may be determined by one of the following methods: In Fig. 98, Primary lead " A " is of opposite polarity to the Secondary lead " C." Connect the primary lead " A " to the Secondary lead " C." Apply one hundred volts, say to the primary " A-B " of the transformer. The voltage measured from " A " to " D " will be greater than the applied voltage. In other words, a transformer connected as shown will act as a booster to the voltage. If the leads " A " and *' B " are of the same polarity, voltage measured from " A " to " D " will be less than that applied at " A-B." If a standard transformer known to have correct polarity and the same ratio as the test transformer is available, the simplest method for testing the polarity is to connect the primaries and secondaries of the transformers in parallel, placing a fuse in series with the secondaries. On applying voltage to the primaries of the transformers if they are of the same polarity and ratio no current should flow in the secondary circuit and the fuse will remain intact. If the transformers are of opposite polarity the connection will short circuit the one transformer on the other, and the fuse selected should therefore be small enough to blow before the transformers are injured. In nearly all transformers there will be a slight current in the secondaries when connected as above. This current is known as the " exchange current " and should be less than 1 per cent of the normal full load current of the transformer. Transformers of large capacity and higher voltage for central station work usually have a polarity opposite that shown in Fig. 98. There is, however, no standard for these transformers. DATA TO BE I>ETJEI*:WLi:VE» I** TJESTS. Partly from a paper by Arthur Hillyer Ford- B. S. I. Copper loss-, to determine the efficiency. II. Iron-core loss, hot and cold, to determine the efficiency : to separate the hysteresis from the foucault current loss. If JF:= watts output, /= watts iron-core loss, C= watts copper loss, . then the Efficiency = 100 - ( y+ ^ +c X 100 ) Foucault currents loss should decrease with an increase in tempera- ture. Hysteresis loss is supposed to be constant regardless of heat. III. Open circuit or exciting current. IV. Regulation, to determine the magnetic leakage. V. Rise in temperature in case and out of case, for no load and full load ; with and without oil. VI. Insulation. 496 THE STATIC TRANSFORMER. Methods. Opposition Method of Avrton and Sumpner. — This method is especially valuable where the transformers to be tested are of large ca- pacity, and a source of power great enough to put them under full load in the ordinary way is unavailable. A supply of current of an amount some- what greater than the total losses of both transformers is all that is neces- sary. Following is a diagram of the connections, by which it will be seen that the transformers are so connected that one feeds the other, or they work in opposition. Fig. SMALL TRANSFORMER Diagrams of Connections for Ayrton and Sumpner Opposition Method of Testing Transformers. In making the test, current is turned on and the resistance R adjusted until full-load current flows in the secondary, as shown by the ammeter A, and the primary current and voltage in A and V is up tojstandard. Then the watts read on W are equal to all the losses in both transformers, and W r ! the losses in the copper of the transformers plus the copper loss in the leads and in the cur- rent coils of Wi and A. The iron loss in both transformers is = W— Wi — A, where A is the loss in the leads and instruments which may be calculated by PR. Method of I>r. Sumpner. Iron liOss. — The following diagram shows the connections for Dr. Sumpner's test for iron losses. The low- \Tl . AOJUSTABL J^VJ I RESISTANC s 1 » R. T fi I Fig. 100. Dr. Sumpner's Test for Iron Losses DATA TO BE DETERMINED BY TESTS. 497 pressure side is connected to a source of current of the same pressure at which the transformer is expected to work, thus producing the same pri- mary voltage in the high-pressure side at which it is expected to work. With the primary circuit open, the iron losses in the transformer are read directly in watts on the wattmeter. Copper Lois. — The next diagram shows the connections for determin- ing the copper losses. The low-pressure side is short-circuited through an ammeter, the high-pressure side being connected to the 100-volt supply- mains. The resistance R is then adjusted to obtain full-load or any other desired current in the secondary, as shown by the ammeter. The reading of the wattmeter will then show the total copper losses in the transformer and in the ammeter plus a very small and entirely negligible iron loss. The ammeter losses and that in the leads may be calculated by 1 2 R. The small iron loss can be separated or determined by disconnecting the ammeter and (Enr Fig. 101. Dr. Sumpner's Test for Copper Losses. adjusting R, until the pressure on the primary is the same as in the copper loss test; the wattmeter will then show the small iron loss. The iron loss is proportional to (ft 1 * 6 and (ft the magnetic density is pro^ portional to the pressure at the terminals of the transformer, therefore the iron loss is equal to JjT.tft 1 - 8 where K is a constant and (ft the voltage. In the iron-loss test the (ft = 1000 and in the copper loss test (ft = 100. K X 10001- 6 = 63,000 K JTX100 1 - 6 = 1,600 K = 2.5% of total iron loss. Heating 1 . — Tests should be made at no load, at full load, and at inter- mediate loads for rise of temperature of the transformers out of their cases, in their cases, without oil and with oil, if full data is wanted. If a strictly commercial test is all that is necessary, a test with the transformer at full load and set up in the condition it is to be run, will be sufficient. Surface temperatures can be taken by thermometers laid on and covered with cotton waste. In oil-insulated transformers, the temperature of the oil should be taken in two places, — inside the coil, and between the coil and case. Leakage Drop. — The drop in the secondary due to magnetic leakage can be found by deducting from the measured total drop in the PR drop due to the resistance of the coil. 498 THE STATIC TRANSFORMER. SPECIFIC AIIOI i* 1 Oil TRAiyiFORMERS. It is almost impossible to enumerate the features to be included in speci- fications covering transformers, because of the wide range of operation and service to which they may be put, necessitating different characteristics for the transformers intended for different kinds of services. For transformers operating from a fairly expensive source of supply, the leading manufacturers have decided on characteristics which, in general, will be covered in the following tabulation. This gives average characteristics of transformers designed for operation on 60-cycle circuits, and the figures given are based on operation of 2000 volts and sine wave alternator. Capacity. Core Loss Watts. Copper Loss Watts. Exciting Cur- rent^,. Regulation % 1 35 30 9.0 2.8 2 45 50 7.0 2.5 3 55 70 3.0 2.3 5 70 105 2.5 2.2 7.5 100 150 2.3 2.2 10 120 180 2.3 2.0 15 155 275 2.2 1.8 20 185 300 1.5 1.7 30 235 475 1.2 1.5 50 335 675 1.0 1.3 AGEOG, Guarantees against serious ageing of iron should cover a period of at least one year. RISE OF TEHPERAT1JRE. The rise of temperature should be referred to the standard conditions of a room-temperature of 25° C, a barometric pressure of 760 mm. and normal conditions of ventilation; that is, the apparatus under test should neither be exposed to draught nor inclosed, except where expressly specified. If the room temperature during the test differs from 25° C. the observed rise of temperature should be corrected by i per cent for each C. Thus with a room temperature of 35° C. the observed rise of temperature has to be decreased by 5 per cent, and with a room temperature of 15° C. the observed rise of temperature must be increased by 5 per cent. The ther- mometer indicating the room temperature should be screened from thermal radiation emitted by heated bodies, or from draughts of air. When it is impracticable to secure normal conditions of ventilation on account of adjacent engine or other sources of heat, the thermometer for measuring the air temperature should be placed so as fairly to indicate the tempera- ture which the machine would have if it were idle, in order that the rise of temperature determined shall be that caused by the operation of the machine. The temperature should be measured after a run of sufficient duration to reach practical constancy. This is usually from six to eighteen hours according to the size and construction of the apparatus. It is permissible, however, to shorten the time of the test by running a lesser time on an overload in current and voltage, then reducing the load to normal, and maintaining it thus until the temperature has become constant. In electrical conductors, the rise of temperature should be determined LOCATION OF TRANSFORMERS. 499 by the increase of their resistance where practicable. For this purpose the resistance may be measured either by galvanometer test, or by drop of potential method. A temperature coefficient of 0.42 per cent per degree C. from and at 0° C. may be assumed for copper, by the formula: Rt = R (1 + 0.0042 and Rt + = R [1 - 0.0042 (* + 0)] where Rt = the initial resistance at room temperature t° C. Tt + = the final resistance at temperature elevation 0°- C. Ro = the inferred resistance at 0° C. These combine into the formula: 'Tt + - 1> 0= (238.1+0 ( T ' + n 1 )°C- For insulation test see report of Committee on Standardization of A. I. E. E., page 514. LOCATION OF TRA^§rORHERS. 1. Where practicable, the transformers should be placed in a boiler iron case, capable of withstanding an internal pressure of 50 lbs. per square inch, the case to be suitably vented. 2. Where a sheet iron construction is necessary, the case should be made practically air tight and provided with a very large safety valve, so that an internal explosion cannot burst the case. 3. Provision should be made for rapidly drawing off the oil in case it becomes necessary to do so. 4. Individual transformer units, or groups of units, should be located in fireproof compartments, such compartments to be suitably drained so that in case the oil escapes from the cases, it can flow out where it can do no harm. 5. Adequate means should be provided for extinguishing fire, and the station attendants should be trained to know what to do in case of emer- gency. An oil should be selected which has a flash point not lower than, say, 175° C. Such an oil, if properly made, will have practically no evapora- tion whatever at 100° C, this temperature being higher than will be found except under the most extreme conditions of temporary overload. Too high a flash test oil is undesirable on account of the viscosity being so great that the power to carry heat from the transformer to the cooler case is greatly reduced, and on account of it being very unpleasant to handle. Where rubber-covered leads are used, the rubber should be heavy (not less than \" wall per 10,000 volts) and of high quality, and a fireproof covering should be used. Extra flexible cable is usually preferable. Rub- ber may be tested for dielectric strength, insulation resistance, etc., but its qualification for important uses is best judged by its mechanical proper- ties. To examine these, remove the braid from the wire for several inches, but without cutting the rubber except at the ends of the space. Here it should be cut (at both ends) down to the wire. It will be found in many makes that there are two joints in the rubber running parallel to the wire. A longitudinal cut along the wire, and down to it, should be made midway between the joints. This will make it possible to easily remove the rubber from the wire. First, test each of the joints by bending them over back- wards. The best joints will show some tendency to open, and for this reason a double layer of rubber, with joints staggered, is desirable. In many (so called) first class wires it will be found that the joints are just slightly stuck together, or break open on the slightest provocation. Such insulation is worthless. The quality of the rubber may be judged by cut- ting long strips, about \" wide, or less, and bending it double and as short as possible. It should show no signs of cracking. Pure rubber is very elastic and strong, and it loses these properties in proportion as it is adul- terated. 500 THE STATIC TRANSFORMER. 6PECIFICATIO]V§ FOR THAPI ORMER Oil* (C. E. Skinner.) In the following will be found a brief specification for a transformer oil. (1) The oil should be a pure mineral oil obtained by fractional distil- lation of petroleum unmixed with any other substances and without sub- sequent chemical treatment. (2) The flash test of the oil should not be less than 180° C. (356° F.), and the burning test should not be less than 200° C. (392° F.). (3) The oil must not contain moisture, acid, alkali, or sulphur com- pounds. (4) The oil should not show an evaporation of more than 0.2% when heated at 100° C. for eight hours. (5) It is desirable that the oil be as fluid as possible and that the color be as light as can be obtained in an untreated oil. The method of making tests to determine the above qualities should be distinctly specified so that there can be no misunderstanding on account of results being obtained by different methods of test. The specification for flash test given above is intended to be low enough so that there will be some leeway to allow for slight variations in the oil and for variations obtained by different observers. It is expected that an oil to fulfill this specification will run something higher than 180° flash test. STANDARDIZATION RULES OP THE AMERI- CAN INSTITUTE OP ELECTRICAL ENGINEERS. (Approved by the Board of Directors, June 27, 1911,) OEA'ERAL PL4\. I. Definitions and Technical Data. A. Definitions — Currents and E.M.F.'s. B. Definitions — Rotating Machines. C. Definitions — Stationary Induction Apparatus. D. General Classification of Apparatus. E. Motors — Speed Classification. F. Definitions — Instruments. G. Definition and Explanation of Terms. (I) Load Factor, Diversity Factor, Demand Factor. (II) Non-inductive and Inductive Load. (III) Power Factor and Reactive Factor. (IV) Saturation Factor. (V) Variation and Pulsation. II. Performance Specifications and Tests. A. Rating. B. Wave Shape. C. Efficiency. (I) Definitions. (II) Measurement of Efficiency. (III) Measurement of Losses. (IV) Efficiency of Different Types of Apparatus. (a) Direct -Current Commutating Machines. (b) Alternating-Current Commutating Machines. (c) Synchronous Commutating Machines. (<2) Synchronous Machines. (e) Stationary Induction Apparatus. (/) Rotary Induction Apparatus. (g) Unipolar or Acyclic Machines. (h) Rectifying Apparatus. (i) Transmission Lines. 0") Phase-Displacing Apparatus. D. Regulation. (I) Definitions. (II) Conditions for and Tests of Regulation. E. Insulation. (I) Insulation Resistance. (II) Dielectric Strength. (a) Test Voltages. (b) Methods of Testing. (c) Methods for Measuring the Test Voltage. (d) Apparatus for Supplying Test Voltage. F. Conductivity. G. Rise of Temperature. (I) Measurement of Temperature. (a) Methods. (b) Normal Conditions for Tests. 500a 500b standardization rules. (II) Limiting Temperature Rise. (a) Machines in General. (b) Rotary Induction Apparatus. (c) Stationary Induction Apparatus. (d) Rheostats. (e) Limits Recommended in Special Gases. H. Overload Capacities. III. Voltages and Frequencies. A. Voltages. B. Frequencies. IV. General Recommendations. V. Appendices and Tabular Data. Appendix A. — Notation. Appendix B. — Railway Motors. (I) Rating. (II) Selection of Motor for Specified Service. Appendix C. — Photometry and Lamps. Appendix D. — Sparking Distances. _ Appendix E. — Temperature Coefficients. Appendix F. — Horse Power. STANDARDIZATION RULES OP THE AMERI- CAN INSTITUTE OP ELECTRICAL ENGINEERS. (As Approved June 27, 1911.) I. OJEJFiA aXIOVS AUD TECHNICAL DATA. 1. Note. The following definitions and classifications are intended to be practically descriptive and not scientifically rigid. A. DEFINITIONS. CURRENTS AND E.M.F.'S. 2. A Direct Current is an unidirectional current. 3. A Continuous Current is a steady, or non-pulsating, direct current. 4. A Pulsating Current is a current equivalent to the superposition of an alternating current upon a continuous current. 5. An Alternating Current or E.M.F. is a current or E.M.F. which, when plotted against time in rectangular coordinates, consists of half-waves of equal area in successively opposite directions from the zero line. 5a. Cycle. Two immediately succeeding half-waves constitute a cycle. 5b. Period. The time required for the execution of a cycle is called a period. 5c. Frequency. The number of cycles per second is called the frequency. 5d. Wave-Form. The shape of the curve of E.M.F. or current plotted against time in rectangular coordinates is ordinarily referred to as the wave- form or wave-shape. Two alternating quantities are said to have the same wave-shape if their corresponding phase ordinates bear a constant ratio. The wave-shape, as ordinarily understood, is thus independent of the scales to which the curve is plotted. be. Simple Alternating Wave. Unless otherwise specified an alternating current or E.M.F. is assumed to be sinusoidal, and the wave a sinusoid, sine- wave or curve of sines. On this account a complete cycle is taken as 360 degrees, and any portion of a cycle may be expressed in degrees from any convenient reference point, such as the ascending zero-point. 5/. A Complex Alternating Wave is a non-sinusoidal wave. A complex alternating wave is capable of being resolved into a single sine wave of funda- mental frequency, with superposed odd-frequency harmonic waves, or ripples, of 3, 5, 7 ... (2 n + 1) times the fundamental frequency, each harmonic having constant amplitude, and a definite starting phase-relation to the fun- damental sine-wave. It is customary when analyzing a complex wave to neglect harmonics higher than the 11th: i.e., of frequency higher than 11 time3 the fundamental. In special cases, however, frequencies still higher may have to be considered. In certain exceptional cases even harmonics are present. bg. Root-Mean-Square Value (sometimes called the Virtual or Effective Value). Unless otherwise specified, the rating of an alternating-current or E.M.F., in amperes or volts, is assumed to be the square root of the mean square value taken throughout one or more complete cycles. This is some- times abbreviated to r.m.s. The term root-mean-square is to be preferred to the terms virtual or effective. The root-mean-square value is indicated by all properly calibrated alternating-current voltmeters and ammeters. In the case of a sine-wave, the ratio of the maximum to the r.m.s. value is v2. bh. Form-Factor of an Alternating Wave. The ratio of the root-mean- square to the arithmetical mean ordinate of a wave, taken without regard to sign, is called its form-factor. The form-factor for a purely rectangular wave is the minimum, 1.0; for a sine-wave it is 1.11, and for a wave more peaked than a sine-wave it is greater than 1.11. bi. The Equivalent Sine-Wave is a sine-wave having the same frequency and the same r.m.s. value as the actual wave. bj. The Deviation of wave-form from the sinusoidal is determined by^ super- posing upon the actual wave (as determined by oscillograph), the equivalent sine-wave of equal length, in such a manner as to give the least difference, and then dividing the maximum difference between corresponding ordinates by the maximum value or the equivalent sine-wave. bk. Phase Difference. When corresponding cyclic values of two sinu- soidal alternating quantities such as two alternating currents or E.M.F.'s or of a current and an E.M.F., of the same frequency, occur at different instants, 501 i 502 STANDARDIZATION RULES. the two alternating quantities are said to differ in phase, their phase difference being the time interval, expressed in degrees or as a fraction of a cycle, between the occurrence of their corresponding values; e.g., their ascending zeros or their positive maxima. bl. Equivalent Phase Difference. If two alternating quantities are non-sinusoidal, and of different wave shapes, the preceding definition of phase- difference is inapplicable, and phase-difference ceases to have exact signifi- cance. However, when the two complex alternating quantities are the voltage E and current / in a given circuit, the effective power P of which is known, it is customary to define the equivalent phase difference by the angle whose cosine is the power-factor, P /EI, of the circuit. See Sections 54 and 324. bm. Single-Phase. A term characterizing a simple alternating-current circuit energized by a single alternating E.M.F. Such a circuit is usually supplied through two wires. The currents in these two wires counted posi- tively outwards from the source, differ in phase by 180 degrees or half a cycle. 5m. Three-Phase. A term characterizing the combination of three cir- cuits energized by alternating E.M.F.'s which differ in phase by one-third of a cycle; i.e., 120°. bo. Quarter-Phase, also called Two-Phase. A term characterizing the combination of two circuits energized by alternating E.M.F.'s whichldiffer in phase by a quarter of a cycle; i.e., 90°. bp. Six-Phase. A term characterizing the combination of six circuits ener- gized by alternating E.M.F.'s which differ in phase by one-sixth of a cycle; i.e., 60°. bq. Polyphase is the general term applied to any alternating system with more than a single phase. 6. An Oscillating Current is a current alternating in direction, and of de- creasing amplitude. B. DEFINITIONS. ROTATING MACHINES. 7. A Generator transforms mechanical power into electrical power. 8. A Direct-Current Generator produces a direct current that may or may not be continuous. 9. An Alternator is an alternating-current generator, either single-phase or polyphase. 9a. A Synchronous Alternator comprises a constant magnetic field and an armature delivering either single-phase or polyphase current in synchronism with the rotation of the machine. 10. A Polyphase Generator produces currents differing symmetrically in phase; such as quarter-phase currents, in which the terminal voltages of the two circuits differ in phase by 90 degrees: or three-phase currents, in which the terminal voltages of the three circuits differ in phase by 120 degrees. 11. A Double-Current Generator supplies both direct and alternating currents from the same armature winding. 11a. An Inductor Alternator is an alternator in which both field and arma- ture windings are stationary. 116. An Induction Generator is a machine structurally identical with an induction motor, but driven above synchronous speed as an alternating- current generator. 12. A Motor transforms electrical power into mechanical power. 12a. A Direct-Current Motor transforms direct-current power into me- chanical power. 126. An Alternating-Current Motor transforms alternating-current power into mechanical power. 12c. A Synchronous Motor is a machine structurally identical with a syn- chronous alternator, but operated as a motor. 12d. A Synchronous Phase Modifier, sometimes called a Synchronous Con- denser, is a synchronous motor, running either idle or under load, whose field excitation may be varied so as to modify the power-factor of the circuit, or through such modification to influence the voltage of the circuit. 12e. An Induction Motor is an alternating-current motor, either single-phase or polyphase, comprising independent primary and secondary windings, one of which, usually the secondary, is on the rotating member. The secondary winding has no conductive connection with the supply circuit. 12/. A Repulsion Motor is an induction motor, usually single phase, in which the magnetic axis of the secondary (a closed coil winding mounted on the rotor) is maintained at a certain fixed angle with respect to the stationary DEFINITIONS AND TECHNICAL DATA. 503' primary coil by means of a multisegmental commutator and short-circuiting brushes. 12a. A Single-Phase Series Commutator Motor is structurally similar to a series direct-current motor, except that it is usually provided in addition with a series compensating winding distributed around the outer air-gap periphery and supported in slots in the pole faces, for the purpose of dimin- ishing the armature leakage reactance. 13. A Booster is a machine inserted in series in a circuit to change its volt- age. It may be driven by an electric motor (in which case it is termed a motor- booster) or otherwise. 14. A Motor-Generator is a transforming device consisting of a motor mechanically connected to one or more generators. 15. A Dynamotor is a transforming device combining both motor and gen- erator action in one magnetic field, either with two armatures, or with one armature having two separate windings and independent commutators. 16. A Converter is a machine employing mechanical rotation in changing electrical energy from one form into another. A converter may belong to either of several types, as follows: 17. a. A Direct-Current Converter converts from a direct current to a direct current, usually with a change of voltage. 18. b. A Synchronous Converter (commonly called a rotary converter) con- verts from an alternating to a direct current, or vice versa. 19. c. A Motor-Converter is a combination of an induction motor with a synchronous converter, the secondary of the former feeding the armature of the latter with current at some frequency other than the impressed frequency; i.e., it is a synchronous converter concatenated with an induction motor. 20. d. A Frequency Changer converts the power of an alternating-current system from one frequency to another, with or without a change in the number of phases or in the voltage. 21. e. A Rotary Phase Converter converts from an alternating-current system of one or more phases to an alternating-current system of a different number of phases, but of the same frequency. 21a. Equalizing Connections are low resistance connections between equi- potential points of multiple-wound closed-coil armatures to equalize the in- duced voltage between brushes. C. DEFINITIONS. STATIONARY INDUCTION APPARATUS. .22. Stationary Induction Apparatus changes electric energy to electric energy through the medium of magnetic energy. It comprises several forms, distinguished as follows: 23. a. Transformers, in which the primary and secondary windings are insulated from one another. 23a. A Primary Winding is that winding of a transformer, or of an induc- tion motor, which receives power from an external source. 236. A Secondary Winding is that winding of a transformer, or of an in- duction motor, which receives power from the primary by induction. Note. The terms "High-voltage winding" and "Low-voltage winding" are suitable for distinguishing between the windings of a transformer, where the relations of the apparatus to the source of power are not involved. 24. b. Auto-Transformers, also called compensators, in which a part of the primary winding is used as a secondary winding, or conversely. 25. c. Potential Regulators, in which one coil is in shunt and one in series with the circuit, so arranged that the ratio of transformation between them ia variable at will. They are of the following three classes: 26. (1) Contact Voltage Regulators, also called Compensator Regulators, in which the number of turns in use of one of the coils is adjustable. 27. (2) Induction Potential Regulators in which the relative positions of the primary and secondary coils are adjustable. 28. (3) Magneto Potential Regulators in which the direction of the mag- netic flux with respect to the coils is adjustable. 29. d. Reactors or Reactance Coils, also called choke coils, are a form of stationary induction apparatus used to supply reactance or to produce phase displacement. 29a. e. An Induction Starter is a device used in starting induction motors, converters, etc., by voltage control, consisting of an auto-transformer com- bined with a suitable switching device. 296. A Leakage Reactance or Series Reactance is a portion of the reactance of any induction apparatus which is due to stray or purely self-inductive flux. 504 STANDARDIZATION RULES. D. GENERAL CLASSIFICATION OF APPARATUS. 30. Commutating Machines. Under this head may be classed the follow- ing: Direct-current generators; direct-current motors; direct-current boost- ers; motor-generators; dynamotors; converters; compensators or balancers; closed-coil arc machines, and alternating-current commutating motors. 31. Commutating machines may be further classified as follows: 32. a. Direct-Current Commutating Machines, which comprise a magnetic field of constant polarity, a closed-coil armature, and a multisegmental commu- tator connected therewith. 33. b. Alternating-Current Commutating Machines, which comprise a magnetic field of alternating polarity, a closed-coil armature, and a multi- segmental commutator connected therewith. 34. c. Synchronous Commutating Machines, which comprise synchronous converters, motor-converters and double-current generators. 35. Synchronous Machines comprise a constant magnetic field and an armature receiving or delivering alternating-currents in synchronism with the motion of the machine; i.e., having a frequency equal to the product of the number of pairs of poles and the speed of the machine in revolutions per second. 36. Stationary Induction Apparatus include transformers, auto-trans- formers, potential regulators, and reactors or reactance coils. 37. Rotary Induction Apparatus, or Induction Machines, include apparatus wherein the primary and secondary windings rotate with respect to each other; i.e., induction motors, induction generators, frequency converters, and rotary phase converters. 38. Unipolar or Acyclic Machines, direct-current machines, in which the voltage generated in the active conductors maintains the same direction with respect to those conductors. 39. Rectifying Apparatus, Pulsating-Current Generators. 40. Electrostatic Apparatus, such as condensers, etc. 41. Electrochemical Apparatus, such as batteries, etc. 42. Electrothermal Apparatus, such as heaters, etc. 42a. Regulating Apparatus, such as rheostats, etc. 426. Switching Apparatus. 43. Protective Apparatus, such as fuses, circuit-breakers, lightning arresters, etc. 44. Luminous Sources. E. MOTORS. SPEED CLASSIFICATION. 45. Motors may, for convenience, be classified with reference to their speed characteristics as follows: 46. a. Constant-Speed Motors, in which the speed is either constant or does not materially vary; such as synchronous motors, induction motors with small slip, and ordinary direct-current shunt motors. 47. b. Multispeed Motors (two-speed, three-speed, etc.), which can be operated at any one of several distinct speeds, these speeds being practically independent of the load, such as motors with two armature windings, or in- duction motors with controllers for changing the number of poles. 48. c. Adjustable-Speed Motors, in which the speed can be varied gradually over a considerable range; but when once adjusted remains practically un- affected by the load, such as shunt motors designed for a considerable range of field variation. 49. d. Varying-Speed Motors, or motors in which the speed varies with the load, decreasing when the load increases; such as series motors. F. DEFINITIONS. INSTRUMENTS. 49a. An Ammeter is a current-measuring instrument, indicating in amperes. 496. A Voltmeter is a voltage-measuring instrument, indicating in volts. 49c. A Wattmeter is an instrument for measuring electrical power, and indicating in watts. 49d. Recording Ammeters, Voltmeters, Wattmeters, etc., are instruments which record graphically upon a time-chart the values of the quantities they measure. 49e. A Watt-Hour Meter is an instrument for registering total watt-hours. This term is to be preferred to the term "integrating wattmeter." 49/. A Voltmeter Compensator is a device in connection with a voltmeter, which causes the latter to indicate the voltage at some other point of the circuit. DEFINITIONS AND TECHNICAL DATA. 505 49fir. A Synchroscope is a synchronizing device which, in addition to indi- cating synchronism, shows whether the machine to be synchronized is fast or slow. G. DEFINITION AND EXPLANATION OF TERMS. (I) Load Factor. 50. The Load Factor of a machine, plant or system is the ratio of the aver- age power to the maximum power during a certain period of time. The average power is taken over a certain period of time, such as a day or a year, and the maximum is taken over a short interval of the maximum load within that period. 51. In each case the interval of maximum load should be definitely specified. The proper interval is usually dependent upon local conditions and upon the purpose for which the load factor is to be determined. (II) Diversity Factor. 51a. Diversity Factor is the ratio of the sum of the maximum power de- mands of the subdivisions of any system or part of a system to the maximum demand of the whole system or of the part of the system under consideration, measured at the point of supply. (III) Demand Factor. 516. Demand Factor is the ratio of the maximum power demand of any system or part of a system to the total connected load of the system or of the part of the system under consideration. (IV) Non-inductive Load and Inductive Load. 52. A non-inductive load is a load in which the current is in phase with the voltage across the load. 53. An inductive load is a load in which the current lags behind the voltage across the load. A load in which the current leads the voltage across the load is sometimes called a condensive or anti-inductive load. 53a. When voltage and current waves are sinusoidal but not in phase, the voltage may be resolved into two components, one in phase with the current and the other in quadrature therewith. The former is called the effective component (sometimes the energy component), and the latter the reactive component of the voltage. The current may be similarly subdivided with respect to the voltage, and the two components similarly named. (V) Power-Factor and Reactive Factor. 54. The Power-Factor in alternating-current circuits or apparatus is the ratio of the effective {i.e., the cyclic average) power in watts to the apparent power in volt-amperes. It may be expressed as follows: effective power _ effective watts _ effective current _ effective voltage apparent power total volt-amperes total current total voltage * 55. The Reactive-Factor is the ratio of the reactive volt-amperes (i.e., the product of the reactive component of current by voltage, or reactive com- ponent of voltage by current) to the total volt-amperes. It may be expressed as follows: reactive power _ reactive watts _ reactive current _ reactive voltage apparent power total volt-amperes total current total voltage * 56. Power-Factor and Reactive-Factor are related as follows: If p = power-factor and q = reactive-factor, then with sine-waves of voltage and current, p 2 + q} = 1. With distorted waves of voltage and current, g ceases to have definite sig- nificance. (VI) Saturation-Factor. 57. The Saturation-Factor of a machine is the ratio of a small percentage increase in field excitation to the corresponding percentage increase in voltage thereby produced. The saturation-factor is, therefore, a criterion of the degree of saturation attained in the magnetic circuit at any excitation selected. Un- less otherwise specified, however, the saturation-factor of a machine refers to 506 STANDARDIZATION RULES. the excitation existing at normal rated speed and voltage, It is determined from measurements of saturation made on open circuit at rated speed. 58. The Percentage of Saturation of a machine at any excitation may be found from its saturation curve of generated voltage as ordinates, against excitation as abscissas, by drawing a tangent to the curve at the ordinate cor- responding to the assigned excitation, and extending the tangent to intercept the axis of ordinates drawn through the origin. The ratio of the intercept on this axis to the ordinate at the assigned excitation, when expressed in percentage, is the percentage of saturation and is independent of the scale selected for excitation and voltage. This ratio is equal to the reciprocal of the saturation- factor at the same excitation, deducted from unity. Thus, if / be the satura- tion-factor and p the percentage of saturation, (VII) Variation and Pulsation. 59. The Variation in Prime Movers which do not give an absolutely uniform rate of rotation or speed, as in reciprocating steam engines, is the maximum angular displacement in position of the revolving member expressed in degrees, from the position it would occupy with uniform rotation, and with one revo- lution taken as 360°. 60. The Pulsation in Prime Movers is the ratio of the difference between the maximum and minimum velocities in an engine-cycle to the average velocity. 61. The Variation in Alternators or alternating-current circuits in general is the maximum difference in phase of the generated voltage wave from a wave of absolutely constant frequency of the same average value, expressed in elec- trical degrees (one cycle equals 360°) and may be due to the variation of the prime mover. 62. The Pulsation in Alternators or alternating-current circuits, in general, is the ratio of the difference between maximum and minimum frequency during an engine cycle to the average frequency. 63. Relation of Variation in prime mover and alternator. If p = number of pairs of poles, the variation of an alternator is p times the variation of its prime mover, if direct-connected, and pn times the variation of the prime mover if rigidly connected thereto in the velocity ratio n; so that the speed of the alternator is n times that of the prime mover. IJC. PERFORMANCE SPECIFICATIONS A\« TESTS. A. RATING. 65. Rating by Output. All electrical apparatus should be rated by output and not by input. Generators, transformers, etc., should be rated by elec- trical output: motors by mechanical output, and preferably in kilowatts. 65a. The following four classes of rating are recognized and recommended: they do not cover the rating of railway motors, which is treated in Appendix B, and there are other large though less definitely definable classes of service in which each case must be treated by itself. Some of these may be later reduced to fairly simple terms and introduced into these Rules. 656. (1) Continuous Rating in which under load there is the attainment of approximately stationary temperature, and no other limit of capacity is ex- ceeded. 65c. (2) Intermittent Rating in which one minute periods of load and rest alternate until the attainment of approximately stationary temperature and no other limit of capacity is exceeded. 65d. Note. Since the temperature depends upon the losses and the capacity of the apparatus to emit them, a constant load may be substituted for the intermittent load in determining the temperature, provided the losses are equivalent. 65e. (3) Minute Rating in which under load for one minute, no mechanical, thermal, magnetic, or electrical limit of capacity is exceeded and no permanent change is wrought in the apparatus. 65/. (4) Variable Service Rating. It is desirable here to recognize this class of rating which is intended to cover the rating of motors for machine- tool and similar service, in which the thermal absorptive capacity plays a part. The specifications for this rating have not been fully determined at the time that this edition of the Rules goes to press. PERFORMANCE SPECIFICATIONS AND TESTS. 507 66. Rating in Kilowatts. Electrical power should be expressed in kilo- watts, except when otherwise specified. 67. Apparent Power, Kilovolt-Amperes. Apparent power in alternating- current circuits should be expressed in kilovolt-amperes as distinguished from effective power in kilowatts. When the power-factor is 100 per cent, the ap- parent power in kilovolt-amperes is equal to the kilowatts. 68. The Rated (Full-Load) Current is that current which, with the rated terminal voltage, gives the rated kilowatts, or the rated kilovolt-amperes. In machines in which the rated voltage differs from the no-load voltage, the rated current should refer to the former. 69. Determination of Rated Current. The rated current may be de- termined as follows: If P = rating in watts, or volt-amperes if the power- factor be other than 100 per cent, and E = full-load terminal voltage, the rated current per terminal is: P 70. I = — amperes, in a direct-current machine or single-phase alternator, 1 P 71. I = — — — amperes, in a three-phase alternator. v3 & 1 P 72. I = - — amperes, in a quarter-phase alternator. 73. Normal Conditions. The rating of machines or apparatus should be based upon certain normal conditions to be assumed as standard, or to be specified. These conditions include voltage, current, power-factor, frequency, wave shape and speed; or such of them as may apply in each particular case. Performance tests should be made under these standard conditions unless otherwise specified. 74. a. Power-Factor. Since the inherent capacity of alternating-current generators, synchronous motors, and transformers, depends upon their voltage and their current, they should be rated in kilovolt-amperes. If the apparatus is rated in kilowatts without specification as to the power-factor, a power-factor of 100 per cent shall be understood. If rated in kilowatts and a power-factor other than 100 per cent be specified, this should be understood as defining only the nature of the load, and not as implying an increase in the ampere rating of the apparatus, which should be based upon the kilowatt rating at 100 per cent power-factor. 75. 6. Wave Shape. In determining the rating of alternating-current ma- chines or apparatus, a sine- wave shape of alternating current and voltage is assumed, except where a distorted wave shape is inherent to the apparatus. See Sees. 79-80. 76. Fuses. The rating of a fuse should be the maximum current which it will continuously carry. 77. Circuit-Breakers. The rating of a circuit-breaker should be the max- imum current which it is designed to carry continuously. 77a. Note. In addition thereto, the maximum current and voltage at which a fuse or a circuit-breaker will open the circuit should be specified. It is to be noted that the behavior of fuses and of circuit-breakers is much influ- enced by the amount of electric power available on the circuit. 78. Indicating Meters should be rated according to their full-scale reading of volts, amperes, or watts. In wattmeters the rated volts and rated amperes should also be included; i.e., the volts and amperes which can be safely and continuously carried by the voltage and current coils respectively. 78a. Watt-Hour Meters should be rated in volts and amperes. B. WAVE SHAPE. _ 79. The Sine Wave should be considered as standard, except where a de- viation therefrom is inherent in the operation of the apparatus. 80. A Maximum Deviation of the wave from sinusoidal shape not exceeding 10 per cent is permissible, except when otherwise specified. See Sees. 5;, 81, 82, 83. See Sees, he to hi. C. EFFICIENCY. (I) Definitions. 84. The Efficiency of an apparatus is the ratio of its output to its input. The output and input may be in terms of watt-hours, watts, volt-amperes, amperes, or any other quantity of interest, thus respectively defining energy- 508 STANDARDIZATION RULES. efficiency, power-efficiency, apparent power-efficiency, current efficiency, etc. Unless otherwise specified, however, the term is ordinarily assumed to refer to power-efficiency. An exception should be noted in the case of luminous sources (see Sec. 346). 86. Apparent Efficiency. In apparatus in which a phase displacement is inherent to their operation, apparent efficiency should be understood as the ratio of net power output to volt-ampere input. 87. a. Note. Such apparatus comprises induction motors, synchronous phase modifiers, synchronous converters controlling the voltage of an alter- nating-current system, potential regulators, open magnetic circuit transformers, etc. 88. b. Note. Since the apparent efficiency of apparatus delivering electrio power depends upon the power-factor of the load, the apparent efficiency, unless otherwise specified, should be referred to a load power-factor of unity. (II) Measurement of Efficiency. 89. Methods. Efficiency may be determined by either of two methods, viz.: by measurement of input and output or by measurement of losses. 90. a. Method of Input and Output. The input and output may both be measured directly. The ratio of the latter to the former is the efficiency. 91. b. Method by Losses. The losses may be measured either collec- tively or individually. The total losses may be added to the output to derive the input, or subtracted from the input to derive the output. 92. Comparison of Methods. The output and input method is preferable with small machines. When, however, as in the case of large machines, it is impracticable to measure the output and input, or when the percentage of power loss is small and the efficiency is nearly unity, the method of deter- mining efficiency by measuring the losses should be followed. 93. Electric Power should be measured at the terminals of the apparatus. In tests of polyphase machines, the measurement of power should not be con- fined to a single circuit but should be extended to all the circuits in order to avoid errors of unbalanced loading. 94. Mechanical Power in machines should be measured at the pulley, gear- ing, coupling, etc., thus excluding the loss of power in said pulley, gearing or coupling, but including the bearing friction and windage. The magnitude of bearing friction and windage may be considered, with constant speed, as inde- pendent of the load. The loss of power in the belt and the increase of bearing friction due to belt tension should be excluded. Where, however, a machine is mounted upon the shaft of a prime mover, in such a manner that it cannot be separated therefrom, the frictional losses in bearings and in windage, which ought, by definition, to be included in determining the efficiency, should be excluded, owing to the practical impossibility of separating them from those of the prime mover. 95. In Auxiliary Apparatus, such as an exciter, the power lost in the auxiliary apparatus shouldjnot be charged to the principal machine, but to the plant con- sisting of principal machine and auxiliary apparatus taken together. The plant efficiency in such cases should be distinguished from the machine efficiency. 96. Normal Conditions. Efficiency tests should be made under normal conditions herein set forth, which are to be assumed as standard. These con- ditions include voltage, current, power-factor, frequency, wave shape, speed, temperature and barometric pressure, or such of them as may apply in each particular case. Performance tests should be made under these standard con- ditions unless otherwise specified. See Sees. 73-75. 97. a. Temperature. The efficiency of ail apparatus, except such as may be intended for intermittent service, should be either measured at, or reduced to, the temperature which the apparatus assumes under continuous operation at rated load, referred to a room temperature of 25° C. See Sees. 267-292. 98. With apparatus intended for intermittent service, the efficiency should be determined at the temperature assumed under specified conditions. 99. 6. Power-Factor. In determining the efficiency of alternating-current apparatus, the electric power should be measured when the current is in phase with the voltage, unless otherwise specified, except when a definite phase difference is inherent in the apparatus, as in induction motors, induction gen- erators, frequency converters, etc. 100. c. Wave Shape. In determining the efficiency of alternating-current apparatus, the sine-wave should be considered as standard, except where a difference in the wave form from the sinusoidal is inherent in the operation of the apparatus. See Sec. 80. PERFORMANCE SPECIFICATIONS AND TESTS. 509 (III) Measurement of Losses. 101. Losses. The usual sources of losses in electrical apparatus and the methods of determining these losses are as follows: (A) Bearing Friction and Windage. 102. The magnitude of bearing friction and windage (which may be con- sidered as independent of the load) is conveniently measured by driving the machine from an independent motor, the output of which may be suitably determined. See Sec. 94. (B) Commutator Brush Friction. 103. The magnitude of the commutator brush friction (which may be con- sidered as independent of the load) is determined by measuring the difference in power required for driving the machine with brushes on and with brushes off (the field being unexcited). (C) Collector-Ring Brush Friction. 104. Collector-ring brush friction may be determined in the same manner as commutator brush friction. It is usually negligible. (D) Molecular Magnetic Friction and Eddy Currents. 105. These losses include those due to molecular magnetic friction and eddy currents in iron and copper and other metallic parts, also the losses due to currents in the cross-connections of cross-connected armatures. 106. In Machines these losses should be determined on open circuit and at a voltage equal to the rated voltage + Ir in a generator, and — Ir in a motor, where I\ denotes the current strength and r denotes the internal resistance of the machine. They should be measured at the correct speed and voltage, since they do not usually vary in any definite proportion to the speed or to the voltage. 107. Note. The Total Losses in bearing friction and windage, brush fric- tion, magnetic friction and eddy currents can, in general, be determined by a single measurement by driving the machine with the field excited, either as a motor, or by means of an independent motor. 108. Retardation Method. The no-load iron, friction, and windage losses may be segregated by the Retardation Method. The generator should be brought up to full speed (or, if possible, to about 10 per cent above full speed) as a motor, and, after cutting off the driving power and excitation, frequent readings should be taken of speed and time, as the machine slows down, from which a speed-time curve can be plotted. A second curve should be taken in tha same manner, but with full field excitation; from the second curve the iron losses may be found by subtracting the losses found in the first curve. 109. The speed-time curves can be plotted automatically by belting a small separately excited generator (say & kw.) to the generator shaft and connecting it to a recording voltmeter. (E) Armature-Resistance Loss. 110. This loss may be expressed by pl 2 r; where r = resistance of one arma- ture circuit or branch, / = the current in such armature circuit or branch, and p = the number of armature circuits or branches. (F) Commutator, Brush and Brush-Contact Resistance Loss. 111. It is desirable to point out that with carbon brushes these losses may be considerable in low-voltage machines. (G) Collector-Ring and Brush-Contact Resistance Loss^ 112. This loss is usually negligible, except in machines of extremely low voltage or in unipolar machines. (H) FIELD-EXCITATION LOSS. 113. With separately excited field, the loss of power in the resistance of the field coils alone should be considered. With either shunt- or series-field wind- ings, however, the loss of power in the accompanying rheostat should also be included, the said rheostat being considered as an essential part of the machine, and not as separate auxiliary apparatus. (I) Load Losses 114. The load losses may be considered as the difference between the total losses under load and the sum of the losses as above specified and determined 510 STANDARDIZATION RULES. 115. a. In Commutating Machines of small field distortion, the load losses are usually trivial and may, therefore, be neglected. When, however, the field distortion is large as in commutating-pole machines, or, as is shown, for instance, by the necessity for shifting the brushes between no load and full load on non-commutating pole machines, these load losses may be consider- able, and should be taken into account. In this case the efficiency may be determined either by input and output measurements, or the load losses may be estimated by the method of Sec. 116. 116. b. Estimation of Load Losses. While the load losses cannot well be determined individually, they may be considerable and, therefore, their joint influence should be determined by observation. This can be done by operating the machine on short-circuit and at full-load current, that is, by determining what may be called the "short-circuit core loss." With the low field intensity and great lag of current existing in this case, the load losses are usually greatly exaggerated. 117. One-third of the short-circuit core loss may, as an approximation, and in the absence of more accurate information, be assumed as the load loss. (IV) Efficiency of Different Types of Apparatus. (A) Direct-Current Commutating Machines. 118. In Direct-Current Commutating Machines the losses are: 119. a. Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102. 120. b. Molecular Magnetic Friction and Eddy Currents. See Meas. of Losses (D), Sec. 105. 121. c. Armature Resistance Losses. See Meas. of Losses (E), Sec. 110. 122. d. Commutator Brush Friction. See Meas. of Losses (B), Sec. 103. 123. e. Commutator, Brush and Brush-Contact Resistance. See Meas. of Losses (F), Sec. 111. 124. /. Field-Excitation Loss. See Meas. of Losses (H), Sec. 113. 125. g. Load Losses. See Meas. of Losses (7), Sec. 114. 126. Note, b and c are losses in the armature or "armature losses"; d and e "commutator losses"; / "field losses." (B) Alternating-Current Commutating Machines. 127. In Alternating-Current Commutating Machines, the losses are: 128. a. Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102. 129. b. Rotation Loss, measured with the machine at open circuit, the brushes on the commutator, and the field excited by alternating current when driving the machine by a motor. 130. This loss includes molecular magnetic friction and eddy currents, caused by rotation through the magnetic field, I 2 r losses in cross-connections of cross-connected armatures, I 2 r and other losses in armature-coils and arma- ture-leads which are short-circuited by the brushes as far as these losses are due to rotation. 131. c. Alternating or Transformer Loss. These losses are measured by wattmeter in the field circuit, under the conditions of test b. They include molecular magnetic friction and eddy currents due to the alternation of the magnetic field, I 2 r losses in cross-connections of cross-connected armatures, I 2 r and other losses in armature coil and commutator leads which are short- circuited by the brushes, as far as these losses are due to the alternation of the magnetic flux. 132. The losses in armature-coils and commutator leads short-circuited by the brushes can be separated in b and c from the other losses by running the machine with and without brushes on the commutator. 133. d. I 2 R Loss, other load losses in armature and compensating winding and I 2 r loss of brushes, may be measured by a wattmeter connected across the armature and compensating winding. 134. e. Field-Excitation Loss. See Meas. of Losses (//), Sec. 113. 135. /. Commutator Brush-Friction. See Meas. of Losses (£), Sec. 103. (C) Synchronous Commutating Machines. 136. 1. In Double-Current Generators, the efficiency of the machine should be determined as a direct-current generator, and also as an alternating- current generator. The two values of efficiency may be different, and should be clearly distinguished. 137. 2. In Converters the losses should be determined when driving the machine by a motor. These losses are: 138. a. Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102, PERFORMANCE SPECIFICATIONS AND TESTS. 511 139. b. Molecular Magnetic Friction and Eddy Currents. See Meas. of Losses (D), Sec. 105. 140. c. Armature-Resistance Loss. This loss in the armature is ql 3 r, where / = direct current in armature, r = armature resistance, and q, a factor which is equal to 1.47 in single-circuit single-phase, 1.15 in double-circuit single- phase, 0.59 in three-phase, 0.39 in two-phase, and 0.27 in six-phase converters. 141. d. Commutator-Brush Friction. See Meas. of Losses (B) , Sec. 103. 142. e. Collector-Ring Brush Friction. See Meas. of Losses (C), Sec. 104. 143. /. Commutator, Brush and Brush-Contact Resistance Loss. See Meas. of Losses (F), Sec. 111. 144. g. Collector-Ring Brush-Contact Resistance Loss. See Meas. of Losses (G), Sec. 112. 145. h. Field-Excitation Loss. See Meas. of Losses (H), Sec. 109. 146. t. Load Losses. These can generally be neglected, owing to the absence of field distortion. 147. 3. The Efficiency of Two Similar Converters may be determined by operating one machine as a converter from direct to alternating, and the other as a converter from alternating to direct, connecting the alternating sides together, and measuring the difference between the direct-current input and the direct-current output. This process may be modified by returning the output of the second machine through two boosters into the first machine and measuring the losses. Another modification is to supply the losses by an alternator between the two machines, using potential regulators. (Z>) Synchronous Machines. 148. In Synchronous Machines, the losses are: 149. a. Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102. 150. b. Molecular Magnetic Friction and Eddy Currents. See Meas. of Losses (D), Sec. 105. 151. c. Armature-Resistance Loss. See Meas. of Losses (E), Sec. 110. 152. d. Collector-Ring Brush Friction. See Meas. of Losses (C), Sec. 104. 153. e. Collector-Ring Brush-Contact Resistance Loss. See Meas. of Losses (GO, Sec. 112. 154. /. Field-Excitation Loss. See Meas. of Losses (H), Sec. 113. 155. g. Load Losses. See Meas. of Losses (/), Sec. 114. (E) Stationary Induction Apparatus. 156. In Stationary Induction Apparatus, the losses are: 157. a. Molecular Magnetic Friction and Eddy Currents measured at open secondary circuit, rated frequency, and at rated voltage — Ir, where / = rated current, r = resistance of primary circuit. 158. b. Resistance Losses, the sum of the Pr losses in the primary and in the secondary windings of a transformer, or in the two sections of the coil in a compensator or auto-transformer, where J = rated current in the coil or section of coil, and r = resistance. 159. c. Load Losses, i.e., eddy currents in the iron and especially in the copper conductors, caused by the current at rated load. For practical pur- poses they may be determined by short-circuiting the secondary of the trans- former and impressing upon the primary a voltage sufficient to send rated-load current through the transformer. The loss in the transformer under these conditions, measured by wattmeter, gives the load losses + I 2 r losses in both primary and secondary coils. 160. In Closed Magnetic Circuit Transformers, either of the two circuits may be used as primary when determining the efficiency. 161. In Potential Regulators, the efficiency should be taken at the maximum voltage for which the apparatus is designed, and with noninductive load, unless otherwise specified. (F) Rotary Induction Apparatus or Induction Machines. 162. In Rotary Induction Apparatus, the losses are: 163. a. Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102. 164. b. Molecular Magnetic Friction and Eddy Currents in iron, copper and other metallic parts; also I 2 r losses which may exist in multiple-circuit windings, a and b together are determined by running the motor without load at rated voltage, and measuring the power input. 165. c. Primary I 2 R Loss, which may be determined by measurement of the current and the resistance. 512 STANDARDIZATION RULES. 166. d. Secondary PR Loss, which may be determined as in the primary when feasible; otherwise, as in squirrel-cage secondaries, this loss is measured as part of e. 167. e. Load Losses; i.e., molecular magnetic friction, and eddy currents in iron, copper, etc., caused by the stray field of primary and secondary cur- rents, and secondary PR loss when undeterminable under (d). These losses may for practical purposes be determined by measuring the total power, with the rotor short-circuited at standstill and a current in the primary circuit equal to the primary energy current at full load. The loss in the motor under these conditions may be assumed to be equal to the load losses + Pr losses in both primary and secondary coils. (G) Unipolar or Acyclic Machines. 168. In Unipolar Machines, the losses are: 169. (a) Bearing Friction and Windage. See Meas. of Losses (A), Sec. 102. 170. (b) Molecular Magnetic Friction and Eddy Currents. See Meas. of Losses (E), Sec. 106. 171. (c) Armature-Resistance Losses. See Meas. of Losses (E), Sec. 110. 172. (d) Collector-Brush Friction. See Meas. of Losses (C), Sec. 104. 173. (e) Collector Brush-Contact Resistance. See Meas. of Losses (G), Sec. 112. 174. (/) Field-Excitation. See Meas. of Losses (H), Sec. 113. 175. (g) Load Losses. See Meas. of Losses (/), Sec. 114. (H) Rectifying Apparatus, Pulsating-Current Generators. 176. This division includes: open-coil arc machines and mechanical and other rectifiers. 177. In Rectifiers the most satisfactory method of determining the efficiency is to measure both electric input and electric output by wattmeter. The input is usually inductive, owing to phase displacement and to wave distor- tion. For this reason the power-factor and the apparent efficiency should also be considered, since the latter may be much lower than the true efficiency. The power consumed by auxiliary devices, such as the synchronous motor or cooling devices, should be included in the electric input. 178. In Constant-Current Rectifiers, transforming from constant potential alternating to constant direct current, by means of constant-current trans- forming devices and rectifying devices, the losses in the transforming devices are to be included in determining the efficiency and have to be measured when operating the rectifier, since in this case the losses may be greater than when feeding an alternating secondary circuit. In constant-current transforming devices, the load losses may be considerable, and, therefore, should not be neglected. 179. In Open-Coil Arc Machines, the losses are essentially the same as in direct-current (closed coil) commutating machines. In this case, however, the load losses are usually greater, and the efficiency should preferably be measured by input- and output-test, using wattmeters for measuring the output. 179a. In alternating-current rectifiers, the output should, in general, be measured by wattmeter and not by voltmeter and ammeter, since, owing to pulsation of current and voltage, a considerable discrepancy may exist between watts and volt-amperes. If, however, a direct-current and an alternating- current meter in the rectified circuit (either a voltmeter or an ammeter) give the same reading, the output may be measured by direct-current voltmeter and ammeter. The type of alternating-current instrument here referred to should indicate the effective or root-of-mean-square value and the type of direct- current instrument the arithmetical mean value, which would be zero on an alternating-current circuit. (J) Transmission Lines. 180. The efficiency of transmission lines should be measured with non- inductive load at the receiving end, with the rated receiving voltage and frequency, also with sinusoidal impressed wave form, except where expressly specified otherwise, and with the exclusion of transformers or other apparatus at the ends of the line. (./) Phase-Displacing Apparatus. 183. In Synchronous Phase-Modifiers and exciters of induction generators, the determination of losses is the same as in other synchronous machines. PERFORMANCE SPECIFICATIONS AND TESTS. 513 184. In Reactors, the losses are molecular magnetic friction, eddy losses and Vr loss. They should be measured by wattmeter. The losses of reactors should be determined with a sine wave of impressed voltage except where expressly specified otherwise. 185. In Condensers, the losses are due to dielectric hysteresis and leakage, and should be determined by wattmeter with a sine wave of voltage or by an alternating-current bridge method. 186. In Polarization Cells, the losses are those due to electric resistivity and a loss in the electrolyte of the nature of chemical hysteresis. These losses may be considerable. They depend upon the frequency, voltage and temper- ature, and should be determined with a sine wave of impressed voltage, except where expressly specified otherwise. D. REGULATION. (I) Definitions. 187. The Regulation of a machine or apparatus in regard to some charac- teristic quantity (such as terminal voltage, current or speed) is the ratio of the deviation of that quantity from its normal value at rated load to that normal value. The term "regulation," therefore, has the same meaning as the term "inherent regulation," occasionally used. 188. Constant Standard. If the characteristic quantity is intended to remain constant {e.g., constant voltage, constant speed, etc.) between rated load and no load, the regulation is the ratio of the maximum variation from the rated-load value to the no-load value. 189. Varying Standard. If the characteristic quantity is intended to vary in a definite manner between rated load and no load, the regulation is the ratio of the maximum variation from the specified condition to the normal rated-load value. 190. (a) Note. If the law of the variation (in voltage, current, speed, etc.) between rated load and no load is not specified, it should be assumed to be a simple linear relation; i.e., one undergoing uniform variation between rated load and no load. 191. (6) Note. The regulation of an apparatus may, therefore, differ according to its qualification for use. Thus, the regulation of a compound- wound generator specified as a constant-potential generator will be different from that which it possesses when specified as an over-compounded generator. 192. In Constant-Potential Machines, the regulation is the ratio of the maximum difference of terminal voltage from the rated-load value (occurring within the range from rated load to open circuit) to the rated-load terminal voltage. 193. In Constant-Current Machines, the regulation is the ratio of the maximum difference of current from the rated-load value (occurring within the range from rated-load to short-circuit, or minimum limit of operation) to the rated-ioad current. 194. In Constant-Power Apparatus, the regulation is the ratio of maxi- mum difference of power from the rated-load value (occurring within the range of operation specified) to the rated power. 195. In Constant-Speed Direct-Current Motors and Induction Motors, the regulation is the ratio of the maximum variation of speed from its rated-load value (occurring within the range from rated load to no load) to the rated-load 196. The regulation of an induction motor is, therefore, not identical with the slip of the motor, which is the ratio of the drop in speed from synchronism to the synchronous speed. 197. In Constant-Potential Transformers, the regulation is the ratio" of the rise of secondary terminal voltage from rated non-inductive load to no load (at constant primary impressed terminal voltage) to the secondary terminal voltage at rated load. 198. In Over-Compounded Machines, the regulation is the ratio of the maximum difference in voltage from a straight line connecting the no-load and rated-load values of terminal voltage as function of the load current to the rated-load terminal voltage. 199. In Converters, Dynamotors, Motor-Generators and Frequency Con- verters, the regulation is the ratio of the maximum difference of terminal voltage at the output side from the rated-load voltage to the rated-load voltage on the output side. 200. In Transmission Lines, Feeders, etc., the regulation is the ratio of the 514 STANDARDIZATION RULES. maximum voltage difference at the receiving end, between rated non-inductive load and no load, to the rated-load voltage at the receiving end (with constant voltage impressed upon the sending end). 201. In Steam Engines, the regulation is the ratio of the maximum varia- tion of speed in passing slowly from rated load to no load (with constant steam pressure at the throttle) to the rated-load speed. For variation and pulsation see Sees. 59-64. 202. In a Hydraulic Turbine or Other Water-Motor, the regulation is the ratio of the maximum variation of speed in passing slowly from rated load to no load (at constant head of water; i.e., at constant difference of level between tail race and head race) to the rated-load speed. For variation and pulsation see Sees. 59-64. 203. In a Generator-Unit, consisting of a generator united with a prime- mover, the regulation should be determined at constant conditions of the prime-mover; i.e., constant steam pressure, head, etc. It includes the inher- ent speed variations of the prime-mover. For this reason the regulation of a generator-unit is to be distinguished from the regulation of either the prime- mover, or of the generator contained in it, when taken separately. (II) Conditions for and Tests of Regulation. 204. Speed. The Regulation of Generators is to be determined at constant speed, and of alternating apparatus at constant impressed frequency. 205. Non-Inductive Load. In apparatus generating, transforming or transmitting alternating currents, regulation should be understood to refer to non-inductive load, that is, to a load in which the current is in phase with the E.M.F. at the output side of the apparatus, except where expressly specified otherwise. 206. Wave Form. In alternating apparatus receiving electric power, regu- lation should refer to a sine wave of E.M.F., except where expressly specified otherwise. 207. Excitation. In commutating machines, rectifying machines, and synchronous machines, such as direct-current generators and motors, alter- nating-current and polyphase generators, the regulation is to be determined under the following conditions: (1) At constant excitation in separately excited fields. (2) With constant resistance in shunt-field circuits, and (3) With constant resistance shunting series-field circuits; i.e., the field adjustment should remain constant, and should be so chosen as to give the required rated-load voltage at rated-load current. 208. Impedance Ratio. In alternating-current apparatus, in addition to the non-inductive regulation, the impedance ratio of the apparatus should be specified; i.e., the ratio of the voltage consumed by the total internal im- pedance of the apparatus at rated-load current to its rated-load voltage. As far as possible, a sinusoidal current should be used. 209. Computation of Regulation. In synchronous machines, the open- circuit exciting ampere-turns corresponding to terminal voltage plus armature- resistance-drop and the exciting ampere-turns at short-circuit for rated-load current should be combined vectorially to obtain the resultant ampere-turns, and the corresponding internal E.M.F. should be taken from the saturation curve. E. INSULATION. . (I) Insulation Resistance. 210. Insulation Resistance is the ohmic resistance offered by an insulating coating, cover, material or support to an impressed voltage, tending to produce a leakage of current through the same. 211. Ohmic Resistance and Dielectric Strength. The ohmic resistance of the insulation is of secondary importance only, as compared with the dielec- tric strength, or resistance to rupture by high voltage. Since the ohmic re- sistance of the insulation can be very greatly increased by baking, but the dielectric strength is liable to be weakened thereby, it is preferable to specify a high dielectric strength rather than a high insulation resistance. The high- voltage test for dielectric strength should always be applied. 212. Recommended Value of Resistance. The insulation resistance of completed apparatus should be such that the rated terminal voltage of the apparatus will not send more than of the rated-load current through PERFORMANCE SPECIFICATIONS AND TESTS. 515 the insulation. Where the value found in this way exceeds one megohm, it is usually sufficient. 213. Insulation Resistance Tests should, if possible, be made at the pressure for which the apparatus is designed. (II) Dielectric Strength. (A) Test Voltages. 214. Definition. The dielectric strength of an insulating wall, coating, cover or path is measured by the voltage which must be applied to it in order to effect a disruptive discharge through the same. 215. Basis for Determining Test Voltages. The test voltage which should be applied to determine the suitability of insulation for commercial operation is dependent upon the kind and size of the apparatus, and its normal operating voltage upon the nature of the service in which it is to be used and the severity of the mechanical and electrical stresses to which it may be sub- jected. The voltages and other conditions of test which are recommended have been determined as reasonable and proper for the great majority of cases and are proposed for general adoption, except when specific reasons make a modification desirable. 216. Condition op Apparatus to be Tested. Commercial tests should, in general, be made with the completely assembled apparatus and not with individual parts. The apparatus should be in good condition and high- i voltage tests, unless otherwise specified, should be applied before the machine is put into commercial service, and should not be applied when the insulation resistance is low owing to dirt or moisture. High-voltage tests should, in general, be made at the temperature assumed under normal operation. High- voltage tests considerably in excess of the normal voltages to determine whether specifications are fulfilled are admissible on new machines only. Unless otherwise agreed upon, high- voltage tests of a machine should be understood as being made at the factory. 217. Points of Application of Voltage. The test voltage should be suc- cessively applied between each electric circuit and all other electric circuits including conducting material in the apparatus. 218. The Frequency of the alternating-current test voltage is, in general, immaterial within commercial ranges. When, however, the frequency has an appreciable effect, as in alternating-current apparatus of high voltage and considerable capacity, the rated frequency of the apparatus should be used. 219. Table of Testing Voltages. The following voltages are recom- mended for testing all apparatus, lines and cables, by a continued application for one minute. The test should be with alternating voltage having a virtual value (or root mean square referred to a sine wave of voltage) given in the table, and preferably for tests of alternating apparatus at the normal frequency of the apparatus. 220. Rated Terminal Voltage of Circuit. Rated Output. Testing Voltage. Not exceeding 400 volts Under 10 kw. . . 1,000 volts 4444 44 4t 10 kw. and over . 1,500 " 400 and over, but less than 800 volts . . Under 10 kw. . . 1,500 " 44 " 44 44 ,4 " . . 10 kw. and over . 2,000 * 4 800 M 44 44 1,200 " . . Any 3,500 " 1,200 " u M 2,500 " . . Any 5,000 " 2,500 44 44 Any . . Double the normal rated voltages. 221. Exception. — Transformers. Transformers having primary pres- sures of from 550 to 5,000 volts, the secondaries of which are directly con- nected to consumption circuits, should have a testing voltage of 10,000 volts, to be applied between the primary and secondary windings, and also between the primary winding and the core. 222. Exception. — Field Windings. The tests for field windings should be based on the rated voltage of the exciter and the rated output of the ma- chine of which the coils are a part. Field windings of synchronous motors and converters, which are to be started by applying alternating current to the armature when the field is not excited and when a high voltage is induced in the field windings, should be tested at 5,000 volts. 223. Rated Terminal Voltage. — Definition. The rated terminal volt- age of circuit in the above table means the voltage between the conductors of the circuit to which the apparatus to be tested is to be connected; for in- stance, in three-phase circuits the delta voltage should be taken. In the 516 STANDARDIZATION RULES. following specific cases, the rated terminal voltage of the circuit is to be de- termined as specified in ascertaining the testing voltage : 224. (a) Transformers. The test of the insulation between the primary and secondary windings of transformers is to be the same as that between the high-voltage windings and core, and both tests should be made simultaneously by connecting the low-voltage winding and core together during the test. If a voltage equal to the specified testing voltage be induced in the high-voltage winding of a transformer it may be used for insulation tests instead of an in- dependently induced voltage. These tests should be made first with one end and then with the other end of the high-tension winding connected to the low- tension winding and to the core. 225. (6) Constant-Current Apparatus. The testing voltage is to be based upon a rated terminal voltage equal to the maximum voltage which may exist at open or closed circuit. 226. (c) Apparatus in Series. For tests of machines or apparatus to be operated in series, so as to employ the sum of their separate voltages, the testing voltage is to be based upon a rated terminal voltage equal to the sum of the separate voltages except where the frames of the machines are separately insulated, both from the ground and from each other, in which case the test for insulation between machines should be based upon the voltage of one machine, and the test between each machine and ground to be based upon the total voltage of the series. (B) Methods of Testing. 227. Classes of Tests. Tests for dielectric strength cover such a wide range in voltage that the apparatus, methods and precautions which are essen- tial in certain cases do not apply to others. For convenience, the tests will be separated into two classes: 228. Class 1. This class includes all apparatus for which the test voltage does not exceed 10 kilovolts, unless the apparatus is of very large static capacity, e.g., a large cable system. This class also includes all apparatus of small static capacity, such as line insulators, switches and the like, for all test voltages. 229. Method of Test for Class 1. The test voltage is to be continuously applied for the prescribed interval (one minute, unless otherwise specified). The test voltage may be taken from a constant-potential source and applied directly to the apparatus to be tested, or it may be raised gradually as specified for tests under Class 2. 230. Class 2. This class includes all apparatus not included in Class 1. 231. Method of Test for Class 2. The test voltage is to be raised to the required value smoothly and without sudden large increments and is then to be continuously applied for the prescribed interval (one minute, unless otherwise specified), and then gradually decreased. 232. Conditions and Precautions for Class 1 and Class 2. The follow- ing apply to all tests: 233. The Wave Shape should be approximately sinusoidal and the apparatus in the testing circuits should not materially distort this wave. 234. The Supply Circuit should have ample current-supply capacity so that the charging current which may be taken by the apparatus under test will not materially alter the wave form nor materially affect the test voltage. The circuit should be free from accidental interruptions. 235. Resistance or Inductance in series with the primary of a raising trans- former for the purpose of controlling its voltage is liable seriously to affect the wave form, thereby causing the maximum value of the voltage to bear a different and unknown ratio to the root mean square value. This method of voltage adjustment is, therefore, in general, undesirable. It may be noted that if a resistance or inductance is employed to limit the current when burn- ing out a fault, such resistance or inductance should^be short-circuited during the regular voltage test. 236. The Insulation under test should be in normal condition as to dry- ness and the temperature should, when possible, be that reached in normal service. 237. Additional Conditions and Precautions for Class 2. The follow- ing conditions and precautions, in addition to the foregoing, apply to tests of apparatus included in Class 2. 238. Sudden Increment of Testing Voltage on the apparatus under test should be avoided, particularly at high voltages and with apparatus having considerable capacity, as a momentarily excessive rise in testing voltage will result. PERFORMANCE SPECIFICATIONS AND TESTS. 517 239. Sudden Variations in Testing Voltage of the circuit supplying the voltage during the test should be avoided as they are likely to set up injurious oscillation. 240. Good Connections in the circuits supplying the test voltage are essen- tial in order to prevent injurious high frequency„disturbances from being set up. When a heavy current is carried by a small water rheostat, arcirig may occur, causing high-frequency disturbances which should be carefully avoided. 241. Transformer Coils. In high- voltage transformers, the low- voltage coil should preferably be connected to the core and to the ground when the high-voltage test is being made, in order to avoid the stress from low-voltage coil to core, which would otherwise result through condenser action. The various terminals of each winding of the high-tension transformer under test should be connected together during the test in order to prevent undue stress on the insulation between turns or sections of the winding in case the high- voltage test causes a breakdown. (C) Methods for Measuring the Test Voltage. 242. For Measuring the Test Voltage, two instruments are in common use, (1) the spark gap and (2) the voltmeter. 243. 1. The Spark Gap is ordinarily adjusted so that it will break down with a certain predetermined voltage, and is connected in parallel with the in- sulation under test. It ensures that the voltage applied to the insulation is not greater than the breakdown voltage of the spark gap. A given setting of the spark;gap is a measure of one definite voltage, and, as its operation depends upon the maximum value of the voltage wave, it is independent of wave form and is a limit on the maximum stress to which the insulation is subjected. The spark gap is not conveniently adapted for comparatively low voltages. 244. In Spark-Gap Measurements, the spark gap may bo set for the re- quired voltage and the auxiliary apparatus adjusted to give a voltage at which this spark gap just breaks down. The spark gap should then be adjusted for, say, 10 per cent higher voltage, and the auxiliary apparatus again adjusted to give the voltage of the former breakdown, which is to be the assumed voltage for the test. This voltage is to be maintained for the required interval. 245. The Spark Points should consist of new sewing needles, supported axially at the ends of linear conductors which are each at least twice the length of the gap. There should be no extraneous body near the gap within a radius of twice its length. A table of approximate striking distances is given in Appendix D. This table should be used in connection with tests made by the spark-gap methods. 246. A Non-inductive Resistance of about one-half ohm per volt should be inserted in series with each terminal of the gap so as to keep the discharge current between the limits of one-quarter ampere and 2 amperes. The pur- pose of the resistance is to limit the current in order to prevent the surges which might otherwise occur at the time of breakdown. 247. 2. The Voltmeter gives a direct reading, and the different values of the voltage can be read during the application and duration of the test. It is suitable for all voltages, and does not introduce disturbances into the test circuit. 248. In Voltmeter Measurements, the voltmeter should, in general, derive its voltage from the high-tension testing circuit either directly or through an auxiliary ratio transformer. It is permissible, however, to measure the volt- age at other places, for example, on the primary of the transformer, pro- vided the ratio of transformation does not materially vary during the test; or that proper account is taken thereof. 249. Spark Gap and Voltmeter. The spark gap may be employed as a check upon the voltmeter used in high-tension tests in order to determine the transformation ratio of the transformer, the variation from the sine wave form and the like. It is also useful in conjunction with voltmeter measure- ments to limit the stress applied to the insulating material. (D) Apparatus for Supplying Test Voltage. 250. The Generator or Circuit supplying voltage for the test should have ample current carrying capacity, so that the current which may be taken for charging the apparatus to be tested will not materially alter the wave form nor otherwise materially change the voltage. The Testing Transformer should be such that its ratio of transformation does not vary more than 10 per cent when delivering the charging current required by the apparatus under test. (This may be determined by short- circuiting the secondary or high-voltage winding of the testing transformer 518 STANDARDIZATION RULES. and supplying fa of the primary voltage to the primary under this condition. The primary current that flows under this condition is the maximum which should be permitted in regular dielectric test.) 251. The Voltage Control may be secured in either of several ways, which, in order of preference, are as follows: 252. 1. By generator field circuit. 253. 2. By magnetic commutation. 254. 3. By change in transformer ratio. 255. 4. By resistance or choke coils. 256. In Generator Voltage Control, the voltage of the generator should preferably be about its approximate normal rated-load value when the full testing voltage is attained, which requires that the ratio of the raising trans- former be such that the full testing voltage is reached when the generator voltage is normal. This avoids the instability in the generator which may occur if a considerable leading current is taken from it when it has low voltage and low field current. 257. In Magnetic Commutation, the control is effected by shunting the mag- netic flux through a secondary coil so as to vary the induction through the coil and the voltage induced in it. The shunting should be effected smoothly, thus avoiding sudden changes in the induced voltage. 258. In Transformer Voltage Control, by change of ratio, it is necessary that the transition from one step to another be made without interruption of the test voltage, and by steps sufficiently small to prevent surges in the testing circuit. The necessity of this precaution is greater as the inductance or the static capacity of the apparatus in the testing circuit under test is greater. 259. When Resistance Coils or Reactors are used for voltage control, it is desirable that the testing voltage should be secured when the controlling resistance or reactance is very nearly or entirely out of circuit in order that the disturbing effect upon the wave form which results may be negligible at the highest voltage. F. CONDUCTIVITY. 260. Copper. The conductivity of copper in annealed wires and in electric cables should not be less than 98 per cent of the Annealed Copper Standard, and the conductivity of hard-drawn copper wires should not be less than 95 per cent of the Annealed Copper Standard. The Annealed Copper Standard represents a mass-resistivity of 0.153022 ohm per metergram at 20° C. or 873.75 ohms per mile-pound at 20° C; or using a density of 8.89, a volume- resistivity of 1.72128 microhm-cm., or microhms in a cm. cube, at 20° C, or 0.67767 microhm-inch at 20° C. G. RISE OF TEMPERATURE. (I) Measurement of Temperature. (A) Methods. 261. There are two methods in common use for determining the rise in tem- perature, viz.: (1) by thermometer, and (2) by increase in resistance of an electric circuit. 262. 1. By Thermometer. The following precautions should be observed in the use of thermometers: 263. a. Protection. The thermometers indicating the room temperature should be protected from thermal radiation emitted by heated bodies, or from draughts of air or from temporary fluctuations of temperature. Several room thermometers should be used. In using the thermometer by applying it to a heated part, care should be taken so to protect its bulb as to prevent radiation from it, and, at the same time, not to interfere seriously with the normal radiation from the part to which it is applied. 264. b. Bulb. When a thermometer is applied to the free surface of a machine, it is desirable that the bulb of the thermometer should be covered by a pad of definite area. A convenient pad may be formed of cotton waste in a shallow circular box about one and a half inches in diameter, through a slot in the side in which the thermometer bulb is inserted. An unduly large pad over the thermometer tends to interfere with the natural liberation of heat from the surface to which the thermometer is applied. 265. 2. By Increase in Resistance. The resistance may be measured either by the Wheatstone bridge, the Thomson or Kelvin double bridge, the potentiometer method, or the ammeter and voltmeter method. If a tem- perature coefficient must be assumed, its value for copper may be taken to be 0.00394 per degree C. from and at 20° C. or 0.00428 per degree C. from and at PERFORMANCE SPECIFICATIONS AND TESTS. 519 0° C. This value holds for average commercial annealed copper. If the copper wire is hard-drawn, or if the conductivity is known, a different value of tem- perature coefficient should be taken, according to the explanation and dis- cussion of the temperature coefficient in Appendix E. The temperature rise may be determined either (1) by dividing the per cent increase of initial resistance by the temperature coefficient for the initial temperature expressed in per cent; or (2) by multiplying the increase in per cent of the initial resistance by T plus the initial temperature in degrees C, and then dividing the product by 100. (— T is the "inferred absolute zero temperature of resistance" and is given in the last column of the table in Appendix E. For average commercial annealed copper it is 233.8.) 266. 3. Comparison of Methods. In electrical conductors, the rise of temperature should be determined by their increase of resistance where prac- ticable. Temperature elevations measured in this way are usually in excess of temperature elevations measured by thermometers. In very low-resistance circuits, thermometer measurements are frequently more reliable than measure- ments by the resistance method. Where a thermometer applied to a coil or winding indicates a higher temperature elevation than that shown by resistance measurement, the thermometer indication should be accepted. (B) Normal Conditions for Tests. 267. 1. Duration of Tests. The temperature should be measured after a run of sufficient duration for the apparatus to reach a practically constant temperature. This is usually from 6 to 18 hours, according to the size and con- struction of the apparatus. It is permissible, however, to shorten the time of the test by running a lesser time on an overload in current and voltage, then reducing the load to normal, and maintaining it thus until the temperature has become constant. 268. 2. Room Temperature. The rise of temperature should be referred to the standard condition of a room temperature of 25° C. 269. Temperature Correction. If the room temperature during the test differs from 25° C, correction on account of difference in resistance should be made by changing the observed rise of temperature by one-half per cent for each degree Centigrade. Thus with a room temperature of 35° C, the observed rise of temperature has to be decreased by 5 per cent, and with a room temperature of 15° C, the observed rise of temperature has to be increased by 5 per cent. In certain cases, such as shunt-field circuits without rheostat, the current strength will be changed by a change of room temperature. The heat-production and dissipation may be thereby affected. Correction for this should be made by changing the observed rise in temperature in propor- tion as the PR loss in the resistance of the apparatus is altered owing to the difference in room temperature. 270. 3. Barometric Pressure. Ventilation. A barometric pressure of 760 mm. and normal conditions of ventilation should be considered as stand- ard, and the apparatus under test should neither be exposed to draught nor enclosed, except where expressly specified. The barometric pressure needs to be considered only when differing greatly from 760 mm. 271. Barometric Pressure Correction. When the barometric pressure differs greatly from the standard pressure of 760 mm. of mercury, as at high altitudes, a correction should be applied. In the absence of more nearly accurate data, a correction of one per cent of the observed rise in temperature for each 10 mm. deviation from the 760-mm. standard is recommended. For example at a barometric pressure of 680 mm. the observed rise of temperature • * u ^i a u 760- 680 is to be reduced by — = 8 per cent. (II) Limiting Temperature Rise. 272. General. The temperature of electrical machinery under regular service conditions should never be allowed to remain at a point at which permanent deterioration of its insulating material takes place. 273. Limits Recommended. It is recommended that the following maxi- mum values of temperature elevation, referred to a standard room tempera- ture of 25° C, at rated load under normal conditions of ventilation or cooling, should not be exceeded. (A) Machines in General. 274. In commutating machines, rectifying machines, pulsating-current gen- erators, synchronous machines, synchronous commutating machines and 520 STANDARDIZATION RULES. unipolar machines, the temperature rise in the parts specified should not exceed the following: 275. Field and armature, 50° C. 276. Commutator and brushes, by thermometer, 55° C. 277. Collector rings, 65° C. 278. Bearjngs and other parts of machine, by thermometer, 40° C. 279. (B) Rotary Induction Apparatus. The temperature rise should not exceed the following: 280. Electric circuits, 50° C, by resistance. 281. Bearings and other parts of the machine, 40° C, by thermometer. 282. In squirrel-cage or short-circuited armatures, 55° C., by thermometer, may be allowed. (C) Stationary Induction Apparatus. 283. a. Transformers for Continuous Service. The temperature rise should not exceed 50° C. in electric circuits, by resistance; and in other parts, by thermometer. 284. b. Transformers for Intermittent Service. In the case of trans- formers intended for intermittent service, or not operating continuously at rated load, but continuously in circuit, as in the ordinary case of lighting •transformers, the temperature elevation above the surrounding air-tempera- ture should not exceed 50° C, by resistance in electric circuits and by ther- mometer in other parts, after the period corresponding to the term of rated load. In this instance, the test load should not be applied until the trans- former has been in circuit for a sufficient time to attain the temperature elevation due to core loss. With transformers for commercial lighting, the duration of the rated-load test may be taken as three hours, unless otherwise specified. 285. c. Reactors, Induction- and Magneto-Regulators. Electric cir- cuits by resistance and other parts by thermometer, 50° C. 286. d. Large Apparatus. Large generators, motors, transformers, or other apparatus in which reliability and reserve overload capacity are import- ant, are frequently specified not to rise in temperature more than 40° C. under rated load and 55° C. at rated overload. It is, however, ordinarily undesirable to specify lower temperature elevations than 40° C. at rated load, measured as above. CD) Rheostats. 287. In Rheostats, Heaters and other electrothermal apparatus, no com- bustible or inflammable part or material, or portion liable to come in contact with such material, should rise more than 50° C. above the surrounding air under the service conditions for which it is designed. 288. a. Parts of Rheostats. Parts of rheostats and similar apparatus rising in temperature, under the specified service conditions, more than 50° C, should not contain any combustible material, and should be arranged or in- stalled in such a manner that neither they, nor the hot air issuing from them, can come in contact with combustible material. (E) Limits Recommended in Special Cases. 289. a. Heat-Resisting Insulation. With apparatus in which the in- sulating materials have special heat-resisting qualities, a higher temperature elevation is permissible. 290. b. High Air Temperature. In apparatus intended for service in places of abnormally high temperature, a lower temperature elevation should be specified. 291. c. Apparatus Subject to Overload. In apparatus which by the nature of its service may be exposed to overload, or is to be used in very high voltage circuits, a smaller rise of temperature is desirable than in apparatus not liable to overloads or in low-voltage apparatus. In apparatus built for conditions of limited space, as railway motors, a higher rise of temperature must be allowed. 292. d. Apparatus for Intermittent Service. In the case of apparatus intended for intermittent service, except railway motors, the temperature elevation which is attained at the end of the period corresponding to the term of rated load should not exceed the values specified for machines in general. In such apparatus, including railway motors, the temperature elevation should be measured after operation, under as nearly as possible the conditions of service for which the apparatus is intended, and the conditions of the test should be specified. VOLTAGES AND FREQUENCIES. 521 H. OVERLOAD CAPACITIES. 293. Performance with Overload. All apparatus should be able to carry the overload hereinafter specified without serious injury by heating, sparking, mechanical weakness, etc., and with an additional temperature rise not ex- ceeding 15° C, above those specified for rated loads, the overload being applied after the apparatus has acquired the temperature corresponding to rated-load continuous operation. Rheostats to which no temperature rise limits are attached are naturally exempt from this additional temperature rise of 15° C. under overload specified in these rules. 294. Normal Conditions. Overload guarantees should refer to normal conditions of operation regarding speed, frequency, voltage, etc., and to non- inductive conditions in alternating apparatus, except where a phase dis- placement is inherent in the apparatus. 295. Overload Capacities Recommended. The following overload ca- pacities are recommended: 296. a. Generators. Direct-current generators and alternating-current generators, 25 per cent for two hours. 297. b. Motors. Direct-current motors, induction motors and synchro- nous motors, not including railway and other motors intended for intermittent service, 25 per cent for two hours, and 50 per cent for one minute. 298. c. Converters. Synchronous converters, 25 per cent for two hours, 50 per cent for one-half hour. 299. d. Transformers and Rectifiers. Constant-potential transformers and rectifiers, 25 per cent for two hours; except in transformers connected to apparatus for which a different overload is guaranteed, in which case the same guarantees shall apply for the transformers as for the apparatus con- nected thereto. 300. e. Exciters. Exciters of alternators and other synchronous machines, 10 per cent more overload than is required for the excitation of the synchro- nous machine at its guaranteed overload, and for the same period of time. All exciters of alternating-current, single-phase or polyphase generators, should be able to give at their rated speed, sufficient voltage and current to excite their alternators, at the rated speed, to the full-load terminal voltage, at the rated output in kilo volt-amperes and with 50 per cent power-factor. 301. /. A Continuous-Service Rheostat, such as an armature- or field- regulating rheostat, should be capable of carrying without injury for two hours a current 25 per cent greater than that at which it is rated. It should also be capable of carrying for one minute a current 50 per cent greater than its rated-load current, without injury. This excess of capacity is intended for testing purposes only, and this margin of capacity should not be relied upon in the selection of the rheostat. 302. g. An Intermittent Service or Motor-Starting Rheostat is used for starting a motor from rest and accelerating it to rated speed. Under ordinary conditions of service, and unless expressly stated otherwise, a motor is assumed to start in fifteen seconds and with 150 per cent of rated current strength^ A motor-starter should be capable of starting the motor under these conditions once every four minutes for one hour. 303. (a) This test may be carried out either by starting the motor at four- minute intervals, or by placing the starter at normal temperature across the maximum voltage for which it is marked, and moving the lever uniformly and gradually from the first to the last position during a period of fifteen seconds, the current being maintained substantially constant at said 50 per cent excess, by introducing resistance in series or by other suitable means. 304. (6) Other Rheostats for Intermittent-Service are employed under such special and varied conditions that no general rules are applicable to them. Ill, VOLTAGES A\I> lREUlEXHE«i. A. VOLTAGES. 305. Direct-Current Generators. In direct-current, low-voltage gen- erators, the following average terminal voltages are in general use and are recommended : 125 volts. 250 volts. 600 volts 522 STANDARDIZATION RULES. 306. Low-Voltage Circuits. In direct-current low-voltage circuits, the following terminal voltages are in general use and are recommended: 115 volts. 230 volts. 550 volts. In alternating-current low-voltage circuits, the following terminal voltages are in general use and are recommended: 110 volts. 220 volts. 440 volts. 550 volts. 307. Primary Distribution Circuits. In alternating-current, constant- potential, primary-distribution circuits, an average voltage of 2,200 volts, with step-down transformer ratios $> and z Vi * s m general use, and is recom- mended. 308. Transmission Circuits. In alternating-current constant-potential transmission circuits, the following impressed voltages are recommended: 6,600 11,000 22,000 33,000 44,000 66,000 88,000 110,000 309. Transformer Ratio. It is recommended that the standard trans- former ratios should be such as to transform between the standard voltages above named. The ratio will, therefore, usually be an exact multiple of 5 or 10, e.g., 2,200 to 11,000; 2,200 to 44,000. 310. Range in Voltage. In alternating-current generators, or generating systems, a range of terminal voltage should be provided from rated voltage at no load to 10 per cent in excess thereof, to cover drop in transmission. If a greater range than ten per cent is specified, the generator should be con- sidered as special. B. FREQUENCIES. 311. In Alternating-Current Circuits, the following frequencies are stand- ard: 25 cycles. 60 cycles. 312. These frequencies are already in extensive use and it is deemed ad- visable to adhere to them as closely as possible. IV. ftJEXEIt AJL llECOnHEADATIO^I. 313. Name Plates. All electrical apparatus should be provided with a name plate giving the manufacturers' name, the voltage and the current in amperes for which it is designed. Where practicable, the kilowatt capacity, character of current, speed, frequency, type, designation and serial number should be added. 314. Diagrams of Connections. All electrical apparatus when leaving the factory should be accompanied by a diagram showing the electrical con- nections and the relation of the different parts in sufficient detail to give the necessary information for proper installation. 315. Rheostat Data. Every rheostat should be clearly and permanently marked with the voltage and amperes, or range of amperes, for which it is designed. 316. Colored Indicating Lights. When using colored indicating lights on switch-boards, red should denote danger, such as "switch closed" or "cir- cuit alive"; green should denote safety, such as "switch open" or "circuit dead." 317. When white lights are used, a light turned on should denote danger, such as "switch closed" or "circuit alive": while the light out should denote safety, such as "switch open" or "circuit dead." Low-efficiency lamps should be used on account of their lesser liability to accidental burn-out. 318. The use of colored lights is recommended, as safer than white lights. 319. Grounding Metal Work. It is desirable that all metal work near high potential circuits be grounded. 320. Circuit Opening Devices. The following definitions are recom- mended. 321. a. A Circuit-Breaker is an apparatus for breaking a circuit at the highest current which it may be called upon to carry. 322. b. A Disconnecting Switch is an apparatus designed to open a circuit only when carrying little or no current. 323. c. An Automatic Circuit-Breaker is an apparatus for breaking a cir- cuit automatically under an excessive strength of current. It should be capable of breaking the circuit repeatedly at rated voltage and at the maxi- mum current which it may be called upon to carry. APPENDICES AND TABULAR DATA. 523 V. ArM3I!¥l>i:CJES AND TABILAH DATA. Unit. volt ampere ohm mho b, " c, watt farad henry- maxwell I, gauss gilbert per cm. cm. or inch t, gm. or lb. second or hour APPENDIX A. NOTATION. ilie following notation is recommended: Name of Quantity. Symbol. «24. Voltage, E.M.F., potential difference E, e. Current I, i, Resistance R, r, Reactance X, x, Impedance Z, z, Admittance Y, y, Conductance G, g, Susceptance B, Power P, Capacity C, Inductance C» Magnetic flux

. Flux and intensity of illumination are connected by the following relation: Illumination ■ Flux Surface or E Mean horizontal intensity is the average intensity in all directions in the horizontal plane passing through the source. In case of an incandescent lamp this plane is taken perpendicular to the axis of the lamp. Mean spherical candle-power is the average candle-power in all directions in space. It bears the following relation to the total luminous flux from the source, Mean hemispherical candle-power is defined as the average candle-power in all directions in a hemisphere having the source of light at its center. The spherical reduction factor is the ratio of the mean spherical candle* power to the mean horizontal candle-power. Trotter gives in the following table the intrinsic brightness of different sources of light. Intrinsic Brightness of Different Sources of JLig-lit. (Trotter.) Platinum (Violle standard) . . . Sun's disk Sky, near sun Albo carbon on edge White paper, horizontal, exposed to summer sky, noon ....... White paper, sun 60° high, paper fac- ing sun Albo carbon, flat Argand Black velvet, summer sky, noon . . White paper, reading without strain- ing C.P. per Sq. In. Red. Green. 120 487,000 120 73.5 16.5 8.25 10.5 6.8 0.0333 0.0018 120 1,000,000 120 60.7 35.2 17.2 8.7 5.29 0.07 0.0024 C.P. per Sq. Cm. Red. 18.5 75.500 18.5 11.4 2.56 1.28 1.63 1.05 0.0052 0.00028 Green. 18.5 155,000 18.5 9.4 5.45 2.67 1.35 0.82 0.0109 0.00037 Sperm candle Moon, 35° above horizon Moon, high Batswing (whole flame) Methven standard Incandescent carbon filament (glow lamp) Crater of electric arc White. I White. 2 2 3 2.26 4.3 120 45,000 0.31 0.31 0.46 0.35 0.666 18.5 7,000 530 ELECTRIC LIGHTING. Units and Standards of I.ig-bt. The Intensity of a Source of tig-lit is expressed in terms of that of some specified unit or standard of reference. No very satisfactory standard for all purposes has as yet been produced, but those listed below are among the best in use or proposed. a. The British standard candle, a spermaceti candle seven-eighths of an inch in diameter, weighing one-sixth pound, and burning at the rate of 120 grains per hour. In case the rate of burning of the candle does not equal 120 grains per hour but falls within the limits of 114 to 126 grains per hour, the value of the light is to be determined by simple proportion assuming that the intensity of the candle light varies in proportion to the rate of consumption of sperm. This standard, in spite of many defects, is still in extensive use and is legalized in many states. It nominally furnishes the unit of measurement in this country. b. Harcourt 10 candle pentane standard. This lamp, which is one of the best of modern standards, is shown in Figs. 1 and 2 in the form in which it is constructed by the American Meter Co. Its fuel is a gas com- posed of a mixture of pentane vapor and air. The pentane is a light distillate of petroleum passing over at a temperature between 25° and 40° C. The pentane is contained in the vapor- izer at the top of the lamp, from which it flows by its own weight down through the small tube to the base of an Argand burner, where it forms a flame inside a metal chimney. The base of the chimney is adjusted accurately to a height of 47 mm. above the top of the burner, and it is only the portion of the flame which comes be- tween the burner and the base of the chimney which falls on the photometer. All the light from the ragged upper portions of the flame is cut off. The flame is adjusted to a definite height by observing it through a mica window in the chimney. The exposed portion of the flame is protected from draughts by a conical shield open on one side. The lamp should be used in a well-ventilated room free from avoid- able draughts. According to Paterson of the National Physical Laboratory the candle-power of the lamp is expressed by the equation: c.p. = 10 + 0.066 (10 - c) - 0.008 (760 - 6), in which € is the number of liters of moisture per cubic meter of dry air, and b is the baro- metric height in millimeters. The quantity € is IP found from the equation c = X 1000, b — e — ei in which e equals the vapor pressure of the water and ei the vapor pressure of the CO2 pres- ent. The constant 10 represents the average hygrometric condition in London for a period of three years. . , the principal French standard, burns 42 grams of purified colza oil per hour, the flame being 40 mm. high. MM. Regnault and Dumas have proven by experiments that when the consumption of colza is at a rate between 40 and 44 grams per hour, the light emitted by this standard is proportional to the weight of colza burned. Following is a table showing the proper dimensions of this standard. Fig. 1. The Carcel lamp, LIGHT. 531 PENTANE10 C. P. LAMP HARCOURT TYPE Fig. 2. 532 ELECTRIC LIGHTING. Dimensions of Carcel Lamp. External diameter of burner Interior diameter of inner air current . . . Interior diameter of outer air current . . . Total height of chimney Distance from elbow to base of glass . . . Exterior diameter at level of bend .... Interior diameter of glass at top of chimney Mean thickness of glass 23.5 17.0 45.5 290 61 47 34 2 Use lighthouse wick weighing 3.6 grams per decimeter and woven with 75 strands. This standard is quite satisfactory if carefully used. d. The platinum standard proposed by Violle is the light emitted by one square centimeter of platinum at its melting-point. Violle shows that the light emitted by this unit is equivalent to 19 1 to 19f British candles. This standard has never been reduced to practice. The French bougie d^cimale is supposed to equal the 20th part of the Violle platinum unit. e. Hefner Amyl Lamp. The legal standard in Germany is the so-called Hefner unit, which is the light given by the Hefner- Alteneck amylacetate lamp. This lamp has been exhaustively investigated by the Reichsanstalt, which certifies to the accuracy of lamps submitted to it ; its intensity is about 10 per cent less than that of the English candle, and its normal flame is 40 millimeters high. It is very uniform and reproducible, and owing to the fact that lamps of certified value can be so readily obtained it is widely used, not only in Germany, but elsewhere. Careful instructions are issued with each lamp, and when used in accordance with these instructions the errors of measurement are not more than half those met with in the use of standard candles. The color is somewhat against this unit, being a distinctly reddish orange, which is a rather serious objection when used as a working standard in measurements of Welsbach burners or incandescent and Nernst lamps. Even with its faults though, it is probably the best primary standard that we have, as it can be reproduced accurately to a most unusual degree. This lamp has of late come into very general use as a reliable, moderate- priced and easily reproducible standard. It has been recommended by the American Institute of Electrical Engi- neers and the German Reichsanstalt. A cylindrical base contains the amyl acetate, which is drawn up through a wick tube of Ger- man silver in a specially prepared wick. The height of this German silver tube and the height of the flame are of vital importance. To secure the proper adjustment at the time the lamp is used, an optical flame gauge is provided, consisting of a small camera with lens, and ground glass plate. On this ground glass plate a horizontal line determines exactly the point at which the top of the flame should be kept. An error of 0.2 of a millimeter in the height of the flame produces an error of i of 1 per cent in the candle-power, so their setting must be made closely. In using this lamp special care should be taken that fresh air in abundance is supplied, but the room must be perfectly free from draughts or air currents, and it should be watched by a person at a distance from it. If the flame does not burn steadily the wick should be carefully trimmed, making itsome- what crowned. Never char the wick by burning it too high; after continued use it should appear to be only slightly browned. Fig. 3> LIGHT. 533 With a little experience it will be found that the flame can be kept accur- ately on the line of the optical flame gauge and quite steady. The variations of temperature, humidity and barometer height affect the candle-power of the 20 30 40 50 60 70 80 90 CHANGE IN INTENSITY FROM HUMIDITY AT DIFFERENT TEMPERATURES Fig. 4. lamp to a certain extent, but these fluctuations have been investigated fully, and corrections are given in the accompanying diagrams (Figs. 4 and 5). +5 TJ JO o z o I > m / BAROI rfETER HEIGH r MMS 500 550 600 650 700 750 800 850 900 CHANGE IN INTENSITY WITH BAROMETER HEIGHT Fig. 5. Incandescent Lamp* as Secondary Standards. Carbon fila- ment lamps which have been seasoned by burning them a few hours until their initial period of rise of candle-power at constant voltage has been passed, furnish secondary standards of light of remarkable constancy. It should be understood, however, that no single lamp can be relied on abso- lutely, but rather the average value given by a group of such lamps. The uni- formity of results which is obtained in the photometry of incandescent lamps in present practice in this country is due in no small measure to the fact that incandescent lamp standards, practically all of which emanate from the same laboratory, are in nearly universal use. These sub-standards have been standardized not by direct reference to a primary standard, none of which is entirely constant, but by reference to a series of incandescent lamp secon- dary standards, whereby a constant value for the unit is obtained. An invariable unit of luminous intensity has been maintained by such a series of lamps by the Electrical Testing Laboratories in New York for upwards of ten years. The standardization value for these lamps was derived from a similar series in the possession of the Edison Lamp Works, which were in turn standardized originally by reference to lamps standardized in the Reichsanstalt. The basis of this original standardization was the assump- tion that the Hefner unit equals 0.88 candle-power. This ratio has since received the sanction of the A. I. E. E., and more recently the Bureau of Standards in Washington has established its unit of luminous intensity 534 ELECTRIC LIGHTING. on the same basis. Thus it has come about that photometric measure- ments in this country which are nominally based on the British candle as a unit are actually, as far as electrical measurements are concerned, based on an invariable unit representing one of the values which the variable candle may assume, which is maintained by standardized incandescent lamps, and which is reproducible only through the intermediary of the Hefner standard lamp. Standardized lamps are furnished by the Elec- trical Testing Laboratories in New York of any required candle-power and voltage and for use either stationary or rotating. A special type of lamp has been developed for use in making stationary standards. These lamps have two horse-shoe shaped filaments in the same plane, one inside the other. The standard direction in these lamps is at right angles to the plane of the filaments, as indicated by vertical lines etched in the glass. Lamps are also standardized and certified by the Bureau of Standards. On account of the adoption of the Harcourt 10 candle pentane lamp as the official standard by the Metropolitan Board of Gas Referees of London and the introduction of this standard into practice in this country, chiefly in the photometry of illuminating gas, a discrepancy has arisen between the candle of the electric industry and the candle of the gas industry. Recent international determinations of the ratio between the Hefner unit and the pentane unit have shown that the Hefner equals 0.915 candle-power, the candle being defined as the one-tenth part of the intensity of the pentane unit. As has been said, the value of the Hefner in terms of the candle of the electrical industry and of the Bureau of Standards is 0.88. The matter of this discrepancy is now (Dec, 1907) under advisement by a joint committee of the Illuminating Engineering Society, the American Institute of Electrical Engineers, and the American Gas Institute. The following is a table giving the values of the various standards and units in terms of each other. This table is compiled from the most recent data on the subject. Hefner unit 10 c.p. pentane . . . Carcel Bougie de*cimale . . . Candle unit. U. S. A. Unit National Physical Lab- oratory. London Unit Laboratoire Central d'El£ctricite\ Paris PHOTOMETERS. A photometer is an apparatus for measuring the intensity of a source of light or of an illumination in terms of a standard. Incase the apparatus is intended for the latter purpose only, it is sometimes called an ' illumi- nometer. " All photometric measurements are made by a visual compar- ison of the source to be measured with some standard. The eye cannot tell us how many times brighter one light is than another. It can say only that one illuminated field is just as bright as another. A photometer consists, then, of two essential parts: first, an arrangement whereby two fields are obtained in juxtaposition to each other, one of them being illumi- PHOTOMETERS. 535 nated by the standard light, and the other by the light which is to be measured ; second, of an arrangement whereby the brightness of one or both the fields can be varied continuously according to a known law, from which the rela- tive intensities of the sources can be computed as soon as the conditions have been discovered under which the fields are equal in illumination. In an illuminometer a further part must be provided; namely, a standard plate for the reception of the illumination which is to be measured. The law of variation which is most commonly employed is that which states that illumination from a punctiform source of light varies inversely as the square of the distance to the source. A common form of photometer is as shown in Fig. 6. The light to be measured and the standard light are set up at opposite ends of a bar on which the sight-box containing a photometric screen or disk for testing the equality of illumination can be moved. When a setting has been made, the intensities of the two sources Fig. 6. Photometer, Queen & Co. of light are directly proportional to the squares of their respective distances from the photometric screen in the sight-box. The forms of sight-box which are most commonly employed are the Bunsen and the Lummer-Brodhun. The latter is unexcelled by any other photometric device when the lights to be compared are of the same color. When color differences are present, the Bunsen is to be recommended, especially so when it is equipped with the Leeson star disk. In Hansen's photometer a piece of white paper — certain kinds of draught- ing paper are good — with a grease spot in its center is placed between the two lights with its surface at right angles to the rays. Behind the paper in the sight-box are placed two mirrors at an angle of about 140 degrees with each other so that both sides of the disk can be viewed simultaneously. The box is moved along the bar between the lights until the grease spot is seen with equal distinctness on both sides. This indicates an equality of illumination, and the general law given above is used to compute the relative intensities of the lights. The Leeson star disk is in some respects superior to the grease spot disk, and is made as follows : A star-shaped figure is cut out of a piece of moderately heavy paper, and the latter is pressed between two pieces of tissue paper of the proper degree of transparency. The out- Bide pieces may be pasted fast to the middle piece. 536 ELECTRIC LIGHTING. In the Lummer-Brodhnn photo- meter, diagram and cut of the carriage of which are shown below, the rays of light from the two sources under com- parison enter at the sides so as to strike the surfaces of the opaque gypsum screen. Diffused light from these white surfaces reaches two parallel mirrors (inside) at an angle of 45°, and is reflected to right- angled prisms which have the outer portions of their hypothenuse surfaces cut away and coated with asphalt varnish to secure complete absorption. Light entering the prisms from the mirrors is Fig. 7. Diagram of Lummer- either transmitted or totally reflected at Brodhun Photometer. their surface of contact, so that an ob- server at the telescope tube sees a circu- lar disk of light from one side of the gypsum screen surrounded by an an- nular ring of light from the other side, the boundary line between the two being sharply defined. Fig. 8. Lummer-Brodhun Photometer Carriage. Romford's photometer compares the shadows of an opaque rod thrown on a white screen by two lights. When the shadows are of equal density, H In Ritchie's photometer two equal white surfaces are placed at an angle with each other, and with the line of light and their brightness com- pared, moving back and forth on the line of light until both surfaces are alike in illumination ; the relative intensities of the lights are then the same as with the Bunsen instrument. In Joly's photometer, two slabs of paraffin wax, or translucent glass about 3" X 2" X V, are fastened together back to back by Canada balsam, a sheet of paper or silver foil being first interposed, after which the edges and surfaces are ground smooth. This slab is placed between the two lights, with the plane of the joint at right angles to the line between the lights, and moved back and forth on that line until the observer looking at the edge of the slab finds both sides equally illuminated, when the relative intensities are as before. By revers- ing the slab, a check can be had on the observation. PHOTOMETERS. 537 * The Test Mate. — Preston S. Millar. In general work the intensity of the light incident upon a given surface is the only quantity which it is practicable or even desirable to measure. This is not proportional neces- sarily to the illuminating effect, which varies as well with the point from which the surface is viewed, with the color of the light and with the color and character of the surface. The criterion by which the light intensity is judged must be strictly proportional to the light incident upon the test plate, and must be inde- pendent of each of the other improper variables just mentioned, if the results of the observation are to show the intensity of the light incident upon the surface. Whether or not the light falling upon the photometric device varies only with that incident upon the test plate, depends upon the design and loca- tion of that plate. The requirements for a theoretically correct test plate are: First, a plain white surface which, when viewed from the point of photo- metric observation, obeys Lambert's law of the cosines with reference to intensity of illumination produced by light incident upon its surface at any inclination and from any direction. Second, a material which will not introduce errors due to color differences. Third, a plate which may be placed at any angle. Fourth, a location such that neither the body of the observer nor instru- ment parts shall obstruct light which would otherwise fall upon the plate. It is, of course, desirable to measure all of the light which would be inci- dent upon an object at the point to be considered. In all interior lighting systems there is more or less diffused light, all of which has some illuminat- ing value. In order to measure all of the effective light, there must be no objective interference with light incident upon the plate at any angle. This means that all instrument parts, as well as the observer, must be beneath or behind the surface of the test plate. This is possible only when transmitted light, instead of reflected light, is measured. The only color which is practicable is white, of as great purity as may be obtainable, and as free as possible from selective absorption. With such a test plate, lights of different colors are credited with approximately their true intensities, when the test plate is viewed from the photometric device. Prof. Ii. Wefoer has invented a photometer, as follows: The apparatus consists of a tube, A, about 30 cm. long, which can be moved up and down and swung in a horizontal plane on the upright, c. The standard light, S, a benzine lamp, is contained in a lantern fastened to the right end of the tube, A . Within the tube, A , a circular plate of opal glass can be moved from or towards the light, S; its distance from E is read in centimeters on the scale, s, by means of an index fastened to the pinion, P. At right angles to tube, A, a second tube, B, is fastened. This tube can be rotated in a vertical plane, and its position in reference to the horizontal is read on the graduated circle, C, A Lummer-Brodhun prism contained in tube B in its axis of rotation receives light from the opal glass plate in tube A, and reflects this light towards the eye-piece, O, so that the outer half of the field of vision is illuminated by this light; the inner half is illuminated by the light entering the tube, B, through g. In making measurements, the tube B is pointed toward the source of light to be measured. The light has to pass through a square box, g, in which may be inserted one or more opal glass plates, in order to diminish the intensity of the light, and thus to make it comparable with the standard light. The apparatus permits the measurement of light in the shape of a flame, as well as the measurement of diffused light. Since the measurement of diffused light interests us most at present, a short description of the method will not be out of place. A white screen, the surface of which is absolutely without luster, fur- nished as part of the apparatus, is placed in a convenient position, either horizontal or vertical, or at any desired inclination, toward the source of light. The photometer having been located at a convenient distance from the screen, the tube B is pointed to the center of the screen. The distance of the photometer from the screen can be varied within very wide limits, the only restrictions being that the field of vision receives no other light than * Trans. Illuminating Engineering Society, October, 1907. 538 ELECTRIC LIGHTING. that emanating from the screen. The necessary precautions for adjustment having been observed, the opal glass plate in the tube A is moved until both halves of the field of vision appear equally illuminated. The distance, r, of this glass plate from the standard light at the moment of equal illumina- Fig. 9. Prof. L. Weber's Portable Photometer R=10,00OMM /^ Fig. 10. tion is read on the scale on tube A in millimeters, and the intensity of illu- mination on the white screen is calculated from the formula, 1O000 , r* /- -K. PHOTOMETERS. 539 The constant K is previously determined as follows: A standard candle or its equivalent is placed exactly one meter distant from the white screen, and the tube, B, of the photometer is pointed towards the screen, so that the center of the screen, which is marked by a cross, is seen in the center of the field of vision. As indicated in Fig. 6, the photo- meter must be so placed that the eye, looking through the eye-piece, sees nothing but the white screen. The angle of inclination under which the screen is observed may be varied within wide limits without influencing the result; it should, however, not exceed 60 degrees from the normal to the screen. Equal illumination of both halves of the field of vision having been ob- tained by means of adjusting the opal glass plate in tube A, the constant, fiC, is found by calculation: r 2 K = R^ Since r is read in millimeters, and R is made 1 meter or 1000 millimeters, 1000 instead of 1 must be taken in the formula for calculating the intensity of illuminating in meter-candles. A second method permits of measurements of diffused light without the intervention of a screen; but for further details the reader is referred .to the description of the apparatus by Professor Weber, Elektrotechnische Zeit- schrift, vol. v., p. 166. The whole apparatus can easily be taken apart, and packed in a box about 24 X 8 X 12 in hes. In some cases the benzine lamp might well be replaced by a small incandescent lamp, provided this lamp is standardized before and after each set of experiments. Such miniature lamps have been Found very convenient, and quite sufficiently constant in candle-power for several hundred observations. Sharp-Millar Universal Portable Photometer. —This instru- ment is designed for making all the various measurements of candle-power Fig. 11. and illumination which the Weber is fitted for, while it is more portable, convenient, and accurate than the latter instrument, and less complicated and expensive. The instrument is illustrated in Fig. 11.* Integrating- Photometers. — Photometers can be constructed, so that they will measure directly the mean spherical candle-power of lamps. Such photometers have been designed by Professor Matthews both for arc and incandescent lamps. (Trans. A. I. E. E.) * See Electrical World, LI. p. 181, Jan. 25, *08. Electrical Review, LII. p. 141, Jan. 25, '08. Electrician (London), LX. p. 562, Jan. 24, '08. 540 ELECTRIC LIGHTING. A simple form of this type of photometer is the Ulbricht sphere photo- meter. This consists of a large sphere coated on the inside with dull white paint and furnished with a small window of diffusing glass. The lamp is introduced into the interior and a screen is so placed that the direct rays of the lamp cannot fall on the window, which is consequently illuminated by reflected rays alone. The theory shows that the intensity of such illu- mination is proportional to the total luminous flux, or the mean spherical candle-power of the source within, so that it is necessary only to photo- meter the light issuing from the window to have a measure of these quan- tities. The sphere must be calibrated by the "substitution method," using an incandescent lamp standardized for mean spherical candle-power. Rating* of Illuminants. — Illuminants are rated according to their candle-power and their volts, amperes or watts. Differences occur in practice as to what is meant by the candle-power, that is, in what direction the candle-power is to be measured. In the earliest days incandescent lamps were rated by their maximum candle-power; now, however, the most common practice is to use the mean horizontal candle-power. In compar- ing lamps having differently shaped filaments this is in general not a fair basis, since two lamps might give the same total flux of light and yet one of them might have a much smaller mean horizontal candle-power than the other. These differences are recognized by the differences in the spher- ical reduction factors of the two. A small difference in spherical reduc- tion factor may have a very large influence on the results obtained in a life-test. The fair way is to use the total flux of light or the mean spherical candle-power as the basis for comparing lamps or illuminants of different types. The American Nernst lamp is usually rated by its maximum candle- power, that is, the candle-power immediately below it. The intensity in this direction is increased considerably by the light reflected from the heater coils and other parts of the lamp. No standard method for candle-power rating of arc lamps has ever been adopted in America. In Germany the mean lower hemispherical intensity is chosen for this purpose. raCAWIMBSClSira JLAIttPS. Watts per candle. — The condition of operation of an incandes- cent lamp is usually specified by the watts per candle, meaning, ordinarily, the watts per mean horizontal candle. The efficiency of a lamp is inversely {)roportional to its watts per candle. The life history of a carbon filament amp is characterized by a small initial increase in candle-power lasting for about 50 hours in the case of a 3.1 watt per candle-lamp and then by a unform decrease in candle-power until the lamp fails. This is accom- panied by a regularly increasing blackening of the bulb. It has been shown (Sharp, Electrical World, Vol. 48, p. 18), that the age of a lamp may be estimated by an examination of the degree of bulb blackening. The light from frosted lamps decreases more rapidly than that from un- frosted ones, an effect which has been shown (Millar, Electrical World, April 20, 1907) to be due to the increased absorption of that portion of the light which suffers multiple reflections. Any lamp may be operated at any watts per candle simply by raising or lowering the impressed voltage, but the life of a lamp decreases very rapidly with decreased watts per candle. In opera- tion it is necessary to strike a balance between increased efficiency and in- creased cost of lamp renewals. The standards are 3.1, 3.5 and 4.0 watts per candle. Closely regulated voltage is essential to successful 3.1 watts per candle operation. After a lamp has reached a certain point in its decline in candle-power and efficiency, it is more economical to replace it with a new one than to consume energy in a wasteful device. The period of the life at which this condition is reached is called the ''smashing point," of the lamp. The smashing point may be computed, but it is found in practice that it is most satisfactory to assume uniformly that its point has been reached when the candle-power has decreased 20 per cent from the initial value. This con- stitutes by common consent the close of the "useful life" of a carbon fila- ment lamp. Spherical Candle-power and Distribution Curves. — A lamp filament giving a certain total flux of light may be made to give a greater or a smaller proportion of this in the horizontal direction. There- INCANDESCENT LAMPS. 541 fore the mean horizontal candle-power is not a true basis for comparing the performance of lamps of different types. The "spherical reduction fac- tor," or ratio of mean spherical to mean horizontal candle-power must be taken into consideration. The following curves and table give values for this factor for different types of lamps and the axial distribution of candle- power about the same types. The curves show also the Rousseau diagrams for the lamps, that is, curves the area enclosed by which is proportional to the mean spherical candle-power. The data were obtained at the Elec- trical Testing Laboratories. Lamp Type. Description. Double loop. Oval. Small spiral; single turn. Large spiral; single turn. Medium spiral; single turn. Short-legged spiral; double turn. Elliptical spiral, double turn, axis of ellipse horizontal. Lamp Type. Watts. End-on c.p Mean horizontal c.p. . . . Mean spherical c.p. . . . -p f - . Mean spherical c.p. Mean horizontal c.p. Mean spherical c.p. End-on c.p. End-on c.p. Mean horizontal c.p. Watts per mean spherical . Watts per mean horizontal Watts per end-on .... Ratio: Ratio: 1 49.6 5.06 16.00 12.82 0.802 2.54 0.316 3.88 3.10 9.8 2 49.6 3 63.5 7.3 16.0 13.19 0.825 1.81 0.456 3.76 3.10 6.78 7.7 16. C 13.42 0.840 1.74 0.481 4.73 3.97 8.26 4 56.6 16.0 13.63 0.854 1.42 0.602 4.15 3.52 5.90 5 53.8 9.31 16.0 13.78 0.862 1.48 0.582 3.91 3.36 5.78 59.3 11.4 16.0 14.07 0.880 1.23 0.712 4.22 3.70 5.20 7 64.74 15.9 16.0 15.72 0.983 0.864 0.992 4.09 4.02 4.04 542 ELECTRIC LIGHTING. i s • e s 9 i 9 ft 8 ft I -2 5 i- eg o u .a 9 >» .Q ft a s i g £ 5 i 8 © o tO CM .8 .CO 'CM ' tO >o o oo to .oo .o o oo ©CO . . i-i CO . . coco ©© - .CM '• . . . .0 .t> .8 .CO =2 .co .10 1—1 *~ 1 "* C-OQ .WHUJ . . . .rHT^CO . . . 00 00© ;©coio '. ' ' CM CM CM :©cso ; ; : CO . . . 00 . . . O .r-» . ... .05 © .r>» .00 IO .CM .00 CM©iO . .©CO© . . . .90H . . . .00© r-l . I>COCO ; ;oo©© ; ; r-l CO© . . .CM"* . . . . .00© . . ©r^ . . .05© . . . . .*>© . . I>© ! . .*>» ' . . .©00© . . t>©lO . .rj© . . . .COI>00 . . ©CO© . .CM"tfiO . . . .COI>00 . . *o rH "* . . . coio«#cooo co-^iocot^ . . rJHiOOiCOiO t^COCM©CM . . CO -"* iO CO l> . . lOl>l^-rHlO I>rH©00© . . co*o»Oko© . . CO •* . . . CO . . . HOJIM . .10000 . . NNN ; ;^T} • • • 'CMCOCO • • ©©CO • -i>r-ICO • • • -CMCOCO • • 00 00 00 . .10©^ • • • -CMCMCO • • CO • © • • CO©© coos co . .1-1^00 • • • -CMCMCM • • t^OO^ • • ©cor-- • • • -CMCNCM • • © . .10 • • -r-1 • . . .^^ . . . . .©co . . . . .*-• . . . . .r-l CM . . . . . .09 . . . . . .00 . . . . . .»H . . . . . .rH . . ! ! * ;oo : ; . . . .CO . . . . .CO . . .co CO . t^ . i-t . 00 . . . . . .co : 1-1 . *. ! '. \ '.co ; CM . co . r-l . © . CM . ! ! ! '. !** in Candle-power. — The drop in candle-power is a characteristic of an incandescent lamp always to be borne in mind. The relative drop or loss of candle-power, other things being equal, determines the comparative value of different lamps. We may have a lamp that loses 50 per cent in candle-power inside of 200 hours on a 3- watt basis. Considered from the standpoint of life only, such lamps are INCANDESCENT LAMPS. 545 excellent, because their filaments deteriorate to such a degree that it is practically impossible to supply enough current to brighten them up to the breaking point, but no discerning station manager would want such dim lamps, even with unlimited life. As in the selection of incandescent lamps so in their use — the exclusive consideration of life leads to poor results Loss of candle-power in a lamp sooner or later makes it uneconomical to continue in use. A customer cares little how efficiently a station is operated, but is much concerned about the quality of light furnished. Some means of keeping the average life below 600 hours should be adopted by every lighting company that has any regard for the economical production of light, or the satisfac- tion of their customers. A simple method is to fix the average life at 600 hours or less, and then determine from the station record how many lamps should be renewed each month to keep the average life within this limit. The required number of lamps should be renewed each month. If, for example, a station decides on an average life not to exceed 600 hours and the station records show that on the average 60,000 lamp hours of current are supplied monthly, then it would be necessary to renew ■ ' or 100 lamps a month. The Importance of Good Regulation. Proper Selection and Use of Transformers* — Poor regulation of voltage probably results in more trouble with customers than any other fault in electric lighting service. Some central station managers act on the theory that so long as the life of the lamp is satisfactory, an increase of voltage, either temporary or per- manent, will increase the average light. The fact is that when lamps are burned above their normal rating the average candle-power of all the lamps on the circuit is decreased; and if the station is on a meter basis, it increases the amount of the customers' bills. Evil* of Excessive Voltage. — Excessive voltage is thus a double error — it decreases the total light of the lamps, and increases the power consumed. The loss of light displeases the customers and discredits the service. If light is sold by meter, the increased power consumption dissat- isfies the customers; if light is sold by contract, the additional power is a dead loss to the station. If increased light is needed, 20 candle-power lamps should be installed, instead of raising the pressure. Their first cost is the same as 16 candle-power lamps; they take but little more current than 16 candle-power lamps operated at high voltage, and give greater average light. Increased pressure also decreases the commercial life of the lamp; and this decrease is at a far more rapid rate than the increase of pressure, as shown in the following table. This table shows the decrease in life of standard 3.1 watt lamps, due to increase of normal voltage. Per Cent of Normal Voltage. Life Factor. 100 1.000 101 0.795 102 .615 103 .49 104 .40 105 .34 106 .29 From this table it is seen that 3% increase of voltage halves the life of a lamp, while 6% increase reduces the life by two-thirds. Irregular pressure, therefore, necessarily results in the use of lamps in which the power consumption per candle is greater than a well-regulated pressure would allow. The result is reduced capacity of station, and reduced station efficiency. 546 ELECTRIC LIGHTING. These remarks apply with special force to alternating-current stations, since we have here two sources of possible irregularity in voltage — the generator and the transformer. Poor regulation is most apt to occur in the transformers, and the utmost care should, therefore, be taken in their selec- tion and use. The efficiency of the average lamp on alternating systems is nearly 4 watts per candle. With good regulation obtained by the intelli- gent use of modern transformers, the use of lamps of an efficiency of 3.1 watts per candle becomes practicable. It is thus possible to save 25 % in power consumption at the lamps, and increase the capacity of the station and transformers by the same amount. The general adoption of higher voltage secondaries gives smaller loss in wires, and permits the use of larger transformer units, thus greatly improv- ing the regulation. On this account 50-volt lamps are gradually going out of use. The replacement of a number of small transformers by one large unit, and of old, inefficient transformers by modern types, has also been of immense advantage to stations. A large number of stations, however, still retain these old transformers, and load their circuits with large numbers of small units. Such stations necessarily suffer from loss of power, bad regulation, and a generally deteriorated lighting service. Simply as a return on the investment, it would pay all such stations to scrap their old transformers and replace them with large and modern units. Proper care in the selection of transformers considers the quality and the size. Quality is the essential consideration, and should have preference over first cost. No make of transformer should be permitted on a station's cir- cuit that does not maintain its voltage well within 3 per cent from full load to no load. The simple rule regarding size is to use as large units as possible, and thus reduce the number of units as far as the distribution of service permits. Every alternating station should aim to so improve regulation as to permit the satisfactory use of 3.1- watt lamps. Good regulation is eminently important to preserve the average life and light of the lamps, to prevent the increase of power consumed by the lamps, and to permit the use of lamps of lower power consumption, so that both the efficiency and capacity of the station may be increased. Constant voltage at the lamps can be maintained only by constant use of reliable portable instruments. No switchboard instrument should be relied on, without frequent checking by some reliable standard. Owing to the varying drop at different loads, constant voltage at the station is not what is wanted. Pressure readings should be taken at customers' lamps at numerous points, the readings being made at times of maximum, average and minimum load. Not less than five to ten readings should be made at each point visited, the volt-meter being left in circuit for four or five min- utes, and readings being taken every fifteen seconds. The average of all the readings gives the average voltage of the circuits. Lamps should be or- dered for this voltage, or if desired, the voltage of the circuits can be re- duced or increased to suit the lamps in use. The practical points are to determine the average voltage at frequent periods with a portable volt- meter at various points of the circuits, and then to arrange the voltage of the lamps and circuits so that they agree. Candle-Hours — The It emulation of lamp Value. The amount of light given by lamps of the same efficiency is the only proper measure of their value. The amount of light given, expressed in candle-hours, is the product of the average candle-power for a given period by the length of the period in hours. Many of the best central station managers consider that a lamp has passed its useful life when it has lost 20 % of its initial candle-power. In the case of a 16 candle-power lamp, the limit would be 12.8 candle-power. The period of time a lamp barns until it loses 20 % of its candle-power may therefore be accepted as its useful life. The product of this period in hours by the average candle-power gives the '* candle-hours " of light for any given lamp. The better a lamp maintains its candle-power under equal conditions of comparison the greater will be the period of "useful life," and therefore the greater will be the "candle-hours." This measure is, therefore, the only proper one with which to compare lamps and determine their quality. INCANDESCENT LAMPS. 547 The practical method of comparison is as follows : Lamps of similar candle-power and voltage are burned at the same initial efficiency of 3.1 watts per candle on circuits whose voltage is maintained exactly normal. At periods of 50, 75, or 100 hours the lamps are removed from the circuits and candle-power readings taken, the lamps being replaced in circuit at the end of each reading. Readings are thus continued until the candle-power drops to 80 % of normal. The results obtained are then plotted in curves, and the areas under these curves give the "candle-hours" and the relative Value of the different lamps. Variation in Candle -power and Efficiency. In the following table is shown the variation in candle-power and effi- ciency of standard 3.1 watt-lamps due to variation of normal voltage. Per Cent of Normal Voltage. Per Cent of Normal Candle-power. Watts per Candle. 90 53 4.68 91 57 4.46 92 61 4.26 93 65 4.1 94 69* 3.92 95 74 3.76 96 79 3.6 97 84 3.45 98 89 3.34 99 94* 3.22 100 100 3.1 101 106 2.99 102 112 2.9 103 118 2.8 104 124* 2.7 105 131* 2.62 106 138* 2.54 Example: Lamps of 16 candle-power, 105 volts, and 3.1 watts, if burned at 98 % of normal voltage, or 103 volts, will give 89 % of 16 candle-power, or 14J candle-power, and the efficiency will be 4.34 watts per candle. lamp Renewals. The importance and necessity of proper lamp renewals applies forcibly to all stations, regardless of the cost of power, and whether lamp renewals are charged for or furnished free. The policy of free-lamp renewals at the present low price of lamps is, however, preferable for both station and cus- tomer. Free lamp renewals give a station that full and complete control of their lighting service so requisite to perfect results. Points to be Remembered. That a constant pressure at the lamps must be maintained. That the lamps are not to be used to the point of breakage — they should be renewed when they become dim. That satisfaction to customers, and the success of electric lighting, are dependent upon good, full, and clear light, which old, black, and dim lamps cannot give. 548 ELECTRIC LIGHTING. That to furnish a good, full, and clear light is as much a part of the light* ing company's business as to supply current to light the lamps. That a company should always endeavor to keep, the average life of lamps within 600 hours. That to renew dim lamps properly en the free renewal system, inspectors should examine the circuits regularly when the lamps are burning. If lamp renewals are charged to customers, induce them to exchange their dim lamps. Ltmiiiio*it v of Incandescent Lampi. As showing the quality of incandescent light, we present here a curve showing the relative luminosity of an incandescent lamp at different regions of the visible spectrum. On this subject Prof. E. L. Nichols states the following : " The most important wave-lengths, so far as light-giving power is con- cerned, are those which form the yellow of the spectrum, and the relative LUMINOSITY OF INCANDESCENT LAMP CFERRY.) 400 RELATIVE WAVE LENGTH RED ORANGE' YELLOW Fig. 13. Regions of Spectrum. luminosity falls off rapidly both toward the red and the violet. The longer waves have, however, much more influence upon the candle-power than the more refrangible rays. " Luminosity is the factor which we must take into account in seeking a complete expression for the efficiency of any source of illumination, and the method to be pursued in the determination of luminosity must depend upon the use to which the light is applied. If we estimate light by its power of bringing out the colors of natural objects, the value which we place upon the blue and violet rays must be very different from that which would be ascribed to them if we consider merely their power of illumina- tion as applied to black and white. In a picture gallery, for instance, or upon the stage, the value of an illuminant increases with the temperature of the incandescent material out of all proportion to the candle-power, whereas candle-power affords an excellent measure of the light to bo used in a reading room. INCANDESCENT LAMPS. 549 Metallized Carbon or Gem lamps, — The so-called " metal- lizing" process as applied to carbon filaments consists in heating the filaments to an enormously high temperature both before and after flashing, using a carbon tube electric furnace for the purpose. The term "metallized" is applied on account of the positive temperature coefficient which the filaments acquire in the process. The useful life of the metallized filament lamps at 2.5 w. p. c. is said to be the same as that of the ordinary carbon lamp at 3.1 w. p. c. label Hating* of Gem Lampi. The style of label employed for Gem Lamps is as here shown. These labels are printed for all the voltages from 100 to 130 and for the various sizes of lamps. As shown in the cut of label, only the total wattage of lamp and the volts are printed. Candle-power values are not given, as these values vary with the different forms of reflectors. (See candle-power distribution curves.) The voltage markings are arranged to show three voltages in steps two volts apart, and this provides a ready method of varying the efficiency and life of lamps to suit different conditions. The values at each of the three voltages are shown in the following table : Lamps should, of course, be ordered at the " Top" or first voltage (VI) whenever possible, so as to secure the full lighting value and maximum efficiency and brilliancy. Fig. 14. Table of Values at 1st, 2nd and 3rd Voltages. Voltage of Circuit. Per cent Total Watts. Per cent of c. p. Values. Eff. in w. p. c. (mean horizontal c. p.) Useful Life in hours. Same as " Top" or 1st Voltage (VI) Same as "Middle " or 2nd Voltage (V2) Same as ■ • Bottom ' ■ or 3rd Voltage ( V3) 100% 95% 90% 100% 90% 80% 2.5 2.65 2.8 500 700 1,000 TA]¥TAirM LAMP. The filament of this lamp is a fine wire of metallic tantalum. The high melting point and low vapor pressure of this metal make it possible to operate the lamps at 2.0 w.p.c. with a life comparable with that of the ordinary lamp at 3.1 w. p. c. The life on alternating current is much shorter than on direct current and is a function of the frequency. Fig. 15 shows free-hand drawings of microscopic views of the tantalum filament as affected by rise on alternating and direct current. The vertical distribution of intensity changes during the life of the lamp, the horizontal intensity diminishing more rapidly than the spherical, due chiefly to more rapid bulb blackening in the horizontal zone. On this account the spherical reduction factor also changes. (See Fig. 16.) Characteristic life curves of tantalum lamps manufactured in Germany in about 1904, are shown in Figs 17 and 18. These tests were made in the Electrical Testing Lab- oratories. (See Sharp, New Types of Incandescent Lamps, Proc. A. I.E.E- 10Q5, p. 809.) 550 ELECTKIC LIGHTING. ALTERNATING CURRENT 130 rv 300 HOURS ALTERNATING CURRENT 60 rv t57 HOURS ^T^F ALTERNATING CURRENT 25 *V 467 HOWRS jTy*^ yrnr ^^7rfrir^^\ ^rf>m m ' ■■ ..i.* m i l .. >f f " T *' -* ;.■ >, . . ■ ■■ t — - OIRECT CURRENT 492 HOURS NEW LAMP Fig. 15. Microscopic Views of the Tantalum Filament. Fig. 16. INCANDESCENT LAMPS. 551 fl v> 200 300 HOURS 400 600 6oa 700 81 30 29 28 27 h i i LIFE CURVES OF REPRESENTATIVE TANTALUM LAMPS TE6TED ON DIRECT CURRENT AT RATED VOLTS ELECTRICAL TESTING. LABORATORIES h \ 1 \ \ \ \ A \ -■ LAMP NO. 3 9 26 25 ff 24 s v §3* \ X C h A \ ^ h V z m \ "Sjw VTTS *•* fc 16 < v*« "? «U 45 Ik "■•V .--. -■*C *"■**« ^ y ^ L \ * s 17 lfi 3.0 <^ •< X HFBI ;al I .-P> 50 U^ ?^- J*? ^ *>c; ■'- —. — .--* ^ [2.fi 05 S J^ --* 3B^ " 18 2rT 1( )0 2 JO 34 to u p »0 6( ?( )0 Fio. 17. 552 ELECTRIC LIGHTING. HOURS 200 300 100 200 300 100 500 HOURS Fig. 18. Curves of Tantalum and Carbon Lampe. INCANDESCENT LAMPS. 553 TuaresMsar iamp. The metal tungsten has such a high melting point (about 3200° C. f according to Waidner and Burgess) that when it has been worked into lamp filaments by special processes, the lamps so produced can be operated at very high efficiencies. The most favorable condition of operation is at about 1.25 watts per candle, which is the point most commonly selected by makers in this country. Even at this low value the rate of deterioration in candle-power and efficiency is very slow until near the end of the life when it may become very rapid. Because of the high conductivity of the metal it has as yet been found impracticable to make 100-volt lamps of lower candle-power than 25, but tungsten lends itself admirably to the construc- tion of heavy current, low-voltage lamps for use in series on constant current street-lighting circuits. Life curves of six German- made tungsten lamps (Osram lamps) are shown in Fig. 19. Fig. 19. Tests of Six Tungsten Lamps. Effect of Chang-es in Voltug-e. Change with 5 per cent increase in voltage above normal. Candle-power. Watts per Candle Carbon +30% -15% Metallized +27% -13% Tantalum +22% -11% Tungsten +20% -10% 554 ELECTRIC LIGHTING. 95 100 105 PERCENTAGE VOLTS Fig. 20. Characteristic Curves of Tungsten, Tantalum, Metallized and Carbon Lamps. FfG. 21. Tungsten 60-Watt Lamp. These data show that Tungsten lamps are less affected by voltage fluctuation than are other lamps. The operating temperature of the Tung- sten is so high that the light is of peculiar and agreeable whiteness, much better fitted for the matching of colors than is that of the carbon lamp. Lamps are now on the market rated at 100, 60, and 40 watts for use on circuits of 100 to 125 volts. Series burning lamps for street lighting are also available. The life of a Tungsten multiple lamp at 1.25 watts per candle is said to be 800 hours. The following statement and table are from a Bulletin of the Engineering Department of the National Electric Lamp Association. Operating* Cost. — Table 5 shows the total cost of operating various lamps on various costs of power. The combined cost of power and lamp renewals, for a period of 1000 hours, shows the saving effected by the use of high efficiency lamps, when the cost of power is high. At costs of power greater than two cents per kilowatt hour the Tungsten lamp is the cheapest, considering both the first cost of the lamp and the cost of power, excepting only the higher candle-power Tantalum lamp. The latter is cheaper than the Tungsten at costs of power below two and one half cents per kilowatt hour on 60-cycle alternating and four cents per kilowatt hour on direct current. INCANDESCENT LAMPS. 555 WHEN Atfl> HOW J[]¥CA]¥I>Ef8C 1 E]¥T LAMP! ARE I ISJEI*. By Mortimer Norden. The following data have been collated to show the yearly consumption of current per 16 c.p. lamp on the circuits of a large central station company, giving the yearly average of current used in kw.-hours. The data represent ten plants all operated by the one company : Totals of JLverag-e Consumption, Showing* Yearly Con- sumption per IG-c.p. Lamp Connected. 1 Green house 24 Colleges and schools 127 Churches Q 3 Parks 1343 Residences 64 Dentists' and physicians' offices 344 Factories 8 Signs 14 Public halls 6 Dressmakers 1 Grain elevator 102 Municipal buildings, hospitals, armories and city halls 104 Clubs and lodge rooms 147 Nine o'clock stores 401 Seven o'clock stores 449 Eight o'clock stores 137 Livery stables and stables 26 Eleven o'clock stores 287 Office buildings and offices ...'.... 10 Theaters 9 Road houses 45 Banks and insurance companies 11 Ten o'clock stores 2 Cold storage companies 4 R. R. terminals and docks 180 Drug, confectionery and cigar stores . . . 640 Saloons, restaurants and concert halls . . 327 Six o'clock stores 22 Wholesale butchers 25 Commission dealers 8 Twelve o'clock stores 3 Steamship docks 5 Hotels 23 Railroad stations 2 All night stores 4904 customers, Lights. Kw.-hours. 54 1.33 2,863 5.70 11,616 7.75 416 9.24 40,095 10.73 1,066 15.10 21,936 15.53 365 18.48 1,781 18.81 111 20.24 24 20.75 14,654 24.79 7,391 24.82 4,433 26.35 17,623 26.55 13,228 27.10 1,775 29.56 624 30.01 7,363 30.65 10,581 32.13 305 32.70 3,322 33.80 339 38.34 158 40.82 854 42.14 4,370 42.44 17,592 43.62 23,584 45.61 1,012 46.92 518 48.06 170 52.44 2,293 61.71 1,099 65. 909 118.98 410 218.06 214,934 Grand average 27.28 556 ELECTRIC LIGHTING. o a 33 r cd V © s cti o o a ©T* bB a> •g P DO r= © +J ~ > er. 43 ■ci C^ 0) cti s- V >£ Js tH -r W qj O 2 <»^ ^ as Hgce c$ ^ •-— ft Cos o|| 32a P of © ee £ © < -> v 2 2:5 si* •**,* © q 0Q-£ *gg c 1 -' © ^ © © ■OS" G O O c3 S cc © s3 H~ . 3-- 88 8- HnN Ci Cq CM dOlCl 9^ © © t- ^ COO© ohm coco-* m©t» i o o i t>* oc od cs os OOO O *"■«« 558 TH tQ Tt< O H/ 1 CSTfOT^CSTf CS t(* CS "# CS ^ IO 74 O t>- O C t- b- t- t- t- t- t> b- l> t- t- t»I>t- X XX 000000 000000 000000 000000 o oi hj cq oq o ct>-t- t-t-t- t>l>t- ooo ooo ooo ooo ooo COL-00 OOtH CNCO'* OCOt- 00CSO CO CO CO tj5 o r-t O CO 8 Ot^cn 00 TH "H^ CO tH CO tHOOO Ol-O THrHCq cq "* t- cqcqcq cq©o StHC^ cq'coeo cq o co © 00© cocotj" CNtHtH W -A ~A co co O CJ t>- csco CS CS o CO ?J s T* OOO CO O OS CS O T-H CO rt* COtH ooo o t>- p OOO CO o oo oo tH COO cq'cq cq* o cocq 00 CO 00 t-- O H cqci co ©ooo CO t- t» CO o t- thcoco po co* oi* OCO o o' 80 00 CO r^COCO 283 00OO ©CO© T* IO t- th cq co cq^o coo © ^uqi^ oo©cq oc p© rHTHCq* CO T ~l y ~l tHCO CO cqo CS t>" •^ CO o CO rH cq t- cq op t> t-coco co co co 00 o o CO ©© t-h cq cq cq cq th CO CO CO ■^ o© ©csoo CO lQ o cq co ■* o© t- oo I ©THcq corj- ©no © co 10 qoh-* t> © co © cd cd cd tj5"^tj5 ioid»o ooo ot^b-" t^cdoc od © co two 000 000 000 000 000 © © t- 00 © CO 1 rf O 00 © q^O 00 © CO rf p 00 © co co co co co co cd cd ^ ■* «^ -^ ^00 id id id © S 8 © 1 cd co" cd •&T*ui id id © ot>t^ t^ od od oioioJ © © ^ §£8 883 888 888 888 888 8 cocon; poop cort

OS i-h COiOt> OS th CO lO t> OS i-J rH i-i r-i CO CO CO CO CO CO COCOCO CO^*rJ* Ti" Ti* Tji id COCOCO COCOCO COCOCO COCOCO COCOCO COCOCO CO ■^>*^ ©«* © "fan* ©rt< os <*©<*> © -* © rj< rHOH CO © 00 1-HCOp 00 rH CO OOOr-t CO © CO r* © ©' i-i co cd ■<* o't^od os i-i co" cd ■<* cd t^ od os i-" TH t-< r-t rlHH i-H t- tH 1-1 CO CO CJCJ(N (M(N^ CO ©">* © ©«3< CO © t-"* CO ©b- M< CO © t~"*C0 © t- u0 CO © © H/i t^i-HUO © CO © © Tfi t- 1-1 lO © CO p © tjh co co ci hho © © oo oq t> o p iq -* rjj t>ooo» os © i-5 cicd"^* rj5 id o t> oo" os dine" ci 1— I i— I wHrl HWH r-l i— I i— I CO CO CO CO cococo oo co oo co oo co ooeooo coooco oocooo co «*©<* ©©i-i t*h©OS ^->*0 ©i-irf O © tH ^ HMO CO CO lO th t> CO OOO? CO lO t-h b- CO © © id id id © ©" t> cd cd © © © ^- i-H c* cd co tj" id id b-b-t>- t~b-b- b-b-t- b-b-b- t~t>£r b-b-b- b- 00 00 00 000000 000000 00 00 oo -co 00 00 CO 00 00 CO © CO ^ © "^ OS rji © "* © H- OS nj © *# © -^ OS rj< TjHrJ*rji rj" id id © ©" b* t^COod ©OS© ©i-ii-" co 1-1 b- o 1C b-b-b- b-b-b- b-b-b- b-b-b- b-b-b- b-b-b- b- S88 88S 888 283 828 882 8 C? co co co co" cd cococo nHT^^ -^idid id»do o MrtOrllNM ^Ot- 00 © O 0Q CO «* lO O 00 OS H^t>U«HC0 U0C0© ONH oooo? ©pco © COiqt>CO©r^ ©Oth t^coos 1*00 ^"^^ P ^"^tj* idido t^t^od odosoi di-*i-5 oScnco co U5rJ IOCOtH CSt>-UO CO 1-1 OS t-icco ^©t>- o JS co co o os o? rjj «>; © co p oq i-j rt^ t> oNiq oq co ifNO co©'* co © co ©^co ©99P 5M lOOlO ©10© -tic:'* CS-#00 CO0CCO OONt- OJ © J-H CO S l> © OJ Th C^ © CJ TJJ t> © - © CO cococo cococo cd cd cd ed-^rt" ^^id ididid o •SJtiOH 0001 *°J jo ^soo patuqiiioo jo 9iq^x lOOO ©lA© ift©UO ©ic© o © o rtiiouo ppt^ £r2£S£ sssp ©r^r" ©©© ©©© J9d J8AV0^ JO ^SOQ o w 8m «o 5§ 558 ELECTRIC LIGHTING. COOPER-HEWITT ^lEHCI HY VAPOR LAMP, Characteristic*. — This lamp is an arc lamp rather than an incandescent lamp, the arc having a mercury cathode and passing through vapor of mercury at low vapor tension. The light probably results from the electro-lumines- cence of the mercury vapor and not from any very high temperature pro- duced either at the anode, the cathode or in the arc stream. Being produced in this way, the light shows not a continuous spectrum of ail the colors, but a discontinuous or line spectrum characteristic of mercury. The per- centage of the electrical energy which is converted into light is relatively high, and the lamp is very efficient. It would constitute for many purposes an almost ideal source of light were it not for the unfortunate fact that in the spectrum of mercury red is almost entirely lacking. The result is that the light of this lamp has a tint which to most people is very distasteful, namely a strongly greenish hue. Red objects look black or purple in it, and all colors containing red are falsely rendered. When this character- istic is not objectionable the Cooper-Hewitt lamp may be used to good advantage. It is asserted that the light is very favorable for the eyes, causing little fatigue. It has been used in draughting rooms to some ex- tent. Its actinic powers are high, due to the presence of bright violet lines in its spectrum, hence it Is a desirable source of light for night photography, for copying, blue-printing, etc. Photometry. — The photometry of the Cooper-Hewitt lamp is attended with considerable difficulties due to the large linear dimensions of the lamp, and to the wide divergence of the color of its light from that of other sources. As a result of its large linear dimensions it is necessary to place the lamp at a considerable distance from the photometer. For distances which are small in comparison with the length of the lamp, the intensity of the light does not diminish as the square of the distance. The difficulties due to the color of the light are two-fold. In the first place photometer settings are difficult and uncertain to make unless a nicker photometer is used, and the personal equation of the operator is a large one. In the second place what is known as the Purkinje Phenomenon plays an important part in the results. This is a physiological effect, according to which if a reddish and a greenish or a bluish light appear equally bright when the intensity of each is high, the reddish light appears much fainter than the other when the intensity is greatly diminished. It follows from this that the apparent candle-power of the Cooper-Hewitt lamp when photometered against an ordinary standard, such as an incandescent lamp, is higher the farther the lamps are removed from the photometer or the dimmer the illumination on the photometer disk. In order to get even approximately accurate results in the photometry of this lamp a standard illumination on the photometer disk must be chosen and adhered to. No such standard illum- ination has as yet been designated and established. The following matter is condensed from an article in the Electrical Age. The Cooper-Hewitt lamp consists essentially of a glass tube, from which all the air has been exhausted, but which contains a small amount of liquid mercury and is filled with mercury vapor. At the ends of the tube are means for introducing the electric current. At the positive end the tube swells out, forming a chamber, which is called the condensing chamber. A platinum wire is sealed into each end of the lamp. At the positive end the wire connects either with a small puddle of mercury or a piece of iron, according to the type of electrode used, and this constitutes the positive electrode, or anode. At the negative end the wire connects with a small puddle of mercury constituting the negative electrode, or cathode. The lamp may be made of such dimensions as to make it suitable for a direct-current line of any assigned voltage. Most lamps are designed to run at a pressure of about 115 volts. A lamp about 4 feet in length and 1 inch in diameter would be suitable for this voltage and would work best on a current of about 3 amperes. Before being started, the electrical resistance of a mercury vapor lamp is very high. This negative electrode resistance to starting may be almost totally destroyed in various ways. One method consists simply in tilting the lamp until the two electrodes are brought into connection by a thin stream of liquid mercury along the length of the tube; then, upon tilting back, INCANDESCENT LAMPS. 559 an arc is started which prevents the high cathode resistance from making its appearance, and the lamp continues to operate until the current is turned off. Another method of starting is to send a small, momentary high- tension current from an inductance coil through the lamp, which at the same time is connected with the low-voltage mains. This high-tension current penetrates the high cathode resistance, and the current from the low-voltage mains follows, and if this latter current be great enough the high cathode resistance does not again make its appearance until the cur- rent is turned off; and if it is desired to relight the lamp the same procedure has to be repeated. To facilitate the starting of the lamp by this method the so-called "starting band" is employed. This is simply a narrow, thin, metallic band attached to the outside surface of the lamp in the neighbor- hood of the cathode, and connected by a wire to the positive terminal of the lamp. In the latest model of automatic lamps this operation is accomplished by the use of a "shifter." This consists of an evacuated glass bulb con- taining mercury which is shifted by the action of an electromagnet when the circuit is closed and which interrupts the current. Thereby a high potential is induced which starts the lamp. A view of this lamp is shown in Fig. 22, of the interior of the auxiliary box in Fig. 23, and a diagram of connections in Fig. 26. ^ Fig. 22. Type " P " Lamp. The resistance which a mercury arc offers to the passage of electric current may be separated into three distinct parts: — First, the resist- ance encountered by the current in passing from the anode into the vapor ; second, the resistance of the vapor column itself; and third, the re- sistance encountered by the current in passing from the vapor into the cathode. In the commercial lamp the potential drop over the anode is about eight volts and is approximately independent of the magnitude of the current flowing and the diameter of the tube. The anode resistance, then, varies inversely with the current. The potential drop over the cathode is about five volts and is approximately independent of the diameter of the tube and of the magnitude of the current flowing, provided that the current is above a certain minimum value, depending upon the inductance and re- sistance in series with the lamp. If the current falls below this minimum value, the cathode resistance immediately becomes enormous and the lamp is extinguished. A certain amount of inductance and resistance is usually placed in series with the lamp, as this has a beneficial effect, caus- ing the lamp to operate more steadily. In fixing the resistance of the vapor to the passage of the current, four quantities predominate, namely, the length of the tube, the diameter of the tube, the magnitude of the current, and the density of the vapor. The results can be roughly expressed as follows: — The resistance of a lamp increases directly with its length; it decreases with increase of its diameter and at a greater rate when the current and diameter are small and the vapor density large; it decreases with increase of the current and at a greater rate when the current and diameter are small and the vapor density large; it increases with increase of the vapor density and almost 560 ELECTRIC LIGHTING. directly, although at a certain value of the density (varying -with different currents and different diameters), the rate of increase changes somewhat abruptly and is less for values of the density greater than this value than it is for lesser values. When the vapor density is quite high, say for values greater than those corresponding to a pressure of three millimeters of mercury, the luminous column no longer fills the tube; and when the density is very high it is of very small cross section and passes along the axis of the tube. The vapor 08E THESE POSTS FOJ» 110-1.20 VOLTS T00-T10 « Fig. 23. Auxiliary for "P" Lamp (casing removed). 1. Ceiling Plate. 2. Nipple. 3. FlG 24 Wiri Diagram. Insulating Joint. 4. Binding Type "P" Lamp Post for Main. 5. Inductance JH H * Coil. 6. Ballast. 7. Shifter. 8. Actuating Armature. 9. Terminal Block. pressure of a lamp operating under normal conditions is in the neighbor- hood of one millimeter of mercury. It has been observed by Dr. Hewitt that there is a value of the vapor density at which the light efficiency of a lamp is greatest, and lamps are designed to run at this density when they are to be operated under com- mercial conditions. In order to maintain the density at the proper point the condensing chamber mentioned at the beginning of this article is em- ployed. This chamber usually, though not necessarily, surrounds the positive electrode at the upper end of the lamp. By virtue of its size it has a considerable radiating surface exposed to the air, and consequently the temperature within it, except in that portion of it which is quite close to the electrode, is low, compared with that in the other parts of the lamp. In consequence of this the pressure also is low in this region, and the mer- cury vapor from the main part of the tube rushes into the chamber and con- denses there. The effect of all this is to keep the vapor density in the con- ducting column at a lower value than it would otherwise assume. By making the condensing chamber of the proper dimensions, the vapor density can easily be made that corresponding to the greatest light effi- ciency. In connection with all this, it should be remembered that the mer- cury at the cathode is continually vaporizing, owing to the heat produced INCANDESCENT LAMPS. 561 by the current. After condensing in the condensing chamber the mercury falls back into the cathode end, and after a while again takes its turn at being vaporized. The efficiency is said to be somewhat higher than that of the arc lamp, and much higher than that of the incandescent light. Fig. 26. Diagram Illustrating the Method of Op- erating Lamps in Series. * -MMH QUICK-BBEAK SWITCH RESISTANCE — \smsmsisuir— INDUCTANCE COIL STARTING BAND & Fig. 27. Diagram illustrating the method of starting by high-tension dis- charge. To light the lamp, the main switch, which is mounted on a small panel board, is closed, and then the lever handle on the quick-break switch is pressed down, thus completing a circuit through the series resistances and inductances, charging the coil. On releasing the handle the quick-break switch automatically opens the circuit and the discharge of the coil passes through the lamp, breaking down its resistance and establishing a path for the main current. 562 ELECTRIC LIGHTING. THE * E1J \*T IAMP. Early in 1898 Dr. Walther Nernst exhibited in this country his new type of incandescent electric lamp. Mr. Westinghouse purchased the patents and placed at work upon it a staff of engineers, who have developed it into the present commercial form in this country. The light -emitting element of the lamp as developed by the Nernst Lamp Company of Pittsburgh, is termed a "glower." It is made by pressing through a die, a dough composed of the oxides of the rare earths mixed with a suitable binding material. The porcelain-like string thus formed is cut, after drying, into convenient lengths. It is then baked, and ter- minals are attached, by means of which a current of electricity may be passed through the glower. The glower of a standard 220-volt Nernst lamp is about 1" long by 3Y' in diameter. It is an oxide incapable of further oxidation, therefore „ HOLDING SCREW I ALUMINUM PLUG I 1 ARMATURE SUPPORT/ — L.POST> ballastI ■-- cut out coil 1 armature 1 ~-tct silver contact stop " housing i contact sleeve porcelain/ globe holding screw holder porcelain) HEATER PORCELAINf 0.0^ HEATER TCIBEl" GLOWER' GLOBE Fig. 28. operative in the open air. The presence of oxygen is essential. Glowers are insulators when cold, but become conductors when hot, hence they must be heated before they will conduct electricity sufficiently well to maintain themselves at a light -emitting temperature. The characteristic of the glower with reference to voltage is as follows: — As the current traversing the glower is increased, the voltage across its ter- minals rises, at first rapidly and then more and more slowly to a maximum; it then drops off with increasing rapidity as the current through the glower and the resulting temperature continues to increase. The glower is oper- ated on the ascending part of the curve at a point just preceding that of maximum pressure. Beyond this point the rapid decrease in the resistance of the glower makes the current difficult of control without a steadying resistance in series with it. This ballasting is accomplished by means of a fine iron wire mounted in a small glass tube filled with hydrogen. The diameter of this wire in a 0.4 ampere ballast is about .045 mm. Iron wire possesses, on reaching its so-called critical temperature, the property of increasing its resistance with great rapidity with rising temperature. THE NERNST LAMP. 563 The negative resistance temperature coefficient of a glower may thus be more than counter-balanced by the temperature coefficient of the iron *vire ballast placed % in series with it. For a 10% rise in current the resist- ance in the ballast increases 150%, so that a glower thus protected at once becomes operative through a wide range of voltage. The construction of a commercial lamp requires a device to heat the glower in starting. The heaters consist of thin porcelain tubes wound with fine platinum wire which in turn is held in place and protected from the intense heat of the glowers by a refractory paste. The automatic lamp is constructed with a cut-out to disconnect the heater from the circuit as soon as the glowers light. A general idea of the construction of the lamp and of its principal parts together with an understanding of its electrical connections may be gained from a study of Figs. 28, 29, and 30. Lamp Utrm/nals Fio. 29. SGlower Lamp Fig. 30. The action of a Nernst lamp when the switch is turned on is as follows: (1) The current passes through the heater, bringing it to a white heat; (2) the proximity of the glower to the heater results in the glower becom- ing a conductor, through which the current then passes; when the current through the glower has reached a predetermined amount; (3) the cut-out coil becomes energized by virtue of the glower current passing through it; (4) the armature of the cut-out which had heretofore closed the heater cir- cuit is attracted; and (5) this opens the heater circuit, leaving only the glowers in operation until the next time the lamp is turned on. Opening the switch which controls the lamp circuit allows the cut-out armature to fall into place again, thus connecting the heaters ready for starting. Efficiency. — It may be noticed that the efficiency of the Nernst lamp increases as the number of glowers increases. This is due to the fact that the glowers in the multiple glower lamps are operated in a highly heated atmosphere by virtue of the mutual heating effect of the several glowers. This causes the glowers to have a much lower voltage at the normal 564 ELECTRIC LIGHTING. current in the multiple glower lamp than is the case when they are operated in the open air, this difference amounting to about 16 volts in the six- glower lamp. Photometric Vests of Various Illuminants by National Electric JLiprlit Association. Illuminants. Multiple D.C. Arc. Multiple A.C. Arc. Nernst 6-Glower. Globes and Shades. Opal. Inner. Clear Outer. Opal. Inner. Clear Outer. Clear Globe. Opal. Globe. Clear H. C. Opal. Shade. E.M.F Current Watts Power Factor Mean Spherical c.p. . . Mean Hemispherical c.p. Watts per Spherical c.p. Watts per Hemispher. c.p. 110 4.9 529 1 182 239 2.90 2.25 110 6.29 417 .6 140 167 3.02 2.53 226 2.4 542 1 163.9 289 3.30 1.88 226.5 2.4 543 1 168.6 258.6 3.22 2.10 226 2.4 542 1 155.8 264.2 3.48 2.05 Illuminants. Nernst 3-Glower. Nernst 1 -Glower. Globes and Shades. Clear Globe. Sand Blasted Globe. Clear H. C. Opal. Shade. Clear Globe. Sand Blasted Globe. E.M.F Current Watts Power Factor Mean Spherical c.p. . . Mean Hemispherical c.p. Watts per Spherical c.p. Watts per Hemispher. c.p. 218.8 1.2 262 1 65.1 112.6 4.04 2.33 219.5 1.2 263 1 61.5 96.9 4.28 2.72 220 1.2 264 1 68.5 118.3 3.86 2.23 223.7 0.4 89 1 21.8 38.7 4.11 2.31 220.5 0.4 88 1 20.5 31.8 4.3 2.78 The British unit of c.p. used in above. The arc lamp figures were taken from the Report of the Committee for Investigating the Photometric Values of Arc Lamps, read before the National Electric Light Association in May, 1900. The Nernst lamp data were obtained from the report of the same committee which was presented at the Twenty-Sixth Convention in May, 1903. maintenance. — The frame and connections of the Nernst lamp form a permanent structure having an indefinite life, but its perishable parts have from time to time to be renewed. Of these, the ballast has a life averaging 25,000 hours. The heater has a life averaging about 8 months in ordinary use. The glower, however, like the incandescent lamp filament, has a practically definite term of use at the end of which it would be advisable to replace it whether burnt out or not. 800 hours are given by the company as the guaranteed life on 60 cycles. Behavior on Alternating- and IMrect Current. — Unlike the carbon incandescent lamp the life of glowers is not the same on direct current as on alternating current, and is affected even by the frequency of the latter. The American glower was constructed originally for use on alternating current only, while in Europe direct current lamps predominated. The direct current lamp in this country is a comparatively recent development and its glower life is shorter than that of the glower used with alternating current. THE MOORE VACUUM TUBE LIGHT. 565 THE TlOOltE VACCTM TUBE JLHSMT. The Moore Vacuum Tube Lighting System, invented by D. McFarlan Moore, has been in commercial service since 1903 and consists essentially of a glass tube about If inch diameter and of any form or length desired up to 200 feet. The tube is attached to the ceiling or walls by supporting XUBE DISTRIBUTED IN ANY FORM DESIRED TO LENGTHS OF 200 FT. DIAGRAM SHOWING ES'S^ENTfAL [ FEATURES OF THE MOORE LIGHT Fig. 31. fixtures at intervals of about 8 feet. A graphite electrode (3, Fig. 31) is sealed into each end of tube and the two electrodes enclosed within a steel terminal box (2) which also contains a static trans- former (4) and regulating device (6) called a feeder valve (see Fig. 32). The completed tube is exhausted to a pressure of about 1/10 mm. of mercury. The feeder valve performs the important service of feeding the tube some pure gas to take the place of that which is used up by the passage of the electric current through the tube. All vacuum tubes or bulbs through which current passes tend to attain a higher vacuum due to solidification or combination of the residual gases. There is a critical pressure, about 0.08 mm. of mercury, at which the con- ductivity is a maximum and the greatest current will flow. The pressure of maximum light efficiency is, how- ever, slightly higher, i.e., 0.1 to 0.12 mm., hence the feeder valve is adjusted to maintain the pressure at this point which it does as follows: A carbon plug (8, Fig. 31) is cemented into the mouth of the small bore tube (9) which connects to the lighting tube. This plug is normally covered by mercury the level of which is varied by the glass displacer (7) which also carries the iron core of the solenoid coil (5) which is connected in series with the transformer (4). As the pressure in the lighting tube falls the conductivity and therefore the current increases, and the plunger rises, allowing the surface of the mercury to fall and expose the tip of the carbon plug. A minute quantity of air FEEDER VALVE or whatever gas is supplied to the feeder valve, filters p IG 32 through the porous carbon plug and finds its way to the lighting tube, the action continuing until the pressure is brought back to 566 ELECTRIC LIGHTING. normal. The device is capable of very close adjustment. The transformer is usually supplied with alternating current at 220 volts and raises the voltage to 2000 volts or more, depending on the length of tne tube. The tube is self-starting and responds at full brilliancy instantly upon closing the switch. The intensity may be made anything desired from 5 to 50 candle-power per lineal foot, the normal commercial brilliancy being 12 candle-power per foot, the radiation being uniform in all directions in planes perpendicular to the axis of the tube. The efficiency is said to vary from 1.4 to 2 watts per candle-power depending upon the length of tube, the light intensity, etc., and is not affected by variation of supply voltage. See Fig. 33. In practice, tubes are said to have a life of from 3000 to 5000 hours and then can be renewed at small cost. The efficiency is said to remain constant after the first 50 hours' run. «,« S B^ 4 5* I 1 oC^ +£ in 4 ^> ■z % 15 3 "6000 z *T 10 2 4000 p i^v vt\^ p« D \ s / ^ ^ o 4 /, >• W/ ^TTS f ER H :fner C J 5 1 2000 / 25 50 75 100 125 150 175 ,200 225 LENGTH OFTUBE IN FEET Fig. 33. The color depends upon the gas supplied to the feeder valve. It i3 exactly the same shade of white diffused daylight when fed with pure nitro- gen, and orange-pink when fed with air. The intrinsic brilliancy is claimed to be the lowest of any known illumi- nant and therefore is extremely soft and agreeable to the eyes and does not require to be shaded or diffused to avoid glare but may be reflected to obtain any distribution desired. An intensity of 0.66 candle-power per square inch corresponds with 12 candle-power per lineal foot. Efficiency of Tioore Tube. Early in 1907, Sharp & Millar conducted a series of tests on a Moore tube that had been installed in Assembly Room, No. 7, of the United Engineering Societies Building, and reported the following results. The tube was 176 feet long and approximately 1| inches diameter. It was fed with nitrogen gas, and operated as a 60-cycle system. Total watts consumed by tube system 3451 Line volts 220.3 Amperes 21 . 5 Volt-amperes (apparent watts) 4736 Power factor . 73% Total lumens produced 17,400 Efficiency as light producer — lumens per watt .... 5.5 Lumens per apparent watt 3 . 68 Watts per equivalent mean spherical candle-power ... 2.49 Apparent watts per equivalent mean spherical candle- power 3.41 THE MOORE VACUUM TUBE LIGHT. 567 This installation of Moore Tube was compared with three installations of incandescent carbon filament lamps in the same room; they were as follows: No. 1, Installation Moore Tube. Installation No. Lamps under Tube. Installation No. Lamps in Rectangles. 3.—: Installation No. 4.- Lamps with Reflectors. Moore Tube, 176 feet long, running around the room close to the cove. One hundred 16-c.p. lamps placed horizontally 5 inches beneath the tube, and equally spaced. Eighty-four 16-candle-power lamps bare, ar- ranged in equal rectangles, 15 feet, 4 inches, above the floor. Same as No. 3, except that the lamps were equipped with Holophane distributing reflec- tors No. 7381. Results of the Comparative Tests. Instal- lation Number. Number of Lamps . Mean Horizon- tal c.p. Mean Spherical c.p. Watts per Horizon- tal c.p. Watts per Spherical c.p. Total Watts. 1 2 3 4 1 100 84 84 (per ft.) 8.1 13.82 11.31 11.11 (per ft.) 7.9 11.41 9.33 9.16 2.39 3.48 4.26 4.32 2.48 4.21 5.16 5.23 3451 4810 4040 4027 . [llumination Values Efficiency Values. 31 Foot Candles. Lamp. Gross . Net. a Maxi- mum. Mini- mum. Mean. Vari- ation. Lumens per Watt. Lumens Effective per Watt. Lumens Effective per Lumen Generated. i 2 3 4 4.38 3.27 2.10 2.51 3.18 2.28 1.16 1.26 3.69 2.69 1.71 1.97 16.2% 18.4 27.5 31.7 5.05 2.98 2.44 2.40 2.08 1.08 0.82 0.95 41.2% 36.2 33.6 39.6 The above table shows that, with regard to the uniformity of the dis- tribution of illumination, the Moore Tube performance was very good, but that the performance of the incandescent lamps arranged beneath the tube was practically the same. A disadvantage from which the Moore Tube suffers is that it flickers in unison with the alternating current which feeds it. On 60-cycle current this flickering is not noticeable, except when the eye is moved rapidly or when an object is moved rapidly before the eye. It then becomes notice- able, and for certain work is very objectionable. It, however, has the great advantage of throwing a very soft light of low intrinsic brilliancy, which does not need to be diminished by diffusing glasses in order to make it entirely bearable for the eye. The test shows that its efficiency, while not equalling that of the tungsten lamp, is about equal to that of the tantalum lamp, and greater than that of any other incandescent lamp. 568 ELECTRIC LIGHTING. ARC LAM IP* AN» ARC IIGHTIIira Revised by J. H. Hallberg, Consulting Engincm The arc lamp is an electrical apparatus in which an electric j»rc is struck and maintained between two or more electrodes, giving a brilliant illumi- nation, the color and intensity of which depends upon the composition and diameter of the electrodes, the kind of current supplied and the watts consumed. Owing to the extremely high temperature of the electric arc (varying between 2500 and 4000° C.) the electrodes must have a high volatilization point in order to obtain sufficient life from one set of them to make the lamp practical. Carbon has been found to be the most suitable material for the purpose. A pair of carbon electrodes of proper diameter to maintain a steady arc with a given current strength and voltage drop, will consume at the approximate rate of 1.25 inches per hour in open arc lamps, and .16 inch per hour in those of the enclosed type. If cross section of the carbon be too large, the arc crater will cover a comparatively small part of the carbon point. The shifting of the arc moves the crater to a cooler point, which makes a considerable change in the resistance of the arc. This change is so rapid that the lamp mechanism cannot compensate for it as quickly as required, hence a variation in the candle-power of the lamp which makes the use of carbons of large diameter impractical. With car- bons of too small cross section, the candle-power is greater, and the arc is very steady, but the life of the electrodes is too short for practical purposes. In Europe, the practice is to use carbons of comparatively small diameter, of extra length, or to trim often in order to secure perfectly steady illumi- nation at maximum efficiency. In the United States, the practice has been to use carbons of larger diameter, giving longer life with one trim and limiting the length of the carbon to about twelve inches, thereby reducing the cost of the carbons and labor required, but sacrificing steadiness of illumination and efficiency. Developments have been made in the manufacture of carbons for the flaming arc for open arc lamps, which have more than doubled their effi- ciency, and give four times the efficiency of the enclosed arc. The intro- duction of arc lamps with electrodes placed points downward at an angle to each other (instead of one above the other as in the old style of lamp) makes it possible to use carbons over twenty-four inches long, if necessary, without making the lamp impracticably long. The metallic oxide electrode has also been successfully developed, and open arc lamps commonly known as "magnetite" lamps have been put on the market and show a marked increase in efficiency over that of the enclosed arc. There are seven governing factors to be considered by the designer of arc lamps: 1. Steadiness of the light. 2. Watt consumption per useful candle-power. 3. Maximum practical length of the electrodes. 4. Length of life with one trim. 5. Cost of the electrodes. 6. Cost and reliability of the lamp. 7. Adaptability of lamp to the several systems of electrical distribution in general use. (ri^IIKATlOX Of ARC liAUEPS. Open Arcs, Direct Current : Ordinary open arc lamp with carbon electrodes. Series or multiple, 6 to 10 amperes, 45 to 50 volts at terminals for constant current series; 50 to 60 volts at terminals for constant potential multiple or multiple series operation. Life of carbons, 10 to 14 hours, approximately .6 watt per candle-power, clear globe. "Magnetite" arc lamp with metallic oxide electrodes in series only od ARC LAMPS AND ARC LIGHTING. 569 constant current, 4 amperes, 75 to 80 volts at terminals. Life of elec- trodes, ) 60 \iours, approximately .3 watt per candle-power, clear globe. ''FlarniLq:'' arc lamp, carbon electrodes with chemical core filling. Series or multip/1, 8 to 12 amperes, 45 to 50 volts at terminals for constant current series; 5^ to 60 volts at terminals for constant potential multiple or mul- tiple ser'.BS operation. Life of carbons, 10 to 18 hours, approximately .22 watt p^r candle-power yellow flame, approximately .3 watt per candle- powe", white flame, clear globe. Open Arcs, Alternating: Current: Ordinary open arc lamp with carbon electrodes in multiple only, 10 to 16 amperes, 40 volts at terminals — minimum practical frequency - — 60 cycles. Life of carbons 1\ to 12 hours, approximately .75 watt per candle-power, clear globe. "Flaming" arc lamp carbon electrodes with chemical core filling. Series or multiple, 10 to 14 amperes, 40 to 45 volts at terminals for constant current series; 50 to 60 volts at terminals for constant potential multiple or multiple series operation; minimum practical frequency, 25 cycles. Life of carbons, 10 to 16 hours, approximately .25 watt per candle-power, yellow flame; approximately .33 watt per candle-power, white flame with clear globe. Enclosed Arcs, IMrect Current : Ordinary enclosed arc lamp with carbon electrodes. Series or multiple, 3 to 7^ amperes, 75 to 85 volts at terminals for constant current series; 100 to 250 volts at terminals for constant potential multiple or multiple Beries operation. Life of carbons, 75 to 150 hours, approximately 1 watt per candle-power, clear globes. Enclosed arc lamp with inclined electrodes of pure carbon. Multiple operation, 8 to 10 amperes, 100 to 120 volts at terminals. Life of carbons, 30 hours, approximately .45 watt per candle-power, clear globe. Enclosed Arcs, Alternating* Current t Ordinary enclosed arc lamp with carbon electrodes. Series or multiple, 4 to 1\ amperes, 75 to 85 volts at terminals for constant current series; 100 to 120 volts at terminals for constant potential multiple, or multiple series operation; minimum practical frequency, 40 cycles. Life of carbons, 70 to 100 hours, approximately 1.33 watts per candle-power, clear globes. Enclosed arc lamp with inclined electrodes of pure carbon. Multiple operation, 10 amperes, 100 to 120 volts at terminals; minimum practical Frequency, 40 cycles. Life of carbons, 20 to 25 hours, approximately .6 watt per candle-power, clear globe. OPEI ARC I^AIflPS. Iow>Ten»on l^aiiip requires for most successful results high-grade carbons, cored positive and solid or cored negative. Lamps with either shunt or differential carbon feed-control, operate 2 in series on 100 to 125 volts, direct-current circuits with any current adjustment between 6 and 12 amperes. The arc should be set for an average of 42 volts, and sufficient resistance must be introduced in series with each pair of lamps to make up the difference between the required lamp voltage and the voltage of the supply circuit. Attempts have been made to operate from 4 to 10 lamps in series on constant potential circuits of 200 to 600 volts, but with only partial success. On alternating current the low-tension open-arc lamp requires a very high grade of carbon both cored and of the same diameter and length. The following are the best dimensions for the carbons: Ten amperes, 94 X \ inches; 14 to 16 amperes, 9£ X I' inches giving about 10 to 12 hours' life. The alternating current, open arc lamp requires about 30 volts at the arc with 35 to 40 volts at the terminals. The carbon feed is controlled by a simple magnet connected in series with the arc. The lamp is, therefore, a strictly multiple, 35 to 40-volt lamp, and requires special means for pro- 570 ELECTRIC LIGHTING. viding this pressure. For large installations a special transformer reduc- ing to about 35 volts is used. Where only a few lamps are required a small ("economy") single-coil transformer with taps for one, two, or three lamps is used. The illumination from the open arc, alternating current lamp has never been altogether satisfactory, mostly on account of low candle-power, exces- sive amount of violet rays, and noise. The low-tension, open arc lamp has not been in general use in the United States since 1900, having been superseded by that of the enclosed type. In Europe, however, this form of lamp has been in use until quite recently, as the enclosed arc was never very generally adopted there. The flaming arc lamp is now, however, replacing many of the other forms of open arc lamps. Hisrti-TViiMion lamp requires ordinary grade carbons, both of which may be solid, although in some cases it is of advantage to use a cored posi- tive. The usual carbon dimensions are, for 6 to 7 amperes, 12 X /* inch upper and 7 X ie inch lower; and for 9 to 10 amperes, 12 X \ inch upper and 7 X i inch lower. This is a strictly constant current series lamp operating any number in series up to the capacity of the generator. Con- stant current series arc generators have been built for single circuits of 175 to 200 lamps, requiring as much as 10,000 volts. Later practice is to build generators for 100 to 150 lamps, but bringing out leads for several circuits, thus reducing the maximum potential of the system and still securing the benefits due to the use of fewer and larger generators of higher efficiency. The brush multi-circuit arc generators, as built by the General Electric Company, represent the latest development in large arc-lighting units for direct current series lighting. The high-tension lamp has either shunt or differential carbon feel and is built for 6.8 amperes with 42 to 45 volts at the arc, usually rated at 1200 nominal candle-power; and for 9.6 amperes with 45 to 50 volts at the arc, rated at 2000 nominal candle-power. The high-tension series open arc lamp, operating on direct-current arc generators was the standard for street lighting in the United States until about 1900, since which time many of them have been replaced by enclosed arc lamps. The " magnetite " Arc Lamp is of the high-tension, direct-current, open-arc type metallic oxide electrodes. It is especially designed for outdoor lighting, to which it is limited on account of the fumes and heavy deposit from the electrodes. The positive electrode is made of pure copper, or from copper in combination with small non-conducting particles. Another form of positive electrode for this lamp is made of convoluted strips of laminated copper and iron, and is 1 inch long by \ inch diameter. The nega- tive electrode consists of a steel tube, tightly packed with a fine powder, the principal ingredients of which are: oxide of iron (magnetite), oxide of titanium and oxide of chromium. The steel tube serves as a conductor for the current to the crater and is also the holder of the oxide powder, making a binder unnecessary. The oxide of iron gives conductivity to the fused mixture when cold, the other oxides being conductors only when hot. The titanium oxide has the property of rendering the arc luminous. The oxide of chromium prevents too rapid consumption, thus giving long life to the electrode. Unlike all other arc lamps the maximum illumination in the "Magnetite" lamp comes from the negative end of the arc. The General Electric Com- pany have designed their "Magnetite" lamp with the negative electrode below the positive, while the Westinghouse Electric . C. Flaming- Arc lamp requires 50 to 60 volts at the terminals and is adjusted for 45 volts at the arc. One lamp operates in multiple on 50 to 60 volts, two lamps in series on 100 to 125 volts, four on 250 volts, ten on 500 volts, twelve on 600 volts, and fifteen on 750 volts. When more than two lamps are to operate in series, an external automatic cut-out with equalizing resistance must be put in multiple with each lamp to protect it against excessive voltage. The standard amperage is 10 to 12; the positive carbon is 10, and the negative 9 mm. in diameter. A pair of carbons 500 mm. long give 12 hours' life outdoors and 13 to 14 hours indoors; the 600 mm. carbons give 16 hours outdoors and 18 hours indoors. The Constant Potential A. C. Flaming* Arc lamp requires 50 to 60 volts at the terminals and is adjusted for 38 to 40 volts at the arc. One lamp operates in multiple on 50 to 60 volts, or two in series on 100 to 120 volt circuits. When one lamp is to operate in multiple on 100 to 120- volt circuit, a small auto-transformer is required to reduce the voltage to 50 or 55. Similar auto-coils should be used when lamps operate on 200 to 460- volt systems. When a large number of lamps are to be used a regular three- wire system can be installed with 55 volts between each outside wire and the center wire. One large transformer reducing from the primary poten- tial to 110 — 55 volt three- wire system — should be installed, allowing the flaming arc lamp to operate in multiple on 55 volts without loss and extra expense for separate auto-transformers or other compensators. The flam- ing arc lamp will operate successfully on any frequency from 25 to 140 cycles. Below 40 cycles, lamps should always be operated in multiple on 55 volts. The standard current adjustment is 12 amperes. The carbons are both 9 mm. in diameter. The 500 mm. carbon gives 10 to 11 hours outdoors and 11 to 12 hours indoors. Carbons 600 mm. long give 13 to 15 hours outdoors and 14 to 16 hours indoors. The alternating current lamp is practically noiseless and gives a very steady illumination. The efficiency of the alternating current flaming arc lamp on constant potential is about 80 per cent and the power factor about 90 per cent. The efficiency and quality of the illumination compares favorably with that of the direct current lamp, which is an important point in favor of the flaming arc lamp for alter- nating current circuits. The Constant I>. C. Series Flaming* Arc Lamp requires 45 volts at its terminals and is adjusted for 43 volts at the arc. The lamp is iden- tical in construction with the direct current constant potential lamp, but requires no resistance in series with the arc. An automatic cut-out is used with each lamp to shunt the current in case the carbons should stick or be prematurely consumed. The lamp can be operated in series on the regular 9.6 ampere arc dynamos used for the ordinary high-tension open arc lamps. The mercury arc rectifiers with constant current transformers can also be used to supply current for the direct current flaming arc lamp. As a matter of fact, it may be operated in series with the old style, high-tension, open arc lamp. The size and life of the carbons is the same as for the direct current constant potential lamp. The Constant A. C. Series Flaming* Arc Lamp requires 40 volts at the terminals and is adjusted for 38 volts at the arc. The constant current lamp is practically the same as that for constant potential, but is provided with an automatic cut-out to shunt the current. The lamp oper- ates with 10 to 12 amperes in series on constant current circuits controlled by constant current transformers or automatic reactive coils. As present alternating current series circuits for street lighting carry only 4 to 7£ amperes, it is necessary to install with each lamp on such circuits a small series transformer or series auto-coil which will deliver from its secondary 10 to 12 amperes at 40 volts to the lamp. In conjunction with series Tung- 18 to K.JKJH. 20 30" 35 50" 60 75 " 90 90" 100 ,25 '• 150 ENCLOSED ARC LAMPS. 575 sten lamps, operating on the same circuit, the entire street lighting field can be covered, furnishing both large and small units from the same wires. The size and life of the carbon is the same as for the constant potential alternating current lamp. The 500 to 600 watt direct current flaming arc lamp, with yellow flame carbons, gives approximately 2700 mean spherical candle-power; white flame carbons give about 2000 candle-power. The candle-power of the alternating current flaming arc lamp is about 10 per cent less than that given for the direct current lamp of the same watt consumption. Searchlight Projectors and focusing lamps for theatrical use and for photo-engraving, etc., take large and varied quantities of current, as they are always connected across the terminals of constant potential cir- cuits, with a regulating resistance in series with the lamp. The General Electric Company state in one of their bulletins the following as being the approximate currents taken by the different sizes of searchlights: Diam. of Projector. 12 inch 18 " 24 " 30 " 36 " 60 " EICLOMEO ARC IjAJKPS. It has been found that by enclosing the arc in a small globe, more or less approaching air-tight conditions, combustion of the carbons is practically complete, leaving no dust, and takes place at a slow rate, burning with a 12 X i-inch carbon 75 to 100 hours without attention. The enclosed arc cannot be properly maintained below 65 volts, and 70 to 75 volts is the usual arc potential for alternating current lamps, and 75 to 85 for the direct current lamp. The minimum current is 3 and the maximum for enclosed arcs is 1\ amperes. The long arc, low amperage and enclosing globe all tend to lower the illuminating efficiency of the enclosed arc lamp, but notwithstanding this it has superseded most of the open arc lamps for general illumination. The long life of the carbon has greatly reduced the cost of trimming and the cost of carbon renewals. It permits the use of very simple mechanism, actuating a clutch which operates directly on the carbon. Enclosed arc lamps are made for all commercial circuits. Constant Potential ». C. Enclosed Arc Lamp requires 100 to 250 volts at the terminals with 75 to 160 volts at the arc. The minimum amperage is 2\ and the maximum is 6. The 2\ to 4-ampere lamps use A to f-inch carbons. The 5 to 6-ampere lamps use & to ^-inch carbons, 12 inches long, giving 75 to 150 hours life. Each lamp is fitted with a resistance coil, and is a complete unit for multiple connection on 100 to 125 volts with 75 to 85 volts at the arc, or on 200 to 250 volts with 140 to 160 volts at the arc. The constant potential lamp is controlled by a series magnet. If the lamp is provided with differential clutch controlling mag- net, automatic cut-out and equalizing resistance, it can be connected in series on constant potential circuits, as follows: 2 on 220 volts, 5 on 500 volts, and 6 on 600 volts. Constant Potential A. C. Enclosed Arc Lamp requires 100 to 125 volts at the terminals, and is adjusted for 70 to 80 volts at the arc. The amperage maybe anywhere between 4 and 1\. The alternating current constant potential lamp is not operated in series. The power factor of the lamp is about 70 per cent. The minimum frequency giving satisfactory illumination, is 50 cycles; and the maximum frequency, in general use, for which this style of lamp is built, is 140 cycles. The carbons are usually 10 inches long X f to \ inch in diameter and give from 65 to 100 hours' life. When alternating current constant potential lamps are to operate on 576 ELECTRIC LIGHTING. voltages above 125, an auto-transformer or other converter for reducing the voltage should be used. A reactive coil is also put in the top of the alternating current lamp. Constant I>. C. Series [Enclosed Arc Lamp requires 75 to 80 volts at the terminals. The arc is set for 73 to 78 volts. The amperage is between 5 and 7, depending upon the candle-power desired. The lamp has differen- tial feed and is provided with automatic cut-out to shunt the current, if the carbon sticks or is consumed. The lamps operate in series on any constant current source of supply. The carbon is 12 X 2 inch and lasts about 100 hours. Constant A. C. Series Enclosed Arc lamp requires 75 to 80 volts at the terminals. The arc is set for 72 to 77 volts. The minimum amperage is 4 and the maximum is 1\. The feed control may be either shunt or differential. The carbon is 10 X \ inch and lasts 75 to 100 hours. Each lamp has an automatic cut-out. The lamps operate in series on con- stant current circuits, usually controlled by constant current transformers or automatic reactive coils. The efficiency of a complete system, including transformer and lamps, is about 85 per cent, and the power factor is between 70 and 80 per cent at full load. The system operates on any frequency from 50 to 140 cycles. Method* of Regulation in Arc Lamps may be classified as follows: Carbons lifted or separated by direct or main magnet; shunt magnet acting on a variable resistance to cut out the main magnet in feeding. Carbons lifted by main magnet as before, and shunt acting to put the main magnet (made movable) into position for feeding. Carbons separated by main magnet armature; shunt circuiting magnet acting to divert or shunt the magnetism of the main magnet from its arma- ture. Carbons separated by main magnet and shunt acting to free the carbon- holder, independently of the support given by the main magnet. Carbons separated by a spring allowed to act by the main magnet lifting a weight which otherwise holds the spring from acting; shunt magnet acts against the spring, to feed and regulate the length of arc. One carbon, generally the lower, separated by main magnet, while the other holder is released for feeding only, such feeding being under the con- trol either of a differential system or a shunt magnet only. Carbons separated by main magnet, which lifts the shunt and its arma- ture together, while the shunt magnet armature, acting on the feeding mechanism, controls the arc and feed of the carbons. Carbon feeding mechanism independently attached to main magnet arma- ture and to shunt armature so as to receive opposite movements of separa- tion, and feed from each respectively. Carbons separated by a feeding mechanism moved by the main magnet, and fed by a further movement of said mechanism, causing release or re- turn of same under the accumulated force of both shunt and main magnets, acting in the same direction. Differential clock gear for separation and feed of carbons under control of the regulating magnet system, either simple or differential. Some of the older clock-work lamps embodied this principle. Carbons controlled by armature of a small electric motor under control of a differential field which turns the armature in one direction for separating and in the other or reversed direction for feeding the carbons. Carbons controlled by a motor running at a certain speed when the arc is of normal length, and varying in speed when the arc is too short or too long, combined with a centrifugal governor on the shaft of the motor, acting on variations of speed to gear motor shaft to screw carbons together or apart, as needed to maintain the normal arc. This mechanism has been applied to large arc lamps, such as naval searchlights, and has the advantage of great positiveness, and an ability to handle heavy mechanism. There are also a considerable number of modifications of these principles. ENCLOSED ARC LAMPS. 577 Tents for Arc Ug-lit Carbons. For Open Arcs. The satisfactory working of arc lamps is largely dependent upon the quality of the carbons used. If carbons are made of impure materials, they will jump and flame badly. If not baked properly, they may cause annoy- ance by excessive hissing or flaming, or become too hot because of high resistance. If the material of which they are made has not been properly prepared in its preliminary stages, the carbons will have either too short a life, through giving a good quantity and quality of light, or will have good life, but will burn with an excessive amount of violet rays, hence with poor illumination. For indoor use a free-burning, uncoated carbon of medium life should be used, so as to give a good quality and quantity of light. If longer life is desired they may be lightly coated with copper without materially interfer- ing with the light. (About 1| lbs. to 2 lbs. of copper per thousand, /g" x 12" carbons, and a half pound more for \" x 12" carbons will give good results, increasing the life from an hour to an hour and a half.) For out-door use a more refractory burning carbon may be used to advan- tage, giving a longer life, as the quality of the light is not so important. Copper-coated carbons are also usually employed, and may have about four pounds of copper per thousand for /g" x 12" carbons, and five pounds for |" x 12". Other sizes in proportion. All plain molded carbons, and most of the forced carbons, deposit dust when burned in the open arc. Those depositing the most dust give out the most light, but have the least life. Those depositing the least dust usually have the longest life, but the light is of inferior quality on account of the increase in the proportion of violet rays. The quality of any carbon may be very quickly tested in any station by using the following method, which has been largely employed by carbon manufacturers. The important points to be determined are therang-e, including the hiss- ing, jumping, and flaming points, the resistance, and the life. The Xtangre is found by trimming a lamp with the carbons to be tested, allowing them to burn co good points and the lamps to become thoroughly heated; then connect a voltmeter across the lamp terminals, and very slowly and steadily depress the upper carbon until the lamp hisses, when the voltage will make a sudden drop. This is called the Hissing-Point, and varies according to the temper of the carbon. It should be between 40 and 45 volts — preferably 42 volts. Then lengthen the arc somewhat, and allow it to become longer by the burning away of the carbons. Presently the arc will make small jumps or sputters out of the crater in the upper carbon. This is the Jumping*- Point, and should be not less than 58 or 60 volts. Let the arc still increase in length, carefully watching the volt- age, and in most carbons there will soon be a decided naming. This is the flamingr-Point. This should not be less than 62 to 65 volts. Very im- pure carbons will commence to jump and flame almost as soon as the volt- age is raised above the hissing-point, and even the hissing-point in such cases is very irregular and difficult to find. The Range is important as being a practical test of the purity of the material used in the manufacture of the carbon, an increase of a quarter of one per cent of impurity making a very decided reduction in the extent of the Range. The hissing-point should be 4 or 5 volts below the normal adjustment of the lamp to insure steady burning. Resistance. — The resistance is measured on an ordinary Wheatstone bridge. Care must be taken that the contact points go slightly into the carbon. A T 7 g " x 12" plain carbon should have a resistance of between .16 and .22 ohms, and £" x 12" between .14 and .18 ohms. T 7 B " x 12" carbons coated with three pounds of copper per thousand, have a resistance between .05 and .06 ohms, and \[* x 12" with four pounds of copper between .04 and .05 ohms. JLife. — The life of a carbon is most easily tested by consuming it entirely in the lamp, observing, of course, the current and average voltage during the entire time. A very quick and accurate comparative test of dif- ferent carbons can be made, however, by burning the carbons to good points, then weighing them, and let them burn one hour, then weigh them again. The amount burned by both upper and lower carbons shows the rate of consumption which will accurately indicate the comparative merits of the carbons tested as to life. 578 ELECTRIC LIGHTING. To calculate the life from a burning test of one hour, both carbons should be first weighed, the upper carbon broken off to a 7-inch length, in order to make the test at the average point of burning, and with the lower carbon, burned to good points, weighed again, and after burning one hour in a lamp that has already been warmed up, taken out and weighed. The amount of two carbons 12 inches long consumed in a complete life-test is 63 per cent of the combined weight of both upper and lower carbons. There- fore 63 per cent of the weight of the two carbons, divided by the rate per hour obtained as above, will give the life approximately. Dust. — The dust from burning carbons can be collected in the globe, or better, in a paper bag suspended below the lamp. In an ordinary plain molded carbon this dust amounts to 4 per cent of the weight of the upper carbon. A variation below this amount will indicate good life, but inferior light. An excessive amount of dust would show a short life, but usually a good quantity and quality of light. Coating a carbon with copper eliminates this deposit of dust entirely. Enclosed Arc Carbons. Carbons for enclosed arcs can be very conveniently tested as to their rel- ative values in an open arc lamp as described above. As their diameters regulate the admission of air to the inclosing globe, thus greatly affecting their life, they should be carefully measured with micrometer calipers. A greater variation than .005 inch from the required diameter should not be permitted. The deposit on the inside of the inclosing globe is caused by impurities, principally in the core. The relative injurious amount of this deposit can be measured by carefully taking the globes off the lamps after burning, and measuring the amount of light absorbed by them with an ordinary photometer, using an incandescent lamp as a source of light, and cutting the light down by means of a hole in a screen so that it will pass through the part of the globe to be measured. Twice the light so measured through the globe, divided by the amount coming through the unobstructed hole, will give the per cent of the light transmitted through the globe from the arc. That carbon whose globe absorbs the least amount of light is, of course, the most desirable. The resistance of forced carbons, whether cored or solid, used in inclosed arc lamps, is very important. Carbons of high resistance are difficult to volatilize, and hence there is trouble in establishing the arc where small currents are used, and in case of any interruption in reestablishing it after- wards. This is especially true of carbons used in alternating arcs, and of cored carbons. The resistance of forced carbons is usually much higher than that of molded, ranging from two to four times as much. This will undoubtedly be corrected when the manufacturers become more familiar with the requirements. The lower the resistance the better the quality of the light and the operation of the lamp. Sizes of Carbons for Arc Lamps. Open Arcs. Continuous Current. Upper. Lower. 6.8 amperes 9.6 " 9.6 " 9 . 6 amperes * 9.6 M 12 in. X A in. 12 •• X I " 12 •• X f " 12 in. X A in. X I in. llf "Xi"Xl u 7 in. X i 7 s in. 7 " X * " 7 ** X 1 '* 6f in. X t 7 s in. X 1 in. 7i " X * " XI " Alternating Current. 15 amperes 9£ in. X f in. | 9£ in. X S in. Enclosed Arcs. Continuous Current. 5 amperes 3 amp eres 12 in. X h in. 12 " XI " 5i in. X £ in. 6 "XI" * These are elliptical in cross section, for higher candle-power and longer burning. ENCLOSED ARC LAMPS. 579 Carbons Recommended for Searchlight Projectors. (Columbia or Hardtmuth or Schmeltzer.) Size of Lamp. Positive. Cored. Negative. Cored or Solid. 9 inch 5i in. X h in. 3i in. X fs in. 13 " 6 1 X f " 4* 4I X * 44 18 44 8* • X H" 5 " X f " 24 44 12 4 X 1 " 7 " X t 44 30 ■• 12 • X 1* M 7 44 X I 44 36 " 12 • X U " 7 " XI 44 48 44 15 4 xitt" 12 "XI A 44 60 " 15 " X 2 " 12 " XU" Carbons Recommended for Automatic and Hand-Feed focusing: Lamps. Continuous Current. Amperes. Positive. Cored. Negative. Solid. 5 to 10 10 " 18 18 " 20 25 " 30 6 in. X & in. 6 " X f " 6 " X f " 6 " X I " 6 in. x t 7 5 in. 6 " X h " 6 " X I " 6 " X 1 " Alternating Current. 5 to 10 10 " 18 18 '• 20 25 " 30 6 in. X & in. 6 " X * " 6 " X f u 6 " X i " Same as for Positive. Candle-power of Arc Lamp*. The candle-power of an arc lamp is one of the most troublesome things to determine in all electrical engineering ; the variations being great the arc unsteady, and the implements for use in such determination being so liable to error. Again, what is the candle-power of an arc lamp, or rather, what is the meaning of the term ? When the lamp was first put forward, for some reason, now in great ob- scurity, the regular 9.6 ampere lamp was called 2000 candle-power, and it has always since been so called, although the word " nominal " has been tacked on to the candle-power to indicate that it is a rating, and not an actual measurement. The candle-power of the arc varies with the angle to the horizon on which the measurement is made ; in continuous current arcs the maximum can- dle-power is at a point about 45 degrees below the horizontal if the upper carbon is the positive, and of course above the horizontal if the negative carbon is above. In alternating current lamps there are two points of maximum light, one about 60 degrees above the horizontal, and the other about the same angle below the line, and the mean horizontal intensity also bears a greater ratio 580 ELECTRIC LIGHTING. to the mean spherical intensity than in the direct current arc. In the alternating current arc much of the light is above the horizontal plane, and it is necessary to arrange a reflector above the arc to throw that portion of the light downward. Mean Spherical Candle-power is the mean of the candle-power measured all over the surface ot a sphere of which the arc is the center, usually about one-third of the maximum candle-power. In practice the spherical candle-power is seldom fully determined, but a fair approximation may be had by the following formula : Let Then S = mean spherical candle-power, H— horizontal candle-power, M = candle-power at the maximum. H M s ~-2 + T' In a test of arc lamps in November, 1889, for the New York City Bureau of Gas, Captain John Millis found the following results in his trial of the Thomson-Houston lamps. The same lamp was used, but connected to the different street circuits, all measurements were made at 40 degrees below the horizontal, and ^g-mch copper-plated carbons were used. Ten readings were taken on each of four sides of the lamp when con- nected to each circuit, with the following results : Candle-power. Watts Circuit No. 1. 2072.7 482.88 " " 2. 1981.0 485.10 " " 3. 2048.5 493.22 44 41 4# 2000.2 494.40 44 44 5. 2067.0 495.36 Means 2033.9 490.19 Mean current, amperes Mean volts .... . . e 10.36 . . . 47.32 The results of tests of candle-power of arc lamps at the Antwerp Exposi- tion, shown in the table below, would tend to verify the above trials. Maxi- mum C.P. Upper Lower Am- peres. Volts. Horizon- tal C.P. Hemi- sphere Hemi- sphere. Mean C.P. Watts. Mean C P. Mean C.P. 4 37.2 390 74 17 119 136 157 6 46.2 1090 168 63 298 361 259 6.8 46 1240 240 65 320 385 313 8 46 1550 334 70 385 454 350 10 45.5 2070 421 102 640 750 491 Arc liig-ht Efficiency. — The light efficiency of an arc lamp is the ratio of its mean spherical candle-power to the watts consumed between the lamp terminals. Some energy is used up in the lamp-controlling mechan- ism, in the carbons themselves, and the remainder is used on the arc. Arc lamp efficiency is sometimes described as the ratio of the watts used in the arc to the watts used between the lamp terminals. This is true of the lamp as a machine; but the first statement is the correct one, as it is fight that is turned out, and not watts consumed in the arc that is the object of the lamp, and the two depend so much on quality and adjustment of carbons, even with the same consumption of current, as to make the latter method erroneous. ENCLOSED ARC LAMPS. 581 Seat and Temperature Developed by tbe Electric Arc. The temperature of the crater, or light-emitting surface of the arc, is the same as the point of volatilization of carbon, and therefore constant under constant atmospheric pressure. This temperature is variously stated by different investigators: Dewar gives it as 6000° C; Rosetti, the positive as 3200° C, and the negative 2500° C. The carbon in the crater is in a plastic condition during burning; and with the same adjustment of carbons, as to length of arc, the light per unit of power increases with the current. Hissing, naming, and rotating of the arc are some of the defects. Hissing is due to a short arc, and was a constant accompaniment of the low poten- tial, high current arc so prevalent during the earlier days of arc lighting. Flaming and rotating in open arc lamps are due to long arcs and to impure carbons, or carbons not properly baked. With good carbons the length of arc, or distance between carbon tips for open arcs direct current, continuous current lamps, should be, for 6.8 ampere lamp, & inch; and for 9.6 or 10 ampere lamps, is to 3 j inch. Balancing' Resistance for Arc Stamps on Constant Potential Circuit. As the ordinary arc lamp takes but 45 to 50 volts, when used on constant potential circuits of more than 50 volts, it is necessary to introduce a cer- tain resistance in series, in order, first, to take up part of the voltage, and second, to act in a steadying capacity to the arc; in fact, until the dead resistance was introduced in series with the arc lamp on constant potential circuits, such lamps were entirely unsuccessful. Prof. Elihu Thomson says, "a certain line voltage as a minimum is abso- lutely necessary in working arc lamps on constant potential lines, whether they be open arcs or enclosed arcs. Thus two 45- volt arcs in series, with uncored carbons like the brand known as 'National,' cannot be safely worked below 110 volts on the line without resistance in series with them. More than 100 volts should, of course, be maintained for safety of the service. M The tests show, also, that with a cored upper carbon, the limit is lowered several volts on the average, and it is known that the voltage of the arcs may be safely reduced somewhat when cored positives are used. "It is also shown that a 75 to 80- volt enclosed arc, run upon a constant potential line, is stable at a considerably less line voltage than the open arc. It would appear, also, that with either open or enclosed arcs at ordinary current strengths of from 5 to 10 amperes, the steadying resistance in the branch is required to cause a drop of about 15 to 20 volts, or waste energy at the rate in watts of 15 to 20, multiplied by the amperes of current used in the lamp." Let E = E.M.F. or difference of potential between the circuit leads. e = E.M.F. required at arc lamp terminals. i = current required by the arc lamp. R = dead resistance to be put in series. r = resistance of the arc lamp burning. r' = total resistance of dead resistance 4- lamp. Then r = -. (1) % r, = -r (2) x R = r, - r. (3) As the E.M.F. of most of the circuits on which lamps of this type are used is more than 100 volts, it is customary, and in fact economically necessary, to place two arc lamps in series, and the formula (3) then becomes, R = r, - 2r. 582 ELECTRIC LIGHTING. Street rig-htiiigr t>J Arc lamps. For good illumination, distance apart of arc lamps should not exceed six times height of arc from ground. For railroad yards, 10 ampere arc lamps 30 feet from the ground and about 200 feet apart are found to give good results. The following table shows some arrangements of arc lamps in foreign cities: Arc Lamps in Foreign Cities. Amperes per Arc. Distance Apart in Ft. Height of Arc in Ft. City of London Streets . Glasgow Streets 10 10 10 • 15 10 15 10 10 10 10 15 115 160 300 137 80 to 100 90 180 60 to 80 75 90 33 41 17.6 18.0 Hastings Streets 18.0 Berlin Streets . .... 26.7 Milan Streets . 25.0 Charing Cross Railroad Station Cannon Street Railroad Station St. Pancras Railroad Station . Central Station, Glasgow St. Enoch's Station, Glasgow . Edinburgh Exhibition, 1886 Edinburgh Exhibition, 1886 18.0 35.0 14.0 19.5 12.6 18.0 ] Dr. ►ff B] by Grlol»c ELL. ,S. With respect to porcelain and glass, the following table gives the general results obtained by several experimenters on the absorption of various kinds of globes, especially with reference to arc lights. Per cent. Clear glass 10 Alabaster glass 15 Opalescent glass 20 to 40 Ground glass 25 to 30 Opal glass 25 to 60 Milky glass 30 to 60 Too much importance should not be attached to this large absorption, since it has already been shown that in most cases, so far as useful effect is concerned, diffusion and the resulting lessening of the intrinsic brilliancy is cheaply bought, even at the cost of pretty heavy loss in total luminous radiation. The classes of shades commonly used for incandescent lamps and gas lights have been investigated with considerable care by Mr. W. L. Smith. The experiments covered more than twenty varieties of shades and re- flectors, and both the absorption and their distribution of light were inves- tigated. One group of results obtained from 6-inch spherical globes, in- tended to diffuse the light somewhat without changing its distribution, was as follows, giving figures comparable with those just quoted: Per cent. Ground glass 24 . 4 Prismatic glass 20.7 Opal glass 32.2 Opalescent glass t 23.0 ENCLOSED ARC LAMPS. 583 The prismatic globe in question was of clear glass, but with prismatic longitudinal grooves, while the opal and opalescent globes were of medium density only. Etched glass has considerably more absorption than any of the above, the etching being optivally equivalent to coarse and dense grinding. Their diffusion is less homogeneous than that given by ordinary grinding, so that they may fairly be said to be undesirable where efficiency has to be seri- ously considered. Trimming; Arc LampN. One trimmer can handle the following number of lamps per day: Walking. Riding. Regular open double carbon street arcs .... 80 100 to 120 Magnetite lamps 80 100 " 120 Flaming arcs 80 100 " 120 Enclosed arcs 50 100 The number of commercial lamps which one man can trim depends so much upon local conditions that it is not possible to give any general figure. ILLUMINATING ENGINEERING. Revised by Dr. C. H. Sharp. The problem of the illuminating engineer may be stated in general terms as follows: to obtain the illuminating effect desired in any case with the maximum economy, having due regard to the protection of the eyes from disagreeable or harmful effects and to architectural and aesthetic consider- ations. Illumination may be direct, coming straight from the lamps which then are visible, or indirect, as when the lamps are hidden from view by a cornice and the illumination is due to the light reflected from a cove above. Measurements of candle-power values are horizontal, vertical and normal illuminations, according to the position of the plane of reference, horizontal, vertical or normal to the light rays. Curves of illumination have as their abscissas distances from the source of light measured along a horizontal line and as their ordinates intensities of illumination. If the vertical distribution curve of the source of light is known the corresponding illumination curves can be computed according to the following equations, in which E is the illumination, h the height of the lamp above the plane of reference, I the distance from the point in question to the point immediately beneath the lamp, and Iq the intensity of the lamp at an angle with the vertical : A* + p Iq COS Iq h Iq cos 8 Eh = Ev = Iq sin IqI I h 2 + P (h* + P)t ~ h sin* In considering the availability of any source of light due regard must be given to the proper selection of shades, reflectors, etc., which may be used in connection with it. These appurtenances serve the following purposes: to direct the light most advantageously; to diffuse the light, decreasing the apparent specific intensity of the source and thereby safeguarding the eyes; pure decoration. The efficiency of an illumination installation often depends to a very great degree on the selection of proper auxiliaries. The illumination on a surface is equal to the luminous flux in lumens per unit area of the surface, e.g. the foot-candles are equal to the lumens per square foot. The average illumination on a plane of reference is equal to the lumens through the plane divided by its area. Hence we have the following definitions: The net efficiency of an illumination installation is equal to the ratio of lumens through the horizontal plane of reference to the total lumens generated by the lamps. The gross efficiency of an installation is the ratio of the watts supplied to the lamps to the lumens on the plane of reference. The net efficiency depends only on the method of installing the lamps, on the reflectors, etc., used, and on the coefficient of reflection of the walls, ceiling, floor and contents of the room. If we represent this average co- * The values of sin 3 and cos 3 are given in Table I. 584 ILLUMINATING ENGINEERING. 585 efficient by fc, multiple reflections theoretically increase the illumination 1 * in the ratio _ . • In practice this is found to be modified by many conditions. A general knowledge of the value of the net efficiency to be expected in any case enables the illuminating engineer to form a very ready estimate of the number of lamps required. Table I. 0° to 29 o 30° to 5S °. 60° to 89°. e. Cos 3 9. Sin 3 9. 9. 30 Cos 3 9. Sin 3 9. 9. 60 Cos 3 9. Sin 3 0. 1.0000 0000 . 6495 1250 0.1250 6495 l . 9994 0000 31 .6299 1366 61 .1139 6690 2 .9982 0000 32 .6098 1488 62 . 1035 6882 3 .9958 0001 33 .5900 1615 63 .0936 7073 4 .9928 0003 34 .5697 1749 64 .0843 7261 5 .9886 0007 35 .5498 1887 65 .0755 7444 6 .9836 0011 36 .5295 2031 66 .0673 7623 i .9777 0018 37 .5093 2180 67 .0596 7800 8 .9712 0027 38 .4893 2334 68 .0526 7971 9 .9636 0038 39 .4693 2492 69 0460 8137 10 .9551 0052 40 .4495 2656 70 .0400 8298 11 .9458 0069 41 .4299 2824 71 .0345 8452 12 .9357 0090 42 .4103 2996 72 .0295 8604 13 .9251 0114 43 .3913 3172 73 .0250 8745 14 .9135 0142 44 .3722 3353 74 .0209 8883 15 .9011 0173 45 .3535 3535 75 .0173 9011 16 .8883 0209 46 .3353 3722 76 .0142 9135 17 .8745 0250 47 .3172 3913 77 .0114 9251 18 .8604 0295 48 .2996 4103 78 .0090 9357 19 .8452 0345 49 .2824 4299 : 79 .0069 9458 20 .8298 0400 50 .2656 4495 80 .0052 9551 21 .8137 0460 51 .2492 4693 81 .0038 9636 22 .7971 0526 52 .2334 4893 82 .0027 9712 23 .7800 0596 53 .2180 5093 83 .0018 9777 24 .7623 0673 54 .2031 5295 84 .0011 9836 25 .7444 0755 55 .1887 5498 85 .0007 9886 26 .7261 0843 56 .1749 5697 86 .0003 9928 27 .7073 0936 57 .1615 5900 87 .0001 9958 28 .6882 1035 58 .1488 6098 88 .0000 9982 29 .6690 1139 59 .1366 6299 89 .0000 9994 * Values of k are given in Table II. 586 ILLUMINATING ENGINEERING. Table II. Showing; the Intensity of the Illumination in Foot Candles Produced at Various Points in Horizontal Planes by a liig-ht Source of I.C. P. : the Angle Made by the Xiig-ht Ray and a Fine Perpendicular to the Horizontal Plane. From a Pamphlet by the National Electric Lamp Association. Horizontal Distance in feet from Point Directly under Lamp to Point where Intensity of Illumination is desired. 72 o ♦a 09 a 2 4 6 8 10 3 "3 Foot 5 Foot "3) Foot 15b Foot fi Foot 12) Foot — d Candles < Candles 1 Candles 1 Candles < Candles -51 Candles o9 o ' o / o / o / o / Ch 2 .250 45 .0883 63 25 . 02240 71 35 . 00790 76 . 00355 78 40 .001907 4 .0625 26 35 .0447 45 . 02206 56 20 .01064 63 25 . 00560 68 10 . 003220 6 .02775 IS 25 . 02365 33 40 .01600 45 . 00980 53 5 . 00602 59 . 003802 8 .01563 14 .01428 26 35 01119 36 50 . 008015 45 . 00552 51 20 .003815 10 .010 11 20 .009417 21 50 . 007997 31 . 00630 38 40 .004757 45 . 003530 Angle Foot Candles Angle Foot Candles Angle Foot Candles Angle Foot Candles Angle Foot Candles 3 S G > z 1 .9 a 6 o9 O +a A U « o / 80 35 71 35 63 25 56 20 50 10 45 40 40 36 50 33 40 31 28 35 26 35 24 45 23 10 21 50 .001109 .001975 .002485 . 002665 . 002623 . 002450 . 002220 .002001 .001781 .001575 .001398 .001240 .001108 .000991 . 000889 o / 81 50 74 5 66 50 60 15 54 30 49 25 45 41 10 37 55 35 32 36 30 15 28 20 26 35 25 . 000722 .001436 .001689 .001913 .001960 .001900 .001801 .001665 .001517 .001375 .001240 .001118 .001008 .000911 . 000826 o / 82 55 76 69 25 63 25 58 53 5 48 50 45 41 40 38 40 36 5 33 40 31 35 29 45 28 5 . 000473 . 0008875 .001207 .001402 .001490 .001506 .001455 .001380 .001288 .001189 .001088 .001000 .000915 .000834 . 000765 o t 83 40 77 30 71 35 66 60 55 56 20 52 10 48 25 45 42 39 20 35 50 34 45 32 45 31 .000341 000631 . 000876 .001050 .001149 .001181 .001178 .001142 .001090 .001025 000955 . 000890 .000821 .000758 . 000700 o / 84 15 78 40 73 20 68 10 63 25 59 55 51 20 48 45 42 20 39 50 37 35 35 35 33 40 . 000242 .000476 . 000654 . 000805 . 000897 . 000950 . 000965 . 000954 . 000927 . 000883 .000835 .000785 . 000736 .000686 .000640 GRAPHIC ILLUMINATING CHART. 587 Graphic Illuminating* Chart. A. E. Parks, Trans. I. E. S., Oct., 1907. The equation upon which the chart is based is the well-known one, T C 3 W 2 c a ' Where / = Illumination in foot-candles normal to the plane to be illumi- nated. C = Candle-power reading from a photometric curve, a = Angle made by reading C with normal to plane illuminated. H = Minimum distance source of illumination to this plane. Solving this equation by logarithms consists, as is well known, of finding log of C, log of cos 3 a, adding same together and subtracting log of H 2 t the remainder giving the logarithm of the result desired, this being exactly the graphic method followed in working the chart. In Fig. 1, if the distance A-B be laid off representing log C, and 'A—C a distance representing log cos 3 a, completing the rectangle will give point D. It is desired to add the length of A—C to the length A—B, however, and fortunately we may do this graphically if from D we draw a line D-E at an angle of 45 degrees till it outs the line A-B produced. A-E now represents log C + log. cos 3 a We now wish to subtract from A-E a distance equal to log H 2 . Laying off vertically from E such a distance E-F, we may, by means of a 45-deg?ee line through F, subtract from A-E this distance E-F, giving us the point G, A-G then representing the solution of the problem or A-G = \og C + log cos 3 a — log H' 2 . If now the diagonal G-F be properly labeled, ill values of E-F falling on this line will have the same foot-candle readings, And for every other foot-candle reading there will be a diagonal parallel to F-G. While a chart constructed exactly as per the foregoing description may *)e conveniently used, the form here presented is somewhat different in arrangement, for by a proper manipulation of axes, one set of diagonals may be made to do duty for both D-E and F-G functions, and considerable saving in space and complexity results. A few samples will elucidate the working of the chart. Say that from a photometric curve we get 50 candle-power in a vertical lirection, and 100 candle-power at an angle of 45 degrees. It is desired to find the illumination on a plane at six feet below the source of light. Taking first the 50 candle-power reading. As a in this case is 0, we find JO on the top candle-power scale, and follow the diagonal lines to the right hand margin, giving the point 5. We now follow horizontally toward the left to the vertical through the point 6 found on the lower inclined margin. Following a diagonal again to the right hand margin we find for the value required 1.40 foot-candles. Again from 100 candle-power on the top scale we follow vertically to the horizontal line through 45 degrees found on left hand margin, from this intersection follow diagonal to right hand margin to 3.5. Proceed toward the left horizontally to vertical through 6 as before, and again along a diagonal fromthis intersection to the right hand margin, giv- ing 1 foot-candle as the desired result. As an example of the reversibility of the chart, the following problem will be solved. Let it be required to construct a photometric curve that will Eroduce a uniform illumination of 1.5 foot-candles upon a plane seven feet elow the light source. Find the intersection of the diagonal from 1.5 on right hand margin with vertical through 7 on lower scale. Follow horizontally to the right to right hand margin, continue from this point along a diagonal toward the top, and where this diagonal cuts the several degree lines, will be found the candle-power readings required at these angles. As 205 candle-power at 45 degrees, 165 candle-power at 40 degrees, 132 candle-power at 35 degrees, 110 candle-power at 30 degrees, 96 candle-power at 25 degrees, etc. etc., to 72 candle-power at zero degrees. 588 ILLUMINATING ENGINEERING. Caudle Power s="C" o o o <: ooooooooooo 12345678 9 1011121314 151617 181920212223242526 272829303132333435 Horizontal Distance Fig. 1. GBAPHIC ILLUMINATING CHART. 589 Table III. Required Illumination for Various Classes of Service. From a pamphlet by the National Electric Lamp Association. Class of Service. Light Intensity in General illumination of: Foot-Candles. Auditoriums 1 to 3 Theaters 1 to 3 Churches 3 to 4 Reading 1 to 3 General illumination of residences 1 to 2 Desk illumination 2 to 5 Postal service 2 to 5 Bookkeeping 3 to 5 Stores, general illumination 2 to 5 Stores, clothing 4 to 7 Drafting 5 to 10 Engraving 5 to 10 Table IV. Snowing- Saving; by the Use of Hig-h Efficiency Lamps. From a pamphlet by the National Electric Lamp Association. Carbon. Carbon. Gem. Tanta- lum. 1 Candle-power 20. 20. 20. 20. 2 Watts per candle, nominal . . . 3.5 3.0 2.5 2.1 3 Watts per candle, actual . . . 3.48 3.04 2.5 2.1 4 Total watts 69.6 1040. 60.8 520. 50.0 560.0 42 5 Hours total life 600. 6 Cost of lamp SO. 16 $0.16 $0.20 $0.54 7 Cost of renewals per year of 1000 hours 0.154 0.308 0.36 0.90 8 Cost of power per year of 1000 hours at 10 c. per k.w. hour . . . 6.96 6.08 5.00 4.20 9 Cost of power and lamp renewals per year of 1000 hours .... 7.11 6.39 5.36 5.10 10 Saving over 3.5 W. P. C. lamp . 0.72 1.75 2.01 11 Saving over 3.0 W. P. C. lamp . 1.03 1.29 Line 5 gives our best knowledge of the life of our lamps with good volt- age regulation. A slight difference in standards, a variable regulation or a poor regulation will cause lamps to average better or poorer than these figures. Line 6 shows the cost of lamp in 10,000 quantity. 590 ILLUMINATING ENGINEERING. f*-^, 80 # C XN \> 70* ^ O ^ 60? Fig. 2. DATA ON ILLUMINATING VALUES. 591 ?\g. 3. 592 ILLUMINATING ENGINEERING. Experimental Data on Illuminating* Values. From paper by Sharp & Millar before Edison Association. This auditorium is equipped with a cove-lighting installation and with an arrangement of ceiling lamps and side brackets. The Edison Company undertook the work of arranging such temporary installations as were required for the purpose of the test. These installations were selected at the suggestion of the advisory committee in such a way as, first, to bring out the relative illumination efficiencies obtainable with similar illuminants, variously arranged and variously equipped with reflectors, etc.; second, to give a basis for reliable comparisons -of the illuminating efficiencies of illuminants of different types. The fact should be emphasized, however, that the results here given apply in all strictness only to the room in question, and that in using these data in connection with other installations, proper consideration should be given to this fact. The sixteen candle-power carbon incandescent lamps which were used in the installations requiring such lamps, were new lamps taken from a package which had been purchased recently subject to the inspections of the Elec- trical Testing Laboratories, and which could therefore be considered as well-rated lamps. These lamps were burned about fifty hours before the first test was undertaken. The frosted lamps were selected in a similar manner. The actual candle-power and watts of these lamps were deter- mined by selecting a considerable number of representative ones and photo- metering them in the laboratory, at the actual voltages used in the tests. The deterioration of these test lamps in successive tests was also deter- mined in this way. It is desirable, also, to know what ratio of the total light which is emitted by the lamps in a room may be expected to fall on a plane of reference, i.e., the horizontal plane on which measurements of the intensity of the illumination are commonly made. This ratio of the light generated to the light utilized on the plane of reference gives a value for the net efficiency of the installation. However, in order to arrive at an expression for this efficiency, it is necessary to employ some unit in which the total light from the lamps and the total light falling on the plane of reference can be expressed. For this purpose the notion of the flux of light is used, and the unit in which luminous flux is measured is introduced. This unit is the "lumen," which is defined as the flux of light emitted by a source of one candle-power in a unit solid angle. The total luminous flux from a source of light is equal to 47r, or 12.57 times its mean spherical candle-power. We can measure in lumens not only the output of the lamps, but also the flux of light through the plane of reference, and the ratio of the lumens through the plane of reference to the lumens yielded by the lamps gives the net efficiency of the installation. In a similar way the efficiency of the lamps may be measured by their lumens per watt; and the gross efficiency of the illumination instal- lation can be measured by the lumens on the plane of reference per watt expended in the lamp. The lumens on the plane of reference are deter- mined by multiplying the intensity of illumination on this plane, as ex- pressed in candle-feet, by the area of the plane in square feet, i.e., the flux through a plane is equal to the intensity of the illumination on the plane multiplied by the area of the plane, or the illumination on the plane is equal to the density flux of the light falling on that plane. In measuring the illumination, forty-five stations were selected, equally spaced over the floor of the auditorium. The values of illumination were then plotted on a map of the floor area, and then all points having the same illumination were connected by lines. This gives a set of lines which we have called equilucial lines, by analogy with equipotential lines of an elec- trostatic or a magnetic field. If the lines are plotted representing in all cases the same percentage variation of illumination, the closeness of the lines to each other represents the illumination gradient, or the rate at which the illumination is changing from place to place on the plane of reference, and consequently the lack of uniformity in the illumination. Diagrams of this character have been prepared for the various tests. A number of such diagrams are given on pages 590 and 591 . These, in DATA ON ILLUMINATING VALUES. 593 each case, show the arrangement of the lamps and a condensed description of the type of installation is given. These diagrams show lines of uniform illumination for various types of installation. The equilucial lines show differences in intensity of ten per cent. Diagram 1 shows the effect of the cove lighting alone; 2, ceiling lamps and brackets frosted; 3, concentrating prismatic reflections, high level; 4, mirror reflectors, high level; 5, distribut- ing reflectors, low plane; 6, gem lamps; 7, tungsten lamps; 8, arc lamps, with diffuser shades. In a general way the tests made were intended to show, first, the compar- ison between the various permanent installations in the auditorium; second, the increase in illumination efficiency resulting from equipping the ceiling lamps with various reflectors, and the effect of using frosted instead of clear bulb lamps; third, the effect of lowering the same equipment to a point nearer the floor. Furthermore, gem lamps, tungsten lamps, Nernst lamps and arc lamps were installed with the idea of obtaining comparative data on their illuminating values as used in a room of the dimensions and char- acteristics of this auditorium. These varying results are summarized in the accompanying table. By a comparison of the lumens which become effective on the plane of reference with the lumens which are generated by the lamps, we get a value for the net efficiency of the installation. The value of this efficiency indi- cates the degree of skill with which the installation has been planned and carried out. It is totally unaffected by the efficiency of the lamps employed and refers only to the illumination installation as such, irrespective of the illuminants used. It is, however, largely affected by the character of the room which is illuminated, as is also the gross efficiency of the installation. Coefficients of Reflections. Bell. Many experiments have been made to find the absolute loss of intensity due to reflection. This absolute value of what is called the coefficient of reflection, that is to say, the ratio of the intensity of the reflected to that of the incident light, varies very widely according to the condition of the reflecting surface. It also — in case the surfaces are not without selective reflection in respect to color — varies notably with the color of the inci- dent light. The following table gives a collection of approximate results derived from various sources. The figures show clearly enough the uncertain char- acter of the data. Material. Coefficient of Reflection .92 .70 to .85 .70 " .75 .60 " .70 .60 .60 " .80 .50 " .55 .40 " .50 Highly polished silver . . Mirrors silvered on surface Highly polished brass . . Highly polished copper . Highly polished steel . . Speculum metal .... Polished gold Burnished copper .... Smooth papers and paint give a very considerable amount of surface reflection of white light, in spite of the pigments with which they may be colored. The diffusion from them is very regular, except for this surface sheen, and may be exceedingly strong. When light from the radiant point falls on such a surface it produces a very wide scattering of the rays, and an object indirectly illuminated therefore receives in the aggregate a very large amount of light. A great many experiments have been tried to determine the amount of this diffuse reflection which becomes available for the illumination of a single object. The general method has been to compare the light received directly from the illuminant with that received from the same illuminant by a reflection from a diffusing surface. 594 ILLUMINATING ENGINEERING. Table V. Comparative Values of Illumination and Installation. Equipment. s a o s u -1 43 c <0 5 » o b£)o 'd +a o V © •r. u (3 ► a % 29.7 52°. 1.48 2"o2 J* h3 ft Gross Lumens Effective per Watt. Net Lumens Effective per Lumen Generated. 16.28 6210 2.27 l.n 1.72 3.38 0.8 % 23.7 244 14.8 12.2 3.26 3.95 11600 7.10 3.92 6.18 25.7 5.52 3.16 1.54 48.8 365 14.05 11.59 3.38 4.05 17300 8.41 5.60 7.58 21.0 6.72 3.06 1.265 41.3 244 13.86 11.41 3.48 4.36 11930 6.53 3.70 5.65 25.1 5.07 2.89 1.365 47.3 365 3.12 3.84 17760 8.00 3.80 5.57 T98 7.23 3.30 16.85 28.5 6.42 3.03 1.18 39.0 100 15.5 12.78 4910 3.28 1.91 58.2 100 15.33 12.64 3.22 3.89 4930 5.47 2.08 4.07 36.8 3.23 2.35 72.7 100 15.17 12.5 3.26 3.95 4940 5.16 2.08 4.02 38.3 3.19 2.31 72.4 100 15.43 12.71 13.28 3.17 3.06 3.85 3.72 4900 7.69 4.15 1.50 "2^38 4.92 3.50 62.9 25.3 3.26 2.86 87.7 100 16.11 4940 3.38 2.02 59.8 100 14.72 12.13 3.30 4.0 4865 5.83 1.85 3.94 50.9 3.13 2.30 73.5 100 15.4 12.69 3.22 3.9 4951 5.92 1.91 4.28 46.9 3.22 2.46 76.4 100 15.33 12.63 3.19 3.87 4892 8.20 1.12 4.45 72.4 3.25 2.59 79.6 98 14.9 12.28 3.29 3.99 4798 7.07 .83 66.4 4.70 3.76 2.79 88.3 98 15.25 12.58 3.19 2.67 2.43 3.87 3.14 2.93 4775 6.82 4.61 .64 2.08 3.33 76.9 37.8 3.92 3.26 2.33 71.5 «( 39.1 105.8 32.9 87.8 4328 4.21 2.0 52.2 24 79.3 63.4 1.19 1.49 1694 4.75 1.88 3.29 30.2 8.46 5.52 65.2 24 37.0 3.15 2802 3.08 1.14 2.09 46.4 3.98 2.12 53.2 24 36.9 3.16 2798 3.55 .95 2.24 58.0 3.98 2.28 57.3 9 22.9 2.7 5530 7.88 2.07 4.31 S7.5 4.54 2.22 48.9 9 . 22.9 2.7 1 5530 6.46 1.73 4.01 58.6 4.54 2.06 45.3 596 ILLUMINATING ENGINEERING. The following table gives an aggregation of the results obtained by sev- eral experimenters, mostly from colored papers: Material. White blotting paper White cartridge paper .... Ordinary foolscap Chrome yellow paper Orange paper Plane deal (clean) Yellow wall paper Yellow painted wall (clean) . . Light pink paper Yellow cardboard Light blue cardboard Brown cardboard Plane deal (dirty) Yellow painted wall (dirty) . . Emerald green paper Dark brown paper Vermilion paper Blue green paper Cobalt blue Black Deep chocolate paper . . . . French ultra-marine blue paper Black cloth Black velvet Coefficient of Diffuse reflection. .40 .82 .80 .70 .62 .50 to . .40 .40 .36 .30 .25 .20 .20 .20 .18 .13 .12 .12 .12 .05 .04 035 012 004 50 Interior Illumination. Bell. To illuminate a room 20 ft. square and 10 ft. high on the basis of a mini- mum of 1 candle-foot, will require from 80 to 144 effective candle-power, according to the arrangement of the lights, if the finish is light, and half as much again, at least, if the finish is dark. The floor space being 400 sq. ft. it appears that the illumination is on the basis of about 3 to 5 sq. ft. per effective candle-power. The former figure will give good illumination under all ordinary conditions; the latter demands a combination of light finish and very skillfully arranged lights . For very brilliant effects, no more than 2 sq. ft. per candle should be allowed, while if economy is an object, 1 c.p. to 4 sq. ft. will furnish a very good groundwork of illumination, to be strengthened locally by a drop-light or reading lamp. The intensity thus deduced may be compared to advantage with the results obtained by various investigators, reducing them all to such terms as will apply to the assumed room which is under discussion. Just deduced Uppenborn Piazzoli . . Fontaine . . 1 c.p. per 3 sq. ft. 1 c.p. per 3.6 sq. ft. 1 c.p. per 3.5 sq. ft. 1 c.p. per 7.0 sq. ft. (approximation). In very high rooms the illumination just indicated must be materially increased, owing to the usual necessity for placing the lamps rather higher than in the case just given, and on account of the lessened aid received from diffuse reflection. The amount of this increase is rather uncertain, but in very high rooms it would be wise to 'allow certainly 1 c.p. for every 2 sq. ft., and sometimes, as in ball-rooms and other special cases requiring the most brilliant lighting, as much as 1 c.p. per square foot. Perhaps the most important rule for domestic lighting is never to use, indoors, an incandescent or other brilliant light, unshaded. Ground or frosted bulbs are particularly good when incandescents are used, and opal INTERIOR ILLUMINATION. 597 shades, or holophane globes, which also reduce the intrinsic brilliancy, are available with almost any kind of radiant. Ornamental shades of tinted glass or of fabrics are exceedingly useful now and then, when arranged to harmonize with their surroundings. The table below is intended as a hint about the requirements for domes- tic lighting, and while it is laid out for a fairly large house, containing twenty rooms and three baths, its details will furnish suggestions appli- cable to many cases. An 8-c.p. lamp of the reflector variety should be placed in the ceiling of every large closet, and controlled by a switch from the room or by an automatic switch, turning it on when the door is fully opened. Room. 8 c.p. Hall ....... Library Reception room . . Music room . . . Dining room . . . Billiard room . . . Porch Bedrooms (6) . . . Dressing rooms (2) Servants' rooms (3) Bathrooms (3) . . Kitchen Pantry Halls Cellar Closets (4) . . . . Total .... 64 16 c.p. 14 4 3 3 30 32 c.p. | Sq. Ft. per c.p 4.7 3.1 7.0 3.0 2.7 2.3 7.0 4.7 9.4 5.0 Remarks. 8-c.p. reflector lamps Eight reflector lamps 32 c.p. with reflectors Reflector lamps Watt* at lianip Terminals Per Square Foot floor Space for Hig-lt Class Arc JLig-liting-. (By W. D'A. Ryan.) Building. Range. Average Condi- tions. Machine shops; high roofs, electrically driven machinery, no belts Machine shops; low roofs, belts, other obstruc- tions .5 to 1 .75 to 1.25 .5 to 1 .75 to 1.25 1 to 1.5 .9 to 1.3 1.1 to 1.5 1.25 to 1.75 1.5 to 2 .75 1 Hardware and shoe stores .75 Department stores; light material, bric-a-brac, etc 1 Department stores; colored material .... Mill lighting; plain white goods Mill lighting; colored goods, high looms . . . General office; no incandescent s Drafting rooms 1.25 1.1 1.3 1.6 1.75 ( Note : Energy based on watts at lamp terminals. 598 ILLUMINATING ENGINEERING. S e Z v ft .a a iS o eo O •a a £ o 03 O .s a .«3 o ^NWNO "5©§NM OOgJeJOgJ v IO05 tH iO CO t^- 00 © CO CO 00 OiCOlNiOiO ^ Tj-, i>oa2t>i> co^- . -csJw CO o3*G 'V CNCNq Q w 1—1 §£ I>© © 0(NiOOr3r^V v - a •^ 00 -o •coco •■<* 1 . ■P ^ 00 2 m O- co^S^o ^* 13 , Sl^gguj GENERAL ILLUMINATION. 599 General Illumination. The subject of illumination has been divided by Mr. E. L. Elliott, to whom we are indebted for many suggestions, into, the following sub-divisions: Intensity or brilliancy, distribution, diffusion, and quality. Intensity of Brilliancy. — The average brilliancy of illumination required will depend on the use to which the light is put. "A dim light that would be very satisfactory for a church would be wholly inadequate for a library, and equally unsuitable for a ballroom." The illumination given by one candle at a distance of one foot is called the "candle-foot" or "foot-candle, " and is taken as a unit of intensity. In general, intensity of illumination should nowhere be less than one candle- foot, and the demand for light at the present time quite frequently raises the brilliancy to double this amount. As the intensity of light varies inversely with the square of the distance, a 16 candle-power lamp gives a candle-foot of light at a distance of four feet. A candle-foot of light is a good intensity for reading purposes. Assuming the 16 candle-power lamp as the standard, it is generally found that two 16 candle-powder lamps per 100 square feet of floor space give good illumination, three very bright, and four brilliant. These general figures will be modified by the height of ceiling, color of walls and ceiling, and other local conditions. The lighting effect is reduced, of course, by an increased height of ceiling. A room with dark walls requires nearly three times as many lights for the same illumination as a room with walls painted white. With the amount of intense light available in arc and incandescent lighting, there is danger of exceeding " the limits of effective illumination and producing a glaring intensity," which should be avoided as carefully as too little intensity of illumination. Distribution of Ug-lit. — Distribution considers the arrangement of the various sources of light, and the determination of their candle-power. The object should be to " secure a uniform brilliancy on a certain plane, or within a given space. A room uniformly lighted, even though compara- tively dim, gives an effect of much better illumination than where there is great brilliancy at some points and comparative darkness at others. The darker parts, even though actually light enough, appear dark by contrast, while the lighter parts are dazzling. For this reason naked lights of any kind are to be avoided, since they must appear as dazzling points, in contrast with the general illumination." The arrangement of the lamps is dependent very largely upon existing conditions. In factories and shops, lamps should be placed over each ma- chine or bench so as to give the necessary light for each workman. In the lighting of halls, public buildings, and large rooms, excellent effects are obtained by dividing the ceiling into squares and placing a lamp in the center of each square. The size of square depends on the height of ceiling and the intensity of illumination desired. Another excellent method con- sists in placing the lamps in a border along the wall near the ceiling. For the illumination of show windows and display effects, care must be taken to illuminate by reflected light. The lamps should be so placed as to throw their rays upon the display without casting any direct rays on the observer. The relative value of high candle-power lamps in case of an equivalent number of 16 candle-power lamps is worthy of notice. Large lamps can be efficiently used for lighting large areas, but in general, a given area will be much less effectively lighted by high candle-power lamps than by an equiva- lent number of 16 candle-power Tamps. For instance, sixteen 64 candle- power lamps distributed over a large area will not give as good general illumination as sixty-four 16 candle-power lamps distributed over the same area. High candle-power lamps are chiefly useful when a brilliant light is needed at one point, or where space is limited and an increase in illuminat- ing effect is desired. Diffusion of Lig^ht. — "Diffusion refers to the number of rays that cross each point. The amount of diffusion is shown by the character of the shadow. Daylight on a cloudy day may be considered perfectly diffused ; it produces no shadows whatever. The light from the electric arc is least diffused, since it emanates from a very small surface ; the shadows cast by it have almost perfectly sharp outlines. It is largely due to its high state of diffusion that daylight, though vastly more intense than any artifi- cial illumination, is the easiest of all lights on the eyes. It is a common I 500 ILLUMINATING ENGINEERING. and serious mistake, in case of weak or overstrained eyes, to reduce the intensity of the light, instead of increasing the diffusion." Quality of* Eig-ht. — " Aside from difference in intensity, light pro- duces many different effects upon the optic nerves and their centers in the brain. These different impressions we ascribe to difference in the quality of the light. Thus, 'hard light,' 'cold light,' * mellow light,' 'ambient light,' etc., designate various qualities. Quality in light is exactly analogous to timbre or quality in sound, which is likewise independent of intensity. The most obvious differences in quality are plainly those called color. But color is bv no means the element of quality. The proportion of invisible rays and the state of diffusion, are highly important factors, but on account of not being directly visible, they have been generally overlooked, and are but imperfectly understood." Xiie Correct "Use of I^igJis. How to Avoid Harmful Effects on the Eyes. — An objection frequently urged against the incandescent lamp is that it is harmful to the eyes and ruins the sight. This is true only in so far as the lamp may be im- properly used. Any form of light as frequently misused would produce the same harmful results. Few people think of attempting to read by an un- shaded oil lamp, and yet many will sit in the glare of a clear glass incan- descent lamp. Incandescent lamps are more generally complained of, because, unlike oil or gas, they can be used in any position. Bookkeepers and clerks are often seen with an incandescent lamp at the end of a drop hanging directly in front of their eyes — an impossible position of the light from gas or oil. The first hygienic consideration in artificial lighting is to avoid the use of a single bright light in a poorly illuminated room. In working under such a light the eye is adapted to the surrounding darkness, and yet there is one spot in the middle of the eye that is kept constantly fixed on the very bright light. The brilliancy of the single light acting on the eye adjusted to dark- ness, works harm. There should be a general illumination of the room in addition to any necessary local light. If sufficient general illumination is provided, the eye is adjusted to the light, and the local light can be safely used. The ideal arrangement provides general illumination so strong that a pencil placed on the page of a book casts two shadows of nearly equal intensity — one coming from the general light and the other from the local light. Care should also be taken to prevent direct rays from striking the eye. The light that reaches the eye by day is always reflected. In reading or writing, to avoid shadows, the light should come over the left shoulder. Only the reflected rays can then reach the eye. Another point to be avoided is the careless, general use of clear glass, unshaded lamps. Frosted bulbs should be used in place of clear glass where soft light for reading is required. The intensity of light reflected from a small source is increased, and intense light injures the eye. With a clear glass globe the whole volume of light proceeds directly from the small surface of the lamp filament. With a frosted bulb the light is radiated from the whole surface of the bulb, and while the total illuminating effect is practically undiminished, the light is softened by diffusion, to the great comfort and relief of the eyes. Finally, the use of old, dim, and blackened lamps, giving but a small fraction of their proper light, is very often a source of trouble in not supply- ing a sufficient quantity of light. Users of lamps are not otfen aware of the loss in candle-power a lamp undergoes, and so it happens that lamps are retained in use long after their efficient light-giving power has vanished. Proper attention to lamp renewals on the part of Central Stations is neces- sary to correct this evil. The correct use of light requires : That there should be general illumination in addition to the light near at hand. That only reflected light should reach the eye. The light should be so placed as to throw the direct rays on the book or work, and not in the eye. That the light should be placed so that shadows will not fall on the work in hand. That shades and frosted bulbs should be used to soften the light. That lamps be frequently renewed to keep the light up to full candle- Dower. CONCEALED LIGHTING SYSTEMS. 601 Distribution of B.ig-ht l>\ Incandescent Lanip<«. The best form of lighting interiors is to have single lamps uniformly dis- tributed over the ceiling; unless the room has been especially designed with this in view, it is sometimes difficult to accomplish. Another method giving most excellent results, but requiring more candle- power, is the arrangement of lamps around the sides of the room close to the ceiling. If the walls and ceiling are of a light color, this method is quite satisfactory, and easier to wire. If the chandeliers, or more correctly in this case, electroliers, are used, it is best to have but one main or large one in the room, balancing the light by side brackets. All such suspended lights should be above the line of vision as far as convenient. » The most economical distribution, as far as candle-power necessary, is the first mentioned, where lights are evenly distributed over the ceiling. To obtain the same luminosity by using clusters of lamps more widely distrib- uted instead of single ones, will require much more candle-power. The 16 candle-power lamp is the universal standard in the United States when rating lamps or illumination, and following are given some ratings on which illumination of different classes of buildings is figured. Ordinary illumination, 1 lamp, 8 feet from floor for 100 square feet, as in gheds, depots, walks, etc. In waiting-rooms, ferry-houses, etc., 1 lamp for 75 square feet. In stores, offices, etc., 1 lamp for 60 square feet. Of course the above must be varied to suit the circumstances, such as dark walls or other surroundings requiring more light, as the walls reflect little of that furnished; and in rooms with dead white walls the reflection approaches 90 per cent, and less lamps would be required than in interiors having worse reflecting surfaces. A very ingenious and satisfactory method of illuminating high arched and vaulted interiors, developed first by Mr. I. R. Prentiss of the Brush Company, is to place a number of lamps around the lower edge of the arch or dome, with reflectors under them, and so located behind the cornice as to be invisible to the eye from the floor. The dome or arch will reflect a large part of the light so placed, giving a very fine, even illumination to the whole interior, without shadows, and very restful to the eye. Of course the arch must be of good color for reflecting the light, or much of it will be wasted. Concealed Lig-hting- Systems. The elements of inefficiency of systems in which the lighting is by con- cealed sources of light, or different lighting systems, have been classified by Millar* under four heads as follows: 1. Light absorbed by ceilings and walls. 2. Loss due to unnecessary intensity at unimportant points. 3. Ineffectiveness of sharply inclined rays. 4. Higher intensity necessary with diffused lighting. Some of his experimental data illustrating these elements quantitatively are given in the following tables. * Millar, Trans. Illuminating Engineering Society, Oct., 1907. i 602 ILLUMINATING ENGINEERING. Table VII. Millar. Temporary Installation at Electrical Test- ing Laboratories. Harlem Office of New York Edison Company. System. System. Direct. Diffused. Direct. Diffused. Total flux of light, lumens .... Flux on working plane, lumens . . Efficiency of light utilization . . . Efficiency of illuminants (lumens per watt) „ , . a. „ Diffused Relative en. of systems; ~. Sacrificed to secure diffusion . . . 424 180 42.3% 2.92 28 pei 72 pei 4824 579 12.0% 2.01 • cent. ' cent. 13938 30532 6642 4689 47.7% 15.4% 3.34 3.34 32 per cent. 68 per cent. Table VIII. Illumination Intensity Required for Reading-. Millar. Foot-Candles. Angle of Paper Observer. with Horizontal. Diff. in Per Direct. Diffused. Cent of Direct. H. E. Allen .... 46° 2.5 4.7 184 Night watchman 42° 3.7 4.8 130 Dynamo tender . 35° 1.85 2.7 144 H. E. Allen . 47° 3.0 5.3 180 W. S. Howell . 47° 2.95 6.3 217 C. H. Sharp . 44° 3.6 5.0 140 Z. N. Corraz 49° 2.3 3.1 135 P. S. Millar . 46° 2.75 5.0 181 F. M. Farmer 49° 2.1 5.0 237 E. Fitzgerald . 49° 2.9 2.6 100 2.7 4.45 165% Note. — The last value obtained, in which the experimenter required the same intensity of illumination with the diffused lighting system that was desired for the direct lighting system, differs from all the other values- Subsequently it was learned that this observer was influenced by the bright- ness of the walls to select the stated intensity upon the paper, feeling that greater brightness upon the walls would be annoying and unpleasant LIGHTING SCHEDULES. 603 Millar's conclusions are as follows: "The conditions of the installations were such that the increase in inten- sity required for reading with diffused lighting was probably larger than may be considered a representative value. The factor is a function chiefly of the brightness of the walls and of the extent to which the walls and other brightly illuminated objects come within the angle of vision. "It was found that if a placard was viewed at a distance of eight or ten feet, thirty times as much light was required to enable an observer to read it as well with the diffused lighting as with the direct lighting arrangement. In this test large portions of the walls were within the angle of vision, and exercised a powerful influence upon the eyes of the observer with both light- ing systems. With the direct lighting system the walls were relatively dark, influencing the pupilary action of the eye so that a low intensity upon the placard appeared satisfactory. With the diffused lighting system they were brilliantly illuminated and so affected the eye that a very intense illumination was required upon the placard. "From the foregoing, the writer has drawn the following conclusions: In diffused lighting systems of the class considered, where the illumination of a working plane is one of the prime objects, a large proportion of the light is lost; that which is not lost becomes less effective; brilliant illumination is produced where it is useless and even undesirable; and conditions are established which create a demand for an unduly high intensity of illumi- nation on objects viewed. "These effects are present in varying degree in all systems in which con- trol of any large proportion of the light is lost. Among such are cove light- ing, lighting with skylight effects, tube lighting, and all systems in which the brilliancy of the light source is reduced by diffusing surfaces used with- out any directing adjuncts. Lighting with large sources is more liable to these effects than lighting with small sources. "The facts indicate the need for devoting as much care to securing suit- able minimum intensities, as is generally expended in striving for maximum values. In certain classes of lighting where more light is asked for, the requirements may be served by reducing the intensity of illumination on unimportant objects which are unnecessarily well illuminated. By taking advantage of opportunities to minimize intensities at unimportant places efficiency is gained, and, in the opinion of many, good lighting as well." MftHTINC} §(HEI)iLES, General Rule for Construction of Schedules. JVEoonligrht Schedules. — Start lamps one half hour after sunset until fourth night of new moon; start lamps one hour before moonset. Extinguish lamps one hour before sunrise, or one hour after moon-rise. No light the night before, the night of, and the night after full moon. During summer months there will be found nights near that of full moon when, under the rule, the time of lighting would be very short. It may not be positively necessary to light up during such times. If better service be desired, but not full every night and all-night service; lamps can be started at sunset and run to 12 or 1 o clock on full-time sched- ule, and after 12 or 1 on the moonlight basis. The above rules by Alex. C. Humphreys, M.E., have been modified by Frund as follows: Light every night from dusk to 12 o'clock; after 12 o'clock follow Humphrey's rule for moonlight schedule, excepting there will be no light after 12 o'clock during the three nights immediately preceding full moon. AU-Nig-ht, Every-Hig-lit Schedule. — Start lamps one half horn after sunset, and extinguish them one half hour before sunrise every day in the year. Full schedule commonly called 4000 hours for the year. All the above rules serve to make schedules for any locality, and such schedules must be based on sun time for the locality, and not on standard time. Permanent average schedules are used in New York City, but for other cities they are usually made up fresh every year. Following will be found New York City time tables, also another set by Humphreys that is a good average for sun time in any locality. 604 ILLUMINATING ENGINEERING. •Sni euiix JNO^^MNOOOJOOOOt-t-WOOlOlQiaiO^^^lfllOiajfiJflOtO • Cooqoqqqququqiqiqiquqiqiqiqiqiqiqiqiqiqiqiqiqiqiqiqiq « •qsm3 ■ui}Xfl; g O O O C O C O O O lO ^ O C r. Ifl O LO iO O lO lO O O li J^^^^eoooeowcocccococoeoeoMcocococccocoeoco'eo^^"^^^ . •^qSiT; ;SiooLu:oL:oioqqqqocoqqqcqqooqqo q q q Ji^L*^c^t^i>t^t^t^odo6<»o6odo6o6odo6Qdco(»odododo6ododoc5*(Nooo(OTjiNHOcoh.ianH05NCOlCT*t^t^t^t>t-t-t-t-t-t-t>t>t>t>t-t>t>t-t-t-t-t-c^t-^- •Sax ©rajx rt©t^ t> t>t^ t>t^ t^t^ . -nana 9UIIX iOOOOOOOOOOOOOOO ~CN^CSOOt^i©Tt«Ct~iOCOCN^©COCO^eOC9©COCOlO g q q i-q iq to o iq iq iq ^ "* rjj t*< t* tj< cq cq cq cq cq cq cq cn cn cn cn cn ^ '-J '-J '-J H'cboioio^ia'ioidujoio'io^'io'ioioiO'o'ioiooiOJOioioioioioioioio •qsm3 -upxa •^qgii; H?0?OtOOQ?DCD?D?03'<^c4ncnc t- CO 1 •IuStt I G^^iqiqoiqiqiq »o J5 c5 q q q — " • I jM id id id id id id id id id id cd co cd co coco co cd cd cd co cd cd cococd cdcdco * ^OOCOt^Oin^COCNOCiMCOlO^fN^CTSCOrfM^OCO^CO^OSt^ Cr^cocqcqcq cq cqcq cq cqc^cNc irS id id id id id id id id id id id id id id jo *q3itf 1-lCNCO^lOCOt^COOJOTHCNCOifiOCOt'COOS S3S33SSS8833 LIGHTING TABLE. 605 •Sat -tun a GUI IX P M CO M CC CO W M ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ t ^ t ^ ^ # ^ ^ ^T ^T 'T ^ ^ ^r -' eo co eo co co co co co co co" co co co* co' co co co co* co" co co' co co* ci co' co" co* co" co' co" co" •qsmS < 'C0t>.00OOHCSM"*l0Ot'0000ClOOOHrtNN^N«n«C0'fT(<^ jcococoM^^^^^^^^^^^^^ou^cicwioou'ii^ciqioifliA :«o»w«c©-^COTt<©t>-l>00©'--i £cOCOCOTt<^T*T^lOiqiqiO©0©©OrHr^i-HrHi^CNCNC>^ J, CN CN CN* CN CN CN* CN CN ©CN»Ot-©CO?DC5©COlflL-©CO'X>00'-''^OC5^H guoiq©©©©^^r^cNCNCNCNcqcqcqcq^^TjHT^iquqiqio©©©©^ • ^ Oi OS ©' ©©'©'©©'©©©©©©©©©©©©©©*©©©'-' i-« TH rH tH • ^0005©T^CNCO^l0^l>COCi©-^CNCOTt«-^lO©t-CO©©i-iCNCOrflO© * Cioiqqqqqqqqqqq^^i^^^«^rH«rj«c^cscNWNc>iN . •?u3iT -"CO^©00©Tt©TfCO'--iCil>lO'*CNi-i©O0©Tt < C0'--iCi0O©lO S © ©. ©. >q iq >q iq *q ■"* "# ** "# -* ^ cq co cq co c^> cq c^ cn cn cn cn c^rHr-M ih ^t>-t>t>"^©©©"co©©cdo*©©©©'©©'cco©©*©*c^^©©©cD?o *SlII -*OCN^C30iTHCO©CC©CN^I^OiCN^«000©CNlOt--©COt>©i--iTf«005CN ..... ** J3 'odc»'c»o6cx'oda'a"o5©*©'©©'c^ •qsmg -uijxg; •;q§iT: J ■<# rjj rij t# rjj <* « •Sui -a\mg -*00(^©^CNCNCO)^lC©a5C75©^CO^©t>'C»©^COlOr>O5'-wCOin?Dt^C75CN 5 iq iq iq © © © © © © © © © t-j i-j r-j h *■< ,-h h cn cn cn cn cn cn cq cq cq cq cq co N-i ^'^^^odc»cx5ooo6o6odcoa"o6a"odcx"o6a"w . „ © © © © © © © © © © © w-i rH 1-* «-J i-J t-H rH ,-h r-j rn -h *h CN CN CN CN CN CN C - . - UUX^f Iq rj* rj* rl* t}* T** ■<* rt" Tj" rf<" ^ Tj* -# Tt* "tf* tJ* ■<*' rl" rl" -t>*.|^ JO^qSi^ | - 1 e^C0^»O«Or>-Q0C35©^CNCOrH»O©b-00O»©T-i SSSSSSSSSSSS 606 ILLUMINATING ENGINEERING. Summary of New York City I^igrliting* Tahle. January . February March . . April . . May . . June . . July . . August September October . November December Hours for Average. Average the Month. Day. h.m. h.m. 413.10 13.19 18th 355.27 12.15 15th 341.29 11.01 16th 290.17 9.40 16th 264.39 8.32 15th 238.51 7.57 12th 256.12 8.16 17th 286.26 9.14 16th 316.48 10.33 15th 368.50 11.54 16th 392.59 13.05 14th 424.52 13.42 10th Total hours 3950 Shortest Longest Average Note. — Lights started 30 minutes after sunset. Lights stopped 30 min- utes before sunrise. For commercial lighting : add 1 hour for part night lights, add 2 hours for all night lights to above schedule. Table Showing- lumber of Hours Artificial liig-ht i* deeded in Each month of the Year. Dr. Louis Bell. Evening from Dusk to 6 o'clock Dusk to 7 o'clock Dusk to 8 o'clock Dusk to 9 o'clock Dusk to 10 o'clock Dusk to 11 o'clock Dusk to 12 o'clock All night .... Morning from 4 o'clock to dawn 5 o'clock to dawn 6 o'clock to dawn 7 o'clock to dawn 13 44 75 116 217 82124 62 80 92 111 122(142 152173 102 112 155 182 204 189 133 164 307 16 142 186;212 172 217 242 345 421 473 110 80 50 20 235 266 527 137 106 75 44 b c3 a 3 A o u 3 >> 03 a »-5 1-5 fe S < s 65 33 61 4 31 96 4 127 89 62 28 4 158 117 93 58 29 8 189 145 124 88 60 38 220 173 155 118 91 68 251 201 186 148 122 98 512 411 382 295 242 195 137 93 71 28 2 106 70 40 3 75 42 9 44 14 722 472 269 122 LIGHTING TABLE. 607 SI s2 8^ £> S3 S§?SSSSaSSSgS2888SSSSSSSiSSSS5?? hOOOOOOOOOOOOO .3* ^ CO SOOOOOOOOOOOOOOOOOOOQOQQOOOOOOOO ^lOlOU5»CCOCOW^COCO«D«>«0«C>«Ci«0tOCO«0«OOiX>CO^)«OiX><£>iX» SI W5c .©OOOOOOOOOOOOOOOOOOOQQOQOOOO -iqioiowio^^wnwconrtcjNHrjrtHHqqoqiciqifl^ SSSS^T^ooooopoo»qioiq»oiqiou3^^Tj;^T}|TfTj;« j; -i d d c c 0'-t«Meo"Tfif5cot*ooc5 0T-( 608 ILLUMINATING ENGINEERING. o £228388888388883388883833888888 jgoooooowooooooocooooododooooooooooooooooooooooodoocooocooocjo ^eocococo^co^co^co^co^coco^co^co^eo^eoeo^co^eo^eo 3 « o 83 22 * -s. - is 4"!' s s W M S8888SgS§883S888S8888S88SSSSSSSS ^Oso>o6o6o6o6c6o6o6o6oc5»o6o6o6o6odo6o6o6o6o6o6o6o6odo6o6odo6o6 5 3 3 § § § § tf5 i£> »fl io ic io rj« ^ -^ t* rf ! 5j t* ■*« ? tj; <<* Tf CO CO CO CO CO CO co ^'^T^coco^cocococococococococococoeo'cocoeo^co'coco gS8888S8888SSSSSSSSSS833888883S8 w o 3g ^NCO^lOWfc-COOejMgw^gW^wagggggggg.gggg IS S8S2SS888§8?5???S?S888S888SSSSS *CCOCOCOOOCOwiO)0>010)0 05 0)CJClOO)QQOC!ODC) 5000000000000000 83883 05 O cRSeocoeococococococo^TtjT^^^^Tfi^^TfSSoo ^<*w*^®^««©3222;3SS;$S38SSc3&S8&c588 LIGHTING TABLE. 609 c~ «S ,0 v a 1 v c § Is u -z ft w s 9 a © u -° 2 s ° H P (s; bo •6 ^ &q bo dS2SSS®OOOOOOOOOOOOQOOOOQQQOOOO coooqoHH«fjc]N«n««nMioicooooooooHHS S'OOOQQOOOOQQQOOOOOOOOOOOOOQOOQQ _ co w « « « « to ■* •* ^ # ^ t ^ 't ^ ^t ■<* ■"f o 10 >o w o ifl 10 »o o o q o © ^ rj" ^5 ijJ r£ rj* t* t** tjJ "^ rt* -^ rj* -^ t*° -^ -^* ■>* tjI TJ* -^ rj* t*i" ^" "^' t** -& •<* id O O ,c5«o?oddddddddd«o>oddddddddddioioioio»oi©iQ ' i-HCNf0"^i0<0t'Q005OrHCNC0rjt-00C5O^; cncn^c^cncscscncnco ^ooooGOoococooocooJoJdoi^oodaJddddcjjdd^dd S888888§8SSSSSSSSSS8S8S8S^8SS888 rj ^< ^< t^ *& *& t^ ^ "^ "^ "^ "^ ^^ ^^ ^J^ "*<■$■* ^^ '■"J' ^< ^* ^^ ^* ^* *^ *^ ^** *^ ^* ^t* "^ ^^ ^^ SSSSSSSSSS8?8888888^?§SS§§S§55?8 c8 O AS a o .3* riCNeorj4iri«ot-ooo>OT-icqcortQoa>Oj-j OHfNco^mo^ooosOH C^C^CNCNCNCNCNCNCSCNeOCO S888SSSSSSS222SSSS8S8888888888S5 ,£* 00 00 CO 00 00 CO CO 00 00 00 00 co" 00 00 00' 00* 00 00* 00" 00" 00" 00 co' CO* CO* CO* CO* CO* CO* CO* CO ScOCO°§^rj;^5?^^?^§^Tl|rJ*^t>t^r>t^t>"^t^t>^^t>'^i>"^i^^t^t>°^t^t>r^t>"a~c^t^t^^ e8 O l«S rt(NP5^iOWl»CC030HCNn^iC»t>.OOOJO'-i(SCO^ifl«ON«OlQi-i i-iHrtT-(rtilT-lr-lHrtCN(NC^CSCNC5^WCSC4CO?) 610 ILLUMINATING ENGINEERING. Bo 8 ^ Si II g *5 W 8t& ft w s s .coeoeoeoeoeoeococoeocoeoeoeocoeocoeoc^e^ 55 •°5 CcocoMcocococoeococococococococococcco^^ , ''^^^ij , ^^^j^j , ^j , ij^ ^ ci (M c* 5* eo co s88S8gg?g§3gSSSSSSSSS888§8S8§8S rtC^c>ic4COO>©iHe*W^iaOt>aOC^©i^C$$c4e3.M. J 4 5 6 7 8 9 10 11 MI0.1.AM. 2 3 4 5 6 7 8 1 \ 1 i "I""" iii: ■W- * v - ^ 1 p FEB. f MAR. >U , ii ; ' • r '■• ' /a \ . / m APR. 0\ / / JUNE y\ ;' J /l P V ' V / ' J - V OCT. u/y i ^ S\j* \ DEC. / jj j 1 1 < ?' t" ~J he snaded area represents the time during which light la required. The horizontal lines show the months of the year. The vertical lines show the hours of the day and night. The inner dotted lines show the time of sunset and sunrise. The outer lines show the time of lighting up and extinguishing. Each square is an hour month, i.e., 30.4 hours? "HIS )s oum ELECTRIC RAILWAYS. Revised by A. H. Armstrong, C. Renshaw and N. W. Storer. The electric railway motor has made such rapid strides in traction that it has pre-empted the entire urban field, taken most of the traffic from the suburban steam lines and is now appearing as a formidable competitor to the steam locomotive in heavy haulage. In considering, therefore, the appli- cation of the electric motor to traction work, it is necessary to determine its capacity and characteristics for city service and single car operation, and also for electric locomotives hauling heavy trains, either high speed passenger or slow speed freight. Small cars weighing 10 to 12 tons may be fitted with two 35 h.p. motors and be geared for a maximum speed of 25 to 30 m.p.h. Larger cars of the single truck variety weighing close to 15 tons may be equipped with motors of 40 h.p. capacity. Single truck cars are used (to a large extent) for city work, although in this class of work the use of double truck cars is rapidly increasing. Suburban cars weighing 18 to 25 tons and measuring 45 ft. overall may be equipped with four 50 h.p. motors and be geared for a maximum speed of 40 m.p.h. Such cars usually make stops approximately every mile, and a schedule speed of about 20 m.p.h. outside of the city limits. Larger types of suburban cars 50 ft. overall, seating 52 passengers, weigh 28 to 30 tons and are equipped with four 75 h.p. motors geared for maximum speed of 45 m.p.h. These cars usually make a stop every mile and a half, and a schedule speed of 25 m.p.h. for the local and 35 m.p.h. for the express cars outside of the city limits. The largest type of suburban car, of which that of the Aurora, Elgin & Chicago is typical, is equipped with four 125 h.p. motors, is geared for maximum speed of 60 m.p.h. and stops but once in two or three miles, making a schedule speed of about 35 m.p.h. These cars represent the highest type of interurban electric railway and their use seems justified under certain conditions. Orades. — Grades upon city lines may run as high as 13 percent, and to surmount these it is necessary to have every axle on the cars equipped with motors; thus a single-truck car would require two motors and double-truck cars four motors; and even then the cars will be unable to surmount these grades with very bad conditions of track. Surface cars operating over city streets have no option but to use the prevailing grades, hence for city work where heavy grades are liable to be met, the motor capacity per car should be liberal, not so much on account of the danger of overheating the motors, as to prevent undue sparking when surmounting the heavy grades. The tendency of the suburban roads is to operate over private right of way, and grades on these roads do not generally exceed two or three per cent, except for very short runs where they may reach four or five per cent. Grades exceeding these are infrequent, and on the best high speed suburban roads two per cent grade is the maximum allowable. The effect of grades upon the heating of motors is largely compensating as the motors cool off nearly as much in coasting down grades as they overheat when doing extra work in surmounting the grades. Curves. — In city work sharp curves are necessary in rounding street corners and curves of 50 ft. radius are sometimes met. These curves are oftentimes so sharp as to prevent the use of heavy, long double-truck sub- urban cars. Such curves cannot easily be avoided and city cars are designed with short wheel base of trucks, generally not over 6 ft. in order to be able to round these sharp curves. The maximum speed of city cars is limited to about 15 m.p.h., so that these sharp curves cannot interfere seriously with the schedule. Suburban cars operate over much straighter track and have a maximum speed of 25 to 50 miles per hour. It is seldom that the curves are sharp enough to seriously inconvenience the purely suburban class of service. Roads operating over private right of way endeavor to limit the curves to five degrees, which can be rounded at a speed of 35 miles per hour, so that 612 ELECTRIC RAILWAYS. 613 they do not seriously interfere with the schedule. Very high speed suburban roads will not permit curves of more than three degrees, as a sharper curva- ture interferes with free running speed of the cars, which sometimes approaches 60 miles per hour. Sharp curves are more detrimental to the maintenance of high speed than grades of four or five per cent unless the latter be of considerable length. Systems of Operation. — There are four systems of operation now in use for electric railways, each of which has some distinctive advantages warranting its use under certain conditions. 1. D. C. generation and D. C. distribution with the possible use of boosters or floating storage batteries. — This system is pre-eminently adapted to the very congested travel of the more densely populated sections of our larger cities. It is not well adapted to the operation of roads covering large areas and is rapidly becoming obsolete, owing to the great amount of feeder copper required to transmit large amounts of energy at 600 volts, which is the standard potential used. The use of boosters is objectionable for con- tinuous work as they add largely to the fuel expense, while a floating storage battery at the end of a long feeder is oftentimes more expensive to install and operate than some of the other systems described later. The direct- current generating system for larger supply is rapidly becoming obsolete, except in localities where the conditions are very favorable for its retention. 2. Alternating current generation and transmission to rotary converter substations. — This system is being used almost entirely for our suburban roads and larger city systems. Alternating current generation and trans- mission offers the advantage of the ability to transmit great power over long distances at very high potentials, in some cases reaching 60,000 volts, so that the copper expense is relatively small. New York City is fed entirely from rotary converters which receive their power from alternating current generators and alternating current transmission lines at 11,000 and 6,600 volts. The office of the rotary converter substation, which was first used in 1897, is to reduce the high potential alternating current to low potential alternating current, then convert it into 600 volts direct current which feeds into the trolley or third rail, as the case may be. 3. Three-phase alternating current feeding direct into high potential trolley and thence into three-phase motors upon the cars is used on some European roads. 4. The single-phase alternating current commutating motor has been devel- oped in several forms since 1904, and there are now quite a large number of roads' operating in this country and abroad, using this type of motor. This motor is said to be more flexible than the three-phase motor, as it has a variable speed characteristic very similar to that of the direct current series motor. Its application in the railway field is therefore much more general and it will undoubtedly find considerable use in suburban work and in the heavier class of electric railways. Train friction. — The resistance offered by air against the front and sides of a rapidly moving car forms a very important factor and has been the subject of a large number of experiments. The most complete are probably the Berlin-Zossen experiments where speeds of 125 miles per hour were reached and wind pressures noted. A large number of formulae have been introduced by different authorities covering the resistance offered by the air, rails, journals, etc., when operating single cars and trains at different speeds. The formulae developed by steam railroad experimenters using heavy trains of many cars may be discarded as worthless when applied to electric traction using single car units. In the same way the results obtained from the operation of single cars cannot be applied to trains, as the wind friction of the succeeding cars is not as great as that of the leading car. These train friction results will be treated and commented on later on in this chapter. Wind friction plays a very important part in determining the power consumption of electric cars operating at high speeds, and both the energy consumption and capacity of the motive power plant must be carefully determined with a full experimental knowledge of wind friction in view. [ Car Equipments. — Car equipments have increased from motors of 25 h.p. for small single-truck cars on city streets to motors of 550 h.p. each, as in the "Mohawk" type of electric locomotive designed for the New York Central Railroad. Electric motors can be designed to meet practically any conditions of operation, but the standard lists of manufacturers run from I 614 ELECTRIC RAILWAYS. 25 h.p. to 200 h.p. in about 25 h.p. steps, in the larger sizes, and less differ- ence in capacities in the smaller sizes. It is better to refer to the manufac- turers when a motor is to be selected for a given class of service which differs materially from a known service upon which full data is at hand. With such a wide range in capacity of motors it is necessary to study the conditions very carefully in order to properly determine the correct size of motor to use. Some general curves are given later from which reasonably correct approximations can be made, but these should be verified by con- sultation with experts in motor design. locomotives. — Electric locomotives have been built for a variety of purposes from yard shifting to the hauling of passenger trains weighing 900 tons at speeds approaching 60 miles per hour. Nearly all these electric locomotives so far have been equipped with direct current series wound motors operating at 600 volts. A number of locomotives in Europe, how- ever, have been equipped with three-phase alternating current motors and a few with single-phase motors. In this country there are now in operation on the Spokane & Inland Railway, 1907, six 50-ton locomotives, each equipped with four 150 h.p. single-phase motors arranged to operate on either 600 volts direct current, or 6600 volts single-phase alternating current. The Westinghouse Electric & Manufacturing Company, who built these locomo- tives, have recently completed thirty-five 88-ton electric locomotives, each equipped with four 250 h.p. single-phase motors arranged to operate on either 600 volts direct current, or 11,000 volts single-phase, alternating current for the New York, New Haven & Hartford Railroad, and also six 60-ton locomotives, each equipped with three 240 h.p. motors for operation on 3300 volts alternating current for use by the Grand Trunk Railroad in the Sarnia Tunnel. The use of electric locomotives is rapidly increasing as the economic operation and other advantages of their operation are appreciated. ]>esirable Points in motors and Car Equipment. — It is desir- able that motors should be electrically sound, i.e., that their insulation should be high, mechanically strong, and waterproof. It is of great advantage in this connection if the entire frame of the motor can be insulated from the car truck and consequently from the ground, thus relieving the insulation of the armature and fields of half the strain. The mechanical difficulties in the way of accomplishing this, however, go a great way towards counter- balancing the advantage gained. A high average efficiency between three quarters and full load should be obtained if possible, but mechanical points should not be neglected to obtain this. A motor should run practically sparkless up to § of its rated capacity. A low starting current obviously is desirable, and for obtaining this nothing is better for continuous current operation than a multiple series controlling device, which cuts the starting current in half. This device also enables cars to be run at a slow speed with good efficiency. Mechanically, the motor should be simple. The fewer the parts, and especially the wearing parts, the better. It should be well encased in a cover- ing strong enough not only to keep out water, pebbles, bits of wire, etc., encountered on the track, but to shove aside or slide over an obstruction too high to be cleared. At the same time, the case should be hinged so that by the removal of a few bolts access can be had to the whole interior of the motor. The brush holders and commutator should be easily accessible through the traps in the car floor at all times. As much of the weight of the motor as possible should be carried by the truck on springs ; if practicable all of it. This arrangement saves much of the wear and tear on the tracks. A switch in addition to the controlling stand should always be provided, by which the motorman himself can cut off the trolley current, in case of accident to the controlling apparatus. Roads having long, steep grades should have their cars provided with a device for using the motors as a brake in case the wheel brake gives out. There are several methods of accomplishing this, but limited space prohibits any description of them. Last, but by no means least, all wearing parts should be capable of being easily and cheaply replaced. WEIGHTS OF RAILS. weights OF RAIIS. 615 Pounds per "Weight per Mile. Weight per 1000 7 . Yard. Long Tons. Long Tons. 640 986.7 25 39 2240 320 39.286 7 ^240 2080 7.441 30 47 2240 47.143 8 2240 933.3 8.929 35 55 1920 55 10 2240 2026.6 10.417 40 62 2240 1600 62.857 11 2240 880 11.905 45 70 2240 960 70.714 13 2240 635.5 13.393 48 74 2240 1280 74.428 14 ~2240 1973.3 14.284 50 78 2240 1600 78.571 14 2240 1066.7 14.881 52 81 2240 960 81.714 15 2240 826.6 15.477 55 86 2240 86.428 16 2240 16.369 56 88 320 88 1604.4 16 2240 586.7 16.667 58 91 2240 2080 91.143 17 2240 ^1920 17.262 58} 91 2240 640 91.928 17 2240 920 17.411 60 94 2240 960 94.286 17 2240 1013.3 17.857 62 97 2240 97.428 . 18 2240 1680 18.452 63 99 1760 99 18 2240 2013.3 18.75 63} 99 2240 1rtO 320 99.785 18 2240 773.3 18.899 65 102 2240 103 1600 2240 102.143 iy 2240 1440 19.345 66 103.714 19 2240 19.643 irv. 1120 1773.3 66} 104 2-240 104.5 19 ^2240" 2106 19.792 67 105 2240 1920 105.286 19 2240 533.3 19.940 68 106 2240 106.857 20 ^240 2000 20.238 70 110 111 280 110 ' 20 2240 293.3 20.833 71 2240 111.125 21 ~2240 21.131 616 ELECTRIC RAILWAYS. WJEIOHTS OF It AIL§ — Continued. Pounds per Yard. Weight per Mile. Long Tons. Weight per 1000 '. Long Tons. 72 75 77 78 80 82 85 90 91 98 100 320 113 2240 1920 117 2240 121 122^ 2240 1600 125 2240 1920 129 2240 1280 133 2240 960 141 2240 143 154 „320 157 2240 113.143 117.857 121 122.143 125.714 129.857 133.571 141.428 143 154 157.143 960 21 2240 720.2 22 2240 2053.3 22 2240 480 23 2240 1813.3 23 2240 906.6 24 2240 666.6 ^2240 1760 26 2240 ^186.6 27 2240 373.3 29 2240 1706.7 29 2240 21.429 22.322 22.917 23.214 23.810 24.405 25.298 26.786 27.083 29.167 29.762 For iron or steel weighing 480 lbs. per cubic foot : Cross-section in square inches =: weight in lbs. per yard -7- 10. Gross tons of rails in 1 mile single track — weight per yard X 11 RADII OF CITRVEi FOR DIFFEREUT Dt^KHKM OF CURVATURE. CO +a ■js « GO ^J P P O 2 <» «.g CO -+J P <£> i 5730 ii 522 21 274 31 187 41 143 2 2865 12 478 22 262 32 181 42 140 3 1910 13 442 23 251 33 176 43 136 4 1433 14 410 24 241 34 171 44 133 5 1146 15 383 25 231 35 166 45 131 6 955 16 359 26 222 36 162 46 128 7 819 17 338 27 214 37 158 47 125 8 717 18 320 28 207 38 154 48 123 9 637 19 303 29 200 39 150 49 121 10 574 20 288 30 193 40 146 50 118 ELEVATION OP OUTER RAIL ON CURVES. 617 GRADES 1 % PER CEIT A N 1> RISE 1 I¥ FEET. Rise in Feet at Given Distances. Per Cent Grade. 500 Feet. 1000 Feet. 5,280 Feet (1 Mile). * 2.5 5 26.4 1 5 10 52.8 1.5 7.5 15 79.2 2 10 20 105.6 2.5 12.5 25 132 3 15 30 158.4 3.5 17.5 35 184.8 4 20 40 211.2 4.5 22.5 45 237.6 5 25 50 264 5.5 27.5 55 290.4 6 30 60 316.8 6.5 32.5 65 343.2 7 35 70 369.6 7.5 37.5 75 396 8 40 80 422.4 8.5 42.5 85 448.8 9 45 90 475.2 9.5 47.5 95 501.6 10 50 100 528 11 55 110 580.8 12 60 120 633.6 13 65 130 686.4 14 70 140 739.2 15 75 150 792 Note No. 1. — For other distances interpolate the table by direct multi- plication or division. ELEVATIOIV OF OUTER UAIL OI¥ CURVES. o . Speed in Miles per Hour 10 15 20 25 30 35 40 45 50 60 a « Elev? ition c )f Outer Rai L in Ir ches. i 5730 A ft i 4 i 7 * s 41 1ft If iH 2* 2 2865 ¥ >? A ,* 1* ift 2ft 2| 2ft 4H 3 1910 tIt ± +i Aft i* 2* *ft 4* 5* 7# 4 1432 i * A }tt 2* 3f *i 4ft m 9f 5 1146 S % i* « % d. 3 5ft 6J 8* 12ft 6 955 T'* ,** H *ft ?** 5 <*ft 8* 10ft 7 818 * {ft ltt 3 !ft 5£ 1% 9ft llf 8 716 ft i 2 T 3 s 3ft 3* 6H ffi 10fr 9 10 636 573 h J* 2* 2f 3fi 4* 7* 8ft 10| 12H 11 12 521 477 1* 3 3ft 3fif 4Ii 5* 4 7 T 5 ^ 9* 14 409 H 2 T 3 S 5H 8 T 9 TT 111 16 18 20 358 318 286 1A if Ol 2f 3ft 4i 5& 6H 7f 8£ m 10| 12 Note No. 1.— When E = elevation in inches of outer rail above the hori- zontal plane: V: R: : velocity of car in feet per second ; - radius of curve in feet ; V 2 Therefore E = 1.7879 —when gauge of track is 4'-8£" R 618 ELECTRIC RAILWAYS. §PIKE§. Size. No. per Keg of 200 Lbs. Lbs. per Spike. Spikes per Lb. 4*X£ 533 .3752 2.66 5 Xt*s 650 .3077 3.25 5 X* 520 .3846 2.6 5 Xi 9 s 393 .5089 1.96 bhXh 466 .4292 2.33 5* X r 9 s 384 .5208 1.92 6 Xft 350 .5714 1.75 6 Xf 260 .7692 1.3 spikes jpjeu looo a\d per mile ii^gle track:, with four spikes per tie. Spacing of Ties. Per 1000 7 . Per Mile. 10 ties to 30 7 rail 1333£ 7040 11 " " " " 1466| 7744 12 •* " " " 1600 8448 13 " «• " " 1733£ 9152 14 " " " " 1866| 9856 15 " " " " 2000 10560 16 " " " " 2133£ 11264 JOINTS PER MILE OE SINGLE track:. Per lOOCX. Per Mile. Joints — 30 / rails Angle bars 66§ 133| 266§ 400 533£ 800 352 704 Bolts — 4 hole bars 6 " " 8 » " " 12 " " 1408 2112 2816 4224 TIES PER lOOO A]¥» PER MILE, BOARD FEET, CUBIC EEET, AID SQX T ARE EEET OE BEARING SUREACE PER TIE. Size. Board Feet. Cubic Feet. Bearing Surface 5" X 5" X V 14.56 1.213 2.91 5" X 6" X V 17.5 1.458 3.5 5" X 7" X V 20.41 1.7 4.08 5" X 8" X V 23.33 1.944 4.66 6" X 6" X V 21 1.75 3.5 6" X 7" X V 24.5 2.041 4.08 6" X 8" X V 28 2.333 4.66 6" X 9" X 7' 31.5 2.625 5.25 6" x 10" x 7' 35 2.916 5.83 0" X 8" X 8' 32 2.666 5.33 C" X 9" x 8' 36 3 6 6"xl0" x 8' 40 3.333 6.66 PAVING. 619 REPORT OF U. S. DEPARTMENT OF AGRICUI- TURE ON DrRABILITY OJP RAILROAD TIES. White oak 8 years. Chestnut 8 « Black locust 10 " Cherry, black walnut, locust 7 « Elm G to 7 " Red and black oaks 4 to 5 " Ash, beech, and maple 4 « Redwood 12 " Cypress and red cedar 10 " Tamarack 7 to 8 " Longleaf pine 6 " Hemlock 4 to 6 " Spruce 5 « PAVMG. Paving prices vary so that it is difficult to state even an approximate cost chat will not be dangerous to use. Prices are not at all alike for asphalt, even in cities in the same localities ; other styles vary according to prox- imity of material, cost of labor, and amount of competition. Square yards of paving between rails, 4' 8|" gauge, less 4" for width of carriage tread : Per 1000 / run = 485.89 sq. yards. Per mile run = 2565.5 Square yards paving for 18" outside both rails : Per lOuO 7 run rz 333£ sq. yards. Per mile run = 1760 " Approximate Cost of Paving (Davis.) PAVEMENT Cost of all Material and Labor. Cost of Tearing up Existing Pavement and Repla- cing as Found. u JO B O e« u <0 O cu.5 "So U PM Granite blocks on gravel foundation Gravel blocks on concrete foundation Asphalt on concrete foundation . . Vitrified brick on broken stone . . . Wood without concrete Cobble without concrete Macadam $ 2.80 3.60 3.80 2.15 1.50 2.00 1.00 $ 2.24 2.88 3.04 1.72 1.20 1.60 .80 $ 12000 15500 16000 9000 8000 8500 4500 $ .35 .45 .45 .30 .50 $ 1900 2400 2400 1600 2700 ESTIMATE OF TRACK IAYOG FORCE, One engineer, 1 rodman, 1 foreman of diggers, 1 foreman of track-layers, tspikers, 20 laborers, 2 general helpers. Such a gang can lay from 400 to 900 feet of single track per day. In case it is desired to proceed more rapidly, the above number of men 620 ELECTRIC RAILWAYS. should be increased proportionately, omitting the engineer and rodman, as these two will be able to handle any ordinary number of gangs, no matter how widely scattered, if a horse and buggy is placed at their disposal. Tools for Track Gang* as Above. — One portable tool-box pad- locked, 1 small flat car, 1 portable forge, 4 cold chisels, 2 ball pein hammers, 6 lbs.; 1 sledge,12 lbs.; 2 axes, 2 adzes, 1 cross-cut saw, 1 large double-handled saw, 6 track wrenches, 2 monkey wrenches, 1 complete ratchet track drill with bits, 1 track "Jimmy" for bending rails, 1 reel line cord, braided; 30 picks, 15 extra pick-handles, 25 long-handled, round-nose shovels, 6 short handled, square-nose shovels, 10 tampers, 5 wheelbarrows, 2 track gauges, 1 level, 1 straight-edge, 4 pair rail tongs, 6 spiking hammers, 3 crow-bars, one end sharp, the other end chisel-pointed, 2 spike claw-bars, 1 engineer's transit, 1 leveling-rod, 10 surveyor's marking-pins, 1 steel tape, 10 red lan- terns, 1 box lump chalk, 1 squirt oil-can, 1 quart black oil, 5 gals, kerosene, I flag-rod, 1 paper of tacks, 1 broad-blade natchet. RAILWAY TIRXOUTS. By W. E. Harrington, B. S. For example, assume a railway to operate 4 cars, the distance between terminals four miles, the time of round trips 60 minutes, and the headway 15 minutes, with a lay over at each end of five minutes. Take a piece of cross-section paper, and make the vertical lines represent distance, and the horizontal lines represent time. The time necessary to run from terminus to terminus is half of 60 minutes, less £ of ten minutes (the layover time), or 25 minutes. Let each division on the ordinate axis represent the distance traversed by a car in one minute, which in the above case is 844.8 feet per minute,as- suming that the car is to run at the average speed of 9.6 miles per hour. Let each division on the axis of ab- scissas represent five minutes. The first car will travel from terminus to terminus as represented by the diag- onal line OA. This line shows the car's position at any instant of time, assuming, of course, that the car is running at a uniform rate of speed. The car upon its a nival at the other terminus will have a lay- over of five minutes as repre- sented by the horizontal space AB Fig. 1. Location of Street Railway Turnouts. Upon the expiration of the time of lay-over the car starts upon its return run. This determines the locus of the several turnouts, as the car has to pass each of the remaining cars. The line of the return run is represented by the line BC. Upon the arrival of the car at the original terminus and a lay-over of five minutes, the cvcle of trips will be repeated. During tne time the first car is running its round trip the other cars are leaving at in- tervals of 15 minutes, as represented by the lines DE. FG, and HI. Where these three lines intersect the line BC turnouts must be located, as the cars meet and pass at these points. The distance apart of the turnouts, as well as th'eir distance from the starting terminus O, may be readily determined by projecting the intersections on the axis of ordinates OY. 1. The number of turnouts for a given number of cars is one less than the number of cars running. RAILWAY TURNOUTS. 621 2. The time consumed running between turnouts must be the same between all the turnouts. For instance, if it is found necessary to irregu- larly locate turnouts for any reason, then the time consumed by a car run- ning between these two turnouts farthest apart determines the time the cars must run between the remaining turnouts, even though two or more of the turnouts be only a slight fraction of the distance apart of the two greater ones. 3. The time consumed running between two consecutive turnouts is one- half the running time between cars. For determining the distance apart of turnouts without the aid of graph- ical methods : Rule. — To the length of the railway from terminus to terminus add the distance a car would travel running at the same rate of speed as running on the main line, for the time of lay-over at one terminus. Divide the above result by the number of cars desired to be run, the result is the distance between turnouts. Multiply this latter result by two less than the number of cars, and deduct the result obtained from the length of the line from ter- minus to terminus, and divide by two. The result is the distance from either terminus and the first adjacent turnout. To operate more or less cars on a railway than it is designed for is a ques- tion most frequently met in railway practice. Rule 1 tells us that we must have one turnout less than the number of cars running. In Fig. 1 we have four cars and three turnouts. If we pro- pose running three cars we would use two turnouts, by omitting the middle turnout. The result is at once apparent ; for according to Rule 2, the time to run between turnouts is determined by the time consumed in running between those two turnouts farthest apart. Since the distance is doubled, the time consumed is doubled. Where with four cars, with fifteen minutes between cars, and sixty minutes for the round trip, with three cars the time between cars as by Rule 2 is thirty minutes, and the time of round trip is ninety minutes, making at once a very pronounced loss. . The better plan, and the one usually pursued by railway managers, is to run the lesser number of cars on the same trip time as the railway was designed for. In our example above, the three cars would be run as if the four cars were running, with the exception that the space which the car should be running in will be omitted, leaving an interval between two of the cars of thirty minutes, giving only the loss occasioned by the omission of one car. Another method to pursue, especially so where additional cars will be run at times, such as holidays, excursions, and other times of travel requir- ing more than the regular number of cars to accommodate the travel, is to provide and locate more turnouts. The expense of doubling the number of turnouts, while they would be a great convenience, would not be warranted without the railway were doing a large and growing business, with a fluctu- ating number of cars in service. Two cases should be considered. First — If a certain fixed number of cars are to be operated for the greater portion of time and the extra cars for odd and infrequent intervals, locate the turnouts to suit the regular business. Second— In the case of a railway running an irregular number of cars — for instance, a railway running a heavy business at certain times of the day — as the lesser number of cars are subordinate to the greater number, locate the turuouts to run the greater number of cars the most efficiently. In conclusion, we might state that the grades, the running through crowded business streets, stoppages occasioned by grade railroad crossings, and varying business, all enter in and must be considered while designing. 622 ELECTRIC RAILWAYS. i:le(thi( hailway automatic block SICHJTAXJLIJra. By Charles F. Hopewell, S.B. Block signalling on single-track railways accomplishes two purposes, namely, that of ensuring safety and of obviating delays in traffic necessitated by cars always meeting at predetermined turnouts. Electric Railway Signal Systems have three positions of signal display, viz.: normal; safety, indicated by green; danger, indicated by red. The red signal is at the leaving end of a block and the green signal operating in unison with it is at the entering end of the block. Were it not so, a car entering a block could not determine if it set a danger signal or the same was set by another car entering from the other end. This requires the TypaA. Douh/eTracks ^\poubk Acting Tnilley Switch =5 ia HO OeuhleActmo f > -*c OoubleActing < Sinale Track. MyS mtct L/ Turn^^out. ^ro lle ^wifch^ Swirc'hi'/nel] p'iff'n'o/I/n?s"\ rs'rsK'n'L'ines. -JOOfeet *-" b? ' , M *ft— Signa/ 3c a Signal oca Swikhi'nes 1 j" loo fee t - A * —/aofect-^ SlflMlSoj Sin gle Act ing TrvS/eyJWitcne*. >eCorLenf~f?i TypeB. Single AcringTrvlleySwhthes - Sing It Track> trleng/fcd 1 \ bjN/^„ -O^ .l— I *>rn-vt r — V^ S#itc'nLin*s. ] XS/'g no/ lines Swirch L ■', —won •Si'nqleActinq TrollejSWJTtAi Br^-g? Double fenV^'^ Signal Bo*. SignalBbx, TypeC Single ^ Trac, dt S/gnetBax. Oovb/'Actinf Tnli/Stfi*y Turrtciif -Noft- Arro#jSh0\ CarDirtcTion. socft. ^1 — J ^T_J- /aoA. £ ggTxy ^i $fah Jw'i/cnUne's' ' SifnolBo/. SicjnalBax Typical Methods of Blocking. UNI SIGNAL CO. BOSTON 'Mastf Fig. 2. normal position of signalling to be when no car is in the block. A motorman of a car approaching a block may have one of three indications signalled to it: No distinctive signal or light, indicating that the block is clear; a green signal indicating that a car has entered the block proceeding in the same general direction as the observing car, and a red signal indicating that a car has entered the block from the distant end and is coming towards the observer. There are three distinctive methods of blocking a single track for operating in both directions. These are represented in Fig. 2. Type A shows the trolley switches which operate the signalling mechanisms located at each end of the section between turnouts or double tracks. The signal boxes are set one pole stretch in advance of the trolley switches. This type requires two differentiating double-acting trolley switches per block. A condition sometimes happens that a car has a red or danger signal set against it just before it passes under a trolley switch, due to a car entering the block from the distant end. Under this condition the car could not be stopped before it had passed under the switch. It will, therefore, be neces- saiy that when the car backs out it must have its trolley pulled down, and BLOCK SIGNALLING. 623 coasts under the switch, otherwise it would restore the signal set by the car already in the block. Type B. — In this type the trolley switches are located on the double tracks or turnouts. These switches are single acting and will only set or restore the signal as arranged for. This type requires four switches per block, but has the advantage that a car can pass under the switch in the reverse direction without restoring the signal. It requires that the cars shall take the turnouts in one fixed direction. Type C represents a combination of Type A and Type B, and can be used to meet special conditions of road and travel. The Requirements of a jSig-nal System are as follows: Mechanical and electrical simplicity of all signal movements and appliances Must be automatic, non-interfering and interlocking; Must be incapable of wrong indications under any of the following men- tioned conditions, and must not permit restoring to normal except under normal conditions of operation, otherwise it could be set or reversed by another car entering the block. Loss of current on signal lines. Cross of signal lines. Ground on the setting signal lines or on the restoring signal lines. Cross with the trolley wire between the setting or restoring signal lines. If the signal is set in one direction and the line then opened, it must be incapable of being set from the other direction, i.e., the signal must be interlocking If a car should run under a trolley switch when the signal is set against it, it must not restore the signal, i.e., it must be non-interfering. It should employ as few wires as possible. It must be impossible to get two safety signals should cars operate the switches at each end simultaneously. In this case both signal move- ments would set, and it is desirable that they may be automatically restored by the car leaving the block without being required to be manually reset. The installation of an electric railroad signal requires at each end of a block which passes cars in both directions the following, with the necessary connections. A signal movement and a lighting and extinguishing switch. The wires required are these: A lighting switch wi^e from same to signal box. An extinguishing switch wire from same to signal box. The signal line wires. Generally a lighting and an extinguishing signal line wire running between the signal boxes at each end of the block. A ground connection between signal movement and rail. A permanent feed connection between signal movement and trolley. A lightning arrester should be attached to the permanent feed wire and one each to the signal line wires. It should be remembered that the trolley is connected to the ground when- ever the signal is set and thus a path of low resistance and inductance is provided for any lightning discharge which may take place on the trolley lines. The above is based upon the signal systems that are in practical operation to-day on trolley roads, and does not apply to systems as used upon elevated railroads. The latter are operated by track instruments and give only clear and danger indications. The manual system consists simply of a group of lamps at each end of a block, and a switch to light and extinguish the same. This system operates ina manner similar to the automatic system referred to in the first part of this article but requires the stoppage of the car to set the same or to restore the signal, and in practice it has been found that the signal has at times been tampered with by people who are able to reach the switches which are located on poles alongside the track. The trolley switches in use are of two types. One consists of a parallel way upon which the trolley runs and in so doing connects the two sides of the switch. One side is permanently connected to the trolley wire and the other to the signal movement. This switch will not differentiate in 624 ELECTRIC RAILWAYS. direction and must therefore be placed upon turnouts and not upon the main line. The other type is a mechanically operated switch which has a pendant lever hanging down and straddling the trolley wire. The trolley wheel strikes this and moves it in the direction in which the car is going. As the pendant arm is about four inches long it remains in contact with the trolley wheel only about one-fifth of a second for a car speed of a mile per hour and proportionally less for higher speeds. This requires that all switches have a retarding device to keep the contacts closed longer than would the trolley wheel. The most common switches to-day use a pallet and wheel escape- ment as retarding devices. Typical Automatic Two-JLine Wire. Non-Interfering* Block Signal. The following description of the Block Signal System made by the Uni Signal Co. of Boston, Mass., is illustrative of what such a signal must accom- plish. Fig. 3 shows the wiring for a complete block and Fig. 4 the detail wiring at each end of the block. The signal movement consists of iron back plate upon which are mounted three magnets known respectively as the lighting magnet, extinguishing magnet, and locking magnet. The first two mentioned are of 70 ohms resistance while the third is of 10 ohms resistance. The magnets are of the well known semaphore type. The lighting and extinguishing magnets have notched iron cores in which loosely play one arm of a switching lever. In the extinguishing magnet there is also an additional magnet core which when down closes a pair of contacts. The other two contacts are shown in Fig. 4 directly above the large magnets and are circular contact discs loosely mounted upon a rod between stops. These rods rest directly upon the magnet cores and are moved to open or close the contacts as the move- ment operates. The armature of the locking magnet is attached directly to the rod over the extinguishing magnet and is so adjusted that it is against its seat when that contact is made and the rod in its lowest position. The lamps are of 110 volts and one-half ampere and the resistance plate of 600 ohms is clearly shown. The operation of the signal is as follows, and can be seen by reference to Fig. 3. t . When a car enters a block it causes current to pass from the trolley wire through the lighting magnet and resistance plate to ground at that end. This causes the switch lever to be thrown over to the left hand contact, thus causing current to be taken from the leaving end of the block, passing through the red lamp, locking magnet at that end, and then through the lighting signal line to the entering end, where it traverses the green lamp and resistance plate to ground. To extinguish the signal, current is taken from the trolley at the leaving end of the block through the extinguishing magnet at that end, thence through extinguishing line to the entering end and through the extinguishing magnet at that end to ground through the resistance plate. It might appear at first sight that there would be current through both magnets at the enter- ing end, and under such condition impossible for the switch lever to be restored to its normal position. Examination, however, will show that as soon as current is established in the extinguishing circuit the gravity armatures, so called, at their lower end, are raised, and the one in the leaving end of the block cuts off the current of the lighting magnet in the entering box, thereby allowing the extinguishing magnet in that box to operate.^ By taking the permanent feed from the leaving end and also opening that circuit at that end, it will be apparent that grounds on the lighting line will not prevent the restoration of the signal. A cross between the signal lines will not restore the signal, but will extinguish the green signal, which will, however, relight as soon as the cross is removed. Grounds on either lines will not restore the signal when set. Ground over 1500 ohms resistance will not affect the operation of the signal even if on both signal lines at the same time. Tlri s . is equivalent to £ ampere leak while the normal current in the signal circuit is only \ ampere. Loss of current will not restore the signal BLOCK SIGNALLING. 625 when set and when the current is returned the signal will indicate the same as before. Should the lighting circuit be open after the signal is set, for instance by a lamp being burned out, and another car at distant end should enter the block, it will be seen by Fig. 3 that the switch lever in that signal move- ment at that end would be thrown over to the left-hand contact as in the box shown at the left hand, the result being that the permanent feed is cut off at both ends and no signal is obtained. Lack of green signal on entering is construed as a danger or cautionary signal. Suppose that a car should pass under the lighting switch at the red lamp end of a block, as represented by the movement at the right hand side of Fig. 3, it will be seen that current will be taken through the lighting magnet at that end and thence through the resistance plate to ground. This Trolley Wire, light ivg M 1 Y-£icr>/to(//shtni SnteringBtd ofB/ock Jig not Set Green LamfiL rghted. L ft Lighting L.A± ightntng Arrestor. j^ LeavingEndofB/ddf SignatSet 1La Cqmpletefflock Signal fo<* Lamp Lighted. Double AcringTrolley Switches Fig. 3. would tend to move the lever or switch arm over to the left hand contact, and thus put out the signal were it not for the locking magnet whose sole function is to prevent this movement. As soon as the lighting circuit has been established the locking magnet at the red end is energized and its core being against its seat at that time it is held there. To the core is attached a tail rod at the other end of which is one of the contact discs mentioned before. This tail rod pressing against the lever arm prevents the lighting magnet from operating it. It will be noted that the locking magnet is instantaneous as it has no moving part to operate before locking, and on account of its closed magnetic circuit is more powerful than the lighting magnet, whose armature is retracted at that time, and has a large air gap in circuit. The signal thus is made non-interfering. 626 ELECTRIC RAILWAYS. In Type B signal made by the same company, the wiring is the same except that the resistance plate is placed in the permanent feed, and two additional graphite resistance rods of 600 ohms are placed in each trolley switch leg. Each lamp is further protected by a paper shunt which closes the circuit when the lamp burns out. Furthermore there is a magnet opera- ting a red, and one operating a green semaphore disc signal, which are cut into the circuit adjacent to the red and green lamps. The trolley switches are double acting and differentiating, operating as EtiinguishinqSignalLirtt. ■Srr/ftS>. 2 /t/n/t,fi/4* Bex v F*$r/n orttnt Fits*. F~vs*2#mp$. X. rfMrnq firmafrr*. ] WIRING Fo^2 WIRE Automatic Block Signal Uni Signal Co Switch Csntocts. Fig. 4. Signal Set at Entering End of Block, Green Lamp Lighter. follows: The first blow of the trolley wheel hits a pendant hanging over the wire and brings the switch contacts into mechanical lock. At the same time it winds up a pallet escapement, which, when it runs down, kicks the lock off and allows the contact to open after a predetermined time. The working parts are in balance and made as light as consistent with strength. There are two contacts, but only one common escapement. The switch lights when passed under in one direction, and restores the signal when operated in the other direction. A time element is necessary, as it requires about \ second for the signal mechanism to operate. The power required to operate the signal switch is 2h pounds pull, while the tension on a trolley wheel to hold it against the trolley wire is over twenty pounds. BLOCK SIGNALLING. 627 [Distributed Signal Block System. (Developed by R. D. Slawson, Electrical Engineer of Easton Transit Co.) This is a manual system, and is used by the Easton Transit Company on the Easton, Palmer and Bethlehem division, and differs from others in having the signals distributed along the line between turnouts. There are two sets of signals, one being used for out-bound and one for return cars. The signal lamps .are enclosed in galvanized iron boxes, attached to poles along the line. Signal poles are also painted with two 12 inch bands of white, and a band of either red or green, as the case may be. Switches are located at each end of the turnouts on poles and the covers are marked "Throw on " Trolley Wire ~^~ Fiq. 5. Diagram of Connections of Slawson's Distributed Signal Block System for Single-Track Railways. and "Throw off," and each conductor is responsible for maintaining his own right of way. No. 14 insulated iron wire is used for the signal circuits. 16 c.p. 110 volt lamps are used for signals, and as the signal boxes are triangular, the lamp can be seen from almost any position. The red lamps are used for out-bound, and the green for return cars. The operation of the system is as follows: The conductor of a car leaving a terminal out-bound, first throws the switch marked "Throw on." This lights the five lamps in the red boxes in the section ahead of him, and he proceeds to the first turn- out, and, if there is no green lamp burning at that place, he throws off the red signals behind and sets the red lights in the section ahead. If a lamp should burn out while the car is running be- tween turnouts, warning of the fact is given by the absence of the red light, and by watching the green signals the motor- man can tell when a car is coming in the opposite direction. If the out-bound car, coming to a turnout, finds the red signal burning for the section ahead, showing that the section is occupied by a car going in the same direction, it must wait until the section is cleared by the car ahead. The signals may then be reset, and the car can proceed. Switch Used. Should a crew find that they are unable to light the red signals, they may use the reverse, or green signal, to the next turnout. On the return the green signals are used in the same manner as described above for the red signals and an out-bound car. If signal switch boxes are placed about a car's length outside of the ends of turnouts, cars will always approach at slow speed, which is quite desirable in running into a turnout. \ Slate Q 1 101 id Q Fig. 6. 628 ELECTRIC RAILWAYS. LIST OE MATERIAL REQUIRED EOR O^E Tf ITE OF OVERHEAD LI1¥E FOR ELECTRIC STREET RAILWAY. 1 Mile Overhead. Curve Overhead Material. Anchor- Material for Railway Construction. Cross Suspen- sion. Bracket Suspen- sion. Main Line. Branch Line. © O H age. H © O u H © 02 H © o ft u H © 13) .9 u H © o H S3 •J H © O ft © g H O No. B. & S. H. D. Trolley Ft. Lb. 5280 1685 10560 3369 5280 1685 10560 3369 250 80 z No. OB.&S. S.D. F'd'rT'ps Ft. Lb. 400 154 500 192 90 35 180 69 7 strand No. 12 span Ft. Lb. 3600 756 3600 756 800 168 800 168 800 168 800 168 200 42 400 84 600 122 7 strand No. 15 guy Ft. Lb. 3000 300 4500 450 1500 150 2000 200 100 10 100 10 100 10 100 10 Plain ears .... Strain ears .... Splicing ears . . . Feeder ears .... 45 1 10 90 2 20 45 1 10 90 2 20 5 2 10 4 5 1 15 2 4 Insulating caps . . Insulating cones . . 45 45 90 90 45 45 90 90 7 7 4 4 6 6 17 17 4 4 g| Straight line . . «,§ Single curve . . £-3 Double curve . HI Bracket . , . 45 90 45 90 3 4 3 11 3 3 5 12 4 Sti Tu Se Ft Fr Ha Ey Ca Ga Cr< ain insulators . . rnbuckles . . . ition insulators 3gS 90 90 2 45 90 45 90 90 4 45 90 45 2 45 45 48 4 90 90 48 4 4 2 4 4 2 2 2 1 2 2 2 2 1 2 2 2 2 1 2 2 2 2 og crossings . . . irdwood pins . . e bolts 2 st-iron brackets . s-pipe arms . . . >ss arms (1|"-18) . Cr< Bo La e La a La )ss-arm braces J"X8") Its for brackets &"X4") 90 45 144 90 45 144 45 45 90 90 y screws for brack- ts ($"x7") • • • g screws for cross rms (§"x3") . . % screws for braces Poles, 125-ft. apart . 90 90 45 45 2 2 2 2 2 2 Bonds 400 800 400 800 • Lightning arresters . 3 3 3 3 Section switch boxes 2 2 2 2 1 STANDARD IRON OR STEEL TUBULAR POLES. 629 ESTIMATE OE COST TO PRODUCE OAE II B LE OF DOUBLE TltACK OVERHEAD TROLLEY CONSTRUCTION EOR CTT1 STREETS. (Report of Bion J. Arnold, November, 1902.) 100 Iron poles, set in concrete, at $28 $2,800.00 50 4-pin iron cross arms, with pins and ins., at $3.95 . . . 197.50 100 Small Brooklyn insulators for spans, at 50c 50.00 100 Globe strain insulators for spans, at 22c 22.00 90 Straight line hangers, at 32ic 29.25 10 Feed-in hangers, at 50c 5.00 140 Soldered 9-inch ears, at 16c 22.40 12 Live cross-overs (estimated), at $3 36.00 8 Insulated cross-overs (estimated), at $6 . 48.00 8 2- way frogs (estimated), at $3 24.00 3000 Feet 5-16 inch galv. strand wire for spans, at $10 per M. 30.00 6 Strain plates (strain layout), at 32c 1.92 12 Small Brooklyn (strain layout) at 50c. ...... 6.00 12 Globe insulators (strain layout) at 22c 2.64 1500 Feet i-inch galv. strand wire (strain layout), at $7 . 25 per M . 10 . 88 20 Double hangers (2 double curve layouts), at 44c. ... 8.80 20 Single hangers (2 double curve layouts), at 35c. ... 7.00 1000 Feet ^-inch strand wire (2 double curve layouts), at $.725 per M. 7.25 4 Heavy Brooklyn (2 double curve layouts), at 70c. . . 2.80 10560 Feet 2-0 trolley wire, 4246 pounds, at 13ic 562.59 2 00 splicing ears, at 50c 1 . 00 Labor, placing spans, trolleys, etc 225.00 Total cost exclusive of feeder wire $4,100.03 Cost of feeder wire estimated average per mile 4,000.00 $8,100.03 STANDARD IRON OR STEEL TURULAR POLES. Tubular poles for electric railway lines are made up of the regular pipe sections, both standard and extra heavy. The combinations in common use are : Pole made of standard tubing. Pole made of extra heavy tubing. Pole made with bottom section of extra heavy tubing, and other sections of standard weight. Pole made with bottom and middle sections of extra heavy tubing, other sections of standard weight. Standard lengths are 28 feet end to end for side or line poles, and 30 feet for corner or strain poles. The standard joint insertion is 18 inches, and total weights can be calculated from regular standard pipe list (see pages 1426-1427). Two section poles are most commonly made up of 6 and 5 and 7 and 6 inch-pipe, for side or line poles; and 8 and 7 inch pipe for corner or strain poles. Three section poles are 6 and 5 and 4 or 7 and 6 and 5-inch pipe for side or line and 8 and 7 and 6-inch for corner and strain poles. 630 ELECTRIC RAILWAYS. Standard Pole JLine Construction. For most urban and all interurban or suburban lines, wooden poles are used, and are either octagon or shaved. The following cuts show common standards of dimensions and arrangments of cross arms, brackets, etc. % x 7 Lag Screws Fig. 7. Standard Pole Line Construction of the Union Traction Company of Indiana. DOUBLE TRACK CENTER POLE CONSTRUCTION. 631 Double Track Center Pole Construction. Electric roads use a greater distance between track centers than do steam roads hence permitting center pole construction, with less cost per mile than would be the case if double pole bracket or cross suspension construc- tion were used, although the latter is often preferred. Barb Wire Mach.Bolt^ '_ Tfl"7£- IJfi Top Grove Insulator . — .I.BOHfif.x"' 6 "S. %"x MvJV "it* "IsProvo Glass Insulator 3/A iW P^12^=18^i8:fe^27fe Special Insulator Pins 78 _ „.??.— !e=^V U V ± ^J Q • , «__«_„ a. Carriage Bolts' %xl2Machine^olt y Special Pine Cross Arm 6x4% x'6% Galv. Iron Braces 24x1 }^'x % S(«cial Pine Cross Ana ,Jr?e Seconds Fig. 27. est potentials used or contemplated, also incidentally affords marked me- chanical improvement which is important with the high speeds of modern suburban and interurban operation, and steam railroad electrification. The catenary system which is equally applicable to bracket or cross span construction, consists essentially of an arrangement of a slack messenger CATENARY TROLLEY CONSTRUCTION. 641 cable and suitable hangers so distributed as to maintain the trolley wire practically without sag between suspension points, or to limit the sag as may be necessary for various conditions of operation. The blow of a collector passing suspension points at high speed is thus greatly reduced. The shorter distance between hangers necessitates less stress in the trolley wire and reduces danger of break in the line. The catenary system, therefore, offers the mechanical advantages of a longer pole spacing and a natter trolley wire, and a flexibility in the line which obviates the hammer blow of the collector at suspension points, and reduces danger of mechanical breakage. The three-point suspension in which, with 150 ft. pole spacing, the 5 re 9 to jD$f/ec6/o/7 /r?c/7e 01 00 OJ bf, 8 3j Ifl s fc ^ ©3.2 £ 3 o © 2 © be P w bC a a> 1? -M I> H« l© t- © bo c a X <# s © > 'boa 53 <>> 648 ELECTRIC RAILWAYS. "Where three or more tracks are equipped as on the New York, New Haven & Hartford Railroad, the trolley wire is generally supported from two catenary cables, which are carried on steel bridges, placed 300 feet apart. Heavier bridges are used at intervals to anchor the system, and views of one of these anchor bridges are shown in Figs. 34, 35, and 36." BRIDGE SUPPORT FOR TROLLEY. 649 r Fig. 35. End View of Bridge for Supporting Catenary Hung Trolley, N.Y..N.H. &H.R.R. 050 ELECTRIC RAILWAYS. 9UJ7 A\n?!|ixrtV' « « Fig. 36. Plan View of Bridge for Supporting Catenary Hung Trolley, NY., N.H. &H. R.R. CATENARY CONSTRUCTION. 651 Fig. 37. Detail of Catenary Construction, Spendersfelds Line Fig. 38. T-Iron Bracket with Main Insulator and Steady Strain. The future development of the A. C. motor is in no way handicapped by the ability of the trolley construction to withstand high potential, as A. C. trolleys have been worked successfully at 10,000 volts and 15,000 volts. 652 ELECTRIC RAILWAYS. E\ERG1 COHSIJMPTIOI. Power Curves. — For convenience in quickly ascertaining the horse- power required to propel a car of known weight under known conditions of speed and grade, the curves shown below have been calculated. The left-hand portion of the lower horizontal line represents the speed in miles per hour; the right-hand portion of same line, the h.p. per car; the oblique lines in left-hand side of cut, the per cent grade as marked on each line; the oblique lines on right-hand side of cut, the weight of car as marked; while the vertical line in center of cut represents the h.p. per ton. This curve is based upon a flat friction rate of 20 lbs. per ton (2000 lbs.) for all speeds and weight of cars, and is approximate only. V \ \ \ Q N \ \ tt "S^ \ \ \ D \ \ < ^ > ^_ \ Ul ^ \ Q. CO N \ b. O V n o ** ^ C i\ a H ""^^ s, i> J i\ i\ Z *)! '3\ A Y C \ V \ & o % °>> « ^s X °6^ \ \ 1 u • t \ o ** \ \ lll l °>y^ , \ z X ^. \ \ 00 dp < o A/Oj^ ? ^~-~~-. III »- < cc NOi. U3d "d'H fcf "* ci5 CM T- O C > « r- .o * c i J^^^/7 o ^ Wfl D VY 7/1 cc / 1 a ^y y /, / / > cc 5-^^ y^ , f cc o n£ _<— "^ / 1 h o _Uo^2 *^S>' *^ ^ /^ y / 1 5 s kri^r. ^ ^ uy / / 1 o ^■"T^^r*^ \ 7 / j 111 ^ ^* ^ *> / a CO" Ul o =^ a. --^1" y' S h -A / I-' ^C *' k*/ \ J / 2 < ^ y V^ / / X o ^ T y u r / 1 / ( t ^1 y / / / < _^/f. yV / f / _ / O / f / / / ! 3 / / Fig. 39. HORSE-POWER OF TRACTIOX. 653 $ SR fcg ^ 888 BBi OS rHTf< © OO r © CO 00 (Nt>^ t~ CN ,_,,_,,_, rH CM CM CM CM CO CO CO CO * * * IfllflO Ot» SCO CO O CO CO O CO CO O CO CO O CO CO O CO CO O CO CO h>0 © 00 OS © CM CO * © L- COOr- cm* © os cm * — 5 00 O CM * t- OS rH CO IC b- OS CM * CO 00 CM CO i-i CO £8 HCM CMCMCM CMCOCO COCO'* * lO © lO © 5CDO CO ■5 00 00 l> m-> t 5 r-tCM CO LO t- CO CO O CO CO © CO CO O CO COCO CO -" COCMO © 00 © £ * CM hOM t- • rH CO *©00 ©CM* ©00© rH O CO CO O CO COO CO CO OOOiC CM O CO b- i-H rt< 00 CM CO CMCMCM CMCMCM COCOCO * * * lOO © CM 00 * © © CM00O CM * © GOO E^ OS © CM * lO l> 00 Oj lO 00 rH * CO CM CM* CO CO CO * *y %H OO OQO OOO OOO OOO ooo ooo ©oo ©o _ CM O0 ^OC0 ©CM* ©00© CM*© CC © CM * © © * 00 CM © © ©i-JrH CM CO * CO t> 00 OS © CM CO * O CO 00 OS Cr|rJH © 00 rH CO© HH HHH HHH CNC^ji (NN« MM CO©© CO©© CO©© CO© °2 52 S5 53S3S £2 3 i2 coo© co©© co©© co © co ©©co © © ooo j3£2£2 c 2^'* tfs © © t- oo oo r^ © © himcq * © t- oo© OHH CM CO * O CO t>- oq OS © rH CM CO * CO t>; 0C OS i-H CO lO t- OS CM H Ht-Ih ririrH ^rt« Nfid CN M © CO © © © CO © © CO © © CO © S2i22 £ s £2 SSS^S «2oco ©©co © © co ©©© co©© co© * os * oooot- oo co lq * * co cm cm r- © © © xxo 1.0 * cm ©© ©©-H rH CM CO * O © t^OOOS © rH C* CO** LO © 00 © CM * ©00 HHr! Hrtr! rH rH rH CM* CM CM* CM CM © © © © © © © © © © © © © © ©©© ©O© ©O© ©©© ©©© ©©© ©©© ©©© ©© *00CM ^*£ ©COCO *CM© 00©* CN©0C ©*© © CM O0 *© ©©r-J rH CM CO * * iq ©t-oq 00©© rH CM CM CO*© t- © © CM* ' ■ ' "rt rtHH rtrlrf rlrHC* «C* CO © © CO © © CO ©© CO © © CO © © CO CO©© CO©© CO©© CO©© CO©© CO©© CO © CO ©©CO ©© CO©© CO©© CO©© CO©© CO©© co © © co © co ©©CO © © ©> ©? r-J rH CM CM CO * * lO © © t> 00 00 OS © © rH CM CO * © C-; 00 © "rHrH rHrHrH rHrHrH rH CM ©CO© ©©CO ©©CO ©©CO ©©CO ©©CO ©© ~© ©CO© ©©CO ©©CO ©©CO ©©CO ©©CO ©©© CO©© CO© CM lO 00 ©©rH © CM t- CMOOCO O0*© * © LO ©©© t^OOOO © © © © © rnrHCM CMCOCO * * LO O © © l> 00 00 OS OS © rH CM CO * © 888 888 888 BBB 888 888 888 88© ©© CM*© O0CM© ©rfOO CM©© * 00 CM CO©* 00 CM © 00©* CM© © © © © rH rH CM ?1 CM CO CO * * * lO lO CO © © t> 00 00 OS © rH CM CO©© COO© CO©© CO©© CO©© CO©© CO © CO ©©CO © o CO©© CO©© COO© CO©© CO©© CO©© COCOCO CO © CO ©© Si! 88°, S3 §53 si^ $%% fel§ 3S8 S$8 ss MB 888 888 888 888 888 888 888 88 BBS §3S S%8 SS3 385 S8S? 3$li 883 co§ rHrHCM CMCOCO * * LO lO © © t-t^O0 00 © © rHCMCO * lO Wn , H. p. — -^_ (ir+2000 sin 0). W 37o Load in tons. n = Speed in miles per hour, = Wn X .0026| (K + 2000 sin 0). K— Resistance in lbs. per ton. K'—^q Hz= Constants of power required to move ONE ton on level at speeds in table with K— 10. E'— Constants of additional power required to raise one ton on grader and at speeds given. Hx WK f — T&. P. required on levels alone for speeds given. i? 7 X JF = H. P. additional on grades alone for speeds and % given. W(K'H± BO = total H. P. required. Example: Given a motor car, total weight 9 tons, to ascend a 7 per cent grade at a speed of six miles per hour. What is the estimated horse- power required, with K= 30 lbs. ? 654 ELECTRIC RAILWAYS. 30 H for 6 miles per hour is . 16, which, multiplied by 9 Xtq, = 4.32 h.p., in overcoming the track resistances alone. H' = 2.240, which, multiplied by 9, = 20.16. The sum of the two will give the total theoretical, i.e., 24.48 h.p. required. Allowing 50 per cent as the combined efficiency of motors and gearing, to operate this car would require a draft of 48.96 h. p. upon the line. HOR§£-POW£R OF TRACTIOI, (Davis.) c3 Speed in Miles per Hour. T3 c3 o 4 6 8 10 12 15 20 25 30 35 40 50 60 3 % Horse-Power Required to Propel One Ton at Various Speeds up £ Various Grades. .32 .48 .64 .80 .96 1.20 1.60 2.00 2.40 2.80 3.20 4.00 4.80 i .53 .80 1.07 1.33 1.60 2.00 2.66 3.33 4.00 4.66 2 .74 1.12 1.49 1.87 2.24 2.80 3.63 4.66 5.60 3 .93 1.44 1.92 2.40 2.88 3.60 4.80 6.00 4 1.17 1.76 2.34 2.93 3.52 4.40 5.47 5 1.39 2.08 2.77 3.46 4.16 5.20 6 1.60 2.40 3.20 4.00 4.80 7 1.86 2.72 3.62 4.53 8 2.02 3.04 4.05 9 2.24 3.36 4.48 10 2.47 3.68 4.90 11 2.67 4.00 12 2.88 4.32 13 3.09 14 3.29 15 3.52 Note No. 1. — The h.p. required to propel a car equals the total weight of car plus its load (in tons) multiplied by the h.p. in table corresponding to assumed grade and speed. STRKKI RAILWAY, Tractive Force, F. E. Idell, M. E. On Good Track. - To start car 116 lbs. per ton. To keep in motion at 6 miles per hr. 15.6 lbs. per ton. On Bad Track. — To start car 135 lbs. per ton. To keep in motion 32 lbs. per ton. On Curves. — To start car from to 6 miles per hour . 284 lbs. per ton. average, 264 feet per minute. TRACTION. 655 TRACTION. (Davis.) Load of Trailer Cars in Tons which a Motor Per cent Grade. Tractive Force in Pounds per Ton. Car of one Ton will Haul. Snowy Kail. Wet Rail. Dry Rail. 30 8.50 12.33 16.00 1 50 4.70 7.00 9.00 2 70 3.07 4.21 6.14 3 90 2.17 3.44 4.55 4 110 1.60 2.63 3.54 5 130 1.19 2.07 2.84 6 150 0.90 1.66 2.33 7 170 0.70 1.35 2.00 8 190 0.50 1.10 1.63 9 210 0.35 0.90 1.38 10 230 0.24 0.74 1.17 11 250 0.14 0.60 1.00 12 270 0.05 0.48 0.85 13 290 Wheels slip. 0.38 0.77 14 310 0.30 0.61 15 330 0.21 0.51 16 350 0.14 0.43 17 370 0.08 0.35 18 390 0.02 0.28 19 410 Wheels slip. 0.22 20 430 0.16 21 450 0.11 22 470 0.06 23 490 Wheels slip. Note No. 1 . — Multiply figures in table by weight of motor car (in tons) to get weight of trailer (in tons) that said motor car will haul up corre- sponding grades. HEVOLlIIO\S PER ]HII¥UTE OF YARIOVi §IZED WHEEL! TO MAKE VARIOUS SPEERS. Miles per Hour. Diameter 2 4 6 8 10 15 20 25 30 40 of Wheel. Feet per Minute. 176 352 528 704 880 1320 1760 2200 2640 3520 24 in. 28 56 84 112 140 210 280 350 420 560 26 in. 26 52 78 103 129 194 258 323 388 517 28 in. 24 48 72 96 120 180 240 300 360 480 30 in. 22 45 67 90 112 168 224 280 336 448 33 in. 20 41 61 82 102 153 204 255 306 408 36 in. 19 37 56 75 93 140 187 234 280 374 42 in. 16 32 48 64 80 120 160 200 240 320 656 ELECTRIC RAILWAYS. POWER KiailREI) FOR BOiBLE AND ftl'XGJLi: truck: cars. Wattmeter placed on car. (McCulloch.) DQ o tJO e3 u m u * TRAIN PERFORMANCE DIAGRAMS. 665 Grades. — A grade of one per cent means a change in altitude of one foot for each 100 feet of track on the grade, and this is equivalent to a tractive force of 20 pounds per ton, which will be positive, or to be added to the tractive effort per ton, if the train is going up grade; or to be deducted from the same if the train is on down grade. Then if g = grade per cent X 20 the formula becomes T = (91 . 1 aw) 4- (t ± g) W and T - (t ± g) W a ~ 91.1 w. Curves. — Values of railway curves are expressed in terms of the central angle subtended by a chord 100 feet long; thus a one degree curve means one such that the angle at the center end of the radius will be one degree, or a radius of 5730 feet, thus de e = 5730 # radius in feet Experiment shows that the effect of curves is to introduce a resistance of about .6 pound per ton per degree of curve; thus a two degree curve will require a tractive effort of 1 . 2 pounds per ton of train to overcome the resistance. If c = tractive effort of a curve at d, degrees, the formula will become T = (91.1 aW) + (t+cyW; T - (t+c)W a ~ Ql.lu;. A combination of a grade and a curve will make the formula: T = (91 . 1 aW) + (* + c ± g) W; = T - (t + c ± g) W a 91.1 w The use of the polar planimeter will very much facilitate the construction of these diagrams. The method of constructing the speed-time curve as described below is about as simple as can be made and was used by Mr. H. N. Latey in laying out the work of the Interborough Company in New York. For purpose of explanation the following example of train performance is given: Example 1. for Train Performance Diagram. Train 3 motor cars, 2 trail cars. Schedule 20 miles per hour. Stops 2 per mile. Acceleration a = 1.25 m.p.h. per second. Braking b = 1.5 m.p.h. per second.^ Tractive effort t = 13 pounds per ton of train. Fly-wheel effect 10 per cent of train weight. Motors 4 for each motor car. Motor cars weigh 60,000 pounds each. Trail cars weigh 40,000 pounds each. Weight of train W = 130 tons. Fly-wheel effect =13 tons Total =143 tons. Weight on drivers all motor cars = 180,000 pounds. Tractive effort due to weight on drivers 18 % = 32,400 lbs. Tractive effort, T = (a X w X 91 . 1) + tW, or, T = (1.25 X 143 X 91.1) +13 X 130 = 17,974 lbs. and T per motor = 17974 -i- 12 = 1498 lbs. From motor curve, Fig. 41, 1498 lbs. = 20 miles per hour at a = 1.25. 20 miles per hour at 1.25 miles per hour per second is the first point p on curve 666 ELECTRIC RAILWAYS. Other points, P\, P2, P3» etc., are determined by the formula, T — t W a = , where T is taken from the motor curve at the miles per hour 91.1 w the train is moving. Then, let T — tW = B, and a = , from which the following table may be constructed for the diagrams: Table I. M.P.H. T. Motors T. tW B a at 20 = 1498 lbs. X 12 = 17974 - 1690 = 44 22 = 1100 " X 12 = 13200 - 1690 = " 24 = 840 44 X 12 = 10080 - 1690 = " 26 = 700 44 X 12 = 8400 - 1690 = " 28 = 590 " X 12 = 7080 - 1690 = " 30 = 500 " X 12 = 6000 - 1690 = " 32 = 420 " X 12 = 5040 - 1690 = 16284 = 1.250 11510 = .840 8390 = .644 6710 = .515 5390 = .414 4310 = .331 3053 = .257 Coasting after shutting off current = tW 91.1 w 13X130 1690 91.1X143= 13027= ' 129 mph ' per second. Table II. Amperes per motor. Amperes per train at 20 m.p.h. = 134 X 12 = 22 " = 108 X 12 = 24 44 = 93 X 12 = 26 " = 83 X 12 = 28 44 = 75 X 12 = 30 " = 68 X 12 = 32 44 = 62 X 12 = 1608 1296 1116 996 900 816 744 Construction of Speed-Time Curve. — An inspection of Fig. 42 will show that the speed- time curve is divided into four parts: (a) the acceleration due to starting the motors and bringing the train up to the speed that will be given by cutting out all resistance, and leaving them in multiple connection. This is shown on the diagram by o.P. (b) the acceleration in multiple, running from P to s; (c) at which point the cur- rent is cut off and the train allowed to coast for the distance indicated between s and n; and (d), where brakes are applied, and from n to g the curve is diagonally downward, assuming that the train retards at a regular rate, which obviously is never the case, but is near enough so to be indi- cated by the straight line as shown. Referring to Fig. 42: The straight part of the curve, from o to p, is laid on the drawing at an angle determined by the rate of acceleration, which in this case is 1.25 miles per hour per second. The example shows that at this rate of acceleration and for the weight of train given, and at a tractive effort of thirteen pounds per ton, a total tractive effort per motor of 1498 pounds will be necessary, and by reference to the curve of tractive effort in Fig. 41, it is found that 1498 pounds correspond to a speed of twenty miles per hour, which becomes the first point P on the acceleration curve. At this point the resistance of the controlling devices is all cut out and the motors are in multiple from this point on, to the point s. When the current is cut off for coasting, the speed will be accelerated at a gradually decreasing rate as shown. The lines between the points p, p 1t vi Va and p 4 represent the average rate of acceleration for speeds of 22-24-26-28 and 30 miles per hour, and in each case start from a point half way between the lines which TRAIN PERFORMANCE DIAGRAMS. 667 ni?!igT(j ^89J •} •el' i\r i * Po ■♦j. br X 1 •+J aJ p$ 1 ■ ^ i "* \ / J \ V- \ \ \ C7 V 1 $ ° \ %\ / fe* ft *4 l -N b — c s, \ *1 \ \ / ) A o / \ •«*< / '\ pn k o / A > \ . jA J ^ +?' \ a and D = d + d x + d 2 + d 3 + d 4 , etc., etc. If the speed-time curve is very irregular it is more convenient to use a polar-planimeter in getting the average rate of speed, but in cases like that shown in Fig. 42, where the sections of the curve are drawn in straight lines, the average rate of speed will be at the center point of each section, and the time interval t is the time space covered between the ends of the section. For instance, to locate the first point on the distance-time curve at t, the average speed for the time interval of 10 seconds is 12.5 -5- 2 = 6.25, then 6.25 X 10 X 1.467 = 91 feet and this value laid off on the sheet over the time 10 seconds, and at a value of 91 feet on the scale of " distance feet" shown at the right, gives the point t. The average speed on the speed-time curve between 12.5 miles per hour and 21 miles per hour, is 16.75 miles per hour for the time interval t, between the two points shown, of 6.5 seconds; then 16.75 X 6.5 X 1-467 = 159, and X) = 91, + 159 = 250, or the point t } on the distance-time curve. Again DISTANCE-TIME CURVE. 669 the average speed between the next two points p and p t is 22 miles per hour, and the time interval is 2.5 seconds, thus, 22 X 2.5 X 1.467 = 80 and D = 250 + 80 = 330, which is the location of point fo. The above described process is repeated to obtain each point on the curve. Table III has been constructed in this way in order to show the progressive value of D. Great care should be exercised in plotting both speed-time and distance- time curves as errors of location are cumulative, and when many points are used the error at the end may throw the result quite out of line. Table MI. — Data For Distance-Time Curve. Total v = t = Time Interval. Total 1.467v* Distance Point Average Time = in feet Numbers. Speed in from Distance from M.P.H. Start. Intervals. Starting Point. 1 6.25 10 10 91 91.0 2 16.75 6.5 16.5 159 250 3 22 2.5 19 80 330 4 24 3.0 22 105 435 5 26 3.50 25.5 133 568 6 28 4.75 30.25 195 763 7 29.7 5.25 35.5 228 991 8 30 4.5 40 197 1188 9 29.5 5 45 215 1403 10 28.7 5 50 210 1613 11 28 5 55 204 1817 12 27.5 5 60 200 2017 13 26.7 5 65 195 2212 14 26.2 3 68 113 2325 15 22 5 73 158 2483 16 14.2 5 78 103 2586 17 6.7 5 83 48 2634 18 1.5 2 85 4 2638 Current Curve. — From the speed curve on Fig. 41, the current, taken at a speed of 20 miles per hour, is found to be 134 amperes, which for 12 motors will be 1608 amperes for the train. Point c is thus located, and the current taken with motors in multiple is twice that required for series running, which locates point d. At 22 miles per hour the curve shows that the motor will require 108 amperes, or 1296 for the 12 motors, which locates point c. Table II gives the location of all the points on the current curve, having been made up from the curves on Fig. 41. Voltagre Curve. — It is only possible to plot this curve from actual test, though in estimating, it is common practice to assume an average voltage in order to work out the power curve. rower or Kilowatt Curve. — This curve is plotted from a combination of the current curve and the voltage curve, the instantaneous values of each being multiplied to obtain the value of the power at the point taken. For simplicity neither of the last two curves are plotted here. In practice the kilowatt ourve is ordinarily plotted by using the average line potential together with the current curve. Example ]¥o. M. — This run is of the same length as that in Example No. I, i.e. one half mile, but instead of being all straight and level track, includes several grades and curves with a portion of track which is straight and level. At the right of Fig. 43 is shown the profile and contour of the line giving the length of each change, and opposite each section will be found the tractive effort per ton necessary to overcome the various condi- tions, thus: it requires 13 pounds per ton to overcome the train resistance on straight and level track; grades require an additional 20 pounds per ton for each per cent of change, and the values are shown in column g. In the 670 ELECTRIC RAILWAYS. SUOX 08T DISTANCE-TIME CURVE. 671 third column are shown the various efforts per ton necessary to overcome the resistance of the curves, at the rate of . 6 pounds per ton per degree. The fourth column shows the combined values of all the tractive efforts for each division of the run, and in the last column are given the total tractive effort for the train of 130 tons weight. Table IV. — Data for Speed-Time Curve, Fig-. 43. T. T. M.P.H. Per No. for Motor. Motors. Train. 20 1498 12 17974 21 1300 15600 22 1100 13200 23 960 11520 24 870 10440 25 760 9120 26 700 8400 27 640 7680 28 580 6960 28.5 560 6720 28.7 550 6600 29.7 500 6000 Coast 29 540 6480 6110 Braking Coast Braking tw. B. B a 13027 - 1690 16284 1.250 - 1690 13910 1.068 - 1690 11510 .883 - 1690 9830 .754 - 1690 8750 .672 - 1690 7430 .570 - 1690 6710 .515 - 1846 5834 .448 - 1846 . 5114 .393 - 1846 4874 .375 - 4290 2310 .177 - 1924 4076 .313 - 1924 1924 - .148 - 6890 - 410 - .032 - 1690 + 4420 + .340 - 1690 - 1690 - .130 - 2.05 - 1690 - 1690 - .130 - 1.5 Table V. — Data for Distance-Time Curve, Tig-. 43. Point Numbers. v. t. Total Time from Start. 1.467* Distance Intervals. Total Distance from Starting Point. 1 6.25 10.0 10.0 91 91 2 16.50 6.5 16.5 157 248 3 21.00 1.5 18.0 46 294 4 23.00 2.5 20.5 84 378 5 25.00 3.5 24.0 128 506 6 26.50 2.25 26.25 87 593 7 27.50 2.25 28.50 91 684 8 28.13 0.75 29.25 21 705 9 28.60 4.65 33.90 195 900 10 29.70 1.60 35.50 69 1969 11 28.70 5.50 41.00 238 1207 12 29.00 4.75 45.75 200 1407 13 28.70 7.15 52.90 300 1707 14 29.50 7.85 60.75 340 2047 15 27.50 2.5 63.25 100 2147 16 23.70 7.0 70.25 240 2387 17 18.75 5.0 75.25 148 2527 18 11.25 5.0 80.25 83.5 2610 19 3.75 5.0 85.25 28.5 2638 672 ELECTRIC RAILWAYS. The speed-time curve on Fig. 43 is worked out in the same manner as that on Fig. 42, except that while the speed-time curve in Fig. 42 may be plotted without reference to the distance- time curve, in the case of Fig. 43, they both must be plotted together, as care must be taken that the speed-time curve is not carried beyond the point where the tractive effort, and, therefore, the acceleration changes, as at T, T it T it etc. Table VI. — Current Data for Wig. 43. M.P.H. Amps, per Motor. Amps, for Train. 12 Motors. 20 134 1608 22 108 1296 24 93 1116 26 83 996 28 75 900 30 68 816 29 71 852 Tables IV and V are made up as the plotting progresses, and in the former give the values of a at which to lay the speed-time curve, and in the latter show the distance D and the time t u being respectively the distance and time from the starting point o. It requires considerably more care to work out one of these irregular curves for, while the method here explained is probably as short and as simple as any, yet it requires much cut-and-try to make the sections of the two curves fit for time and distance, and the location of the point s, at which current is cut off and coasting begins, requires experience and judgment, in order that the total area of the speed-time curve o, p, s, n, g, may equal that of the schedule o, m, x, y. Both the previous examples have dealt with short runs where the motors are never left in circuit long enough to reach their speed and current limit. In case of long runs as on suburban lines, current is left on in full, and the train is accelerated until the values of T = tW, and B is therefore zero and there is neither acceleration or deceleration, the train moving forward at a level rate^ of speed, as the tractive effort is just enough to overcome the whole train resistance. The values of T and tW will then only be varied by grades and curves, and the prolongation of the acceleration curve will have to be plotted to the point when coasting can begin in order to complete the time schedule. Of course if the track is straight and level, after T = tW, the speed-time curve will be straight and level to the coasting point s, and the current curve also will have reached a constant value and its curve will be a straight line until cut off for coasting. Curves must be plotted for each run, then motors best adapted for all purposes can be selected and the amount of power needed and the best equipment for producing the same can be determined. After all points have been carefully considered, due attention must be given to future needs, and great care be taken that the equipment has not been worked up to s fine a point that no allowances have been made for the idiosyncrasies of the motorman who, in many cases, will entirely undo all the results of fine calculation. Curves like that in Example II are seldom calculated as rolling-stock; being operated in both directions, grades practically neutralize each other, so that a curve like that in Example I for straight and level track is quite accurate enough for all practical purposes. RATING THE CAPACITY OF RAILWAY MOTORS. 673 IMTiVG THE CAPACITY OF R1IIWAT MOTORS FROM PERFORMANCE CURVE*. The limiting condition in rating the capacity of a railway motor is the heat developed in its use. "When a motor is carrying any load, certain copper and iron losses take place in it, which depend upon the load. It is these losses, which appear as heat, that tend to raise the temperature of the windings. Thus a loss of three watts (neglecting radiation) will raise the temperature of one pound —el- -1400 -1300 -1200 -1100 -1000 -900 -800 1 / WESTING HOUSE No. A RAILWAY MOTOR 500 Volts Iron Loss Curves VI ill / //// / r $/// &/// ^AAAA -600 'WAV i/\/%/ v //'a/ / p -800 -200 -100 5 1( K) 15 2C 10 21 Volts 300 3£ 4( 41 « X) 51 )0 Fig. 44. of copper approximately 1° C. per minute, or of one pound of iron approxi- mately .8° C. per minute. The copper loss depends upon the current only, and is proportional to its square, but the iron, or core loss, depends upon both the current and the voltage and does not follow any simple law. The iron loss in the motors in question, when carrying any given current at any given voltage, is shown in Figs. 44 and 45. Its dependence on both current and voltage may be seen in Fig. 44, from the fact that 20 amperes at 500 volts produces the same loss as 105 amperes at 305 volts. Owing to the great mass of metal in its frame, a motor has a considerable amount of heat-storage capacity. Instead of only a few hundred pounds of copper in the windings to be acted on, the temperature of the frame must also be raised ; when cooling, the entire mass must cool off simultaneously- 674 ELECTRIC RAILWAYS. That is, when the temperature of the windings is rising, that of the frame must also rise, and similarly when falling. The actual temperatures of the different parts may, of course, be widely different. Owing to this action, the temperature of the windings of the motor does not fluctuate in accord- ance with the instantaneous losses but rises at a fairly uniform rate depending on their average value. The important factor as regards the effect of the service loads on the motors, provided that the maximum loads are within the proper limits, is thus the average value of the losses, averaged, of course, over the entire time of the cycle. It is evident that the average copper loss in any case is J2 / £ WESTINGHOUSE No. B RAILWAY MOTOR 500 Volts Iron Loss Curves / -5500 -5000 -4500 -4000 -3500 ■3000 -2500 -2000 -1500 -1000 -500 / A. / / / /\f/ fs k s* '/A 5^ P^ 50 100 J« 200 250 300 350 400 450 500 550 U0 ) 1 Volts I I I I I Fig. 45. equal to that which would be produced by the continuous application of a current equal in value to the root mean square of the service currents. Thus, if this current and voltage is applied to the motor for the entire cycle, the average losses in the motors — both copper loss and iron loss — will have the same value and the same distribution as the losses due to the service loads. This voltage may be called the "equivalent" voltage of the service. This method of equating the service loads on a railway motor to simple and intelligible terms was devised by Mr. N. W. Storer, of Pittsburg, and gives a convenient way of expressing the service capacity of railway motors in a usable manner. The limiting capacity of any type of motor may be readily expressed by the manufacturer in terms of the current (root mean square) which it will carry continuously at various voltages (equivalent voltage) with a safe rise in temperature. In choosing a motor for a given service, the root mean RATING THE CAPACITY OF RAILWAY MOTORS. 675 square current and equivalent voltage can be calculated from the speed- time curves and a comparison of these results with the values allowable for the motor in question will determine its fitness. Where motors are already installed, the continuous equivalent of the service can be found by means of comparatively simple tests and the relation of the actual loads carried by the motors, to their safe capacity, thus determined. It has been found that where the equivalent voltage is less than 300, a reduction of voltage, with the same current, makes but little difference in the temperature attained. Even when the equivalent voltage is changed from 300 to 400 volts only a comparatively slight reduction in current is neces- sary in order to maintain the same temperature rise. Thus the capacity need be stated at only one or two different voltages. In many cases where tests or calculations are made to determine the approximate service loads on a motor, the average voltage at the motor terminals is a sufficient indication of the iron losses, and the equivalent volt- age need not be determined. An ammeter in the circuit of one motor and a voltmeter at the terminals of the same motor, read at suitable intervals during a typical round trip over a given route, will thus give sufficient data for determining the loads which a motor is carrying in service. From the current readings, the root mean square current can be found, and from the voltage readings, the average, or the equivalent voltage. The starting current is a most important factor in determining the copper loss, hence it is essential to get an accurate idea of this. On account of the rapid variations of the current while the car is starting and the short duration of the starting currents, readings should be taken at very close intervals, preferably at intervals of five seconds, or less, in order that the large currents used in starting may be duly represented in the results. The capacity of a railway motor is expressed in two different ways : 1st Commercial Rating*. — This is the horse-power output of the motor that will give a temperature rise of 75° C. above the surrounding air after a run of one hour. It also is about the maximum momentary output which the motor is called upon to deliver in service. The commercial or horse- power rating of a motor does not indicate its capacity to do work in regular operation where the demands upon the motive power are very irregular; hence there has arisen the need of a service rating by means of which the proper motor can be selected for a given service without the necessity of going through a mass of tedious calculations. 2nd Service Capacity. — The temperature of a railway motor in service should not rise more than 65° C. above that of the air, as a higher temperature is liable to cause deterioration of the insulation and thus increase the cost of maintenance. The most convenient service rating of a railway motor shows the relation between the rate horse-power (commercial) and the weight of car in tons it can propel at any speed. This should be given also for both single car and train operation in order to comply with the different train friction rates with different composition of trains. Example. — Given a 48-ton car running singly at 45 miles per hour, what capacity motor is required with a four-motor equipment? See curve sheet, Fig. 48, made from "C" friction curve for single car operation. Four- motor equipment and 48 tons gives 12 tons per motor. Follow 12-ton line horizontally until it cuts curve labeled 45 miles per hour, drop to scale at bottom and find 115 horse-power motor required. Select next larger size from standard 'lists of manufacturers. Curve sheets 46, 47, 48 are made for "A" Trains of 10 cars or more. "B" Trains of two cars. "C" Single cars. The four curves on each plate are made for 30, 45, 60 and 75 miles per hour, and these values represent the maximum speed the car will reach with 550 volts on the motors and on a level tangent track. Do not make the mistake of choosing a motor too small for the work to be done, as it will cost more in the end, due to increased cost of maintenance. 676 ELECTRIC RAILWAYS. Lay-over«, — Should there he considerable lay-over at the ends of the run, it may be possible to select the next size smaller standard motor to the one indicated bv the curves. By a considerable lay-over is meant 15 per cent of the running time. Thus a run of 20 miles, requiring 60 minutes for a suburban run, should have a lee-way or lay-over at each end of the run of 10 minutes, in which case it would be feasible to select the next smaller standard size motor than the one indicated by curves. MOTOR CAPACITY CURVES 60°CLtise A- Friction Curve 550 Volts Gross acceleration 120 Lbs. per ton Braking 120 " u kl Duration of stops 15 Sec. Coasting 10 « Level tangent track y 85 80 75 70 bo b0 5b 50 45 40 85 30 25 20 15 10 /i p fr *K> p vV* e36 ■y s -£»•* n r* b .40 60 80 100 120 140 160 Commercial H. P. Rating of Kotor 180 200 Fig. 46. Motor Capacity Curves, 60° C. Rise. A-Ffiction Curve. MOTOR SERVICE CAPACITY CURVES. 677 SERVICE CAPACITY CURVES B-Friction Curve 550 Volts Gross acceleration 120 Cbs. per ton Braking- 130 " " *» Duration of stops 15'Sec. Coasting 10 u Level tangent track f- bo 64 60 52 48 44 30 32 28 24 20 16 12 J.0 12 8 4 n ' ± ? * ~^r £e u. IS ~>u ».B. , 10 20 HO 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 v Commercial H.E. Rating of Motor Fig. 48. Motor Capacity Curves, 60° C. Rise. C-Friction Curve. ENERGY REQUIRED FOR ELECTRIC CARS. 679 GRAPHICAL APPROXIMATION OF ENERGY REdllRED FOR ELECTRIC CARS. Mr. A. H. Armstrong has developed a series of curves, based upon the friction diagram, Fig. 49, from experiments by W. J. Davis, Jr. By the use of these curves a quick approximate determination of power required may be made. The curves shown in Figs. 50, 51, and 52, are referred to curves A, B t C, respectively on diagram, Fig. 49. 100 90 80 70 S 50 10 TRAIN FRICTION CURVES A Ten or more 40 ton ears Two 40 ton cars One 40 ton car -4- —> t z /a / Z -t A y FT 7 z ^ h ^ y y \ / /- I ~f~7^ 14-/ t~tz -4/ - T Mf liY- t t T - \f ._. 10 30 40 50 Lbs. per Ton Fig. 49. Friction Curves. Example: — Given an eight-car train for a schedule speed of 25 miles per hour; to find the maximum speed and watt-hours per ton-mile, at one stop per mile. Look along the bottom of the diagram. Fig. 50, for one stop per mile; vertically above this, opposite 25 miles per hour will be found a curve; follow this curve upward to the left to the zero stops per mile where will be found the maximum speed 45 miles per hour. Again, above the one stop per mile the maximum speed curve of 45 miles per hour crosses, opposite 68 watt-hours per ton-mile in the first column. 680 ELECTRIC RAILWAYS. 190 95 180 90 170 85 160 80 150 75 110 70 130 65 © £ 120 60 B EH no 2 55 100 S 50 A ° /< 4? ^ ^ \ /• 5 w <**%■ ^ c& .v> \ s?^ 5f? bT \ '50 idu] Sen e S pee ! 1.5 2 2.5 Stops per Mile 3.5 Fig. 52. Speed and Energy Curves. Referred to C-Friction Curve of Fig. 49. ENERGY REQUIRED FOR ELECTRIC CARS. 683 The controlling factor in all of these curves is the friction curve, which includes track, rolling, journal and wind-friction. The constants assumed in calculating the above curves are those pertain- ing to average high-speed suburban work as follows: Gross accelerating rate 120 lbs. per ton Braking effort (average) 120 lbs. per ton Duration of stop 15 seconds each. Track assumed to be perfectly straight and level. In the above curves, due consideration is given to all the losses occurring during acceleration with the standard series-parallel controller and direct- current motors. 70 1 Car Train 60 H & 40 •-I 1 V k 30 B S jyj H 20 ^ *%&& & ^| $&#*^ ,£&£*£. £0322 8 10 1 0' 2 3 4 6 i 80 Speed M. P.H. Fia. 63. - Train Resistance Curves for 1 Car Train The inertia of the rotating parts of the equipment generally amounts to 5 per cent and this value is taken throughout, being perhaps a little high for the higher speeds and low for the lower speeds. The speed curve of a standard 125 horse-power motor is used throughout. The energy curves given are somewhat affected by the amount of coasting done, although this is not so determining a factor in high-speed work as it is in slow-speed accel- erating problems. In order that the energy curves should be conservative, they are plotted with only 10 seconds of coasting permitted and therefore the schedule speeds given are nearly the maximum possible, and the energy curves given are also practically the maximum possible with the maximum speeds assumed. Should power be shut off earlier and more coasting be 684 ELECTRIC RAILWAYS. permitted, the energy consumption would have been decreased and the schedule speeds decreased somewhat also, especially with the more frequent stops per mile. .„ , . An inspection of these three sets of curves will bring out the very great effect of the wind-friction when using trains of one or two cars at very high speeds* in fact at 75 miles per hour maximum speed the operation of single car trains becomes impracticable with light 40-ton cars of standard construc- tion, and even at 60 miles per hour is questionable. To quote from the curves, it requires an energy consumption of 47 watt-hours per ton-mile for a train of several cars, as against 137 watt-hours per ton-mile for a single i Co ,T •ain ~7 / f Davis • Abe inwall / /l :ngi leering / / S ew. i > Blc od / s Ba ..occ mo orl m ;ive s V 'ell ngt 011/ / \t / I one ie^ x^ y /**' ' ~y 10 20 70 80 ) 40 50 60 Speed in Miles per Hour Fiq. 54.-Train Resistance Curves for 5 Car Train Length of Car, 51 V' Height, 89M" Diameter of wheel, 33 * Effective area, 96 square feet. jtfo. of units,5. car operating at 75 miles per hour without stops ; that is, a single car opera- tion would demand 3.7 times the energy per ton that would be required for the operation of a train of many similar cars. Even a two-car train will require but 92 watt-hours per ton-mile, or only 67 per cent of the energy required per ton for single car operation. As these values are for constant- speed running, while more or less frequent stops would obtain, a comparison at say one stop in 4 miles would be nearer the actual results in practice. Here a single car requires 157 watt-hours per ton-mile, a two-car train requires 120 and a train of several cars 79 watt-hours per ton-mile. With one stop in 8 miles it is possible to make a schedule of 61 miles per hour with maximum speed of 75 miles per hour, and a schedule of 28 miles per hou. with maximum speed of 30 miles per hour. If stops be increased MOTOR CHARACTERISTICS. 685 so that they average one per mile, however, the schedule speed possible with a maximum speed of 75 miles per hour is dropped to 29 miles per hour, while the 30 miles per hour maximum speed permits of a schedule speed of 22 miles per hour. Thus while 30 miles is but 40 per cent of the higher maxi- mum speed it permits a schedule at one stop per mile of 76 per cent of that possible with 75 miles per hour maximum speed. The fallacy of using high- speed equipments for frequent stops is forcibly brought out by referring to the energy curves in Figs. 50, 51, and 52. With one stop per mile it requires 200 watt-hours per ton-mile with 75 mile maximum speed equipment, and the 30 miles maximum speed equipment can obtain 76 per cent of the same schedule with an expenditure of only 28.5 per cent of the energy. Figs. 53 and 54 show the comparative values of train resistance as deter- mined by various authorities. Following are several train resistance formulae. Baldwin, R = 3 + Jr o V Engineering News, R = 2 + — Davis (45-ton car), R = 4 + .13 V + ' * T [l + .1 (N - 1)] , Smith, R - 3 + .167 V + .0025 ^ V* Mailloux, R =(~7= + ff) + .15 V + .02 AT + .25 y2 Where R = resistance in pounds per ton. b- = constant depending on diame- V = velocity in miles per hour. ter of wheels and journals (6 to 9). A = cross section of car in square feet, g = constant depending on condi- T = weight of train in tons. tion of track (2 to 5). N = number of cars per ton. n = total number of cars in train. MOTOR CHARACTERISTICS. Railway motor characteristics are generally expressed in curve form as speed in miles per hour for 33 inch wheel, tractive effort at the rim of a 33-inch wheel and efficiency. The efficiency is ordinarily expressed as tlie relation between the electrical input to the motor and the mechanical output from its armature shaft. When the losses in the gears connecting the armature shaft with the car axle are also deducted, the efficiency thus obtained gives the relation between the electrical input to the motor and the output at the rim of the car wheel. This relation is ordinarily referred to as "efficiency with gears." The efficiency with gears is the one most generally used, although it is best to have both given in order to eliminate errors made by determining gear and friction losses by different methods of unequal degree of accuracy. Motor characteristics form the basis of all calculations involving maxi- mum and schedule speeds and are generally determined for 500 volts, although nearly all railway motor are now designed to operate at 600 volts. Several typical motor characteristics follow. It is not practicable to include more, as styles of motors change so rapidly. Note. — In changing gear ratio on the same class of motor the sum of the number of teeth in gear and pinion must always be the same. For example, for GE-58-A-3; GE-58-A-4; the sum of the number of teeth in gear and pinion is always 84. 686 ELECTRIC RAILWAYS. 40 H.P. output at 71 Amp. input Volts at Motor Terminals 500 Diameter of car wheel 33" Armature 3 turns, Field Spools 110.5 turns Pinion 19, Gear 59, Ratio 3.42. 100 50 2000 90 45 1800 * 80 o 1 70 40 Jh35 +.1600 f 1400 S 60 © •S 30 u £1200 © © g 50 ^25 giooo 2 50- 25 "Siooo s a e 2 40* 20 «; 800 s 30 15 600 20 10 400 10 5 200 1 E ffic iet cv / \ ! \ i / / \ / \ -s> / *'• / __£ n^_ (I / y i i 10 20 30 40 50 60 70 Amperes 90 100 110 688 ELECTRIC RAILWAYS. 100 90 80 o I 60 o % 50 | 40 (2 so a 20 10 125 H.P. output at 208 Amp. input Volts at motor terminals 500 Diameter of Wheels 33" Armature 1 turn, Field spools | Large 56 turns i Small 29 turns Pinion 29, Gear 60, Ratio 2.07 2400 2200 2000 1800 article = 1600 o 1400 y y to 14W / \ & ,* / s 4 > / s s | 100 ° **J 800 00 ' 0 MOTOR CHARACTERISTICS. I 40 H.P. output at 72 Amp. input Volts at motor terminals 500 Diameter of car wheel 33' Armature 3 turns* Field spools 110.5 turns Pinion 17, Gear 69. Ratio 4.06 28 2400 26 2200 100 24 2000 90 22 1800 80 20 1600 70 18 1400 60 16 1200 50 14 1000 40 12 800 30 10 600 20 8 400 10 6 200 4 10 20 30 40 50 60 70 80 90 100 110 Amperes Fig. 61. G. E.-80-A-1- 690 ELECTRIC RAILWAYS. 100 90 80 70 60 50 40 30 20 10 36 32 28 24 20 16 12 8 4 I 2000 1800 1600 1400 1200 1000 800 600 400 200 40 ti.P. output at 72 Amp. input Volts at motor terminals 500 Diameter of car wheel 33' Armature 3 turns, Field spools 110.5 turns Pinion 19. Gear 67, Ratio 3.53 _\_ Efficiency 4 ^ 4 7 - /V t% --I \ - -4 ^r- - -t s - -t ^ -t h 4»! <7* 32- ^ - Vy S : \S \ . L 10 20 30 40 50 60 70 80 90 100 110 Fig. 62. G. E.-80-A-3. MOTOR CHARACTERISTICS. 691 £ 40 H.P. output at 72 Amp. input Volts at motor terminals 500 Diameter of car wheel 33' Armature 3 turns, Field spools 110.5 turns Pinion 22. Gear 64. Ratio 2.91 100 40 90 36 80 32 70 28 60 24 50 20 40 16 30 12 20 8 10 4 % •t: •3 2200 i \ 2000 \ 1800 \ \ t -ffi cm nrv 1600 \ / v 1400 / V / s 1200 i / 1000 / / 800 fy ee L / 1 600 j P 400 fii e> UjS / 200 n 10 20 30 40 50 60 70 80 90 100 110 Amperes Fig. 63. G. E.-80-A-4. 692 ELECTRIC RAILWAYS. I c ! 3400 3200 3000 2000 26 2600 24 2400 22 2200 '00 20 2000 so /a /300 60 /6 /600 70 /4 MOO €0 /2 /200 50 /O /ooo 40 e €00 30 V 600 20 4 400 /O 2 200 O €OAf.& output at /05 Amp. input Vo/Cs at motor terminaAs 0~00 Diameter of car pvneeis 33" Armature 2 turns. r~/e/c? spooAs 87. S turns Pinion /6~ Gear 7/. Ratio <4*44 T ~X- -X- z J- -1 -J. T z 2 -l / L / V / 3 -/ L_ Z I /. A f -\ -f Fff/'c/'enct/ xr~ sL " = - ^ \ z -/ ^ / X ^^ X -i ?-- t >/ """" t ~t(°U £a .&/ JM- ~7<$- 2£ 7 7 J 7 • : _. O /O 20 30 40 SO 60 70 80 SO /OO //O /20 /30 /40 /30/60 Amperes Fig. 64. G. E.-87-A or B-l. MOTOR CHARACTERISTICS. 693 t 1 I X 33 1 | f&OO 36 /700 54 /600 32' /500 SO /400 28 /300 26 /200 24 //OO wo 22 /OOO 90 20 300 80 /a aoo 70 /6 700 60 /4 600 50 /2 300 40 /O 400 30 a 300 20 e 200 /o 4 /OO 2 O eOtf.Routput at /03/lmp. input Vb/ts at motor ter/n/naAs 300 O/ameter of car wneete 33" Armature 2 turns. r7e/aTspoo/s 67.3 turns P/n/on 23. Gear 64 . f?at/o 2. 73 z L ■/■ I -/ X t J J- H U -f t z IE -/- X f A Z c j. X y "Y z~ V • / r: - , A r s Z ^ . JL~ Nv ' " " -S. z — ^-v — -/ £ff/c/enc(/ s>**" 2 "-^ *" ! -^ -^ ' =» •*_ < ^ > e_ _ s 1 ^cn S 4 ^*Z * t z^z».z t~ i j!li£^- *$t? • J>3&- tf'i? wu ]+? ^ -<*- . ^: CO < 6 & d 2 O /O 20 30 40 50 60 70 80 30 /OO f/0 /20 /30 MOTOR CHARACTERISTICS. 695 Per Miles Lbs. cent per trac- 35 H.P. output at 62 Amp. input Armature 3 turns, Field effici- hour tive Volts at motor terminals 500 spools 110.5 turns ency effort Diameter of car wheels 33" Pinion 14, Gear 69, Ratio 4.93. 66 24 Z400 ZZ ZZOO 100 ZO zooo so •/6 1600 60 16 1600 70 Id 1400 60 IZ IZ00 50 10 WOO 40 a 600 30 6 600 zo 4 400 10 2 ZOO _/_ r ^ I 7 T / / r JL 5 7^ \ ^f^c/en^. J*r -9 / s / _r \ / t N ^ -/ / t ^v t 2^ \U ^P&<*~, A&C- U- 4 \$- Jdr t ^ / 7 t Z 7 7 i. t d /0 ZO 30 40 50 60 70 60 90 100 I/O Amperes nf^t Sr S trao ' 35 HP - output at 62 Amp. input Armature 3 turns, Field tJt. t t- Vnlta ot lYintnr torryiinoU Hflrt snmilfi llft.S turns effici- hour tive ency effort Volts at motor terminals 500 spools 110.5 turns Diameter of car wheels 33" Pinion 21, Gear 62, Ratio 2.95. 1300 I ZOO 1100 If00 4C 1000 90 36 900 60 3Z 600 10 za 100 60 Z4 600 60 ZO 500 40 16 400 30 IZ 300 ZO 6 ZOO 10 4 100 ! 1 /I IWJL 1 t % / - 7 _ T t V I / %cz. . ~r-£&&g*l ^ 7 =p- 4 \ -/ i v -f t \ z h ^w ^5^ =^*ep^. a / ^rfr— */ " " ~W- & \ £ ^ ^ • -n 40 - V A / -30- li lot c lo- W ^Jiu. |5J /^ r Who UR ~20~ 10- V ^ft- -10- -5- s& \s fc* £■ _20_ L-S 8E r ROM 75° r — ) i i X) JO U 4 an] >er< I iS 6 < £ ►0 Fiq. 70. MOTOR CHARACTERISTICS. 697 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II 1 WESTINGHOUSE No. 92 A RAILWAY MOTOR 500 VOLTS GEAR RATIO, 18 TO 66- 83* WHEELS CONTINUOUS CAPACITY 30 AMPERES AT 300 VOLTS H u 28 " " 400 » & 0) A s W 1 I 1 2G0 ^ 240 a +3 220 @ W fa 100 — 200 a I o o9 180 51 \- E r F/C /EN sy c7F -80 3 ^ S PR 3X» JXr I s Icy <£U1 G£, RW // 7 \ 1 ''"0- JLOU -70 v / I \ / L40 A / 4 / -60 ^(\ 7 V <^v V I 500 L-20 / \ ^n \ L , * ^ <>/ r J & 100 m r ' $ r^ p/ y^- iS* 40 ^n -* V i m 400 ,, 0D_ © -0- g as J. H 240 220 "(V m g u h : 3 ! P S3 -o- * f \ < Ci H & \ a s 100 -25- \ I \ 25 \aj aw \ \ ISO t~ \- bFF ICIE Ntt l_Wr 40| T -££ ^s -80 20 / K- ' \ A ^R 3Xlfl *AT : — . "ff? GF^ J 20 )0-l60 / / H \ C/£ NC y W/ T ^T"~ RS 1 y \ \-A L40 70 / \ \rti 4> -60 [5 V A X ♦ )0-120 n C> / ^> y $s 50 K} > LOO \ *tf : . V x> ^r SP Etn -40 ■10 IP V.' £*r ^ 10 30- -80 v°' &J -50 v X • 20 -5 V" K>- -40 te 5( -10 \? \t 20 ^O £_ rtge FF OM 75 V. - T 5 7 5 1 X) 1 15 1, X) "Amperes Fig. 72. MOTOR CHARACTERISTICS. 699 S -aj- Oh WESTINGHOUSE No. 92 A RAILWAY MOTOR 500 VOLTS CONTINUOUS CAPACITY, 30 AMPERES AT 300 VOLTS " " 28 " " 400 ■■ -1500 1 e -1400 \ B -1300 I 203 -©- p -1200 1 210 i -1100 220 \ 100 -1000- \ Q ^ -90 -900 - 180 V Fr ^Cl :NC -80 -800 y 800 160 \ -70 -700- / \ 140 / 700 -60 -600- / l-n i\T >^ soo — 120 % ^ (V) 1m % ^ ^r 500 — 100 ^ \ -40 -400 50 \° '&} k ? ? 1 80 1> too -30 -3 00 4v -60 \-2 JcS >■ K ow -20 -200 'A < V ^ te 200 dp* s£5 c. -10 inn t s& WL ^f. -20 ^ -51 St PROM ?b° ( 1 1 20 30 40 50 c 7 80 90 100 1 L0 | Alnr er es Without Gears Fig. 73. 700 ELECTRIC RAILWAYS. 1 1 II 1 1 1 1 1 1 II i 1 II o -13- WESTINGHOUSE No. 93 A RAILWAY MOTOR 500 VOLTS CONTINUOUS CAPACITY 50 AMPERES AT 300 VOLTS i( n 4 6 ii tt 400 ii o< g BQ o- _£_ H Ph P^ / h2 _ a ti / 14w -13 X)- r / / X«A» t / / hi t / / -12 ± / / S £ -1100- 1 ^ «„L Sr u q // / -1000- f 200 I ^ b/ -90 -90 -900- \ EFF ICH= A &/ 90" ion > i <&/ f -80 -SO -800- / A \ 4/ oojteo 7 \ \ w 8 -70 -70 r f 7 V XH.40 < a>\ .<£ // i -60 -60 -6 $y [K)-120 -» ■60 -50 -5 1m \ti Afe ^ bP£ ■£».f>. M -40 40 -4 . I 33 ** X)- -80 X) \q -30 -30 -300- ^ »~ -60 4yfe fcc -20 20 1 //^ T V H oo- -40 , 2™ (TV / / ^ k -10 -10 / ?fe ?l y\ , 30- 20 / / , ^ N < s^ & RfS c l / —- j / 6 c 2 40 60 8 l! )0 120 i^ 10 1 30 li 30 2( X) A mpei i 1 Fig. 74. MOTOR CHARACTERISTICS. 701 1 1 ii i i 1 . i i .i i ii i ii 1- — i .WESTINGHOUSE No, 112 RAILWAY MOTOR 500 VOLTS GEAR RATIO, 16 TO 73 - 33" WHEELS CONTINUOUS CAPACITY 60 AMPERES AT 300 VOLTS . « 55 - " "400 « ' Cm' .o 2 \ 6 . IT -30 I A > MO 1- \ • | s' \ | 220 ,100 X -5000- 1 200 2o / -90 180 FF FICIEN CY w TH EAT ,S__ 180 f ' 4 .1 ■fr* >ol Ml b+ nc ENCY -VIT H~J ^S -40 oo- 160 -80 IbO f -TO WO <2j -35 )0- 140 Ico A^ ^/> ft* -60 120 -15 d SI r\ -xj f / A <> 30- 120 j \c **S y> -50 m b u -25 :o- 100 4° $* ~~— fo Ml* s*£ -40 -80 -10 ra w- -80 Vo 0. ^ *k n 10 & fa ^ It -10 IM ftr »- -20 C^ £'S£ F i -/ *£ a X -* £ 4 -_ _£X -^- - - y. -j4- a 2 2- v S S - * l3- zx &fe -dz._ _ _ j :___:d: - o o % _Z £ ->z~ o - 7 Tl t* r rC 13 =_ / 2500" ""2 ~?£~ £ — " n - 7 S -F 6 ^ / £ / .i-._ #-- B- - - -- EF .p«i.lEN0Y-_. W4+Oj4- 7 &?,rr* 1 :S-:::z:::^:::;:iEffe|::::: — 1 •■ H-^- =s»-r?rE := F- EFF| c ENCY ,'%IIH^'"-Wrji ''"onnn 160 80 40 ^:V- ~ f ~ ZaZ S& -<$?/' &?' ikaA 120 60 30^ ~Y A ^ oPpT ^A>^ 1500 ^.y -^3& cfc y <**J' S&Z- -,**$> rfyT egX 100 -25 V /\ .. it^r ' o^^ I\. / V /flXjrf Ny f \ "80 20 X' y\ V? 1000 *>J- -L 5£fc -Sc oO 15 / s y __?St:z==^~^Jii^ : : :::::+::: zz-t?\itz :: : V^A,4- =*-»-. ZE. ZS \? 5fc,>-o AO, X -fc S -+40 1 - "1( £%' s Jg 50(!> 7 ^ ^^^ ^> "Sj?,*.& <2n 20 " of f JS.. - _ - z — ^ 60 100 -450 _900-L _i250 Z I J ! 1 1 I J..J..., , Fig. 77. 704 ELECTRIC RAILWAYS. WESTIT K|a 11Q RAI 1 WAY MOTOR << " i 5 " " 400 " &~ a _j_ _ _ — . S „ xJ] 2 112 ■ - ■ j , 'b § : ~v~ h: fa^ IT ^ H^ ^, tr 240 S_ Li__ . _ L Q _. ._ 5T -H- -t - V - Z Hi X X .2- X 4 - t i 4 - XXX 7 2750 22 i: ri M 4 / I rj -- n i it ?! 2X itit -F X 4- 7* -t -t- ~ + 2o()^~ l(X)t~ 1U(X) "i / // 2500 200 t - \ -•225r--90 900- — t-'-V— 4-i LU ^ ficienc-J— f--/+-^-l 2250 — 4:80- 1 A~~r y\\ \ J z n i 400 "oO oOO ~ T ^ >5 // 2000 160 L _K " t ^ gA~// w^/A 175 rfO <00 "TT y ' .«-^££ ■■ - /ly ^^u^ 14o oO" 1- 500 r 7 * /L nV^" t" " *"*"•» 125C ^00' \ / / <£/ i 1 ■,-... /°/ i X VII 100 -10 400 r?x&~ ' ' 1000 80 it L* // I£^o a -,-/% „« <0 oO oOO "t J ' ~r~ "^. 5£ - 750 60 ^ 1 -~^^r 500 40 ^ /\o " $p7" "' " ^V^" •^5 10 1007 / ^'<3Tt" ***>>^ 250 20 _ _ __ _ «. _, _ -vv - r/ 2 - N & :::::::::::??:::::::::::::;:;: =f^^^R^irW>;r ^ 100 2 IN MTTPl 1 ' 1 1 1 1 ' ' 1 00 ^0O 400 - - --Ami i Mill T Fig. 78. MOTOR CHARACTERISTICS. 705 - ■■ ■ i n 111 ii 1 1 1 1 1 1 1 1 1 1 I I I ' II l i 1 1 I II 1 1 1 1 1 1 1 I I 1 1 1 1 1 1 ii 1 1 1 1 1 1 ii M I I I i I I 1 1 | I I I I 1 1 1 1 II I ~:: _~:::_ westi nJGHOUSE z i: ; ::: no. 119 rah -WAY MOTOR BZ « _ I_ 550 VOLTS g fj CONTINUOUS CAPACITY 5 5 AMPERES AT 300 VOLTS _^T _T? M (i 35 » " 400 " te Jc~ S -13 r j H Fh 'in / >i _ v • , oUUU Z40 4- 5 ZMZZ _ "~ 7 -- -- 0) T^ K-t / & ■/ ^7 / OoUU 2z0 4- * -^r- : i -t - 4 "t/A-,2 X ±1 iTZ 41 250 100 ,50 1 o//* 5000 200 ~ -£LT~ Zl ii L-fc" "" -£>5 w it X2: & "' 175] 70 \ - tr ■ \| ^Yfh / ^3500 140- I ' " ■ \ X I ±nS^ " " "1 "it "" \ - \ zsJLzmZ- x 11 &l2^ ± 150 60 30 \ \ ~t S7 77 " 3000 120 ZZZt Z ~ "y.-t \ ~& X S.o ' <$ I Y [ / 125 5G* r V r 7 "2500 - 100 ::: : ts jz ^ Ka X-^2 o \ >nv> IIXJ 40 zO^ ' i° " yNy/Y 2000 80 x) ~? * @ <0 oU A ' I ""■■'Gg.^ " 15(i0 60- 1 ^ 1{j -¥ " ? \/«, I-H ■ 50" 20" L // A<^ SKo ~ 10(0- -40- xx r_ xgn /£^ 2 \cT _j_ s ^g_ _j_^ _j_ i^7 V V - ^>K 1/1- - -,{- ' -,£ VA o """^P/lr ' Lnn k/\ 25 10 77 ^tf ^t^T 500 ~^ 2 > / £le / / >>.fi Sf *, -^ ;* S ft I _ x z -Mttsij-Br- cC- 1m < !t l x 3 idaii 11 11 T : :__:„:-., - Jill II II 1 Fig. 79. 706 ELECTRIC RAILWAYS. DETERMOATIOl OF EXFROTT. Gotshall gives the following as a method of approximating the demand for energy of an electric railway. Let W = maximum weight of loaded car, or train unit, in tons of 2,000 pounds each. D = length of road. T =time in minutes occupied in running between termini = single trip. K = energy consumption in watt hours per ton mile. N = number of cars or train units on the road during time of maximum service of minimum headway. Then, W X D = ton mile per trip = P. 1* V K = energy per trip in kilowatt hours. 1000 P XK 1000 P XK 1000 X -^ = mean rate of energy input per car or train unit. .. 60 X N = A = total maximum average energy required at the car motors for maximum service condition. If to the foregoing, 25 per cent be added for transmission losses and heat and light, 60 X P X K X N X 100 n no PXKxN . — nnr . „ — = 0.08 — = maximum average demand = R. JLUUU X J- X . (O 1 To R must be added the fluctuations, which will vary from . 2 R to . 33 R, as the number of train units in regular service are great and the average load consequently relatively high, or as the number of train units in regular ser- vice are few and far apart, and the consequent relative increase of the load during certain hours relatively great. In the foregoing, the quantity K is the important quantity. K will vary with the schedule and the location, the distance between, and number of stops and stations, as well as with the alignment and gradients. Table VII, has been compiled from data showing relations between schedule speed and energy consumption in watt hours per ton mile. These figures are based upon approximately straight and level roads. As the effect of grades upon energy consumption is, to a large extent, compensating, the data may be used with safety. The compensating effect above referred to is due to the fact that while a car going up-grade is consuming more energy, per contra a car going down-grade consumes much less or none, thereby equal- izing the effect of, or compensating for, the gradients. Table VII. Watt Hours per Ton Mile for Schedule Speeds of Stops. 40 miles per hr. 35 miles per hr. 30 miles per hr. 25 miles per hr. 20 miles per hr. 15 miles per hr. Miles. 3 2* 2 H 1 i ! Feet. 15,840 13,200 10,560 7,920 5,280 2,640 1,320 110 121 142 80 90 99 123 78 83 86 95 128 65 74 80 85 90 145 53 54 60 68 74 119 40 40 41 43 50 56 120 Train friction in pounds per ton 35 30 27.5 25 20 15 ALTERNATING CURRENT SYSTEMS. 707 The breaking effort or retardation is taken at 150 pounds per ton. The stops are taken at 15 seconds each, except in the case of the 15 miles per hour schedule, where the stop is taken as 10 seconds. The foregoing figures are for cases of approximately level and approxi- mately straight roads. For a schedule of 40 miles per hour the speed attained will be between 60 and 65 miles per hour. A schedule of 25 miles will require speeds of from 40 to 50 miles per hour, etc. The rate of acceleration for the long runs varies from 75 to 110 pounds per ton, going as high as 210 pounds per ton for short runs. The foregoing applies to single car units. If units of more than one car be used, the friction in pounds per ton will decrease and with it will also decrease the energy consumption in watt hours per ton mile. IL\fwLE.PHl§E ALTERHATOO CUR REIT SYSTEMi OF RAILW AY IHOTOIIS. The use of the single-phase commutator type motor for electric traction was first seriously advocated by the Westinghouse Electric & Manufacturing Company, and a description of a single-phase system, proposed by that company for the Washington, Baltimore & Annapolis Railway was read by Mr. B. G. Lamme before the American Institute of Electrical Engineers in October, 1902. The development of this type of motor was at once taken up by other manufacturers including the General Electric Company in this country and a number of prominent companies in Europe. The first rail- way to employ the system on a large commercial scale was the Indianapolis & Cincinnati Traction Company, which began operation over a short portion of its track on December 30, 1904. t • Practically all manufacturers employ a laminated field, an armature wind- ing similar in general to that used in direct-current machines, and an auxiliary or compensating vending on the field, to neutralize the armature reaction. In general, also, the single-phase motors of all manufacturers are designed for operation on 250 volts or less. A frequency of twenty-five cycles has been used exclusively in this country. In Europe, however, some roads employ this frequency, some lower and some higher frequencies. Lower frequencies are now being advocated in the United States. Sizes of motors up to 250 horse-power have been built. Those in service at the present time range from 40 to 150 horse-power and are used in both two and four-motor equipments. One of the essential advantages of the single-phase system is the economy of feeder copper which is secured, due to the use of a high trolley voltage. The higher the voltage the greater the saving thus effected. On the other hand, the greater the trolley voltage the greater the difficulty of insulating the line. Trolley voltages of 3300, 6600, 11,000 and as high as 13,000 are in use. No attempts have been made to standardize trolley voltages at present, but the general tendency seems to be toward the use of 6600 volts for ordinary trolley roads and of 11,000 volts for the electrification of existing steam railways. Single-phase equipments in general include, in addition to the motors, a specially designed trolley to collect the high-voltage current, a transformer to reduce the voltage for use at the motors, and the necessary controlling devices to regulate the supply of the current and control the speed of the car. These latter devices consist of drum-type controllers for small equipments and single car operation and unit switches operated by inde- pendent power for large equipments, or where multiple unit service is desired. The single-phase alternating current motor will operate equally well on direct current of the proper voltage and by connecting two or more motors in series a single-phase car equipment can be arranged to run from an ordinary direct-current trolley as well as from a high voltage single-phase trolley. With such an arrangement, cars can be run over the same tracks as ordinary city cars when entering a town. 708 ELECTRIC RAILWAYS. Figure 80 shows a diagram of connections for a double equipment oi 50 horse-power single-phase motors with hand control as supplied by the Westinghouse Electric & Manufacturing Company. It will be seen from the diagram that there are five different notches on the controller, by means of which the motors may be connected to live different points on the trans- former and that the motors may be run continuously on any notch, thus giving five different car speeds. When running at less than the maximum speed, the power required is reduced in approximate proportion to the speed. ALTERNATING CURRENT SYSTEMS. 709 Figure 81 shows a schematic diagram of a car equipment for multiple unit operation on either direct or alternating current. In this equipment the main circuits are opened or closed by unit switches operated by com- pressed air from the brake system in the same way as those employed in the Westinghouse unit switch system of control for direct-current motors. The main switches are controlled by means of magnet valves operated through auxiliary circuits from a master switch. The auxiliary circuits are carried from car to car by flexible connections in the usual way so that the operation of the master switch on any car operates the main switches on all motor cars simultaneously. See Figs. 81 and 82. The auxiliary circuits between the master switch and the main switches .A.C. Trolley d.C. Trolley Sequence of Switches 2 n-j ♦ tfff •a 2 • H •s • >* ss i. . • • IM !r on ► • 1 • » i°: - 1 " Ti -,•- --" * —Hlr •l»l !• •. f • • • 5"*' b ;• ! 1 « O Running Notches Fia. 81. Schematic Diagram of Westinghouse A. C— D. C. Car Equipment. are led through an automatic change-over switch, which normally remains in the position for direct-current operation but which changes to the position for alternating-current operation whenever alternating current is supplied to the car transformer. By this arrangement operating the same master controller closes different main switches, according to whether direct current or alternating current is being used by the car. For the sake of clearness the auxiliary circuits are not shown on this diagram. Figure 83 shows a schematic diagram of a car equipped with four 50 horse- power single-phase motors for operation on 3300 volts. Figure 84 shows diagram of connections for a quadruple equipment of 75 horse-power motors with hand control, as supplied by the General Electric Company for operation on alternating current only, and figure 85 shows diagram of connections for the same equipment with multiple unit control for operation on both alternating current and direct current. Figure 85 shows performance curves of typical single-phase motors manufactured by the Westinghouse and General Electric Companies. 710 ELECTRIC RAILWAYS. £ C»lt*KltrSK*tt itft* Chimin \\ysiBfiaKAw l mOuMtt Afasfer n/>;rp.tt„e,r Cwt/vlter tr/la/UPvmp. Q Fig. 82. Diagram of Apparatus for Unit Switch System of Multiple Control, A. C. Equipment. \Confri1ke Ma/ n Cylinder LoA^Cy/indtr^^^ockw/thOum /bA/'r/Teserva/r > Conduit. orHandPurnp. -V I Contro/Je/r V te=e> Si Pre yen/in /^revet/titr Co/'/ RtsistoKt. Main Auto-Trvnsror/nit Fig. 83. Diagram of Apparatus for Hand Control, A. C. Equipment- General Electric Company's Hand Potential Control System. This being a system of hand control for alternating current running only- it is less complicated and somewhat lighter than a train system. The General Electric potential control is also used for combined alternating current and direct current running by the addition of starting resistances and a commu- tating switch, whose office is to make the necessary change in connection. This potential control gives a higher efficiency equipment than is provided by any form of resistance control. ALTERNATING CURRENT SYSTEMS. 711 ^ „ ^f)|. fj«j ^ d /? 80 40 • %^ £<;/£ _i -£°2«f? ;cm, «FAR 4 ON 70 14 "iO ■*^£ §£* cv- > FRICT > LOSS 4 13 DO 60 30 1200 1100 B0 1000 900 40 20 800 700 a o < BE 100 50 20C 90 180 — ££ S£*> (S4t 80 40 16C c foft \RC£ fr a ^vc ]vj f^x 70 HO 60 30 12C }, w I / J?/ 50 IOC & / 40 20 8C 30 6C 20 10 4C i'P£ ED 10 20 4 B 12 160 AMPERES 2 )0 2- 21 Fig. 87. SINGLE-PHASE MOTOR CHARACTERISTICS. 715 fc o £ u K 90 a 160 80 140 70 120 60 100 60 40 20 10 #132 RAILWAY MOTOR 225 Volts 3000 Alts. \ 36"Wheels Gear Ratio 20:63 \ Performance of Westinghouse \ 100 H . P . Single Phase M otor \ operating at normal voltage \ on alternating current. "1 Ffficie """ s — i? 1^ o*. i? f/ 4 V ^ / 4000 3000 400 Amperei Fig. 88. 716 ELECTRIC RAILWAYS. o L80 w 90 X 60 80 40 140 70 100 60 80 40 20 60 80 40 20 10 #132 HAILWAr MOTOR 150 Volts Direct Cucrent 36"Wheels Gear Ratio 20:63 Performance of Westinghouse 100.H.P.. Single Phase Motor operating four in series on 600 volts direct current. Fig. 89. SINGLE-PHASE MOTOR CHARACTERISTICS. 717 a*- LOO 80 400 60 300 SO 20 100 #130 RAILWAY MOTOR 220 Volts— 3000 Alts. ^ Single Phase Gearless 62 "Wheels Performance oi~250H..P. ~Westinghouse gearless Single Phase Motor foT New York, New Haven & Harftord JEtaUtaoad. o 8 g 1 5 sooo **^Po ^£aei ^fficiejj or cy*"-—— 4000 3000 u^V «

03 £ £ ^ > > > C S 0Q fl ^ 02 d^- 02 d- Jl 02^ d- ^ «,£ S3- «> UMaJ Om ~ Oi-rl « O Wi_H ~ O « »— | to o II ^ Q, q co q ^ 05 o oj o i-h o oq oq to q o q w rH c^j q co ^' OJHON iH t-S O " r-I o od l> lO lO i-H rH I— I i-i i— 1 i— 1 rH J, oa P33 ooooooooooooooooooooo oqqoooooooooooooooooo rHco"oo"oo"io*ocDioo"co^W^i^o*^*o*o"croo'"oo* ^CO«DiOiOOCO!>iO«DOO(NC^C^(NCOTtT^tDcC) +a 09 P c-V-"o^^oooV>*oo*t-~oV-^>o"oo^"o ^ i CO fi oi M c3 S C J4 C o "oi 03 o *3 8 J* 09 I ti O o CD o3 03 *CP 1 CP 03 09 g "8 M o3 r£ '% CP eg to CP > P o o O* IO CO on 1 09 CP o3 Pi ■*a *CP > t) 03 o rH M CP o -r»" bo CP 03 C CP (h CP T3 o3 I eg rfi I '3 0> CO CP O o o s CP >> a o3 S-5 be rH o bfi "3 CP CP rH rG -o b 00 E © q TtH q oq i> iO CO q q co # CO OS Cs' ,-H CM CO OS OS* CO* Tt" id r-i r-i CO CO CO CO 1—1 CO CO CO CO CO CO <* ^ £ ID hL CP^ 1 iO t» tH ^ o O iO o o 00 o iO o iO OS os iO IO 00 CO CM 00 T^ o o CM CO 00 CO -^ o "fl Tf «* 00 o 1^ 1^ S^S CO CM CM CM co I> CM CO CM CM CO co CM CM £ ° £ s 5? si t^ q Tj* iO o q iO rH q CM CO CM r-f t^ CO co" os' iO o" id CO IO os OS s CO* id CO* 03 00 co i> CO co OS CO OS CO t^ 00 t^ 1> £ * o o o O o O © O o o o o o o £ 5 o o o o o o o o o o o o o o °i q o_ q_ °°. " co" r^." N* OS* CO" OS* CM* Tt* OS* CM* CO* CM CM CM 1—i CM OS O 00 00 mber of ops. "tf i> 1> CM OS T^ CO CO CM CM iO CO q -cH t^. t^ "*! q CO «o r-i rH TjJ ^ 00 "*' *d ,_J CM CM rH id Tt< 00* CO* CO o* 6 so rH r-t co CO CO CO s ill . o iO iO o O o O o o iO o iO g iO CO 1— 1 rH CO CM >o CO r-i CM CO CO «I H ^ 00 1^ -tf CO t> 00 tH CO 00 O -tf CO rH r-i rH rH iO iO Tf "* 722 ELECTRIC RAILWAYS. I\rEHiniM\ CAR TESTS. By W. E. Goldsborough and P. E. Fansler. Trans. A. I. E. E. Tests Made upon Cars of the Union Traction Company of Indiana. The cars used measure 52 feet 6 inches over all and weigh 63,100 pounds- The motive power equipment consists of two number 50 C Westinghouse motors, which are mounted on the forward truck and are nominally rated at 150 horse-power each. The motors are geared with the ratio of 20 to 51 and are geared to 36-inch wheels. Records were obtained from 10 cars of this type. The following tables give the results for several different cars used on various routes, a special test of three cars, and a table showing the personal factor of different motormen: Table X. Train liOgr. Train No. Car No. Direction. K.W.H. K.W.H. Per Car Mile. 1 12 19 28 35 246 246 246 246 246 East West East West East 131.2 128.5 125.6 134.8 119 2.32 2.28 2.21 2.38 2.11 Average, East 2.21 Average, West 2.33 9L 18L 25L 34L 250 250 250 250 East West East West 107.4 123.8 108.5 119.6 1.9 2.19 1.92 2.11 Average, East 1.91 Average, West 2.15 39 32 44 252 252 252 East West West 128.7 139.5 113.1 2.27 2.46 2.00 Average, East 2.27 Average, West 2.23 6 13 22 29 41 42 254 254 • 254 254 254 254 West East West East East West 142.5 137.6 139.2 162 119.0 126 2.52 2.43 2.46 2.86 2.10 2.23 Average, East 2.46 Average, West 2.40 INTERURBAN CAR TESTS. Table X. Train I^og*. — Continued. 723 Train No. Car No. Direction. K.W.H. K.W.H. Per Car Mile. 10L 17L 26L 33L 255 255 255 255 West East West East 101.0 96.0 106.0 101.0 1.77 1.70 1.87 1.78 1.74 Average, West 1.83 2 7 15 ]6 23 38 260 260 260 260 260 260 West East East West East West 122.4 130.6 127.5 114.2 133.5 128.5 2.16 2.30 2.25 1.85 2.35 2.27 Average, E ast . 2.30 2.09 31 8 24 261 261 261 East West West 156.5 142.0 132.8 2.59 2.51 2.34 Average, East : 2.59 Average, W r est 2.42 30 3 14 21 37 263 262 262 262 262 West East West East East 127.0 111.0 122.0 123.0 112.5 2.24 1.96 2.15 2.17 1.98 Average, East 2.03 Average, West 2.19 11 20 27 40 43 4 263 263 263 263 263 263 East West East West East West 124.5 135.5 94.5 134.0 118.5 140.0 2.20 2.39 2.48 2.37 2.09 2.48 Average, East . 2.26 Average, West 2.41 724 ELECTRIC RAILWAYS. Table XI. Comparison of Car Tests. Number of car 255 252 252 Service, west bound semi-limited local limited Weight 63,100 63,100 63,100 23:48 122 20:51 156 20:51 Total time trip, min 126 Time urban work, min 44 40 34 Time interurban work, min. . . 78 116 92 Average speed for trip, m.p.h. . 28 22 27 Average urban speed, m.p.h. . . Average interurban speed, m.p.h. 8 9 10 39 26 33 18 5 44 15 12 Urban starts 7 Interurban starts 13 29 5 Maximum speed, m.p.h 64 52 Running speeds 50-55 40-45 40-45 Running currents 173 145 145 lbs. per ton 27.7 19.9 19.9 Time to reach 25 m.p.h 30 30 30 Acceleration current, max. series 280-340 200-300 200-300 Acceleration current, max. par . 320-540 250-300 250-300 Consumption, k.w.h., p. cm., west 2.20 2.44 2.10 Consumption, k.w.h., p. cm., east 2.38 2.80 2.32 Consumption, watt-hour per ton mile, west 69.7 77.5 66.7 Consumption, watt-hour per ton mile, east 75.5 89.0 73.5 Sq. root mean sq. current, west . 95.6 92.1 78.0 Sq. root mean sq. current, east . 105.5 98.4 87.2 Running factors, west 43.5 37.8 36.2 Running factors, east 43.3 31.5 37.6 Average voltage, west 485 429 Total consumption k.w.h., west . 124.9 138.0 118.8 Total consumption k.w.h., east . 134.3 176.2 131.2 Ta1>le XII. Personal Factor of Hlotormen. Local Runs. East. Total K.W.H. West. Total K.W.H. Trips. Name. Min. Average Max. Min. Average Max. East West Eller 122 135 148 114 125 136 6 6 Lee . . 116 121 126 124 129 130 4 4 Robbins 122 131 138 119 124 128 4 4 Green 113 123 131 126 134 141 3 3 Young . 118 122 128 112 128 145 3 6 Griffin . 124 130 140 127 131 134 3 4 Embry . 108 126 154 134 135 135 3 2 127 130 26 29 INTERURBAN CAR TESTS. 725 § Z 2 fa 5 fa 9 « s e « fa S i 3 fa w i* s I i s « p c < (a u «* o> p. ~ So h r^ cq 5: * I O A C3 s <* oc o 5F o CO CO o (-. c f £ c 5 c3 1 *3 CD o -M _M .S pq 5 X O •S -2 03 *c s* ^ 3 g O 0? c M u u -S ,o ? ^ s o* .« u '-2 •id S: t o O X -r tN o3 03 c o ** N a c U3 iO * ^ 8 1 iO co CO • "S o 6 OQ o OC c>^ M CO CO S © p £ 73 <- CO 8 c3 b) a ffl T3 ^ 73 H | 1 cp S faO c 2 ^ C a M £ £ *** G -^ B 0> u c H -*> «3 o C © s w S b9 PL CO T— 1 CO * s CO «3 T^ cq k . -w .* J i^ CO o Christensen. Independen Acting Aii £ o3 c OS oo o 4 cq CO w o3 W a 5 c a iO .1 N PQ a c 5 1 pq a t> c M O d T3 a 9 c * s; »d ^ „; o C ^ N a n «3 -^ CM s IO »o f-H M 6 C C 5 T3 CO g J G s oo 3 J bO .9 1 CO M a C c LO co CO CO CO 1 1 . £ C $ B 2 M 3 c3 o » g -2 d *3 c g BO M Q O to 6 CD O CP S 1 ►3 4* O 3 .2 DE u 1 C c *o q •s 3 0) CO S ■3 -a * o3 i. e2 a a E 2 9 8 i s c a d S +a iO CO IO 00 ^ ,H CM © © J J "3 J> o iO CM iO CM CO OS CM CM © IO M © 00 CO iO £ CO < iO CO IO ^ © cS © t^. 9 M © rH I> o iO CM iO CM CO CO 00 rH © © CM 00 © CO © CO CO rH t>. IO iO CM CM © 00 IO © >- d \ l> w o iO CM iO CM CO CO 00 rH © rH CM 1> CO CO © CM O 3 © > CM ^ CD CO o © 00 CO © © co -*i "0 © "H iO 1> CO CM o CO CO OS © rH 00 00 CM © © IO 00 rH o rH ^f Tj< iO © CM t^ iO © # fl 33 o- — CM cm rH w iO co CM o co 00 00 © CM CM 00 CM CO © IO iO 00 § U Nl 3 © iO o IO tH o CM © © IO CM £ CM O CM CM CM o CO CM 00 CM CM 1> iO rt rH 00 CM tH © CM 1 © 03 w IO O iO iO t^ Tt< *H CM © © 1 .a 8 T—{ o o CM o CM O CO TP t^ tH CM °2 iO y-< § s CO rH rH © © q fcs CM id rH CO CO CO CO q rH CM rH rj< 00 © oq CM* IO ©' © 00 00 ^H CO rH CM 3j 3 o rH ^ rH iO l-< 00 00 IO © © rH w IO 1> CO CM co CM o © CM 00 00 CM y—i © CM • IO IO rH CM © CM £ iO O IO _ iO CM © o © © a o 4 CM O CM CM CM »o CO OS o © CM CM CM © © CO tH CM CM © (N w •O O iO »o iO 00 CM © © • IO IO E rH o o CM o CM CO OS OS © CM 00 IO ^ CM 00 IO CO © § 8 © T— 1 © a QQ £ o CO CM CM © CO CM CO* CO CO CO IO OS CM CM © id CM CM iq © © © © rH CM 1) © OS © OS CO IO CM © rH O 00 a © w to CM rH CO CM co CO 00 00 CO CM CM CM © 00 iO ^H CM P © £ CO rH o CO o ^ ^ © CM © tH 7-1 rH IO CO CM iO CM 00 CM CO CO OS rH CM t^ r^ CM S 00 ^H £ I> OS <* CO CO o 00 © CM iO tH 9 3 a? rH w 10 CM rH iO CM CM CO o OS CM CM IO CM © oo © iO a o S i-H © £ 00 «* CM CO o ■<* CM CO iO 00 2 -2 'ea- rn' rH rH rH o iO CM CD CM TH CO CO OS CM co CM 00 IO rH CM © w 10 -tf OS co o ,_, CM © © CO s o •*> y-i O CM rt iO CM CO CM CO OS 00 CO CM © CM © 1> iO CM S © rH © £ OS rH CO co iO © rH CM iO iO CO S O <4S * CM rH IO CM CM CO IO OS CM © 00 CM CM © iO a © w © -* c CO © ^ © © rH © 00 H rK rH CM CM M iO CM CM CO CM 00 iO CM 3 CM © © 6 EI 8 © -♦J 1 s g a 9 v» Ch *o o 53 d m £rHrH O -r» o H 01 JO 1 05 a s a> p 0) a . tH a> o B W i k. to t-< o . -i CU ■rs t---r 1 OS a - o QQ INTERURBAN CAR TESTS. r27 rH » CO © iO lO 00 iO CO CM CM CM CO CM iH CO OS "tf CO <* 1> iO iO l> CO o lH 1> "V. t^ w o iO CM iO CM oo CM OS iO 00 CM 00 -3 rH CD «D 00 ^ ^ £ CO "tf t^ iO o "* CD iO 1> CD 00 iO OS OS ^ T— ( tH t> w o CN iO CM 00 CM CM CO CM IO IO rH 00 1> CD rH CM © CD* 3 w CD CM CD CM 00 CM CO CM CO CO iO CM CM \ OS £ iO O iO iO iO OS o OS 1> iO CM CM CM o CM CM CM 00 CM CO CD CD CM iO •"•. OS 00 IO CM CM © w iO O iO CD ip CM CD CM O -* CM o o CM o CM 00 CM CM CD CM OS rH 00 t^ ^ t^ © u O ft CM o CM T* O T* o O o o i—i © ► o CO CM rH CO co CO 00 CM O CD 00 CM 1> iO CM o CD l> iO OS © rH © faO OS O OS Tf< IO CM CM iO rH iO o c3 w iO CM i— i CO CM CO 00 CM CD iO O CO CD CM iO CD iO CD OS 09 I> & CD T* o •* iO CM 00 © O O 1> >> OS rH iO CO CM IO CM 00 CM 00 CM iO OS CM 1> IO CM CM 1> OS iO [> OS th CO "* iO rH 00 00 O TjH OS rt w IO CM i—i iO CM CM 00 CM 00 iO os CM rH CD CM* o 00 CO CD Q M e9 O o s 1> 00 "# CM "* iO CD O OS iO •tf OS •"■ £ rH o IO CM CD CM 00 CM IO OS (M iO •"• CM CD iO 00 I> w iO -tf OS tH IO CD CD CO iO ,_, OS 1-1 O CM rt >o CM CD CM 00 CM OS iO 00 CM 1> iO O CM* iO 1> 00 iO OS CO CD OS ^ CO Tf< iO tH O o OS r-( £ iO CM rH IO CM CM 00 CM 00 IO OS CM CD w CD ^ o rfi CM O o iO o CO OS rH CM CM rt IO CM CM 00 CM CO IO CM CO CM 1> CM OS 00 S & 6 S £ £ £ o o H 3 o 3 S s o 728 ELECTRIC RAILWAYS. "* CO >— ( CO ^ t^ °P iO CO CO tq q CO V. t- ! 6 to* to* rH (N* 00* d CO* d d rH - 1 CO CM iq oo q o oq \ t- l d id id rH rt* l> ^ oq Tj* d to - 1 oq Oq r-t o oq ^ to CO tH t^ CO q cq CO tq o oq \ t- d to to* rH 1> d d CO* d oq* CO *-■ Oq Oq y-i CO oq -^ to i> CD rH CO ^ I> 00 t» Tj< q CO CO \ t~ d to* id rA oi 00* co" oq* TjH 00* ^ rH oq Oq ^ 00 r-i ^ Oq **. q 00 q q q q q q oq OS % to" i> S3 rH id CO CO ^ rH ** c TjH J to CO CM 3 w o tF "*! 00 t^ OS q cq to oq to t>* Oq' rH to* d d CO* od oq' CO CO r^ r-i r>- oq CD Oq \ a £ to o to 00 tq 00 cq q q "*! CM* d * CO Oq Oq rH OS y—< os Oq # © C* OS* to* d d V. CO » °i "* CO 00 q CNJ oq o q 00 iH rH 1 to rH to* 1> r- * id I> i> "* d od \ CM (M oq rH T^ oq O os rH (-1- o iO P-! I> 00 T^ OS* to* rH t>* «* ^ 1-1 - o OS *~ (S to CM rH to* CM rH i-i OS* CO CO # -* q 00 CO y—t CNJ o o •* ,_ « CM ,_* to* I>* ^ co" 00 t^" ■*' 00* oq' \ CM CM HORSE-POWER. H.P. per Lb. Applied at Periphery at 100 Rev. per Min. Diameter Wheel. 26" 28" 30" 33" 36" H.P. .02062 .02221 .0238 .02618 .02856 Pounds at Periphery per H.P. at 100 Rev. per Min. Diameter Wheel. 26" 28" 30" 33" 36" Lbs. 48.481 45.018 42.017 38.197 35.014 Lbs 126050.9 Diam. X H.P. X Rev. H.P. = .00000793 X diam. wheel X rev. X lbs. at periphery. H.P. per lb. at periphery at one mile per hour = .002667. Lbs. at periphery per H.P. at one mile per hour = 374.9. Note on Emergency Braking* of Cars. In case of emergency, motormen often reverse the motors, which brings the car up with a severe jerk, and is quite apt to strip gears. This is not necessary, and should never be done unless the canopy switch is first thrown off, then when the motors are reversed and the controller handle thrown around to parallel, the motors will act as generators and will bring the car to an easy stop with no harm to the apparatus. In case circuit breakers are used in place of the plain canopy switches, the reversal of the motors will draw so much current from the line that the circuit breakers, if properly adjusted, will open the circuit and the controller can then be used as suggested above. COPPER WIRE EUSES EOR RAIIWAY CIRCUITS. B. &S. Gauges. 17 16 15 14 13 12 11 10 9 8 7 Fuse Point in Amperes. 100 120 140 166 200 235 280 335 390 450 520 732 ELECTRIC RAILWAYS. 8< §8 g § s 8 8 § g s •sqrj 'Apog; rfiO ^s © «© IQ 5 3 § 3 &- jo *q8i9^ O 4a O «a ~— . ^— V ^*-i g *~~ ~v~ SS 4 b0£> C3 V CN CN to CO 9 i- u 33 53 AC § 8 § O g« w ss-d '©is ^5 ^ o ©+3 B o o ■ , — ■ — . © s ^~ ^~^ v-vw H g CO -*> u a 09 a? © ! - s © - OQ «M § * O o pd *3 © ~ 5 ^ ; d s bo G CD S - 5 * a - •s^s CN CN cn CN 00 t- O t- t jo J8qum^ " v « So j ^ h CO b CN | CO •tWPTAl " " ^ « « t«- t- *- co t^ CO b- * ^ c w d Lob w fi! fa n episai ^qSi9H 00 w 9 t~+» 5 e S •spoqAV JO ozig CO CO O S = 2 2 = 2 3 I ^ 1 V N •iiv ?b»H - 2 ; » ~ ^ - CO « J8AO )qS|9H COO vh ih i »-( -*a T ~ l s s wv— , •snuo^BU 00 Is JO T^SuQri r»* Th ^f< ^f« CN rt* •s^soj CO ob 00 c^ I9AO THSu8 r X cb 00 CN 8 Jj CN ^ 8 •suijoji^u o CO o> OS M 9 joao itfSuaq CN CO CN cn CO §3 OS CN CO CN ^ ^ CM 8 ! — *-^ —- -s QQ ~* — , ~*~ N •d .2 fi • s . © . o . £^* •d a © •d © &o O 1? 03 •d a © © o so •d G © G © P. O G • © O •d * © o « © © . •d o 5 3c +3 50 ■♦» •d •d © • ,a •d © 1 -d © 2 "2 go •2 "2 d ,© 0~ g s •d js"8 o © 00 © CO GO la o rt © © © a O-G- 6 3 Q p © O O O a Q APPROXIMATE DIMENSIONS OF ELECTRIC CARS. 733 i I § © 1 o 8 8 CO 8 8 8 © 1 1 1 co CO ■*f T* 35 "«* X* CO CO lO ?s CO CO CM ^ CO CO cc K X5 : CO CO o UO g o OS CM CD "** .a CD-M .a So i © © © to" oc" M o OS co,q CD 3 — -N ,~*-V ~— , c3 oj © © aj •*h)m CO CD >> cd >, '5 3 c3 §* •s© CD J2 >>'3 4* go ©— < e3 co ©"3 3 S-l E fl ® 2 fi >.2 CD -2 tH CO CO co b- CO I- co CO £ CO CO ^ CO l> CO CO CO v CO CO CO CO v cb O o 2~£ „ „ „ c^ "* " " " U3 >b * H 5: co co CO CO ib CO CO CO CO CO CO IO CO 1—1 1 5 ^ o 5: CO t> CO CO C5 t~ CO IO oc CO CO CO CO ,H v 5: © v CO © CO s CO 8 - 3 co CO co CO 3 CO 2 l^ CO ^— s ~^~. .—»— V ,— V — ^— . © O • S . U ' 33 s-. . CD 1 3 • CO CD o -d * CD CD M • CD CD • 73 >, 'd • CD o CD . c3 CD be C3 bD be 08 X5 CD CD 3 c CD CD • •a t3 •d o o -d c3 *1 «- is CD ** a fl a £-2 3>> X5-d ©IS is © CD II CD © CO A 55 9 9 © o fl 2 S 2 ~ -= o< s o O. G, a a © <* O-© o © o © ^ ^ X fc O O o Q Q A q 4< «& S3 734 ELECTRIC RAILWAYS. M w M 8 a s V e a M H * & I 8 •a 3 « e3 o OS o P-< «-' o ■M c3 a a ^ w 3 O -t-3 pq 1-8 | n* fl S C W) o s H ofl g: : ^ 3 « g c3 : eJ oj o8 Pm H » (h © -O O O o o "3 t 8 ,D CU » m flH O cj O^ o o "*B CP c O S Q o o O Pm > O o o o o o o o o o o O © O o c 88 o o o o o o o o o o iO o to c to CO "^ to tH CO CD CO CM »0 lO co to co CM CO tO CM r^ ^ TJH rt< "tf ^ ^ to ^ ^ . CO ^ CO ^ CO "tf CO fSH CM CM 09 a fl o ^ o o 3 '•5 1> '-IS o H 8 M M t CM oo g o CD H-g §H d "H d '43 fl : : : : : : : § 2 fe £ ^ d £ H '3d wi . - - ^ a 3 X ,2 12 fl 3 '*< CD* fH g C ! : d fl c3 c d O c3 ic a3 S aa ■*a « oooooooo O O O'O o o o 88 o o O O Weigh Body Lbs. oo^oo^ooo o o o o CO o o o o o o O O Oi p_ iH CO CO, Cs, iq oo, oo, O, to, tP iq p. ^ CD <* ^ °l iO* CO* CO* t>»* I> 00 1> 00 J> fH l> 00~ os oo" |> co" 00* OS ^q" O* r-4 ?H iH i— i "^ ^ . 5;%:S:S;S:%*% S: *: S; 5: % * * % % ** >> ©OOCNOOOO O O CO CO O O ^H o o O o O OS M^ fl o CDCO0000OCDOO O O h o O CM CM to to 10 LO to to H HHH(Mrt(N(M CM CM CM CM CM CM CM CM CM N tM CM CM =35 * % % * * % * % % ^ % % ^ S: % % o O CO o o o o Tj< 1> O O CM b CM o ^ ^ k M r-t > fl CNJ CO CD CO 00 00 00 00 OS as b cm eb to o to to o3 CM CM CM CM CM CM CM .CM CM cm co co eo CO PQ co co . . 6 . . . . • -•••• 08 . . Q . . . g ' * d (J 0) 8 . 9 . 09 09 * . 8 1 fl o go f =2 a-c ft"-" =3 S o fl fl *fi .2 8 5 ? 2 : ■is Si oq W > H >> 5 a' © o 2 © & 5 a t 03 © ,• gO o3 H g J 13 ; ^ o3 1 J" G >j © d © Oi c o js i -+3 CO GO o3 r >> tf "2 ft to & © M • -d -2 °8 "S CQ ^ IS _* s Tf * T3 5 . .r d -f •g > © © d fa M -e tfg bHj '§ d * fl ■ ft ft rn CS' £ » S d « trC/J «j t— t oo3 r Q. * - g ft ©03 © JtJ © 05 d § * O PQ © >Jp2 5 - - X £ O O H 5 «° ft g M o3 S -^ -^ ^-i © O CQ £h o o o o o o o o o o M (M 05 CO (N CO t^ CO lO CO o o o o o *o o o iO h IM CO rt< »0 CO CD o o o o o © TjH Ift © O © © >C ON 0)0 h lO >0 lO lO iO N ft >> H © t g ^ csi ^ d o pH ©* 2 © ©* . 2 .S 2 3 >< V X V V 03 O o3 O O o o3 H d PQ -i> 00 J* d (N • o 2 ■ o d 43 o & £ I fc jj" © Eh j» 2 3 . ^ vox? o o d o o o Q d : o Q © © © 2 3 3 d d d o o o QflQ © 6 3 £ © o © © © lO CO © 1> CO CO 1> 00 © CO © © © © © © © © © tH CD "* o © © © © © •^f © CO © iO © N 00 h h 00 00 O CO CO »0 H ,-h IM CO 00 CO *C CO »0 N CO >0 O +3 >» bCT3 d o © lO © © © tH 03 -rf -a S3 O o PQ '3 .3 '3 o — : Is 5 : A d - CQ ^ CQ 736 ELECTRIC RAILWAYS. I 2 23 bC-^3 c o ®pq O^ W) I J t>0 ^ 3 d o O _• g OQ ^ o -° •* ^ Sm 3 2 eJ -£ . £ « tf 5 o 5 % 5: % -f »o o o o o X . c> tH 00 b ;_, c ro CO ^ TF |Q io If i— i i— I O O y- 1 * CI w o ^d^ bo o o o o o o o o o o o o oo w n a h n ^ »C M »C iC ■* c -+■> o 3 o c O S 2 3 . f f-, t" 1 Q „ .2 Jg . .2 3 3 C cj c3 C! O O a3 S E S O Q o o o © o o O O O O ." o* oT ih io* Q O O ;>> IS I s s I I I I I I I I OOOO OQOO Q O © < I ©OO© I OOOO O O I I o 00 00 00 00 0 t- t~ co t- © CO CO CO co CO CO CO O O |Oc t>. b- oo t~ t- 1>- oo t»t-t^t^ t- oo ooe •S - CN i CO CO CO i CN CN CN I COCOCOO I ilj "^"^ CO COCO b- t- f. t- W CO CO Tt< rf co co co en CM 00 00 00 00 OO OOoOOOOO CN MO00 rt ^ £3 2 ^ --:::::: - -:::::: ^ 2^2-- 888 8 ££££8 8 8833 $ eo-3883 £~ z z z z z z z z z z z z - i^^i OOO t>- COOOOO O OOOO CO COOOOOC lOiOlO CO t> t~ 00 00 rH COO HHrt iH HrtHrtM (N C<>u % u - «r3 . HHH U HHHWO .COt-t>-t>t>l>t~t>00 t- t- I l« I I ^ ^ ^ ^ ^ ^ "^ "^ cO ^ r(*MOOOOOO^O^OO«00» O O CN CNJ ^ C4 (M CN (N CO C5 CO C? C^ 53 ^ CO cS l I I I i 131 12 : : 2 §5 CO S s • • u ,§/&' ' • © 93 08 . - '2 fe • • '* ^Sg g" a 5|.|S .fgJ8 tj o»: fl 8= *§ 2 5 ^oooooooooSSSSo'aSvS a ELECTRIC LOCOMOTIVES. 739 ELECTllIC LOCOMOTIVES. The number of electric locomotives in commercial operation is rapidly increasing. The service ranges from yard shifting, for which they are particularly well adapted, up to the hauling of passenger trains of 900 tons at 60 miles per hour. The motor capacity varies from two 50 horse-power motors of the geared type up to the four 550 horse-power gearless motors on the "Mohawk" type of the New. York Central locomotive. The following list is of interest: 1907. t' fl.S 09 7jj .* $ si A Locomotives. IS, S ft ft = X 03 a o 13 > to O a) d *o o O u w 43 d & * fee i-pH — 1 g o3 i '55 ft O W) • o **> 8 o c3 S o Cayadutta .... 1894 l 35 500 500 58/17 40 /r Freight 96-Ton Baltimore & Ohio 1895 3 96 625 720 Gearless 62" Passenger Hoboken R.R, . . . 1897 1 28 500 560 54/17 40" Freight Buffalo &• Lockport . 1898 2 36 500 *300 f600 900 52/21 36" Freight Paris & Orleans . . 1898 8 55 575 78/19 49" Passenger Compagnie Francais Thomson-Houston 1899 1 38 500 600 56/17 42" Freight St. Louis & Belleville 1901 2 50 500 360 56/17 33" Freight G. E. Co. 30-Ton Yard Locomotive. 1902 1 30 250 J150 §300 72/17 36" Switching 160-Ton Baltimore & Ohio 1903 2 160 625 1600 81/19 42" Passenger Bush Terminal Co. . 1904 1 50 500 360 52/21 33" Freight G. E. Co. 40-Ton Yard Locomotive. 1904 1 40 250 J340 §680 59/18 33" Switching N.Y. C. & H. R. R.R. G. E. Co 1904 95 625 2200 Gearless 44" Passenger N.Y., N.H., & H. R.R., W. E. & M. Co 1906 II 88 600 1000 Gearless 62" Passenger * Motors in series. J 250 Volts, t Motors in multiple. § 500 Volts. Operates also on 11,000 volts, A. C. No standard electric locomotive design has been reached, although many locomotives equipped with geared motors have the general shape shown in Fig. 92 (G. E. Co.). The motors, four in number, are geared to the axle by single reduction gears and are mounted on two bogie trucks, having about six-foot wheel base. The main cab contains the controller, and the sloping ends contain the necessary starting resistances. While this design using bogie trucks is suitable for locomotives of small capacity, it is not adapted to withstand the strains to which the larger locomotives are sub- jected, hence there has been developed a type having a solid cast-steel frame containing the motors of which the later B. & O. locomotives are 740 ELECTRIC RAILWAYS. Fig. 92. Typical Electric Locomotive of G. E. Co. typical. A cross section is shown in Fig. 93 of a half unit of the B. & O. locomotive. This locomotive has a rigid wheel base and contains four geared motors of a total capacity of 1600 horse-power. It is well adapted to stand the shocks of the most severe service and handles all passenger trains in the tunnel at Baltimore. Fig. 93. Electric Locomotive used in Baltimore tunnel by B. & O. R.R. The 6000 or "Mohawk" type of locomotive adopted by the New York Central R.R., shown in Fig. 94, differs from others in having four gearless motors mounted directly upon the axles. The armatures are not even spring suspended, but are keyed solidly to the axles. The dead weight f)er axle is said to be less than. in the case of the larger types of steam ocomotives. The fields are bipolar and are so arranged that the same flux passes through the four sets of fields in series, returning partly through .the side frames and partly through an overhead longitudinal frame. The departure from the previous methods of construction, using geared motors, is pronounced, and exhaustive tests seem to prove its wisdom for the pro- posed service. In Fig. 95 are given the motor characteristics of the 550 ELECTRIC LOCOMOTIVES. 741 W ON CO it i° is o O I-} 03 CO 3-9 •8 m p oS ^^ 742 ELECTRIC RAILWAYS. 2 , 1 N> X. Lo CO ^10 riv E ( IE. 84 n ore )F? i $ J % § Pfl ni *m ?te ri sti C ( ?u rr es $ N & 2 VAi 10, r w, V£l is 44 //K 1 y " s 1 6 oo I/O. f .T6 J3 £ 1 1 t I 3 I \ / \ 1 f / \ / \l \ , / f \ - / \ V / \ \ L ty c > s Y *s^^ tgl "y^ Jb / ^2 A V i 4 / 7 / / / / i ' £c >0 4C SC 10 Si W /c 00 /£ 00 /IA7* 9 £fi £S P£6 'Sft ?ra ? Fig. 95. ELECTRIC LOCOMOTIVES. 743 horse-power motor. Fig. 96 gives a specimen speed run of the 6000 loco- motive hauling a train of 336 tons or a total train weight of 431 tons, in- cluding the locomotive itself. The speed reached, 63 miles per hour, has since been greatly exceeded, one run being made during which a speed of 84 miles per hour was recorded. A locomotive which is of particular interest is that shown in Fig. 97. This is equipped with four 250 horse-power single-phase gearless motors, which are arranged for operation on either 600 volts direct current or 11,000 volts single-phase alternating current. This locomotive is the first of thirty-five (1907), which the Westinghouse Electric and Manufacturing Co. has supplied to the N.Y., N.H. & H. R.R. It is of the double-truck type and has two swiveling trucks with a wheel base of 8 feet each and a distance between truck centers of 14 feet 6 inches. P "el frn in ir r ' ¥ eed f R un J \ NY C. -OC -.Of OT ivE *e po V 'PC* ■ss -. N 8C \Loc 1J? on //V 33 9 6 r 3 7/V3 s \ , ■'' \ 0/4 v o 'or, U JfS ^3 4 t. h • \ \ y\ otr 4o to & w J. 'O Jt. o +< ■>o Si C t >/v, 73 Fio. 96. Preliminary Speed Run of N. Y. C. Locomotive 6000. November. 1907. This arrangement eliminates any danger of nosing and insures easy riding. The armatures of the motors are built on hollow shafts or quills which surround the axles and are connected to the driving wheels on each side by seven driving horns. These driving horns are surrounded by springs and fitted into pockets in the wheel hubs. The frames of the motors are spring supported from the journal boxes. The wheels are 62 inches in diameter and thus raise the center of gravity of the locomotive high above the track, similar to that of a steam locomotive. This arrangement, to- gether with the spring supports of the motors, makes the locomotive par- ticularly easy on the track. The general shape is such that ample room is obtained in the cab for mounting all apparatus so that it is readily The tracks of the N.Y., N.H. & H. R.R. are equipped with an 11,000 volt overhead trolley wire, but those of the New York Central Railroad, 744 ELECTRIC RAILWAYS. INSTALLATION OP ELECTRIC CAR MOTORS. 745 over which the trains must run from Woodlawn Junction to Grand Central Station, are equipped with a direct current third rail. For this reason these locomotives are arranged to operate from either of these conductor! and to change from one to the other without slackening speed. The motors are cooled by means of an air blast forced through them by motor-driven blowers in the cab and on this account they are capable of developing 200 horse-power each continuously, although an ordinary rail- way motor of the same nominal rating could operate continuously at only about 110 horse-power. The performance of the motor is shown by the curves in Fig. 90, p. 717. The weight of the locomotive complete is approximately 88 tons. A single unit is capable of handling a train of 200 tons in local service or a train of 250 tons in through service, and two or more units may be readily coupled together and operated as one for handling heavier trains. IA8TA114TIOA OF EIECTRIC CAR MOTORS. (General Electric Company.) In General. In locating the various parts of the equipment and in wiring the car, par- ticular attention should be taken to secure the following results : 1. Maintenance of high insulation. 2. Exclusion of all foreign material, particularly grease, dirt, and water, from the electrical equipment. 3. The avoiding of fire from arcs, naturally occurring at fuse-box, light- ning arrester, etc. 4. The prevention of mechanical injury to the parts. 5. The placing of the parts so as to be accessible for operation and inspec- tion, and yet out of the way of passengers. Preparation of the Car Body. The floor should be provided with a trap-door of such size as to allow as free access as possible to the motors. Particular attention is called to the advisability of having the bar across the car between the trap-doors remov- able, in order that the top of either motor can be thrown back. The roof should be provided with a trolley board which strengthens it, and protects in case the trolley is thrown off; it also deadens the noise. A firm support should be provided for the light clusters. Grooves should be cut for the leading wires in the roof molding, and also in two of the corner posts, one for the trolley wire, the other for the ground wire of the lighting circuit. On a closed car four 2-inch holes should be bored through the car floor under the seats, one as near each corner of the car as possible. On one side of the car, four f-inch holes should be bored in a line, and 4 inches apart, to receive the taps from the cable to the leads of motor No. 1. The exact location of these holes depends on the type of motor used. The distance from the center of the axle to the center of this group of holes should be about two and one-half feet for G. E. motors. On the same side of the car, and in the same line, four other f-inch holes should be bored 4 inches apart, to receive the taps from the cable to the resistance boxes. On the other side of the car three f-inch holes in a line and 4 inches apart should be bored to receive the taps from the cable to the leads of motor No. 2, and on same side of car and in the same line five other f-inch holes 4 inches apart should be bored to receive the taps for the trolley, resistance, and shunt for Motor No. 2. Reference should be made to diagram in order that each set of holes shall be on the proper side of the car, and at such a distance from side-sills as to be out of the way of wheel throw. 746 ELECTRIC RAILWAYS. Measuring about 38 inches from the brake-staif and a suitable distance inside of the dash rail, an oval hole 5 in. x 2| in. should be cut in each plat- form to receive the cables. On an open car no holes need be bored for the floor wiring except those through the platform. Installing: Controllers. In the standard car equipment one controller is placed on each platform on the side opposite the brake handle, in such a position that the controller spindle and the brake-stall shall not be less than 36 inches, nor more than 40 inches apart. The exact position depends somewhat on the location of the sills sustaining the platform. The feet of the controller are designed to allow a slight rocking with the spring of the dasher. Two one-half inch bolts secure the feet to the platform. An adjustable angle iron is furnished to be used in securing the controller to the dash-rail. A wire guard is also furnished, to be secured to the platform in such a position that the cables pass through it into the controller. A rubber gasket is furnished with each controller, to be placed between the wire guard and the platform,to exclude water. For dimensions of controller, see Figs. 104 and 105. Wiring. This work can be conveniently divided into two parts; namely, roof wiring- and floor wiring*. Roof -wiring* includes the running of the main circuit wire from the trolley through both main motor switches down the corner posts of the car to a suitable location for connecting to the lightning arrester and fuse box ; also wiring the lamp circuit complete, leaving an end to be attached to the ground. Whenever wires lie on the top of the roof, they need not be covered with canvas or moulding, except to exclude water where they pass through the roof. In such cases a strip of canvas the width of the moulding, painted with white lead, should be laid under the wire, and over this and the wire should be placed a piece of moulding extending far enough in either direction to exclude water. The moulding should be firmly screwed down and well painted. The above wiring should be done if possible while the cars are being built. Floor wiring* may be done after the car is completed without injuring the finish. IfEade np cables give far better protection to the wiring, and are easier to install than separate wires, and should be used in the floor wiring if possible. The simplest way of installing them on box cars seems to be as follows : After the car bodies are prepared according to the above instructions, the cables (one on each side of the car) should be run through holes in the plat- form, and the connections made to the motors and controllers. After making connection to the controllers, all slack should be pulled up inside of the car under the seats, and held in place, preferably against the side of the car, by canvas or leather straps. Motor taps should project through the sills for attachment to the flexible motor leads just far enough to permit easy connection, leaving as little chance as possible for vibration. No rubber tubing will be required on taps, as they all have a weather-proof, triple-braided cotton covering outside of the rubber insulation to prevent abrasion. All joints should be thoroughly soldered and well taped. The portions of the cables passing under the platforms should be supported by feather straps screwed to the floors or sills. Cables should never be bent at a sharp angle. The ground wire should run under the car floor rather than under the seats. On open cars all wires and cables must be run under the car, and should be well secured to the floor with cleats or straps. A good joint can be made by separating the strands of the tap-wire, and INSTALLATION OF ELECTRIC CAR MOTORS. 747 wrapping the two parts in opposite directions around the main wire. Both Okonite and rubber tape are furnished. It is desirable that Okonite should be used first and rubber tape put over it, as the latter will not loosen and unwrap as Okonite will. All openings in the hose should be sewed up as tightly as possible around the wires. Separate wires can be installed if necessary, observing the following directions : ' The floor wires on box cars should be placed under the seats as much as possible. In the few places where it is necessary for wires to cross, wood should intervene in preference to a piece of rubber tubing or loop in the air. This rubber tubing is not necessary where wire is cleated under the floor (as on open cars), if it does not pass over iron work, or is not ex- posed to mud and water. Where so exposed, it should be covered with moulding, but where moulding is used it should be carefully painted inside and out with good insulating compound to exclude water. The wire passing to the fuse box should be looped downward to prevent water running along the wire and into the box. Care should be taken to avoid metal work about the car in running the wires, and that nails or screws are not driven into the insulation. In general it is not desirable to use metallic staples and cleats for car- wiring, except about the roof, or inside the car. Where wires are subject to vibration, as between the car bodies and motors, flexible cable must al- ways be used. A certain amount of slack should be left in the leads from the motor to the car body, depending on their length. On cars with swivel- ing trucks a greater amount of slack is necessary. As slack gives greater opportunity for abrasion, care should be taken to leave only what is abso- lutely necessary. Operation and Care of Controller. When starting, regulate the movement of the handle from point to point so as to secure a smooth acceleration of the car. Do not ran between points. The resistance points 1st, 2d, 3d, 6th, and 7th, are intended only for the purpose of giving a smooth acceleration, and should not be used contin- uously. For continuous running, use the 4th, 5th, 8th, and 9th points, which are shown by the longest bars on the dial. When using the motor cut-out switches be sure that they are thrown up as far up as they will go. In case the trolley is off and the hand-brakes do not hold the car, an emergency stop may be accomplished by reversing the motors, and turning the power-handle to the full speed, or next to full speed point. To examine the controller, which should be done regularly, open the cover, remove the bolt with wrench attached, and swing back the pole-piece of the magnet. The contact surfaces and fingers should be kept smooth, and occasionally treated with a small amount of vaseline to prevent cutting. All bearings should be regularly oiled. A repellent compound, paraffine, rosin, and vaseline, equal parts by weight, placed in the water-caps of the power and reversing shaft, is an eflicient protection against water. Dirt must not be allowed to collect inside of the controller. Diagrams of Car Wiring*. In general car wiring is carried out in about the same manner for all styles and sizes of car, more particular description being given above. Wir- ing differs mainly in details, governed by the number, style and horsepower of motors used. 748 ELECTRIC RAILWAYS. Diagrams of standard wiring for two motors per car and for four motors per car follow in Figs. 98, 99, 100, 101. They are all from the G. E. Co. lists, as controllers made by that Company are almost universally used, although many of older design by other companies are still in the held. **«--- -turnip * s CO •" £ 9 £3« o * WIRING DIAGRAM OF ELECTRIC CAR MOTORS. 749 o O 2 O "J ■a ■ ^Hfs^® I 1 FOB IN *.' 750 ELECTRIC RAILWAYS. CO ca O oS s- °l fS §2 2* 32g -I z S og . O H o- *1 £s O Q 2S CCO > cc > <3 K < O WIRING DIAGRAM OF ELECTRIC CAR MOTORS. 751 752 ELECTRIC RAILWAYS. EaUIPMEM LISTS. The following is a list of material required for the electrical equipment of one car fitted with two motors: QUANTITY. 1 Trolley pole. 1 Trolley base. 2 Motor circuit switches. 1 Lightning arrester. 1 150 ampere magnetic cut-out (fuse-box). 1 Resistance box. 1 Resistance box. 1 Core for kicking coil. 2 Controllers (includes wire guard and gasket, supporting bracket, cap screws, and washers for fastening to dasher). 1 Controlling handle. 1 Reversing handle. One of each of these handles is always shipped with each pair of controllers unless specified to the contrary. No. 6 B. & S. strand wire (7-. 061 in.) for roof-wiring. 100 or 150 ampere fuses. Two-way connectors, i-inch hole, No. 6. Brass corner cleats, T Vinch slot. Brass flat cleats, r Vmch slot. £-inch No. 4 R. H. brass wood screws for brass cleats. Wood cleats, J-inch slot. Wood cleats, f-inch slot. li-inch No. 8 R. H. blued wood screws for wood cleats. Solder. f-inch Okonite tape. 1-inch adhesive tape. Material for set. of cables as follows: No. 6 B. & S. strand wire (7-. 064 inches), single braid. No. 6 B. & S. strand wire (7-. 064 inches), triple braid for taps. Brass marking- tags. 1^-inch cotton hose. Rubber tape. Paragon tape. Solder. This material can be procured made into a "set of cables" with- out extra cost. Car-lighting equipment. 75 ft. 20 10 30 25 1 1 n 11U 25 25 i r\r\ 1UU lib. 1 lb. 1 lb. 480 ft. 100 ft. 41 64 ft. 1* lbs 4 lbs. H lbs. CONTROLLERS. 753 CONTROLLERS. Under this heading are included all that type of appliance used for starting and stopping the motors and controlling the speed of the same. As almost all the old forms of rheostat with different steps have been abandoned for the so-called series-parallel controller, it is not necessary to describe any other here, nor will any detailed description of those now in use be attempted. But one form is now in general use, viz., the magnetic blow-out type, made by the General Electric Company and used also by the Westinghouse Electric and Manufacturing Company. The principle of the magnetic blow-out type was first developed by Prof. Elihu Thomson, i.e., that an electric arc in a strong magnetic field is blown out of line and extinguished or cut in two. This fact is taken advantage of in the controller of the General Electric Company by using a strong electro- magnet to extinguish the arcs formed at the contact-points, when the circuits are broken. The construction is shown in the cut of series-parallel controller, Form K2, following. Controllers are now made in so many forms and varieties that it is im- possible to give more than a few of the combinations which are practically the same everywhere in the United States. Fig. 102. Series-Parallel Controller, Form K2. General Electric Company. Used also by the Westinghouse Electric and Manufacturing Company, and others. 754 ELECTRIC RAILWAYS. The General Electric Company manufactures controllers for all conditions of electric railway and power service. They are divided for convenience in designation into five general classes, each designated by an arbitrary letter. Type JB Controllers may be of either the series parallel or rheo- static type, but always include the necessary contacts and connections for operating electric brakes. Type K Controllers are of the series parallel type and include the feature of shunting or short circuiting one of the motors when changing from series to parallel connection. Type JL Controllers are also of the series parallel type, but completely open the power circuit when changing from series to parallel. Fig. 103. " R " Type of Rheostatic Controller. Type It Controllers are of the rheostatic type and are designed to control one or more motors by means of resistance only. The Type Iff Control System developed by the General Electric Company with particular reference to the operation of motor cars in trains, is also suitable for operation of large equipments, where the size and weight of a cylinder type controller are objectionable. This system of control consists essentially of a number of electrically operated switches called "contactors" that close the various power and motor circuits, and which are in turn controlled by small master controllers which are called upon to carry only the current for the operating coils of the contactors. The motors are reversed by electrically operated reversing switches also controlled by the master controller. Where equipments are operated together in trains, the control circuits are connected between adjacent cars by suitable couplers and the operation of the contactors and reversers on all the cars in the train are controlled simultaneously from any master controller on the train. CONTROLLERS. 755 Series Parallel Controllers. Title. Capacity. Controlling Points. Remarks. K-2 Two 40 h.p. Motors. 5 Series. 4 Parallel. For motors using loop or shunted field only. K-4 Four 30 h.p. Motors. 5 Series. 4 Parallel. For motors using loop or shunted field only. K-6 Two 80 h.p. Motors or Four 40 h.p. Motors. 6 Series. 5 Parallel. K-10 Two 40 h.p. Motors. 5 Series. 4 Parallel. K-ll Two 60 h.p. Motors. 5 Series. 4 Parallel. Similar to K-10 but has connecting wires and blow-out coil of larger capacity. K-12 Four 30 h.p. Motors. 5 Series. 4 Parallel. Similar to K-ll but has reversing switch arranged for four motors. K-13 Two 125 h.p. Motors. 7 Series. 6 Parallel. K-14 Four 60 h.p. Motors. 7 Series. 6 Parallel. K-27 Two 60 h.p. Motors. 4 Series. 4 Parallel. Similar to K-ll but is arranged for oper- ation on metallic circuit, having con- tacts for opening both sides of the circuit. K-29 Four 40 h.p. Motors. 6 Series. 5 Parallel. Similar to K-6 but is arranged for operation on metallic circuit, having contacts for opening both sides of the circuit. K-31 Four 30 h.p. Motors. 4 Series. 4 Parallel. Similar to K-27 except has reverse switch arranged for four motors. K-32 Two 40 h.p. Motors. 4 Series. 4 Parallel. Similar to K-27 except has connecting wires and blow-out coil of smaller capacity. L-2 Two 175 h.p- Motors. 4 Series. 4 Parallel. L-3 Four 150 h.p. Motors. 8 Series. 7 Parallel. L-4 Four 100 h.p. Motors. 4 Series. 4 Parallel. Similar to the L-2 but with additional reversing switch parts for four motors. L-7 Four 200 h.p. Motors. 9 Series. 6 Parallel. Electric Brake Controllers. Title. Capacity. Controlling Points. Remarks. B-3 Two 40 h.p. Motors. 4 Series. 4 Parallel. 6 Brake. Superseded for general use by the B-13. B-7 Two 100 h.p. Motors. 6 Series. 5 Parallel. 6 Brake. Has separate brake handle. B-8 Four 60 h.p. Motors. 6 Series. 5 Parallel. 7 Brake. Has separate brake handle. B-13 Two 40 h.p. Motors. 5 Series. 4 Parallel. 7 Brake. Supersedes the B-3 from which it differs in that the braking connections are such as to render the skidding of the car wheels practically impossible. ELECTRIC RAILWAYS. Electric Brake Controllers.— Continued. Title. Capacity. Controlling Points. Remarks. B-18 Two 40 h.p. Motors. 4 Series. 4 Parallel. 6 Brake. Similar to B-3 but arranged for rheostatic braking only. B-19 Four 40 h.p. Motors. 5 Series. 4 Parallel. 7 Brake. Similar to B-8, having separate handles for power and brake. Supersedes B-6. B-23 Two 60 h.p. Motors. 5 Series. 4 Parallel. 7 Brake. Similar to the B-13 but has connecting wires and blow-out coil of larger capac- ity. B-29 Two 60 h.p. Motors. 5 Series. 4 Parallel. 7 Brake. Similar to B-23 but has separate brake handle. Electric braking is made little use of owing to the fact that it adds con- siderably to the heating of the motors. The conditions are such that the motors are already over-taxed and the use of brake controllers necessitates an increase in the size of motor required. Air-brakes are in almost uni- versal use on the heavier cars owing to their smaller expense of installation. Rlieostatic Controllers. Title. Capacity. Controlling Points. Remarks. R-ll One 50 h.p. Motor. 6 For motors using shunted field for running points only. R-14 Two 35 h.p. Motors. 5 Very short and specially adapted to mining locomotives. Motors connected perma- nently in parallel. R-15 Two 80 h.p. Motors. 6 Motors connected permanently in parallel. R-16 Four 40 h.p. Motors. 5 Similar to R-15 but has reversing switch arranged for four motors. Motors con- nected permanently in parallel. R-17 One 50 h.p. Motor. 6 R-19 Two 50 h.p. Motors. 6 Similar to R-17 but has reversing switch arranged for two motors. Motors con- nected permanently in parallel. R-22 Two 50 h.p. Motors. 5 Similar to R-14 but has connecting wires and blow-out coil of larger capacity. R-29 Four 25 h.p. Motors. 6 Similar to R-19 but has reversing switch arranged for four motors. Motors con- nected permanently in parallel. R-37 Two 50 h.p. Motors. 5 Similar to R-22 but has extra contacts on the reversing switch for connecting the motors either in series or parallel. R-38 Two 35 h.p. Motors. 5 Similar to R-37 but has connecting wires and blow-out coil of smaller capacity. R-48 Four 75 h.p. Motors. 8 R-55 Two 150 h.p. Motors. 7 Has series parallel reversing switch same as R-37. It is specially adapted to mining locomotive service. These controllers are used with single motor equipments or for loco- motive work where the speed is very low, as in yard shifting service. CONTROLLERS. 757 8 ft ►4 & >> 3 HceNaowHt . . H-* . H»HH . H«*w|ceio|ao .... OOt> . . .00 . . »0 . . .00 . . W N . 05 H 00 . . . . ^H • , n<l»»4»HlM«H' ^ '. H^H-Hr-Jr-.*^ iONM^HfO^O5CDCON»O00C0 , "5 N ?D ^ CO T|H CO?DN^OOOO . WJ N CO rt< CO ^ IN (N H H M H<» wh* * * °V< ! 1 csh*eoNM CD «* CO rj( (N (N H H o 3 C0t>*O00O0 . iO I> CO "iO(ON"500CO . UJ N <0 ■* CO «* r^r*< 758 ELECTRIC RAILWAYS. |*o Fig. 104. Type K. Fig. 105. Type L. J D. j i r . i i H| I ^^° Fig. 106. Type B. Type R. Diagrams for Dimensions of Controllers. CONTROLLERS. 759 o ft >> CO h N CO . . .05 . .cONiOOOOO^l^O^ . CM CM tH tH CM CM " ' 00 ' "tONiOOOOOiON©^ CM CM rH rH 05 1 m K°|2hqoio|qo HNnlSlSKHSiolooHNeoHi HrnV^HSrlSHS | | WNW^HCOH00i CM . ' b- ' ©N»O0Cl00iON«O , 1 ( CM CO i-i i-i o 3 CO N N . . . rH . .CDNiOOOOOiON©^ . . CO T* rH rH i P3 "PH^Hao ; ; ; Ha ; ; >4x>h CM . . .00 . .CONiOOOOOiCNCOi 1 . . CM CM r-l rH s CONMtJIHCOHOO^ONiOOOOOiONO^ . . CO ^N (N H H o M s PQ cu a >» H 05 CM 1 pq HrH (»N(lOl00 HllNHlrHwIoOiHleOHlC^rHHrWlH* r^H^lOfoOr^pHoOnrHCCl^ 0005CO^rHCOCOOCOrH(MC00005iOCOi0^iOOO CO ^t CO M IM (N CO CM 1 pq rH|"#ifl|oo Hc^H^ojooHooiHlc^Hoofoh* co|aoH^Hoo ">h | CX>ffl^^HCOOO«H(NiOOJ05iOINiOiO . . CO «tf CO H ^p *>< M|H< COO5COTHrHCOTj»O0505»OCNiOi0 . . CO tH CO ^^^S 760 ELECTRIC RAILWAYS. MOTOR COMBINATIONS RES. MOTOR 1 MOTOR 2 "2-mO— O— AW/J— O — WA-I— i-^Ql}— O— WW^-O w^J— -^QD- po— ^v^f -Q — w-l- X^- pO — ww-Jp o — wA-1— -L-Cnj ^-o — "w^p o — w^l— --Ml[} i T O n -W^|-3|m^[ C<^^ 17^ r-O-W^-,1 f-Q-W^-, I Fig. 108. SERIES L2 CONTROLLER MULTIPLE RES. MOTOR 1 MOTOR 2 B — O — VW — O — W/>— g O — WW— O — WW— ' OPEN CHANGES TO MULTIPLE 8EE NEXT COLUMN Fig. 109. B K HUh HK>-Wtf- r jj rO-W-l B K HW-i g u o-w*- r r-O-W^-L. ELECTRIC MULTIPLE UNIT CONTROL. 761 THE SFRAGUE GENERAL ELE( THK MULTIPLE OIT CONTROL. The multiple unit control is designed primarily for the operation of motor cars in trains. Motor cars and trail cars may be coupled in any combination and the whole operated as a unit from any controller on the train. The system may also be used to advantage on individual equipments and loco- motives. The control apparatus for each motor car may be considered as consisting essentially of a motor controller and a master controller. The former comprises a set of apparatus, — usually located underneath the car, — which handles directly the power circuits for the motors, con- necting them in series and parallel and commutating the starting resistance in series with them. This motor controller is operated electrically and its operation in establishing the desired motor connections is controlled by the motorman by means of the master controller. The latter is similar in con- struction to the ordinary cylinder controller and is handled in the same manner, but instead of effecting the motor combinations directly, it merely controls the operation of the motor controller. The latter consists of a number of electrically operated switches, or " con- tactors" which close and open the various motor and resistance circuits, anol an electrically operated "reverser" that connects the field and armature leads of the motors to give the desired direction of movement of the car. Both the contactors and reverser are operated by solenoids, the operating current for which is admitted to them by the master controller. In addition to the motor and master controllers, each motor and trail car is equipped with train cable consisting of nine or ten individually insulated conductors connected to corresponding contacts in coupler sockets located at each end of the cars. This train cable is connected identically on each motor car to the master-controller fingers, and the contactor and reverser operating coils ; and the train cable is made continuous throughout the train by couplers between the cars, connecting together corresponding terminals in the coupler sockets. All wires carrying current supplied directly from the master controller fprm the "control circuit;" those carrying current for the motors, form the "motor" or "power circuit." Inasmuch as the motor controller operating coils are connected to this control train line, it will be appreciated that energizing the proper wires by means of any master controller on the train, will simultaneously operate corresponding contactors on all the motor cars, and consequently establish similar motor connections on all cars. In case the "power" circuit is momentarily interrupted for any reason, the system of control provides for the immediate restoration of the motor and resistance connections, which were in effect immediately preceding such interruption. Should the motorman remove his hand from the operating handle of the master controller, the current will immediately be cut off from the entire train, thus diminishing the danger of accident in case the motorman should suddenly become incapacitated. The system must be supplied with a potential of at least 300 volts to insure successful operation. The approximate total weight of control equipments, exclusive of supports is as follows: Aggregate H.P. of Motors. Weight of Equipment in Pounds. 125 1500 250 2000 400 3000 500 4500 800 5000 The approximate weight of the apparatus for each trail car, which included tram cable, coupler sockets and connection boxes, is 100 pounds. ; The position of the handle on that master controller which the motorman is operating always indicates the position of motor-control apparatus on all cars. The motor controller which handles all the heavy arcing is located underneath the car. 762 ELECTRIC RAILWAYS. Apparatus. Contactors. — The contactors are the means of cutting in and out the various resistance steps, of making and breaking the main circuit between trolley and motors and of changing from series to parallel connection. Each contactor consists of a movable arm carrying a renewable copper tip which makes contact with a similar fixed tip, and a coii for actuating this arm when supplied with current from the master controller. The contactor n&c£./or->s of /''or- r*vo aOO H. /=> //totorj Cot//D/(?r JocA«ts Fig. 110. is so designed that the motor circuit is closed only when current is flowing through its operating coil; and gravity, assisted by the spring action of the finger, causes the arm to drop and open this circuit immediately, when the control circuit is interrupted. In order to save space and eliminate interconnections as much as possible, several contactors are mounted on the same base. The contactors should preferably be located under the car, and boxes are therefore supplied which facilitate installation, protect the contactors from brake-shoe dust and other foreign material, and provide the necessary insulation. Reverier. The general design of the reverser is somewhat similar to that of the ordinary cylindrical motor-reversing switch with the addition of electro- magnets for throwing it to either forward or reverse position. In general construction, the operating coils are similar to those used on the contactors, but in order to secure reliability of action the coil is given full line potential. The reverser is provided with small fingers for handling control-circuit connections and when it throws, the operating coil is disconnected from ground and is placed in series with a set of contactor coils, thus cutting the ELECTRIC MULTIPLE UNIT CONTROL. 763 ^ b* 764 ELECTRIC R'AILWAYS. operating current down to a safe running value. These coils are protected by a fuse, which will open the circuit if the reverser fails to throw. If the position of the reverser does not correspond to the direction of movement indicated by the reverse handle on the master controller, the motors on that car cannot take current. While the motors are taking current the operating coil is energized, and the electrical circuits are interlocked to prevent possi- bility of throwing. * L 0/m€>03/ons of C-2S Cont rotter form A iOn// Fig. 112. Master Controller Sprague G. E. Multiple Unit System. Master Controller, — The master controller is considerably smaller than the ordinary street-car controller, but is similar in appearance and method of operation. Separate power and reverse handles are provided, as experience has led to the adoption of this arrangement in preference to providing for the movement of a single handle in opposite directions. An automatic, safety, open-circuiting device is provided, whereby, in case the motorman removes his hand from the master-controller handle, the control circuit will be automatically opened by means of auxiliary contacts in the controller, which are operated by a spring when the button in the handle is released. This device is entirely separate and distinct in its action from that of the main cylinder. Moving the reverse handle either forward or backward makes connections for throwing the reverser to either forward or backward position. The handle can be removed only in the intermediate or off position. As the power handle is mechanically locked against move- ELECTRIC MULTIPLE UNIT CONTROL. 765 ment when the reverse handle is removed, it is only necessary for the motor- man to carry this handle when leaving the car. When the master controller is thrown off, both line and ground connections are cut off from the operating coils of important contactors, and none of the wires in the train cable are alive. The current carried by the master controller is about 2.5 amperes for each equipment of 400 horse-power or less. Fig. 113. Details of Top of Master Controller Sprague G. E. Co., Multiple Unit System. Master Controller Switch. — A small enclosed switch with magnetic blow-out is used to cut off current from each master controller; and it is supplied with a small cartridge fuse enclosed in the same box. When this switch is open all current is cut off from that particular master controller which it protects. Bridge Connection. — A noteworthy feature of the control is the method of accomplishing the series-parallel connection of the motors. This is by the so-called "Bridge" method of connections, which are so arranged that the circuit through the motors is not opened during the tran- sition from series to parallel and substantially the full torque of both motors is preserved at all times, from the series to the full parallel connection. This connection does away with any serious falling off in the rate of acceleration which is sometimes noticed when the motor circuit is interrupted during transition from series to parallel in other methods of control. The " Bridge connection is therefore particularly adapted to high rates of acceleration which can thus be sustained throughout the accelerating period without causing discomfort to passengers. 766 ELECTRIC RAILWAYS. 7/7o-£,or- C/'rca/d; Com^/r-j or*" . sS'proyc/e- Senero/ £/ectr/c /77v/t,//=>/G Ury/^. Gr"tct$e 70't.0/-? r -<7Hvv L -'-^r^^ ?-w~© — L-ophpku^ir — i*~c/// 5er/ o «^vvHi)--^nWphrHr^^ r ' Fig. 114. WESTIXCWHOI'**: OIT SWITCH SYSTEMS OJF JJKITLTIPJLE E-aUIPUIEIVT. Rates Stated toy Chicag-o City Railway in u Street Railway Journal," Dec, l&OS. Engines, 8 per cent ; Boilers, 8 per cent ; Gene- rators, 3 per cent ; Buildings, 5 per cent. Cable machinery, 10 per cent ; Cables, 175 per cent. Rails, 5.5 per cent ; Ties, 7 per cent. Granite, 5 per cent ; Cedar blocks, 16 per cent ; Brick, 7 per cent ; Asphalt, 7 per cent ; Macadam, 6 per cent. Car bodies, 7 per cent ; Trucks, 8 per cent. Armatures, 33 per cent ; Fields, 12 per cent ; Gear cases, 20 per cent ; Controllers, 4 per cent ; Com- mutators, 33 per cent. Wiring and other electrical equipment, 8 percent. Iron poles, 4 per cent ; "Wood poles, 8 per cent ; In- sulation, 12 per cent ; Trolley-wire, 5 per cent ; Trolley insulation, 7 per cent ; Bonding, 8 per cent. All based upon renewals and per cent of wear. CAR MEATlirO RY ELECTRICITY. Test on Atlantic Avenue Railway, Rrooklrn. Power-Station. Cable Jflachinery Roadbed. Paving*. Cars, Rolling* Stock Line Equipment. Cars. Temperature F. Watts Doors. Windows. Contents, Cu. ft. Outside. Average in car. Consumed. 2 12 850£ 28 55 2295 2 12 850£ 7 39 2325 2 12 808£ 28 49 2180 2 12 913£ 35 52 2745 4 16 1012 7 46 3038 4 16 1012 28 54 3160 TRACK RETURN CIRCUIT. 771 TRACK RETURN CIRCUIT. It goes without saying that the return circuit, however made, whether through track alone or in connection with return feeders, should be the best possible under the circumstances. Few of the older roads still retain the bonds and returns formerly considered ample and good enough. Electrolysis and loss of power have compelled many companies to replace bonds and return circuits by much better types. The British Board of Trade paid especial attention to the return circuit in the rules gotten out by them (see page 781), and many American railroads would have been much in pocket to-day if such rules had been promulgated in the United States at the beginning of the trolley development. With few exceptions the practice of engineers has been to connect the rail joints by bonds, both rails of a track together at intervals, and both tracks of a double-track road together. To this has sometimes been added track return wires laid between the rails, and in other cases return feeders from sections of track have been run to the power house on pole lines or in ducts underground. The writer favors the full connection return with frequent insulated over- head return feeders where there may be danger from electrolysis of water and gas pipes ; in fact, ample return circuit has been proved time and again to be the only preventive of that trouble. On elevated railways where the structure is used for the return, the ends of abutting longitudinal girders are likewise bonded together at the expan- sion joints. Tests have shown that the riveted joints, where well riveted, have a conductivity nearly equal to that of the girder itself, hence it is not necessary to bond them. The return circuit of the New York Subway is designed for an extreme drop of five volts. Careful and continuous attention should be given to bonds from the moment cars are started on a line. SINGLE TRACK TURNOUT DOUBLE TRACK TURNOUT r* 4 ^ %*s ^ ^ J I ] I .) i L ^ »1 7^*^ * «; J Z2 i_ / \ . , t \ • 3T\ ) ( =^~ £ r = T JLL= CROSSING OF TWO ELECTRIC ROADS CROSSING OP ELECTRIC AND STEAM RQAD$ Fig. 116. Showing Cable Connections for Bonding Around " Special Work." Dr. Bell gives the following ratios of track return circuit to overhead system as being average conditions. Let R\ = resistance of track return circuit, and R = resistance of overhead system. Then R x = . 1 to . 2R. R x = . 2 to . 3R. R t = . 4 to . QR. R t = .2 to .3R. R t = .3 to .7R. R t = .7 to I. OR. Exceedingly good track and very light load. Good track and moderate load. Fair track, moderate load. Exceptional track and large system. Good track, large system. Poor track, large system. 772 ELECTRIC RAILWAYS. In exceptional cases track resistance may exceed that of overhead system. It is sometimes assumed that R ± = .25R, but this is rather better than usual. Under ordinary conditions R x = . 4R is nearer correct. If formula for copper circuit = cm. = ^ — : then for R t = . 4R, the constant 11 should be increased to between 14 and 15 in order that copper drop may bear correct proportion to that of the ground return. Type of lloiulv (By F. R. Slater.) Bonds are divided into two general classes. (1) those which are fastened to the surface of the rail or girder to be bonded, commonly called "soldered" bonds, and (2) those having terminals with a shank which is expanded into a hole in the rail or girder to be bonded, commonly called "riveted" bonds. In both classes that portion which is attached to the rail is called the terminal, the remainder the body of the bond. Soldered Bonds. — These are formed in various ways but in general by a series of thin strips of annealed copper bent in the form of an ~w i°a m Fig. 117. Soldered Bond. Fig. 118. Bond Attached to Base of Rail by Soldering only. arch for the greatest degree of flexibility, with a pair of feet or terminals to provide contact surface. The strips of each foot are soldered or welded together, making a solid terminal, while the intermediate strips of the arch are free and unattached to each other so that they can readily take up vibra- tions. Figs. Ill and 112 illustrate this type. Shawmut Soldered Bond. — This bond is constructed of copper laminations .023 inch thick, the ends separately tinned, clamped together Fig. 119. Soldered Bond Applied to Head of Rail. Fig. 120. Soldered Bond Applied inside of Angle Bar. TYPE OF BONDS. 773 X and dipped, then covered with a tinned wrapper, thus insuring perfect union when heated, and the form of construction assuming a great degree of flexibility. In applying soldered bonds too much care cannot be exercised. The rail must be cleaned perfectly at the point of application and then tinned. The bond is then clamped in position and heat applied to both feet at once by means of a double burner gasolene torch, the solder being applied with zinc chloride flux. Bonds can be applied to the ball, web, or base of a rail, and each of the feet of the bond should be able to withstand a mechanical strain of two thousand pounds shearing stress, the electrical resistance not exceeding that of more than three feet of the rail to which it is applied. Fig. 121. Soldered Bond Applied to Base of Rail. Q Fig. 122. Soldered Bond Applied to Head of Rail. Fig. 123. Soldered Bond Applied to Flange of Rail. Figure 124 shows the result of tests made on three sizes of soldered bonds 250,000 cm., 370,000 cm., and 640,000 cm., to determine at what current the bond would melt off. The rise in temperature at the terminals and center of the bond is given. The 640,000 cm. bond melted off at 5,500 amperes, melting both at the terminal and at the arched portion. The great difference in the heating of the two terminals of the 370,000 cm. was due to the imperfect soldering of one of them. 150 37 ft ft n M R IN T) .-I 5. 130 IfiflfiJVM ? 25 00< Q IL M B M T) A 120 Xi IB MTS A JL.AREA^^SG • 1 X. s > 1000 Aft IP J N ~< .,- < 00 | ^o M H m J a ca ± fe £ ^ on 5 < • A -- H - f- fc 90 : A ~> Q 7 / / 5 : 3 S- z t / / ; / / — C* = -< O / / a 70 / J / / 1 1 / / / I / / / t 1 / 1 } 7 t / t / J V 1 / / // / / > / r ^ X / • 3 L 10 <: -— t ^ f *» 9 34 38 42 46 60 54 56 62 TEMP$— °C. Fig. 124. 774 ELECTRIC RAILWAYS. Riveted Bonds. — These are formed of a length of wire or cable having a copper terminal pressed or welded to its ends. Solid wire bonds of this type break easily from track vibration if short, and are used most largely for connecting around special work. This type of bond is sub- divided into several styles, according to the way the shank of the terminal is fastened into the hole in the rail. 1. Bolt Expanded Terminal. — In this one the shank of the terminal is made with a hole through its center. Through this hole is passed a steel bolt which is threaded on one end and has a beveled shoulder Fig. 125. on the other. After the shank is fitted into the hole, it is expanded by pull- ing the bolt through the terminal by means of a nut, the tapered shoulder expanding the shank into the hole. This is shown in Fig. 125. £. Pin Expanded Terminal. — In this type the terminal is made with a hole through the center of the shank which is fitted into the hole it is to occupy and a beveled steel pin is driven through its center, expanding the shank to a tight fit. This is shown in Fig. 126. Fig. 126. Fig. 127. These two types are used principally for bonding the channel rails of the conduit system of electric railways. In both types the shank of the terminal should practically fit the hole before the pin or bolt is driven in. 3. IKachine Hiveted Terminals. — In this type the shank of the terminal is made solid and is compressed into the hole by means of mechanical or hydraulic pressure (Fig. 127). Terminals of bonds should never be riveted by hammer as the shank is Fig. 128. Poorly Riveted Terminal. Fig. 129. Well Riveted Terminal. not properly expanded into the hole (Fig. 128). An imperfect contact increases the resistance besides making the bond liable to further deteriora- tion by reason of the accumulation of moisture between the shank and the hole. By means of the compressor the back of the terminal is first held securely against the face of the rail, then the shank of the terminal is ex- panded, forcing the soft metal back toward the base, making a uniform contact throughout the thickness of the rail, filling the hole so completely as TYPE OP BONDS. 775 to fill even the tool marks of the drill, and moreover, greatly increases the area of contact between the bond and the rail on account of the button head caused by the compressor (Fig. 129). This contact surface is an essential feature, and the efficiency of the bond depends upon this connection being made in the best possible manner. Tests show that it takes twice the power to turn the compressed terminal in its hole that it does to turn the pin-driven terminal. As the only resist- ance against turning is the friction between the copper in the terminal and the sides of the hole, the compressed terminal must have much the superior contact. Fig. 130. Fig. 131. Fig. 132. Figures 130 and 132 show respectively the double-screw and hydraulic compressors which have been successfully used on bonds in the web of the rail, and Fig. 131 shows a hydraulic compressor used successfully for putting bonds in the base of the rail. The requirements for a good bond are: 1. Terminal should be made as an integral part of the stranded or body portion, in such a manner as to form practically a molecular union and thereby introduce a minimum resistance between the two. 2. Its terminal should be so proportioned as to have contact surface with the rail sufficient to carry the same amount of current as the body portion of the bond. 3. Its body portion should be so constructed as to possess sufficient flex- ibility to withstand all vibrations to which it may be subjected, such as Fig. 133. Fig. 134. hammer blows, of passing car wheels on the track, and expansion and con- traction of the rails due to temperature variations. 4. A method cf applying the bond which will insure the permanency of the contact with the steel and reduce depreciation to a minimum. In all cases it is desirable to have the bonds as little exposed as possible both for appearances, and to prevent their being stolen. This is particu- larly true of those in the return circuit. Bonds should also be made as short as possible to make their cost a minimum. For these reasons it is highly 776 ELECTRIC RAILWAYS. desirable that the bonds be placed under the splice plates whenever possible. In new installations standard splice plates are now procurable which have ample space between their inner surfaces and the rail to allow for the bonds, and in changing over old installations the saving in the initial t cost of the bonds and the saving from loss by theft will go far towards paying for new splice plates. With the idea of placing the bonds under the splice plates, manufacturers have designed them in suitable shapes, either by flattening the strands, Fig. 135. or the use of flat wires in the strands. Figures 133 and 134 show girder rails with bonds under the splice plates, and Fig. 135 shows a standard "T" rail similarly bonded. Resistance of Bonds. — The total resistance of a bond is com- posed of three factors, the resistance of the copper in the bond, the resistance between the body of the bond and the terminal, and the contact resistance between the terminal and the rail. The following table gives the resistance of some of the more common sizes of bonds used: Size of Bond. Length of Bond. 5" 6" 7" 8" 9" 10" 00 000 oooo .000047 .000039 .000033 .000028 .000056 .000046 .000038 .000032 .000064 .000052 .000043 .000036 .000072 .000059 .000048 .000040 .000081 .000053 .000053 .000044 .000089 .000072 .000059 .000048 For any given size of bond the only variable factor in its resistance with the length is the resistance of the copper in the bond, the other two factors remaining constant. Hence the resistance of different sizes can be plotted as is done in Fig. 136, using resistance in ohms and length in inches as ordinates. At least i inch extra length of short bonds should be allowed for extreme contraction of rails due to changes in temperature, and bonds shorter than 9 inches are liable to excessive breakage due to vibration. The most common practice has been to have the bond holes drilled at the rolling mills. Hence, when it is desired to do the bonding, the holes are rusty and will need to be reamed out until clear and bright. The cost of having the holes drilled at the mill at the current price ($1.00 per ton of rail) usually amounts to about 20 cents per hole, and the reaming to about 5 cents per hole — a total of 25 cents per hole, while if the holes are drilled just as the bonding is done, they will cost about 1\ cents each, including tools and supervision. Punched holes cost about 4 cents each. These costs will vary with conditions and rates of wages, the above being based on $2.00 for a day of eight hours. There is no material disadvantage in drilling the holes with oil. RESISTANCE OF BONDS. 777 RESISTANCE IN OHMS b 2 c> * & b 1 § 8 S 8 ►=* M t-* to to g S § g § w \\ l\ \ N 1 \ \ 1 \ \\ \ \ \ lh\ \ \ \\ \ o v A \ \ \ \\ \ \ \ r \ m \\ \ \ \ \\ \ \ V \ V \ \ • \ RESISTANCE OF COPPER BONDS AT 75°F \ \ \ \ \ \ \ \ \ \ \\ n \ ) v£ c W Y? V W< \ \to \s % Y- -1 \' ^ U V if Wto. 3s \« \> i\§is_ I© YsL ^ \ Li In r t\ 1 \a£,\§' } t \, \ gw? \ \ \ i \ \ \ \ ^ ~ j. \ \ \ 1 \ r l| \ \ \ \ \ la \ \ \ \ _p_L \ \ \ \ \ \ r \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 1 m Fig. 136. 778 ELECTRIC RAILWAYS. Care should be taken to see that the holes are free from all moisture, as its presence greatly reduces the efficiency of the bond, hence bonding should never be done during damp or wet weather. After the holes have been properly prepared the surface of the metal directly around the hole should be reamed so as to provide a bright, clean surface for the base of the terminal on the one side, and the button head, when riveted, on the other. If the shank of the terminal becomes oxidized or dirty it should be cleaned before being put into the rail. Third Hail .Bonding*. — The practice in third rail bonding has been to bond the rail slightly in excess of its conductivity, in order to make the rail nearly a uniform conductor. In order to accomplish this it has been necessary to bond the base of the rail as well as the web. Special malleable iron splice jplates are used which allow sufficient space for the bonds. Fig. 137 shows the bonding of the third rail of the Interborough Rapid Transit Company (New York Subway). VFelded Joints. — On many systems where the rails are imbedded they are made practically continuous by the use of welded joints, and but 3-^2 ,6% L-l)*/ Drilled Hole Punched Hole Fig. 137. little trouble is experienced by broken joints or bent rails. These are not practicable on third rails or track rails that are not embedded and thus exposed to all temperature changes. In the electrically welded system an iron plate is welded across the joint on each side of the rail web by means of heavy current of electricity applied by special low voltage machinery. The cast weld joint is simply a large lump of steel cast about the joint in a mould after the rail ends have been cleaned. Voynow Joint,— (Street Railway Journal.) The Voynow joint con- sists of what may be called two special channel bars which are riveted to the ends of the rail. These plates are not made to fit the fishing sec- tion of the rail; on the contrary, spaces are left under the head, tram and around the foot of the rail. The flat surfaces of both sides of the rails and of the joint bars having been previously cleaned by sand-blast, these spaces are filled with molten zinc, which enters into and fills out all the irregularities of the rolled surfaces, thus giving a continuous bearing through- out the whole length and width of the flanges of the plates. The adhesion of the molten zinc to the rails and plates, together with the body-bound rivets, holds the joint permanently tight, and at the same time prevents expansion, thus making rails continuous. As the rail ends and inside of the plates are cleaned to the metal by sand-blast, the joint is also of the best, electrically considered. Thermit Rail- Welding*. — The thermit process is a purely chemi- cal operation, based upon the fact that metallic aluminum, under proper conditions, will reduce many of the other metals from their compounds to their simple form; as, for instance, if aluminum is mixed with oxide of iron and the mixture is ignited, the aluminum will unite with the oxygen of the oxide, forming aluminum oxide (which is commercial corundum), leaving the iron free. As the process of reduction liberates a great amount of heat, the temperature of the mixture during the reaction rises rapidly (to about 5000° F.), changing the iron to a molten low-carbon steel. Expressed in RESISTANCE OF TRACK RAILS. 779 chemical terms, the equation, according to which the reaction takes place, would toe Fe 2 3 + 2A1 = Al 2 3 + 2 Fe. This is the process utilized in welding rails. The oxide of iron is mixed with powdered aluminum in the right proportion, and introduced into a crucible lined with magnesia, or with material obtained from a previous fusion. In order to set off the con- tents of the crucible, a small quantity of ignition powder (barium peroxide and pulverized aluminum) is put in a small heap on top of the mixture, and is ignited by means of a match or red-hot iron rod. The reaction propagates itself quickly through the whole mixture, with the result that in a few seconds the whole charge is a mass of white-hot fluid material. The contents of the crucible have separated into two layers, the molten metal reduced by the aluminum being at the bottom and the molten aluminum oxide above it. In the application to rail-welding, a cone-shaped crucible, with magnesite lining, is mounted on a tripod over the joint to be welded, a properly pre- pared iron sand clay mould having been previously clamped around the joint. The conical crucible has a hole in the bottom, and before the operation a small iron rod or pin is placed in this hole with its end projecting several inches below the crucible. Above the head of the pin in the bottom of the crucible is first carefully fitted an asbestos washer, and on top of this is placed a solid circular metal washer to hold it in place. About 15 pounds or 20 pounds of powdered aluminum and oxide iron are then poured into the crucible. This mixture is known as "Thermit," and is furnished properly mixed and ready for use in small bags by the manufactures. On top of the mixture is placed a quantity of ignition powder, about enough to cover a 50-cent piece, When all is ready, a match is applied to the powder and a conical cover with a central opening is hastily placed on the crucible. In a few seconds the reaction commences, and within thirty seconds the contents of the crucible become a seething, boiling mass of molten metal. As soon as the reaction has reached its height, a man strikes the pin projecting from the bottom of the crucible with a rod or small shovel, driving the pin upward, thus freeing the hole and allowing the molten metal to flow down into the mould around the joint, depositing a mass of metal around the joint and welding the ends of the rails into one piece. Resistance of Track Rails. The resistance of the commercial steel track rails is about thirteen times that of copper. On this basis the following table of resistances of rails is computed. Weight Sectional Equivalent Resistance of Area Cir. Mils per Mile Rail. Sq. Inch. of Copper. Ohms. 45 4.4095 431,883 . 13074 50 4.8994 479,884 .11766 55 5.4874 536,034 . 10502 60 5.8794 575,505 .09806 65 6.3693 623,887 .09051 70 6.8592 671,825 .08404 75 7.3491 719,380 .07844 80 7.8392 767,763 .07354 85 8.3291 814,873 .06922 90 8.8190 863,766 .06537 95 9.3089 911,767 .06193 100 9.7988 1,072,068 .05883 Area in cir. mils Equivalent cir. mils of copper 1,000,000 X weight per yard 10.2052 X Area in cir. mils t 13 7854 780 ELECTRIC RAILWAYS. EXPEHI^IEXTS IFOR DETERMIXATIOX OJF THE RELATIVE VAMJJE OJF RAJTJLS l\I) BONDED JOIVI*. (W. H. Cole.) Fifteen rails were used, giving three joints for each of the five different classes, and in making the tests and observations an average of the results for the three rails of its class was given. Micrometer calipers were used in measuring the wear of the rails each month, three different measurements were made at each place, and an average was calculated from these three measurements, viz.: A. At a point at or near the gage line. B. At a point in the center of the tread. C. At a point near the outside of the rail. The joints that were bonded were fished with standard fish plates, bolted with eight 1-inch bolts, screwed up tight; the rail ends butting each other were laid, fished and bonded in the maximum heat of the day, and immedi- ately covered and paved around them. No. 1. Three joints fished as above and bonded around the fish plates with standard Chicago bonds No. 00 B. & S. gage, two bonds to each joint. No. 2. Bonded with "Crown" concealed bonds, with two bonds of a section equal to two No. 00 copper B. & S. gage, and the fish plates bolted over them. No. 3. No. 2 plastic bonds, made by Harold P. Brown, and carefully installed according to instructions, by a man formerly experienced in this work. No. 4. Three joints welded by the Falk process. No. 5. Three joints welded by the Goldschmidt thermit process. The rails were laid continuously so the same cars passed over the same section containing the different types of joints. The subjoined tables give the results, from which the writer has arrived at the following con- clusions: That for electric street railways under average traffic conditions, rails should give a life of about forty years if the joints are made continuous, and are composed of Carbon 55 to .58 Silicon 10 or under Phosphorus 08 or under Sulphur 06 or under Manganese 83 or under Ingredients of Rails Under Test. Carbon. Carbon Silicon . . Phosphorus Sulphur . Manganese , Iron . . , Soft. .284 .061 .105 .085 .784 1.299 98.701 Medium. .572 .235 .052 .078 .981 1.918 98.082 100.000 100.000 100.000 Hard. .591 .057 .098 .060 .830 1.636 98.364 Note. — Metalloids ignored. BOARD OF TRADE REGULATIONS. 781 The following would be the electrical efficiency and loss at the beginning and end of the first year: Class of Joint. Chicago bonds Crown bonds Plastic bonds Falk cast weld . . . . Goldschmidt thermit weld Electrical Per Cent Efficiency at Begin- ning of Year. 89.51 86.71 89.72 101.16 101.14 Electrical Efficiency at End of Year. 74.43 73.72 77.84 86.53 100.39 Per cent below Equal Sec- tion of Rail. 29.57 26.28 22.16 10.44 100.39 + BOARD ©J? TRADE HEGlLATIO.\§. for Oreat Britain. Regulations prescribed by the Board of Trade under the provisions of Section of the Tramways Act, 189 — , for regulating the employ- ment of insulated returns, or of uninsulated metallic returns of low resist- ance; for preventing fusion or injurious electrolytic action of or on gas or water pipes, or other metallic pipes, structures, or substances; and for min- imizing, as far as is reasonably practicable, injurious interference with the electric wires, lines, and apparatus of parties other than the company and the currents therein, whether such lines do or do not use the earth as a return. Definitions. In the following regulations : The expression " energy" means electrical energy. The expression "generator" means the dynamo or dynamos or other electrical apparatus used for the generation of energy. The expression " motor " means any electric motor carried on a car and used for the conversion of energy. The expression " pipe " means any gas or water pipe, or other metallic pipe, structure, or substance. The expression " wire " means any wire apparatus used for telegraphic, telephonic, electrical signaling, or other similar purposes. The expression "current" means an electric current exceeding one- thousandth part of one ampere. The expression " the company " has the same meaning or meanings as in the Tramways Act, 189—. Regulations. 1. A.ny dynamo used as a generator shall be of such pattern and con- struction as to be capable of producing a continuous current without appre- ciable pulsation. 2. One of the two conductors used for transmitting energy from the gen- erator to the motors shall be in every case insulated from earth, and is hereinafter referred to as the M line"; the other may be insulated through- out, or may be insulated in such parts and to such extent as is provided in the following regulations, and is hereinafter referred to as the " return." 3. Where any rails on which cars run, or any conductors laid between or within three feet of such rails, form any part of a return, such part may be uninsulated. All other returns or parts of a return shall be insulated, unless of such sectional area as will reduce the difference of potential be- tween the ends of the uninsulated portion of the return below the limit laid down in Regulation 7. 4. When any uninsulated conductor laid between or within three feet of the rails forms any part of a return, it shall be electrically connected to the rails at distances apart not exceeding 100 feet, by means of copper 782 ELECTRIC RAILWAYS. strips having a sectional area of at least one-sixteenth of a square inch, or by other means of equal conductivity. 5. When any part of a return is uninsulated it shall be connected with the negative terminal of the generator, and in such case the negative termi- nal of the generator shall also be directly connected, through the current- indicator hereinafter mentioned, to two separate earth connections, which shall be placed not less than twenty yards apart. Provided that in place of such two earth connections the company may make one connection to a main for water supply of not less than three inches internal diameter, with the consent of the owner thereof, and of the person supplying the water ; and provided that where, from the nature of the soil or for other reasons, the company can show to the satisfaction of an inspecting officer of the Board of Trade that the earth connections herein specified cannot be constructed and maintained without undue expense, the provisions of this regulation shall not apply. The earth connections referred to in this regulation shall be constructed, laid, and maintained so as to secure electrical contact with the general mass of earth, and so that an electromotive force not exceeding four volts shall suffice to produce a current of at least two amperes from one earth connection to the other through the earth, and a test shall be made at least once in every month to ascertain whether this requirement is complied with. No portion of either earth connection shall be placed within six feet of any pipe, except a main for water supply of not less than three inches in- ternal diameter, which is metallically connected to the earth connections with the consents hereinbefore specified. 6. When the return is partly or entirely uninsulated, the company shall, in the construction and maintenance of the tramway (a), so separate the uninsulated return from the general mass of earth, and from any pipe in the vicinity ; (b) so connect together the several lengths of the rails ; (c) adopt such means for reducing the difference produced by the current be- tween the potential of the uninsulated return at any one point and the po- tential of the uninsulated return at any other point ; and (d) so maintain the efficiency of the earth connections specified in the preceding regulations as to fulfill the following conditions, viz.: (1.) That the current passing from the earth connections through the in- dicator to the generator shall not at anytime exceed either two amperes •per mile of single tramway line, or 5 per cent of the total current output of the station. (2) That if at any time and at any place a test be made by connecting a galvanometer or other current indicator to the uninsulated return, and to any pipe in the vicinity, it shall always be possible to reverse the direction of any current indicated by interposing a battery of three Leclanche cells connected in series, if the direction of the current is from the return to the pipe, or by interposing one Leclanche cell, if the direction of the current is from the pipe to the return. In order to provide a continuous indication that the condition (1) is com- plied with, the company shall place in a conspicuous position a suitable, properly connected, and correctly marked current indicator, and shall keep it connected during the whole time that the line is charged. The owner of any such pipe may require the company to permit him at reasonable times and intervals to ascertain by test that the conditions specified in (2) are complied with as regards his pipe. 7. When the return is partly or entirely uninsulated, a continuous record shall be kept by the company of the difference of potential during the work- ing of the tramway between the points of the uninsulated return furthest from and nearest to the generating station. If at any time such difference of potential exceeds the limit of seven volts, the company shall take imme- diate steps to reduce it below that limit. 8. Every electrical connection with any pipe shall be so arranged as to admit of easy examination, and shall be tested by the company at least once in every three months. 9. Every line and every insulated return or part of a return, except any feeder, shall be constructed in sections not exceeding one half of a mile in length, and means shall be provided for insulating each such section for purposes of testing. BOARD OF TRADE REGULATIONS. 783 10. The insulation of the line and of the return when insulated, and of all feeders and other conductors, shall be so maintained that the leakage cur- rent shall not exceed one-hundredth of an ampere per mile of tramway. The leakage current shall be ascertained daily, before or after the hours of running, when the line is fully charged. If at any time it should be found that the leakage current exceeds one-half of an ampere per mile of tram- way, the leak shall be localized and removed as soon as practicable, and the running of the cars shall be stopped unless the leak is localized and removed within twenty-four hours. Provided, that where both line and return are placed within a conduit this regulation shall not apply. 11. The insulation resistance of all continuously insulated cables used for lines, for insulated returns, for feeders, or for other purposes, and laid be- low the surface of the ground, shall not be permitted to fall below the equivalent of 10 megohms for a length of one mile. A test of the insulation resistance of all such cables shall be made at least once in each month. 12. Where in any case in any part of the tramway the line is erected over- head and the return is laid on or under the ground, and where any wires have been erected or laid before the construction of the tramway, in the same or nearly the same direction as such part of the tramway, the com- pany shall, if required to do so by the owners of such wires or any of them, permit such owners to insert and maintain in the company's line one or more induction coils, or other apparatus approved by the company for the purpose of preventing disturbance by electric induction. In any case in which the company withhold their approval of any such apparatus, the owners may appeal to the Board of Trade, who may, if they think tit, dis- dispense with such approval. 13. Any insulated return shall be placed parallel to, and at a distance not exceeding three feet from, the line, when the line and return are both erected overhead, or 18 inches when they are both laid underground. 14. In the disposition, connections, and working of feeders, the company shall take all reasonable precautions to avoid injurious interference with any existing wires. 15. The company shall so construct and maintain their systems as to secure good contact between the motors, and the line and return respec- tively. 16. The company shall adopt the best means available to prevent the oc- currence of undue sparking at the rubbing or rolling contacts in any place, and in the construction and use of their generator and motors. 17. In working the cars the current shall be varied as required by means of a rheostat containing at least twenty sections, or by some other equally efficient method of gradually varying resistance. 18. Where the line or return or both are laid in a conduit, the following conditions shall be complied with in the construction and maintenance of such conduit : (a) The conduit shall be so constructed as to admit of easy examination of, and access to, the conductors contained therein, and their insulators and supports. (b) It shall be so constructed as to be readily cleared of accumulation of dust or other debris, and no such accumulation shall be permitted to remain. (c) It shall be laid to such falls, and so connected to sumps or other means of drainage as to automatically clear itself of water without danger of the water reaching the level of the conductors. (d) If the conduit is formed of metal, all separate lengths shall be so jointed as to secure efficient metallic continuity for the passage of electric currents. Where the rails are used to form any part of the return, they shall be electrically connected to the conduit by means of cop- per strips having a sectional area of at least one-sixteenth of a square inch, or other means of equal conductivity, at distances apart not ex- ceeding 100 feet. Where the return is wholly insulated and contained within the conduit, the latter shall be connected to earth at the gen- erating station through a high resistance galvanometer, suitable for the indication of any or partial contact of either the line or the return with the conduit. 784 ELECTRIC RAILWAYS. (e) If the conduit is formed of any non-metallic material not being of high insulating quality and impervious to moisture throughout, and is placed within six feet of any pipe, a non-conducting screen shall be interposed between the conduit and the pipe, of such material and dimensions as shall provide that no current can pass between them without traversing at least six feet of earth; or the conduit itself shall in such case be lined with bitumen or other non-conducting damp- resisting material in all cases where it is placed within six feet of any pipe. (/) The leakage current shall be ascertained daily before or after the hours of running, when the line is fully charged, and if at any time it shall be found to exceed half an ampere per mile of tramway, the leak shall be localized and removed as soon as practicable, and the running of the cars shall be stopped unless the leak is localized and removed within 24 hours. 19. The company shall, so far as may be applicable to their system of working, keep records as specified below.' These records shall, if and when required, be forwarded for the information of the Board of Trade. Daily Records. Number of cars running. Maximum working current. Maximum working pressure. Maximum current from earth connections (vide Regulation 6 (1) ). Leakage current (vide Regulation 10 and 18/.). Fall of potential in return (vide Regulation 7). Itloiitnly Records. Condition of earth connections (vide Regulation 5). Insulation resistance of insulated cables (vide Regulation 11). Quarterly Records. Conductance of joints to pipes (vide Regulation 8). Occasional Records. Any tests made under provisions of Regulation 6 (2). Localization and removal of leakage, stating time occupied. Particulars of any abnormal occurrence affecting the electric working of the tramway. Signed by order of the Board of Trade this day of 189 Assistant Secretary, Board of Trade. • OVERHEAD CONDUCTING SYSTEM. 785 CAlCrLATIl¥G THE OVERHEAD COIDUCTOG llSTtn Ol EJLFCTJR JLC RAILWAYS. Dr. Louis Bell gives the following steps as the best to be followed in entering upon the calculation of the conducting system of a trolley road: Extent of lines. Average load on each line. Center of distribution. Maximum loads. Trolley wire and track return. General feeding system. Reinforcement at special points. It must be said at once that experience T skill, and good judgment are far better than any amount of theory in laying out the conducting system of any road. Much depends upon the character of the load factor, i.e., the ratio of average to maximum out-put ; and this, varying from . 3 to . 6, can only be judged from a study of the particular locality, the nature of its industries and working people, the shape of the territory, and the nature of the sur- rounding country. Map out the track to scale, noting all distances carefully, and dot in any contemplated extensions, so that adequate provision may be made in the conducting system for them. Note all grades, giving their length, gra- dient, and direction. Divide the road into sections such as may best sug- gest themselves by reason of the local requirements, but such as will make the service under ordinary conditions fairly constant. The average load on each section will depend, of course, upon the number of cars, and the number of cars upon the traffic. This can only be arrived at by a comparison with similar localities already equipped with street railway, and even then considerable experience and keen judgment of the general nature of the towns are necessary in arriving at anything like a correct result. If the road has been correctly laid out as to sections, the load on each will be uniform and may be considered as concentrated at a point midway in each section. Now, if a street railway were to be laid down on a per- fectly level plain where the cost of real estate was the same at all points, and wires could be run directly to the points best suited; then it would only be necessary to locate the center of gravity of the entire system, and build the power station at that point, sending out feeders to the center of each section. Unfortunately for theory, such is never the case; and cost of real estate, availability of the same, convenience of fuel, water, and supplies will govern very largely the selection of a location for the power-house. Even when all the above points necessitate the placing of the power-house far from the center of gravity of a system, it may be possible to use such center as the distributing point for feeder systems, and even where this is not possible, it is well to keep in mind the center, and arrange the dis- tributing system as nearly as possible to fit it. All this relates, however, to preliminary determinations for the system as determined at the time, and in large systems will invariably be supple- mented by feeders, run to such points as the nature of the traffic demands. A basebail field newly located at some point on the line not known to the engineer previous to* the installation, will require reinforcement of that particular section; and often after a road has been running for some time, the entire location of traffic changes, due to change in facilities, and feeder systems then have to be changed to meet the new conditions, so that after all, location of the center of distribution depends largely on judgment. The maximum current will rise to four or five times the average where but one or two cars are in use; will easily be three times the average on roads of medium size, while on very large systems it may not be more than double the average. If speeds are maintained on heavy grades the maxi- mum is still further liable to increase. Another point to be considered in connection with maximum load is the location, not only of heavy grades, but of parks, ball-grounds, athletic fields, cemeteries, and other such places for large gatherings of people that are liable to call for heavy massing of cars, many of which must be started 786 ELECTRIC RAILWAYS. practically at the same time, and for which extra feeder, and in some cases extra trolley capacity, must be provided. Having determined the average current per section of track, the maximum for the same, and the extraordinary maximum for ends, park locations, etc., as well as the distances, all data are obtained necessary for the determina- tion of sizes of feeders. The selection of the proper size of trolley wire is somewhat empirical, but the size may be governed by the amount of current that is to be carried. It is obvious that with given conditions the larger the trolley wire the fewer feeders will be necessary, and yet with few feeders the voltage is liable to vary considerably. In ordinary practice of to-day No. OB. &.S. and No. 00 B. &. S. gauge, hard-drawn copper are the sizes mostly in use, the latter on those roads having heavier traffic or liable to massing of cars at certain localities. On suburban roads using two trolley wires in place of feeders, 0000 B. & S. gauge will probably be best. Track return circuit has been treated fully in a previous chapter (see page 771); and all that is needed to say here is, that some skill in judgment is necessary in settling on the value of the particular track return that may be under consideration, in order to determine the value of the constant to be used in the formula for computing the size of wire or overhead circuit. In ordinary good practice this value may be taken as 13, 14, or 15, according as the bonding and rail dimensions are of good type and large. It is quite obvious that the current-carrying capacity of the feeder must be taken into consideration, in spite of any determination of drop; and this can be found in the chapter on Conductors. Sizes of conductors are also governed to some extent by convenience in handling, and it is found that 2,000,000 cm. is about the largest that can be safely handled for under- ground work, while anything larger than 500,000 cm. for overhead circuits is found to be difficult to handle. CONTINUOUS CURRENT fKEUERK LOAD I)KTE«. MIIATIOA T . The first step towards determining the load is to draw a train diagram from the proposed time-table or schedule of trains. Such a diagram, having as abscissae the length of the line and as ordinates the hour of the day, shows in a graphic form the course of every train and the number of trains on the line at any time. The stops may be omitted if they are very short compared to the runs, but in any case it is usual to show the course of each train by a straight line over each run, variations of speed being ignored unless of considerable duration and magnitude. An example of such a train diagram is given in Fig. 138, in which each train is indicated by a special kind of line in order to illustrate how it travels to and fro. The load at any time is estimated by counting how many train curves cut the line representing that particular time. Knowing the average amperes per train the total amperes are easily estimated for any time of day and may be plotted in the form of a load diagram. The average value of amperes for this purpose is obtained by plotting the curves of current for each run and adding the ampere hours of all these runs. The total ampere hours divided by the total number of hours occupied by the runs, is the average current taken by a train. The method of plotting the current curves is described on page 667. Economical Hesig*n of feeders. — The investment in a system of feeders may be expressed as an initial cost, or as an annual interest or percentage thereof. The value of the kilowatt-hours lost in the feeders is most conveniently expressed as an annual expense. The sum of these two annual items is the total annual expense of the feeders. If the cost of feeders be proportional to the amount of copper and if the energy loss be computed for exactly the same part of the system as the first cost expense, the total cost will be a minimum when the interest and energy items are equal. This is known as Kelvin's Law. Unfortunately the conditions which are necessary for the correct application of this rule are not usually met with in practice. The cost of conductors is seldom proportional to the amount of copper owing to the existence of such items as cost of manu- facture, installation and insulation. When, however, it is desired to find the most economical size of feeder to connect to a trolley wire or contact CONTINUOUS CURRENT FEEDERS. 787 1 1 ' 1 1 | I 1 Hi l \[ I 1 1 1 1 1 f • I j j | 1 1 1 1 p 1 1 1 H 1— 1 1 1 j l 1 1 1 1 1 1 1 1 1 1 1 1 ii ■ ii ! 11 1 1 1 1 1 1 — 1 — 1 1 1 I l i 1 1 1 ! 1 -w— ! 1 1 1 rt t 1 1 i _j , — 1 I l 1 1 i j » 1 1 1 1 1 1 y 1 1 II 1 i li . i 1 | 1 1 1 1 1 Ii i 1 1 1 H 1 1 1 1 1 jl U -H 1 !! I| ii i l| 1 1 i i ! i ■ ii 4 5 6 7 8 j i 11 A.M. Train Diagram Fig. 138. 788 ELECTRIC RAILWAYS. rail, the total energy loss in the combined system is more important than the loss in the feeder wires alone, so that in this case it is advisable to make a minimum the sum of the energy loss in the whole system and the interest and depreciation on part of the system, and the most economical case must be worked out by trial. A table showing how to do this is given herewith and should be used in connection with that on "Distribution of Copper," which is given below. In the former table the system of most economical distribution (Case 3) of the latter table, is assumed to be used, but this is not necessary, and is not even applicable if there is no drain of current from the conductors. Volts drop to end of line V. Kw, hrs.lost per annum with R.M.S. current 5.2 aV Annual cost of energy at n cents per kw.-hrs. $.052 naV. Total C.M.- feet 4 oL 2 9 k * V Exist- ing con- ductors CM.- Ft. Extra C.M.- Ft. reg.'d Feet of C.M. cable req.'d Total cost of new cables Interest main- tenance and depreci- ation on cable at • - •% Total annual ex- pense, sum of third and last items. a = square root of the mean of the currents squared. Limiting- Potential Brop.- The total drop in the positive and negative feeders is regulated by several conditions some of which, unfor- tunately, may be contradictory. The line voltage must always be high enough to supply current for starting a car on an up-grade, and to keep the lights bright. For a multiple-unit system, the line voltage must be sufficient to operate the contactors and air compressors with certainty The General Electric Company's type M system of control should have at least 300 volts. The permissible drop is also influenced by considera- tions of economy, and in grounded feeders is often required not to exceed a certain limit fixed by law, this limit varying according to the locality. In England the maximum drop allowed in the grounded conductors is seven volts, whereas in most American cities no limit at all exists, it being only necessary for the railway company to take whatever precaution may be requisite to prevent electrolytic trouble. Two Classes of feeders. — Any direct current feeder system consists of two parts, the conductors which carry the current to the line and the line conductors (trolley wire) which serve as contact media to con- vey current to the cars. One set of conductors may be so designed as to fulfil these two functions, or the lines from the power station may be quite distinct from the contact rail or wire. In this latter case, the conductors from the power station carry the same current along their entire length, so that problems relating to drop, etc., may be treated by Ohm's Law. The contact conductors in either the first or second case mentioned above require somewhat different treatment owing to the fact that the current depends on the distribution of cars on the line. CONTINUOUS CURRENT FEEDERS. 789 Varioua arrangements of feeder and contact conductors are shown in Figs. 139, 140, 141, 142, and 143. Fig. 139 shows the simple ladder system in which the feeders and trolley wire are joined at intervals so as to form vir- TROLLEY WIRE TRACK RETURN CIRCUIT Fig. 139. tually a single conductor. In its best form the cross section of the feeder is tapered according to the rules given below. Fig. 140 shows a modification of the last scheme. In this case the trolley wire is cut into sections, so that while losing the extra conductivity of the continuous trolley, each section TROLLEY IN SECTIONS I) L" TRACK RET.UR8 CIRCUIT Fig. 140. may be cut out in case of trouble without depriving the remainder of the system of current. Each section may be protected by a fuse and switch or a circuit breaker, but it is a disadvantage to have such apparatus scat- tered along the line. Fig. 141 shows a system where the current leaves the Fig. 141. TRACK RETC0B8 CIRCUJT Fig. 142. 790 ELECTRIC RAILWAYS. station by several lines, thereby enabling a number of small circuit breakers to be used instead of the large one required by the other systems. It, how- ever, has the disadvantage of being uneconomical in copper, as the long lines carry very little of the load near the generators. The system shown in Fig. 142, is in many respects ideal from an operating standpoint, but it is very uneconomical in copper and energy. Each section of the trolley wire . Station Bus .Kail Fig. 143. or third rail may be controlled by a circuit breaker in the power station thus giving the operators complete control in case of overload, short-circuit, or accident of any kind. It is also quite advantageous to replace a large circuit breaker by a number of small ones where thousands of amperes have to SUB-FEEDER HEAVY GRADES 40 TRACKRETURN CIRCUIT Fig. 144. be transmitted. A combination of the last two systems is where the sections are connected by switches which can be opened in case of accident, but are normally kept closed. Fig. 143 shows a system that is useful for nega- tive return conductors in cases where it is important to keep down the drop .1 BALL PARKAT END OF LINE TRACK RETURN CIRCUIT Fig. 145. in the grounded rails. The numerous taps drain off the current in their neighborhood and so prevent the current in the rails being great at any point. The drop of potential in these insulated feeders will be considerable, but in the grounded ones it will be very little. This is in some cases more economical and certainly more simple than a "negative booster." CONTINUOUS CURRENT FEEDERS. 791 caicciatiox ojf nini:\^iov^ ojf conductors. The problem of determination of the proper size of conductors to be used in distributing the current for an electric railway is somewhat com- plicated by the fact that the load is moving or changing its location all the time, and more so by the always changing condition of the resistance of the ground return, due to load, to track bending, condition of the earth return, and nearness of water and other underground pipes. Owing to this changing condition of the ground return part of the circuit it is necessary to assume some arbitrary value for it, in comparison with that of the over- head or insulated portion. The resistance of the ground return is seldom as high as that of the overhead part, nor is it often as good as .25 of that value ; these values change with the load and track conditions, and it is now most universal to use the factor 14 as a number which represents the value of both overhead and return conductor, in place of 10.8, the resistance per mil-foot of copper, and that value is therefore used in the formulae for calculating the sizes of overhead conductors, and has been found to produce good results in practice. Let d == distance from switchboard to end of conductor. CM = cir. mils area of the conductor. V = drop in volts at far end of line. / = current. W = watts. E = volts at switchboard. 10.8 = resistance of arc mil-foot of commercial hand drawn copper wire at 20° C or 68° F. 14 = resistance factor, including track return. % = per cent expressed as a whole number, as 10 or 20. Then for plain feeders between switchboard or other source of supply and the attaching point to the system, CM — V 1400 Xd XI CM = r = v = % XE 1400 X^X watts % XE* 14 X d x I CM % XE 100 The above formulae can be used for nearly all practical determinations of feeder and other conductor sizes, but must always assume the load to be concentrated at one point or center. For other formulae for calculation of the size of conductors see chapter on conductors. Distribution of Current. — It is usual to assume the drain of current from the contact conductor to be uniform, so that the current at any section is given by the ordinates of a straight line sloping down from the power station. The error in this assumption is decreased on account of the motion of the cars as this causes the load to act as if more distributed. Distribution of Copper — As the feeders carrying the same current along their entire length can be treated by the simple formulae shown above, it is only necessary to consider those along which there is a uniform drain of current. Four typical cases are shown in the table with their respec- tive formulae for circular mils, CM. ft., watts lost, and potential drop. The following abbreviations are used. Where conductors of iron or aluminum are used it is best to reduce them to equivalent sections of copper. The volts drop given by the formulae are from the far end of the line ; in order to get the drop from the power station, the values obtained by the formulae must be subtracted from V. 792 ELECTRIC RAILWAYS. Uniform Drain of Current. AMPS.jv CM, Fig. 146. Case 1. Conductor Uniform. 10.8X1X1 Watts lost = - IV. 10.8 X Txd' Fig. 147 Case 2. Conductor Uniformly Tapered. CM. = CM. ft. = 10.8 X IXd V 10.8 XlXP 2 V Watts lost — -IV. Volts drop 10.8 XlXd CM. CONTINUOUS CURRENT FEEDERS. 793 AMPS. CM, CM. — CM. ft. = Fig. 148. Case 3. Conductor Most Economically Tapered. 2xl0.8X/xVlx^ 3 V 4 X 10.8 X / X I 2 9 V Watts lost — - IV. 5 Fig. 149. Case 4. Conductor Uniform. Current X at Station and i at Distant End. CM. = CM. ft. = 10.8x(/+i)* 2 V 10.8X(/-HW 2 2 V Watts lost = Total drop, V— 10.8xlX(I 2 +H + i 2 ) CM. X 3 10.8x?X(/+i) CM. X 2 794 ELECTRIC RAILWAYS. In case 3, the formula for CM. gives the most economical distribution of copper to produce a certain drop V to the far end of the line. It is, of course, impossible to get this exact arrangement in practice as conductors of definite size must be used. The conductors are, therefore, arranged in steps of AMPS. CM. •s, ' — ~* % ^ v v^ \ Fig. 150. decreasing area as shown in Fig. 150, each of which may be treated as an example of case 4. Miscellaneous formulae. — Watts lost, assuming uniform drain of current. Watts = amperes per foot X area of "Drop" curve in volt-feet. Potential drop in uniform conductor with any distribution of current. Volts = ohms per foot X area of current curve in ampere-feet. Most economical distribution of copper with any distribution of current. Cross section of copper proportional to * current. Note. — Do not connect trolley wire to feeder too close to power line or sub- stations, as if done this will cause frequent opening of circuit breakers. Drop and .Loss, etc., in Line between Two Substations of Unequal .Potential. Assumptions. One train moving between S.S. with constant speed and constant current. / = current per train. L = distance between sub-stations. R = resistance of line per mile of track. E x = potential of S.S. No. 1. E 2 = potential of S.S. No. 2. S.S. 1. $= S.S. 2. Fig. 151. IMPEDANCE OF STEEL RAILS. 795 .Haximiim Drop at Train. Dmax j / Dmaz 2 \ IRL 4 ~ E x - 2 E 2 + ill 2 - E 2RL 61 L + E 2 " 1 -E 2 2IR Averitgre Drop at Train. Dave i 1 Dave 2 ) 6 " E x - 2 E 2 !;! / + E t 2 - — E2 RL Averagre XiOss between S.S. _ PRL (Ei ~ - E 2 ) i (E x - E 2 )* 4JRL RL IMPEDANCE OF STEEI HAII8 TO AI^TEIti^ATINO CU«IIJE]¥T. The impedance of iron or steel conductors to alternating currents is a complicated phenomenon which varies with the frequency of the current flowing with the area and the shape of the perimeter of the cross section and the permeability; and the permeability depends upon the current in the con- ductor; therefore statements of the impedance of iron or steel conductors to alternating currents convey little true meaning without a statement of all the conditions named above. Owing to the complexity of these con- ditions it is practically impossible to compute the values which must there- fore be determined by experiment. Following are tables showing the results of experiments upon steel track rails. Experimental Determination of Impedance of Steel Rails* (A. H. Armstrong, G. E. Co.) 45-pound Rail. Measured cross section — 4 . 26 square inch. Perimeter — 15 . 875 inches. Direct current resistance of 180 feet — .00371 ohm. Cycle Amps. Volts Power Factor Imped- ance Watts Efif. Res. React. 25 25 25 223.2 332 438 4.18 6.75 8.85 .834 .852 .864 .01875 .0203 .0202 776 1910 3350 .0156 .01735 .01747 .0103 .0106 .0102 40 40 40 223.2 332 438 5.37 8.8 11.47 .826 .876 .889 .0241 .0265 .0262 990 2560 4450 .0199 .0233 .0232 .0136 .0129 .0120 60 60 60 223.2 332 438 6.88 11.06 14.46 .850 .901 .877 .0308 .0334 .0330 1308 3305 5550 .0262 .0300 .0289 .0162 .0145 .0158 796 ELECTRIC RAILWAYS. OO-pound rail. Measured cross section — 6 square inches. Perimeter — 18.75 inches. Direct current resistance of 180 feet — .00185 ohm. Cycles Amps. Volts Power Factor Imped- ance Watts Eff. Res. React. 25 25 25 296 398 622 6.32 8.64 11.73 .826 .849 .861 .0213 .0217 .0189 1545 2920 6280 .01765 .01841 .01625 .0120 .01145 .00961 40 40 40 296 398 622 7.95 10.98 15.4 .896 .871 .870 .0268 .0276 .0248 2110 3800 8340 .0241 .0240 .02155 .0119 .01355 .0122 60 60 60 296 398 622 10.13 13.74 19.15 .901 .916 .869 .0343 .0345 .0308 2700 5010 10350 .0308 .0317 .0268 .0149 .0138 .01525 SO-pound rail. Measured cross section — 7.77 square inch. Perimeter — 21.5 inches. Direct current resistance of 180 feet — .002035 ohm. Cycles Amps. Volts Power Factor Imped- ance Watts Eff. Res. React. 25 25 25 392 620 820 6.1 10.01 12.83 .796 .756 .834 .01555 .0162 .01565 1905 4700 8760 .0124 .01225 .0130 .0094 .0106 .00863 40 40 40 392 620 820 7.61 12.98 17.35 .816 .837 .866 .0194 .0209 .0212 2440 6720 12300 .0159 .0175 .0183 .0112 .001145 .0106 60 60 60 392 620 820 10.15 17.03 21.65 .863 .898 .853 .0259 .0275 .0264 3430 9460 15150 .0223 .0246 .0225 .0131 .0121 .0138 Experiment on Interworks Tracks of Westingrhouse E. & M. Co. "In order to determine the drop in voltage in a circuit composed of a trolley wire and a pair of track rails and to determine also the effect of the addition of a feeder, the following tests were made on the Westinghouse Interworks Rail- way, in March, 1905. The section of the road selected was 4000 feet long and consisted of 1200 feet of double catenary construction and 2800 feet of single catenary construction. The trolley wire was No. 000 and the track rails were 70 pounds. The trolley wire was 24 feet above the track on the double catenary portion and 22 feet on the single catenary. The messenger cable consisted of yVmch stranded steel cable. A No. 0000 feeder was located approximately 3 feet above and 8 feet to the side of the trolley wire, as indicated in sketch (Fig. 152). -4000 <- D Fig. 152. EXPERIMENT ON INTERWORKS TRACKS. 797 With the end of the trolley wire grounded to the track and an alternating current of 25 cycles applied at the points B, C, the following results were obtained, with the aid of the No. 0000 feeder used as a voltmeter lead. Total Volts Volts Total Im- Power Amperes volts B - C A - B A - C pedance B - C Factor 50 23.5 15.5 8 .47 .646 100 46.2 .465 .637 150 68.5 45 22 .456 .639 200 89.6 63.2 29.5 .448 .63 300 138.4 97 44 .448 .62 Average .457 .634 On direct current the average resistance of the total circuit B-C was .248 ohm; of the portion B-D, .219 ohm; and of the portion C-D, .0266 ohm. It will be .seen from the above that the drop in voltage in this circuit, composed of trolley and track, was 45 . 7 volts per 100 amperes and that approximately two-thirds of this was due to the trolley wire and one-third due to the rails. In the second set of tests, current was supplied to the No. 0000 feeder and trolley wire in parallel and with 25 cycles alternating current, the following results were obtained. Total Amps. Amperes in trolley Amps, in feeder Voltage Imped- ance Power Factor 100 150 200 51.5 72.7 95.3 48.5 77.3 104.7 32.5 48.4 63.2 Average .325 .323 .316 .321 .553 .544 .54 .542 On direct current the resistance of this circuit was . 1298. It will be seen from these results that the addition of the No. 0000 feeder, which reduced the resistance from .248 ohm to .1298 ohm, or nearly cut it in half, reduced the drop with alternating current from 45 . 7 volts per 100 amperes to 32.1 volts per 100 amperes or only about one-third. This indicates that for single-phase railways the most economical use of copper is to place it in the trolley wire only and to so locate the feeding points that proper voltage will be obtained. In general, with a circuit consisting of No. 000 trolley and a pair of 70- pound rails, the drop in voltage with 25 cycle alternating current is approx- imately 60 volts per 100 amperes per mile, but only from 60 to 65 per cent of this voltage represents a loss of energy. With the alternating current system using a trolley and track return, there is an inductive drop in the trolley and rails, with an additional loss in the latter case due to eddy currents and hysteresis. Measurements made upon the Ballston line indicate an apparent trolley resistance of 1.3 times the ohmic resistance, and a rail resistance 6 . 55 times the ohmic resistance. 798 ELECTRIC RAILWAYS. Comparative A. C. and ■>. C. Resistance Trolley and Track, Per UEile of Circuit. ^ p.c. Resistance A.C. Resist. 25 Cycles v + - A.C. Two trolleys in series One trolley and double track .... Two trolleys and double track . . . Double track alone Ohms. .318 .167 .088 .0174 Ohms. .417 .259 .155 .114 1.31 1.55 1.76 6.55 The impedance of an electric railway conducting system consisting of a trolley wire overhead, placed in some sort of location above the two track rails, is a still further complication, and this impedance comprises the resist- ance and reactance of the trolley wire, and if of catenary construction, the messenger wires; the resistance and inductance of the rails; the inductance of the circuit bounded by the rails and the trolley wire, and the mutual inductance of the currents in the two rails. The calculation of this imped- ance is therefore hardly possible and in all cases its value must be deter- mined by experience. TESTS OE STREET RAILWAY CIRCUITS. The following tests are condensed from an article by A. B. Herrick in the Street Hallway Journal, April, 1899. The following instruments will be required : A barrel water rheostat to take say 100 amperes. A voltmeter reading to 600 volts. A voltmeter reading to 125 volts. An ammeter reading to say 150 amperes. A pole long enough to reach the trolley wire, with a wire running along it having a hook to make contact. Use one generator at the station, and have the attendant keep pressure constant. Test for Drop and Resistance in Overhead Lines and Returns. The car containing the above equipment of instruments is run to the end of the section of conductor which it is desired to test, where a line circuit- breaker divides the sections. The instruments are then connected as shown in Fig. 153. It is clear now that if the switch G be closed, current will flow through the rheostat and be measured by the ammeter. We now have the trolley and feeder B for a pressure wire back to the station, and the reading of voltmeter C therefore gives the drop between the station and the point A in the feeder and trolley carrying the load. Voltmeter D shows the drop across the rheostat ; and if the sum of readings C and D be deducted from th« station pressure, the difference will be the drop in the ground return. TESTS OF STREET RAILWAY CIRCUITS. 799 Fig. 163. The station pressure can be taken by changing the lead of voltmeter C down to F as shown by the dotted line. The drop on A and its resistance having been found, the trolley-pole can be swung around and the same data be determined for the circuit B. To Read the Ground Return Drop Directly. Open the station switch on that feeder that is being used as pressure wire, and ground the feeder to the ground bus through a fuse for safety. Connect the instruments as shown in the following cut ; then when the switch G is closed and current flows, the drop from A to F read on voltmeter C will be the drop in the ground return from F to X. Pig. 154. 800 ELECTRIC RAILWAYS. To Determine Drop at End of line. For use on double-track lines only, unless a pressure wire can be run to the end of line from the last line circuit-breaker. Break all cross connections from feeder to trolley-wire for one track, as at n ; connect this idle trolley to the next one back toward the station, as at C, then make the tests as in the two methods described above, connections being shown in the following cut. Fig. 155. To Determine the Condition of Track Bonding:, and the Division of Return Current tbroug-h Rails, Water or Ga§ Pipes, and Ground. The cut below shows the connections for this test as applied to a single track, or to one track of a double-track road. Ground the feeder A at the station, or rather connect it to the ground bus through a fuse. Then connect the track at C to A by the pole E through the ammeter M. The drop between points F and D will be the drop through the rail circuit between C and D, due to the current flowing. If connection be made to a hydrant, or other water connection, and to a gas-pipe, as at X, still retaining the rail connection at C, more current will Fig. 156. TESTING RAIL BONDS. 801 flow through ammeter M, due to providing the metallic return through A for the water-pipe, and the first reading of the ammeter M is to the second reading as the resistance of the water-pipe is to that of the rail return, and the current returning to the station will distribute itself between the two paths in proportion to the readings mentioned. If ammeter G be read at the same time, the difference between its reading and the sum of the other two readings will be the amount of current returning by other paths than the rail and water-pipe. If C is near the station it may be necessary to break the ground connection between rails and bus, so that all current may return over the metallic circuit A. To determine condition of bonds, move the contact C back towards D, and the decrease in drop as shown by the vm. will be very nearly proportional to the length of track, except where a bad or broken bond may be located, when the change will be sudden. XESTI\(» Hill ltO\B>*. It is not commercially practicable to measure the exact resistance of rail joints, as such resistance is small under ordinary circumstances, and all the conditions vary so much as to prevent accurate measurement being made. The resistance of rail joints is therefore measured in terms of length of the rail itself, and there are numerous instruments devised for the purpose, nearly all being based upon the principle of the wheatstone bridge, the resistance of the rail joint being balanced against a section of the rail, as in the following diagram. MILLI-VOLTMETER , CENTER-ZERO < Vl 0HM Fig. 157. Diagram of Method of Testing Rail Joints. A Weston or other reliable milli-voltmeter, with the zero point in the mid- dle of the scale, is the handiest instrument for making these tests. The points b and c are fixed usually at a distance of 12 inches apart, the point a is then moved along the rail until there is no deflection of the needle when both switches are closed. The resistance of the joint or the portion between the points b and c is to that of the length, x, inversely as the length of the former is to that of the latter, all being in terms of the length of rail, or, Let then, x = distance in inches between points a and c, y = distance between the points c and 6, v = resistance of joint in terms of length of rail, 802 ELECTRIC RAILWAYS. and if x — 36 inches and y = 12 inches, then 36 v = — = 3 times its length in rail. Another scheme for testing rail joints is pointed out by W. N. Walmsley in the M Electrical Engineer," December 23, 1897. In the following cut, the instrument is a specially designed, double milli- voltmeter, both pointers having the same axis, and indicating on the same scale. DOUBLE MILIVOLTMETER WAL'MSLEY'S RAIL, TESTER Fig. 158. The points ab are at a fixed distance d, the point c being movable along the rail. Points a and b are set on the rail astride the joint, as shown ; the point c is then moved along the rail until the pointers on the instrument coincide, indicating the same drop. Then the resistance of x is the same as dj in terms of the size of rail used. Harold P. Brown has devised an instrument for testing rail joints with little preparation. It consists of two specially shielded milli-voltmeters of the Weston Company's make, put up in a substantial wooden case, the top of which is made up in part of two folding legs which, when unfolded, cover six feet of rail. These legs form one length, which is divided by slots into two lengths, one of one foot, the other five feet long. The instrument is placed alongside the track in such position that the leg rests on the rail, and the joint to be tested is between the ends of the shorter branch or leg, while five feet of clear rail are included between the ends of the longer leg. The instrument terminals are connected to small horseshoe magnets, that fit into the slots in each leg, and when rested on the rail always make the same pressure of contact, the poles being amalgamated and coated with a special soft amalgam, called Edison Flexible Solder. With the five feet of rail as a shunt, the instrument will read to 1500 am- peres. There are several separate resistance coils and binding-posts supplied for different sizes of rail in common use, so that the dial of the milli-voltmeter needs but one scale. The second milli-voltmeter measures the drop around the one foot of joint, and has coils so arranged to permit of reading .15, 1.5, 15. volts. A reading of the current value is taken from the five feet of rail, and a simultaneous reading of the drop across the joint and one foot of rail is also made. The resistance of the latter is then found by ohm's law, B = E TESTING RAIL BONDS. 803 A B C Fig. 159. Brown's Rail-bond Testing Instrument. Street Railway Motor Vesting*. Barn test for efficiency : — Put a double-flange pulley on the car axle for the application of a prony brake, pour water inside the pulley to keep it cool. Use common platform scale, as shown in cut. Fig. 160. Then let D = distance from center of axle to point on scales in feet, measured horizontally. n — 3.1416, R ■=. revolutions per minute, E = voltage at motor, Iz=. amperes at motor, Tzz force applied to balance scales, in pounds. Then B. H. P. = 2tt BR T 33,000 B. H. P. at 500 volts = EI [2,DRX^\T E J 33,000 =z E.H.P. supplied to motor. 500 / 746 Efficiency of motor = : E.H.P. supplied to motor at 500 volts. B.H.P. X B.H.P. at 500 volts E.H.P. ~ E.H.P. at 500 volts Draw-bar Pull and Efficiency Test Without Removing 1 motor from Car. Rig up lever as shown in cut, being sure the fulcrum A is strong enough to stand the pull. Posts, as shown, make good fulcrum ; have turn buckle F for taking up any weakness. 804 ELECTRIC RAILWAYS. Fig. 161. Let D = diameter of car wheel in feet. 7T = 3.1416, T== force on scale in pounds, L == length of long arm of lever, L, ■=. length of short arm of lever, R = revolutions per minute. Place a jack-screw under each side of the car, and lift the body until there is only friction enough between wheels and rail to keep the speed of revolu- tions down to the normal rate. Then Draw-bar pull := T and V T-^-DnR B.H.P. : & 33,000 and the efficiency is the same as before, B.H.P. i.e. E.H.P. : efficiency. Mr. A. B. Herrick has devised a testing-board for street-railway repair shops that will greatly assist in making all inspection tests, and which is described in the " Street Railway Journal " for January, 1898, pages 11 and 12. Testing 1 Drop in Railway Circuits. — For this test use can be made of any car that is in good order, and it should be carried out after the last car is in the barn, and the track is clear. Run the car over the line starting from the point nearest the power house, making the test at any points that may be selected. The following cut No. 162 shows the arrangement of instruments. Fig. 162. E = drop a to b without load, and in clear dry weather this should be same as at the switchboard. In wet weather or with poor insu- lation the drop without load may be considerable. FAULTS AND REMEDIES. 805 Ei = drop a to 6 taken with the brakes set and the controller on the first notch. / = amperes of current under conditions E\. E — E\ = e — drop in circuit due to current /. R = j — resistance of entire circuit of trolley wire; feeders, and rail returns. Ri = resistance of feeders and trolley wire as calculated from their known dimensions. R — Ri = resistance of the return circuit. FAULTS A XU REMEDIES. Car Will not Start: a. Turn on lamps ; if they burn, trolley and ground wires are all right and current is on line. b. If lights die down when controller is thrown on, trouble may be poor contact between rails and wheels, or car may be on '• dead " track. c. If car works all right with one controller, fault may be open circuit, or poor contact in the other. Throw current off at canopy, or pull down the trolley and examine the controller. d. See that both motor cut-outs are in place. e. Fuse may be blown ; throw canopy switch and replace. /. See that motor brushes are in place and intact, and make good contact. a. Car maybe standing on "dead" or dirty rail ; in either case connect wheels to next rail by wire. It is better to open canopy switch while con- necting wire to wheels, or a shock may be felt. h. Ice on trolley wheel or wire will prevent starting. Sparking* at Commutator Brushes: a. Brushes may be too loose ; tighten pressure spring. b. Brushes may be badly burned or broken,' and therefore make poor con- tact on the commutator. Replace brushes with new set, and sandpaper commutator surface smooth. c. Brushes may be welded to holder, and thus not work freely on commu- tator surface. d. Commutator may be badly worn and need renewing. e. Commutator may have a flat bar, or one projecting above the general surface ; commutator must then be turned true in lathe. /. Dirt or oil on commutator may produce sparking ; clean well. Flame at the commutator may be produced by : — a. Broken lead wire or coil, producing a greenish flame, and burning two bars usually diametrically opposite each other. If left too long the two bars will bebadly burned, as will also the insulation between. Temporary relief can be had by putting a juniper of solder or of small wire across the burned bar, connecting the two adjacent bars to each other ; one juniper is enough. b. A short-circuited field coil, or a field coil improperly connected, will produce flare at commutator. Short-circuited coil can be found by volt- meter test across terminals showing drop in coil. Wrong connection can be detected by pocket compass. Incandescent lamps sometimes burn out or break. Replace with new ones. If they do not burn when switch is on, a. Examine each for broken filament. 6. Examine for poor contact in socket. c. Examine switch for poor contact or broken blades. d. Examine each part of circuit, switches, line, and sockets with magneto, which will locate opening. The wire may be broken at ground or trolley connections. Brakes fail ta Operate: In great emergency only, throw controller handle to off, reverse reversing- switch, and turn controller handle to first or second notch. 806 ELECTRIC RAILWAYS. In sliding down grades, or when there is time, proceed as follows : a. Throw controller handle to off point. b. Throw canopy switch off. c. Reverse reversing-switch. d. Throw controller handle around to last notch. Both methods are more or less strain on the motors, but the second is somewhat less so than the first. Grounds : Either on field or armature coils will nearly always blow fuse ; it can then be tested out. Bucking 1 : When running along smoothly, a car will sometimes com- mence jerky, bucking motions, and should be thoroughly examined at once. It may be due to a ground of field or armature that may short-circuit one or the other, either fully or intermittently. Injured motor may usually be located by smell of burning shellac, and can be cut out at the controller, •and the car run in with the good motor. Mud and water splashing on commutator will sometimes produce bucking, and often a piece of wire caught up from the track may do the same. Miscellaneous "Xote. Experiments show that four arresters per mile of trolley wire are plenty for safety. Green wooden poles should not be painted for at least a year after they are set, as the paint will peal off and not give good results. Loose ornamental joint caps frequently used on iron or steel poles collect moisture and rust out the pole. Wiring- Diagrams for JLightiiig; Circuits on Street Cars. trLH i^^^^ Head Light i Fig. 163. Diagram for two Circuits Fig. 164. Diagram of Wiring to Headlights, Platform Lights and permit use of 32-p. Headlight. Sign Lights Interchangeable. r X — pn rf==^ L/ • Three Point ^SJ^ I j: | j j jf-— ' Switch ^ 1 \J^' Head. Light Head Light '■3 F Fig. 165. Diagram of Wiring where Fig. 166. Same as above but three- Headlights are placed on Hoods. point Switch located on Trolley End of Car. SPECIAL METHODS OF DISTRIBUTION. 807 Fig. 168. Same as above three-point Switch. except Fig. 167. Diagram of Wiring for five-light Circuit with four-point Switch for Headlights and Plat- form Lights. Special methods of Distribution. For cases requiring excessively large currents carried a considerable dis- tance, or for ordinary currents carried excessive distances, it is usually economy to adopt some special method ; and among those most commonly mentioned are : the three-wire system, the booster system, the substation system. Three-Wire System. This system, patented some time ago by the General Electric Company, has been seldom used, and where used has met with little success, owing to the difficulty met in keeping the system bal- anced. The diagram below will assist in making the method plain. Two 500-volt generators are used, as in the lighting system of the same type. The rail return is used as the neutral conductor; and if both trolley wires could be made to carry the same loads, and to remain balanced, then the rail return THREE WIRE SYSTEM Fig. 169. Three-Wire System. would carry no current, and no trouble would occur from electrolysis. The overhead conductors could also be very much smaller, as currents would be halved, and the full voltage would be practically 1000. A balanced three-wire system has been proposed and is in limited use abroad in which the car carries two trolley poles, making contact with both trolley wires. The motor equipment is in duplicate, thus each set of motors is fed from 600 volts making the current through the return practically zero, and the whole equipment forming a balanced three- wire system in itself. This system is the only practical three- wire system and offers some advan- tages for transmitting large amounts of power over considerable distances. The Booster System. — Where current must be conveyed a long distance, say five to ten miles, and be delivered at 500 volts, it is hardly good economy to install copper enough to prevent the drop ; and if the volt- 808 ELECTRIC RAILWAYS. age of the generator be raised sufficiently to deliver the required voltage, the variations due to change of load will be prohibitive. In such cases a "booster" can be connected in series with the feeder, and automatically keep the pressure at the required point, as long as the generator delivers the normal pressure. The "booster" is nothing more than a series-wound dynamo, connected so that all the current of the feeder to which it is attached flows through both field and armature coils, and the voltage produced at the armature terminals is added to that of the line, and as the voltage so produced is in proportion to the current flowing, it will be seen that the pressure will rise and fall with the current. This is now used in many instances, both in lighting and for railway feeders, and especially in feeding storage batteries, and has met with entire success. The following cut is a diagram of the connections. 6 TO 1tt M1UE8 OVERHEAD RETURN BOOSTER SYSTEM Fig. 170. Return Feeder Booster. — Major Cardew, Electrical Engineer for the Board of Trade, some time ago devised a method of overcoming exces- sive drop in track return circuits by the use of insulated return feeders, in series with which he placed a booster. The booster draws current back toward the station, adding its E.M.F. to that in the feeder. Cardew used a motor generator, the series field of which was separately excited by the outgoing feeder for the same section of road. Thus the volts " boosted " were in direct proportion to the current flowing. H. F. Parshall, in adopting the return feeder booster for some of his work in England, used a generator in place of the motor generator of Major Cardew, exciting the field by the current flowing out on the trolley feeder, thus producing volts in the armature in proportion to the current flowing. The following diagram shows Parshall's arrangement. TROLLEY WIRE SEPARATELY EXCITED) '-\- VGENERATOR A BUS BAR* AT 6TATJOI* T* GENERAT0R8 Fig. 171. Modification of Major Cardew's System of Track Return Booster for Preventing Excessive Drop in Rail Return Circuits. ELECTRIC RAILWAY BOOSTER CALCULATIONS. 809 Electric Railway Booster Calculation*. (H. S. Putnam.) The following method of calculating the size and characteristics of electric railway boosters, and the graphic representation of the results will be found useful. A\ A 2 , A 3 , A*, etc., = load in amperes at various points along the line. These loads should be taken from schedule, and should ordinarily represent an average maximum condition. R 1 , R 2 , R 3 , R*, etc., = feeder resistance (including trolleys) to the corre- sponding load points. 2 = drop in volts to the point at which it is proposed to feed into the system with the booster. V = allowable volts drop in feeder system with the booster in circuit. / = amperes in booster. E = volts boost. E V = r = ratio of volts boost to amperes boosted. Rb = resistance of booster feeder. R = resistance of feeder system to point selected for the booster feed. Then assuming that all the load beyond the point at which it is proposed that the booster should feed into the system is concentrated at the latter point, 2 = A* R* + A 2 R 2 + A 3 R\ etc. —A b R. 2- V I R V Rb =j + p. E = / X p. V - Rb -f These equations give the necessary data to determine the required size and ratio of the booster and its feeder. In case it is desired to install a negative booster, the same method is followed. In case the load is uniformly distributed over the line, or is assumed as distributed in that manner, the voltage drop at any desired point on the line is found from the equation: v (2 L - d + 1) dIR 2= ^ ' in which L = total length of line in feet. d = distance to point selected. I = amperes per foot. R = resistance of feeder system per foot. If desired these units can be expressed in 1000 feet or miles or any other unit of distance. When the drop to the end of the line is desired, this equation becomes: 810 ELECTRIC RAILWAYS. It is often desirable to represent these calculations graphically. Special cases are shown in Figs. 172, 173 and 174, in which the potential diagram is shown for different conditions and schedules. In the preparation of these diagrams it will be found convenient to plot the schedule and feeder and return resistances on the same sheet. In Fig. 172 it is seen that a negative booster is not required though one is included. Fig. 173 shows a system in which a booster is used at either end. Fig. 174 illustrates a different and more severe operating condition than shown in Fig. 173. ^&- >A £ C E ^_- T S -^i 906 v "^ Boo 3tei Diagran AJe, Maximum' Conditions 1 a BC ars each car }£ fullj A< c. Current Single ' rrack £ 600 -!< *>.< I ou lie Tr >lley 1 -en !* •4 .«.<■ OFt, ) Positive Joo'ster **- J&J J 395 Ac i P . - 295 V. 3S Vm p = d7~ 75 R = 4 "" .0 500 li - • 1 05 Am p. - -95 At ap. ^V H ■J^t " .5 ij nn An 2-25 M Tr uii &x A*n W Ai a p. 400 ^ " .4 300> ^ r .8 ) Generator ^+ \ce 8Qf)A mp. - 55C V .-* ** ck Re lur at 200 *> — * J 100 - " ar + 1 )rc i9i )V lit J .-* •n _- S -^ _ ^ E :b£ L Iracl :B ail — - 1 r 2 j— - - - 5 00 10 X)t louoo 20DO0 86 JQO 30000 35C 00 40( M q, 1 ** Forn tula C- A JD P B .-: er i r Res Bolster Fee ler _V + 1 ) Negatii el JooBterj ■§F xb* P- Ratio Voits to Am >eres i a 1 soostei 9 )1 Am P- 361V ff\ C *• i £ - Volts Drop witbou Boos tor to 6 5 fc 300 p rlOt) - &!0 B- Resi stance j •(• P "I 6 , r"~ 7 a Allov able Volts Drop at 6 1 Mill Fig. 172. Kelvin's Law can be applied to the booster distribution as well as to other methods of distribution. In most cases, however, it will be found that the voltage requirements will govern. The question as to whether a booster, more feeder copper or a sub-station shall be employed, is one which must be determined from the annual charges against the investment and the cost of the power lost in each method. In calculating the cost of the power lost, the load factor must be considered. In selecting a booster care must be exercised that its overload capacity shall be sufficient to take care of the maximum operating condition which occasionally arises in any system where boosters are likely to be employed, namely, when all the cars are accelerating at once. As such occasions may be rare, it is only necessary that the voltage shall be maintained above the minimum voltage at which contactors will operate, if such contactors are employed, that the booster motor shall carry such overload, and that the machines shall properly commutate at the overload current. By varying the value of "p" the ratio of the volts of boost to the amperes boosted, the size of the booster feeder and the amount of power lost in the booster system is changed. By Kelvin's Law the annual charges on the boos- ter feeder and booster should equal the annual cost of the power lost in the booster system. ELECTRIC RAILWAY BOOSTER CALCULATIONS. 811 -- ia6oq tnoA w» v 9 CO 1 s 3 *i 1 i § i § i § i § * 812 ELECTRIC RAILWAYS. © ?. oj <-• O S ■* a> §2 ELECTRIC RAILWAY BOOSTER CALCULATIONS. 813 Series Boosters for Railway Service. — The amount of varia- tion allowable in the voltage characteristic of a series booster for 500- volt railway service is an important factor in its design, as it largely determines the amount of material required and therefore the cost. The actual voltage characteristic of commercial series boosters is not a straight line but a curve, which at partial load will be above the theoretical line as shown in the accompanying diagram. The amount of variation from the straight line is principally affected by the saturation of the magnetic circuit ; if the saturation is high, the variation of the voltage characteristic will be great. By increasing the amount of iron in the magnetic frame and there- fore keeping the saturation low, the voltage characteristics can be made to more nearly approximate a straight line; but, obviously, a machine so designed is more costly than a booster having a voltage characteristic departing further from a straight line. These facts are particularly im- portant in cases where high voltage boosters are used, as may be seen from the following example: In the accompanying diagram of a 200-kilowatt, 400-volt booster, the potential at half load is 240 volts, that is, 40 volts, or 10 per cent of the full 400 * / soo / / / 200 / / / / 7 V / 125 250 375 500 Amperes Fig. 175. Characteristics of a 200-Kw. 400- Volt Booster. /oad voltage, higher than a theoretical straight line characteristic. Had less iron been used in the magnetic circuit the potential at half load would have been higher, and estimating an inctance of 320 volts, the potential would be 120 volts too high, or, at 695 volts, assuming a generator potential of 575 volts. This high voltage might burn out the car lights and would increase the speed and subject the motors to a severe strain. While a straight line characteristic is not essential, the variation from a straight line must be kept within reasonable limits. Unless otherwise specified, the voltage characteristic variations of all series boosters of different potentials should not exceed the following values at partial current load and at constant speed: Full Load Voltages of Boosters. 50 to 100 volts 100 to 150 volts 150 to 250 volts 250 to 500 volts Maximum Variation of Full Load Voltage at Partial Load. 20 per cent 15 per cent 12^ per cent 10 per cent 814 ELECTRIC RAILWAYS. Temperature. — After a run of twenty-four hours at full rated volts and amperes, the temperature of no part of the machine should be more than 40° C. above the temperature of the surrounding air, provided the conditions of ventilation are normal and the temperature of the sur- rounding air does not exceed 25° C. If the temperature of the surrounding air differs from 25° C., the observed rise in temperature is to be corrected by one-half per cent for each degree centigrade that the temperature of the surrounding air differs from 25° C. The booster should be capable of standing an overload of 25 per cent of the full load ampere and volt capacity of the machine for one-half hour; and as this corresponds to the 25 per cent voltage overload, the overload capacity in kilowatts will be about 50 per cent. The boosters should be capable of standing a momentary overload of 50 per cent of the rated capacity in amperes at full load or about 100 per cent in kilowatt rating. §IB.«TAT10V SYSTEM!. Where traffic is especially heavy, and a railway system widespread, it is now the practice to use one large and economical power station with high- pressure generators, now invariably polyphase alternators, and to distribute this high-pressure alternating current to small sub-stations centrally located for feeding their districts, and there changing the current by means of static transformers and rotary converters into continuous current of the requisite pressure, in the case of railways 550 to 600 volts. The following diagrams will assist in making the system plain. SUB*8TATI0M NO. 2 STATIC TRANS- FORMERS SUBSTATION, NO. 1 DISTRIBUTION FROM SUB-STATIONS Fig. 176. The universal use of rotary converters has led to many similar designs of sub-stations. It is customary to install the rotaries in buildings designed for the purpose and Figs. 178 and 179 show a typical station in plan and eleva- tion. As each sub-station is in reality a complete supply station.it is neces- sary to install suitable protective devices for both high-tension alternating SUB-STATION SYSTEM. 815 816 ELECTRIC RAILWAYS. current and 600 volts direct-current circuits. The necessary connections are shown in Fig. 180. . In Fig. 181 is shown a cross section of one of the latest types (1907) as developed for the United Railways and Electric Company of Baltimore, by Mr. L. B. Still well. This station has an unusually large capacity for one center of load. In designing sub-stations, their equipment should be based upon taking care of the maximum load of the stations, while a central power station operating through rotaries may be designed to take the average load only. PLAN Fig, 178. Rotary Converter Sub-Station. SUB-STATION SYSTEM. 817 Fig. 179. Rotary Converter Sub-Station. 818 ELECTRIC RAILWAYS. *s[swiNGINa!c.S.F. MNElj c.S.R. PANEL I A - T ' RJ » BRACKET! $oov AMp 600V. 300 K.W. JS0OO0VJ ■♦■RAILWAY BUS TO BELL LOwfl VOLTASE — RELEASE CONNECTIONS SHOWN DOTTED SHOULD BE MADE ON I. C.S.R. PANEL 0*lY AJTJTfAHZl 80000*. /"~ 60000 V u i.e. bIssl INCOMING LINE Fig. 180. Diagram of Connections for Proposed Rotary Converter Sub-Station. SUB-STATION SYSTEM. 819 Fig. 181. Cross section of typical large sub-station (1907) 12,000- kilowatt 13, 000- volts alternating current, 575- volts direct current. L. B. Still well, Engineer. Portable Sub-Stations. — Many roads have a heavy traffic on certain lines for a portion of the year only, thus making it hardly feasible to expend a large sum in a permanent sub-station. For such cases, the porta- ble sub-station has been designed, consisting of a box car containing step- down transformers, rotary converter and all necessary protecting devices. Such a sub-station can be run out on any line having a transmission system connected up, and put into service in a very short time. It therefore forms a reserve sub-station. A plan, elevation and diagram of connection, of a typical portable sub-station is shown in Fig. 182. A portable sub-station having as high as 1000-kilowatts capacity is in use, see Street Railway Journal, November 4, 1905 and June 23, 1906. 820 ELECTRIC RAILWAYS. W & 2 THIRD RAIL SYSTEMS. 821 THIRD RAIL SYSTEMS, {By F. R. Slater.) For certain classes of electric railways, such as elevated, interurban and underground, a steel conductor insulated from and alongside the track, commonly called the third rail, is much used in place of the copper over- head trolley wire. This conductor is easily installed, cheaply maintained, presents a large surface area for conducting and collecting the current, and is, therefore, particularly suitable for high speed and heavy service. With costs cal- culated on the basis of equal conductivity in rail and trolley wire, the third rail is the cheaper, except where the necessary trolley wire would be of considerable less conductivity than would be obtained with the smallest size of steel rail that would ordinarily be used. Even in such cases the lower cost of maintenance, together with the advantage of adaptability (particularly in the case of terminals, yards and very heavy high speed service), will frequently offset the higher first cost of the third rail and make it the preferable means of conducting the current from the power station to the car motor. With the coming of the heavy high-speed service of the past few years, the resistance of standard "T" rails has been found to be so high, that rails of higher specific conductivity were sought, and specifications have been drawn, usually based on the fact that the conductivity of a metal is generally directly proportionate to its purity. Resistance of Rails with Varying* Composition. — Mr. J. A. Capp, of Schenectady, conducted a series of tests of steel for electric conductivity. He says in part: "In most cases the purity of the iron specified for such rails has been so high, that not only was it difficult to obtain, but the iron was also correspondingly high in price. One of the factors governing the choice between a third rail and a trolley wire is the relative price of steel and copper, allowance being made for the difference in conductivity. Hence a balance must be struck between high conduc- tivity (which is equivalent to saying a high degree of purity or freedom from the usual metalloids associated with iron) and the cost of producing the steel of the composition necessary for the conductivity required. "Table XVII below states the electrical resistance and the chemical com- position of 47 samples of steel, and Table XVIII similar data on 7 samples of wrought or refined iron: 822 ELECTRIC RAILWAYS. + + lO "*C0 1> 00 10 COOi'*T-lCOOOO'-U>(M05TH05COOC01>COC505rHCO rHrH^HT-lCq(Mi-l(M(MT-lrHOq^lOOrHOOrH^HrHiH CO ^rfH COt>- C5C0 iO CO tH CO t^ CO I> 00 00 COa50COOOO^TH0500Ct^Ttll>CO^COlOOOOCO'^CAlOCOCO>-lrHr-l00500 05 rt^cq cdi> „ COOOi-l CO ©00 coLo^cqoo^^iocoT-iTjias^coT-iT-iT-Hr-icDt^co^o t-hOOOoOOOOOOOOOOOOOOOOi-i I Ci CO>OC5 CD CO ^ 00 CO CO 1> Tj< lO iO050^rHOOOO^T-iC00005T-icD00iOtO00t^t^00 OOOOTHT-H^HrHOOOOOOOOOOOOOO §■11 o CDOC io CD Ci Oi 00 00 -^ ^ ^ CO CO # Oi CO* CO* (N CM* t-J rH r-i t-I H CD O O Oi Ol Oi 05 OS Ci C> OS Ci 00 1-8 l! So COO J* .2° ft CO00r^^iO00O>CiCOO5t>>CDCDCD^T}HlCiCiCN)Oia500 "tf »0 CuOOiO>05Ci0505050500iOiCD C-trHTHr-lT-lTHTHT-lC* s o + + OOcO 00 00 tH ^ OS t^ ^totoTtC0C00300O>— (COtOCOO'^'-trHCOCOt^CO'^OSOiOO OiOiooi>oooooc^ooi>t^oo^t^i>^^cocDi>coco # ^»oco in CD .... ROCOCO 00t^ rH COOOOi . ^ . NNM»hhHNONOiOHOHO^HOO^CO CO qqqHBHHooqqqoqqqqqqooho . o • QQ tJh CO COOS tO tH tJH 00 00 ^ «cH t^ lO lO rH rH CNJ CO *■« lO t-( CO l> ^ "# CO "tf rH tO OOOOOCOWOOOOOOOHOOOOOOOOhO Ph' 00 T*0 OStOrHCO00cO1^00tOrH00l>tOt^OSl>rH00©c>q00tOr-i(NOS -* ^ ^ # ^ # 'tf "3 to "* to rJH tjh lO r^OOI>00«DtOCT>r-(iO CCM(NHH(MHHHH(M(NHNNHCqMO(MHHHINO 6 1 to '55 4) • ftrH &II P ococqoqtocqcocO'^cqiococq^cO'HOcocDOoooooooo 05 tjh tjh tjh co NHoqqoj os os i> t^ i> r* # co co # co ^ # ^ # co t>* ^ 1> t>* t>* 1>* I> 1> 1> t>* 1> 1> t»* CO :g O S3 a o O a . 0>T3 02 lL 1-8 ^05 rHiHrHrHTHC^CNCQC^C^C^CQC^C^OQO^C^COCOCOCOCOCOCO^O o a a OQ "53 • 0) o "o 9 ad o o o o o to toto _ to iO o .o ..ooo .ooooo .oooooooooo OiOSOCOCOOSOSOSOSOSOSOSTHOSCiOSOSOSOSCSOOOSOSOS THrHC^t>lC^THrHrHTHrHrHTH(^rHrHrHrHrHrHrHCN CD rH O 00 00 00 CO CO Oi CO CO CN CNJ h rH O 00 t> O»0 tO* tH t** Ti* rJH tJ* «rj* CO* CO* CO* CO* CO* CO CO* CO* CO* CO* CO* CO CO* CN CO *-* cq 'I T— ( q q -H O J^ 00 o CO o ^ CO o 00 «* t^ oo io d *0 Tjl **. CO CO CO o T^ r- CO *o 10 CO o CO t>. o zn i— 1 i—i q q q q a a 8 • 00 >o CO q rH q c5 o Ih fl c3 CO 00 rH rH CO i> CO +a 00 "*. "*. •H I> rH 1—1 SB 1> t>* t»* 1> CD* CD* CO* "8 o p* St3 CQ m TO 03 00 1> o r^ O «* ^l CO CO CO* CO* o oo rj* CO CO* CO CD 13 fl o c5 O. . o o o fl a -*5 QQ 1 ad iq o o »o »0 *o So »o* CO CO »o* iO* ^ »o* H assage of any of their rolling stock, either freight or passenger. This ocation is as follows: "The third or conductor rail shall be located outside of and parallel to the track rails so that its center line shall be 27 inches from the track gauge line and its upper face 3^ inches above the top of the track rail." THIRD RAIL INSULATORS. 831 From Top of From Track Relative Location of Third Rail Third Rail Gauge Line on Different Railway Systems. to Top of to Center of Track Rail. Third Rail. General Electric Railroad, Schenectady 3" 28" Met. West Side Elevated, Chicago .... 6J* 20£" Lake Street Elevated, Chicago 6¥ 20£" South Side Elevated, Chicago e>r 20i" Northwestern Elevated, Chicago 6i" 2or Brooklyn Elevated, Brooklvn 6" 22|" Manhattan Elevated, New York .... 7V 20f* Albany & Hudson, New York 6" 27" Boston Elevated, Boston 6" 20|" Aurora, Elgin & Chicago, 111 6 5-16" 20i" 27" Columbus, Buckeye Lake & Newark. Ohio 6" Columbus, London & Springfield, Ohio . . 6" 27" B. & 0. R.R., Baltimore 2f* 24" N.Y., N.H. & H. R.R., Connecticut . . . W Center Central London. England w Center THIRD RAIL 1.\H'L1T01|K, The requirements for a third rail insulator are: (a) That it shall have sufficient strength to carry the weight of the rail and not crush under the vibration of passing trains. (b) That its insulating body shall be made of a thoroughly vitreous material, practically impervious to heat and moisture, and having its exposed surface well glazed. (c) That its resistance shall, when wet over its entire surface, be 1 megohm at least. (d) That it have a drip edge between the rail and ground. (e) That the portion upon which the rail rests shall allow free move- ment of the rail, laterally and longitudinally _ to allow for expansion and contraction, and vertically to allow for depression of ties during the passage of trains. (/) That it must be capable of easy and quick renewal. Those here illustrated show the two general types which have been most widely used (Fig. 184 and Fig. 185). Fig. 184 consists of a metal base surrounded by an insulating body of vitreous material to which are clamped the clips which hold the rail. Fig. 185 is practically the same, except that in place of the clips clamping the insulating body there is a metal cap setting over it, having ears which may or may not be bent over the rail. Fig. 184. Fig. 185. These insulators are usually placed 10 feet apart, except on sharp curves, where they are generally placed on 5-foot centers in order to keep the rai' up to gauge, to allow for the expansion and contraction. The rail is usually anchored at the two center insulators, any movement being taken up at the joints where a sufficient distance has been left between rails for 832 ELECTRIC RAILWAYS. the purpose. This is either done (1) by making the portion of the insu- lator upon which the rail rests in such way that it may be bolted to the web of the rail, or (2) by making the portion of the insulator upon which the rail rests with a lug that fits into a slot punched in the bottom flange of the rail. Where the shoe or current collector leaves the third rail at the ends on straight track and at the side at switches and crossovers, suitable in- clines must be provided, because the shoes normally hang lower than the top of the third rail. (See Fig. 186.) ^ Tbircl Rail Shoe. — These shoes are of practically but two types viz., the link shoe and the slipper shoe. The link shoe is shown in Fig. 187, and is attached to the coil spring seat of the truck, and the shoe proper is suspended by two links from the yoke Motor Lead Fig. 187. Link Shoe, used on Manhattan Elevated Railway. which is in turn bolted to castings on the shoe beam. This type of shoe is not entirely satisfactory because it has a tendency for the shoe to ride on its nose when the speed is high, and does not permit of adequate pro- tection of the rail from the weather. THIRD RAIL INSULATORS. 833 The slipper shoe shown in Fig. 188 is also carried from the shoe beam, which in turn is fastened to the spring seat. This type of shoe is quieter Fig. 188. Shoe. sparks less under heavy currents and allows of the use of a top guard, usually a plank or wide channel section of light steel. (See Fig. 189.) P Z\W I^S .rq Tie Fig. 189. Direct metallic connection is maintained from the shoe to the motor in the types of shoes shown, by copper terminals bolted to both yoke and shoe, these two being connected to each other and to the motor by extra flexible copper leads. 834 ELECTRIC RAILWAYS. Hew York Central Third Rail. — This arrangement of contact rail is the joint invention of W. J. Wilgus and Frank J. Sprague, and as will be noted in the illustration, is supported every eleven feet by iron brackets, which hold the insulation blocks by special clamps. These blocks, which are in two pieces, are six inches long by f inch in thickness, and are inter- changeable. Between supporting brackets the upper part of the rail is If 2 6 To Gauge Line Laterally and Longitudinally Section of Protection Sheathing Fig. 190. Details of Third Rail Construction, New York Central R.R. covered by wooden sheathing, which is applied in three parts and nailed together. At the joints where the third rail is bonded, and at the feeder taps, the wooden sheathing is mortised. This rail is given a little play in the insulators for expansion and contraction, except at certain central points, where it is anchored. It weighs 70 pounds per yard; is of special section and composition; and has a resistivity between seven and eight times that of copper. The under or contact surface is placed 2f inches above the top of the service rail, and its center is 4 feet 9 1 inches from the center line of the service track, or 2 feet 5 inches from the gauge line of the near rail. CONDUIT SYSTEMS OP ELECTRIC RAILWAYS. 835 AJPrROXIlflATE X§TIiVATED COST OF OSTE mi fi ow so«L£ track: or PROIECTED third RAIL. (W. B. Potter.) 6-Inch Channel Iron Protection. 5260' 75-lb. 3" X %¥ conductor rail at $43 per ton (66 tons) . . $2,840.00 528 Reconstructed granite insulators, clamps and lag screws at 40 cents per set 211.00 352 No. 0000 GE 9" Form B bonds at 38 cents _ 134.00 $3,185.00 5280' 31^-lb. 6" channel iron guard for conductor rail at $45 per ton (27.71 tons) $1,248.00 792 Malleable-iron guard supports at 36 cents 286.00 176 Malleable-iron fish plates and bolts at 25 cent 44.00 $1,578.00 Approximate labor for installation, including drilling rails and channels 900.00 Total cost $5,663.00 8-Inch Channel Iron Protection. 5280' 75-lb. 3* X 2\" conductor rail at $48 per ton (66 tons) . . $2,840.00 528 Reconstructed granite insulators, clamps and lag screws at 40 cents per set 211.00 352 No. 0000 GE 9" Form B bonds at 38 cents 134.00 $3,185.00 5280' 48-lb. 8" channel iron guard for rail at $45 per ton (42.24 tons) $1,900.00 792 Malleable-iron guard-rail supports at 36 cents 286.00 176 Malleable-iron fish plates and bolts at 25 cents 44.00 $2,230.00 Approximate labor for installation, including drilling rails and channels 900.00 Total cost $6,315.00 8-Inch Wood Protection. 5280' 75-lb. 3" X 2£" conductor rail at $43 per ton (66 tons) . . $2,480.00 528 Reconstructed granite insulators, clamps ana lag screws at 40 cents per set 211.00 352 No. 0000 GE 9" Form B bonds at 38 cents _ 134.00 $3,185.00 5280' Ash plank H" X 8" at $48 (M board feet) in the rough, 5280 board feet $253.00 792 Malleable-iron guard-rail supports for wooden guard plank at 39 cents 308.00 176 Malleable-iron fish plates and bolts at 25 cents 44.00 $605.00 Approximate labor for installation, including drilling rails . . . 750.00 Total cost $4,540.00 CONDUIT SYSTEMS OJF EIECTRIC RAILWAYS. Previous to 1893 many patents were granted on conduit and other -sub- surface systems of carrying the conductors for electric railways, and hun- dreds of experiments were carried on; but it has been only since that year that capitalists have had the necessary courage to expend enough money to make a really successfully operating road. The work was put into the hands of competent mechanical engineers, who perfected and improved the mechanical details, and the electrical part of the problem was by that means rendered very simple. 836 ELECTRIC RAILWAYS. The Metropolitan Street Railway Company of New York, and the Metro- politan Railroad Company of Washington, decided, in 1894, that, by build- ing a conduit more nearly approaching cable construction, the underground electric system could be made a success. The former contracted for its Lenox Avenue line, and the latter for its Ninth Street line. The New York road was in operation by June, 1895; the Washington road by August of the same year ; and they continue to run successfully. While modifications have been made in some details since these roads were started, yet the present construction is substantially the same. These roads were the first to avoid the almost universal mistake of spending too little and building unsubstantially where new enterprises are undertaken. The history, in these particulars, of the development of overhead trolley and conduit roads is to-day repeating itself in the third-rail equipment of branch and local steam roads. The Metropolitan Railroad, in Washington, used yokes of cast iron placed on concrete foundations, and carrying the track and slot rails. The slot rails had deep inner flanges, with water lips to prevent dripping on con- ductors. The conductor rails were T bars 4 inches deep, 13 feet 6 inches long, 6 inches apart, and were suspended from double porcelain corrugated insulators filled with lead and mounted on cast-iron handholes. A sliding plow of soft cast iron collected the current. During the first few months of its operation there were but few delays, mostly due to causes other than electrical defects. Some trouble came from short-circuiting of plows, which was remedied by fuses on plow leads, and a water rheostat at the power- house. The flooding of conduits did not stop the road, although the leakage was 300 to 550 amperes. Under such circumstances the voltage was reduced from 500 to about 300. The average leakage on minus side, when tested with plus side grounded, was one ampere over 6,500 insulators. The positive side always showed higher insulation than the negative, possibly due to electrolytic action causing deposits on the negative pole. The Lenox Avenue line of the Metropolitan Street Railway was the first permanently successful underground conduit line in the United States. The cast-iron yokes were similar to those used on their cable lines, placed 5 feet apart. Manholes were 30 feet apart, with soapstone and sulphur ped- estal insulators located under each, carrying channel beam conductors, making a metallic circuit. At first the voltage was 350, but it was gradually raised to 500. The pedestal support was afterwards abandoned, and sus- pended insulators used every 15 feet, at handholes. At one time iron-tube contact conductors were tried, but they proved unsatisfactory. The details of track construction for underground or sub-surface trolley railroads are essentially of a special nature, and are determined in every case by the local conditions and requirements. They belong to the civil en- gineering class entirely, and will not be treated here in any way other than to show cuts of the yokes and general construction. The requirements of the conduit for sub-surface trolley conductors are first, that it shall be perfectly drained, and second, that it be so designed that the metallic conductors are out of reach from the surface, of any- thing but the plow and its contacts. Another requisite is that the conduct- ing rails and their insulated supports shall be strong and easily reached for repairs or improvement of insulation. The conducting rails must be secured to their insulating supports in such a manner as to provide for expansion and contraction. This can be done by fastening the center of each section of bar solid to an insulated support at that point, and then slotting the ends of the bar where they are supported on insulators. The ends of the bars will be bonded in a manner somewhat similar to the ordinary rail bonding. The trolley circuit of the sub-surface railway differs from the ordinary overhead trolley system in that while the latter has a single insulated con- ductor, and return is made by the regular running rails, the former has a complete metallic circuit, local, and disconnected in every way from track return. The contact rails must be treated like a double-trolley wire, and calculations for feeders and feeding in points can be made after the methods explained for overhead circuits and feeders earlier in this chapter. Feeders and mains are usually laid in underground conduits for this work, and the contact rails may be kept continuous or may be divided into as many sections as the ser- vice may demand, taps from the mains or feeders being made to the contact CONDUIT SYSTEMS OF ELECTRIC RAILWAYS. 837 838 ELECTRIC RAILWAYS. Fig. 193. Drainage at Manhole of Conduit. Metropolitan Railroad, Washington, 1895. h |-\ ^.A-J^j- J °J_ END ELEVATION OF CLIP Pl*N OF CLIP Fig. 194. Clip and Ear for Conduit, Metropolitan Railroad, Washington, 1895. rails at such points as may be determined as necessary. All the insulated conductors should be of the highest class: may be insulated with rubber or paper, but should in any case be covered with lead. Especial care should be taken in making joints between the conducting rail and copper conductor so that jarring will not disturb the contact. Other than the above few general facts it is difficult to say much regard- ing this type of electric railway, for it is so expensive to install that it can be used in but a few of the largest cities, and in every case will be special, and require special study to determine and meet the local conditions. The reader is referred to the files of the street railway journals for complete descriptions of the few installations of this type of electric railway. CONDUIT SYSTEMS OF ELECTRIC RAILWAYS. 839 Following are a number of cuts showing the standard construction of electric conduits as designed and built by the Metropolitan Street Railway Company, of New York. The system of railway may be said to use all the latest methods, including wire-carrying conduits along side or under the tracks, as will be seen by the next cut. The porcelain insulator here shown for supporting the contact rails is very substantial in design and construction, and by its location at a hand- hole is easily reached for cleaning, repairs, and replacement. The plow has also received careful attention, and those now used as standard by the Met- ropolitan Company leave little to be desired. Fig. 195. Section of Conduit, Metropolitan Street Railway, New York. — Standard Work, 1897-98. Fig. 196. Section, Side and End Elevation of Plow, Metropolitan Street Railway, New York. — Standard Work, 1897-99. 840 ELECTRIC RAILWAYS. TOP OF TRAM. RAILJ Fig. 197. Plan and Elevation of Plow Suspension from Truck, Metropolitan Street Railway, ISIew York. — Standard Work, 1897-98. *£^=£*=:=F3SP mg.198. Section and Elevation of Insulator, Metropolitan Street Railway, New York. — Star dard Work, 1897-98. SURFACE CONTACT OH 1 LF(TIH>..tI tCXETIC iYSTEKIS. The development of surface contact systems began even earlier than the use of the overhead-trolley wire, and many patents have been issued on the WESTINGHOUSE SURFACE CONTACT SYSTEM. 841 same. Most of these failed through ignorance of the requirements, and timidity of capital in taking up a new device answers for others. The Westinghouse Electric and Manufacturing Company and the General Electric Company finally took the matter up, and being equipped with vast experience of the requirements, and the necessary engineering talent and apparatus, have each developed a system that is simple to a degree, and is said to cost but half as much to install as the conduit system, ai^d to oifer advantages not known to that or other systems. I quote as follows from a bulletin issued by the Westinghouse Electric and Manufacturing Company. Some Advantages of the System. No poles, overhead wires, or troublesome switches are employed. The streets, yards, and buildings are left free of all obstructions. The facility with which freight cars can be drilled in yards and through buildings, without turning the trolley whenever the direction of a motor car or locomotive is reversed, and the absence of the necessity of guiding the trolley through the multiplicity of switches usually found in factory yards and buildings, is of great advantage, permitting, in fact, the use of electric locomotives where otherwise electricity could not be used. The only visible parts of the system, when installed for street railway work, are a row of switch boxes between the tracks, flush with the pave- ment, and a double row of small contact buttons which project slightly above the pavement, and do not impede traffic in any way. This system can be used in cities where the use of the overhead trolley is not permitted, and if desired the continuation of the road in the suburbs can be operated by the cheaper overhead system. It would only be neces- sary to have a trolley base and pole mounted on the car, the pole being kept down when not in use. There are no deep excavations to make. The system can be installed on any road already in operation without tearing up the ties. The cost is only about one-half that of a cable or open conduit road. The insulation of all parts of the line, the switches, and the contact but- tons is such that the possibility of grounds and short circuits is reduced to a minimum. The system is easy to install, simple in operation, and reliable under all conditions of track and climate. Finally, the system is absolutely safe. It is impossible for anyone on the street to receive a shock, as all the contact buttons are " dead " except- ing those directly underneath the car. Requirements. In devising this system the following requirements of successful working were carefully considered. The insulation must be sufficient to prevent any abnormal leakage of current. The means for supplying the current to the car must be infallible. The apparatus must be simple, so that inexperienced men may operate it without difficulty. The system must operate under various climatic conditions. Finally, absolute safety must be assured. WESTHtfGHOUSE §¥STEH. This system includes the following elements. First. Electro-magnetic switches, inclosed in moisture-proof iron cases. Each switch is permanently connected to the positive main or feeder which is laid parallel to the track. Second. Cast-iron contact plates or buttons, two in each group, placed between the rails and electrically connected to the switches. A separate switch is provided for each group of buttons. Third. The conductor forming the positive main or feeder. This is com- pletely inclosed in wrought-iron pipe, and is connected to the various switches. 842 ELECTRIC RAILWAYS. Fourth. Metal contact shoes or bars, suspended from the car trucks ; two bars on each car. Fifth. A small storage battery carried upon the car. The operation of the system is described as follows, and is illustrated by cuts making plain the text. Fig. 199. Diagram of Switch Connections. OAR WIRING CONTROLLER Dj^- STORAGE BATTERY' A r Fig. 200. Diagram of Car Connections. Electro-magnetic switches, X lf X 2 , X 3M inclosed in water-tight casings, are installed at intervals of about 15 feet along the track to be operated. Each switch is provided with two windings, I and H, which are connected by the wires N and M to two cast-iron contact buttons, 1 and 2, which are mounted on suitable insulators and placed between the rails. Each car to be operated on this system is provided with two spring- mounted T steel contact bars, Qj and Q 2 , and a few cells of storage battery in addition to the usual controllers and motors. The contact bars are mounted at the same distance apart as the contact pins, 1 and 2, so that as the cars advance along the track the bars will always be in contact with at least one pair, as the length of the bar exceeds the distance between any two pairs by several feet. Suppose a car is standing on the track over the switch X 2 , the contact bars, Qi and Q 2 , being then in connection with the buttons 1 and 2 respec- tively. The first step is to " pick up " the current, i.e., render the buttons 1 and 2 alive. Switch A is first closed ; this completes the circuit from the storage bat- tery, D, through the wiring, R, contact shoe, Q 1? button No. 1, and shunt coil, H, to the ground. The current passing through H magnetizes the core, S, which in turn attracts the armature, P, closing the switch and es- tablishing connection between the 500-V main feeder K, and button No. 2, through the contacts, JJ, coil I, and wiring N. Switch C is now closed and switch A opened ; the switch X, is kept closed, however, by the current flowing from button No. 2 through bar Q 2 , connection T, resistance L, con- nection R, bar Q,, button No. 1, connection M, coil H to ground. The car now proceeds on its way, current from the main passing through connection T, to the controller and motors. When the car has advanced a short distance the contact bars make connection with the pair of buttons connected to switch X 3 . Current then passes from bar Q, through the shunt coil of this switcli. The operation described above is then repeated. As soon as the bars leave the buttons 1 and 2, current ceases to pass through the coils I and H of switch X 2 , and this switch immediately opens by grav- WESTINGHOUSE SURFACE CONTACT SYSTEM. 843 ity, leaving the buttons connected to it dead and harmless. As connection with the main has already been established through switch X 3 , there will be a continuous flow of current from the feeder, and no flash will occur either at the button or the switch. It will be observed that all the current passing to the car from the main through switch contacts J J passes through the series coil, I, holding the switch firmly closed and precluding all possibility of its opening while cur- rent is passing through the contacts, even should the circuit through coil H be interrupted. Although the act of "picking up the current" requires some time to describe, it takes in practice only a few seconds. Two separate switches, A and C, are shown in the diagram; but in practice one special switch of circular form is provided, and the necessary combina- tions required for " picking up the current " are made by one revolution of the switch handle. The battery need only be employed to lift the first switch; for after thac has been closed, the contact shoes bridge the main voltage over from one set of pins to another, as described, thus closing the successive switches, with- out further attention from the motorman. The battery is charged by leaving switches A and C closed at the same time. The Switch. Fig.201 shows the general arrangement of switch, bell, and pan. The switch and magnet are mounted upon a marble slab, which is secured in the bell by means of screws to the bosses, B B. The switch magnet, M, is of the iron-clad type. It is secured to the upper Fig. 201. Section of Switch, Bell, and Pan- side of the marble base, and is provided with a fine (shunt) winding for the " pick up " current, and a coarse (series) winding through which the work- ing current passes. When magnetized the poles attract an armature attached to a bridge piece, J,each end of which carries a carbon disk, N. R, R, are guides for the bridge piece, J. Directly above each of the carbon disks, N, is a stationary disk, O, mounted upon a marble base. One of the disks, O, is permanently con- nected by means of one of the contact cups, Gj, as explained later, to the positive main cable, and the other, through the series coil and cup, G„ to the positive contact button. 844 ELECTRIC RAILWAYS. The pan, C, is provided with four bosses, S, to support the vertical split pins, F, which are insulated from the pan. These pins slide into recepta- cles, G, on the switch base. The pins, F, are provided with connectors, I, for the purpose of making connection with the several cables, H, which pass through the holes in the under side of the pan. The pan is completely filled with paramne after the connections are made, thus effectually keeping out all moisture. The object of the bell, A, and the pan, C, with the split pins, F, and the cups, G, is to provide a ready means of examination of the switch without disconnecting the wires. The bell can be lifted entirely free of the pan. In replacing it, it is only necessary to see that a lug, T, on the side of the cover, fits into a slide, U, on the frame. When in this position the split pins make connections with their corresponding cups, G. The bell, A, is provided with lugs, L, to facilitate handling ; and also a double lip, W. The inner portion of this lip fits into and over the annular groove, D, of pan C. This groove is filled with a heavy non-vaporizing oil. The outer portion of lip, W, prevents water from entering the groove. The object of the groove, D, and the lip, W, is to make a waterproof joint to pro- tect the switch and cable terminals without the necessity of screw joints or gaskets. The bells are all tested with 25 pounds air pressure ; they may be entirely submerged in several feet of water without affecting the operation of the system. The Contact Buttons are made of cast iron. They are about 4£ inches in diameter, and, when installed on paved streets, project about five-eighths of an inch above the pavement and offer no obstruction to traffic. This is sufficiently high to enable the collector-bars to make contact, and at the same time to entirely clear the pavement. For open-track installations they are substantially mounted in a combination unit as described below. Fig. 202. Section of Combination Unit. The Combination Units. The bell and pan are entirely inclosed in a cast-iron switch-box. This box and the contact buttons are made into a complete unit as shown in Fig. 101. Each unit consists of three separate castings. The cylindrical cast-iron box, which incloses the switch, bell, and pan, is bolted into a recess provided for that purpose in the bottom of the spider-like structure, which is a sep- arate casting, consisting of box rim, receptacles for the button insulators, and supporting arms. The removable lid is the third casting. The insulators, A, Fig.202,are made of a special composition, and are ce- mented into the tapered cups, B, and supported by the iron plates, C. The contact buttons, E, are mounted on top of these insulators and stand, when installed, about one inch above the rail. > - The four arms, G, are secured to the ties by means of the bosses, H, thus reducing to a minimum the labor of leveling the boxes and avoiding the necessity of special ties. WESTINGHOUSE SURFACE CONTACT SYSTEM. 845 Mains and Wiring-. The positive main or feeder is incased in a lj-inch iron pipe, and passes directly through each switch-box, and a tap is made to each switch, the switch-boxes being all connected by the iron pipe, as per cut below. n n n Fig. 203. Track Equipped for Track Return Circuit. No additional wires are used to interconnect the coils or contacts of ad- jacent switches. The Contact Bars are of steel, of ordinary T section. They are sup- ported from the car trucks by two flat steel springs and adjustable links. These bars are inclined at the ends so that they may readily slide over the buttons and over any ordinary obstacle. Insulated Return JLinc*. In case it is considered best not to use the rails as the return line, insu- lated mains for this purpose may be included in the system. It is only necessary to install another row of contact buttons, another collecting bar, Fig. 204. Track Equipped for Insulated Return Circuit. and to use double-pole switches. Fig. 204 illustrates an installation of this kind. For all ordinary work, however, the ground return is satisfactory. Modifications of the System. The description given on the preceding pages applies to the system as in- stalled for yard and similar work. Modifications can be made and detail matters arranged according to the requirements of each case. Street Railway Work. The foregoing description applies to installations where the track is open (unpaved),and where it is unnecessary to make provision for traffic crossing the tracks except at certain points. For street railway work, the switch- boxes are preferable installed outside the track, while the buttons are placed between the rails and mounted on a light metal tie, as shown in Fig. 205. 846 ELECTRIC RAILWAYS. The operation of the system is exactly the same as in open-track work. Connecting wires pass from the buttons under the tie to the switch-boxes. For double-track work the switches are installed between the two tracks, and the boxes may be built to hold two switches, one for each track. UNE_or_PAvmG CHANNEL IRON. Fig. 205. Section of Track Equipped for Street Railway Service. "When, as is sometimes necessary, the buttons are placed in a single row, it is necessary that the "pick-up" current should be of the same voltage as that of the main circuit, and consequently the car-wiring indicated in Fig. 206 is used, instead of that shown in Fig. 200. STORAGE BATTERY Fig. 206. Diagram of Car- Wiring. Referring to Fig.206, the method of "picking up" the current is as fol- lows : Switch A is first closed ; this completes the circuit from a storage battery D, through a small 500-volt motor-generator F, which immediately starts. As soon as it is up to speed, which only requires a few seconds, switch B is closed ; current then passes from F through the wiring R, to contact shoe Q, and then through the switch magnet, as explained on page 538. Switches A and B are then opened, thus stopping the motor-generator, which need only be used to operate the first switch. The successive switches are closed, as described on page 842. This arrangement of a high-voltage "pick-up " may also be used advan- tageously with two rows of buttons where the track is liable to be obstructed by mud or snow. Sectional Rail Construction. For suburban railway or similar service two light rails may be substituted for the two rows of contact buttons, as shown in Fig. 207. The cars are then equipped with contact shoes instead of bars. These rails are insulated from the ground, and may also be insulated from each other wherever desirable, thus breaking them up into sections, which are each controlled by a single switch. The sections may be made of any desired length to suit the conditions. For example, between stations they may be 500 or more feet long, while near stations or crossings, where anyone is liable to come in contact with the rail, the length of a section may be reduced to 50 feet or less. The electrical operation of two-rail installations is the same as when two rows of buttons are used. The sectional switches along the tracks are entirely under the control of the motorman, and the rails may be rendered 11 dead " at any moment should occasion arise. G. E. CO. SURFACE CONTACT SYSTEM. 847 Fig. 207. Sectional Rail Installation. GENERAL ELECTRIC SYSTJEIfl OE SURFACE CONTACT RAILWAY. Following is a description of the surface contact system, as developed b> the General Electric Company, and practical application of it has been made at Monte Carlo, and at the company's works at Schenectady. The description is from a report made by W. B. Potter, Cf . Eng. of the Railway Department, and written by Mr. S. B. Stewart, Jr. In the operation of electric cars, by the closed conduit surface plate con- tact system of the General Electric Company, the current is collected for the motor service by means of two light steel shoes carried under the car, making contact with a series of metal plates, introduced along the track between the rails, automatically and alternately energized or de-energized by means of switches grouped at convenient places along the line ; the method of the switch control being such that in the passage of the car, in either direction, it is impossible for any plate to become alive except when directly under the car body. In ordinary street car practice, the contact plates are spaced approxi- mately ten feet apart, positive and negative plates being staggered, as shown in Eig. 208, which admits of but three plates ever being covered at any one time by the shoes, which are so designed as not to span more than two plates of the same polarity. In grouping the switches it is customary to locate them either in vaults constructed between or near the tracks, or in accessible places along the side of the street, the location and spacing of groups and number of switches in each group being based upon a comparative cost between the style of vault or other receptacle, and the amount of wire with ducts be- tween the contact plates and their corresponding switches. The main generator feeder is carried to each vault or group, and auxiliary feeders from it are distributed to each switch, the track rail being utilized for the return circuit. 848 ELECTRIC RAILWAYS. The operation or performance of this system can be readily traced out by reference to Fig. 208. It will be seen that the current in its passage to the motor from the positive generator conductor passes to contact A of switch No. 2 through the carbons on its magnet armature (which has been lifted by the energized coil G) to contact plates B and C, through the contact shoe D to the controller and motor, coming out at contact shoe E to the contact plate F, when it passes through the coil of the automatic switch G, ener- gizing it and returning by the track-rail H ; thus maintaining contact at switch No. 2 armature carbons as long as the shoes remain on the contact plates C and F. It should now be noted that contact plate B is energized MOTOR <^q /^ SB 3EI Fig. 208. Diagram of Connections for Surface Contact Railway Plate System, General Electric Co. as stated above. As the car proceeds, the shoe D spans the plates B and C» thereby keeping the coil of switch No. 2 energized after shoe has left plate C, and until shoe E comes in contact with plate J, which immediately ener- gizes coil No. 1, thus making the preceding contact plate energized, prepara- tory to the further advance of the car. It will be noted in the above description of the performance of the system, that we have assumed switch No. 2 on Fig.208as closed; it should therefore be understood that an aux- iliary battery circuit is necessary in starting or raising a first switch, pre- paratory to its armature being held in contact position by the generator current, which current energizes the preceding contact plates consecutively as described above. The battery current is brought into the automatic switch circuit momen- tarily during the period of first movement of handle of the controller in starting a car, the transition of the controller cylinder also bringing the generator current in connection with the battery for a short period cf time, thus replenishing the elements sufficiently to operate the switches. The battery is also used to supply current for lighting the car, the generator circuit being disconnected while the car is at rest. Surface Contact Plates. The surface contact plates are made of cast iron, with wearing surfaces well chilled, designed to be leaded into cast-iron seats in such a manner that they are thoroughly secure, but can be readily removed by special tongs for the purpose. The seat is imbedded in a wooden or composition block set into a cast-iron box, the latter being spiked or screwed to the tie. A brass terminal is fastened to the seat for the reception of the connecting wire from the switch. See Fig. 209. Gr. E. CO. SURFACE CONTACT SYSTEM. 849 As stated above, the plates are usually located 10 feet apart for straight line work, but somewhat closer on curves, depending upon the radius of the curve and length of contact shoe. The negative and positive contact plates are staggered with a uuiform angular distance between them, situated not less than 10 inches from the track rails. Fig. 209. Plan and Section of Track, Monte Carlo, Europe. General Electric Company's Surface Contact System, 1898, Surface Contact Switch. The automatic switches are constructed on the solenoid principle, the armature or core of which is employed in closing the contacts as shown in Fig. 210, Automatic Switch for Open Conduit, Burface Plate Contact System. 850 ELECTRIC RAILWAYS. Fig 210. The end of the armature core is provided with a pressed sheet- steel carbon-holder, for the purpose of supporting the carbon contacts which are held in place by bronze clips and cotter pins which can easily be re- moved. The pressed-steel carbon-holder can also be detached with little trouble by removing the end holding it to the core. Copper plates are se- cured to the slate base for contact surfaces and the attachment of feeder- wires. The wire of the solenoid is wound on a copper &pool and placed in a bell-shaped magnet frame, and a pole-piece, slightly recessed to receive the end of the armature core when the switch is in a closed position, is at- tached to the top cover, and extends part way down through the winding. The recess in the armature increases the range of the magnet, making the attraction uniform except at the point of contact where the power increases rapidly, thus securing an excellent contact. A blow-out magnet coil is con- nected in series with the feeder current, and so situated that the influence of its poles is used to rupture any arc that might be formed while the switch is opening ; however, this blow-out magnet is used simply as a precaution- ary device, as under ordinary conditions there is no arcing, the succeeding automatic switch closing the circuit before it is opened by the preceding one. Each vault or group of switches should be provided with cut-outs or an automatic circuit breaker to protect them in the event of short circuits. Surface Contact Shoes. The contact shoes are made of " T " steel of light section, the suspension for which is an iron channel beam extending longitudinally with the truck frame directly under the motors, with a substantial wooden cross-arm at- tached to each end for the shoe-supporting casting, the shoes being attached to these supporting castings by a spring equalizing device for maintaining the shoes at the proper height, and also for making them flexible enough to meet any slight variations in the contact plates and track rails. The shoes when in their correct position should never drop over one-fourth inch below the surface contact plates, and are designed so that they may raise three- fourths of an inch or more above them. See Fig. 211. Fig. 211. Collecting Shoes, Monte Carlo, Europe. General Electric Company's Surface Contact System, 1898. A screw adjustment is provided to lower the shoes as they wear away, or to take care of any other discrepancies due to wear of parts, etc. ; if they are allowed to drop too low they will interfere with rail crossings, causing short circuits. Storage Batteries. It requires for closing the first automatic switch when starting, and for lighting the car approximately, ten storage battery elements capable of 35 amperes rate of discharge for five hours. G. E. CO. SURFACE CONTACT SYSTEM. 851 The batteries are only slightly exhausted in making the initial connec- tions through the automatic switch, as it only takes approximately 15 am- peres momentarily to perform this work, the battery is immediately recharged by current which has passed through the motors. The battery serving as a rheostatic step, this momentary charging does not represent any extra loss of energy. The circuit connections of the battery are accomplished in the controller and require no attention on the part of the motorman. Car Lighting-. The amount of recharging derived from the motor circuits is sufficient to operate the automatic switches, but where lighting of the car is done from the same battery, an additional recharge is required. Assuming that 10 20-volt lamps are used for lighting a car, the batteries will need to be recharged every night about five hours, at an approximate rate of 25 amperes. It is customary to run leads from both the positive and negative terminals of the batteries to charging-sockets attached to the under side of one of the car sills in a convenient place for connection to the charging-wire. A small generator of low potential (30 volts) driven by a motor or other method is required for supplying current for recharging the batteries where the desired low-potential current is not accessible, and the wiring from the charging source should be run to a location in the car-house most convenient for connections to the battery sockets. These locations may be fixed either in the pits or on posts at the nearest point to where the cars will be sta- tioned, and there should be flexible lead wires attached to plugs for connect- ing to the battery circuit on the car. In wiring the car-house for the battery connections, it would be found convenient to designate the polarity of the various wires either by different colored insulation or tags, and the plugs at the ends of the flexible leads should be marked plus and minus to avoid mistakes in making connections with the car battery receptacle. Motors and Controllers. The motor and controller equipment used with the surface plate contact system is standard apparatus as ordinarily employed for electric car service, with the exception that provision is made in the controller for cutting in and out the storage battery while starting the car. Care of Apparatus. As success in the operation of the contact plate system depends largely on the care of the apparatus, a few general remarks on the subject will not be out of place here. Care should be taken that the contact plates are kept clean, and they should be frequently inspected, the roadbed being well drained. Any small quantity of water temporarily standing over the tracks, however, would do little harm, as the leakage through the water would not be sufficient to create a short circuit, although this condition should not be allowed to exist any length of time. The automatic switches should be carefully inspected and all cast-iron parts thoroughly coated with heavy insulating paint, and a test for insula- tion or grounds be made frequently, and all the parts kept clean and free from moisture. The contact shoes, in order to prevent leakage, should have their wooden supports well protected with a coating of an insulating paint, and should also be occasionally cleaned. The storage batteries should be properly boxed and should have the cus tomary care which is necessary to keep them in good working order. DETERIORATION OP UNDERGROUND METALS DUE TO ELECTRO- LYTIC ACTION. Revised by A. A. Knudson, Electrical Engineer. In view of the different phases and effects of electrolytic action herein presented, it seems essential, where a clear insight of the subject is desired, that a reference to the causes which underlie the principles of such action should first be given. To this end the following is abstracted from the Report of the Electrical Bureau of the National Board of Fire Underwriters, Pamphlet No. 5, dated August, 1896, viz : This -deals with early discoveries and represents the gist of opinions given by several authorities on this subject at that time. The balance of this article is treated in a purely practical manner. Recent reports show that the destructive effects of electrical currents on subterranean metal pipes are becoming sufficiently marked in many parts of the country to seriously interfere with the service the pipes are intended to perform. Underground water mains have broken down, because of faults unques- tionably due to electrolytic action; and smaller service pipes have been weakened to such an extent as to break at critical moments, when excess pressure is put upon them at intervals during a fire. Measurements show that conditions unquestionably exist in nearly every district in the United States covered by a trolley road, which are favorable for destructive action on the subterranean metal work in the vicinity, and pipes taken up in many of these districts show unmistakable signs of harmful effects. The general nature of this action, and the causes which bring it about, are too often seen to need elaborate description. Briefly it may be compared to the action which takes place in an electro-plating bath. The current which enters the bath through the nickel or silver metal sus- pended therein, flowing through the bath and out through the object to be plated, ultimately brings about the destruction of the suspended piece of metal. Similarly, _ the current from a grounded trolley system flowing through the earth in its course from the cars back to the generating station selects the path of least resistance,* which is generally for the whole or a part of the way the underground mains, and at points where it leaves the pipes to reach the station the iron of the pipe wastes away until at points the walls become too thin to withstand the pressure of the water, and a breakdown ensues. The difference of potential necessary to bring about this action is very small, — a fraction of a volt, — and consequently in all districts where potential differences are found between water-pipes and the surrounding earth, such actions can be assumed to be taking place, for dampness, and the salts necessary to produce electrolysis, are present in all common soils. Whenever, then, a reading is shown by an ordinary portable voltmeter registering tenths of a volt with the positive binding-post in electrical con- nection with a water-pipe or # hydrant, and the negative binding-post in elec- trical connection with an adjacent lamp-post, car track, or metal rod driven in the earth, electrolytic action will be found upon examination to be tak- ing place at that point which will ultimately result in the destruction of the water-pipe, provided that the resistance of the soil is sufficiently low to conduct current. Referring to the diagram shown in Fig. 1, it is seen that the current will gass from the generator out over the trolley line, through the motor to rail, ack to the power house. There are obviously two paths open for the * The correct statement would be that the current follows the law of divided circuits taking all paths offered, rails, earth, pipes, etc., in inverse proportion to their respective resistances. 862 ELECTROLYTIC ACTION. 853 current. One a return through the rail, the other a return through the earth and any existing gas-pipes, water mains, or other metallic structures that may be in its path in the earth. The current flowing through these two paths in parallel is plainly inversely proportional to the resistance of these two paths. Therefore, in a general way the current will leave the rails at A, flowing into the water-pipe at B, and will again leave the water-pipe at C and enter the rails. Here, then, is an electric current flowing between metallic structures that may be called electrodes at places in the return path from the motor to station. All that remains, then, to promote electrolytic action is the presence of some solution which will act as an elect rolvte Observation has shown that the earth, especially in the larger cities, con- tains a large percentage of metallic salts in solution, which will readily act as electrolytes upon the passage of electric current. It can be seen, then, referring to this diagram, that if there exists in the ground sufficient moist- ure of some metallic salt, electrolytic action will take place between the electrodes A and B, and between the electrodes C and the rails. In the earlier electric roads the positive terminals of the generators were connected to ground. This arrangement of the polarity of . the street railway has a tendency to distribute the points of danger on water-pipes, gas-pipes, cable- sheathing, or any other underground metallic structure throughout a large and extended territory. By reversing the polarity of the railway generator, i * i r imm Fig. 1. bringing the positive terminal to line and negative to ground, the points where the current leaves these metallic structures will be brought much nearer the power station, and will be localized in a much smaller area. From the electric railway standpoint, the prohibitive expense of the requisite addition of copper to make a complete circuit is advanced, to- gether with the impracticability of a double-trolley system that is appar- ently a necessary concomitant of the metallic return; and these arguments have a certain weight. There is no question but that the complete metallic return is in the beginning a more expensive installation, but per contra few railway companies have any idea of the energy now expended in returning the energy delivered by the power station through the poor conductivity of the average railway track with its surrounding earth. Destructive Effects. — In the process of electrolysis upon under- ground pipes there are two distinct phases of action considered as follows: A, the lateral effect which is most common, illustrated by Figs. 2, 3, 4 and B, the joint effect as shown in Fig. 6. A. Where the current is leaving a cast iron main and passing into the soil the iron is usually removed in spots, causing pittings of varied size and depth, and in aggravated cases, furrows and holes. The pittings are small at first, being 1-16 to £ inch in depth and varying in diameter at the surface from I to 1 in.; those more advanced are from £ to 1 in. or more in depth, with corre- spondingly larger surfaces. When a section of cast iron pipe containing such pittings has been removed from the soil and exposed to the sun, the graphitic carbon and impurities, of which the pittings are filled, become dry and hard and drop out or are easily removed. In appearance they are flat, or nearly so, at the surface of the pipe and oval in depth, as in Fig. 2. These are | of the actual size and shape taken from a pipe. In weight they are about the same as dry wood of equal dimensions. Where electrolytic action has been severe and the main has burst, the most of these impurities will have become detached or washed out by the force of 854 ELECTROLYSIS. escaping water, and the spots and holes are plainly revealed. Fig. 3 is an example of severe action and represents a section of a 6-in. cast iron water main taken from a street in Brooklyn, N. Y. The water from this Section i ACTUAL SIZE Section action /*^ Plan Section Plan o Plan Fig. 2. leak escaped into a canal, did not appear on the street, and the leak was only discovered by accident. The length of time the water was running to waste is not known. Fig. 4 is a 6-in. section from Reading, Pa. Fig. 5, also from Reading, Pa., replaced Fig. 4 and failed again in about one year. Fig. 3. Section of 6-inch Cast Iron Water Main Destroyed by •'Electroly- sis," removed from Wallabout Place, East of Washington Avenue, Brooklyn, N. Y., January 21, 1903. Fig. 4. Fig. 5. ELECTROLYTIC ACTION. 855 B. Joint Effect. — This is caused by electric currents flowing through or along the pipes lengthwise, and by reason of resistance at the joints, elec- trolytic action takes place. Resistance is caused partly by the coating of asphalt varnish upon both the inside and outside of the pipe, making a partial insulation; and partly by corrosion due to the continued presence of water upon the inside, and moisture upon the outside. In such case the current shunts the joint, the damage occurring at points where it leaves, causing pittings in the iron close to the lead, softening of the lead, resulting in leaks. Fig. 6 — the spigot end of a cast iron pipe — shows cause of a leak through disintegration of the iron near the lead of the joint ; the furrow of pittings — between chalk-marks — extend half way around the pipe; the left end of the pipe softened three-eighths of an inch deep was cut with a pocket knife. The extent of joint damage depends upon the strength of current flowing in a given time. The action upon wrought iron or steel pipes differs somewhat from that upon cast iron. In the reduction of wrought iron by the process, there is Fig. 6. a seamy, or shredded appearance, with but little residual carbon. Upon steel such as the base of steel rails, or rail chairs (the latter now little used), the effect is a melting away of the metal, leaving sharp edges at their bottom portions. This effect is found where rails are positive to pipes. The action upon lead service pipes, or lead covering upon cables, is some- what similar to that upon cast iron so far as pittings and furrows are con- cerned, but instead of the graphitic residue there is left in the pittings and the surrounding soil a whitish matter consisting of the oxide or residue of lead. Increase of Current Flow upon Ufa in* due to Bonding- Same to Rails or to Negative Conductors. Measurements in different cities under varying conditions sbow the in- creased flow of current through mains after bonding the mains to the rails, from four to ten times above the normal at points near the bonds, in some cases very much higher. In one case where 5 amperes maximum was found flowing through a 6" main a temporary connection with ammeter and leads was made between main and P. H negative with result of over 150 amperes. The flow in excess of normal is generally less as the distance is increased from a bond. 856 ELECTROLYSIS. The following tables represent actual measurements made in different cities. Measurements made near the bonds, except in No. 3, Table 1. Table I. Flow in Amperes. No. of Notes. Test. Normal. Connected. 1 21.0 41.7 2 21.0 60.2 3 bonds. 3 30.5 4.3 3000 ft. from bond. 4 5.0 128.0 In negative district 5 mile? from P.H. 5 6.0 32.0 Geneva, Switzerland. 6 11.5 37.5 7 80.0 125.0 8 27.7 45.1 9 9.8 30.5 10 6.6 10.5 Table II. Three Cases Difference of Potential in Average Volts. No. 1 No. 2 No. 3 Connected. 0.25 0.3 2.5 In one city examined by the writer two water mains in front of a power house were connected by copper cables directly to the negative bus bar of the switchboard. The estimated amount of current flowing by this path was found at times to be oyer 1000 amperes; a very much smaller flow has been known to damage the joints of mains. Current Movements upon Underground Tlaim. — The flow of current upon underground mains is proportional to the traffic upon the car lines. When railway traffic is heavy mornings and eveningi more current output is required at the power house than during hours oi light loads. Such changes are faithfully reflected by current flowing in the mains. This is illustrated in curve sheet, Fig. 7, where the load line of a 24-hour log of a power house is shown, and directly above it is placed the line of current strength flowing through a 36-inch water main. It will be noticed that the rise and fall of current strength upon the water main takes Elace at the same hours of the twenty-four as the load changes at the power ouse. This effect is more or less common in all cities where electric railways with the usual ground return prevail. Many instances of railway currents flowing through and across waterways have been discovered, where, as is often the case, the power house is located upon the banks. mr? ne mstance °f such action was discovered at Bayonne, N.J., November, 1904. At that time current was supplied from the power house in Jersey krtVf five miles from the central part of Bayonne. The city is nearly sur- rounded by salt water. Mains in streets near the shore and in salt marsh ELECTROLYTIC ACTION. 857 COMPARISON CURVES SHOWING CURRENT VARIATIONS ON 36"WATER MAIN 24 HRS. AND ALSO POWER STATION LOAD 24 HRS. ENDING I 2 MDT. 1?3456 7 8 9 10 11 12 1 2 3 4 5 6' 7 8 9 10 11 12 130- N DON o f) A V1DT. 120 - 1 o CO -J UJ u ft. J) 110 " Ul 1 ii 1 0- 1/ / o - UJ O f\ \ Hi A < 90- I 1/1 ' J / \ ' / \ i / i Z 80-- < / l 1\ A A t :v L / i' 1 A 70 - DC Ul T i 1 / k I / ] 20 TO 00 Vl I I ' < 1 1 o I A V z 40 ^ Ul \ cc UJ h 4 ^ h 90Ou V D o 20 ._ £ ' \ i 80 r \ L OAC L I TOOu \ - 4 6000" I \ \ fa 30 20 J I DO \ / > N( DOt^ 00 VWDTj 12 3^56739 10 11 12 12345 6 7 8 9 10 1 1 12 Fig. 7. 858 ELECTROLYSIS. have been destroyed by the returning railway currents delivering at such grounds, causing a heavy loss in piping property to the city by electrolysis. There was no point in the city where mains were positive to the rails; the flow was from rails to mains, thence to shore and to power house. A similar case was discovered by the writer in 1906 during a survey in the city of Toronto, Canada, where mains adjacent to the shore of Lake Ontario, 2 to 4 miles distance from the power house, were badly damaged. The conditions in Bayonne have been changed by the placing of a sub power station in that city. Such returning currents usually enter the power house through pipes used for condensing. Cases have been found where much damage has been caused to apparatus in the steam plant. Other current movements may be cited where metal bridges cross a river as in map, Fig. 8, as was discovered in the city of New York. The power house is located near the Navy Y'ard, in Brooklyn. A portion of the returning currents, as shown by arrows, flows over the New Y r ork and Brooklyn Bridge to Manhattan, thence north to the new Williamsburg Bridge by way of underground mains, subway structures and other metals, and passes over that bridge back to Brooklyn, thence through mains, to Fig. 8. rails and negatives, to power house. In this case damage may be expected at three points, viz., where currents leave bridge metals on the Manhattan side, where they leave pipes to enter Williamsburg bridge, where they leave same bridge fof pipes on Brooklyn side. When the two bridge structures are connected in Manhattan as proposed, then there will be further changes in this direction of current. Before the new bridge was built, these currents recrossed through the river bed, leaving mains all along the docks on the Manhattan side, for the river, and leaving the river for mains or other metals along the docks of the Brooklyn side. Traces of these currents have been found as far north as 23d St., a distance of over two miles from the Brooklyn Bridge. Since the Williamsburg Bridge has been built, nearly all traces of these currents flowing north of it have disappeared, showing that the mass of metal composing the structure acts as a "short circuit " or path of lower resistance and now carries practically all of the returning currents flowing from Manhattan back to Brooklyn. Electrolytic Effect* upon "Water TOeters. — This is a compara- tively recent discovery, and is due to the location in which manv meters are placed. Those found damaged by electrolysis in one citv examined have in every case been taken from pits in the cellar bottoms of dwellings, stores, stables, and near water fronts, where tide water had access. The meter pits in many cases are constructed of boards at the bottom and sides, with a loose fitting wooden cover; this pit, being the lowest point in a cellar, acts as a catch basin and collects the drainage when water is present, partially or wholly submerging the meter very often in stagnant; water. ELECTROLYTIC ACTION. 859 The quality of such liquid makes a convenient electrolytic for any current of electricity. Railway or other current passing to the meter through the service pipes, and out of the meter into this liquid, in time causes a rupture of the thin iron shell of the small sizes where the top is iron. The actual weight of iron lost through electrolysis by a 4-inch meter located in a ferry house and subject to tide water was in about six years 15 pounds. This meter was near a power house where the p. d. at times reached 25 volts, with mains positive to rails. These severe electrical conditions have since been modified by the railway company improving their track return. Meters constructed of bronze have had holes eaten through their base where resting on damp soil in cellars. Such grounds often attract trolley current through the service pipes.* Danger from Tire or JE x.plosions. — Currents entering buildings which contain explosives, through water or gas mains, are dangerous owing to sparks when gas mains are separated or the cross-connecting and discon- necting of pipes containing current, by movable metals is made. The usual course of such currents is to enter a building on one pipe and pass out upon another when a cross-connection is made between the two systems anywhere inside of a building. When the connection is broken the spark appears, and it may appear at any point in the building, possibly in the presence of explosives. Bonding the pipes together where they enter the building has proved effective as a temporary remedy in some cases. As no two cases are alike, no particular rule can be laid down as a remedy. Where the conditions are considered dangerous the services of a specialist should be engaged. Electrolysis in Steel Frame JSuilding-s. — While no instance of serious damage to a steel structure through the disintegration of supports caused by electrolytic action can be cited, still this question is now receiving attention by architects and others, and methods for safeguarding against such corrosive effects are being applied. One such instance of protection is the new New York Times building. In one of their publications the fol- lowing is stated in reference to this structure: "The danger that in case of the steel frame rusting the disintegration of electrolysis would hasten the process of dissolution so much as to make structures of this kind prematurely unsafe through the destruction of their supports, was recognized in time to permit of ample safeguarding in the case of the steel frame of the Times Building. " It is axiomatic that columns to which moisture has no access will not be impaired by rusting, and that those effectually insulated from vagrant electrical currents will not be affected by electrolysis. The first considera- tion was to keep the basements dry; hence the thorough waterproofing and draining of the retaining walls already described, which was also carried under the floor of the pressroom, occupying the great area of the sub- basement. As a further safeguard, all the steel members up to the street level are incased in Portland cement mortar to the minimum thickness of three-fourths of an inch. This is effectual protection against rust deteri- oration. Under these conditions electrolytic disintegration is deemed impossible, but the probabilitv of its occurrence in even microscopic degree is rendered still further remote by as perfect insulation as can be provided. There is sufficient grounding to relieve any electrical tension which may exist in anv part of the steel frame by drawing off the current at points where electrolvtic action cannot be set up. This also makes i it lightning- proof to the extent to which it is possible to impart that quality to a Duiid- ing. " For results of experiments by the writer upon metals in concrete, see February, 1907, Proceedings of the A.I.E.E. in a paper. entitled Electro- lytic Corrosion on Iron or Steel in Concrete." discussion in April numoer. Current Swapping-.— The transfer of currents between the tracks ot different companies through underground routes, often by way of mains, is of frequent occurrence, particularly if the lines parallel even for a short distance. Th's is more noticeable at the terminus of suburban lines, but also pre- vails in cities. * Case illustrated in abstract of the writer's report for Providence, R.I. in Water and Gas Review, N.Y., March, 1907. 860 ELECTROLYSIS. One case in a city where the termini of two different lines were but a few feet apart, showed upon measurement a heavy delivery at times, leaving tracks of one company for tracks of another, soil conditions continually wet, consequently a large percentage was flowing through soil and the water mains. Another case near suburban terminals of two railway lines about 600 feet of 6-inch water main with a number of service pipes were practically destroyed by electrolysis; the main acted as an inter- mediate conductor; the pipes were destroyed under the tracks of one road by the currents from the other. An attempt to remedy was made by bonding the two tracks together. This method cut the potential difference be- tween mains and rails from 6.7 volts down to about 2 volts. After six months' standing no further breaks in the mains have occurred. This plan was considered of value in affording temporary relief, but is not now of importance as the tracks of the two lines have been joined by new tracks in a cross street. Current swapping is more frequent than generally supposed, and is caused largely by local conditions, such as swamps, rivers or other waterways to which a company's tracks connect and are grounded, offering paths which attract their own as well as foreign currents. In the case cited of damaged mains, the flow was from newly constructed tracks, seeking grounds on another road where rails were in wet soil. Usually, however, the cause is due to opposite reasons, viz., currents seeking a track return of lower resistance. A well^constructed road bed on suburban lines will often avoid such opportunity for grounds, and current swapping. Alternating-Current Electrolysis. The possibility of damage to underground structures by alternating currents has been investigated by several authorities both in this and foreign countries. As no actual damage has yet been discovered so far as known to the writer, these investigations are necessarily confined to labor- atory experiments. The following abstracts from a few papers give a fair idea of what is known of the subject, and where further information may be obtained. The Ultimate Solution of the Electrolysis Problem by S. P. Grace, paper before the Pittsburg, Pa., Branch A.I.E.E., read December 12, 1905: " Our many hundreds of laboratory tests have shown us that the electrol- ysis to be expected from alternating currents is by no means negligible, and that while it is far less than that encountered with direct currents, in practice we should anticipate that it is only a question of time until its action would destroy many millions of dollars of underground metallic structures." From transactions of the Farady Society, Volume I, February, 1906, Part 4. Alternating -Current Electrolysis as shown by Oscillograph Records, by W. R. Cooper, M.A.B. Sc, read October 31, 1905: Photographic reproductions of oscillograph records are given illustrating results of his investigations. The author also gives results of several other investigators of this subject. From transactions of the Farady Society, Volume I, August, 1905, Part 3. Alternate Current Electrolysis by Prof. Ernest Wilson, paper read July 3, 1905: The author gives results upon different metals at different frequencies and in different solutions, and begins by saying, "It is well known that if an alternate current be passed between metal electrodes in an electrolyte, electrolysis may take place." The Electrolysis Problem from the Cable Manufacturers' Standpoint, by H. W. Fisher, paper before A. I. E. E., Pittsburg, Pa. Branch, read December 12, 1905: 44 My experiments have not been very comprehensive, but I have found under certain conditions, destructive electrolytic action may occur with alternating currents operating at a frequency of 60 cycles per second. The solution I employed for the electrolyte was water containing common ELECTROLYTIC ACTION. 861 salt and salammoniac, all of which may occur in and around duct systems. I found that with a current density of 0.1 ampere per sq. in. of lead, there was no electrolytic action. Amperes per sq. in. of Surface. Lead Destroyed per Ampere, per hour, per sq. in. 3.04 11.8 17.9 .004 Grammes. .136 .237 * " with a frequency of 25 cycles per second, the alternating current action would probably be greater than shown by my tests." This latter statement agrees with Prof. Wilson's tests above referred to, where he says, " It will be seen from the table that the total diminution in weight, which was equally distributed between the two plates, in a given cell is nearly twice as great at low frequency as it is at high frequency." Remedies. — Several methods have been suggested for counteracting the evil effects of electrolysis. The insulated metallic circuit. The underground, known as the "slotted conduit," has been in success- ful practical use in the borough of Manhattan, city of New York, some ten years, and for a still longer time in the city of Washington, D. C. The double overhead trolley has been in successful practical use in the suburbs of the city of Washington for some years, and in the city of Cincinnati, Ohio, since 1889, and more recently has been established in the city of Havana, Cuba. Both outgoing and return conductors of either construction are insu- lated; where there is no connection to the rails or ground the currents which propel the cars are confined to their respective conductors, consequently no damage to underground metals is possible. Improved Track Return. Next to the double trolley, this method is probably the best, although a modification of the trouble. In some cities a large amount of copper for returns has been placed for this purpose, as well as heavy double bonding at the rail joints. The expense involved in providing copper returns sufficient to give a fair degree of pro- tection to mains, would in most cases be considered unnecessary by the railway companies, unless compelled by law. Bonding Mains to the Track Circuit. This has been done in some cities for the purpose of protecting a positive area where electrolysis was found to be acute; usually this is near a power house. Some effects of such bonding have been mentioned. While this may protect from injury the immediate area where such con- nections are made, it is likely to aggravate joint corrosion by the increased flow which has been pointed out. Meters. — A remedy for exterior electrolysis upon meters is to place them in iron or other receptacles under a sidewalk where they will be free from liquids or damp soil. Such methods are used in the cities of Cleveland, Ohio; Richmond, Va.; and Louisville, Ky. Official reports show in such case they are in no danger from electrolysis, or from freezing, and are easily accessible for reading, and removing when desired. Insulating* Joints in Mains. — This is a further attempt at remedy, and much attention has been given to this phase of the subject by rail- way companies in Boston, Mass., with the Metropolitan Water Works cooperating. The Metropolitan Official Report dated January, 1905, contains much information on this and other attempts to stop the current action which * In this case a large hole was eaten through the lead, and the surface exposed to electrolytic action was nearly a square inch. 862 ELECTROLYSIS. was causing great damage to their mains. Several insulated joints have been set, and are found to be fairly efficient in arresting the flow of current through a main. Usually, however, it is at the expense of diverting flow into other mains. In one case an experiment was tried of two joints in a 48-inch main, one insulated with wood and the other with rubber. A measurement made when the writer was present showed the one with wood insulatioD than that of the rubber after six months' use. The following sketch will illustrate the tests. *kEi WOOD /INSULATION No. I RUBBER INSULATION Fig. 9. Ammeter test between A and C gave 60 to 110 amperes, representing the flow if there were no joints. Between A and B, flow passing through No. 2 (rubber) 0.6 to 1.0 ampere. Between B and C, flow passing through No. 1 (wood) 0.1 ampere. This reading should not be taken as the true value for all cases owing to varying conditions. The efficiency of either one for stopping current was in this particular case very good. Fig. 10 represents a pair of insulated joints ready to place in a 6-inch main. They are made up of wood slats driven in the hubs; a flange of wood rests at Fig. 10. the bottom of the hub. The three screw posts are for wires which are led to the surface for testing efficiency of each joint. Fig. 11 shows the same joints connected in the main at the bottom of the pit, and wirer run to ammeter. Before the pit was filled in, wires were run through small pipe to the surface of the street, the ends being secured by cap, for future testing. A test with low reading ammeter failed to show any sign of current pass- ing through either joint, when first set. After two years one joint shows leak of 0.1 ampere; the other perfect, short circuit around both joints shows 5 amperes. A water pressure of 110 lbs. to the square inch was put on this main, and neither joint leaked. Two joints were used in case one failed, and to pro- vide opportunity for testing efficiency of either one. Experience in Boston is, joints of wood are preferable to those of rubber, on the ground of expense, and equally efficient for stopping current flow. Surface Insulation. — Wrapping a 48-inch main with burlap saturated with asphalt cement applied hot, is another attempt to stop electrolytic action near a power house. After two years' trial, results show, after careful examination, this method to be unsuccessful, and it has been abandoned. This class of insulation has long been known by electricians to be no -protection to metals where subject to continual moisture. ELECTROLYTIC ACTION. 863 Fio. 11. Summary. 1. The tendency of return currents on long lines five to ten miles from a power house is to leave the tracks near a terminus and seek "grounds." This may be by way of other tracks, by way of underground mains, or by water routes. Recent tests show that a very good return construction will not wholly prevent such diversion of currents. 2. Low spots in a company's road bed, where rails are in contact with wet soil, offer an attractive outlet for their own, or foreign, currents. 3. Bonding rails to mains always invites heavier flow of currents to the mains, with corresponding increase of damage at joints. 4. All establishments manufacturing or carrying explosives should be often examined, particularly if contiguous to electric railways, and if metal pipes of any kind pass to them the passing of straying currents into and through such establishments is quite possible and oftentimes dangerous. 5. Protection of metal foundations of important structures, such as tall office buildings, bridges, etc., from electrolytic action should be well con- sidered before their construction and occasionally tested after construction. 6. Current Swapping 1 . — The cause for current swapping between railway tracks should be sought out and removed where possible, especially in cities or towns where underground mains are likely to be included as conductors to their detriment. In one case bonding of tracks of two companies together afforded relief. 7. Insulated .Joint* in water mains have proven effective to stop cur- rent flow in some cases, but often at the expense of diverting it to other mains. 8. No complete cure for electrolysis has been discovered where the grounded return is in use. TRANSMISSION OP POWER. Revised by F. A. C. Perrine. The term " Transmission of Power," as used by electrical engineers, has come to have a conventional meaning which differentiates it from what must be considered its full meaning. Any transmission of electric current, for whatever practical purpose, whether for lighting, heating, traction, or power-driving, must of course be a transmission of power ; but the conven- tional meaning of the term as now used by electrical engineers and others eliminates many of these objects, and is held to mean simply the trans- mission of electric current from a more or less distant point or station to a center from which the power is distributed, or to power motors at different points in a factory or other installation. While the distances over which electric current is transmitted for arc lighting in some large cities and in many small places far exceed the length of line of the ordinary or average power transmission, yet the former is never alluded to as transmission of power. The same condition obtains with traction, the transmission of cur- rent covering miles of territory, and yet it is only alluded to as power transmission when the current is transmitted from a central point to vari- ous sub-stations from which it is distributed. Many engineering features of transmission of power will be found treated under the separate heads in their respective chapters, and the following is a short resume of the subject matter. Building-. Structural conditions and material. Iflotive Power. Water power : Turbines, etc. Steam power : Boilers and appliances. Engines and appliances. Shafting and pulleys. Belting and rope drive. Generators. Dynamos : Direct current. Alternating current. Double current. Transmitting* Appliances. Switchboards. Transformers, step up. liotaries. Cables and pole lines. Conduits, etc. Distributing* Appliances. Sub-stations and terminal houses. Transformers, step down. Switchboards, high tension and secondary. Rotary converters. Direct current motors. Synchronous motors. Induction motors. Frequency changers. Distributing circuits. 864 DISTRIBUTING APPLIANCES. 865 Much has been written regarding the relative values of the different methods of transmitting power, and comparison is often made between the following types, i.e., a. Wire rope transmission. b. Hydraulic transmission, high pressure. c. Hydraulic transmission, low pressure. d. Compressed air transmission. e. Steam distribution for power. /. Gas transmission. g. Electrical transmission. All of the first six methods listed have so many limitations as to distance, efficiency, adaptability, elasticity, etc., that electricity is fast becoming the standard method. The matter of efficiency alone at long distances is one of the best arguments in its favor, and we take from Prof. Unwin's book, " Development and Transmission of Power," the following table of the effi- ciencies such as have been found in practice. i System. Per Cent Efficiency at Full Load Half Load Wire rope 96.7* 55 50 51 75 73 93.4 * Hydraulic high pressure Hydraulic low pressure Pneumatic Pneumatic reheated virtual efficiency . . Electric 45 50 44 64 65 For short distances out of doors, transmission by wire rope is much used both in the United States and Europe, and where but few spans are neces- sary, say less than four, it is obvious that the efficiency is very high. Hydraulic transmission is in considerable use in England, but except for elevator (lift) service is in little use in the United States. Pneumatic transmission is in wide use in Paris, but not so for general distribution in the United States, although for shop transmissions for use on small cranes and special tools is making good progress, the principal usage being for the operation of mining drills, hoists and pumps. Electrical transmission is so elastic and so adaptable to varied uses, and has been pushed forward by so good talent, a not small factor, that its pro- gress and growth have been simply phenomenal. In one place alone, that of traveling cranes for machine shops, it has revolutionized the handling of material, and has cheapened the product by enabling more work to be done by the same help. Indeed the great increase in size of units which is such a distinguishing characteristic of modern engineering has been ren- dered possible by the capacity of the electric traveling crane for lifting great weights. Electric Power Transmission may be divided into two classes, i.e., long distance, for which high tension alternating current is exclusively used ; and local or short distance transmission, for which either direct current or polyphase alternating current are both adapted, with the use of the former largely predominating owing perhaps to two factors: a, the much earlier development of direct current machinery, and b, to the fact that a large number of manufacturers are engaged in the building of direct current machinery. Both types of current have their special advantages, and engineering opinion is, and will probably remain, divided as to which has the greater value. * Per span. 866 TRANSMISSION OF POWER. Long distance transmission is now accomplished by both three-phase three-wire, and by the two-phase four-wire systems, with the former pre- dominating for the greatest distances, owing to economy of copper. Every case of electric transmission presents its own problem, and needs thorough engineering study to decide what system is best adapted for the particular case. Limitations of Voltage. — While 10,000 volts pressure was used with some distrust for a time previous to 1898, since that time voltages up to 70,000 volts have been and are still in use with substantial satisfaction, and plants using voltages of 80,000 and 100,000 are under construction. Properly designed glass or porcelain insulators, made of the proper material and tested under high pressure conditions, cause little trouble from puncture or leakage. The latter is its own cure, for the reason that the leakage of current over the surface of the insulator dries up the mois- ture. Dry air, snow, and rain-water are fairly good insulators, and offer no difficulties for the ordinary high voltages. Dirt, carbon from locomotive smoke, dust from the earth, and such foreign material that may be lodged on the insulators, are sure to cause trouble. In the West and some sections of the East many insulators are broken by bullets fired by the omnipresent marksman. At the lower voltages glass makes a satisfactory insulator, as the eye can make all necessary tests ; but it is so fragile that porcelain is more com- monly used. It is not safe to accept a single porcelain insulator without a test with a pressure at least twice as great as that to be used. Mr. Ralph D. Mershon of the Westinghouse Electric & Manufacturing Company made a long series of tests at Telluride, Col., on the high-pressure lines in use there. With a No. 6 B. & S. copper wire he found that at 50,000 volts there will be a brush discharge or leakage from one wire to the next that can be seen at night, and makes a hissing noise that can be heard a hundred feet or more. This brush discharge begins to show at about 20,000 volts, on dark nights, and increases very rapidly, as does also the power loss at 50,000 volts and higher. This loss depends upon the distance apart of the conductors and their size. For these reasons, wires should be kept well apart and be of as large size as other properties will allow. The wave form of E.M.F. used also influences the brush discharge, being the least in effect for sine wave curves of E.M.F., and being much increased by the use of the sharp, high forms of curve. In regard to the frequency to be adopted for power transmission, one has to be governed by the case in hand, and the commercial frequencies avail- able at economical cost. SPECIAL FEATURES OF DESIGN »XJE TO TRAIS- IKISSION LOE REi|UIRE9[Ei\I§. While the general requirements for the design of a power plant and line for long distance power transmission are practically similar and theoreti- cally identical with those for other electrical installations, at the same time special features are important. These are due to the character of service required, the size of the plants, high voltage, and location of the plants. The general features of design have already been considered in this book, and a short resume is given on page 864. Below, attention is called to special requirements to be considered in power transmission instal- lations. Building^. — Transmission generation stations are commonly located in relatively inaccessible locations, and the size of unit is therefore limited, whereas the total capacity of the station may be great and the current is transmitted at high potential. Transportation and labor conditions must be carefully studied, as the neglect of this precaution may readily involve an underestimate of no less than 25%, and has often so resulted in estimates otherwise correct. This is especially true as regards the use of patented or special building con- struction, which might result in savings where competent workmen are to be had, but which actually result in excessive cost where the amount of work to be done is not sufficient to import men familiar with the type of construction. SPECIAL FEATURES OF DESIGN. 867 Roofing*. — The buildings should be entirely fireproof, and whereas this Is easily taken care of by avoiding wood altogether in the interior construc- tion, supports and walls of the building, a mistake is often made in choos- ing a roofing which must be laid upon planks. Such construction has frequently resulted in disastrous fires at power plants otherwise inde- structible. Heating-. — Where temperatures do not fall to less than 10° F. the waste of energy from the machines is commonly sufficient for heating ; where lower temperatures are encountered, special provisions must be made for heating. Boilers for steam- or water-heating fired in cellars accessible from the outside of the building only are the best. Outlet* for Higrh-Tension Wires. — In buildings where the tem- perature falls below freezing, sewer pipes with large openings for high- tension wire outlets should not be used on account of the excessive draft through these openings. A number of systems for high-tension wire out- lets are described in Transactions of American Institute of Electrical Engineers, Vol. 22, p. 313 ; Vol. 23, p. 578 ; Vol. 25, p. 865. Special methods for carrying out some of these plans have been designed and are described in the catalogues of the porcelain insulator manufacturers. Eig-htning* Arrester Protection. — Arresters should be considered as belonging to the line and not to power house, and lightning] arresters should not be installed in the power house itself, but in a separate neigh- boring enclosure especially erected. Arresters are to be considered as a means for preventing line disturbances entering the power house in any manner. Separating* Generator and Transformer Rooms. — The only reason for attempting to separate generator and transformer rooms is on account of the oil contained in the transformers which may become the source of fire hazard. If, however, the oil transformer is properly enclosed, separate buildings are unnecessary. See Transactions of American Insti- tute of Electrical Engineers, Vol. 23, p. 171. Auxiliary Building's. — No estimate on an isolated transmission power house is complete which does not include houses for the married employees, a central mess house with reading room, assembly room and offices, and stables for the accommodation of horses. Unless these features are properly taken care of, it will be difficult to retain satisfactory em- ployees and to operate the plant economically and continuously. MOTIVE POWER. Water Power, — Load factor and total capacity are closely related in questions of design and revenue. The effect of yearly load factor on revenue is shown by the curves below. By reducing all yearly load rates to a K.W.H. basis we are enabled, through the use of these curves, to determine the total revenue to be derived when we know the total yearly K.W.H. that any variable water supply may sell when applied to the operation of any set of variable loads, and hence the value to the plant of an annual storage. In variable loads there is a variation in the daily load factor as well as in the annual load factor. STORAGE RESERVOIRS. Apart from Plant. — These reservoirs serve to aid in properly sup- plying variable annual load factor, but on account of plant distance, cannot take care of daily variation in load factor. Adjacent to Plant. — When a daily variation of load factor is to be met, revenue may be increased by reservoirs near the plant that may be called upon for conserving water flowing at low power periods and deliver- ing it at peaks, which cannot be done by distant storage. Auxiliary Power. — The value of any plant should be based, not upon the total maximum or minimum capacity, but upon the K.W.H. sala- ble, and in obtaining the maximum K.W.H. capacity it is often possible to increase this by auxiliary machinery to be used at the low water periods or 868 TRANSMISSION OF POWER. at periods of customer's peak. Neglecting the study of this factor often results in estimates of plant value unnecessarily low. Auxiliary power may be obtained from steam, water, or gas, as is obtain- able at the most satisfactory cost, not necessarily the lowest price. The most satisfactory cost is that which yields the greatest annual K.W.H. output from the total plant at the lowest cost. / / 1 / / / 1 j r 0- J ' / f 1 i / &/ / > 1 ! i y/ / ' / / 1 Vi / A t / / 1.90 V -W / / / // & z / ' 1 1 1 1.80 Hy / 1 ! n N ;&> z I t-70 ■? y\ / 1 / I h 1.60 F fef 1 ',& 1 / r \?/ 1 1 f j J. 50 u^> J.4© / % [nibV 1 f 4 / 1 3 / $ Y f 1 4 °y 1.20 T Ov V «s l.lO %/ a ,0/ J TY 2) which, when immersed in dilute sulphuric acid, form a voltaic couple. Its action differs in no wise from that of the ordinary primary battery, except that when it has given out all the energy that the chemicals present enable it to supply, instead of having to put in new chemicals, the cell can be regenerated or brought back to its original condition by passing current into it in a direction opposite to that in which the flow took place on dis- charge. Obviously, there are many combinations which can be used as storage batteries, but with the exception of the lead-sulphuric acid battery, none has proven commercially practical, unless it be possibly the Edison battery, which has lately appeared. This battery has for one of its elec- trodes, nickel oxide, and for the other, finely divided iron or iron sponge, these being immersed in a solution of sodium hydrate. Up to the present, however, these cells have not been used for power work, and therefore the discussion will be confined to the lead battery. The plate on which the lead peroxide is carried is termed the positive plate, and the lead sponge plate is termed the negative, the reason being that on discharge, current flows from the lead peroxide plate and returns to the battery via the lead sponge plate. The condition, however, is the opposite of this inside the cell, as the current flows from the lead sponge plate to the lead peroxide plate. Therefore, considered as a voltaic couple, the lead sponge plate is the positive; considered as a source of electric current, however, the lead peroxide plate is the positive, since it is from this electrode that^the current flows out. r JTneo2*ies. — The first and oldest theory is that on discharge hydrogen, which is released at the lead peroxide plate (Pb02), combines with some of the oxygen in the peroxide, forming water, and reducing the oxidization of the PbC>2 by one molecule of oxygen, bringing it to a state of lead oxide, or PbO. At the sponge lead plate, oxygen is released (these released gases coming, of course, trom the electrolytic decomposition of the water in the electrolyte), and this oxygen (O) combines with the sponge lead (Pb), and oxidizes it, causing it also to become lead oxide (PbO). Thus the two plates tend to approach the same chemical composition. If lead oxide (PbO) be immersed in sulphuric acid, it will be chemically attacked, inde- pendently of any current flow, and change into lead sulphate, the chemical reaction being PbO + HzSO* - PbS0 4 + H 2 0. Thus the active material on both the plates tends to approach the condi- tion of lead sulphate. On charge, the reverse condition takes place, the hydrogen being released at the negative plate and the oxygen being released at the positive, the hydrogen reducing the oxide in the negative plate and carrying it back to its original condition of sponge lead, and the oxygen at the positive increas- ing the oxidization of the positive plate and returning it to its condition of lead peroxide, PbOg. The later theory is that the plates do not pass through the intermediate stage of being changed to lead oxide, but, on discharge, change directly from their respective states to that of lead sulphate. This theory is doubt- less the correct one, for the reason that in the chemical change from lead oxide to lead sulphate, heat is released, which represents lost energy, and if this energy loss should take place it would be impossible to get from the storage battery a large proportion of the amount of energy which might have been put into it on charge. 872 THEORY AND GENERAL CHARACTERISTICS. 873 The foregoing is set forth by the following reversible equation, which shows the action that takes place: charge (1) Pb02+H 2 S04 = PbS04 + H 2 + (2) Pb + H2SQ 4 = PbS0 4 +H 2 (3) = (1) + (2) = Pb0 2 +Pb + 2H2S04 = 2PbS04-f-2H 2 0. discharge The first equation shows the reactions which take place at the positive plate; the second shows those which occur at the negative; and the sum of these two, the third, is the combined effect and is the fundamental equa- tion of the storage battery. Reading from left to right the reactions are those which take place on discharge, while read from right to left the reactions are those which take place on charge. Chi«ng*e in Electrolyte. — The reversible equation of the storage battery shows that some of the SO 3 in the sulphuric acid (which may be looked on as being made up of H 2 4- S0 3 ) goes into chemical combina- tion with the plates on discharge, and a definite amount of SO3 is abstracted from the electrolyte from each ampere hour of discharge, and therefore the concentration of the electrolyte decreases and is lower at the end of dis- charge than at the beginning. The amount of S0 3 abstracted per 100 ampere hours is 298 grams, and therefore, with a given quantity of electro- lyte and acid density, the final density at the end of discharge after a cer- tain number of ampere hours has been taken out, can be computed. The formula for computing the quantity of electrolyte required, when the initial and terminal densities are given is „ _ 1290 - 10.53 d X = number of ounces avoirdupois of electrolyte per 100 ampere hours of discharge. D = percentage of H2SO4 in the electrolyte at the beginning of discharge. d = percentage of H 2 S04 in the electrolyte at the end of discharge. For discharge other than 100 ampere hours, multiply the computed value of X by the actual discharge and divide by 100. ^ 1290 + d(X- 10.53) Also D = ( and d = X 1290 - XD 10.53 - X ' Sulphate. — Lead sulphate, which is a white substance, has no con- ductivity whatever, and if too much sulphate be allowed to form on discharge, it is difficult to bring the battery plates back to their original condition because the regenerating current cannot be made to flow through the sulphated masses. If the plates are only partially sulphated, the high conductivity of the active material with which the sulphate is mixed will afford a path for the current which can easily reduce the sulphate back to sponge lead or lead peroxide. This is one of the reasons why discharge should never go beyond the point where the voltage per cell is 1.8 with normal outflowing current. Change in Volume. — Another reason for avoiding overdischarge lies in the increase in volume of the active material when converted into lead sulphate. If too much of the active material be converted into lead sulphate, the increase in volume sets up strains in the plates, tending to buckle them, and causes the active material to crack or shed and fall away from the sup- porting grid, thus reducing the amount of available active material, the capacity of the plates, and shortening their life. 874 STORAGE BATTERIES. Voltage. — The voltage of lead peroxide against sponge lead in dilute sulphuric acid is about 2 volts, varying with the concentration of the acid. The actual voltage for any concentration may be computed by Streintz's formula: E = 1.850 4- 0.917 (S-s), in which E = E.M.F. of cell. S = Specific gravity of the electrolyte. s = Specific gravity of water at the temperature of observation. In practice it is generally assumed as 2.05 volts, this being the E.M.F. on open circuit when the battery is fully charged; that is, both electrodes being free from any lead sulphate. As the battery discharges, the voltage gradually decreases, so that when the battery is nearly discharged its voltage is less than at the beginning of discharge. The reasons for this will appear hereafter. Appearance of Plates. — The battery plates are distinguishable both by their appearance and hardness, the peroxide plate being of a reddish brown or chocolate color and hard like soapstone, and the sponge lead plate is a grayish color, and can readily be cut into with the thumb nail. Requirements. — Neither lead sponge nor lead peroxide possess any mechanical strength, and therefore in order to make them into suitable electrodes it is necessary that they be attached to a supporting plate or grid, and since lead is the only metal except the so-called "noble metals" which resists the action of sulphuric acid, the supporting grid is always made of it. In order that a storage battery should work satisfactorily the current must be distributed equally over the surface of the plate and pass through, practically, all the molecules of the active material both on charge and discharge, and it is essential that batteries be so designed as to attain this condition; otherwise portions of the plate will be overworked and will dis- integrate, while other portions may be left in good condition. Types of Mates. In the production of battery plates there are three general methods: One is known as the Plante process, which consists in chemically or electrochemically forming sponge lead or lead peroxide directly on the surface of a lead plate, this active material being produced from the lead of the plate itself. The second method consists in taking certain oxides of lead, principally litharge and red lead, and mechanically applying them to a previously prepared leaden grid — generally under pressure — and afterwards reduc- ing these oxides to sponge lead or lead peroxide. The third method, which is not much used now, is to prepare pellets of sponge lead or other lead compounds which may easily be reduced to sponge lead, placing them in a mould, and casting the supporting grid around them. In the Plante type of battery the layer of active material produced is comparatively thin, and in order to obtain a sufficiently large quantity to give each plate a reasonable capacity, it is necessary that the area exposed be made as large as possible. This is accomplished by some method which raises grooves or webs in the plate, or by making up the plate of narrow ribbons of lead, which are folded backwards and forwards until an electrode is finally produced, the thickness of which is equal to the width of the lead ribbon, the length and breadth of the plate being anything that may be desirable. The comparative value of these different types of batteries will be taken up after discussion of various characteristics of batteries in operation. Capacity. — The unit of storage battery capacity is the ampere hour, that is, the ability to discharge one ampere continuously for one hour. The capacity is dependent on the rate of discharge; the temperature; the quantity of active material present; the quantity of electrolyte in the cell, and the exposed surface of the plate. Theoretically, .135 oz. of active material per negative plate, with .156 oz. per positive or .291 oz. for both electrodes will, in the presence of suffi- cient electrolyte, give a discharge of one ampere hour. In practice about TYPES OF PLATES. 875 uJ a - cc q£i uj x l < X X CO ^^ " — 4s \ c ~\ •£ Us M r o r i \ \ \ \ i \ \ \ o AllOVdVO dflOH 8 JO ±N30H3d five times this much, or 1.45 oz. for both plates is required The reason of this is that the active material is not completely reduced, the discharge being stopped before the point of zero voltage is reached and the gradual formation of sulphate as discharge proceeds, tends to close up the pores and prevent access of the electrolyte to the mass of active material. The capacity increases with increase in temperature, being about 1 per 876 STORAGE BATTERIES. cent for each degree Fahrenheit increase in temperature. Theoretically, the ampere hour capacity of a battery should not vary with the current rate. If a battery discharge continuously 100 amperes for 8 hours, giving 800 ampere hours at this rate, theoretically it should discharge 800 amperes for one hour. As a matter of fact, however, the ampere hour capacity of a battery decreases rapidly with increase of rate of current flow. The reason for this decrease in capacity is due to several causes, the most impor- tant one being that as discharge proceeds, the active material begins to turn into lead sulphate. The volume of the lead sulphate is very much greater than the volume of the active material from which it is formed, and since the action takes place most rapidly on the surface of the plates where they are in contact with the electrolyte, the formation of the sulphate also takes place most rapidly at the surface, and this mcrease of volume tends to fill up the pores of the plate and prevent access of the electrolyte to the active material which lies beyond this shielding layer. If the discharge rate be very rapid, the masking layer of sulphate is rapidly built up, and the shielding effect takes place more quickly. In a battery discharged at a low rate the formation of this sulphate layer is so slow that the electrolyte can reach the innermost portions of the porous active material, the chemi- cal action takes place more thoroughly, and a greater amount of current can therefore be taken out. Curve No. 1 shown in Fig. 1 gives the variations in capacity with varying rates of discharge in percentages of the eight-hour rate, and curve No. 2 shows the increase in amperes output with increased discharge rates. Thus if a battery have a capacity of 400 ampere hours, it will discharge 50 amperes continuously for eight hours. If the total capacity be taken out in one hour, the discharge rate will be 200 amperes, and the ampere hours will be 200, this being 50 per cent of the eight-hour rate as indicated by the curve. If the ampere hour capacity of the battery at the eight-hour rate be known, its capacity at any other rate can be determined from this curve, or if its capacity at any rate be known its capacity at the eight-hour rate can be also determined. The curve is an average, and applies approxi- mately to nearly any type of battery, although different characters of batteries will give different curves, but none of them will depart materially from that shown in the figure. Voltag-e Variation. As stated, the voltage depends on the character of the electrodes and the density of the electrolyte. The available potential at the battery ter- minals is further dependent on the internal resistance of the cell. These facts explain the drop in voltage as discharge proceeds, as indicated by the curves in Fig. 2. 2.6 2.4 c HAR GE 2.2 CO / 5 1 , > 2.00 N ^ D!S :ha }GF 1.8 1.6 1 i HO 4 JR8 t r S Fig. 2. ELECTROLYTE. 877 The electrodes gradually change from pure active material to a mixture of active material and sulphate; the formation of the sulphate increases the resistance from the surface of the electrodes to their conducting grid, thereby increasing the internal resistance, and the surface layer of sulphate prevents access of electrolyte to the interior pores of the active material, and the small amount of electrolyte imprisoned in these pores has its SO3 rapidly abstracted from it, greatly reducing its concentration and there- fore the voltage of the cell. To this cause nearly all of the fall in voltage may be attributed. Electrolyte. The resistance of the electrolyte varies with the density of the acid, being a minimum when 30 to 35 per cent of the mixture is acid, and increasing if a greater or less percentage of acid be present. Parts of the plate surface may do more than their share of the work if the plates be very long and the containing tanks deep, this condition aris- ing from a difference in the density of the electrolyte at the top and bottom of such tanks. The containing cells should therefore never be deeper than 20 inches, unless some artificial means of acid circulation be used, such as compressed air introduced into the bottom of the tank through small rubber tubes. With such circulation the electrolyte density is maintained con- stant in different portions of the tank, and the plates will then be worked at equal current densities over their entire surfaces. Conductivity also changes with the temperature, being greater for increase of temperature. The table on page 1229 under caption "Electro- chemistry" shows the changes in electrolyte resistance with variations in density and temperature. The density of electrolyte in storage batteries should never exceed 1.200 when the batteries are fully charged, and there should be ten pounds or more of electrolyte per 100 ampere hours of battery capacity on a basis of the eight-hour rating. The final density at the end of discharge with this quantity of acid and 1.200 initial density, will be about 1.134. In motor car batteries about four pounds of electrolyte per 100 ampere hours is sufficient, and because of the small amount of acid present the initial density must be higher. If the initial density be 1.265 at beginning of discharge it will, with this amount of acid, fall to about 1.137 at the end of discharge. Since there is a definite change in density for a given amount of discharge taken from a cell, the density of the electrolyte is one of the best indications of the state of charge of a battery, provided, of course, that no internal discharge, due to local action, takes place. If, when the cell is charged, it shows a density of 1.200 and when discharged 1.130, the difference, .07, represents the total change. If at any time the density is 1.165, just one half the amount of capacity has been taken from the cell. In order that these observations may be reliable, however, it is necessary to stir the electrolyte well, so that the density is the same all through the tank; also if the discharge has taken place at a high rate, the cell must stand for an hour or more before the electrolyte will completely diffuse so that the density readings are correct. The electrolyte must be made of either distilled or rain water, mixed with pure brimstone acid. Ordinary city or well water will, in all prob- ability, ruin the batteries, and pyrites acid will most certainly do so. The electrolyte should always be tested to discover if harmful impurities are present, which are platinum, iron, chlorine, nitrates, copper and acetic acid. The tests for these are as follows: Platinum. — A complete test for this substance can only be made by an experienced chemist with proper appliances. A good rough test for traces of platinum is to pour electrolyte into a cell and note if gassing takes place on open circuit. If it does, and continues for some time, it is an indication of the presence of platinum, and the suspected electrolyte should then be sent to a chemist for analysis. Never use chemically pure sul- phuric acid which has been refined in platinum stills. Iron. — Take a sample of the electrolyte and neutralize with ammonia. Boil a small portion with hydrogen peroxide, which process will change whatever iron may be present into the ferric state. Add ammonia or 878 STORAGE BATTERIES. caustic potash solution until the mixture becomes alkaline. Iron will be indicated by a brownish red precipitate which will then form. Chlorine. — Take a small sample of the electrolyte, add a few drops of nitrate of silver solution of concentration of twenty to one. A white pre- cipitate will indicate chlorine. This precipitate will be redissolved by addition of ammonia, and can be re-precipitated by the addition of nitric acid. Nitrates. — Place some of the electrolyte in a test tube, and add strong ferrous sulphate solution. Then carefully pour down the side of the tube a small amount of chemically pure concentrated sulphuric acid, so that it forms a layer on top of the liquid. If nitric acid be present it will be shown by a stratum of brown color, which will form between the electro- lyte and concentrated acid. Acetic Acid. — Add ammonia to a sample of electrolyte until it becomes neutral, then add ferric chloride (Fe2Cle). A red color will indicate the presence of acetic acid, which may be confirmed by the addition of hydro- chloric acid, which will bleach the mixture. Local Action. — Certain metallic impurities present in the electrolyte may be, on charge, carried over to the negative plate, and the hydrogen there evolved will turn these impurities into pure metal. The condition then exists of the sponge lead plate having a different metal attached to it, and in electrical connection therewith, and the two immersed in elec- trolyte. If the voltage of such a couple is sufficiently high to decompose the electrolyte, current will begin to flow, the whole acting as a short- circuited battery at the negative plate. This discharges the negative, either wholly or partially, according to the amount of metallic impurities which may be carried over, and it is then not in a proper condition to discharge in company with the positive plate when it is desired to take current from the cell. If this local action continues for some time the negative plate may be so far discharged that it will sulphate, and finally become worthless. Cadmium Test. The condition of the negative and positive plates can best be ascer- tained by measuring the voltage between the plate under examination and a small test electrode of cadmium. This cadmium should be covered FU LL :harg E CA >MI UM ^> '4 t-4 '* FU LL DIS CH ASC E .'.10 CO So. o > .10 .20- .30 - Fig. 3. with rubber, perforated so that the test piece cannot come in contact with any of the battery plates or connections, though the electrolyte may freely penetrate to it. When a cell is fully charged, showing a voltage of 2.5, the voltage between the negative plate and the cadmium should be from .16 to .2 volt. When discharge takes place, this voltage gradually reaches zero, after which a potential begins to rise in the opposite direction, gradually increas- ing with discharge. When the voltage, after passing through zero, reaches a value of .25 volt, the full amount of discharge has been taken from the negative plate, and the current should be cut off regardless of the potential of the cell. Figure 3 shows the way in which the potential between the negative EFFICIENCY. 879 plate and the cadmium changes. The cadmium undergoing no discharge does not change, and its line of potential is therefore horizontal and unchanging, as indicated. The negative plate, however, is discharging, and its potential decreases so that, though it begins to discharge at a poten- tial of .18 volt above the cadmium, it soon reaches a point at which it is the same as the cadmium, the voltage between them then being zero. As the potential of the negative falls further, a potential begins again to appear between the two, but, as is obvious, it is in the reverse direction, as the potential of the negative plate is now lower than that of the positive. On charge, the voltage between the cadmium and the negative plate should be brought up to at least .17, even if continued overcharge after the cell has reached 2.5 volts is necessary to do it. Batteries are so designed that the negative plates work through their proper range of potential with normal change in the cell E.M.F., but over- sulphation, reduction in amount of active material, or, most of all, local action, will destroy this balance, and these cadmium tests are useful in keeping watch over the condition of batteries in service. Polarization. If the voltage of a battery on open circuit be a given amount, say 2 volts, and charging current is sent into it, it would be natural to assume that the potential rise at the battery terminals would be equal to the drop due to the internal resistance of the battery. It is found, however, to be very much greater than this amount — the actual internal resistance of large cells being practically negligible. This increase in drop, when cur- rent passes through a cell, comes from a phenomenon known as polariza- tion, which is, in effect, the production of a counter E.M.F. which opposes the flow of current, and which always takes place whenever current passes from one electrode to another immersed in an electrolyte. This effect also opposes the flow of discharging current, and causes the voltage drop at the cell terminals, which is observable when current is taken from a battery. The principal polarizing agent is hydrogen, which may be considered as an electro-positive element. It always forms at the negative electrode and sets up an E.M.F. opposing current flow. In cells of the same type the drop at any given time rate of charge or discharge is the same for any size of cell. The Voltagre I>rop in cells of a given type is independent of the size of the cell, but varies with the state of battery charge and the rate of discharge. This drop is also fairly constant for various types of cells. The following table gives the fall or rise in voltage from the open circuit E.M.F. when discharge or charge takes place: 8-hour rate 05 volt 6 " " 065 4 " " 09 3 " " 11 2 " " 14 1 " M 2 Efficiency. The efficiency of the storage battery, similarly to that of any other device, is the ratio of the watts output to the watts input. If current be taken out at a high rate, and a resulting small capacity be obtained, it does not follow that the efficiency has been lowered correspondingly, as it will be found that the amount of current required for succeeding charge will not be so great as if a lower rate of discharge had been used, and a greater amount of energy taken from the battery. In other words, there is a relation between the amount of energy derived on discharge and the amount of energy required on subsequent charge to bring the battery back to the condition at which the discharge began. The efficiency of batteries which discharge only a few moments and immediately after receive charge, that is, in which the charge and discharge fluctuate rapidly, 880 STORAGE BATTERIES. and the net amount discharged from the battery in an interval of time is small, is about 90 to 92 per cent. Where used for power storage, a long continuous charge being sent into the battery and followed by a long con- tinuous discharge, the efficiency is from 75 to 80 per cent. The losses in a battery are made up of the I-R, and the gassing at the end of charge, in which the constituent gases which are released by the action of the electric current do no chemical work on the electrodes, but escape into the air, the energy required for this dissociation being lost. There is also the further loss due to the counter E.M.F. of polarization, as has been explained. Comparison of PI ante and Pasted Electrodes. Of the two types of cells mentioned, the Plante and the pasted, each has its particular place, and one is more suitable than the other for its partic- ular class of work. The pasted negative plate is, in general, the best type for nearly every class of work. Pasted positive plates are necessary in batteries where light weight is required, such as in automobile and train lighting batteries. They are also suitable for battery plants which receive long charge, store the energy and discharge over a considerable length of time, such as resi- dence and isolated plants, and central lighting stations. The Plante posi- tive is most suited to those conditions where the battery discharge takes place for short intervals at very high rates, such as regulation of railway and elevator loads, and also when prolonged overcharge is likely to occur frequently. Charging*. In charging, the voltage gradually rises, as shown by the upper curve in Fig. 2, until about 2.5 volts are reached, when, at both the positive and negative plates, gases are rapidly released. Charge should always be continued until both plates gas freely. Full charge will also be indicated by the electrolyte density rising to its proper value. The best way to charge is to send in current rapidly at the beginning and gradually decrease it until at the end of charge the current flow is very small. For instance, in charging a 1,000 ampere hour cell for eight hours, the average rate of flow is 125 amperes. The proper rates at which to charge this cell would be 250 amperes for 1 hour 200 " " 1 " 150 " " 3 " 75 " " 1 " 25 " " 1 " For rapid charging, when a battery has to be charged in four hours, the current should vary as follows: 40 per cent of total 1st hour 25 " " " 2d 20 " " " 3d 15 " " " 4th " For quick charging in three hours the rates should be: 50 per cent 1st hour 33J " " 2d " 163 " " 3d " BATTERY TROUBLES. 881 Whatever the rapidity of charge, never send a heavy current into a battery toward the end of charge. The rapid rates can only be used during the early part of charge. In case of loss of electrolyte from the cells from evaporation or spraying, add only pure water to maintain its level, as the addition of normal elec- trolyte will gradually increase the density of that in the cells, because the added liquid merely takes the place of that which has been carried off as gas or lost from evaporation, which, in either case, is pure water only. High electrolyte densities tend to accentuate all the troubles that can befall a battery, and accelerate the formation of sulphate. The water should be introduced through a rubber hose or lead pipe extending nearly to the bottom of the cell, so that it will diffuse and mix with the electrolyte. If the water be poured in, it, being lighter than the electrolyte, will float and take a long time to diffuse with the liquid in the cell. Removal from Service. To take a battery out of commission it should first be fully charged, then given a good overcharge, and then discharged down to 1.7 volts per cell in the electrolyte, immediately after which the electrolyte should be drawn off, and either distilled or rain water put in the cells. The dis- charge should then be continued until the voltage comes down practically to zero. In most cases it is necessary to short-circuit the cells in order to get them down nearly to zero with pure water as the electrolyte. Dis- charging them in the water has no injurious effect, however, as no sulphate can form. Upon complete discharge the water should be poured out of the cells, and the plates thoroughly washed, generally by running water continuously through the cells. All water is then drawn off, and the plates may then stand for any length of time without injury. ^ When the bat- teries are again to be used, it is only necessary to pour in the electrolyte and give a long overcharge. Battery Troubles. The principal troubles which are encountered in battery operation are loss of capacity, buckling, shedding of active material, sulphation and loss of voltage. JLoss of Capacity usually comes from clogging of the pores in the plate with sulphate which is not visible to the eye because the surface of the plate is maintained in proper condition but the interior portions of the active material have not been thoroughly reduced. This condition can be remedied by prolonged overcharge at low current rates, say about one- fourth the normal eight-hour charging rate. JLoss of ActiVe Material will also reduce the capacity of a plate, and this takes place continuously, but slowly, in every storage battery, and may be considered as the normal depreciation. If the battery be over- worked, however, and especially if discharge be carried too far, the amount of sulphate formed will so expand the active material as to cause it to crack or shed off very rapidly. Buckling-. — Under the action of unequal expansion of the two sides of the plate, or certain portions of the plate, the strains may distort it and cause it to assume a buckled shape, that is, bent so one side is concave and the other convex. This is due, in every case, to over-discharge on either the whole or some portion of the plate, and consequent over-sulphation and over-expansion. In certain battery plates, which are designed to allow this expansion, buckling cannot take place, but in most of them the active material is on an unexpanding framework, and over-discharge is therefore to be avoided. Sulphation. — This is practically the cause of every storage battery trouble, and can only be avoided by stopping the discharge before the voltage of the cells has fallen too low, namely, at l.S volts per cell, with normal discharge current flowing, and by occasional boiling, that is, over- charge which should be given at intervals of about three or four weeks. 882 STORAGE BATTERIES. ' In giving this overcharge the battery should be fully charged at normal rates until it shows about 2.6 volts per cell. The current should then be decreased to about one-half its normal eight-hour rate, and the charge continued until the cells show about 2.65 volts, and about twenty minutes after this potential is reached. This will effectually reduce any sulphate which may have accumulated in the pores of the active material. A bat- tery should never be allowed to stand idle or uncharged after discharge, as the plates will sulphate very rapidly. A charge should be started imme- diately after discharge, or as soon thereafter as possible. JLosa of Voltagre. — It will frequently be found that one or more of a number of cells will show a lower voltage than the others. This generally occurs because of loss in capacity, so that a cell having this lower capacity, and in series with the main battery, would discharge the same amount as the other cells having a higher capacity, and in this way its voltage would drop more rapidly and always be lower than that of the other cells on discharge. Testing*. There are two classes of storage battery tests. One is to determine whether a battery which has been installed meets the conditions of the specifications; the other is to determine all the constants of a battery as compared with others on the market, either for purposes of improving the product of the factory or determining its commercial value. The first class of tests will not be gone into here, as they will be indicated by the conditions of the contract and specifications. In the second class of tests the following are the points to be determined: 1. Weight of complete cell. 2. Weight of the separate component parts, namely, elements, electro- lyte, separators and containing cell. 3. Dimensions of component parts of the cell. 4. Rates of charge, maximum and normal. 5. Rates of discharge, maximum and normal. 6. Capacity at low, normal and rapid discharge rates. 7. Voltage curves of charge and discharge. 8. Internal virtual resistance. 9. Variation in density of electrolyte. 10. Loss on charge with time. These are all determined by test and observation, and from them are deduced: 11. Charge and discharge rates per square foot of positive plate surface. 12. Charge and discharge rates per pound. (a) of complete cell. (6) of element. 13. Capacity per pound. (a) of complete cell. (6) of element. 14. Efficiency at various charge and discharge rates. Weight of Complete Cell and Component Parts. The weight of complete cell is of course found by means of the scales, and in order to determine the weight of the component parts the elements should be partly discharged, then removed from the electrolyte and dried with blotting paper, after which they are weighed. Do not keep the nega- tive plates in the air any longer than necessary. The weight of the elec- trolyte is equal to the total weight, less that of the elements and jar. INTERNAL VIRTUAL RESISTANCE. 883 Dimensions* These are determined by usual are dismantled for weighing, and Fig. 4. and V is a low-reading voltmeter measurements at the time when the cells should include dimensions of separators, height of lower edge of plate above bottom of jar, clearances between adjacent plates and between interior of jar and plates. Also area of plate surfaces and of con- ducting lugs. This latter for the purpose of determining if current densities are within usual practice, namely, about 150 amperes per square inch. The cell may then be reassembled, given a prolonged overcharge, and connected up for testing. Connections for Testing*. — Re- ferring to Fig. 4, R is an adjustable resist- ance by means of which the current to the battery may be kept constant. B is the cell under test; lates in the adjacent cell by burning each of the plates separately to the eaden bus bar, as shown in Fig. 5. In the smaller sizes of cells which .NEG. .PLATE LUG /.BUS BAR .P0S. PLATE, LUG — ^ BURNED JOINT NEG. PLATE LUG -vglass support plate- \lead lining lead lining^- ^wood tank wood tank" Fig. 5. have lead lined tanks, they are generally set on a framework from twenty to twenty-four inches high and rest upon four insulators. The plates of each cell may be joined to those of the next succeeding cell either by burn- ing to a common bus bar, as above mentioned, or by bolting together lead straps which form the cell terminals, the bolts and nuts being, of course, lead covered. ^ If the containing vessels are glass jars, it is usual to set each of them in a shallow wooden box about 1§ inches deep and filled with fine, dry sand. The glass cell beds itself in this sand, giving an equal distribution of pressure over the bottom of the jar, and the sand also catches and absorbs such electrolyte as may be spilled or sprayed out with escap- ing gases. Each sand tray, as these are termed, rests upon four porcelain insulators, and the cells are placed on a framework in one or two tiers, as may be desired. Fig. 6 shows this method of installing. JLead Burning:. — The hydrogen flame has the special property of not oxidizing, or otherwise soiling the lead, and is therefore used for melting together two lead surfaces, notably that between cells and the sheet lead lining of the tanks. Hydrogen gas is generated in a vessel from sulphuric acid and zinc. The gas is collected and passed through a water bottle to a burner, where it is mixed with air that has been forced into the burner by a pump or bellows tne_ mixture being ignited for the welding. 886 STORAGE BATTERIES. + PLATES PLATES ;v^-^JSAND;'^V^^ : ':VS ? >;-y^'v":' rT v ,„ , „„„„„„„„ "S zn Fig. 6. Uses of Batteries. The principal uses are : (1) For propelling electrically driven motor cars. (2) For railway train lighting. (3) As a substitute for the ordinary primary battery in telephone and telegraph work. (4) To carry the load peak on a supply system. (5) To carry the entire load during the periods of light demand, the generating equipment being shut down. (6) To regulate the load on systems where the demand fluctuates widely. (7) To act as an equalizer on three-wire systems in which the gener- ators are connected across the outsides of the system and give a corre- sponding voltage. (8) To reduce the amount of copper required for systems supplying variable loads. (9) To insure continuous service. (10) As auxiliaries to exciter dynamos in large alternating current stations. (11) Combinations of any of above from (4) to (8). The first three applications involve no special engineering knowledge. (4) In case of a supply system on which the load rises greatly during certain hours of the day, as shown by the load curve A, B, C, D, E,F, G, H, in Fig. 7, it is often advisable to install a battery to receive charge during the period of light load, as shown by the shaded area in which the heavy curve is the demand on the station and the light curve, the load on the generating equipment, the difference going into the battery; and to dis- charge in parallel with the generators during the heavy output on peak d, E, e, as shown by the cross-hatched area. Such a battery assists to USES OP BATTERIES. a 1 X 887 c_ 1 %Hk N S? 3 \ '■'■■'■■ ' 1 '■■'• .;..-: J5 c» '-•> -<" K ■ P ^ ■« // Z^^x^^% X" *~t i f y//W//////>' r 5* ^ VJ^ l iJ^ ^•J^jL^^^J-.. a IP i^~lv 1 ■ rO - i c 3 "^ Ik u 1 ■ I--- u UJ O o cc < ■ t J i ' f / < X o CO Q / , ■ V . f 7 say 3dIAI\ p CD o o o o CJI o o o ST o o T o o vo 1 a \ : D 3 C C c 3 > s o Fig. 7. 888 STORAGE BATTERIES. maintain a reasonably constant load on the dynamos, reduces the cost of the generating equipment, and is always ready to take up any excess load on the system, such as may come from a suddenly overcast sky or storm, without the loss of time necessary to fire up additional boilers and start additional engines, as would be the case if the entire load were carried by generating machinery. (5) After the peak discharge is ended and the load on the system decreases below the generator capacity, the batteries may be fully charged by, say, midnight, and the entire plant shut down during the period of light load until, say, five or six a.m. This is also indicated in Fig. 7, where the shaded area e, f, g, h, F, represents charge put in after the peak dis- cbarge, while the cross-hatched area h, k, m, n, indicates battery discharge. If the battery is large enough to do this, the cost of the fuel and the depre- ciation, for the time of shut-down, are saved, and two shifts of station attendants only are required instead of three. (6) In case of a system on wmich the load fluctuates rapidly and between wide limits, such as an electric railway or elevator load, the form of load diagram will be as showD in Fig. 8. Here it will be seen that the A 900 1— I » Sr- OUU q. I II tt V I 1 \ A \ -T ft I moJX t t 44 *- it-, h I 3-3QLJ 1 3 - 1 I 3E roo M""*~TT "" r"H 1— T+Tf H 500 1 _ ±7 t ' I ~W\ "ir H 400 t - 3L I fl "IVr < ~t\ t, 44- t _.I 7SEI I o 00 nt jft L _-__ 3l\t I A- £lAt 200' ' L H-t o READINGS TAKEN AT I SECOND INTERVALS Fig. 8. load varies from 200 to 1,000 amperes, though the average load taken over in diagram shown is 533 amperes. If the system be without a battery the generating equipment, including steam plant generators and accessories, will have to be of sufficient capacity to carry the maximum load, and the moving macbinery will be subjected to excessive shocks and strains due to the sudden loads. The fuel con- sumption is also much more than it would be if the engines could work under steady load. If a regulating battery be used, the generating equip- ment need only be great enough to supply the average load, as the battery will absorb all fluctuations. When the current required to supply the external circuit is small, the additional amount, supplied by a generator working under constant load, will go into the battery and be stored there as charge. When the external load exceeds the average generator out- put, the excess is furnished by the battery discharge. Thus the battery maintains a constant load on the generating equipment regardless of the variations in the external load, and the attendant advan- tages of fuel economy, normal duty only on moving machinery, decreased depreciation and repairs, are realized. (7) In three-wire systems, if the generators give a voltage equal to that between the outside mains, some forms of equalizer are necessary to prevent the unbalancing which may take place. If a battery be connected across the outsides with a sufficient number of cells in series to give an E.M.F. METHODS OF CONTROLLING DISCHARGE. 889 equal to that of the system, and the neutral be connected to the middle of the battery, any excess of current flow, on one side of the system, will be supplied by discharge from the half of the battery connected across that outside and the neutral, while the half of the battery on the other side will receive an equivalent amount of charge. This is a widely used arrangement, as all the other advantages of storage batteries are obtained in addition to the balancing effect. (8) In cases where current is transmitted over a considerable distance, and the load varies either at different periods of the day, such as a lighting load, or rapidly, as a railway load, a storage battery located far away from the station, near the point where the load comes on the system, may be made to maintain the voltage at periods of heavy load when the feeder drop would be excessive, and the useful potential too low for satisfactory service. This is accomplished by the discharge of the battery when the heavy load comes on, reducing the amount of current transmitted and therefore the drop. The battery is charged during the time of light load when sufficient current is transmitted to supply the load, and also charge the battery. In other words, the battery equalizes the load over the line, causing the continuous flow of average current, and reducing the cost of feed or copper. In certain classes of rapidly fluctuating loads this effect is automatic and produced by slight changes in line drop, with small changes in the load over the line. (9) To insure continuous operation of any electrical plant a storage battery is necessary. No matter what may happen to the generating part of an equipment, if a storage battery be connected to the system it will immediately take on the load and carry it a sufficient length of time to enable any quick repairs to be made and the machinery again started up. (10) In large central stations where alternating current is generated and distributed to substations, and a large territory is dependent on the station supply, the failure of an exciting dynamo would cause a shut-down of pos- sibly several minutes, which would be a serious mishap. To insure against this a storage battery is connected directly across the exciter bus bars. It does no work and is never % of any real service unless failure of an exciter takes place, in which case the alternator field excitation is taken up without a break or interval. The insurance against stoppage, even for a moment, by means of the storage battery, is so thoroughly demonstrated that nearly all the large alternating current stations have added this equip- ment to their exciter systems. (11) Combinations of (4) to (9) inclusive can be in part effected by a single battery, such as regulation of fluctuating load, discharging on peaks and carrying the night load alone, or equalizing on a three-wire system, carrying peaks on both sides of the system, and also carrying the light load alone. Many other combinations will suggest themselves to the engineer as the conditions to be met may reqinre. Methods of Controlling: Discharge. In Fig. 2 is shown the change in voltage of a cell when charging and discharging at the normal rate. In order to compensate for this variation, so that the E.M.F. supplied to the discharging circuit may be maintained constant or varied at will to meet external load conditions, the following methods of control are used: (1) The number of cells in series may be altered by means of suitable switching mechanism. (2) Counter cells, or cells connected in opposition to the main battery, may be included in the discharging circuit and the desired voltage obtained by varying the number of counter cells in this circuit. (3) A variable resistance may be interposed in the main circuit to regu- late the discharge. (4> A dynamo-electric machine, termed a " booster," having its arma- ture in series with the battery circuit, its field being variable at will as to either direction or magnitude, may be employed. If any of the first three methods be employed, the total number of cells 890 STORAGE BATTERIES. composing a battery must be such that at the end of discharge, with normal outflowing current, the sum of the voltages of all cells in series is equal to the voltage to be maintained on the supply circuit. When discharging at normal rate, it is usual to stop discharge when the E.M.F. per cell has dropped to 1.8 volts. End Cells and Switches. Fig. 9. As shown in Fig. 2 the E.M.F. at the beginning of discharge is 2.15 volts, and at this point on the discharge curve only 51 cells would be required to give 110 volts; as discharge continues and the E.M.F. falls, the number of cells in series must be increased accordingly, and at the end of discharge, when the cell voltage is 1.8, 61 cells are required in series to supply a 110-volt system, 10 of them being end or reserve cells. The whole 61 cells would be con- nected in a single series, a conductor being connected to each of the ten end cells and to suitable contacts on an end cell switch. The voltage across the dis- charging circuit will be depen- dent upon the number of cells included in the circuit. Figure 9 shows an arrangement of cells, all connected in series, a portion of these being end cells; the voltage when the moving arm M is in the posi- tion shown by the full lines will be that due to all the cells in the main bat- tery, plus the voltage of the two end cells included by the arm. If now the arm be moved to the position shown in the dotted lines, the voltage across the mains L will be increased by the addi- tion of the end cells 4 and 5. In switching from one end cell point to another the discharging circuit must not be opened, neither must the moving arm touch one contact before leaving the adjacent one, since the joining of two con- tacts will short-circuit the cells connected thereto. In general, the form of switch for this pur- pose is essentially that shown in Fig. 10, where the moving arm is provided with a small ad- vance arm, the two being insulated from each other but connected through the resistance X. The spacings of the two arms and contacts are such that when the main current carrying arm is squarely on an end cell contact, the advance or auxiliary arm touches no other contact, but in passing from one point to the next, the ad- vance brush reaches the contact towards which the arm is moving, before the main brush leaves its contact; the resistance X between the two points prevents short-circuiting, and the current to the main circuit is never broken. The conductors joining the end cells to the end cell switch contacts must be of the same sectional area as the conductors of the main circuit, for when any end cell is in use the conductor connecting it to the switch becomes a part of the main circuit. 1000 amperes per square inch, when the battery is discharging at the two-hour rate, is good practice. End cell switches of small capacity are made circular; the larger sizes are, however, made horizontal in form, and both types may be either manu- ally operated or motor driven. End cell switches of large capacity are generally located as near the battery room as possible, to avoid the cost of running the heavy con- FiG. 10. BOOSTERS. 891 ductors, and when such switches are motor driven, the usual practice is to control their operation from the main switchboard. Automatic end cell switches have been used more or less abroad, but have found little favor in this country. The controlling devices for such switches are so arranged as to make the switch automatically respond to changes in the discharging circuit. Counter E.M.JF. Cells. Counter cells or counter electromotive-force cells are merely lead plates in an electrolyte of dilute sulphuric acid; they have no capacity but set up an opposing E.M.F. of approximately 2 volts per cell if current be passed through them. In using these cells for controlling discharge, the total number of active cells in the battery will be the same as if the method of end cell control had been used. The counter cells represent an increase in equipment, the additional expense being 8 per cent or more. Figure 11 shows the method of counter cell control; these cells are con- nected in opposition to the main battery, and conductors are run from each of the counter cells to points on a switch similar to an end cell switch. At the beginning of discharge all the counter cells are in circuit, acting in oppo- sition to the main battery. As discharge proceeds and the battery voltage falls, the counter cells are gradually cut out of circuit. Controlling discharge by counter cells is now nearly obsolete prac- tice, and is scarcely ever to be recommended; the only advantage in this method of control is that the discharge throughout the bat- tery is uniform, but this fact alone does not warrant the use of such methods on account of the addi- tional expense involved, and the energy loss when discharging against counter cells is the same I M I I I I I i I 9 o 9 'discharge C. E. M. E. CELLS Fig. 11. as if resistance had been interposed in the discharging circuit. Resistance Control. The discharge may be controlled by a variable resistance included in the discharging circuit. This method is not used unless the battery is of small capacity and the cost of energy low. Figure 12 shows a diagram for resistance control. In small plants, where the available space for battery auxiliaries is limited — such conditions obtaining in bat- teries for yacht lighting and the FlG. 12. hk e — the resistance control has some merit. M i I I I I I I I I l III l—A/VW VARI NAM VARIABLE RESISTANCE Boosters. A booster consists of a dynamo electric machine, the armature of which is in the battery circuit, its E.M.F. being added to or subtracted from that of the battery to produce discharge or charge. This action of the booster, i.e., the direction and magnitude of its armature E.M.F., may be auto- matically or manually controlled. 892 STORAGE BATTERIES. The Shunt Booster. JUPPLY MAINS As shown by the battery curves in Fig. 2 the maximum voltage per cell at the end of charge is 2.6 volts. As 61 cells are required for a battery operating on a 110-volt circuit, the total charging voltage required is 2.6 X 61 = 158.5 volts, or about 50 volts higher than the voltage of the supply circuit, and to fully charge the battery this additional voltage must be supplied by a booster or by an excess voltage in the charging generator. Figure 13 shows the diagram of a simple charging booster. Its armature should be wound for the normal charging current, and have a maximum voltage equal to the difference between that of the supply circuit and the maximum charging voltage. The field is separately excited, either from the bus bars or the bat- tery, and the voltage at the armature may be varied by the field rheostat. Instead of discharging through an end cell switch or resistance, the current through the booster field may be reversed and varied, so that the E. M.F. of its arma- ture may oppose that of the battery, this E.M.F. being reduced as the battery vol- tage falls, the algebraic sum of the booster and battery E.M.F.'s being always equal to that of the supply circuit. In this case, however, it is usual to put in fewer cells, the available voltage being taken as 2 volts per cell. On dis- charge when the voltage of all cells in series is greater than that of the supply circuit, the booster voltage is equal to the excess battery voltage over the supply circuit potential, and in opposition to the battery voltage: when the battery voltage becomes equal to that of the supply circuit the booster voltage is zero; when the battery voltage falls below that of the supply circuit, the booster voltage must then be~ in a direction to assist the battery, adding its voltage to that of the battery. Automatic Boosters. In batteries which are used for regulation on fluctuating loads, the changes from charge to discharge and vice versa are so rapid that the state of battery charge changes but little. The voltage of the battery, however, changes with these fluctuations, increasing with inflowing and decreasing with outflowing current. In this respect the storage battery has much the same characteristics as a shunt wound generator: with increasing output the battery voltage falls, due to the drop caused by internal resistance and polarization; with decreasing output the voltage rises for the same reasons. These voltage changes are approximately proportional to the rate of current flow causing them. The fluctuations coming with such rapidity and irregularity must be automatically compensated for by changes in booster voltage, which vary both in direction and magnitude with the direction and rate of current flow. There are two generic types of automatic boosters, viz., the non-reversible and the reversible. BOOSTERS. 893 Jfon-Iteversible Booster. In installations where it is desired to supply both an approximately constant and a fluctuating load, from the same generators — such condi- tions obtaining in an office building or hotel, where it is necessary to supply lights and elevators from the same source of supply ~ the fluctuations in the power circuits must not interfere with the lighting circuits, and to prevent this, two sets of bus bars are provided. The generators are con- nected in the usual manner to one set of bus bars, and the lighting circuits are connected across these. Across the other set of bars are connected the circuits supplying the fluctuating load, and the battery is also con- nected directly across these power bars. The power bars are supplied with current from the lighting bars, a non-reversible or so-called " constant current" booster being interposed between the two, as shown in Fig. 14. Since this permits only a constant current to pass from the lighting bus bars, the load on the generator does not vary, although the load on the power busses may vary widely. The connections and operation of this system are as follows: tLEYATOR LAMPS or u Fig. 14. The booster armature and field are in series between one side of the light- ing and power bus bars. A shunt field is also provided, which acts in opposition to the series field. This booster carries a practically unvarying current from the lighting to the power bus bars, regardless of the fluctua- tions of the external load, which current is equal to the average required by the fluctuating load. Except under abnormal conditions, the shunt field always predominates, giving a voltage which is added to that of the lighting bus bars, so that the voltage across the power busses is always higher than that across the lighting by an amount equal to the booster voltage. If an excessive load comes on the power circuits, the increased excita- tion of the series coil, due to a slight increase in current from the lighting to the power bus bars, lowers the booster voltage and consequently reduces the voltage across the power bus bars. The battery discharges, furnishing an amount of current equal to the difference between that required by the load and the constant current through the booster. u If the power load decreases below the normal value, the slight decrease m current in the booster series field increases the booster armature voltage, and the excess current goes into the battery. The booster therefore does not in reality give a constant current, but by proper design the variation may be kept within a few per cent. 894 STORAGE BATTERIES. Fig. 15. Reversible Booster. Figure 15 shows a diagram representing one form of booster for produc- ing charge and discharge in accordance with variations in load, in which S represents a series field winding, and f a shunt field winding. The gen- erator output passes through the series winding, and the current in the coil S is to remain practically constant. The shunt coil / produces a field which opposes the field produced by S, the resulting magnetization being, in direction and amount, the resultant of the two field strengths. The adjustments are so made that when the normal generator current is passing through the series coil S, the shunt field just neutralizes its effect, and the resultant magnetization is zero. Since the open-circuit voltage of the battery is equal to that of the system, neither charge nor discharge takes place. With increased demand on the line, the slight increase in generator current in the coil S overpowers the shunt field, and causes an E.M.F. in the booster armature in such a direction as to assist discharge. If the external load falls below the average demand, the current in the coil S decreases slightly so that the shunt field predominates, producing a booster armature E.M.F. in a direction to assist charge. Although the voltage of the battery falls while discharging by an amount proportional to the outflowing current, the increased excitation due to this current through S is also proportional to it, and the booster voltage rises as that of the battery falls, their sum being always equal to that of the system. In other words, the booster serves to compound the battery for constant potential. Externally Controlled Boosters. The types of boosters before described, depend for their action on the differential relation of shunt and series coils, and produce a constant volt- age change, to charge or discharge the battery, with a given change in generator current. This is not the desired relationship, as the voltage required to effect a given charge or discharge of a battery varies greatly with its state of charge and its condition. Also, such Boosters require large frames for a given kilowatt capacity in order to accommodate the windings. Recently, systems of external control have been devised, which make use of ordinary shunt-wound machines as boosters, the fields being regu- lated to produce the proper voltages for effecting charge or discharge, by an external device which is, in turn, controlled by small changes in gener- ator current. So successful have EXCITER SERIES COIL these later forms been, that they have superseded the differentially wound boosters for both reversible and non-reversible control. One form is that of Hubbard, in which the external controller is a small exciting dynamo. The gen- eral arrangement is diagram mati- cally shown in Fig. 16. The exciter is provided with a single series coil, through which the station output or a proportional part thereof, passes; the armature of the exciter is connected to the excit- ing coil on the booster, and thence across the mains, as shown. With the average current passing through the field coil or the exciter, its arma- ture generates an E.M.F. which is equal to that of the system, and in oppo- Fig. 16. BOOSTERS. 895 Fig. 17. sit ion to it. These two opposing E.M.F.'s balance, and no current flows in the booster field coils. m With an increase in external load above the average, the tendency is for an increase to take place through the exciter series coil, augmenting its field strength and consequently the exciter armature voltage. This latter now being higher than that of tne line, causes current to flow in the booster field coil, in such a direction as to cause an E.M.F. in the booster armature which assists the battery to discharge, and is of a magni- tude to compensate for the battery drop occasioned thereby.^ When the load decreases below the normal, the current in the exciter field is decreased, and its armature voltage falls below that of the system. Current will now flow in an opposite direction in the booster field coil, generating an E.M.F. in the booster armature to assist charge. Since the exciter always gener- ates a voltage in opposition to that of the line, this system is known in the trade as the Counter E.M.F. System. Another type of externally controlled booster is that of Entz. The arrangement and connections are shown in Fig. 17. Ri and R 2 are two resis- tances made up of piles of carbon plates. These resis- tances diminish greatly in value when subjected to pres- sure. L is a lever resting on the tops of the piles, Ri and R2, which is pulled downward to compress them, by the spring at one end and the electromagnet S at the other, as shown. The magnet winding is in series with the current from the generator, and with normal output to the load M.M., the pressures of the spring and the magnet are so related that the resistance of Ri equals that of R 2 . The booster field has one terminal connected to the middle point of the battery, and the other terminal is connected to a wire which joins the upper ends of the two carbon piles. The lower end of R) is connected to the positive side of the circuit, and the lower end of R 2 to the negative side. The drop through R t plus R 2t i.e., from the positive to the negative side, is equal to the potential of the system, and therefore, when R^ is equal to R 2 the drop through either is equal to one-half the potential of the system; hence the potential of the terminal of the field coil /, connected to the upper ends of the resistances, is midway between the potentials of the positive and negative mains. Since the other terminal of coil / is connected to the middle point of the battery, its potential ^ is also midway between the potentials of the positive and negative mains, from which it follows that when Rx and R 2 are equal there is no difference of potential between the field coil termi- nals, consequently no excitation, and the booster potential is zero. If the external load should increase, a small increase in generator current will cause a stronger magnet pull, decreasing the resistance of R 2 and increasing that of R t . The drop through R 2 becomes much less than half the potential across the mains, anc] consequently there is a potential across the field winding f to cause current flow from the middle point of the battery, through the winding, through the diminished resistance R 2 , to the negative main. This produces a booster E.M.F. in a direction to discharge the battery and cause it to assist the generator to supply the load demand. Conversely, if the external load M.M. should decrease, the diminished pull of the magnet due to the slight decrease in generator current allows the spring pull to predominate, and the resistance of Rt is decreased while that of R 2 is increased. The field / becomes excited by current flow from the positive main, through the diminished resistance R lf through field /, to the middle point of the battery. This sets up an E.M.F. in the booster armature to charge the battery, the difference between the normal gen- erator output and the load demand being thus absorbed. Owing to the comparatively small change in the pressures which the magnet S exerts, and the thereby limited size of the carbon piles, this sys- tem is only directly applicable to small boosters. Where large machines 896 STORAGE BATTERIES. are to be controlled, the booster has a small exciting dynamo, its field being controlled as above described. Another form of externally controlled booster is that of Bijur and is shown diagrammatically in Fig. 18. The booster field winding has one terminal connected to the middle point of the battery, the other terminal being connected to the wire join- ing the resistances Pi and P 2 . L is a lever carrying at either end a number of metallic contact points Pi and P2 which dip into troughs of mercury D t and D 2 when one end of the lever moves upward or downward. These points are connected to corresponding points on their respective resistances, and therefore all of the resistances connected to contact points which are immersed in the mercury are short-circuited. The points are of unequal length, being in a step formation, so that they gradually contact with the mercury as the lever is moved. If more of the points Pi than points P2 are immersed in the mercury the resistance P2 is less than Pi, more sections of it being short-circuited. Current will therefore flow from the middle point of the battery, through the booster field /and through resistance P 2 to the negative side of the system, exciting the booster field and producing a booster E.M.F. to charge the battery; while if more of the points Pi are immersed the resistance Pi becomes the smaller, and current then flows from the positive side of the p + system through resistance Pi, through booster field /, to the middle point of the battery, the field excitation and the booster E.M.F. produced being in a direction opposite to the first described, and tending to discharge the battery. When the resistances Pi and P2 are equal there is no potential to send current in either direction through the field coil /. When the load on the external circuit is normal, the lever L is in a hori- zontal position, resistance of Pi is equal to the resistance of P 2 , no current flows through the booster field, the booster E.M.F. is zero, and no current passes into or out of the battery. With increase of external load the pull of magnet S is strengthened by a small increase in generator current passing through the winding. This draws down the left end of lever L, overcoming the pull of the spring. The contacts Pi are immersed to a greater or less degree in the mercury, thereby short-circuiting portions of Pi and decreasing its resistance. This pro- duces a current flow in the booster field to cause an E.M.F. to discharge the battery and assist the generator to supply the load demand. A decrease in external load is attended by a slight diminution in gen- erator current; magnet S is weakened, the pull of the spring predominates, resulting in a movement of the lever to immerse points P2 in the mercury trough Do and thereby reduce the resistance of P 2 , causing excitation of the booster field to produce an E.M.F. to send charge into the battery. The essential difference between this form of regulator and other types is that the design provides for a condition of neutral equilibrium between the pull of the magnet and that of the spring for any position of the moving parts; that is, with a given current passing through S, the pull of the mag- net balances the pull of the spring in any position of the lever L, conse- INSTALLATIONS. 897 £uently, the change in the generator current with change in external load is not proportional to the load but is a fixed amount. This variation is just sufficient to cause such a change in the pull of the magnet that the resulting unbalanced force overcomes the friction of the parts. The lever will begin to move and will continue to move until the current through £ is restored to its normal value, which is accomplished by causing the bat- tery to absorb or discharge current equal to the difference between the normal generator current and that supplied to the external load. The change in the resistances, being made by the immersion of the small con- tact points in mercury, offers no appreciable opposition to the movement of the parts and thus allows a continuous condition of neutral equilibrium to be maintained throughout the travel of the moving parts. Obviously by providing externally controlled boosters with a single vari- able resistance, a non-reversible booster is produced, its action being in effect the same as that described under the heading "Non-Reversible Booster." Comparison of Boosters. Reversible boosters should be used where the average, total current to the fluctuating load is greater than the battery discharge current, and where the potential of the power bus bars must not fall off with increase in load. Electric railway and lighting plants having long feeders are examples of the systems to which reversible boosters are suited. Non-reversible boosters should be used where the average total load is less than the bat- tery discharge current, and where a drop in the potential of the power bus bars is of advantage. Examples of such plants are hotel or apartment houses where electric elevators are operated from the lighting dynamos. Boosters are usually driven by electric motors directly connected to them, though any form of driving power may be used. They are some- times operated by engines or turbines. Installations. Figure 19 shows diagram of connections and Fig. 20 the switchboard of a battery equipment for a residential lighting plant. In the diagram, the voltmeter and voltmeter connections have been omitted. The bus bars on the battery panel are connected directly to the bus bars on the gene- rator panel. In this installation the generators are run during the after- STARTING BOX l|l|l|l|l|l|l|l|l|l|l|l|l|l|l|l|(J Fig. 19. 898 STORAGE BATTERIES. +" DIFFERENTIAL- AMMETER 1 [I ft -4- VOLTMETER SWITCH C L2 CHARGE AND DISCHARGE SWITCH BATTERY SWITCH J EI OJ BOOSTERFIELD SWITCH BOOSTERFIELD RHEOSTAT MOTOR STARTING BOX BATTERY CIRCUIT BREAKER MOTOR SWITCH END CELL SWITCH UNDERLOAD CIRCUIT BREAKER Fig. 20. noon, charging the battery and supplying the load. When the battery is fully charged the generators are shut down and the battery carries the load alone. In this manner the plant gives continuous service, while the generators are run only from five to nine hours per day. The bus bar voltage remains constant at all times, the battery voltage INSTALLATIONS. 899 on discharge being regulated by means of an end cell switch. On charge, the E.M.F. above that of the bus bars, required to bring all cells up to full charge, is supplied by means of a motor driven charging booster, the voltage at the armature being suitably varied by changing the field excitation. Figure 21 shows diagram of connections arranged for charging the bat- tery in two parallel groups and discharging in series, the charge and dis- charge being controlled by variable resistances. In yacht lighting the limited space generally prohibits the use of a charging booster, and in such instances this method of charge and discharge control is the usual practice. In case the generator from which the battery is charged has sufficient range in voltage to charge all cells in series, a charging booster is not AMMEXEfl SHUNT AMMETER SHUNT VARIABLE J RESISTANCE 6 6 n Fig. 21. required, nor is it necessary to connect groups of cells in parallel, as the generator voltage may be varied as charge proceeds. The diagram shown in Fig. 22 permits of charging the battery at one voltage and supplying lights at a different voltage. As may be seen, two end cell switches are required for this plant. The voltage of the supply circuit is adjusted by the number of cells in series on switch S 2 , while 83 is moved to cut out cells as they become fully charged. In this instance the end cells included between the contact arms of the two end cell switches must be of sufficient size to receive the charging current, plus the current to the supply circuit. , If the battery can be charged at times when the generator is supplying no other load, only one end cell switch is required. Figure 23 shows a diagram of connections for a constant current booster system, in which the same generators supply a lighting and a power load, the battery being connected directly across the power bus bars. The diagram further provides for the battery to supply lights at such times as the generators may be shut down. Three-Wire Systems. In three-wire systems it is usual to put in two equipments, one on each side of the system. Figs. 24 and 25 show the general schemes of two different three- wire systems; the one shown in Fig. 24 consists of a com- flete battery equipment and charging booster on each side of the system, n this diagram the generators are connected to the outsides of the line, the neutral being taken from the battery. This makes a good arrange- ment. One side of the battery system will discharge a sufficient current to take up any unbalanced load. 900 STORAGE BATTERIES. Figure 25 is a battery three-wire system in which only one booster is used. The main battery is charged from the outsides of the system, and the booster forms a local circuit of the end cells and gives them the proper charge; the voltage of the system being high enough to charge the cells is the main battery. In the boosters shown in these diagrams the arma- tures only have been indicated, as in nearly every instance boosters on three-wire systems are merely charging machines, the fields being sepa- rately excited from the bus bars or from the battery. , Figures 26 and 27 show clearly the switchboard connections of a central station battery working on three-wire systems. It is obvious that the systems would work just as satisfactorily if the generators were of a poten- - C B - ^1! Ill I III I III I III I CIRCUIT BREAKER .-•«3 6 o-~ o o i — ZJ oooooooo 0S3 066666666 END CELL SWITCH s 2 Fig. 22. tial equal to that of the outsides, and connected directly across the system, as any unbalancing would be taken up by the batteries. Battery Capacity. In computing the capacity of a battery to give a certain discharge, it is necessary to take into account the fact that the capacity of a battery varies greatly with the rate of discharge. This variation in capacity can be computed from the curves, Fig. 1. Taking the eight-hour rating as a basis, it is seen that only 50 per cent of the ampere hour capacity is avail- able at the one-hour rate of discharge. Therefore if 200 ampere hours be required at the one-hour rate, the normal ampere hour capacity must be _ _ = 400 ampere hours. In a like manner the normal capacity required for any other rate may be obtained. In the case of a load curve such as BATTERY CAPACITY. 901 Fig. 23. 902 STORAGE BATTERIES. that shown in Fig. 7, when the peak dEe is to be carried by the battery, it will be seen that the rate of battery discharge changes continually. If the area of the peak be taken above the line of generation supply. i h Sbtl _ Switches C M A C M A C MA End Cell H jV Switch N'o. l T Amnieterel __ Battery Fig. 26. ± Dynamo Bus' Fig. 27. discharge. The portion of the load peak to be carried by the battery is divided into vertical divisions, as indicated by the dotted lines. The ampere hours of each strip, divided by the rate of discharge factor (from curves, Fig. 1), gives the ampere hours capacity, on a basis of the normal rate, required for that particular strip. The sum of all these capacities 904 STORAGE BATTERIES. must be the capacity of the proper battery. If the assumed figure be toe small or too large, a second computation must be made, based on a capa city again assumed, which is greater or less than that just taken according as the result of the first computation is too small or too large For instance, if peak E be divided vertically into areas V, W, X, Y, and a 900 ampere hour battery assumed as the proper size, the normal rate of discharge will be 112.5 amperes. The ampere hours of area V are 75, and the average discharge rate is 210 amperes. Dividing 210 by the amperes of normal discharge, the result is 1.86. Locating 1.86 on the right-hand scale of curve, Fig. 1, and moving horizontally to curve No. 2, and then downwards to the lower scale, it is seen that this corresponds to the 3-i— hour rate. The percentage of the normal capacity at the eight-hour rate, when the dis- charge takes place at the 31 -hour rate, is shown by curve, Fig. 1, to be 76 75 per cent. The capacity required to cover strip V then is -=— = 99 ampere hours. Similarly the ampere hours of strip W are 193, the rate of dis- 340 charge 340 amperes, the factor = 110 K = 3.02 corresponding to the 1£- 112.5 Percentage of eight-hour capacity, .58, and ampere hours = hour rate. In a like manner, the capacity required for area X is 269 ampere hours, and for Y is 237 ampere hours, the sum being 938 ampere hours. The assumed capacity is therefore nearly correct, and a 950 ampere battery will be the proper size in this case. If the battery is also to be used for supplying the light load from 11 p.m. to 5 a.m., the capacity must be computed from the area h, k, m, n, which is 990 ampere hours. The rate of discharge is fairly constant, and extends over six hours. The percentage of normal capacity available at the six- hour rate of discharge is 94 per cent. 990 -q2 = 1050 = ampere hour capacity of battery required to carry the load given from 11 p.m. to 5 a.m. Strength of Dilute Sulphuric Acid of Different Densities at 15° C. (59° JF.) (Otto.) Per Cent of H 2 S0 4 . Specific Gravity. Per Cent of SO s . Per Cent of H 2 S0 4 . Specific Gravity. Per Cent of S0 3 . 100 1.842 81.63 23 1.167 18.77 40 1.306 32.65 22 1.159 17.95 31 1.231 25.30 21 1.151 17.40 30 1.223 24.49 20 1.144 16.32 29 1.215 23.67 19 1.136 15.51 28 1.206 22.85 18 1.129 14.69 27 1.198 22.03 17 1.121 13.87 26 1.190 21.22 16 1.116 13.06 25 1.182 20.40 15 1.106 12.24 24 1.174 19.58 14 1.098 11.42 Ordinarily in Accumulators the densities of the Dilute Acid vary between 1.150 and 1.230. CONDUCTING POWER OF ACID. 905 Conducting: Power of Dilute Sulphuric Acid of Various Strengths. {Matthiessen.) Sulphuric Relative Specific Acid in Temperature. Resistances. Gravity. 100 parts C.° Ohms per by Weight. cubic centimeters. 1.003 . 0.5 16.1 16.01 1.018 2.2 15.2 5.47 1.053 7.9 13.7 1.884 1.080 12.0 12.8 1.368 1 . 147 20.8 13.6 .960 1.190 26.4 13.0 .871 1.215 29.6 12.3 .830 1.225 30.9 13.6 .802 1.252 34.3 13.5 .874 1.277 37.3 . . . .930. 1.348 45.4 17.9 .973 1.393 50.5 14.5 1.086 1.492 60.6 13.8 1.549 1.638 73.7 14.3 2.786 1:726 81.2 16.3 4.337 1.827 92.7 14.3 5.320 1.838 100.0 • • • . . . Conducting- Power of Acid and Saline Solutions. Copper (Metallic) at 66° F 100,000,000. Sulphuric Acid 1 Measure \ Water 11 Measures ( QC n annY . n ^ r : rn „*. , (Equal to 14.32 parts by weight of Acid in 100 ( y *' U approximate. parts of the mixture), at 66° F / Sulphate of Copper, saturated solution at 66° F. 6.1 Chloride of Sodium, saturated solution at 66° F. 35.0 Sulphate of Zinc, saturated solution at 66° F. . 6.4 SWITCHBOARDS. Revised by H. W. Young, B. P. Rowe and E. M. Hewlett. The object of a switchboard is to collect the electrical energy in an installa- tion, for the purposes of control, measurement and distribution. In small stations this is accomplished by concentrating the energy at a single place. In the large modern stations this is often impractical, and it is, therefore, customary to concentrate only the control and measuring apparatus. There are two general types of switchboards: (1) IMrect-Control Panel Switchboards, in which the switching and measuring apparatus is mounted directly on the switchboards. (2) Remote-Control Switchboards, in which the main current carrying parts are at some distance from the contro^ng and measuring apparatus. This type may again be divided into two ivisions, viz.: hand- operated remote-control, and power-operated remote-contro! apparatus. The best modern power-operated apparatus is electrically operated, although there are a few installations which have employed compressed air. The above general types may both be sub-divided into Direct-Current and Alternating-Current Switchboards, and there are numerous and distinct classes in each subdivision. It is customary to mount apparatus and switching devices for low-tension service up to and including 750 volts directly on the face of the switchboard panels unless provided with suitable insulating covers or is out of reach of the operator. If the plant is of small capacity, the switching devices and conductors may be provided for on the rear of the panels. Heavy capacity plants from 2200 to 6600 volts, however, are invariably remote control, and nearly always electrically operated. In all high-tension plants from 6600 to 33,000 volts the switchboard is in- variably remote control, and if of heavy capacity it is invariably electrically operated. In large stations, for pressures above 33,000 volts, switchboards are invariably electrically operated remote control. In small capacity installations where the high-pressure service consists of only one or two incoming lines, which will not warrant expensive remote control switches, a set of simple fused circuit breakers or expulsion fuses are often installed and a switchboard dispensed with. Cut out switches are used, however, in addition, for disconnecting the lines. Design of Direct-Control Panel Switchboards. — In de- signing buildings for control stations or isolated plants, the switchboard should be located in an accessible place, with plenty of room in front and rear. If care is taken in locating the various panels with respect to the machines and feeders to be controlled, much unnecessary expense and com- plication may be avoided. If extensions to switchboards are expected, which is usually the case, panels controlling generators should be together at one end of the switch- board, and those controlling feeders at the other end. When total output panels are used, they are placed between the generator and feeder sections. It is advisable, however, in some special cases, in order to save copper in the busses and simplify the station wiring, to intermingle the generator and feeder switches although even in this case it is desirable to group the gen- erator indicating devices together and likewise those of the feeders. Unnecessary complications and extra flexibility being at the expense of simplicity are always to be avoided. For instance, in a majority of cases it would seem unnecessary to provide more than one set of bus bars. Plainness, neatness, and symmetry in design should be aimed at, and nothing placed on the switchboard which has no other function than orna- mentation. Sufficient indicating and recording instruments should be used to deter- 906 SWITCHBOARDS. 907 mine if the machines are working efficiently, to obtain a record of the output of the feeders, to detect external or internal troubles, and to check with records obtained from outside sources. The degree of accuracy required in the switchboard instruments depends entirely upon the conditions involved, greater accuracy being required where power is bought or sold. Instru- ments which are accurate to within 2 per cent of the full scale deflection will generally fulfill all requirements. Switchboards are now standardized, covering a large range of requirements, and standard panels are advisable for general use, although special conditions may usually be met with small modifications of the standards. For ordinary direct-current switchboards, 4 feet is little enough behind the panels. In any case there should be a clear space between the connec- tions on the panels and the wall of 2\ to 3 feet. For heavy direct-current work and most alternating-current work it is often necessary to have 6 to 8 feet behind the panels. Hand-control panel switchboards may not be advisable in direct-current stations where capacities are large, and in such cases remote-control installa- tions should be considered. It is likewise inadvisable to design switchboards of this class for heavy capacity alternating-current circuits of 2200 volts or upward, as the conductors for such service should be specially isolated. It should be noted especially that heavy capacity conductors arid switch- ing devices for circuits of 4000 alternations and above should be avoided, on account of excessive heating" to be met with due to eddy currents in the conductors. It is doubtful if satisfactory switching devices can be easily procured, which will carry currents of more than 3000 amperes at 7200 alternations or the equivalent, and such devices require special design and expense. In locating switching apparatus it is usually assumed that dynamo leads come up from below, and feeder wires go out overhead except that under- ground feeders naturally. go out below. In order to avoid a very unsightly complication of wiring and apparatus on the rear of switchboards, it is best to locate series and voltage trans- formers apart from the switchboard on the incoming and outgoing cables, if at all possible, and to make all large rheostats operate with sprocket and chain, thus locating the rheostats separate also. Any extensive system of fuses to be supplied on the rear should preferably be provided for on a sepa- rate framework. The material from which panels should be made varies with the service. Plain slate can be used for any panels where the potentials are not above 1200 volts. This slate may be either plain, or oil filled, or it may be given a black finish. The black enamelled slate is very satisfactory for use where oil is prevalent, but it shows scratches easily, and is not easily repaired if chipped. The most popular finish is the natural black oil finish slate, which may be made oil proof, and is a durable dead black. It is easily replaced when damaged. For switch bases and panels not requiring finish, soapstone is often used as it is a better insulator than slate, the latter being liable to contain con- ducting veins. Such slate should be rejected. Marble is largely used for switchboard panels because of its good insulating qualities. Many varieties are available, the most common being the white Italian, pink or grey Tennessee, and several varieties of blue Vermont marble. The colored marbles do not show oil stains as readily as the white varieties, and present a more pleasing appearance. The blue Vermont marbles are more uniform in coloring, and therefore easier to match; but if absolute uniformity in this respect is desirable, it is advisable that all panels be given a black marine finish, as it is often difficult to get new panels with exactly the same shades and markings as those it is desired to match, marble being a natural product. Standard Central Station switchboard panels are commonly made 90 inches high, and composed of two or three slabs. The upper slab of a two-piece panel is usually from 60 to 65 inches high, the lower one being from 25 to 30 inches high. The General Electric Company's three-section panels are upper and middle sections 31 inches each and the lower section is 28 inches, the corre- sponding Westinghouse standard being 65 and 25 inches. The Westinghouse three-piece panel has an upper slab 20 inches high, middle slab 45 inches high and lower slab 25 inches, the 20-inch slab being provided primarily to permit circuit breakers to be directly mounted thereon, and allow of easy removal in case of substitution or repairs. ( 908 SWITCHBOARDS. The General Electric Company also makes panels of any sizes up to 48 inches high and I5 inches thick for isolated plants. Panels 48 inches high are mounted on 76-inch pipe supports, the Westinghouse standard for similar service being 48 inches high and 1£ or 1£ inches thick as required. Each panel is beveled I to \ inch all around the front edges, the dimen- sions being measured from the edges of the panel, and not across the face of the level. Switchboard frames for very heavy panels are often made of channel iron tees or I-beams. The Central Station Switchboard frames are made of steel Fig. 2. Method of Joining Ad- jacent Panels. Fig. 1. Fig . 3 . Channel Foot for Switch- board Frame. angle bars varying from 2J X U X i inches to 3 X 2 X i inches or lj-inch gas pipe. The angle bars are supported in an upright position on a level strip which rests on the floor. This may be of slate, an inverted channel iron, or a hardwood plank. The panels are bolted to the narrow web of the angle bars and the adjacent angles bolted together through their wide webs. (See Fig. 2.) Another method used with panels which carry a moderate weight of appa- ratus is to make a frame of iron piping, secured to the panels by means of suitable iron supporting clamps. SWITCHBOARDS. 909 The framework of all switchboards should be insulated from ground when used on systems of 600 volts or less. In high-tension alternating-current systems, it is necessary to ground all framework to carry off static dis- charges, and to insure safety to the operator, should he touch the frame- work. For securing the frame in a vertical position, rods are used with or without turnbuckles, or else angle iron braces. As a general thing, alternating and direct-current panels should never be intermingled, especially when this involves the mingling of conductors on the rear. It is recommended that illuminating lamps be omitted from the front of switchboards, and that the instruments be illuminated by lamps in front of the same. The copper bars and connections on the rear of switchboards should be ^Gasppe . i A T A £* c/ter oas °° r supports -^p Supports for * rnoQstab pvnen moan teat on tQpo/ia/. ■+-lJfGcrspJpG Fig. 4. Showing Method of Bracing Switchboard Panel to Wall. >. Showing Gaspipe Framework. carefully laid out in order that the current may be carried economically and without overheating, and especially to prevent undue crowding and insure a neat and workmanlike appearance. The best practice requires that bus bars be not placed near the floor. Switches, circuit breakers and other apparatus are connected up with bare copper strap or insulated wire as occasion requires, bent in suitable forms. Where bus bars are not rigidly supported, it is not recommended, as a rule, to have long studs on the appa- ratus, projecting out far enough to connect to the busses, as the strain on the apparatus due to the weight of the busses may affect the adjustment of electrical contacts. Except for small switchboards the bus bars are usually supplied with insulated supports. Bare flat or round copper bars are now used almost universally for con- ductors on low-potential switchboards, the flat bar being usually preferred on account of ease in making connections and the facility with which addi- tional capacity may be provided for. The prevailing thicknesses vary from & to \ inches with widths proportioned to suit the capacity. The size of copper conductor is usually figured out on the basis of 800 to 1000 amperes per square inch of cross section. By properly laminating the bars, even verv 910 SWITCHBOARDS. heavy currents may be provided for on this basis. Contact surfaces should be figured on a basis of 100 to 200 amperes per square inch according to the method of clamping, bolting, or soldering. Steel bolts are used in clamping. Care must be taken, however, with alternating-current circuits to see that iron clamping plates and bolts do not form complete magnetic circuits and cause undue heating, due to eddy currents set up in the iron. Connections and apparatus for carrying current should be guaranteed to carry their normal current at a temperature rise not exceeding 30° C, above the surrounding air. Rolled copper should be used for conductors to secure the best conductivity, but it is often necessary to use copper or brass castings. As their conductivity is usually low, such materials should be avoided as much as possible. Where it is necessary to use castings they should be of new metal only and care should be taken to insist on a standard of conduc- tivity for each piece where such a condition counts. The ordinary mixtures vary from 12 to 18 per cent according to mixture. A conductivity of 50 per cent may be considered high and sufficient, but it is not obtainable in a regular brass casting. The following table from "Modern Switchboards," by A. B. Herrick, gives percentages of mixtures with resulting conductivity as compared with 100 per cent copper: % % Conduc- % % Conduc- Copper. Zinc. tivity. Copper. Tin. tivity. 98.44 1.56 46.88 98.59 1.41 62.46 94.49 5.51 33.32 93.98 6.02 19.68 88.89 11.11 25.50 90.30 9.70 12.19 86.67 13.33 30.90 89.70 10.30 10.21 82.54 17.50 29.20 88.39 11.61 12.10 75.00 25.00 22.08 87.65 12.35 10.15 73.30 36.70 22.27 85.09 14.91 8.82 67.74 32.26 25.40 16.40 83.60 12.76 100.00 27.39 100.00 11.45 All minor connections to bus bars such as switch leads, feeder terminals, or any attachments whatsoever, whether clamped, bolted or soldered, should have ample contact surface contact rated at 100 amperes per square inch, and all round conductors should be cup-soldered to flat lugs leaving proper amounts of contact surface. Cup-soldered connections should enter the sockets from two to three diameters. All permanent joints of this nature should be soldered, as required by the National Board of Fire Underwriters. Where it is essential to leave a joint that may be easily disconnected, the old style sleeve or socket with binding screws can be used, but the connections should enter from four to ten diameters to make a secure connection. An exceedingly clever device to take the place of the connection referred to or to use in place of cup-soldering is the Dossert joint which is quickly and easily applied to the end of a wire or cable, and is so designed as to insure the full conductivity of the conductor to which it is applied. The tables given below furnish the electrical constants of copper and aluminum bars which are most likely to be of use to the switchboard designer. The current which any given section may carry is calculated upon the basis of a load factor of 50 per cent, and the densities given are those which for average conditions of radiation would result in a temperature rise of about 10 degrees Centigrade. Where the load factor is to be 100 per cent, and it is desired to keep the heating within the above limits, the current densities must be halved. The data given show in an interesting manner the relative values of copper and aluminum in switchboard construction. COPPER BAR DATA. 911 Copper Bar Data. The Cutter Company. Size. Amps. Amps. per Square Inch. Circular Mils. Square Mils. Ohms per Foot. Weight per Foot. 1 X i in. ... 433 1732 318,310 250,000 .0000336 .97 H X i in. . 530 1696 397,290 312,000 .0000269 1.21 UXi in. . 626 1669 477,465 375,000 .0000223 1.45 HXi in. . 725 1657 556,400 437,000 .0000192 1.70 H X t in. . 676 1442 596,830 468,750 .0000179 1.82 HXi in. . 798 1418 716,200 562,500 .0000149 2.18 If X f in. . 916 1395 835,600 656,250 .0000128 2.54 2 X 1 in. . 1035 1380 954,930 750,000 .0000112 2.92 2i X f in. . 1154 1367 1,074,300 843,750 .00000995 3.27 24 X 4 in. . 1500 1200 1,591,550 1,250,000 .00000672 4.86 2J X f in. . 1715 1097 1,989,440 1,562,500 .00000537 6.07 2 X 4 in. . 1222 1222 1,273,240 1,000,000 .00000840 3.89 No. 0000 B. & 1 :J. 267 1606 211,600 166,190 . 0000505 .64 4 in. round . 305 1552 250,000 176,350 . 0000428 .76 | in. round . . 426 1388 390,625 305,796 . 0000273 1.18 f in. round . . 560 1267 562,500 441,787 .0000190 1.71 1 in. round . . 861 1096 1,000,000 785,400 .0000107 3.05 Aluminum Bar Data. The Cutter Company. Size. Amps. Amps. per Square Inch. Circular Mils. Square Mils. Ohms per Foot. Weight per Foot. 1 X i in. ... 347 1388 318,310 250,000 .0000534 .291 ] HXi in. . 424 1360 397,290 312,000 .0000428 .362 HXi in. . 500 1334 477,465 375,000 .0000356 .435 If X i in. . 580 1327 556,400 437,000 .0000305 .507 H X f in. . 530 1131 596,830 468,750 . 0000285 .544 14 XI in. . 638 1130 716,200 562,500 . 0000237 .653 If X f in. . 2 X I in. . 733 1117 835,600 656,250 . 0000203 .762 830 1107 954,930 750,000 .0000178 .871 2i X f in. . 925 1096 1,074,300 843,750 .0000158 .980 24 X 4 in. . 1200 960 1,591,550 1,250,000 .0000107 1.45 24 X f in. . 1400 897 1,989,440 1,562,500 .00000855 1.81 2 X 4 in. . 980 980 1,273,240 1,000,000 .0000134 1.16 No. 0000 B. & 3. 211 1266 211,600 166,190 .0000803 .193 4 in. round . 244 1260 250,000 176,350 .0000680 .228 f in. round . , 340 1108 390,625 305,796 .0000436 .355 f in. round . , 448 1013 562,500 441,787 .0000302 .513 1 in. round . . . 690 880 1,000,000 785,400 .000017 .911 912 SWITCHBOARDS. Circuit breakers, if required to open circuits carrying heavy loads, should be mounted at the top of the panels to give the arc plenty of room to rise without scorching the instruments or the panel, and to keep it above the attendant's head. Instruments should be mounted below the circuit break- ers, while the lower portion of the panel should be utilized for switching devices. Switches, circuit breakers and fuses are usually rated at their maximum continuous ampere capacity and for this reason care should be taken in selecting these devices. Take into account the one hour, two hour and three hour overload guarantee on the machines. Indicating instruments should have scales calibrated to read in excess of the overload guarantee of the machines to which they are to be connected. It is usually good practice to have the needle about in the middle of the scale at normal load, but a good reading should be obtained as low as one quarter load. Meters affected by stray fields should be kept away from the influence of connections carrying heavy currents. Panel switchboards for small capacity stations for alternating-current circuits from 480 to 3300 volts are usually supplied with oil switches, mounted on the back of the panels, with handles for manual operation on the front. In large stations, however, these are usually replaced by remote- control switches. Insulation Distances. — In high voltage switchboard work where there are bare conductors, safe distances must be maintained between the conductors and from the conductors to the switchboard structure. The striking distance through air may be somewhat less than the distance over surfaces. The air distance should not be less than two and one half times the striking distance of the given voltage as taken from the curve on page 462, and the surface distance should not be less than three times the air distance allowed for the given voltage. It is obvious that the greater the distance the greater the factor of safety ; and in large capacity stations this greater factor of safety is usually advisable on account of the greater insur- ance given by the use of greater distances. The creepage distance to be maintained in the switchboard depends upon many conditions some of which are: The material of the surface; the con- tour of the surface; the liability to collect dust and the properties of the dust; and the amount of moisture in the atmosphere. aheh^itogcirheit switchboard The instruments, switches, etc., required for the various types of panels are listed below, for assistance to the engineer when designing a switch- board. Each type of panel will be described individually. Equipment of 3-Phaie Generator Panels. 3 Ammeters (one is sufficient for practically balanced loads or may be connected by means of plugs or ammeter transfer switches, so as to read the current in either of the 3 phases) . 1 Voltmeter. 1 Polyphase indicating wattmeter. 1 Field ammeter. 1 Polyphase integrating wattmeter (optional). 1 Wattless component indicator or power-factor indicator (optional). The first instrument indicates the useless watts and the rheostat should be adjusted to reduce them to a minimum. The power- factor indicator is used for the same purpose, but does not give a direct indication of the idle currents at all loads. 1 Voltmeter switch for reading voltage on either of the 3 phases (on balanced systems this is usually omitted and voltmeter per- manently connected to one phase). 1 Synchronizing switch (one synchronism indicator can be used for all generators). ...-. „ ,. 1 Field rheostat with chain operating mechanism (small machines may have the rheostat mounted at the back of the panel). If electrically operated rheostats are used the handwheel would be replaced by a controlling switch. ALTERNATING-CURRENT SWITCHBOARD PANELS. 913 1 Field switch with discharge clips. 1 Discharge resistance for field circuit. 1 Non-automatic main switch (controlling switch required if oil switch electrically operated is used). 2 Current transformers (3 transformers are necessary if neutral oi generator is grounded). Potential transformers (3 potential transformers are desirable if neutral of generator is grounded, but one is required if used only for synchronizing). Both may be omitted on circuits of 600 volts and less, if all meters have their coils wound for operating at generator voltage. 1 Engine governor control switch if governor is electrically controlled . If each alternator has its own exciter the exciter may also be controlled from the alternator panel, by the addition of an exciter field rheostat. < -VOLTMETER AMMETER INDICATING WATTMETER IELD AMMETER -RHEOSTAT HAND WHEEL /CIELD DISCHARGE SWITCH VOLTMETER PLUG RECEPTACLE Fio. 6. 440- and 600- Volt Three-phase Generator Panel. Two-phase generator panels have a similar equipment to the three-phase accept that but two main ammeters, two current transformers and two potential transformers are required. 914 SWITCHBOARDS. &%% gKg — ftgi o3 o T3 C ALTERNATING-CURRENT SWITCHBOARD PANELS. 915 Ammeters f~ie/a//rr, meter- Vo/tmeter V%>ter>tia/ Y*ecejotac/e VfheosCot KO/oc rating \/Wecr>anisrr> ne/cf S*vrtc* Sync/ironiz T'ng \Kece0tocte- 4Af Oi/SwitcH Connections fbr&ig/ne Governor Contra/ Motor ana" Switch when supp/fecf Generator BazA Z/ewofGanet Fig. 8. Two-Phase 2300- Volt Generator Panel. '« -/tm meters-' ■ Vr~/e/cf /7mmet<3r Vo/tmeter \Potentfot \ftecejotac/e \Pheostat fieceptock Operating {Mechanism eic/ Switch ^Synchronizing Wece/otac/e ^ eent ' a A~fps£T_ T. Pi Oi/ Switch SwitchX Ammeter fir filternating\ Current \ Generator J _ &7CA V/eyvbffibnef J Starting \®' Synchronizing , Plugs ' Running Synchronizing Buses Ground Bus Connections forchpine GovernorControf ■ Motor ana 'Switch when supplied Fig. 9. Three-Phase 2300- Volt Generator Panela. 916 SWITCHBOARDS. Equipment of Single-Phase feeder Panel. 1 Main ammeter. 1 Compensating voltmeter (optional). As single-phase panels are invariably used for lighting it is necessary to maintain a constant potential at the point of distribution, and as each feeder circuit is likely to have a different load characteristic, potential regula- tors are frequently installed. The compensating voltmeter com- pensates for the ohmic drop or for both the ohmic and inductive drop in the line at all conditions of load and gives a direct indi- cation of the voltage at the center of distribution. 1 Potential regulator and operating mechanism (optional). 1 Main switch with automatic overload trip or automatic circuit breaker. 1 Current transformer. 1 Potential transformer if voltmeter is used. 1 Time limit overload relay (optional). 1 Single-phase integrating wattmeter (optional). Ammttert [Regulator Operotiryff \^tec^ar>>3rrt O0&3***** CeseJT I ToBusBars 7r/joCo/£^~Zl\ Xfiutomat/c Currerit\ \\\o/l Switch Tronsfo rr \/?mmeter QrouncfBos L Iff htntno Arresters. Notfurn.isheaw/t/i Pane/ Back y/ew^f Pons/ Fio. 10. 2500- Volt Single-Phase Feeder Panels with Primary Ammeters and with Series Trip Oil Switches. ALTERNATINO-CURRENT SWITCHBOARD PANELS. 917 Equipment of Three-Phase feeder Panels. Main ammeters for transmission lines used to detect any unbalancing due to leakage to ground. A single ammeter may be used if desired, with suitable plugs, to indicate the current in either of the three phases. (One ammeter is sufficient on feeders for induction motors and rotary converters, or on incoming lines in a sub-station.) Polyphase indicating wattmeter (optional). For power circuits in mills and mines. This wattmeter gives a sufficient indication of the output without the ammeters. Polyphase integrating wattmeter (optional). Oil break switch with overload trip, or automatic circuit breaker. Current transformers (three transformers are necessary if neutral of three-phase system is grounded). Potential transformers for wattmeters. Time limit overload relay (optional). The number of potential transformers can be reduced for a switchboard containing a num- ber of feeder panels by connecting two potential transformers to the busses and feeding all the wattmeters. r Q it 6Z 1. i 1 | LaT-J TfiO// Switches ,-_ac*_«» L-24'STSwt- _J L/^htninp/ffresters. Not furnished w/tr) Pane/ BacA V/'ewof&we/ Fig. 11. 2500- Volt Three-Phase Feeder Panels with Primary Ammeters and Series Trip Oil Switches. 918 SWITCHBOARDS. Equipment of Two-Phase feeder Panels. 2 Main ammeters. 1 Polyphase indicating wattmeter (optional). 1 Polyphase integrating wattmeter (optional). 1 Oil break switch with overload trip, or automatic circuit breaker. 2 Current transformers. 2 Potential transformers for wattmeters. 1 Time limit overload relay (optional). The number of potential transformers can be reduced for a switchboard containing a number of feeder panels by connecting two potential transformers to the busses and feeding all the wattmeters. Equipment of Induction Motor Panels. 1 Ammeter. 1 Oil break switch with overload trip, or automatic circuit breaker. 2 Current transformers. 1 Time limit overload relay (optional). The various methods of starting induction motors are as follows: 1. By Connecting* them Birectly to the Line. — This is sel- dom done except on motors under 10 horse-power capacity, because it pro- duces variation in the bus voltage unless the busses have considerable energy back of them. 2. By Inserting' an Internal Resistance in the circuit of the motor by means of a switch on the motor shaft. 3. By Introducing* an External Resistance in the rotor cir- cuit through collector rings. This resistance is cut in or out by a controller. 4. By first Connecting* the Motor to Eow-Voltag-e Taps. — If the motor is fed from step-down transformers, it may first be con- nected to low- voltage taps on the transformer and then to the full- voltage connections. 5. By Employing* a Starting* Compensator. — Many compensa- tors have an internal switch for starting; otherwise the panel should be pro- vided with switches to connect and disconnect the compensator. J LP»wL fndkjct/on MoCa* Fig. 12. 2080- Volt Induction-Motor Panel for Controlling Motors having an Internal Resistance. ALTERNATING-CURRENT SWITCHBOARD PANELS. 919 Equipment of Three-Puase Synchronous Motor Panels. 1 Ammeter. 1 Three-phase indicating wattmeter. 1 Field rheostat with operating mechanism. 1 Synchronizing switch. (The synchronism indicator will answer for any number of motors or the generator synchronism indicator may be used.) 1 Main oil switch with automatic overload trip. 1 Field switch with discharge resistance. 2 Current transformers. 2 Potential transformers. 1 Time limit overload relay (optional). A synchronous motor driving a direct-current generator can usually be started from the direct-current side, in which case the synchronizing switch is necessary. If always started as an induction motor the synchronizing switch is unnecessary. The equipment of a two-phase motor panel is the same as for a three- phase, except that two ammeters should be used. Equipment of a Xnree-Phase Rotary Converter Panel. For rotary converters connected in the high-tension side of step-down transformers, the panel for the alternating-current side is the same for three- phase or six-phase machines. 1 Three-phase integrating wattmeter (optional). 1 Ammeter. 1 Power factor meter. 1 Main oil circuit breaker with automatic overload trip. 1 Synchronizing switch (not necessary if rotary is started from the alternating-current side) . 1 Starting motor switch (only used where rotary is started by a starting motor) . 1 Switch for synchronizing resistance (only used where rotary is started by a starting motor). 2 Current transformers. 1 Potential transformer (if rotary is started from the direct-current side or by a starting motor). 1 Time limit overload relay (optional). One method of starting a rotary converter is by connecting the alternating- current side first to fractional voltage taps on the transformers, and then to full-voltage connections. This is accomplished by means of a double-pole, double-throw switch on a separate panel for a three-phase converter, and two triple-pole, double-throw switches on a separate panel for a six-phase conver- ter, Fig. 13. Another method is by the use of a motor on the rotary shaft, as shown on diagram, Fig. 14. The rotary may also be started from the direct-current side. In either of the latter cases it is necessary to synchronize. In case several rotary converters must operate from the same bank of transformers, it is best to have a separate set of secondaries for each rotary. But in case of rotaries which must be parallel on the alternating-current side under such a condition, it is essential that reactances be provided in the circuits to prevent interchange of current between machines, and that switches be provided in the alternating-current leads. These are used as main switches in synchronizing and are usually mounted on the alternating- current panel. For the condition just described, the panel would contain the same list of apparatus mentioned above, except that these switches 920 SWITCHBOARDS. cruses ^LPotent/a/ ' \7ransfbrmer \\b/tmeter\ x \/lmmeter j ~J Main _^— Buses Cou/2//r?& Potentio/ Transformer BuS. t ^Resistance j fleceptac/e »J 4 \Transformer -Lv ^Grounded •Switch Current Transformer Synchronizing Buses Synchronizing Mug *iH* * *AA\. I WW SA> I fteact/ve Co// Two TP. P.F Switches Synchronizing Connections {Shown Dotted) forrfotaries Starting /corn D.C. £nd i0 5 >* 1 To/Vext Transformer fuses To B/ower Motor * f?otary Converter* iflGh 13. Diagram of Connections. Three-Phase Rotary Converter started directly from Alternating-Current Side. ALTERNATING-CURRENT SWITCHBOARD PANELS. 921 would be substituted for the automatic overload main switch and relay, and either automatic protection provided in connection with the switches or else a fuse in each lead. When a rotary converter must run inverted (i.e., to convert direct current FROM TRANFORMER3 III OB 2 FUCE 6 PT.6YN. RECEP. • F.M.J ( VM. J Fio. 14. Diagram of Connections. Three-Phase Rotary Converter with Starting Motor. into alternating current), provision must always be made for starting the rotary from the direct-current side. Two-Phase Itotarr Converter Panels are essentially the same, but require two ammeters instead of one, and four-pole alternating-current switches. 922 SWITCHBOARDS. Equipment of Constant- Current Transformer Panels, for Series Arc or Incandescent liig-hting*. The primaries of these transformers may be controlled by an oil switch, with automatic overload trip, or by plug switches and fuses. The secondaries, being of small capacity, are usually controlled by plug switches. An ammeter should be connected in the secondary side to indicate the current and to detect grounds or open circuits. An integrating wattmeter on the primary side is a valuable adjunct to record the total power consumed. The diagram shown is that of a single- circuit transformer. Various modifications result from using multi-circuit transformers and introducing transfer systems in either the primary or secondary side. Open Circuiting Plug Jtritcnes^ Primory Aug Suvitcnes Tube Expulsion fuses Lightning Arrester Open Circ ult ing f/ug Quitches Constant. Current Transformer ffi Primary Plug S-kthii \ Current Transformer" when AHecCJSOryj Seconoory £Z- union Couplings AotenUof . Transformer Primary Fig. 15. Constant-Current Transformer Panel for Single Circuit. ARC SWITCHBOARDS. This line of switchboards represents an entirely different construction from that of ordinary switchboards. Extra flexibility makes it desirable, and small currents make it possible, to use plug connections instead of the ordinary type of switches. The function of arc switchboards is to enable the transfer of one or more arc light circuits to and from any of a number of generators. This trans- ferring is sometimes accomplished by means of a pair of plugs connected with insulated flexible cable; sometimes by plugs without cables, which bridge two contacts back of the board, or by a combination of cable plugs and plugs without cables. The type using plugs without cables is preferable because danger is eliminated, which would otherwise be possible to attendant, due to contact with exposed or abraded cables carrying high-potential current. The accompanying illustration shows an arc switchboard of the General Electric panel type, arranged for three machines and three circuits. The vertical rows of sockets are lettered and the horizontal numbered. The ends of the vertical bars are connected to the machines and circuits. Each of the bars is broken in three places, and the machine may be connected to its circuit by plugging across these breaks, thus making the bar con- tinuous; by removing any pair of plugs the machine may be disconnected. Cll, Ell and Gil are ammeter jacks, and are used in connection with two plugs connected with a twin cable, for placing an ammeter in the circuit. The six horizontal bars are for the purpose of transferring a machine or a feeder to some circuit other than its own. Each horizontal bar is provided, at one side of the panel, with a socket (A3, A4, A5, A7, A8, and A9) by means of which it can be connected with the horizontal bar on the adjoining panel. All ordinary combinations can be made by means of the bars and ARC SWITCHBOARDS. 923 plugs; but cable plugs are provided with each panel, so that when necessary, machines and feeders can be transferred without the use of the bar. These plugs and cables are intended for use only in case of an emergency. To run machine No. 1 on feeder No. 1, insert plugs in BIO, CIO, B6, C6, Fig. 16. B2, and C2. To shut down machine No. 2, and run feeders Nos. 1 and 2 in series on machine No. 1, insert a plug at C5, D5, C7, and D7, and remove plugs at C6 and D6; this leaves two circuits and two machines in series. Short circuit machine No. 2 by inserting the plug at E7. Cut out machine No. 2 by removing the plug at D10 and E10. Take out plug at D7. 924 SWITCHBOARDS. DIRECT-CrRRE^T SWITCHBOARD PAIEIS, Equipment of I».C. Generator Panels. 1 Overload circuit breaker. 1 Ammeter. 1 Voltmeter switch. (One voltmeter will answer for all generators.) 1 Field switch with discharge resistance (optional). 1 Positive main switch. 1 Negative main switch. (For railway service where the generator series coils are on the negative side, and the negative side is grounded, this switch should be replaced by a circuit breaker mounted near the generator, and connected in the armature lead.) 1 Equalizer switch. (Mounted near the generator. For small capacity generators all three switches may be combined into a triple-pole switch mounted on the panel.) 1 Field rheostat. 1 Recording wattmeter (optional). For small machines, fuses may be substituted for the circuit breakers. ^—POWER FACTOR METER SYNCHRONIZER LAMP SYNCHRONIZER PLUG RECEPTACLE AMMETER PLUG RECEPTACLES RHEOSTAT (IF NOT MOUNTED, "ON D.C. PANEL) SWITCH FOR SYNCHRONIZING, RESISTANCE SWITCH FOR STARTING MOTOR Equipment of A.C. and D.C Rotary Converter Panels. The equipment of a direct-current converter panel may be the same as i direct-current generator panel, but a field switch with discharge resistance is unnecessary and the cir- ^ammet cu jt breaker in the nega- tive on grounded return system should be omitted as the necessary protection is secured on the alternat- ing-current side. The main switches, however, should all be single pole. Rotary converters started from the alternat- ing-current side may build up with reversed polarity, which will be indicated on the voltmeter. To change the polarity back to nor- mal, a double throw field switch is provided (usually mounted on the converter frame) for the purpose of momentarily reversing the field to "Slip a pole." To reduce the destructive in- ductive discharge of the field a multi-pole switch is used, each pole of switch breaking only two or three field spools. Rotary converters oper- ating on grounded return systems may have the neg- ative side connected direct- ly to ground without the interposition of a switch. Rotary converters start- ing from the direct-current side require a field-transfer switch, as well as a starting Fig. 17. Three Phase Alternating Current switch, which are usually Rotary Converter Panel for use with provided with the direct- Rotary and Starting Motor. current panel. A double- reading ammeter is usually provided, or else other provision to prevent damage to the meter by reversal of current. DIRECT-CUKRENT SWITCHBOARD PANELS. 925 TYPE C CIRCUIT BREAKER RHEOSTAT iANDWHEEL NEGATIVE MAIN SWITCH POTENTIAL RECCPTACLE POSITIVE MAIN SWITCH RECORDING WATTMETER Fig. 18. Westinghouse Panel for D.C. Generator or Rotary Started b> Starting Motor. 62' t*. -/•>'!-. «£ Type C Form /< C/rca/t Breaker Ma/nBus Bar * TID rlmmetef* Potent/'o/BusW/rc Support i P/?eoJtat r/a/?c/tv/?ee/ ■Potent/a/Peceptac/e • Card r/o/der Prteastat C/>a/n \ _ Opcrat/'rg Mechan/sm] " -L/$f>t/n$ Smv/'£c/> ' type Q& Porm/4 Stv/tcS) ■Pccord/n$ Wattmeter Wattmeter Pes /-stance - Fig. 19. Direct-Current Rotary Converter Panels. General Electric Panel Rotary Started Direct from A.C. Side. 926 SWITCHBOARDS. Bock view Positive Bus law Vo/toge ffe/eose. — eovott Lamps /Pes/ stance - ■ +►§ — £^J voltmeter Fuse h \ Z-urszzzzz." fuses L ig?~> t mg Sw/'t en j\ To Center stud „1 j of Lighting 5w/ Left j i on adjacent Pone/ j % I i Fuse (J ! Station Lights Rheostat II IX- ) To Aiorm Belt Low voltage Release Bus ) Potential louses £ ^>eedUmtL Oev/CQ" ^//egat/ve Bu$ Qroyndeaf Fig. 20. Connections of a Direct-Current Rotary Converter Panel. Equipment of a Three-Wire (Generator Panel. The Westinghouse three-wire generator combines in its system of con- nections all of the circuits which were required for the usual generating sets of an Edison three-wire system, and a double equipment of apparatus is required, as follows: 2 Ammeters (operating from shunts located in armature leads of generator) . 2 Circuit breakers, each either two pole or supplied with equalizer contacts, to open a main and equalized lead (with operating coil in the main lead) to trip together. 2 Double-pole main switches. 1 Double-pole two-way voltmeter plug receptacle. 1 Field rheostat. 2 Double-pole balancing coil switches. (If the unbalanced load is to be measured, a double-reading direct-current ammeter should be placed in the neutral return.) The connections for such a system are shown in diagram, Fig. 21. DIRECT-CURRENT SWITCHBOARD PANELS. 927 928 SWITCHBOARDS. Equipment of D.C. Feeder Panel. Direct-current feeder circuits should be protected from overloads by cir- cuit breakers or fuses. Circuit breakers should be used if overloads occur frequently, such as on railway and most power circuits. They should also be used for all large ampere capacity circuits — say above 600 amperes. Small feeder circuits may be controlled solely by a double-pole circuit breaker, but on large circuits a switch in series with a circuit breaker is necessary. The equipment should then consist of: 1 Single-pole circuit breaker. 2 Single-pole switches. (On grounded return systems the second switch will be unnecessary.) Ammeters and integrating wattmeters are optional devices. Equipment of D.C. IfEotor Panel. 1 Double-pole automatic circuit breaker. 1 Starting switch and resistance, or 1 Single-pole automatic circuit breaker. 2 Single-pole switches or one double-pole switch. 1 Starting switch and resistance. or 1 Double-pole switch. 2 Inclosed fuses. 1 Starting switch and resistance. Ammeters are optional, but are recommended for motors of large sizes. Either the circuit breaker or the starting switch should have a low- voltage release' attachment. The starting switch and resistance should be so con- nected that the field, when the switch or circuit breaker is opened, will discharge through the armature. Starting switches for motors starting under heavy torque should have at least eight steps. Motor- genera tor sets may properly be started with but three or four steps. As the starting resistances are invariably designed for intermittent service, starting switches, except in power stations where an electrical attendant is in charge, should be provided with a spring or other means to prevent the switch arm from remaining on an intermediate starting point. Hand-Operated Remote-Control (Switchboards. — Wher- ever it is desirable to install a plant of moderate size and obviate the necessity of having any high potential conductors on the rear of the switch- board, a hand-operated remote- control switchboard may be installed. The panels will have the same appearance on the front as any other hand-operated alternating-current switchboard, but the rear of the panels may be made safe and accessible with a neat arrangement of small wiring, inasmuch as all heavy conductors, meter transformers and accessories are mounted apart from the panels. A common method of providing for the switches and transformers mentioned is to mount them on a separate framework in some distant place and control the switches from the switchboard by means of bell cranks, levers and connecting rods. These latter are usually made of gas pipe. The framework used to support the switches is usually utilized to support the bus bars also. As the connections between the panel board and the switching structure are made by small secondary wiring for meters and instruments, and the bell -crank attachments permit of an infinite variety of combinations, the location of the switching devices may be selected to best suit the station wiring so long as the cranks and levers can be arranged to operate suitably ; the total length of any set of bell cranks and levers should not, as a rule, be greater than 12 feet, although longer runs than this will operate successfully under favorable conditions. Central Station Electrically Operated Switchboards. — The concentration of energy in large central stations requires that the measur- ing and controlling devices shall be concentrated also, in order to be under the hand of a single operator and enable him to have absolute control of the whole installation. This end is best attained by the use of electrically operated switchboard apparatus. Electrically operated switchboards may be divided into two classes, namely, alternating-current and direct-current equipment. As large • central stations almost invariably generate alternating current for distribution, the electrically operated switchboard is usually of the latter class. ELECTRICALLY OPERATED SWITCHBOARDS. 929 Circumstances whicn Indicate the Wccesnity of Installing- Electrically Operated Switchboard Apparatus. First, switches used to control the circuits may be so heavy that they cannot be easily operated by hand. Second, the location of these switching devices can be made most conven- ient to the circuits to be controlled and apart from portions of the equipment which are liable to cause trouble, such as steam pipes, etc. Third, in case of accident to any of the apparatus, the operator may be located well away from the seat of trouble and is therefore not so liable to be frightened or lose his head in an emergency. Fourth, the entire absence of dangerous potentials at the center of control provides absolute safety for the operator. Fifth, the number of circuits and amount of power may be such that the control cannot be concentrated within a space of reasonable size unless electrically operated. Sixth, it may be necessary that the operator be located a long distance from the apparatus which he controls. Reliability of Service. — When the choice of an electrically operated switchboard is made, the next consideration is as to how much apparatus to install to insure reliability of service. It is possible to carry this idea to an unnecessary refinement in some cases, where the chances of a shut down are small and the consequences of it are not very disastrous. On the other hand there are some plants where no expense must be spared to provide against the contingency of a shut down even of a very short duration. The latter case requires much duplication of apparatus and great flexibility. Where a large number of feeders are used a circuit breaker is sometimes provided to connect between certain groups of feeders on the bus-bars, and is known as a group circuit breaker. Each feeder circuit of the group has its own individual circuit breaker to open automatically and relieve the group on the overload, but in an emergency the whole group can be switched on or off the circuit by means of the group circuit breaker. The value of this group circuit breaker for a single-throw system is doubt- ful except in cases where transfers of load must be very rapid and a large number of feeders are installed. It is more valuable in such a case on a double-throw system, because it enables the transfers from one set of bus- bars to the other to be made very rapidly and with a minimum number of switches, as one pair of circuit breakers will transfer an entire group of feeders instead of having two circuit breakers for each feeder circuit. There are four systems of connections for bus-bars commonly used. The first is the single-throw system, the second is the relay system, the third is the ring system, and the fourth is the double-throw system. Each of these may be made more flexible by dividing the bus-bars into sections by means of sectionalizing switches. Except in special cases it will be found that where any system is required to provide flexibility, the double-throw system will be most satisfactory. It is considered the best practice to provide disconnecting switches between all bus-bars and oil circuit breakers in order to permit a disabled switch to be isolated and repaired without shutting down the system. As the bus-bars form really the vital part of the system, it is necessary that care be taken to insulate them so that short circuits shall be impossible and that trouble on one set shall not communicate to another. Where absolute certainty must be insured against ^interruption of service, all conductors should be isolated from each other and all adjacent material made as fireproof as possible. In large stations this is attained by means of masonry structures and barriers and flame proof cables, with absence of inflammable material for supporting the cables, using cells for all fuses and apparatus liable to arc and all oil-insulated transformers that are so con- structed that danger from burning oil exists. This includes voltage trans- formers which are oil-insulated. The greater the energy involved the greater is the necessity for isolation, especially in plants of pressures under 45,000 volts. The isolation is most needed in heavy capacity stations of 2,200 to 13,000 volts and in some cases it is advisable up to 45,000 volts, if the use of compartments makes a more consistent layout. Isolation is, however, rarely advisable in stations above 930 SWITCHBOARDS. 45,000 volts, as small isolated conductors well supported in air will in such cases prove quite satisfactory, while barriers or adjacent walls usually serve a3 so many grounds to insulate from. Whenever modern practice reaches such a point that extremely high voltage circuits carry heavy current capacities, however, barriers may be advisable, but this condition is not liable to be met with. Fio. 22. 60,000-Volt Hydro-Electric Generating Station Sectional Elevation. 932 SWITCHBOARDS. BUS-BAR AND BUS-BAR STRUCTURES. 933 BO-BAR A1¥B BUS-BAR STRUCTURES. The bus-bars of a high-tension central station make up the backbone of the installation. As the entire distribution depends upon them, the design of the station as a whole should be executed with this fact in view. The bars should be entirely isolated from all danger from arcs, short circuits or flashes. All large stations should be laid out with a suitable arrangement of bus-bars, to guard against interruption of service from unforeseen causes and to pro- vide a means whereby circuits can be installed and connected with facility. The modern bus-bar structure for 2,200 to 33,000 volts is of brick or con- crete with each bus-bar of opposite potential in its own separate com- partment, well supported on porcelain insulators. The shelves or barriers in such a structure are usually of soapstone or con- crete. Some of these structures are enclosed entirely, one side having removable doors, while others are made with the entire side open for inspec- tion and facility in making connections and alterations. The bus-bars, being well protected and insulated, are usually composed of bare copper. For higher voltages than above mentioned a different form of bus-bar sup- port is generally used, and the connections to the bus-bars are made with wire or cable well supported on suitable insulators. Diagrams of a few typical arrangements of bus-bars and oil switches follow: Operae/s?pAfecfo/>/'s/f •Socpseone orStote *{4)+/4U4f+/4*iSpi/4i^4~ •e :=TJJT — p iA 3E - 1 H ° ^ i If E B 'WMMMMMXa TYPICAL ARRANGEMENTS OP WMJD VOLT BUS-BARS, ELECTRICALLY OPERATED OIL SWITCHES, AND DISCONNECTING SWITCHES IN THREE-PHASE STATIONS Fig. 25. 934 SWITCHBOARDS. fi « 3 2 BUS-BAR STRUCTURES. 935 General Arrangement of Switching- Devices. — In addition to the masonry required for the bus-bars, there must be provided struc- tures for the oil circuit breakers. The elements are contained in struc- tural work of brick or concrete. On account of this construction and the desirability of making connections between the apparatus in the most safe and direct manner, it is generally necessary to build structures in galleries Fig. 27. Double-Deck Oil Circuit Breaker and Bus-Bar Structure. Two Sets of Main Bus-Bars and Two Sets of Auxiliary Bus-Bars. one above the other, or if galleries are not to be considered, then a basement must be provided to take a portion of the gear. The simplest switchboards are usually double decked, while others require three or four galleries. For a given amount of apparatus, a double-decked arrangement requires the longest galleries and more material for bus-bars. It is the simplest, however, and often the most economical when the switchboard apparatus is located near the generators and transformers, and saves long and expensive lines of connecting cables. On the other hand, where the galleries must be small, a three-deck arrangement is more satisfactory. 936 SWITCHBOARDS. In each particular case the conditions of space, accessibility, etc., must determine the most suitable place for the structure and the best relative arrangement of the circuit breakers and bus-bars. The series and voltage transformers for the operation of the oil circuit breakers, meters, etc., in almost every care are placed in the structure, the best arrangement depending upon local conditions. Fig. 28. Double-Deck Oil Circuit Breaker and Bus-Bar Structure. Sets of Three-Phase Bus-Bars. Two Isolation of Conductors. — When barriers are used each conductor is confined to its own compartment and in case of accidental ground or short-circuit the flashing or combustion is confined to the conductor involved and prevented from destroying neighboring conductors. Barriers, while fire-proof, are not necessarily made of insulating material, although, were it not for the expense, they might well be made of such material. They are frequently made of brick, masonry, concrete, or tile, while in places where insulated barriers are desired, soapstone is the most favored material. It absorbs less moisture than marble, but the insulating BUS-BAR STRUCTURES. 937 properties cannot be depended upon. The cost is a little less. Soapstone is readily obtained in any reasonable size or shape, and is easily drilled and nut when fitting is necessary at the place of erection. When the barriers and compartments of the switchboard structure are made from any of the above-mentioned materials, they should be treated as Fig. 29. Three-Deck Oil Circuit Breaker and Bus-Bar Structure. Two Sets of Bus-Bars. grounds with reference to high-tension circuits. It is true that vitrified brick and concrete, when very dry, are more in the nature of insulators than conductors, but the tendency of all such materials, and even soapstone, is to absorb more or less moisture, preventing any absolute dependence bein* f)laced upon them as insulators, and all conductors must, therefore, be msu- ated from them. 938 SWITCHBOARDS. Each bus-bar is in a separate fire-proof structure, and each pole of the oil circuit breaker isan independent fire-proof compartment. Masonry barriers separate the leads from the oil circuit breakers to the bus-bars, and to the outgoing lines. Wherever it is desirable to use disconnecting switches between the circuit breakers and the bus-bars or circuit breakers and the / \ Oz/C/nzc'/f' Fig. 30. Three-Deck Oil Circuit Breaker and Bus-Bar Structure. Sets of Main Bus-Bars and Two Sets of Auxiliary Bus-Bars. Two outgoing lines on circuits not exceeding 13,000 volts, these disconnecting switches can be mounted as shown in Fig. 30, which also illustrates one of the many ways of arranging circuit breakers and bus-bars in two galleries. Cells for Voltag-e Transformers and fuses. — In installa- tions of this nature the voltage transformers are connected to large sources of power, and it becomes necessary to avoid possible damage to the sys- tem by one of them burning out; it is therefore customary to protect BUS-BAR STRUCTURES. 939 them with enclosed fuses, the fuse and transformer being isolated in their own individual cell in keeping with the practice of isolation which has been described. When the fuses are installed as described it is often desirable to close the cells with doors. Hig-h-Tension Conductors. — Manufacturers supply rubber-insu- lated cables for use up to a certain voltage, which can be relied upon for a long time in regard to insulation; but it is a well-known fact that rubber deteriorates with age and the higher the voltage the faster the deterioration, when conditions are favorable; so it is the best practice in all high-tension installations not to depend upon the rubber insu- lation, but to support the conducting cables on porcelain insulators and keep them away from all grounds and other conductors. The insulation on the cable serves, under such conditions, only as a possible preventive of troubles due to acci- dental contact therewith. This does not mean that the insulation is useless, as it might at times prevent loss of life or serious troubles due to accidental contact. Isolated cables laid against the grounded structure or covered with lead are subjected to strains, which might sooner or later break the insulation down. Lead-covered, paper-insulated cables are seldom used in high-tension switchboard structures. Some of the best cables obtainable are insulated with rubber. As the rubber, however, is com- bustible and easily takes fire from flash, manu- facturers supply cables, when required, covered with fire-proof braid of asbestos, or with the outer braid saturated with a fire-proof paint to prevent accidental burning of the rubber cover. For very high voltages, cables insulated with wrap- pings of impregnated cambric may be obtained, with or without a flame-proof covering. The terminals of cables used in the construc- tion of high-tension switchboards can be insulated with any good material such as oiled linen coated with shellac, but this should not be relied upon to prevent {accidental contact with live terminals, and no attempt should be made to insulate for safe handling, as the only time to safely handle a high-tension cable is when it is absolutely dead. Flame-Proof Covering's. — In order to prevent the flame from an arc setting fire to the insulation of a cable and being thereby communicated to other cables or setting fire to the building, flame-proof coverings are often used. These coverings are always supplied by the cable companies, being purchased under specifications which require that they shall meet the requirements of the National Board of Fire Underwriters. When installing such cables they must in every case be supported on insulators, and not carried in ducts, as the flame-proofing is a poor insulator and when saturated with moisture will serve as a conductor. For the same reason the covering must be stripped away from all live terminals a suitable distance for insulation purposes. Auxiliary IMrect-Current Circuits. — The direct current for operating the oil switches and other apparatus may be obtained as follows: From auxiliary storage batteries. From motor-generator sets. From direct-current exciter systems or other direct-current bus-bars. It must be especially noted that where the exciter system is controlled by a Tirrill regulator, the voltage fluctuation is likely to be so great that it cannot be relied upon for standard electrically operated apparatus. In this case either a small storage battery or a motor-generator set must be relied upon to supply the energy. In cases where a storage battery must be employed, owing to such considerations, and no charging current is available, a mercury rectifier may be relied upon to charge the battery. Fig . 3 1 . Three-Deck Oil Circuit Breaker and Bus-Bar Structure. Two Sets of Bus-Bars. 940 SWITCHBOARDS. fT\ xtCN /TX In cases where it is absolutely necessary to operate oil circuit breakers from direct-current exciter systems which are connected up to Tirrill regu- lators, the coils can generally be specially wound so as to operate at a low voltage, and the magnetic circuit be designed to saturate at high voltage so as to prevent the switch closing with too much force. Controlling* and Instrument Switchboard. — Under this head will be considered the installation of controlling switches and accessories that control electrically operated oil switches. In this connection it is essential to make sure that direct current is available at a suitable voltage to operate the electrically operated devices. The standard controlling devices are designed to operate from 125, 250 or 500- volt circuits, but when the potential is liable to drop below 80 volts, operating coils must be specially provided for the low voltage. The controlling appa- ratus can be mounted on the face of the switchboard panel together with the instruments where the system is simple and an inexpensive arrangement is desired. Nearly all large stations have the generator-control apparatus mounted on control desks or pedestals. A feature of some control outfits is the use of miniature bus-bars with lamps and indicators in the circuits. By means of these bus-bars the entire main station connections are embodied in miniature on the controlling desk, and, if the indicators or lamps are placed in the miniature circuits, the switching operations can be seen to take place when the operator moves his controller exactly the same as they occur in the main circuit. When the desk type switchboard is used, it is usually placed directly in front of the instrument switchboard and the operator has his control apparatus arranged as nearly as possible opposite the respective instrument panels. Nearly every large installation starts with a few generating units and increases as the demand for power increases. For this reason it is desirable that the structure used for carrying the control apparatus be so designed to admit of extension to meet future demands, or be made in the form of pedestals carrying the various in- struments. Such controlling table or pedestal should generally contain controllers, indicators and lamps for the oil circuit breakers, synchronizing plugs and lamps, voltmeter plug, electrically operated rheostat controller, a controller for the engine governor to change the speed in synchronizing the generators, and a controlling device to open and close the electri- cally operated alternating-current generator field switch. The usual method of controlling feeder circuits is to place the controllers on the switchboard directly beneath their respective feeder instruments. Generator-Control Pedestals. — For aux- iliary controlled switchboard apparatus, mountings must always be provided for the control apparatus of each generator. The pedestal shown in the illustra- tion is designed for this purpose, and is used in com- bination with an instrument post or panel located immediately in front of it. The pedestal as shown in Fig. 32 is designed to take the following apparatus: Signal lamps. Six oil circuit-breaker indicating lamps. Three oil circuit-breaker controllers. One voltmeter plug and receptacle. Two synchronizing plugs and receptacles. One controller for engine governor motor. One controller for electrically operated field rheo- stat. One control switch for electrically operated field discharge switch. One control switch for engine signal. The controlling devices are not included but must be specified separately, and may be selected to suit the requirements of the installation. Fig. 32. Control- ling Pedestal. CONTROL DESK. 941 Controlling* Desks. — Wherever great concentration of controlling apparatus is necessary, a desk or bench-board is often used. This is usually built of marble or steel, and special conditions sometimes require special designs. This type of controlling desk as shown in Fig. 33, has an iron frame enclosed by paneled steel sides and a marble top. The construction is such that each top panel with its corresponding paneled sides forms a section, and the desk may be extended in either direction by installing additional sections, the end panels and end moulding being remov- able in one piece to provide for inserting the necessary additions. Instrument I*©sts. — The instrument posts used with desks or control pedestals are divided into two general classes, viz.: swivel type and stationary type. Fig. 33. Sectional Controlling Desk. These again may be designed with suitable bases to mount jacks, or recep- tacles, to enable one to calibrate or check up the meters, by comparison with standards whose terminals have plugs to fit the receptacles. A post supplied with receptacles for calibrating meters as described above is shown in Fig. 34. Calibrating- Jacks. — In many installations it is desirable to have jacks or receptacles provided in the series and shunt transformer circuits to enable standard meters with suitable plugs attached to be connected in these circuits for comparing the readings of the switchboard meters. There are two kinds of these receptacles used, one for establishing a loop in a series transformer circuit and used for an ammeter jack or an ammeter plug receptacle, the other being a double-pole receptacle or voltmeter jack for use on shunt transformer circuits. 942 SWITCHBOARDS. Field Rheostats and Field Switchboards. — If the gener- ator control apparatus is located on a panel, the field rheostat can be conveniently operated by means of a hand- wheel geared directly to a face plate on the rheostat by gearing or chain and sprocket. If the rheostats are electrically controlled from a distance through face plates, they should have a small motor geared to the contact arm, the motor being controlled from the operating platform and the field switches electrically operated. Direct- Current Exciter Switch- board. —The switchboard for control of the exciters is sometimes placed in the operating gallery when this is not too remote from the machines. In other cases it is placed on the station floor, as near as convenient to the exciters. It is usually a typical direct-current board, and, while the most serviceable ones have entire panels finished in black marine, a large number of stations are using blue Vermont marble. The circuit breakers used are non- automatic, being used only to trip by hand when the circuit is to be interrupted, to pre- vent the arc from burning the switch. Some station managers prefer reverse-current circuit breakers in the exciter circuits, but the usual practice is to omit protective devices. Station Apparatus. — In addition to providing for station voltmeters, synchroscopes and wattmeters, either on panels or on an instrument post, it is often necessary to install static ground detectors. These are always operated through condensers so located that the wiring is short and the conductors properly separated and far enough from neighboring metal so as not to interfere with the operation of the instruments. This is best accomplished by installing the ground detectors on the station wall or on suitable supports near the condensers if it is difficult to properly run the leads to the operating gallery. Sub-Station Switchboard Equip- ments. — Sub-stations are more commonly used for railway service. The usual equipment of switchboard apparatus for a sub-station is laid out on the same lines as for a generating station, but the arrangement and selection of the equipment is changed to agree with the requirements of the case. As lighting and power sub-stations are more or less special it is impossible to give a description which will be generally applicable. Railway sub-stations, however, fulfill practically the same purpose and in general differ only in number and capacity of the units. The conductors in such a station are usually very heavy and care should be taken to make the runs as short and direct as possible. A number of modifications may be made in the apparatus supplied. For instance, electrically operated direct-current apparatus may be used to save cable and permit of greater concentration, or for small stations the alternating-current switchboard may have hand-operated circuit breakers mounted directly on the panels. For single-phase alternating-current railway systems, the sub-stations are essentially transformer houses and are very simple. Figs. 35 and 36 show a typical sub-station of this character. This apparatus for such a sub-station will vary with the requirements of service. nine Fig. 34. Post with instruments. The three ammeters at the bottom are for a bank of trans- formers. The six instru- ments above are for one generator. The plug switches in the base per- mit testing the calibra- tion of the instruments without removal. SUB-STATION SWITCHBOARD EQUIPMENTS. 943 v/mm a Fig. 35. Single-Phase Alternating-Current Sub-Station or Transformer House — End View. 944 SWITCHBOARDS. Fig. 36. Single-Phase Alternating-Current Sub-Station or Transformer House — Side View. SWITCHBOARD INSTRUMENTS AND METERS. 945 SWITCHBOARD I]¥STIlCnHCE]¥T AID METERS. The following is a list of the various instruments and meters used for switchboard work: Direct Current Alternating Current Indicating ammeter, Graphic ammeter, Indicating voltmeter, Graphic voltmeter, Single-phase indicating wattmeter, Single-phase integrating wattmeter, Single-phase graphic wattmeter, Polyphase indicating wattmeter, Polyphase integrating wattmeter. Polyphase graphic wattmeter. Graphic frequency meter, Graphic power factor meter, Differential voltmeter, Power factor indicator, Wattless component indicator, Frequency indicator, Synchroscope, Indicating compensating voltmeter, Electrostatic ground detector, Electrostatic voltmeter, Automatic synchronizer. Indicating ammeter, Graphic ammeter, Indicating voltmeter, Graphic drawing voltmeter. Integrating wattmeter, Graphic wattmeter. The names of the instruments in most cases describe their use. Inte- grating meters record by means of a dial the watthour output. Graphic meters record on a chart by a line the fluctuation of the voltage, cur- rent or watts of the circuit. Indicating wattmeters indicate the actual watts of the circuit which is equivalent to the volts as shown by the volt- meter multiplied by the current as shown by the ammeter multiplied by the power factor of the circuit, for single-phase circuits. Electrostatic Voltmeters are used only for high-potential circuits, such as 20,000 to 100,000 volts. They are connected directly to the cir- cuit without the interception of potential transformers and do not carry any current. Condensers are sometimes interposed. Alternating'- Car rent Instruments for high-tension circuits are not connected directly to the circuit, but are used in connection with cur- rent and potential transformers. Current transformers are connected in series with the main circuit, but are wound for different ratios of trans- formation so that approximately five amperes is obtained in the secondary, and therefore the instruments may all have five-ampere windings. The use of the current transformer makes it unnecessary to insulate the instru- ment for high voltages and furthermore does not necessitate running the high-tension leads to the switchboard. Ammeters are sometimes connected in series with circuits as high as 2500 volts. Potential transformers are usually wound to obtain from 100-125 volts on the secondary and are used on circuits of above 600 volts for voltmeters and other instruments having potential windings. 946 SWITCHBOARDS. method of JTigruringr Instrument Scales. SINGLE-PHASE GENERATORS: Minimum ammeter scale _ K.W. X 1000 X (1 + per cent overload guarantee) voltage Wattmeter scale = ammeter scale obtained from above X voltage. THREE-PHASE GENERATORS: Minimum ammeter scale _ K.W. X 1000 X (1 4- per cent overload guarantee) voltage X 1 . 73 Polyphase wattmeter scale = ammeter scale obtained from the above X voltage X 1.73. TWO-PHASE GENERATORS: Minimum ammeter scale K.W. X 1000 X (1 4- per cent overload guarantee) voltage X 2 Polyphase wattmeter scale = ammeter scale obtained from the above X voltage X 2. DIRECT-CURRENT GENERATORS: Minimum ammeter scale _ K.W. X 1000 X (1 + per cent overload guarantee) voltage THREE-PHASE MOTORS: Minimum ammeter scale Horse-power X 746 • ( voltage X per cent Eff. X per cent P.F. X 1 . 73 X U + Per U ' ^' } ' TWO-PHASE MOTORS: Minimum ammeter scale Horse-power X 746 X (1 +per cent O.G.). voltage X per cent Eff. X per cent P.F. X 2 DIRECT-CURRENT MOTORS: ii/r- • i Horse-power X 746 w ,, , . _ _, . Minimum ammeter scale = — j- -^ ^ _ „ X (1 4- per cent O. G.) voltage X per cent Eft. THREE-PHASE ROTARY CONVERTER: Minimum ammeter scale = K.W. X 1000 v n -l *f\c'\ voltage X per cent Eff. X 1 . 73 X per cent P.F. X U + P ercent U ' «■'. Wattmeter scale = ammeter scale obtained from the above X voltage X 1 . 73. TWO-PHASE ROTARY CONVERTER: Minimum ammeter scale ' K.W. X 1000 trim voltage X per cent Eff. X per cent P.F. X2 X1 +per Cent u -^ ) ' By per cent overload guarantee is meant the £, 1 or 2-hour overload guar" antee on the generator and not the momentary guarantee, although some prefer to have scales calibrated to read momentary fluctuations. The per cent efficiency and per cent power factor should be taken at full load or overload. The wattmeter scales should theoretically be multiplied by the power factor, but practically the scales work out better as given. Integrating watt- meters have no scales and therefore need only have sufficient current carrying capacity. When the minimum scale is determined from the formula the next larger standard scale, depending on the manufacture, should be selected. P.F. = Power Factor. O.G. = Overload Guarantee, GUIDE FOR SWITCHBOARD SPECIFICATIONS. 947 4 BRIEF GUIDE FOR WRITING IUITCHBOARB SPECiriCAlIONi. The initial and ultimate number of each type of generator, motor and feeder circuit with their voltage, kilowatt and frequency rating should be given. The overload guarantees of the machines and duration of same should also be specified. Other characteristics of the machine, such as "Y" connected three-phase generators with grounded or ungrounded neutral, two-phase generators with inter-connected phases, direct-current generators with grounded or ungrounded negative, should be clearly stated. Plans of the building, or of that section of the building occupied by the switchboard should, if available, accompany the specifications. It is essen- tial to know the construction of the floor supporting the switchboard, and if there is a basement below the floor, when oil switches, rheostats and other similar devices are not to be mounted on the panels. Specifications should be specific as to just what the switchboard contract is to cover. Switchboards as furnished by the manufacturers usually do not include the following, which should, therefore, be furnished by the purchaser unless otherwise specified. Complete flooring, sills for supporting switchboard and other pieces set in the floor or wall for supporting cable racks, oil switch operating mechanism, etc. All false flooring, if any is required. All masonry work for oil switch cells and bus-bar compartments. All openings in walls or floors, with suitable bushings. All clay ducts, iron conduit and other similar material to be laid in the concrete floors. Doors for bus-bar compartments, lightning arrester or static discharge compartments. All cable between switchboard and machines and between switchboard and feeder circuits. All bus-bars not connected directly with the switchboard, such as equal- izer or negative bus-bars near the machines. If the purchaser desires to include any of the above material in the switch- board contract, such material should be clearly specified. A connection diagram showing the proposed main connections, providing they are unusual or complicated, should accompany the specifications. The height and width of the panels should preferably be left to the discre- tion of the manufacturer. The thickness of the panels depends on the size of the panel, the material of the panel and the devices mounted thereon. The design of the supporting framework need not be specified. In general, statements in specifications can be made as follows: 1. "The material of the panels shall be such as to afford the proper insula- tion between live metal parts mounted directly on the panel, for the voltage on which they are used. It shall have a (natural oil), (black enameled) or (polished) finish, and the panels shall harmonize in color and markings and fit together in a neat and workmanlike manner. The panels shall be properly supported on iron framework. Connection bars, bus-bars and wires shall be properly supported and insulated." 2. "All instruments shall be dead beat and protected from stray fields produced by adjacent connections or bus-bars." 3. "Circuit breakers shall be of sufficient capacity to carry the overload ampere capacity of the generator or motor, without overheating. They shall be capable of opening under short circuited conditions without dan- gerously burning the contacts and shall be of such a design as to be positive in action." 4. "Oil switches shall have a kilowatt rupturing capacity based on the ultimate installation of generators as heretofore stated in these specifications. The switches shall withstand for one minute a potential test between con- tacts and frame, of at least twice the rated voltage of the circuit." 5. "All switches shall be of such capacities as to carry the one or two hours overload rating of the circuits to which they are connected without undue temperature rise, and shall be properly designed for the service for which thej' are intended and without defects of workmanship." 948 SWITCHBOARDS. 6. "Connection bars and wires shall be of sufficient cross section so that with maximum load the temperature rise at no point will exceed 40° C. rise above the surrounding air, which may be based on 20° C. Bus-bars shall be of sufficient cross section to carry continuously the total normal load of all the generators feeding in parallel through the busses at various points. The design of the busses shall, as far as possible, permit additions and extensions without materially interfering with the operation at a later date, or changing the existing supports. "Insulated main connection wires or cables should have flame-proof covering, and the insulation should not be wholly relied upon but should be supported by suitable insulators." It is not advisable to specify the contact area, cross section or rating of switches, circuit breakers or connection bars, as this often necessitates special devices, whereas standard devices could have been used if only the temper- ature guarantees were given. If purchaser has determined as to what instruments and switches are necessary, a complete list, giving the equipment of each panel, should be ncluded. Otherwise this equipment should be specified in detail in the manufacturers' proposal and inserted in the specifications forming part of contract. SWITCHING DEVICES. Switching devices in connection with switchboards can be divided gen- erally into the following-named classes, viz.: Switches for low voltage and small current are of the same general form, though differing in details. In the main they consist of a blade of copper, hinged at one end between two parallel clips, the other end of blade sliding into and out of two parallel clips. The clips are joined to copper or brass blocks to which the circuit is connected. There seems to be little uniformity among manufacturers regarding the cross section of metal and surface of contact to be used. Perhaps a cross section of metal of one square inch per 1000 amperes of current capacity is as near to the common practice as any, and a contact surface for bolted con- tacts of at least one inch per 100 amperes or ten times the cross section of metal is also common practice, but will depend somewhat on the pressure between surfaces. For sliding contacts the density per square inch should not exceed 75 amperes. Auxiliary breaks are demanded by the National Code for currents ex- ceeding 100 amperes at 300 volts, and "quick-break" switches are now quite common for pressure as low as 110 volts. The rules on switch design issued by the National Code cover the require- ments well, and they must be followed in order to obtain or retain low insurance rates; all switches must meet the requirements. Blades, jaws, and contacts should be so constructed as to give an even and uniform pressure all over the surface, and no part of the surfaces in contact should cut, grind, or bind when the blade is moved. The workman- ship should be such that the blade can be moved with a perfectly uniform motion and pressure, and the clips and jaws should be retained so perfectly in line that the blades will enter without the slightest stoppage. Sparking; at Switches. — In a paper read before the British Institu- tion of Electrical Engineers, A. Russell and C. Paterson discuss the subject of sparking at switches. In the diagram are given lengths of sparks at various constant voltages. Following are the conclusions arrived at: (1) The spark at break ought to be taken as a guide to the rating of a switch for use on direct-current circuits. (2) The shape of the terminals does not make much difference in the length of the spark. (3) The effect of increas- ing the speed of break above that ordinarily employed is small. (4) The effect of a double break is to make the lengths of the spark the same as the length of a spark with the same current at half the voltage. (5) The dif- ference in the length of the spark when copper, steel, or zinc is used is not great. (6) For small double-break switches for use on circuits of 200 volts and upwards, when the trailing spark just fails to bridge the air-gap, the air- gap should be double the distance at which a permanent arc can bejobtained. (7) For double-break switches for large currents under the same circum- stances the air-gap should be more than double the arcing distance. SWITCHING DEVICES. 949 SOdVduTS ao4Vo(.Y9 ' / / J / 9 / / / / / / > LaoVcLTa f / / -A / / ^ / y >T~ _io>Voi T « i ^ f^ / / | s / / / s P 13 / / y / s / / / / ~J_ / / t / ee^ 'OLI s 3 T / -X / t T 2 / s / V ^ «V *iT' i v 5 o i 3 8 3 A o s o s 3 ? e H J © 6 ia Ampere*. SPARKING AT SWITCHES* Fig. 37. Switching devices used in connection with switchboards can be divided into several classes as follows, viz.: Circuit breakers, automatic. Relays. Lever switches (knife switches). Quick-break switches. Plug switches. Disconnecting switches. Controlling switches. Oil-break switches (oil circuit breakers). Fuses. Circuit Breakers. A circuit breaker is a device which automatically opens the circuit in event of abnormal electrical conditions in the circuit. Automatic circuit breakers are designed for alternating and direct-current circuits. Alternating-current circuit breakers are usually made to operate on overload or low voltage. The usual conditions under which circuit breakers operate are: Overload. Underload. Reverse current. Overvoltage. Undervoltage. Electrically tripped from a distance (shunt trip). 950 SWITCH HOARDS. If no conditions are specified it is always understood that the overload circuit breaker is desired, as reverse current, low-voltage features, etc., are usually in the form of attachments to the standard overload circuit breaker. The Overload Circuit Breaker is used to protect the system against excessive overloads. The overload feature consists of a coil con- nected in series with the main circuit, which operates the circuit breaker tripping trigger by means of its armature. Since this power is obtained from a solenoid connected in series with the circuit breaker it is obvious that the number of turns of wire or bar on the magnet depends on the ampere capacity of the circuit breaker. Circuit breakers of 800 amperes and above may be designed so as to require but one turn which is obtained by encircling one of the studs of the circuit breaker with an iron horseshoe to which is pivoted the armature. In order to provide for a wide variation in capacities without introducing too many sizes, each circuit breaker is designed to cover a large range of current, between the limits of which it may be set to trip at practically any point. The limits of calibration usually range from 50 to 150 per cent of the con- tinuous current carrying capacity. The "Underload Circuit Breaker is similar to that for overloads, except that it acts in event of an underload instead of an overload. This Fig. 38. Type " C," Form " K2," 2000-Ampere, 650- Volt, Automatic Circuit Breaker, as manufactured by the General Electric Company. type of breaker is applied to storage battery circuits to cut off the battery when the current falls to an amount which would indicate that the battery was fully charged. It may also act as a reverse current circuit breaker, because during the reversal the current must fall to zero value. The under- load breaker also acts as a low-voltage breaker, inasmuch as if the source of power is cut off the flow of current will cease. However, it is not always desirable to use an underload breaker for such purposes as it would operate in many cases on small loads when not intended to. The Direct Current Reverse Current Circuit Breaker is essentially an overload breaker, having a potential winding operating magnetically in conjunction with the overload feature so that, the circuit breaker will open in event of a reversal of the direction of the flow of current. Under some conditions the circuit breaker would be required to operate on an overload and a reversal of current. In other cases it may be required to operate only on a reversal of current. Both kinds of circuit breakers are manufactured, but the most reliable method is to apply a reverse current SWITCHING DEVICES. 951 relay as described on page 961 to a standard overload breaker, having a shunt trip or low- voltage attachment. In this case the overload feature may be adjusted independently of the reverse-current attachment, or may be blocked to make it inoperative. The principal uses of the reverse-current circuit breaker are briefly de- scribed under the subject relays on page 961. The low-voltage feature is usually an attachment to a standard overload breaker, and is used chiefly on motor circuits to cut off a motor from the source of power in event of an interruption of current, in order that the motor may be properly started by the attendant, with the aid of a starting Fig. 39. Westinghouse Type C Circuit Breaker Showing Adjusting Mechanism and Terminals for Rear Connections. rheostat or compensator, when the source of power is restored. The low- voltage coil may also "be advantageously used to trip the circuit breaker from a remote place by shunting the coil. Low voltage breakers are operated by opening the circuit of the coil. Shunting the coil short circuits the line. The Application of Reverse Current Circuit Breakers to the Protection of Transmission Line in Multiple. — Where power is delivered to a single receiving point by more than one system of feeders, it will be seen that in the absence of suitable protective devices prop- erly disposed, a short-circuit upon one set of feeders will be fed not only through the portion of the feeder located between the short-circuit and the source of supply, but also by means of the portion of the damaged feeder beyond the short-circuit, with current flowing in the reverse sense from the receiving station. " Overload " circuit breakers at both generating and receiv- ing ends of the cables form a means of isolating the damaged lines. Their use alone, however, is liable to cause momentary interruption of service in the uninjured cables, which will be repeated until the damaged line is finally located and put out of service. Circuit breakers heaving reverse current operation located at the receiving end of the transmission lines will automat- ically sever the damaged cables at this end and prevent the receiving station from feeding back into the short-circuit; this being attained without inter- ruption of the service. In case of a receiving station having a number of feeders of approximately the same capacity, ordinary overload circuit breakers will generally afford ample protection because a short-circuit on one feeder will be fed through its own circuit breaker from all the other receiving station circuit breakers in parallel. This will tend to open the breaker on the short-circuited feeder line first, and relieve the system. If the station has only two incoming feeders, however, this condition obviously does not obtain and reverse circuit breakers are very essential. 952 SWITCHBOARDS. The Application of Circuit Breakers to the Protection of Storagre Battery Boosters. — Boosters of the compound or series type, if left connected with the system when the circuit of the driving motor is inter- rupted, will act as series motors rotating in the reverse direction, and, if not promptly disconnected, will attain a destructive" speed. Similar conditions occur should the booster circuit be closed before the motor has been started, or should the motor for any reason lose its field. Proper protection under these conditions is secured only by having an overload and no voltage circuit breaker in the motor circuit inter-connected with the circuit breaker in the battery circuit in such a manner that the motor circuit breaker must be closed, before the booster circuit breaker can be made to latch, while the opening of the first-named instrument instantly causes the opening of the second. The Application of Circuit Breakers to the Protection of Boosters Supplying" feeders. — Boosters employed to compensate voltage losses in feeders, incident upon transmission over considerable dis- tances, are either series or compound wound; if, therefore, when for any rea- son the driving motor is not receiving current, the booster should be left in connection with the system, it will run reversely as a motor, and in view of its series field-winding will attain destructive speed. This condition may be adequately dealt with by the employment of circuit breakers similar to those prescribed for the previous section. The low- voltage trip coil consists of a shunt winding connected across the circuit in series with a resistance, or may be connected in series with the shunt field of a motor if used on direct current. So long as the voltage remains constant the coil holds up a plunger, but if the voltage drops below a certain limit the plunger is released and the force of the blow trips the breaker. The shunt trip coil is normally open-circuited, and when energized, by means of a controlling switch or auxiliary switch or such device, it actuates the circuit breaker. CIHCriT BREAKER BESIGJi T. — Birect-Current Cir- cuit Breakers are made single, double and triple pole and four pole. The double-pole circuit breakers usually have the overload feature on one pole only, which is sufficient protection, except in case of the three-wire systems where a triple-pole breaker having two or three coils should be provided. Some types of double-pole breakers have a coil to a pole. Alternating-Current Circuit Breakers are made single, double, triple and four pole. The single-pole circuit breaker has one coil ; the double- pole circuit breaker has one coil; the triple-pole circuit breaker may have but one coil if used on a motor circuit, as there is practically no chance of a short circuit between but two of the leads, otherwise it should have two coils, and in cases where the three-phase system has a grounded neutral it should have three coils; the four-pole circuit breaker should have two coils, unless the phases of a two-phase system are interconnected, in which case it should have three coils. The carbon-break circuit breaker has been generally adopted for station work on account of the fact that it requires minimum attention, and will operate many times on short circuits without requiring cleaning or repair of the contacts. The sequence of operation of the various contacts of the carbon-break circuit breaker, is as follows: First, the main contact opens, which shunts the current through the intermediate and carbon contacts, then the inter- mediate contacts separate; this leaves the circuit through the carbon con- tacts, where the circuit is finally broken. The object of the intermediate contact is to prevent an arc forming on the main contact. Where it is desired to definitely direct the arc from the circuit breaker, or the amount of space for the arc is limited, such as would be the case in oar work, magnetic blowout breakers are preferable. Circuit breakers of the carbon break type which are in most common use, are preferably mounted at the top of the switchboard panels, as the arc formed in opening is invariably blown violently upward, and is liable to damage any apparatus mounted directly above it, or blacken and burn the panel. This tendency is not pronounced on small capacity circuit breakers on circuits of 250 volts or less, and this precaution is unnecessary. CIRCUIT BREAKERS.— For Alternating-Current Ser- vice. — The class of circuit breakers required for polyphase circuits largely depends upon individual conditions; the few cases considered here will suffice to indicate the principles which should influence the selection. CIRCUIT BREAKERS. 953 In the consideration of polyphase systems, it must not be forgotten that a large proportion of the generators and motors are made with interlinked windings, and for this reason circuit breakers for the protection of two-phase, four-wire generators and circuits should, regardless of voltage, provide for the severance of all four leads, as a single break in each phase still leaves the two remaining leads subject to a potential difference of not less than seven tenths of the voltage in either phase. This point is made clear by reference to the accompanying cut A, which shows two pieces of two-phase apparatus, as, for instance, generator and motor connected to the same circuit. On account of the windings being interlinked, it will be seen that the passage of current from one to the other is still possible, unless at least three of the four wires are severed.- Where, as is frequently the case, the entire output of the two-phase generator is supplied to single-phase transformers having independent primary windings, then it is true that in the absence of grounds or crosses B C Circuits Connecting Polyphase Apparatus. the generator will be fully relieved of its load by the opening of both phases, each at one point only. Preference to cut B shows, however, that the possi- bility of grounds or crosses is a contingency which in this case needs to be carefully reckoned with, as in the event of either of these conditions involving both of the unsevered mains, the opening of the circuit at one point in each phase does not relieve the generator. Circuits Connecting- JPol jphase Apparatus. — In the event of a short circuit on the mains supplying a synchronous motor this piece of apparatus, kept in motion by its own momentum, acts for the time being as a generator, thus, much increasing the severity of the short circuit. Again upon the opening of the circuit breaker the coincident slowing down of the motor results in its E.M.F. dropping out of phase with that of the generator, thereby very greatly increasing the total electromotive force of the circuit and producing abnormal strains upon opening devices and insu- lation. . . Therefore, the circuit breaker chosen should be such that when it is open, not more than one main of the circuit shall remain in connection with the source of the supply. Motors operating on three-wire circuits of moderate voltage may be adequately protected by double-pole circuit breakers. Those 954 SWITCHBOARDS. on four-wire systems fed from transformers whose secondaries are not in electrical connection may also be protected in the same manner. Four-wire transmission circuits require circuit breakers of not less than three poles, etc., but preferably the circuit breakers chosen for the protection of poly- phase generators and feeders should be capable of severing every main of the circuit, thus securing complete interruption of the current regardless of possible grounds and crosses. The higher the voltage of the circuit the more important this consideration becomes. The protection of polyphase motors is a subject deserving of special con- sideration. The staunch build of this class of apparatus and its known ability to withstand heavy overloads often lead to its being carelessly started Co/t.r£cT/o/r$ or Auto -^TAffTtff Art? CftCo/r QnEAXtn ffenot/t/fiG Circuit 3/tfAkt/t OP£#ATive 0/tj.r. DuRtriG, FturtritrtG ConPrtton* Fig. 40. and otherwise unduly abused. While this may not result in immediate injury to the motor, it causes excessive disturbances in the voltage of the circuit, and undue waste of energy. The heavy starting current required by many types of polyphase motors has in the past constituted a serious objection to the use of overload circuit breakers for their protection. This difficulty is overcome by making the connections between the auto-starter and circuit breaker such that the latter will be included in the circuit of the motor only when the switch of the auto- starter is in the running position. Reference to Fig. 40 shows how this may be effected. When the circuit breaker is connected in the manner there shown it will not be acted upon by the currents passing in the starting posi- tion of the switch, but should the switch be thrown into the running position at once or before the motor has come up to speed, the circuit breaker will open upon the resulting overload, as will also be the case should the motor be unduly loaded. Perhaps the most potent source of damage to polyphase motors is the accidental severance of but one phase of the circuit, due in most cases to the blowing of a fuse, either at the motor or somewhere in the circuit supply- ing it. Where this occurs when the motor is running the latter will, unless very lightly loaded, come to a standstill, and if not promptly disconnected will be seriously injured. CIKCUIT BREAKERS AND RELAYS. 955 Capacity of Circuit Breaker Required for B.C. Generators. The size of a circuit breaker is ordinarily determined by its normal current carrying capacity, and for any generator the capacity of the circuit breaker should be the same as the normal rated capacity of the generator, and the breaker should be calibrated for such a range of overload as is required by the service conditions. Capacity of Circuit Breaker Best Adapted for Motor of ©riven Size. The Cutter Company. The following table indicates the sizes of circuit breakers best adapted for the protection of various sizes of motors of from \ horse-power to 100 horse-power at voltages of 125, 250, or 500. The figures given in the left hand column indicate the horse-power of the motor at full load; the remaining columns show the normal capacity of the circuit breakers required for each of the voltages given. Horse-Power of Motor at Rated Load. For 125 Volts Nor- mal Capacity of Cir- cuit Breaker. For 250 Volts Normal Capacity of Circuit Breaker. For 500 Volts Normal Capacity of Circuit Breaker. \ 4 amperes 1 8 amperes 4 amperes 2 16 or 20 amperes 4 amperes 4 amperes 3 24 or 30 amperes 12 amperes 8 amperes 5 45 amperes 20 amperes 10 amperes n 60 amperes 30 amperes 20 amperes 10 80 amperes 40 amperes 20 amperes 15 150 amperes 60 amperes 30 amperes 20 200 amperes 80 amperes 45 amperes 25 200 amperes 100 amperes 60 amperes 30 300 amperes 150 amperes 60 amperes 40 300 amperes 150 amperes 80 amperes 50 75 400 amperes 600 amperes 200 amperes 300 amperes 100 amperes 150 amperes 100 800 amperes 400 amperes 200 amperes BELAYI, Definition. — A relay is a device which opens or closes a local circuit under pre-determined electrical conditions in the main circuit. Classification. — There are three general classes of relays as follows: 1. Signalling. 2. Regulating. 3. Protective. Signalling- Relays. Function. — The signalling relay acts to transmit signals from a main to a secondary circuit. Application. — They are mainly used in telegraph and telephone work, being known by the terms telegraph or telephone relays, and ao not need further description here. 956 SWITCHBOARDS. Regulating 1 Relays. Function. — The regulating relay acts to control the condition of a main circuit through control devices actuated by a secondary circuit. This control may involve the maintenance of either the voltage, current, fre- quency or power factor of a circuit at a constant value. Application. — The regulating relay finds application in generator and feeder circuit regulators, such as the Tirrill Regulator, etc., in which it forms the main device, all other apparatus being subsidiary and actuated thereby. It differs from the usual protective relay in having its contacts differ- entially arranged, that is, so that contact is made on a movement of the relay to either side of a central or normal position. The regulating relay is usually considered a component part of its par- ticular regulator and for this reason it will not be further considered here. Protective Relays. Function, — Distributing systems requiring more selective and flexible protection than that afforded by the inherent control features of automatic circuit breakers are equipped with protective relays. Protective Relays. — Protective relays are used entirely for the protection of circuits from abnormal and dangerous conditions such as over- loads, short circuits, reversal of current, etc. They act in conjunction with automatic circuit breakers, operating when their predetermined setting has been reached, energizing the trip coils of the breakers and opening the circuit. Auxiliary Relays. — Sometimes a main relay, due to inherent limitations, is not able to fulfill all of the necessary requirements. An "auxiliary" relay is then used in conjunction with the "main" relay and supplies the missing functions. Such missing functions may be for example: 1. Lack of time element feature in the main relay. 2. Insufficient carrying capacity of the main relay contacts. Classification. — Protective relays are sub-divided according to their particular function into the following classes: Over-voltage, overload, overload and reverse current, reverse current, underload, low-voltage and reverse phase. These designations indicate the circuit conditions under which the various classes operate. For example, the over- voltage relay operates when the voltage rises above a predetermined amount; the reverse current relay operates upon reversal of current, etc. Time Element JPeature. — Continuity of service is an essential consideration in all installations, and interruption of the service cannot be tolerated unless the protection of the apparatus demands it. There are, however, certain abnormal conditions of current flow which may exist for a short time on a circuit without causing serious damage, such as swinging grounds, intermittent short circuits, synchronizing cross currents, etc. The simple instantaneous relay would in such cases act instantly and interrupt the service unnecessarily. There has, therefore, arisen the necessity for relays having a retarded or time element action. Refinite Time JLimit Relay. — For certain service it is sufficient that this retarded action have a definite predetermined value independent of the load condition. Such a relay is termed a "definite time" limit relay. Inverse Time JLimit Relay. — For other service it is necessary that this time element vary inversely with the load, that is, with greater load the time element should be less, and vice versa. Such a relay is termed an "inverse time" limit relay. Application of the Instantaneous Relay. — Instantaneous relays are used where it is desired to give protection only at the limiting carrying capacity of the apparatus. Application of Refinite Time .Limit Relay. — Definite time limit relays are used where it is necessary to maintain service on a given circuit at all hazards for a predetermined time. This allows temporary grounds and short circuits to clear by burning themselves out, and prevents synchronizing cross currents from opening the breakers. Most desirable of all, however, it enables instantaneous and inverse time-element relays on CIRCUIT BREAKERS. 957 contiguous circuits of less importance to operate and cut off under dis- turbances without opening the important circuit, even though the latter is temporarily heavily overloaded during the disturbance. Characteristics of the Inverse Time Element Relay. — — Inverse time element relays possess two valuable characteristics as follows: 1. Their operation is inversely proportional to the strain on the system ; the greater the strain, the quicker the relay will operate. 2. By virtue of 1, they act "selectively," those nearer a point of dis- turbance in a system, and which, therefore, receive the greatest load, oper- ating first, cutting out the affected portion and clearing the system while confining the disturbance to a minimum area. As an example, consider a system of three feeders (1, 2, and 3, Fig. 41) connecting a set of power station bus-bars, A, with a set of sub-station bus-bars B, and protected with auto- matic circuit breakers controlled by overload inverse time element relays at D, E, F, and reverse current inverse time element relays at P, Q, R. The overload relays will each be adjusted for operation at the same current; like- wise the reverse current relays will each be adjusted for operation at the same current. Assume now that a short circuit develops in 1 at point X. All three feeders will at once commence to supply current to the short circuit from A . A 8 1 X P E Z Q r w _ 3 K Fig. 41. Illustration of Selective Action of Inverse Time Element Relay. If B is a rotary converter sub-station, the rotaries, by virtue of their enormous fly wheel effect, may tend to supply current also, but as this has no par- ticular bearing on the point to be brought out it will not be further consid- ered. D being nearest the fault X, and therefore in the circuit of least line drop, will receive more current than E and F. By virtue of the inverse time law it therefore operates first or "selectively," cutting off the feeder 1, from A before E and F have time to act. Simultaneously P has been receiv- ing current in the reverse direction through bus-bars B, from feeders 2 and 3, and has cut off feeder 1 from B. Q and R will not operate as they receive current only in the normal direction, and E and F will not operate as the fault has been isolated and they have been relieved of their overload before they have had time to act. In actual practice on alternating-current circuit relays P, Q, R will operate on both overload and reversal of current, and are so designed that the operation on reversal of current is at a much lower value than on overload (about ^ to £ in representative types). If overload and reverse current relays were used at P, Q, R, the relay at P would operate before Q and R, for the reverse fault current flowing through P is the sum of the normal fault currents through Q and R. Where only two feeders exist as, say 1 and 2, P and Q would each receive the same amount of fault current, and the selective action is not so great, but is still amply sufficient to allow P to operate before Q, on account of the difference between their reverse and overload tripping values. 958 SWITCHBOARDS. Similarly to the definite time element relay, the inverse time element relay will allow temporary grounds or short circuits to clear themselves and will prevent synchronizing cross currents from opening breakers. This action is somewhat more limited in the latter on account of the inverse feature, but is quite sufficient for all ordinary conditions. Jleclia ni am of the ^Protective He lav. — Protective relays in their simplest form consist of three elements: 1. The actuating mechanism energized by the line source to be pro- tected. 2. A set of contacts operated thereby. 3. The time element feature (where present). Actuating* Mechanism. — The actuating mechanism assumes the form which will give operation under the desired conditions. It usually involves a motive device consisting of a solenoid and core, a rotating motor or some form of instrument movement. Tripping* Mechanism. — This usually consists of a set of moving platinum, silver or carbon-tipped contacts engaging a corresponding set of stationary contacts. Some relays have single contacts for closing a single tripping circuit; others are provided with multiple contacts for closing two or more tripping circuits, as in the operation of double throw systems where a relay in the main circuit has to operate circuit breakers in each of the duplicate feeder bus-bars. Time Element Mechanism. — In this instantaneous relay all retarding mechanism is eliminated, the relay acting practically instantane- ously with the application of an excessive current. In the definite time limit relay it is the usual practice to employ an air dashpot, such as used in arc lamps, to the piston of which the contact mechanism is attached. Upon the operation of the actuating mechanism the contact mechanism is released and allowed to descend by gravity against the action of the dashpot, thereby making contact a definite interval of time after the disturbance and inde- pendent of the magnitude of the disturbance. In the inverse time limit relay the actuating and contact mechanism is attached directly to an air bellows and upon operating tends to compress the bellows against the action of a specially constructed escape valve in the latter. The amount of the retardation varies inversely with the pressure on the bellows and, therefore, inversely with the magnitude of the disturbance. An alternative arrangement replaces the bellows with a conducting disk cutting a magnetic field, in which the retardation due to the eddy current reaction, induced on moving the disk through the field, varies inversely with the magnitude of the force with which the disk is urged through the field and hence inversely with the disturbance. Shunt Trip Contacts. — The usual arrangement of relay contacts provides for their closure upon the operation of the relay, in which case the relay is spoken of as being provided with "shunt trip contacts." The con- tacts are connected in series with the tripping circuit of the breaker and an independent source of current, and upon closing energize the tripping circuit and open the breaker. The tripping coils are wound for shunt operation from the independent source which is usually a direct-current exciter circuit or a storage battery, and the circuit breaker is spoken of as being equipped with shunt trip coils. The operation of shunt tripping coils from the circuit being protected is inadvisable, owing to the liability of the trip coil failing to operate on the low voltage existing under short circuit and overload conditions. Series Trip Contacts. — Where an independent source of current is not available the circuit breakers are provided with series tripping coils, wound for operation from series transformers in the main circuit. Overload relays are also provided with series trip contacts which differ from the shunt trip contacts in being normally closed instead of open, and opening upon operation of the relay. They are connected in shunt with the series trip coils short circuiting the same. Upon operation of the relay they open, allowing the transformer secondary current to flow through the trip coils and trip the breaker. As there is always sufficient current flowing under overload and short circuit conditions to operate the trip coils, this arrange- ment is as satisfactory as shunt tripping. CIRCUIT BREAKERS. 959 Protection of Alternatingr-Current Systems. — The applica- tion of relays to any given system depends almost entirely upon the local conditions of operation, varying somewhat with each installation. Generator Circuit Protection. — Representative practice rec- ommends the placing on generator circuits of either a reverse current relay, with a time element feature, or else the entire elimination of automatic protection. feeder Circuit Protection. — For feeders at the power station end, overload inverse time element relays are desirable. For feeders at the sub-station end, overload and reverse current inverse time element relays are desirable. Botary Converter Circuit Protection. — With rotary convert- ers, an overload inverse time limit relay in the high tension side of the power transformers will give protection for the alternating-current side. For the direct current side a reverse current inverse time limit relay operating the direct current breakers will be required. Protecting* JF our- Wire Three-Phase System. — An example of the relaying required in a typical four-wire three-phase system is illus- Automatlc Oil Circuit Breaker with relays. \ \ Power Bus// \ Bus * Tr s ans Pis. 3 Phase Generator 7/ Generator \ -H- Neutral , % -*& us\ SW9. A.C. Distribution fea * l\ cQ^^ Hf-^O^ Distri. bution * if o f Botary -Con verier # Catbort breaker with relaj; Fig. 42. Relaying of a Four- Wire Three-Phase System. trated in Fig. 42. Three generators operating with their neutral points grounded through a resistance, feed a common bus system, four sets of feeders, power transformers, rotaries, etc., for alternating-current and direct- current distribution of power. Automatic circuit breakers are inserted operated by relays as follows: At A, A.C. = Overload and reverse current inverse time element relays. At B, A.C. = Overload inverse time element relays. At C, A.C. = Overload and reverse current inverse time element relays At D, A.C. = Overload inverse time element relays. At E, D.C. = Reverse current inverse time element relays. The relays at A are intended for reverse protection only and so have their overload adjustment set at the maximum value. 960 SWITCHBOARDS. Relays Commonly Employed. most commonly employed are: The types of protective relays D.C. Over-voltage relays. D.C. Reverse current relays. D.C. Low- voltage relays. D.C. Underload relays. A.C. Overload relays. A.C. Overload and reverse current relays. A.C. Low-voltage relays. A.C. Reverse phase relays. Fig. 43. Westinghouse Direct-Current Time Limit Relay Definite Time Limit Action Shunt Trip Contacts. Application. Direct-Current' Over- Voltagre Relay. — The direct-current over- voltage relay is used chiefly on battery charging panels, but is also used to protect any direct-current apparatus which would be liable to dam- age from excess voltage. In storage battery work the relay may be used to disconnect the battery from the circuit when it is fully charged, as under certain well-defined conditions the voltage of the battery is a measure of its charge. The voltage of a battery is dependent, however, not only on its inherent characteristics, but also upon its charge and discharge history. Abnormal charging and discharging conditions operate to temporarily or BELAYS. 961 permanently change the law of a battery's voltage curve, and an over- voltage relay set for a given full charge condition may actually operate when the battery is not at full charge. The proper setting of a relay on such a circuit is, therefore, a matter entirely to be determined by the operating conditions and with full consideration being given to the effect upon the full charge voltage of the charge and discharge factors. Direct- Current Reverse Current Relay. — The direct-current reverse current relay is chiefly used for the protection of storage battery installations and rotary converters. When applied to rotary converters operating in parallel the relay serves to protect against short circuits occur- ring on the alternating-current side of the rotary, on the direct-current side between the rotary and relay, or in the rotary itself. Fig. 44. General Electric Alternating- Fig. 45. General Electric Alternating* Current Overload Relay Instantan- Current Overload Relay Instantan- eous Action Shunt Trip Contacts. eous Action Series Trip Contacts. - Short circuits occurring on the direct-current side beyond the relay are taken care of by the circuit breaker overload coils. When applied* to storage battery installations the relay prevents the battery from discharg- ing back into its charging source. Time Element feature. — When synchronizing machines to a system operating a rotary converter, momentary and harmless corrective currents are liable to flow toward the rotary on the direct-current side. In order to prevent interruption of the circuit by such flow, where reverse current relays are present, it is necessary that the latter have a time element. This time element must be of the inverse order to give quick interruption on overloads and short circuits and to give a selective action so as to cut off affected circuits. Overspeeding- of notaries. — Reverse current relays are not a complete protection against the overspeeding and running away of rotary converters such as would result from the opening of the rotary 's field. They should be supplemented by mechanical overspeed devices attached directly to the shaft of the rotary and arranged to close the trip circuit upon opera- tion. Such additional precaution is necessary as very low reverse currents exist under such conditions, only sufficient to supply the losses in the rotary and less than the minimum setting of the ordinary reverse current relay, which will therefore fail to operate and protect the machine. Direct-Current Xow-Voltagre Relay. — This relay is generally used in connection with direct-current motors and operates when the vol- tage of the circuit falls below a predetermined value. 962 SWITCHBOARDS. Direct-Current TTnderload Relay. — This relay is mainly used in the charging of storage batteries to disconnect the batteries when charged. Alternatingr-Current Overload Relay. — This relay is used very extensively, mainly for the protection of feeders, rotary converters, motors and transformers. All three forms exist, viz.: the instantaneous, definite time limit and inverse time limit, each finding its special application as outlined in the preceding pages. Either series or shunt trip contacts are provided depending on the tripping source. Alternating--! lurrent Overload and Reverse-Current Re- lay* — This relay is an important one, very extensively used for generator and feeder protection. It exists only in the inverse time limit form. When used for generator protection the overload adjustment is set at the maxi- mum value to give overload protection only at the maximum carrying Fig. 46. Westinghouse Direct- Current Over- Voltage Relay In- stantaneous Action Shunt Trip Contacts. hotts^^t Fig. 47. Westinghouse Alternating- Current Overload Relay (Cover Re- moved) Inverse Time Limit Action Shunt Trip Contacts. capacity of the generator and a sensitive reverse protection to prevent a return of energy from the line. The selective action of this relay has been covered in the preceding pages. Alternatingr-Current I^ow- Voltage Relay. — This type of relay is used for the protection of induction motors against a fall in the line voltage. Alternatingr-Current Reverse-Phase Relay. — This relay is used to protect synchronous apparatus against a reversal of direction of rotation or phase progression of the alternating-current source. Remote-Control Switches for Equalizer Circuits. — In large power houses of the modern type, considerations of economy as well as of convenience, dictate the placing of equalizer switches close to the genera- tors, for the reason that the cost of cable to connect the generator to the equalizer, if carried from the machine to the switchboard and back again, would be excessive. DISTANT CONTROL SWITCHES. 963 Figure 48 herewith is an illustration of a pair of switches, made by the Cutter Company for a large New York power house, that meet these condi- tions: The right-hand switch has a capacity of 2,000 amperes, and the left, 3,000. The upper terminal of each of these switches connects with the equalizer main, which takes the place of the equalizer bus otherwise required at the switchboard. The lower terminal of each is connected with the appropriate terminal of the series winding of the corresponding generator. The closing of each switch, therefore, completes the equalizer circuit of Fig. 48. Remote- Control Switch for Equalizer Circuits. the generator with which it is connected. These switches are designed for control from a switchboard located at a distance, such control being effected by the movement of a small double-throw switch, bringing into circuit upon one movement the opening coil, and upon the opposite move- ment, the closing coil. Each switch is also provided with a nand-closing mechanism, so that it can be operated at the machine. Lever Switches. — Lever switches are plain knife blade switches and are for use on direct-current circuits up to 250 volts and on alternating- current circuits up to 500 volts. The design of these switches is thoroughly covered by the "Fire Underwriters Code." The accompanying diagram represents typical lever switches. It should be noted that for switches of 964 SWITCHBOARDS. this class, a capacity of 4500 amperes in a single switch is about as high as is practical, as heavier capacity switches are liable to be too hard for the ordinary attendant to operate manually. Fig. 49. S.P., S.T., 250- Volt 200- Ampere Lever Switch. Fig. 50. 6000 Ampere, S.P., S.T. Quick lit calt Switches. — The quick break switch is essentially a lever switch provided with spring-driven follower blade which remains in the clips after the main blade leaves and is opened quickly by means of a Fig. 51. TO y jotted tines, i J -Lint Fig. 59. Diagram of Connections. 3-Pole Electrically Operated Type C Oil Circuit Breaker. CIRCUIT BREAKERS. 971 tii = — ^ — & — s — s — » ft* fiecf/omp. ■•4- 7rlp Indicating Lamps ' ^are enLonyx aearswi-kh COitS connected as shown for I2S» tsovotta. Connections for — —i SOOVO/vnttlori \oreir>ctfecrted \t>yavffod/Snesi 7ow}gCe//f. -line Fig. 60. Diagram of Connections. 3-Pole Electrically Operated Non- Automatic, C Oil Circuit Breaker. Westing-house Type B Oil Circuit Breaker for Potential 3500 to 32,000 Volts. The type B circuit breakers are made in the electrically operated form for potentials of 3500 to 22,000 volts, and in capacities up to 600 amperes. A simple system of toggles and levers is mounted on the top of the breaker, and a powerful electromagnet is arranged with its movable core Fig. 61. Hand-Operated , Automatic, 600- Ampere, 3-Pole, Type B, Oil Circuit- Breaker Mounted on Panel. Tank Removed, Open Position, not Over 22,000 Volts. attached to the lever system, so that when it is drawn into the coil, the circuit breaker will be closed. A tripping-coil is also mounted with the operating mechanism. A small single-pole, double-throw switch is mounted on the breaker, and is operated by the motion of the levers in opening and 972 SWITCHBOARDS. closing the circuit; it controls the tell-tale indicator and lamp which ara mounted in view of the operator. These circuit breakers are operated by 125, 250 or 500- volt direct current, and are calibrated for 25 cycles. The electrically operated type E oil circuit breakers are made both non- automatic and automatic, the latter being operated by means of overload relays. The breaker is made in single-pole units, each being mounted in a brick or concrete compartment. Two, three and four-pole combinations are made by placing these units side by side. The tanks are of a design similar to those of the type C circuit breakers. Oil Switch Structures. — The structural work for types C or E oil switches may be brick or concrete. When the structure is of brick, it is necessary that the anchor bolts pass /XT tmea 7#rf79ffer Fig. 62. Diagram of Connections. 3-Pole Electrically Operated Type E Oil Circuit Breaker. outside of the brickwork. When the switch has a concrete base, however, the bolts are usually anchored in the concrete. The only soaps tone sup- plied with the type C oil circuit breakers is the top slab, the blocks to hold the terminal insulators in the rear, and the soapstone barriers between these terminals. Westing-house Type GA Electrically Operated Oil Circuit Breaker. Westinghouse type GA oil circuit breakers are designed for use on cirouits carrying large amounts of power. The distinctive features of the type G A circuit breakers are: Liberal insulation and breaking distances; open position maintained by gravity; all metal tanks and tank tops; accessibility of parts; long break in clean oil; low first cost. Construction. — Type GA circuit breakers consist of one or more poles self-contained in heavy steel oil tanks with treated linings and provided with CIRCUIT BREAKERS. 973 heavy cast-iron covers to which all of the mechanism for operating each pole is secured. Each pole is entirely separate and distinct from the others, the operating rod being the only connection between them. Fig. 63. Type GA Oil Circuit Breaker, without Closing Mechanism, for Po- tentials of 44,000 to 110,000 Volts. Current-carrying- Capacity. — Type GA circuit breakers are de- signed to carry 300 amperes per pole with a maximum temperature rise of 20° C. They can be built with a larger current-carrying capacity if desired. Voltagres. — These circuit breakers are built for use on circuits of 44,000, 66,000, 88,000 and 110,000 volts. Breaking 1 Capacity. — Type GA circuit-breakers are guaranteed to open any short circuit which may develop on transmission systems of the following capacities: 60,000 kilowatts at 44,000 volts; 80,000 kilowatts at 66,000 volts; 100,000 kilowatts at 88,000 volts; 120,000 kilowatts at 110,000 volts. Tlie breaking* distances, or distances between contacts when the circuit breaker is open, are large and there are two breaks on each pole. The breaking distances of different capacity circuit breakers are given in the follow- ing table: 974 SWITCHBOARDS. Breaking* Distances. Amperes. Voltage. Minimum Dis- tance of Terminal to Case or Ground. Breaking Dis- tance per Break, Inches. Breaking Dis- tance per Pole, Inches. 300 300 300 300 44,000 66,000 88,000 110,000 16 22 30 32 11.5 16.5 20 23.5 23 33 40 47 Fig. 64. Westinghouse High-Potential Fuse-Type Circuit Breaker — Open Position for Potentials not Exceeding 66,000 Volts. High-Potential fused Circuit Breaker, Westingrhouse. — This fuse-type circuit breaker consists of a long hardwood pole on which is mounted a movable arm consisting of a reinforced fuse tube. At the bottom of the fuse tube is a brass expulsion chamber which is connected to the lower ter- minal of the breaker by a flexible copper shunt. Attached to the top of the pole and forming the upper circuit-breaker terminal there is a brass bracket, with a groove along its top, which supports the fuse, and a wing nut to hold the end of the fuse when the breaker is closed. The fuse passes from the wing nut over the bracket and down through the fuse tube to the expulsion chamber when it is attached to the screw-plug terminal shown in the end of the expulsion chamber. The pole of the circuit breaker is provided with spring jaws or clips so that it may be quickly and easily attached to or detached from the line terminals at the base. Adjustment. — First: Remove the mechanism from its base, taking hold of the long pole. Second: Remove the screw-plug terminal from the lower end of the expulsion chamber and attach the fuse to the terminal . Third: After passing the fuse through the fuse tube replace the screw plug and attach the other end of the fuse to the wing nut on the bracket at the upper end of the long pole, passing a turn or two around the lug, at the same time drawing the moving arm into its proper position against the end of the bracket. Fourth: Replace the mechanism on the base and the breaker is ready for use. Operation. — When the load on the line exceeds the capacity of the fuse the latter blows and the arm of the breaker swings by its own weight away from the upper line terminals, thus giving a positive indication that the fuse has blown. CIRCUIT BREAKERS. 976 Oil Circuit-Breaker Controller. — This controlling switch is of the drum type with a hinged handle, which, when thrown to the open position, may be locked by swinging the handle outward so that it is in line with the drum shaft. It cannot be locked in the closed position. When the handle is raised as described it indicates to the operator that the switch is out of service. The act of raising the handle cuts the current off from the controller and thus extinguishes the lamps. The switch is arranged for switchboard mounting, the dial and handle being on the face of the panel. It may also be provided with an indicator to show the last operation performed. Fig. 65. Controlling Switch, Cover Removed Lamp Indicator for Oil Circuit Breaker. — The indicator consists essentially of a hollow tube with a lamp socket mounted on a porcelain base in one end, held in position by suitable clips. The socket can be easily removed and is intended to hold a 5 c.p. candle-shaped in- candescent lamp which extends into the tube. Suitable holes are pro- vided for ventilation. A colored lens is secured to the front end of the tube. A special feature of the lens is a V-shaped projection which extends across its face, enabling the operator to see the light from any angle within an arc of 180°. Control and Instrument JLeads. — The control wires for the electrically operated circuit breakers are run in conduits, or in some other suitable manner, to the place where the operating switchboard is located. The small size of the controlling and conducting devices permits a large number to be grouped in a comparatively small space where they are easily accessible to the operator. The sizes of conductors usually required where lengths do not exceed 200 feet, are as follows: For series tranformer circuits, each lead equivalent to No. 7 B. •e f > • • • O III o o .• • o • J v §• i e Fig. 1. Non-Arcing Railway Lightning Arrester, Type "K." (For Station Use.) the surface of the charred grooves from one electrode to the other exactly as it would if there were but a simple air-gap. The presence of the charred grooves simply makes the path easier. , There being no room for vapor between the two tightly fitting blocks, no arc can be formed, hence the arrester is non-arcing. Some years ago Prof. Elihu Thomson devised a lightning arrester based on the principle that an electric arc may be repelled by a magnetic field. © \\ /( \Pj! r © ^g--— ^ , g> ,1 m ^|t||i|: willlllb' (o\ ^- — - — - — Fig. 2. Type "A" Arc Station Arrester. In this device, the air-gap, across which the lightning discharges to reach the ground, is placed in the field of a strong electro-magnet. When the generator current attempts to follow the high potential discharge, it is instantly repelled to a position on the diverging contacts where it cannot be maintained by the generator. 986 LIGHTNING ARRESTEES. The magnetic blow-out principle has been employed in the construction of a complete line of lightning arresters for all direct current installations, and in more than ten years of service magnetic blow-out arresters have always been effective in affording protection to electrical apparatus. In designing lightning arresters for the protection of high-voltage alter- nating current circuits, however, different conditions have to be met, since high-voltage arcs are not readily extinguished by a magnetic blow-out. In a recently designed lightning arrester for alternating current circuits, metallic cylinders with large radiating surfaces are found so to lower the temperature of the arc that volatilization of the metal ceases and the arc is extinguished. The variety hi these lightning arresters provides for the protection of all forms of electrical apparatus and circuits. The Type * 4 A" Arrester is manufactured for the protection of arc lighting circuits. Its construction includes a pair of diverging terminals mounted on a slate base with an electro-magnet connected in series with the line. The magnet windings are of low resistance, and therefore consume an in- appreciable amount of energy with the small current used for arc lighting, although they are always in circuit. The single Type "A" Arrester is suitable for circuits of any number of series arc lamps not exceeding seventy-five. For circuits of higher voltage, a double arrester known as the type "AA" is made by mounting two arresters on one base and connecting them in series. One arrester should be installed on each side of the circuit, as shown in the Diagram of Con- nections. For use in places exposed to weather, the Type "A" Arrester is furnished enclosed in an iron case, and designated Type "A," Form *'C." © ij.i © ® jj U' fi J fi ii m , 3. Connections for Type "A" Arresters. Fig. 4. Type "B" Incandescent Station Arrester 300 Volts or Less. The construction of the Type "B" Arrester is similar to that of the Type " A, " but its magnet windings are excited only when a discharge takes place across the air-gap. A supplementary gap is provided in the Type "B" Arrester, in shunt with the magnets, thus providing a relief for the coils from excessive static charge without affecting their action upon the main gap. The magnet coils, carrying current only momentarily, allow the same arrester to be used on circuits of large and small ampere capacity. The Type *'B" can also be furnished with weatherproof case similar to that used with Type "A." The Type "MD" Lightning Arrester has been designed for use on direct current circuits up to 850 volts. While similar to Type "M," Form "C, " Arrester, it is considerably smaller, and is enclosed in a compact porcelain box measuring 7£ inches x 5 inches x 4£ inches. For street car and line use, the arrester is furnished in an additional baK of iron or wood. LIGHTNING ARRESTERS FOR ALTERNATING CURRENT. 987 The arrester has been adopted as standard for railway and all direct current 500-volt circuits. It has a short spark gap, a magnetic blow-out and a non-inductive resistance. CONNECTIONS OF MAGNETIC BLOW-OUT LIGHTNING ARRESTERS TYPE MD. FOR DIRECT CURRENT CIRCUITS OP TO 850 VOLT8. CONNECTIONS FOR LIGHTING OR POWER CIRCUITS. (METALIC CIRCBIT8) REACTANCE COIL CONNECTIONS FOR RAILWAY CiROUIT (ONE StDS GfiOUMDEO) REAOTAKCe COtLJB COMPOSED OF 25 FT. OF CONDUCTOR WOUND ffl A COIL OF TWO OR MORE TURKS AS CONVENIENT. Fig. 5. Connections of Magnetic Blow-out Lightning Arresters, Type "MD," for Direct Current Circuits up to 850 Volts. IIGHTXOG ARRESTERS FOR AJLTERXATXIVO CCRREHT. •The G. E. Alternating Current Arresters have been designed to operate properly with very small gap spaces. The arrester for 1000-volt circuits has two metal cylinders 2 inches in diameter and 2 inches long, separated by a spark gap of about ■£% inch. One cylinder is connected to the overhead line and the other cylinder to the ground, and a low non-inductive graphite resistance is placed in circuit. The large radiating surface of the metal cylinders combined with the effect of the non-inductive resistance prevents heating at the time the lightning discharge passes across the gap, and the formation of vapor which enables the current to maintain an arc is thus avoided. The arrester under normal action shows a small arc about as large as a pin-head between the cylinders. The arrester for 2000-volt circuits is designed with two gaps of approxi- mately 3*5 inch each and a low non-inductive resistance. The G. E. Arresters are now furnished by the General Electric Company for use on all alternating current circuits at practically any potential. For circuits above 2000 volts, the standard 2000-volt double-pole arrester hae been adopted as a unit, and several of these are connected in series to give the necessary number of spark gaps. 988 LIGHTNING ARRESTERS. REACTANCE COIL Fig. 6. G. E. Alternating Current Three-Phase Multiplex Lightning Arresters, 10000 V. ARRESTER CONSISTS CF FOUR 2000 V.D.P. ARRESTERS CONNECTED IN SERIES. [L^j i^jj 20 2000 V.D.P. ARRESTERS CONNECTEO AS 8000 V.S.P. ARRESTERS. Figs. 7, 8, 9. Connections of G.E. Alternating Current Lightning Arresters 2000, 3000, 10,000 Volts. LIGHTNING ARRESTERS FOR ALTERNATING CURRENT. 989 1000 V0LT3 9000 VOLTS LINE ARRE8TCR DYNAMO —rfl Ck-Q- imr jrilfr ME J GROUND Fig. 10. Diagram Showing Electrical Connections for A. C. Lightning Arresters. Fig. 11. Double-Pole Non-Arcing Metal Lightning Arrester, Type "A.' (For Station Use.) The J¥on -Arcing* .ffotal JLig*htning* Arrester. — The non- arcing metal lightning arrester made by the Westinghouse Co. for alternat- ing current circuits is based upon the discovery made by Mr. A. J. Wurts that an alternating current arc cannot be maintained over a short air-gap when the electrodes consist of certain metals and alloys thereof. Types *' A " and "C" arresters, described below, are of the non-arcing metal type. The Type k4 A" Arrester. — The construction of this arrester can be best understood by reference to Fig. 11. It will be noted that there are seven independent cylinders of non-arcing metal placed side by side and separated by air-gaps. The cylinders, which are mounted on a marble base, are knurled, thus presenting hundreds of confronting points for the discharge. The dynamo terminals are connected to the end cylinders, and the middle cylinder is connected to the ground. The arrester is, therefore, double pole, that is, one arrester protects both sides of the circuit. When the lines become statically charged the dis- charge spark passes across between the cylinders from the line terminals to the ground. The non-arcing metal will not sustain an arc or become fused by it; hence with an arrester constructed of this material all possibility of vicious arcing and short circuits is avoided. 990 LIGHTNING ARRESTERS. The Type u C " Arrester. — This is similar to type "A," but instead of being mounted on marble it is enclosed in a weather-proof iron case for line use. The cylinders are placed in porcelain holders, as shown in Fig. 12. Fig. 12. Unit Lightning Arrester, Type"C," Showing Cylinders in Place. Tlie Garton Arrester. — In Fig. 13 a cross- section view is shown of the Garton Arrester. The discharge enters the Arrester by the bind- ing post A, thence across non-inductive resistance B, which is in multiple with the coil F, through conductors imbedded in the base of the Arrester, to flexible cord C, to guide rod D and armature E, which is normally in contact with and rest- ing upon carbon H, thence across the air-gap to lower carbon J, which is held in position by bracket K. This bracket also forms the ground connection through which the discharge reaches the earth. We have noted that the discharge took its path through the non-inductive resistance in multiple with the coil. This path is, however, of high ohmic resistance, and the normal cur- rent is shunted through the coil F, which is thereby energized, drawing the iron armature E upward instantly. This forms an arc between the lower end of the armature and the upper carbon H. As this arc is formed inside the tube G, which is practically air-tight, the oxygen is consumed, the current ceases, and the coil loses its power, allowing the armature to drop of its own weight to its normal position on the upper carbon. The arrester is again ready for another discharge. The S. K. C JLigrhtning- Arrester Equipment, manufactured by the Stanley Electric Mfg. Company of Pittsneld, Mass., consists of three essential parts. The Lightning Arrester proper is two nests of concentric cylinders, with diverging ends held in relative position by porcelain caps, as shown in cross section, Fig. 14. To the innermost cylinder the line is con- nected ; to the outer, the earth. The porcelain caps are provided with Fig. 13. THE GARTON ARRESTER. 991 grooves so placed as to make all spark gaps one-sixteenth, inch wide. Be- tween these grooves are sufficient perforations to allow the free circulation of air between the cylinders. If, on the occasion of lightning, the dynamo current follows the lightning, a current of air is at once established through the perforations between the cylinders, blowing the arc between the flar- ing ends where it is instantly ruptured. Between the line terminal and the ground connection there are three spark gaps, each one-sixteenth inch in width, making a total of three-six- VERTICAL-SECTION OF UGHTNINft ARRESTER Fig. 14. Fig. 15. teenth inch air-gap between either line-wire and the ground. At ordinary frequencies five thousand volts or over are required to jump the gaps of the arrester ; but at the frequency of a lightning discharge the sparking poten- tial is reduced to less than one-half of this. This phenomenon shows that the relative value of spark gaps cannot be expressed by "short" and " long," and their effectiveness as lightning protection cannot be measured by inches. The spark gaps of the arrester described are about double the widths ordinarily used, yet the sparking potential at lightning frequencies is less. The concentric cylinders provide large discharge surface, enabling the arrester to take care of all the heavy discharges, relieving the line completely. The second essential feature of the S. K. C. Lightning Arrester Equipment is a Choke Coil, so wound (Fig. 15) as to possess great opposition to the passage of lightning, yet practically no self-induction with currents of ordi- nary frequency. This coil is to be placed in the circuit between the lightning arrester and the apparatus to be protected. Introducing such a coil between the lightning arrester and the machine will offer practically no dis- turbing effect, either as to magnitude of the output or regulation of the system, and at the same time interposes enormous opposition to the passage of lightning discharges towards the machine to be protected. To remove even the slightest static discharge from the line, an instrument similar to the one illustrated in Fig. 16, called a "Line Discharger," when used with the apparatus above described, discharges the line com- pletely. The S. K. C. Line Discharger is a minute air-gap in series with a tube or tubes, filled with oxidized metallic particles, thus offering practically an infinite resistance to dynamic currents, yet allowing static discharges of extremely low potential to pass readily to earth. The Line Discharger is connected to the line as shown in Fig. 17. The number Fig. 16. 992 LIGHTNING ARRESTERS. of tubes required is determined by the voltage. As the Line Discharger will remove even the small static charge, it prevents the accumulation of such charges on the line which might prove dangerous. LilUUriJgt. 25-239 GROUND GROUND LIGHTNING PROTECTION FOR 6000 VOLT TRANSMISSION* LINE USING 6.k.c. arresters, choke coils and line discharges Fig. 17. Static Dischargers. Where circuits are entirely underground and ground connections are un- necessary, static dischargers are recommended. These consist of a number of gaps in series with very high resistances and are connected directly be- tween phases and adjusted to break down at slight increases over the line potentials. Fig. 18. Connections of Static Dischargers. AHRESTERS FOR HIGH POTENTIAL CIRCUITS. 993 ARRESTERS lOK IB I C. II POlEXTI.il. CIRCUITS. (Abstract of paper by Percy H. Thomas in Franklin Inst. Journ.) A lightning discharge is of an oscillatory character and possesses the property of self-induction; it consequently passes with difficulty through coils of wire. Moreover, the frequency of oscillation of a lightning dis- charge being much greater than that of commercial alternating currents, a coil can readily be constructed which will offer a relatively high resistance to the passage of lightning and at the same time allow free passage to all ordinary electric currents. A more complete method, a method of prevention rather than resistance, which is available for higher voltages, is the use of the static interrupter, which is substantially a magnified choke coil. Its function is so to delay the static wave in its entry into the transformer coil that a considerable por- tion of the latter will become charged before the terminal will have reached full potential. If a very heavily insulated powerful choke coil be placed in the lead of the transformer, when a static wave approaches electricity will begin to pass in small quantity and will pass in gradually increasing quantity at later instants of time, so that the coil will be, comparatively speak- , ing, gradually brought t*> full po- tential; meanwhile the volume of the static wave is being reflected and choked back and perhaps being discharged to the ground if there be a lightning arrester near. It is evident that this choke coil, to be effective, must be so propor- tioned as to delay the incoming wave enough so that the portion of the winding which has become choke coil . CHOKE COIL Fig. 19. Static Interrupter Protecting Transformer. charged when full potential is reached at the terminal shall be sufficient to withstand the strain of the full voltage of the wave. It is evident that such adjustment does not depend directly on the frequency or abruptness of the static wave, since both the transformer and the choke coil are similarly affected by the frequency. But a choke coil sufficiently powerful to accomplish this result satisfac- torily is found to be impracticable on very high potential circuits on account of the size, cost and interference with the operation of the system. How- ever, if the arrangement of the static interrupter be used, that is, if a con- denser be connected between line C H0K . E . C 0IL line anc * ground behind the choke coil -'wnnnp 1 nearer the apparatus to be pro- 1 LT ARR tected, this choke coil will absorb a considerable portion of the cur- rent actually passed by the choke coil, and the time required to pass sufficient electricity to charge the terminal will be much increased. With this arrangement a compara- tively small choke coil may be used. The condenser has a very small electro-static capacity, and has no appreciable effect upon normal operation, and yet has a very powerful effect on the static wave on account of its extremely high frequency. As in the case of the choke coil, the static interrupter must be roughly proportioned to the transformer winding to be protected. The condenser must also be suitable for the voltage between line and ground. If static interrupters be placed in each lead of high tension apparatus which may be injured by local concentration of potential, its windings will be amply protected against danger of short circuits from static wave either positive or negative. Such an arrangement is shown diagrammatically in connection with a transformer and a high tension generator in Figs. 3 and 4. CHOKE COIL umt Fig. 20. Static Interrupter Protect- ing High-tension Generator. 994 LIGHTNING ARRESTERS. Static Interrupters and Low Equivalent Lightning* Ar- resters. — A short description of the salient features of some actual lightning arresters and static interrupters will be given. The Low equivalent A. C Lightning* Arrester consists of a number of & inch air gaps between non-arcing metal cylinders in series with non-inductive resis- tance. A portion of the resistance, called shunt resistance, is shunted by a second set of air gaps called shunted gaps. The object of this arrange- ment is to reduce the amount of the series resis- tance through which the discharge must pass to ground. This arrester is diagrammatically illus- trated in Fig. 22. The series gaps withhold the line voltage and are chosen so as to break down at something between 50 per cent and 100 per cent rise of voltage above that of the earth. A portion of the series resistance is shunted by gaps so that the static discharge can pass around this portion, thus avoiding its resistance. It evidently is then necessary to suppress the arc from the generator which tends to follow through the shunted gaps. It is found that with a number of shunted gaps equal to the series gaps*the arc will be withdrawn from the shunted gaps by the shunt resistance when the shunt resistance does not exceed a proper value, which is a considerable portion of the total resistance in the arrester. _ A very marked gain in the reduction of the resistance offered to the discharge is therefore made by means of the use of shunted gaps and shunt resistance. It must be noted as well that no more voltage is required to cause a rise of potential to jump over all the series gaps and shunt than would be required to jump the series gaps alone, since on account of the shunt resistance the series and shunted gaps are broken down separately one after the other. A Static Interrupter consists of a choke coil in series with the line and a condenser connected between line and ground on the apparatus side of the choke coil. Cables. — The high tension electric cable is in principle no different from the electric air line, but has a different insulating material, paper or rubber Fig. 21. Choke Coil with Support, for Use with Low Equivalent Lightning Arrester. -SHUNTED GAPS- o o o o o o o o — o o o o o o o o \Z\AAHl' SHUNT RESISTANCE Fig. 22. Diagram of Low Equivalent Lightning Arrester. m place of air. This has the double effect of increasing its electrostatic capacity and changing the velocity at which waves progress. Ine in- creased electrostatic capacity tends to decrease the -peed, but as tne inductance of the cable is small this partially compensates tor the in- creased capacity. The differences from the air line are differences in degree only, and do not affect the passage of waves, reflection, resonance, etc. Consequently, no phenomena different from the air lines may be expectea as a result of static disturbances. Since the cable contains no coils of wire, no local concentration ot poten- tial will be found like that in transformer coils, and there is no oooasion tor the use of a static interrupter. ARRESTERS FOR HIGH POTENTIAL CIRCUITS. 995 In a general way it must be expected that the surging about of the energy which is stored whenever a line is charged will cause increased potential at certain points. This should be provided for by placing suitable lightning arresters at all points where important and vulnerable apparatus is located. There will also be local concentration of potential in windings connected to the circuits as the result of all static disturbances. Such transformer or other coils should always be either sufficiently insulated or protected by choke coils or static interrupters or by some other suitable method. Horn Tvpe. — This arrester was invented by Oelschlaeger for the Siemens & Halske, A. G., and like the Thomson arc-circuit arrester, its operation is based on the fact that a short circuit once started at the base, the heat of the arc will cause it to travel upward until it ruptures by attenu- ation. On circuits of high voltage this rupture sometimes takes a second or two, but seems to act with but little disturbance of the line. It has been used little in this country until lately when it has been installed on a few of the high voltage lines on the Pacific coast, and the results are so far highly commendable. The following figures, Nos. 23 and 24, show the application, one as applied to the line, and the other in diagram. The knee-shaped horns are of No. 0000 copper wire, one connected directly to the line, the other through a water resistance and choke coil to the ground. The horns are mounted on the regular line insulators, and for 40,000 volts the distance between the knees varies from 2\ to 3 or 3| inches. The water receptacle should have a capacity of at least 15 gallons, and users differ as to whether the water should have salt added. The water should, however, be covered by a layer of oil about one-eighth inch deep in order to prevent evaporation. The choke coil can be made of about eighteen turns of iron wire wound on a 6-inch cylinder. Fig. 23. Construction of the Horn Type Arrester as used by the American River Electric Company. 996 LIGHTNING ARRESTERS. Care should be taken that the knee is not too sharp, or the arc is liable to reform after being once broken: again, the horns should not lie too flat, or the arc will strike down as shown in Fig. 24. The curve of the knee is not alike for all parts of the line, but depends on the line constants, and will have to be fitted to each case. TO MAIN LINE ^6 INCHES GROUNDED ON PIPE LINE COPPER STRIP ONE INCH WIDE fc^r XZ 6 TOO SMALL ARC HOLDS ON KNEE TOO SHARP ARC STRIKES BACK Fig. 24. Arrangement of the Parts of a Horn Type Lightning Arrester, the Two Small Diagrams to the Right Showing Faulty Construction of the Horns. — N. A. Eckert. ELECTRICITY METERS. Revised by H. W. Young. Meters for measuring the amount of electrical energy furnished to con- sumers are known as recording or integrating watt-hour meters and are made in several different forms to meet the varying conditions. The regis- tration of an integrating meter must be very accurate to meet commercial requirements owing to the fact that any errors which may be present are cumulative and even a small percentage error will, after a lapse of time, become relatively important from a pecuniary standpoint. The accuracy must be especially high at the lower end of the curve owing to the fact that for the larger part of the time the actual load is but a small percentage of the meter's capacity, and a meter which shows inaccuracy at this point cannot be a profitable investment for the central station for the reason that the tendency is to under register rather than over register. Action of Integrating* Meters. — The action and operation of an integrating meter may be likened to that of a small direct-connected motor generator set in which the current and potential coils are considered as the motor element and the disk and the permanent magnets as a magneto- generator with a short-circuited disk armature. The work expended by the motor is absorbed in driving the short-circuited generator and overcoming friction in the bearings and registering mechanism. In a perfect meter (or motor generator) all the work would be expended in driving the disk or generator — friction being absent — in which case a direct ratio would exist between the speed and the energy passing through the motor system, thus giving a meter absolutely accurate throughout its entire range. It is, however, impossible to entirely eliminate friction, but it will be seen that the more perfect the meter is, the greater will be the ratio between the work expended usefully in driving the disk or armature of the generator and that expended in overcoming friction; or, in other words, the "Ratio of Torque to Friction" in the meter will be high. Meter manufacturers, recognizing this essential feature, endeavor to make this ratio of torque to friction very high by efficient design of the measuring elements and reduc- tion of friction in the bearings and registering mechanism. Direct- Current Commutator Type Meters. The best known of the direct- current meters is the commutator type con- sisting of a small motor driving a registering mechanism. There are usually two series coils wound with comparatively few turns of heavy wire and prac- tically surrounding a pivoted armature containing several coils of fine wire suitably connected to a commutator on which bear small brushes. In series with the armature is a comparatively high resistance and a light load or friction compensating coil. The stationary series coils are connected in series with the load and the shunt circuit consisting of the armature and its resistance is connected across the line. The construction employed gives a driving torque proportional to the energy flowing in the circuit, and to secure correct registration it is necessary for a retarding torque to be provided which will be proportional to the driving torque. A controlling force varying directly with the speed is obtained by causing an aluminum or copper disk to pass between the poles of permanent magnets whose fields induce "Foucault" or eddy currents in the disk. The interaction between the fields of these eddy currents and the field of the permanent magnets produces a retarding torque varying directly with the disk speed. With such an arrangement of driving and retarding torques a rotation is produced which is always proportional in speed to the driving torque and, therefore, to the energy passing through the measuring coils. As tne measuring elements do not employ iron and are practically non-inductive, the meters can be used on either A. C. or D. C. circuits. 997 998 ELECTRICITY METERS. Thomson Recording: Wattmeters. (General Electric Company.) These meters (Fig. 1) are of the commutator type previously described, and the salient features claimed are as follows: High torque, direct-reading registers, dust proof construction, small size commutator, gravity brushes, adjustable shunt field coil, inter- changeable on D. C. and A. C, high accuracy, heavy overload capacity, jewel bearings. Bearing's. — The top bear- ing consists of a simple brass plug having a hole of sufficient size to allow free rotation of the armature shaft. The lower bearing consists of a hardened steel pivot made of piano wire and resting on a spring supported sapphire or diamond jewel. This insures a bearing having a low friction value and long life. During shipment the jewels are pro- tected by a special armature locking device which, when the jewel is backed away from the pivot, automatically locks the moving element. Westing-house 1>. C Integrating: JVEeters. These meters (Fig. 2) are of the same general type as the Thomson, but differ in me- chanical construction. The sa- lient features claimed are prac- tically identical with those of the Thomson meter. The lower bearing is. how- ever, of an entirely different type, consisting of a small, highly polished steel ball rest- ing between two sapphire jew- els, one of which is secured in a removable jewel screw. The idea of this form of bearing is to present constantly changing contacts between the ball and its jewels owing to the attendant rolling action, thus securing a long useful jewel life and increased accuracy. During shipment the disk is locked Lin position by a suitable locking device operated from the top bearing. Fig. 1. Thomson Recording Wattmeter (Cover Removed). Duncan Uleters. This meter (Fig. 3) in common with the Thomson and Westinghouse forms, is of the commutator type and practically the same claims are made as for the other forms. It differs in the method in the friction or light load compensation in that the auxiliary field coil is provided with taps brought out to a multi-point switch. This arrangement enables the auxiliary torque to be varied by cutting in or out one or more coil sections. The lower bearing is of the "visual" type designed to permit inspection of the jewel and pivot while the meter is in operation. The pivot is of hardened steel piano wire and is securely held in position. ( During trans- portation the jewel post is lowered, thus locking the disk in position. WESTINGHOUSE D. C. INTEGRATING WATTMETER. 999 Induction Type Alternating* Current Integrating* Wattmeters. Principle of Operation. — The single-phase induction wattmeter is in principle and operation analogous to a single-phase induction motor having a stationary shunt and series winding so related and located as to produce a rotating field acting upon a closed rotable secondary. In the induction meter the secondary consists of a light aluminum disk. The shunt winding, con- sisting of a large number of turns of fine wire wound on a laminated iron core, is highly inductive and its current lags approximately 90 degrees behind the impressed or line voltage. The series winding consisting of but Fig. 2. Westinghouse D. C. Integrating Wattmeter (Cover Removed). a few turns of comparatively heavy wire, has low self induction, and on non- inductive load (such as incandescent lamps alone) the current producing the series magnetic field will be in phase with the impressed or line voltage. Thus the magnetic field produced by the shunt winding will lag approxi- mately 90 degrees behind that of the series winding on a non-inductive load. With this relation of the two fields at the instant of time when the current in the series coil is greatest the current in the shunt coil is the least. (If it were not for the iron loss and small resistance or copper loss in the shunt circuit, the angle would be exactly 90 degrees.) During a portion of each alternation of the circuit the series coil helps the flux of one pole of the shunt field, opposing the other, and during another portion of the alternation it haa the opposite effect; these reactions being combined in such a way as to 1000 ELECTRICITY METERS. give a general shifting of the lines of force in one direction — that is, pro- ducing a rotating field. Rotating* field. — That the shunt and series fields combine to form a rotating field may be more clearly understood by tracing the action or relation of these two fields for a complete cycle by one-quarter periods of the same. Referring to Fig. 4 and noting that the two poles of the shunt coil magnet are designated by the letters A — Ai and C, and the poles of the series coil magnet by B — D, a clear statement of the relation of the fields by one- quarter periods is given in the table shown in Fig. 5. The signs given in this table represent the instantaneous magnetic values of the poles indicated, and it will be observed that both the positive and negative signs move constantly to the left indicating a shifting of the field in this direction, the process being repeated during each cycle. Driving* Xorque. — This continuous motion of the field induces eddy currents in the aluminum disk which react to produce rotation in the same manner as in the rotor of an induction motor. The rotary field being a combination of the series and shunt fields, the torque on the moving element or disk will be directly proportional to the energy flowing in the circuit. Retarding* Torque. — With a driving torque propor- tional to the energy flowing in the circuit it is necessary, in order to obtain steady rota- tion, for a retarding torque to be provided which will be proportional to the driving torque. A controlling force varying directly with the speed is obtained by causing the aluminum disk to pass be- tween the poles of two perma- nent magnets whose fields induce "Foucault" or eddy currents in the disk. The interaction between the fields of these eddy currents and the fields of the permanent magnets produces a retarding torque varying directly with the speed of the disk. With such an arrangement of driving and retarding torque a rotation is produced which is always pro- portional in speed to the driving torque and therefore to the energy passing through the operating coils. Wattmeters on Inductive Circuits. — Assuming that the current producing the shunt field lags exactly 90 degrees behind the line voltage and neglecting for the moment the iron loss and resistance loss in the circuit, it will be seen that when the load is non-inductive (such as offered by incan- descent lamps) the current of the series coil will be in phase with the line voltage and the shunt and series fields will differ in phase by exactly 90 degrees. From the table (Fig. 5) it will be seen that this gives a maximum pull on the disk. If, however, the load is purely inductive having zero power factor, the current in the series coil will lag 90 degrees behind the line voltage and will be in phase with the current in the shunt coil. Under these conditions the relation between the fields for each one-quarter period of a complete cycle is shown in the table on page 1002. Fig. 3. Duncan D. C. Recording Watt- meter (Cover Removed). INTEGRATING WATTMETERS. 1001 Scries Element O „ C L® ^B O / / i O^- -"'O x Shunt Element Fig. 4. — -'O Start A + B c D >i Period - !4 Period - ^ Period - Full Period - 5 I I I Fig. 5. Table Giving Relation of Fields by One-quarter Periods. 1002 ELECTRICITY METERS. At start . . At i period . At h period . At I period . At full period When A is + + B is 4- + Cis Dis + + A x is + As no progression or shifting of the field occurs, there is no rotation of the disk and thus the meter will not record when the current in both the series and shunt coils is 90 degrees out of phase with the impressed voltage; hence, the meter will record true power whether the load be inductive or non-inductive. Power Factor Compensation. In the preceding diagrams it was demonstrated that for correct registration on any power factor, exactly 90 per cent phase relation between the shunt and series fields must be obtained. Consequently, compensation must be made for the small decrease of this angle caused by the copper and iron losses in the shunt circuit. This compensation is usually obtained by placing one or more short cir- cuited turns (or secondary) of conducting material around the projecting pole C of the shunt electromagnet, producing an induced magnetic field which, acting with the shunt magnetic field, produces a resultant field lagging behind the field of the series coil. By varying the position or resist- ance of this short-circuited turn (or secondary) the compensation necessary to obtain the exact 90 degrees phase relation may be obtained. This method of securing the resultant field can be better under- stood by referring to Fig. 6 in which: OA represents the voltage of shunt coils. OY represents current passing through shunt coils. YOA represents angle less than 90 degrees due to iron and copper losses in shunt coils. OS represents induced voltage of short-circuited turn K and exactly opposite in phase relation to that of OA, but very small in value; the current passing through the short-circuited turn K being in phase equal and approximate to OC. This current OC and main current OY have a combined magnetizing effect on the iron core, which effect is found by forming the parallelo- gram OC — XY when OX is the resultant effect now practically at right angles to the impressed E.M.F. of the circuit. By raising or lowering, thus changing the position of the short-circuited turn, the magnetism of the shunt field can be shifted back to the proper angle, giving the 90 degree phase relation and adjusting the meter so as to read correctly under all conditions of power factor. Note. — This power factor compensation holds true only for approxi- mately the frequency for which the meter is adjusted and if highest accuracy is expected, wattmeters should not be used on inductive loads having a frequency variation from normal of more than 10 per cent plus or minus. minimizing* Effect of Voltagre Variations. — It is desirable that induction meters be capable of operating over a wide voltage variation without impairment of accuracy, and freedom from error due to voltage variations is accomplished by the design of the shunt magnetic circuit. By referring to Fig. 4 it will be seen that the shunt magnetic circuit is so arranged that the greater portion of the magnetic lines generated by the shunt winding are shunted across the narrow air gaps FF and do not pass through the disk, thus cutting or damping its action and thereby impairing the accuracy. While the exact leakage across the gaps cannot be accurately determined, it is a large proportion of the total flux generated so that a comparatively wide variation from the normal voltage has practically no effect on the meter's registration owing to the small percentage of damping flux which is produced. Fig. 6. Diagram of Re- sultant Field. "WESTINGHOUSE INDUCTION WATTMETERS. 1003 Figure 7 illustrates a typical voltage curve of an induction wattmeter. It will be noted that a voltage range from 50 per cent to 125 per cent of normal voltage does not materially impair the accuracy. Westing-house Single-Phase Induction Wattmeters. These meters (Fig. 8) are of the rotating field type previously described and the salient features claimed are as follows: High ratio of torque to 1 K DC I7 00 vc LT/ G£ CO ^VE AT CONSTANT LOAD AND FREQU ENCY 3 en 20 40 60 80 100 120 140 1 voLts Fig. 7. friction; high ratio of torque to weight; improved lower rolling ball bearing; improved self oiling top bearing; light load adjustment located in leakage gap of shunt coil and unaffected by flux of series coil; mechanical power factor and frequency adjustment; accurate on non-inductive or inductive loads; freedom from effect of stray fields; permanent magnets magnetically shielded; light rotating element (15 grammes); unaffected by voltage variation from 50 per cent to 125 per cent of normal; unaffected by wide variations in wave form and frequency; freedom from rattling or Fig. 8. Type "C," Westinghouse Single-Phase Induction Meter humming; dust proof; light running, gold plated, non-corrosive registering mechanism; meters shipped ready for installation without preliminary adjustment. Westing-house Polyphase Induction Wattmeters. These meters consist of two single-phase elements which are mounted in ft single case and actuate a common registering mechanism. 1004 ELECTRICITY METERS. Figure 9 illustrates a House Service Polyphase Meter and Figure 10 the Polyphase Switchboard Service Meter. POLYPHASE INTEGRATING WATTMETER. WESTINGHOUSE ELECTRIC 8c MFG. CO. PITTSBURG, PA., U S.A. Fig. 9. Westinghouse Polyphase Induction Meter (House Service). Fig. 10. Westinghouse Polyphase Induction Meter (Switchboard Service). SINGLE-PHASE INDUCTION WATTMETERS. 1005 Thomson High Torque, ft ingle -Phase Induction Wattmeters. (General Electric Co.) These meters (Fig. 11) are of the same general type as the Westing- house, but differ in mechanical con- struction. The salient features claimed are practically identical with those of the Westinghouse meters. The bearings, however, are of a different type, being of the same con- struction as employed in the Thomson D. C. meter. The torque is of high value, thus giving a high ratio of torque to weight. During shipment the armature is locked in position in a manner similar to that of the Thom- son D. C. meter. Thomson Polyphase Induc- tion Wattmeters. ((Q) These meters (Fig. 12) in common with the Westinghouse form, consist of two single-phase elements in a sin- gle case. Fig. 11. Thomson High Torque Single-Phase Induction Meter (Cover Removed). Fig. 12. Thomson Polyphase Meter, Glass Cover. Type " K " Single-Phase Induction Wattmeters. (Fort Wayne Electric Co.) These meters (Fig. 13) are also of the rotating field type, but employ a drum-shaped rotor instead of a disk. The light load adjustment is affected 1006 ELECTRICITY METERS. by an adjustable starting coil which can be shifted to give the necessary compensation for friction effect at light loads. The salient features claimed are practically identical with those of the Westinghouse and Thomson meters. The upper and lower bearings are simi- lar to those employed in the Thomson meter. Type "K" Polyphase Induc- tion Wattmeter*. These meters (Fig. 14) consist of two single-phase measuring elements mounted in a single case and acting upon a single drum-shaped rotor. Hang'amo I>. C Integrating* Meter. This meter (Fig. 15) is a mercury con- tact motor meter of a type that has been used to a greater extent abroad than in this country. In common with all motor type inte- grating meters the Sangamo contains the three necessary elements, namely, a mo- tor producing a driving torque; a gen- erator providing a load or drag varying with the speed, and a registering mech- anism arranged to integrate the instantaneous values of the electrical energy passing through the measuring coils. Fig. 13. Type " K " Meter (Cover Removed). Fort Wayne Elec. Co. Fie. 14. Type "K" Form MAB Wattmeter — Half Front View, Case off. Fort Wayne Elec. Co. • SANGAMO D. C. INTEGRATING METER. 1007 Principle of Operation. — The principle of operation maybe under- stood by referring to Fig. 16, and the following description: A — A are the poles of an electromagnet energized by the potential coil which, through a resistance, is connected directly across the line, thus forming the voltage element of the meter. E is a soft iron bar located just above A — A and forming the air gaps in which the copper disk D is located. This copper disk is connected in series with the line and forms the current element of the meter. In capacities exceeding 10 am- peres the disk only carries a certain por- tion of the main current which is obtained by inserting a shunt in series with the line and allowing but a small portion to pass through the mercury and disk. These & voltage and current elements form the driving motor element of the meter. B is an aluminum disk so arranged that its edges pass between the poles of two per- manent magnets, F — F thus forming the generator or load element of the meter. D and B are mounted on a common shaft which is suitably pivoted or suspended. The third element of the meter, namely, the registering mechanism, is not shown, but, in common with other forms of motor meters, is driven by a suitable gearing actuated by the rotable shaft. Fig. 15. Sangamo Direct-Cur- rent Meter, Case off. 0* Fig. 16. Elementary Diagram of Sang-amo I>. C. meter. From the arrows on A — A it will be seen that the field generated by the potential coil threads the two air gaps and in doing so cuts or passes through the copper disk D. The disk D being in series with the load is, therefore, carrying a current which, due to the position of the leading in contacts, passes across the magnetic fields produced by the magnet poles A — A and 1008 ELECTRICITY METERS. is at right angles to this field. As is well known, a conductor free to move and carrying a current whose direction of flow is at right angles to a fixed field will tend to move out of the fixed field. As the disk moves from its initial position the current enters at a new point on the periphery of the disk which is again impelled forward, and this constant change in point of current entrance to the disk produces a con- tinuous rotation. It will thus be seen that the meter, in common with the Westinghouse D. C. meters, operates as a simple motor driving a magneto- generator having a short circuited armature. The Sangamo meter differs, however, in its construction from that em- ployed in the commutator D. C. meters in that the voltage element is station- ary rather than rotable; the current element being rotable rather than stationary and instead of employing a commutator and brushes to lead current in and out of the rotable element, or armature, it is submerged in mercury contained in an insulating chamber having contact pieces at each edge to which the circuit connections are made. Figure 15 illustrates a meter as actually constructed. The mercury is contained in a dome-shaped chamber and not only serves to conduct the current to and from the armature, but also tends to buoy up the disk and relieve the pressure on the lower bearing. The full load adjustments are accomplished by varying the strength of the magnetic field through which the disk passes, and the adjustment at light load is accomplished by a compounding coil so located as to assist the field generated by the potential coil. Sang-amo A. C Meter. This meter has the same general appearance and operates upon the same principle as the D. C. meter, but differs somewhat in the arrangement of the measuring elements. In the A. C. meter the main current energizes the stationary electromagnet and the shunted or potential current passes through the copper disk. Compensation is provided for light load and inductive load. WRIGHT DISCOUNT MIIXKH. This instrument is for use in connection with a watt hour meter for de- terming the maximum use of current during any given period ; or may be used without the watt-hour meter in connection with any electrical device for which it is desired to know the maximum use of current, either direct or alternating. It is slow acting so as to take no account of momentary spurts, such as starting an elevator or street car, and is rated to record as follows : If the maximum load lasts 5 minutes, 80 % will register ; If the maximum load lasts 10 minutes, 95 % will register ; If the maximum load lasts 30 minutes, 100 % will register. The following figure shows the working parts in theory, which, being of glass and liquid, are placed in a cast-iron case, with a glass front to permit reading. As shown, one leg of the circuit passes around a glass bulb which is hermetically sealed, and connected to a glass tube holding a suitable liquid. rr hw T, Terminals. h w, House wires. r w, Resistance wire. H B, Heated bulb. CW) ff>* A B > Air BulT) * HB j^\ \ it, Indicating tube. L, Liquid. > Direction. Wright Discount Meter. METEK BEAMNGS. 100S The heat due to the current passing in the circuit expands the air in the bulb, which forces the liquid down in the left column and up in the right. Should the quantity of heat be such as to force some of the liquid high enough, it will fall over into the central tube, where it must stay until the instru- ment is readjusted. The scale back of the central tube is calibrated in am- peres on the left and in watts on the right. After reading and recording the indication for any period of time, the liquid is returned to the outer tubes by simply tipping up the tubes, etc., which are hinged at the top connections for the purpose. The readings of the demand meter or discount meter, either of which names are used, together with those of the watt-hour recording meter, furnish a basis for a more rational system of charging for electricity than has been customary. This, subject is being taken up by many of the larger electricity supply companies. The instrument is handy to use in circuit with a transformer to show how the maximum demand compares with the transformer capacity ; also on feeders and mains to show how heavily they may be loaded. Fig. 17. Visual Pivot Type Bearing. Fig. 18. Pivot Type Bearing. WEETER BEARI]¥G§. II EG LITERS A \ I> COIHE- MUlAIOR§, Two forms of lower bearings are in general use in both direct and alter- nating-current meters. Figs. 17 and 18 represent the pivot forms consisting of a hardened and highly polished steel pivot resting on a cupped sapphire, or on a rings tone end-stone or cupped diamond jewel. Figure 19 is a rolling type ball bearing formed by a small hardened and polished steel ball resting between two jewels, one of which is attached to 1010 ELECTRICITY METERS. DISCS OF BILLIARD CLOTH SOAKED IN JEWELER8 OIL the armature shaft and the other to a fixed support. By this construction a rolling action is secured as contrasted to the rubbing action of the pivot bearing. Both types of bearings are exten- sively employed by meter manufactur- ers and each has strong advocates. The pivot form of bearing is invariably supported by a spring suspension, while with the ball bearing the spring is only resorted to in the direct-current meter? having comparatively heavy moving elements. The registering mechanisms of the various types of meters are quite simi- lar in appearance, differing principally in the method of construction. Fig. 20 illustrates a typical form of register- ing mechanism employed in both D. C. and A. C. meters. - To reduce the variable nature of the contact surfaces of the commutator and brushes it is customary to em- ploy non-oxidizing metal in the con- struction of these elements, thus reduc- ing to a minimum changes at this point. Fig. 21 illustrates the damping disk, armature and commutator mounted on the rotable shaft. JEWEL SCREW .CLAMPING NUT Fig. 19. Rolling Type Ball Bearing. Hie Prepayment Wattmeter. The prepayment idea for the purchase of practically all forms of commodities is rap- idly growing, for the vending of practically all forms of com- modities, and is now receiving recognition in the electrical field. Like the installment plan of payments, the prepay- ment meter appeals to a class of people who are accus- Registering Mechanism. tomed to receive and spend their money in small quan- tities. The success of gas companies has been greatly aided and furthered by the prepayment meter, and its use in the electrical field should prove as great a suc- cess as it has proven in this field. Prepayment meters are especially applicable in sup- plying energy to customers whose total consumption is relatively small and the collection of whose bills is a very considerable proportion of the total revenue de- rived. Their use reduces the amount of bookkeeping and unavoidable monetary loss due to poor accounts, for the service is such that before securing light it is necessary that payment be made. This system, there- fore, automatically collects its own bills, registers the actual consumption, and when the energy prepaid for is consumed, automatically disconnects the service. In installations such as flats, dormitories, barber shops, cafe's, saloons, boot-blacking establishments, cigar stands, rented nouses, or in any other installations where Fig. 21. Rotating the volume of energy consumed is necessarily small, Element of D. C. the prepayment meter will be found extremely useful. Commutating Type Central stations supplying towns having a large "float- Meter, ing" population, such as seashore resorts or college THE PREPAYMENT WATTMETER. 1011 towns, where the rapid shifting of population renders difficult the following of accounts, will find the prepayment meters extremely useful. Another use for the prepayment meter is in the collection of old accounts. Central stations frequently have a considerable number of customers who are usually backward in payments, although they ultimately pay their bills. One method of forcing such customers to pay back bills is to threaten discontinuance of service, but this method is only resorted to as an extreme measure, owing to the resulting unpleasantness and very possible loss of a customer. On the other hand, a central station cannot afford to have its legitimate revenue tied up even with customers who will ultimately pay. An effective way to collect these old bills, and at the same time continue the service, might be to install a prepayment meter adjusted for a "higher rate per kw.-hour than the regular rate. For instance, assuming the normal rate to be 10 cents per kw.-hour, the meter may be set at 15 cents per kw.- hour, so that the customer not only pays for the energy being consumed, but also gradually pays up the old bill on the installment plan. The majority of customers would undoubtedly prefer this method of paying up Fig. 22. General Electric Prepay- ment Wattmeter. Fig. 23. General Electric Prepay- ment Wattmeter (Internal View). old bills to being forced by threats of discontinuance of service. After the account has been settled, the meter can be reset for the normal rate per kw.-hour. At the present time many central stations are unable to connect a con- siderable number of relatively small consumers, owing to the fact that the amount of energy used by each customer would be so small as to hardly justify the collection and accounting expense, which would be a very con- siderable percentage of the total receipts. For example, many station managers would hesitate to connect up consumers whose bills would prob- ably not average over $1 per month, and, furthermore, these consumers do not understand and will not agree to a fixed minimum charge. However, assuming that the total revenue from such a consumer would average $12 per year, and assuming the cost of generation and distribution is one-half the gross receipts, it would leave a remainder or profit of $6, less the interest, collection and maintenance cost. While the gross profit would not be very large, yet the percentage is very satisfactory, and there is the additional advantage that a large majority of these new customers would gradually use larger amounts of energy and in time come within the class of desirable customers. The use of electricity increases with the knowledge of its advantages, and 1012 ELECTRICITY METERS. there is no better way of introducing its use especially with the smaller customers, than with the prepayment meter. With the prepayment meter, differential rates can easily be made, owing to the fact that the rate per kw.-hour is not shown on the meter bills and the central station may, therefore, place meters adjusted for dif- ferent rates to meet the various conditions which arise; for instance, a long-hour consu- mer could be supplied through a meter adjusted at a lower rate than the short-hour consumer. This method of differential rates, though not in general use, is feasible for the reason that with a prepayment meter consumers feel they are purchasing light and not kw. -hours. Another use for the prepayment meter is in connection with electric cooking and heating appliances, which frequently are supplied with energy from a separate circuit at a different rate than is charged for lighting. These appliances may be supplied through a prepayment meter, and this system has the additional advantage of permitting the consumer to determine ac- curately just what the electric cooking or heat- ing outfit is costing for the results obtained. By prepaying the meter for a definite amount it can be used as a time switch to automatically turn off arc lamps, electric signs and store win- dow lighting. The construction of several forms of commer- Fig. 24. Prepayment At- cial prepayment meters is shown in the ac- taehment for General companying illustrations. Essentially the pre- Electric and Fort Wayne payment meter consists of a measuring element Wattmeters. used in conjunction with a special register, au- tomatic switching device and coin chute. Fig. 24 illustrates a separate attachment which can be used in conjunction with a specially arranged standard meter and located apart from the meter itself. Fio. 25. Fort Wayne Prepay- ment Wattmeter. Fig. 26. Westinghouse Prepay- ment Wattmeter. INTEGRATING WATTMETER TESTING. 1013 INTEGRATING WATTMETER TESTING. It is quite generally recognized that integrating wattmeters can only be maintained in an accurate and efficient condition by comparing them at certain intervals with known standards, and it is obvious that the standards for this purpose should be highly accurate. To avoid a multiplicity of instruments they should have a wide operating range which may be obtained primarily by a long scale, and where possible the range should be further in- creased by combining several current and potential capacities in one meter. To combine laboratory accuracy with the speed necessary in commercial work, two sets of standards should be provided which may be designated as "primary" standards for extreme accuracy and "secondary" or working standards for use directly with the service meters. Checking* of Secondary Standards. — All secondary standards should be frequently checked with the primary standards, the frequency of such checking varying largely with local conditions. As a rule, however, it is advisable to check the secondary standards at least once a month, especially when such standards consist of indicating meters, owing to the fact that all portable indicating meters are more or less delicate and the rough usage attendant on commercial testing is liable to materially change the calibration. To compare the calibration of a secondary stan- dard indicating wattmeter with the primary stan- dard, it should be connected into the circuit as shown in Fig. 28, having the current coils of the meters in series and the shunt coils in multiple with each other. Care should be taken to have the shunt coil of each meter connected to the same point or source of potential to avoid the possibility of one meter measuring the shunt loss of the other. Testing* Load. — The load for the test can readily be obtained by a bank of incandescent lamps so arranged that any value from zero to full -load current may be easily and quickly obtained. The load should be taken from a source of supply having as little voltage variation as possible on account of the effect of rapid fluctuations on the reading of indicating Fig. 27. Westing- house Prepayment Wattmeter t (Inter- nal View). LOAD PRIMARY 'SECONDARY STANDARD STANDARD RESISTANCE OOOOO Ooo=ooO LINE Fig. 28. Connections for Comparing Secondary with Primary Standard. 1014 ELECTRICITY METERS. meters, it being somewhat difficult to secure accurate readings on a circuit having a badly fluctuating voltage. A convenient arrangement of load is shown in Fig. 29, and consists of a bank of lamps of different candle- power ranging from 4 to 50 C.P., these lamps being arranged in connection with single-pole, single-throw switches so that the smaller sizes may be thrown in circuit individually and the larger sizes in groups. The arrange- ment shown may, of course, be varied to suit local conditions. In circuit with a portion of the lamp bank is placed an adjustable re- sistance or rheostat for use in obtaining exact current values and also to assist in maintaining a constant load. The water rheostat is very con- venient for this class of work as the load can be varied quickly and with perfect uniformity. The resistance of the water rheostat can be readily changed to almost any value by changing the strength of the solution. Having made connections as above, it is now only necessary to take the readings on the portable meter at convenient points and to compare these LAMP BANK SINGLE POLEV //•///// SWITCHES -^ l I I 1 I I I I- ADJUSTABLE RESISTANCE^ *— i PRIMARY SECOND- STANDARD STANDARD RESISTANCE OQOOQ OO ooO Fig, 29. Lamp Bank and Connections for Comparing Secondary with Primary Standards. readings with the true values as given on the primary standard. It is considered good practice to check the portable meter at each of the marked points on the scale, simply estimating the error of the intermediate points, thus showing the error very closely at all points of the scale. Checking* Calibration of Portable Standard Integrating Wattmeter. — If the portable rotating standard meter is used as a second- ary standard, it should be checked with a primary standard wattmeter from time to time and for this purpose should be connected in the same manner as the indicating standard shown in Fig. 28. To make a comparison of the rotating standard with the primary standard it should be properly connected and placed in series with a primary standard of approximately the same ampere capacity. liig-ht Load Test. — The load should now be maintained constant at approximately 4 per cent of full load and the pointer revolutions of the rotating standard timed by a stop watch. Having obtained the time con- INTEGRATING WATTMETER TESTING. 1015 »umed in making a certain number of pointer revolutions, the watts should be computed by the formula applying to the particular meter under test. Ifull-JLoart Test. — The meter may be tested on other loads ranging from the light load to full load, but as the calibration curve of the rotating standard from light load to full load is practically a straight line, it is unnec- essary to take readings at other points than light and full load unless extreme accuracy is required. If this is desired, readings may be taken at several intermediate points, from which readings a curve may be plotted giving the exact calibration of the meter at all points. Selection of ^Primary Standard meter Capacity. — In com- paring secondary with primary standards, care should be taken to select the windings of the primary meter having a capacity nearest that of the meter under test, in order that it may be used at the highest possible part of the scale. This rule also applies to the comparison of service meters with secondary standards. Testing; Service Ifleters. — For the testing of service meters, either the "portable indicating" meters may be employed in conjunction with a stop watch and the reading computed by the use of a calibrating formula, or the meter may be compared with a portable standard integrating wattmeter. To use either of these methods the standard should be connected in circuit with the service meter as shown in the diagrams usually accompanying each meter. Where meters operating from series and voltage transformers are to be tested, it will usually be found advisable to test them as 5-ampere, 100- volt meters without using the transformers. If such meters are to be tested under the running load, the standard may be connected in the secondary transformer circuit of the meter under test, using the 5-ampere, 100-volt coils of the standard. Testing* Service HEeters with Standard Indicating* Meters.— To conduct a test with the indicating meter it will be necessary to hold the load as constant as possible and while noting the reading of the standard, count the revolutions of the disk of the meter under test, taking the time by means of a stop watch. To eliminate personal errors several readings of at least one minute each should be taken and averaged. To compare the reading of the meter with the standard, it is necessary to use a formula pertaining to the particular meter under test. "Use of Stop Watch. — When employing the indicating wattmeter method it should be remembered that the stop watch is not infallible and should be frequently checked by comparing it with the second hand of a good clock. For this purpose a clock in which the pendulum beats seconds or half seconds should be used, starting the watch with a certain beat of the pendulum and having allowed the watch to run several minutes to elim- inate personal errors, it should be stopped on the same beat of the pendulum on which It was sorted. A little practice will enable the operator to check the watch within .1 of a second without difficulty. Testing- Service Meters with Portable Standard Inte- grating' Wattmeters. — If the integrating standard is used for testing single-phase service meters, the operation is much simplified, as the use of the formula and stop watch can be eliminated. To conduct a test by this metnod, the standard should be connected as shown in Fig. 30, and the connections so arranged, if possible, that the capacity of the standard will be the same as that of the meter under test. The proper connections having been made, the load should be adjusted to the desired value and a direct comparison made of the number of revolutions of the meter under test with the number of revolutions shown on the counter of the standard. In common with the indicating standard method, readings should be taken for at least one minute to eliminate personal errors. The percentage of error in the meter under test may be found directly by dividing the number of revolutions of the service meter by the number of revolutions made by the standard meter ; that Is, if the meter under test makes 10 revolutions while the standard meter shows 10.4 revolutions, the ratio would show the meter under test to be approximately 4 per cent slow. The above applies only when the meter under test has the same ful'-ioad speed as the standard. In order that the standard meter may be conveniently employed in testing meters in which the full-load speed is other than twenty-five revo- lutions per minute, the following table has been prepared as applying to Westinghouse, General Electric and Fort Wayne meters. By the use of 1016 ELECTRICITY METERS. 00 -* 00 00 CO t^ OS CON OS rH ihPh © CO 00 tH r-i-i id rH HH Ol* Tj< 0) i o © fi ~ — i-i Ph © -* OOi-H t^TH to CM CM as o CM (N C<| CM CO t^ i> (Mt}< t^CO ,_< »o TjiiO CQ 2pH$ OS tHOS rfHi-i l^i-H CM CO t> OC^f CM rH CM Tfl o © fi *— i P-i « tO OS oil— • COtH CO l^ O^ 05 00 i-i I> t^ i 00 lOtO CMrH CO o CM © COO i-i CM ■fi& CM CO i—l CM . i-H CMrH <* oq CM CO i-H CM CO (MCO ioco 8 IO OSTt <* (N t- o OO lOCM 00 rHlO t^ CO OSfl, 0) id CM -* oCM ^ 00 <* IO •~J CO O cj OSp^ o CO i— 1 i-i CM OOi-I 00 rH COrH to CM 9 ° CM CO 1—1 (M (N CO CM iq oo ioco *S 55 c CO © «5H OJ(M -* 00 cOiO -* CO o Ospn a> CO i-H i-HCM 00 i-i 00 rH COrH id OS OS CO 00 CO CO* ,_j i-IcM Os'i-i 00 !-H COrH id CM CM CO CM CM CO oiad'uiy -ia^ tOOOO tHCMtJh lOOOO rH « c > >> • iO^ >~2 to to »o to iO ' o ; oo o : OO d*X 03 3 asno qSupsa^V * ou )^ia l^iau ^O INTEGRATING WATTMETER TESTING. 1017 CO s rHpH © tH o V CO CM ,_l ■* ,_! r^ Tfl »-l +2 50 M< CO ^ © c CO 05 CO CO © CO LO flCD •do OiQ. o Tp LO <* LO o 5 d CIQl, >■§ tf c3 g oo© OO i| © O CO CO CO coco 2§ -£*! -O'BCl^Q lOOO LOOO LOO LOO •dray rH^H HTf CMLO CMLO 43 M*X„ « >d*£ 5 > sd © oj © © J» 3 * ©73 -^ "Sdd © d© d ^-O 2 g.03^ Sd £ |So d 2 8^4? 9 B 1018 ELECTRICITY METERS. this table any one of the three makes can easily be tested with the one standard. In explanation of the use of this table the following examples are given: (1) If it is desired to test a Westinghouse service meter by using the rotating standard, the two meters should be connected in series and loaded so as to give one revolution of the disk in approximately one minute's time for a light-load test, and for full load, twenty-five revolutions of the disk in the same time. The number of revolutions made for these two loads by the standard — if the service meter is correct — would be one and twenty- five respectively. If the number of revolutions made by the standard is 25.77 the service meter is three per cent slow at full load.^ If the number of revolutions of the standard is 24.27, the service meter is three per cent fast at full load. From this example it will be seen that the accuracy can be determined for any speed within six per cent fast or slow, reading same directly from the table without any calculation whatever. (2) If it is desired to test a five-ampere General Electric meter the load can be adjusted to give say — two revolutions at light load and thirty revolutions of the disk at heavy load in approximately one minute's time. If the meter is correct the standard will show 1 .8 and 27 revolutions respec- tively. If the standard shows 1 .85, the service meter is three per cent slow LOAD PORTABLE STANDARD INTEGRATING WATTMETER LINE Fig. 30. Connections for Checking Service Meter with Portable Standard Integrating Wattmeter. at light load. If the standard shows 1.75 the service meter is three per cent fast at light load. (3) It it is desired to test a five-ampere Fort Wayne meter the load can be adjusted to the same value as with the General Electric meter. If the meter is correct the standard will show 1.5 and 22.5 revolutions respectively. If the standard shows 1.54 the service meter is three per cent slow at light load. If the standard shows 1.45, the service meter is three per cent fast at light load. If it is desired to test three-wire meters, the standard should be con- nected into the circuit with one side of the meter under test, the other side of the circuit being left open. When the test is conducted in this manner the pointer of the standard will revolve at a rate twice as fast as the disk of the meter under test, which has but one-half of its current winding in use during the test. To effect a direct comparison, the number of revolutions made by the meter bing testeed should be multiplied by two. Testing: Meters for Accuracy on Inductive load*. — When it is desired to test meters for inductive load accuracy the necessary load may be obtained in one of several ways as outlined below: For obtaining the inductive load from a single-phase circuit a set of two or more five^ampere reactance coils, such as are used in the multiple A. C. arc lamp, will be found convenient. The coils can be arranged to give almost any current value, when used on a 110-volt circuit, up to 25 am- peres by means of series parallel connections. The taps which are brought INTEGRATING WATTMETER TESTING. 1019 cut at numerous points are useful in obtaining close adjustments of current value. Fig. 31 illustrates a method of connection for use in testing meters on inductive loads, the power factor of which can be directly determined by a power-factor meter or by the use of an ammeter, voltmeter and watt- meter connected in circuit as indicated. Method of Vesting: Service Meter for Inductive Load Accuracy. — To conduct this test, the service meter should be loaded to its full current capacity as indicated by the ammeter. The lamp load and inductive load should be so adjusted as to give a reading on the wattmeter equal to one-half of the volt-ampere reading as shown by the reading of the ammeter multiplied by the voltage of the circuit. If a standard indi- cating wattmeter is used, the watt value is at once apparent. If the rotating standard integrating meter is used, however, the approximate watt value may be obtained by noting the speed of the pointer which should rotate one-half as fast as it would if the same volt-amperes were applied at unity power factor. The full-load speed of the rotating standard operating at the cur- LAMP BANK (k (4 (h(k(k Fig. 31. Obtaining Inductive Load from Single-Phase Circuit. rent and voltage marked upon the dial is 12 J R.P.M., at a power factor of 50 per cent. With this method of testing on inductive load at a power factor of 50 per cent, it is necessary to take comparative readings the same as in the ordinary test of meters. Obtaining* Inductive I^oad from Two-Phaie Circuits. —An integrating meter can readily be checked for inductive load accuracy if a two-phase circuit is available by connecting the current coils of the meter in one phase and taking the potential from the other phase as shown in Fig. 32. The meter should be given normal full-load current and potential and as the current and potential in this case are 90 degrees apart or in quadrature, it is obvious that the meter disk should not move. A standard indicating or integrating meter should be in circuit during this test as a check upon the two-phase current being exactly in quadrature. If the standard shows any load the current should be further lagged by Inserting a sufficient number of lamps in the phase B circuit, or, if desired, an inductance can be inserted in the series circuit of the wattmeter. In order to secure the proper phase relation it may in some instances be necessary to reverse the primary or secondary connection of the transformer in phase 1020 ELECTRICITY METERS. B. When the phase displacement is exactly 90 degrees the standard should not show any load. Obtaining' Inductive JLo.irt from Three-Phase Circuits. — The above condition of zero power factor or quadrature may also be ob- tained from a balanced three-phase circuit by connecting the meter as shown in Fig. 33 with the current coils in phase A, taking the potential across phases B and C, the load being placed between phases AB and AC. This load must be the same (balanced) on each phase to obtain the desired result. Another method of obtaining this condition from a three-phase circuit is to transform from three-phase to two-phase and connect the meter into the two-phase circuit as shown in Fig. 34. This method necessitates the use of special transformers having the "Scott" three-phase to two-phase con- nections, but in some cases this method may be more convenient than the method shown in Fig. 33, as it eliminates the necessity of maintaining the balanced load on the three-phase circuit, it being only necessary to have one lamp bank on one phase of the two-phase circuit for a load. Having obtained a current in quadrature with the potential, the test should be conducted as outlined in the preceding paragraph describing the two- phase method. Testing* D. C Meters. — For testing D. C. meters a testing arrange- ment similar to that shown in Fig. 31 may be employed and the meters tested LAMPS LAMPS PORTABLE STANDARD INTEGRATING WATTMETER INTEGRATING 1 WATTMETER Fig. 32. Obtaining Inductive Load from Two-Phase Circuit and Using Integrating Wattmeter as Standard. by the voltmeter-ammeter method or by the indicating wattmeter method. The reactance coils would not be employed, but in general the method of test is the same as for A. C. meters previously described. Owing to the rate of heating being different for the shunt circuit and the disk, it is neces- sary that the meter be run long enough before test to allow it to reach its normal operating condition, which is approximately 15 minutes. Testing* Polyphase Service Jfleters. — As the polyphase meter is really two single-phase meters having a common shaft and registering mechanism, the general instructions for the single-phase meters will apply to the polyphase meters. The calibration and checking of these meters, however, is necessarily more complicated and the following general in- structions will be of assistance in the testing of this type of meter. Standards for Testing* T*olrphase Meters. — As yet a rotating standard of the polyphase type is not (December, 1907) on the market, and INTEGRATING WATTMETER TESTING. 1021 it is eustomary to use the indicating meter and stop watch method for testing this class of meter, although a single-phase portable integrating standard wattmeter may be used if the method is properly applied. A- B- Current Coil y M ETE R Pot Colt V- LAMPS y 00 Fig. 33. Obtaining Inductive Load from Three-Phase Circuit. To test a polyphase meter it is customary to employ an artificial load and test each side as a single-phase element. To test a self-contained meter using neither series or voltage transformers the connections should be made as shown in Fig. 35, and for testing a meter using transformers CURRENT COIL METER Por.coiL THREE PHASE TWO PHASE Fig. 34. Obtaining Inductive Load from Three- Phase Circuit by Use of " Scott " Three-Phase- Two-Phase Connection. connect as shown in Fig. 36. A three-point switch is provided to cut either series element of the service meter in circuit with the standard. As but one series side of the meter is in service at a time it is either necessary to Fig. 35. Connections for Testing Self -Contained Polyphase Meter Using Single-Phase Standard. 1022 ELECTRICITY METERS. multiply the disk revolutions by two or divide the calibrating constant by two. The test should be conducted in the same manner as when test- ing single-phase meters previously described. It will be noted that both potential elements of the service meter are energized, this being essential in polyphase testing. LINE Fig. 36. Connections for Testing Polyphase Meter Employing Transformers and Using Single-Phase Standard. S£#/£S Trans. POLYPHASE INDICATING WATTMETER Fig. 37. Connections for Testing Polyphase Meter Employing Trans- formers. Testing on Running Load and Using Polyphase Standard. INTEGRATING WATTMETER TESTING. 1028 To test a polyphase meter on the running load, connections should be made as shown in Figs. 37-38 and the test conducted in the same manner as for single-phase testing. Care should be exercised to connect the poten- tial element to the same point to avoid danger of one meter measuring the watt loss of the other. When desired, the single-phase portable standard integrating watt- meter may be used for checking polyphase meters instead of the indicating wattmeter. For this purpose the polyphase meter should be connected as shown in Figs. 35-36 and the standard integrating meter substituted for the indicating meter. When so connected the disk revolutions of the polyphase meter should be multiplied by two and directly compared with the rotating standard, in which case instructions for single-phase testing will apply. If desired the current elements of the polyphase meter may jftfraJ POLYPHASE INDICATING WATTMETER Fig. 38. Connections for Testing Self-Contained Polyphase Meter on Running Load and Using Polyphase Standard. be connected in series, in which case the service and test meter disks v ,-h revolve at the same speed. Note: — In all tests of polyphase meters both potential coils must be connected in circuit and energized. Polyphase meters should be given the same tests at light and full loads as the single-phase meters and the same adjustments apply. Service Connections of Polyphase meters. — Great care should be exercised in the installation of polyphase meters to insure the connections being made exactly in accordance with the proper diagrams. This is extremely important, as it is possible to make incorrect connections pro- ducing excessive errors on inductive loads and still have the meter rotate in the proper direction. It is not a safe plan to try out polyphase meter connections by alternately opening the sides of the measuring elements and noting that the disk rotates in the forward direction in each case, unless the power facter is definitely known. If the meter should be connected to 1024 ELECTRICITY METERS. a three-phase circuit operating at a power factor of less than 50 per cent, one element should cause the disk to rotate backwards, and if the above test alone is depended upon when installing the meter, it is very probable that the average man installing the meter under these conditions would reverse the side rotating backwards, thus introducing an enormous error as the power factor of the circuit changed. It is also possible to so connect a polyphase meter that it will run in either the forward or reverse direction on both elements regardless of the power factor, the meter either running faster or slower than it would on unity power factor, depending upon the phase relation of the particular connection used. The action of two single-phase meters, or the two single-phase elements of a polyphase meter operating upon a three-phase circuit, may be explained by the following vector diagrams. Figure 39 shows the phase relations between the current and potential of each single-phase element when operating on a three-phase circuit at unity power factor, one meter element having its series coil in A and its poten- tial coil across AC and the other element having its series coil in B and its potential coil across BC. From this diagram it will be seen that the cur- rent in phase A is displaced 30 degrees from its respective potential AC and the current in phase B is also displaced 30 degrees from its potential BC, Fig. 39. Fig. 40. but in the opposite direction from that in phase A, thus giving the effect of a lagging current in phase B and a leading current in phase A, the resultant being zero displacement, or unity power factor, on the three-phase circuit. From this it will be seen that at unity power factor on the three-phase cir- cuit each single-phase element of the polyphase meter will operate at the same speed, each element operating at a single-phase power factor of about 88 per cent, or the cosine of 30 degrees. Figure 40 shows the condition existing when the current in the three- phase circuit lags 30 degrees or is operating at a power factor of 86 per cent. From this diagram it will be seen that the current in phase B lags behind its respective potential BC 30 -I- 30 degrees or 60 degrees, while the cur- rent in A has been brought exactly in phase with its respective potential AC. This gives a condition where one single-phase element is operating at a power factor of 50 per cent (or cosine of 60 degrees), while the other ele- ment is operating at unity power factor, its current and potential being exactly in phase. Under this condition one element will run twice as fast as the other. Ox, Oy and Oz show positions of three-phase current with 30 degrees lag. To show phase relation of each current with its respective voltage, Ore is rotated about center A instead of O and falls in phase with its voltage AC. Current Oy is rotated about center B and falls 60 degrees behind its voltage BC. Figure 41 shows the condition met with when the current in the three- phase circuit lags 60 degrees or is operating at a power factor of 50 per cent. From this diagram it will be seen that the current in phase B lags its re- INTEGRATING WATTMETER TESTING, 1025 spective potential BC 60 4- 30 degrees or 90 degrees, while the current in pnase A lags its potential AC 60 — 30 degrees or 30 degrees. This gives a condition where one single-phase element is operating at zero power factor or cosine of 90 degrees, while the other element is operating at 86 per cent or cosine of 30 degrees. Under this condition one element has stopped, the other element doing all the work. For clearness in showing phase relations the centers of rotation of the currents are changed as in Fig. 40. Figure 42 shows the condition met with when the current in the three- phase circuit lags 90 degrees or is operating at a power factor of zero. From this diagram it will be seen that the current in phase B lags its respective potential BC 90 4- 30 degrees or 120 degrees, while the current in phase A lags its^ respective potential AC 90 — 30 degrees or 60 degrees. As the angle of lag in phase B now exceeds 90 degrees, the cosine of the angle is the same as the sine of the difference between the angle and 90 degrees, in this case minus 30 degrees, giving a power factor of minus 50 per cent in phase B and a power factor of plus 50 per cent in phase A. From this it will be. seen that at zero power factor of the three-phase circuit, one single- phase element of the meter will try to operate at half speed in one direction Fig. 42. while the other element is trying to operate at half speed in the opposite direction, the resultant of these two equal forces acting in opposite directions being zero; hence, the meter as a whole will not move. From the preceding explanation of the phase relations of single-phase meters used on a three-phase circuit, it will be apparent that the energy of a three-phase circuit cannot be measured by the use of one standard single- phase meter. It also shows why it is extremely important to have the polyphase meter connected into the circuit in accordance with the proper diagrams as, owing to the fact that one element of the polyphase meter should tend to reverse its direction of rotation on a power factor of less than 50 per cent, it is not safe to depend upon the direction of rotation of each element separately to determine whether or not a meter is connected into the circuit properly unless the power factor is known. The general scheme of connections for correctly connecting a polyphase meter to measure the energy of a three-phase circuit is shown in Fig. 43, the current coil of one element being connected in line A and its poten- tial across A and B, the current coils of the other element being connected in line C and its potential coils across B and C. If a meter should be connected, as shown in Fig. 44, with the current coil of one element in line A and its potential across A and C and the current of the other element in line C with its potential coil across B and C, both elements of the meter will run in either the forward or reverse direction at 1026 ELECTRICITY METERS. all values of power factor at equal speeds, and will be either fast or slow on all power factors other than unity, depending on the phase relations of the particular connection used. This erroneous connection should be care- fully guarded against, and it will be readily seen that this condition cannot be detected by the common method used of opening one side of the meter at a time to determine that the meter runs in the forward direction on each element alone. The effect of the connections shown in Fig. 44 can be seen by referring to Fig. 45. If one series element of the polyphase meter is connected in at A and its potential element connected across AC, and the other series element CurrentCof/ VSAAA— - > Potential Co) I LOAD LINE Potential Coil WW Current Coif Fig. 43. connected in at B with its potential element connected across BA, when operating under 30 degrees lag the currents Ox and Oy will be shifted so that botn will be in phase with their voltage and the meter will run in a forward direction faster than it will at unity power factor of the three- phase circuit. With one series element of the meter connected in at A and its potential element connected across AB and the other series element connected in at B and its potential element connected across BC, the cur- Current Coil — V\A/V A Potential Coil LOAD B LINE Potential Coil Current Coil Fig. 44. rents will be shifted so that both Ox and Oy lag behind their respective voltages and the meter will consequently run slower than it will at unity power factor of the three-phase circuit. Practical Methods of Checking" Connections of Poly- phase Meters. — In cases where it is not positively known that the power factor is above 50 per cent, the following method may be used, which is based on the fact that the sum of the two readings should be positive, so long as the power is in the positive direction. When the currents in the voltage and series coils, as indicated by the clock diagram, are in the same direction, or within 90 degrees of being in the same direction, the meter will read forward. When the current in the series coil is more than 90 degrees out of phase with the voltage, the meter will reverse. First. By proper testing with an incandescent lamp or a voltmeter, obtain three voltage leads, A, B, C, having equal voltages between them. INTEGRATING WATTMETEK TESTING. 1027 Second. Connect these leads to the voltage circuits of the wattmeters as per Fig. 43. Third. Connect the series transformer at A to meter whose potential is connected to AC, and series trans- former at B to meter whose potential is connected to BC. See clock dia- gram (Fig. 46) giving the phase rela- tions. In this diagram, AC repre- sents the voltage on meter connected at A, BC the voltage on meter connected at B, OA the current in meter connected at A, and OB the current in meter connected at B. Fourth. Change voltage connec- tion from AC to AB on meter con- nected at A. If power factor is 100, the readings will be alike with both connections. If the power factor is less than 100 and greater than 50, the readings will differ, but be in the same direction (either both positive or both negative). If equal to 50, one of the readings will be zero. If less than 50, the readings with connec- tions AC and AB will be reversed in direction, with respect to each other. Fifth. The same test may be performed on meter connected at B, by changing the voltage connections from BC to BA. If the power factor Fig. 45. ' I \ A 60 i A 40 A 50 Fig. 46. OA current in meter at A 100 per cent P.F. OA 60 current in meter at A 60 per cent P.F. OA 50 current in meter at A 50 per cent P.F. OA 40 current in meter at A 40 per cent P.F. OB current in meter at A 100 per cent P.F. OB 60 current in meter at A 60 per cent P.F. OB 50 current in meter at A 50 per cent P.F. OB 40 current in meter at A 40 per cent P.F, 1028 ELECTRICITY METERS. is 100, the readings will be alike. If less than 100 and more than 50. the readings will differ, but be in the same direction. If equal to 50, one of the readings will be zero. If less than 50, the readings with connections BC and BA will be reversed in direction with respect to each other. Sixth. If it is found from the above tests that the power factor is greater than 50, connect the series coil of the meters so that both read forward. If the power factor is less than 50, connect the series coil of the slower meter so that meter reads backward, and the series coil of the faster meter so that it reads forward. The above description indicates the use of two single-phase meters, but holds equally true for a polyphase meter consisting of two single-phase meter elements driving the same shaft. nETEH TE8TMG FOJU^Il I,JE. Below will be found the formulae and testing constants to be used in col- junction with the testing methods described on pages 1013 to 1023. Formula for Vesting* the Shallenberg*er Ampere-bout Meter. To Tell the Exact Current Flowing at any Time. Note the number of revolutions made by the small "tell-tale" index of the register dial, in a number of seconds equal to the constant of the meter. The number of revolutions noted will correspond to the number of amperes passing through the meter. For example: the 20-ampere meter constant is 63.3; if the index makes 10 revolutions in 63.3 seconds, 10 amperes are passing through the meter. In order to avoid errors in readings, it is cus- tomary to take the number of revolutions in a longer time, say 120 seconds, using the following formula: No. of Rev. X Meter Constant ~ rr t-5 = Current. No. of Sec. If, therefore, the index of a 20-ampere meter makes 19 revolutions in 120 seconds the current passing is 19 X 63.3 tn — ^r — = 10 amperes. The cover should be left on the meter while these readings are taken. The constants of the different capacity meters are given below: Meter Capacity. Amperes. Calibrating Constant. Meter Capacity. Amperes. Calibrating Constant. 5 10 20 40 22.5 33.8 63.3 126.6 80 120 160 253.1 386 506 Testing* Formula for Shallenlterg-er and Westing*house Integrating* Wattmeter*. The standard formula for testing all types and capacities, when using indicating standards and stop watches, is Watts = ™ K in which: R = Number of complete revolutions in time T. T = Time in seconds required for revolutions R. K = Jonstant. The constant " K " varies with different types and capacities as outlined on the following page. METER TESTING FORMULA. 1029 Rating's. — In all cases the volt and ampere values used with the formula are those marked on the meter. The full-load speed of Types "B" and "C* meters is 25 R.P.M. lulMoad Speed's. — The full-load speed of Shallenberger, Westing- house, Round Pattern and Type "A" Single and Polyphase Wattmeters is 50 R.P.M. The full-load speed of Type "B" single phase and Type "C" single or polyphase wattmeters is 25 R.P.M. For Shallenberger, Westinghouse Round Pattern Back Connected and Type "A" Meters the constant "K" has the following values: 2-Wire Meters {Single Phase). For self-contained meters K = volts X amps. X 1.2. For meter used with series transformer only (but checked without) K = volts (as marked on dial) X 6. For meter used with series and voltage transformers (but checked with- out) K = 600. For meter used with transformers of either or both forms (and checked with) K = volts X amps. X 1.2. 3-Wire Meters (Single Phase) . For self-contained meters K = volts X amps. X 2.4. For meters used with series transformers only (but checked without) K = volts X 6. Type "A" Polyphase Wattmeters. For self-contained meters K = volts X amps. X 2.4. For meters used with series transformers only (but checked without) K = 5 X volts X 2.4. For meters used with series and voltage transformers (but checked with- out) K = 1200. For meters used with transformers of either or both forms (and checked with) K = volts X amps. X 2.4. The Testing: Constant " I£ " of Westing-house Types " JB " and " C " Meters is as follow*: 2-Wire Meters (Single Phase). For self-contained meters K = volts X amps. X 2.4. For meters used with series transformers only (but checked without) K = volts X 5 X 2.4. For meters used with series and voltage transformers (but checked with- out) K = 5 X 100 X 2.4. For meters used with transformers of either or both forms (and checked with) K = volts X amps. X 2.4. 3-Wire Meters (Single Phase) . For self-contained meters K = volts X amps. X 4.8. For meters used with series transformers (but checked without) K = volts (as marked on meter) X 12. Note. — When the voltage marking of Westinghouse three-wire meters covers both the voltage between neutral and outer and the voltage between outers such as 100-200 volts, K = volts (between outside wires) X am- peres as marked on meter X 2.4. Type "C" Polyphase Wattmeters. For self-contained meters K = volts X amps. X 4.8. For meters used with series transformers only (but checked without) K = 5 X volts X 4.8. For meters used with series and voltage transformers (but checked with- out) K - 2400. For meters used with transformers of either or both forms (and checked with) K = volts X amps. X 4.8. 1030 ELECTRICITY METERS. WESTUVGHOUSE DIRECT-CURRENT METERS. For all capacity meters K = volts X amps. X 2.4. Formula for Testing- General Electric Recording: Wattmeters. The standard formula for testing all types and capacities when using indicating standards and stop watches is Watts = 3600 XK X R in which: R = number of revolutions. S = number of seconds in which revolutions is made. K = calibrating constant marked on dial face of "non-direct" reading meters and on disk of "direct" reading meters. Table of General Electric D. C. Type "C6" Testing* Con- stants ** li " and Watts per 11 evolution per Minute. Capacity of Meters in Am- peres. 3 5 10 15 25 50 75 100 150 300 600 100-120 Volts. Testing Constant. .125 .2 .4 1. 2. 3. 4. 6. 12.5 25. Watts Per Revolu- tion per Minute. 7.5 12. 24. 36. 60. 120. 180. 240. 360. 750. 1500. 200-240 Volts. Testing Constant . .25 .4 .75 1.25 2. 4. 6. 7.5 12.5 25. 50. Watts Per Revolu- tion Per Minute. 15 24 45 75 120 240 360 450 750 1500 3000 500-600 Volts. Testing Constant .6 1. 2. 3. 5. 10. 15. 20. 30. 60. 125. Watts Per Revolu- tion Per Minute. 36 60 120 180 300 600 900 1200 1800 3600 7500 Table of General Electric A. C. Type "* " Testing- Con- stants " K " and Watt* per Revolution per .Tlinute. Capacity of 100-130 Volts. 200-260 Volts. 500-600 Volts. Meters Watts Per Watts Per Watts Per in Am- Testing Revolu- Testing Revolu- Testing Revolu- peres. Constant . tion Per Minute. Constant. tion Per Minute. Constant. tion Per Minute. 3 .2 12 .4 24 1. 60 5 .3 18 .6 36 1.5 75 10 .6 36 1.25 75 3. 180 15 1. 60 2. 120 5. 300 25 1.5 90 3. 180 7.5 450 50 3. 180 6. 360 15. 900 75 5. 300 10. 600 25. 1500 100 6. 360 12.5 750 30. 1800 150 10. 600 20. 1200 50. 3000 200 12.5 750 25. 1500 60. 3600 300 20. 1200 40. 2400 100 6000 DUNCAN METERS. 1031 " » 3 " Polyphase UEeters. 100-120 Volts. 200-260 Volts. 500-650 Volts. 25 Cycles 60 Cycles 25 Cycles 60 Cycles 25 Cycles 60 Cycles Amps. Testing Testing Testing Testing Testing Testing Constant. Constant. Constant. Constant. Constant. Constant. 3 1. .4 • 2 .75 5. 2 5 1.5 .6 3 1.25 7.5 3 10 3. 1.25 6 2.5 15. 6 15 5. 2. 10 4. 25. 10 25 7.5 3. 15 6. 40. 15 50 15. 6. 30 12.5 75. 30 75 20. 7.5 40 15. 100. 40 100 30. 12.5 60 25. 150. 60 150 40. 15. 75 30. 200. 75 Note : — Testing constant is actual watt-hours per revolution of disk. formula for Testing- Dnncan Recording* Wattmeter*. The standard formula for testing all types and capacities when using • j. .- j j j . u • w ** Rev. X 3600 X K . indicating standards and stop watches is Watts = ~ — , in which: R = Number of complete revolutions. Sec.= Time in seconds required for revolutions R. K = Testing constant marked on meter disk. Table of Duncan Testing* Constants " K " and Watts per Revolution per Minute. Capacity of 100-125 Volts. 200-250 Volts. 450-550 Volts. Meters Testing Watts per Testing Watts per Testing Watts per m Con- Revolution Con- Revolution Con- Revolution Amperes. stant. per Minute. stant. per Minute. stant. per Minute. 2i i 15 $ 30 1 60 5 A 15 1 30 1 60 7* £ 30 1 60 2 120 10 ^ 30 1 60 2 120 15 1 60 2 120 5 300 25 1 60 2 120 5 300 50 2 120 4 240 10 600 75 3 180 6 360 16 960 100 4 240 8 480 20 1,200 150 6 360 12 720 30 1,800 200 8 480 16 960 40 2,400 300 12 720 25 1,500 60 3,600 450 20 1,200 30 1,800 80 4,800 600 25 1,500 50 3,000 100 6,000 800 30 1,800 60 3,600 160 9,600 1,000 40 2,400 80 4,800 200 12,000 1,200 50 3,000 100 6,000 250 15,000 1,500 60 3,600 120 7,200 300 18,000 2,000 80 4,800 160 9,600 400 24,000 2,500 100 6,000 200 12,000 500 30,000 3.000 120 7,200 250 15,000 600 36,000 4,000 160 9,600 300 18,000 800 48,000 5,000 200 12.000 400 24,000 1,000 60.000 6,000 250 15,000 500 30,000 1,200 72,000 8,000 300 18,000 600 36,000 1,600 96,000 10,000 400 24,000 800 48,000 2,000 120,000 1032 ELECTRICITY METERS. The table given below will be found convenient in showing the per cent fast or slow which a meter is running when employed in conjunction with the following formula : Watts Constituting Load Testing Constant X 60 = Rev. Per Min. Per Cent Error Table for Fifth* of it Second. Time Per Cent Time Per Cent Time Per Cent Time Per Cent in Seconds Fast in Seconds Fast in Seconds Slow in Seconds Slow 40.20 49.25 50.20 19.52 60.20 0.33 70.20 14.52 .40 58.51 .40 19.05 .40 0.67 .40 14.77 .60 47.78 .60 18.58 .60 0.99 .60 15.01 .80 47.06 .80 18.11 .80 1.31 .80 15.25 41.00 46.34 51.00 17.65 61.00 1.63 71.00 15.50 .20 45.63 .20 17.19 .20 1.96 .20 15.73 .40 44.93 .40 16.73 .40 2.27 .40 15.96 .60 44.23 .60 16.28 .60 2.59 .60 16.20 .80 43.54 .80 15.83 .80 2.91 .80 16.43 42.00 42.86 52.00 15-38 62.00 3.22 72.00 16.66 .20 42.18 .20 14.94 .20 3.53 .20 16.89 .40 41.51 .40 14.50 .40 3.84 .40 17.12 .60 40.85 .60 14.07 .60 4.15 .60 17.35 .80 40.19 .80 13.64 .80 4.45 .80 17.58 43.00 39.53 53.00 13.21 63.00 4.76 73.00 17.81 .20 38.89 .20 12.78 .20 5.06 .20 18.03 .40 38.25 .40 12.36 .40 5.36 .40 18.25 .60 37.61 .60 11.94 .60 5.66 .60 18.47 .80 36.98 .80 11.52 .80 5.95 .80 18.70 44.00 36.36 54.00 11.11 64.00 6.25 74.00 18.92 ' .20 35.75 .20 10.70 .20 6.54 .20 19.14 .40 35.14 .40 10.29 .40 6.83 .40 19.35 .60 34.53 .60 9.89 .60 - 7.12 .60 19.57 .80 33.93 .80 9.49 .80 7.40 .80 19.79 45.00 33.33 55.00 9.09 65.00 7.69 75.00 20.00 .20 32.74 .20 8.69 .20 7.97 .20 20.21 .40 32.16 .40 8.30 .40 8.25 .40 20.42 .60 31.58 .60 7.91 .60 8.53 .60 20.63 .80 31.00 .80 7.53 .80 8.81 .80 20.84 46.00 30.43 56.00 7.14 66.00 9.09 76.00 21.05 .20 29.87 .20 6.76 .20 9.36 .20 21.26 .40 29.31 .40 6.38 .40 9.63 .40 21.47 .60 28.76 .60 6.01 .60 9.92 .60 21.68 .80 28.21 .80 5.63 .80 10.17 .80 21.88 47.00 27.66 57.00 5.26 67.00 10.44 77.00 22.07 .20 27.12 .20 4.89 .20 10.71 .20 22.27 .40 26.58 .40 4.53 .40 10.97 .40 22.38 .60 26.05 .60 4.17 .60 11.24 .60 22.68 .80 25.52 .80 3.81 .80 11.50 .80 22.88 48.00 25.00 58.00 3.45 68.00 11.76 78.00 23.08 .20 24.40 .20 3.09 .20 12.02 .20 23.28 .40 23.96 .40 2.74 .40 12.28 .40 23.47 .60 23.45 .60 2.39 .60 12.53 .60 23.66 .80 23.15 .80 2.04 .80 12.79 .80 23.86 49.00 22.45 59.00 1.69 69.00 13.04 79.00 24.05 .20 21.95 .20 1.35 .20 13.29 .20 24.24 .40 21.46 .40 1.01 .40 13.54 .40 24.43 .60 20.97 .60 0.67 .60 13.79 .60 24.63 .80 20.48 .80 0.33 .80 14.04 .80 24.82 50.00 20.00 60.00 0.00 70.00 14.28 80.00 25.00 FOKT WAYNE SINGLE-PHASE METERS. 1033 Example. — If the revolutions to be made in one minute are completed in exactly 60 seconds the speed is correct and the per cent error is zero, but if the revolutions were made in 57 seconds then the meter is running 5.26 per cent fast; if completed in 58.4 seconds it is 2.74 per cent fast. When the time exceeds 60 seconds, the meter is slow. If it requires 63 seconds it is 4.76 per cent slow; if 64.6 seconds it is 7.12 per cent slow. The per cent error will be found in the column after the time in seconds. The seconds columns are divided into fifths of a second so as to conform to most stop watches whose seconds are split to fifths. Formula for Testing- Fort Wayne Type "K" Wattmeter. The standard formula for testing all types and capacities when using indicating standards and stop watch is Watts = '■ ~ • Tables of Values of Constant PEi\, or OPE\ CIRCUIT MFTlf 4»1>. The following diagram shows the connections of one terminal station with the line connecting to the next. The ground plates may be dispensed with if a return wire from the next station is used, thus forming a metallic cir- cuit. This method of connecting Morse apparatus is used mostly in Europe, and has two advantages over the American method . a. The battery is not in circuit except when signals are being sent. b. When the key is closed and the current admitted to line, the coils of the relay are cut out of the circuit, thus lessening the hindrance to the flow of current. 1040 TELEGRAPHS 1041 LINE TO NEXT 8TATION KEY rT MAFN BATTERY LINE TO GROUND OR TO RETURN WIRE ISf Fig. 2. REPEATERS. In practical telegraphy, the high resistance of the line wire between the terminal stations, and imperfect insulation permitting leakage in damp weather, make it inexpedient to attempt to transmit signals over circuits whose lengths have not well-defined limits. But a circuit may be extended, and messages exchanged over longer distances by making the receiving instrument at the distant terminal of one circuit do the work of a transmit- ting key in the next. The apparatus used for this purpose is called a re- peater, and is usually automatic, in a sense which will appear later on. From among the scores of repeaters, selection must be made of repre sentative types, — the three in most general use Milliken Repeater. The following diagram illustrates the theory of the Milliken repeater, which is in general use in the United States and Canada. The essential feature of every form of automatic repeater is some device by which the circuit into which the sender is repeating not only opens when he opens, but closes when he closes. i 1042 TELEGRAPHY. In the diagram is represented the apparatus of a repeating station in which appear the instruments and three distinct circuits in duplicate, viz.: the east and west main line; east and west local (dotted); east and west extra local (dash and dot). Starting with both "east" and "west" keys closed and the line at rest, battery b', whose circuit (dash and dot) is com- plete through transmitter, T\ energizes extra magnet, E', attracts the pen- dent armature, P', leaving the upright armature free, the pendent armature, P, being similarly held by battery, b. In operation, the distant east opens his key, relay, E, opens, then transmitter, T, through whose tongue and post passes the west line, which opens, and would open relay, W, and therefore transmitter, T'\ but at the moment transmitter, T, opens, the extra local circuit (dash and dot) opens, releasing pendent armature, P, which is drawn by its spring against the upright armature holding closed the points of relay, W, and transmitter, T", and therefore the east line, which passes through its tongue and post. When the distant west breaks and sends, the action begins with the west relay instead of east, and follows the same course. Olieg-an Repeater. In repeaters for lines worked single, the characteristic is a device in the repeater which holds closed the main line on which the sending is being done, L& \ V AtLi > Fig. 4. while the distant relay on the second main line records that sending; the parts arranged to effect this result should act quickly on the "break" and a little slowly on the "make" of the main line current — " break " and "make" being the technical terms respectively for the opening and closing of the circuit. A form of repeater intended to effect in a high degree this result, called from its inventor the Ghegan, is shown in theory in the dia- gram, Fig. 4. The characteristic instrument is a transmitter having a second armature-bearing lever placed above the first one in such a position that one electromagnet serves to work both; the upper armature forms a back contact simultaneously with the opening of the transmitter, and it inclines to preserve the contact at U' until the regular local circuit (dotted) has been closed at the local points in relay E; the action is therefore quick or slow as occasion requires. As in the Milliken and Weiny-Phillips, there are three pairs of circuits; the main lines (solid black); the local circuits (dotted); and the shunt circuits (dot and dash). When relay W open it REPEATERS. 1043 releases the armature of transmitter T'\ through its tongue and post passes the west wire which opens, releasing the armature of relay E, and opening its local points. At the same time upper armature U' flies against its back contact and completes a shunt circuit by which battery b holds transmitter T closed; and the wire passing through its tongue and post is kept intact. Reverting to the position of the instruments in the diagram, the distant east is supposed to have opened his key. This opens relay W, which opens transmitter T' (both armatures); the drop in the lower armature opens the west main line, which opens relay E and its local points; but, as just explained, the circuit of battery b is now complete through the dot and dash lines, so that transmitter T is held closed and the east line is kept intact by its tongue against the stop. When the distant west breaks, the armature of relay E remains on its back stop, and, on the first downward stroke of the upper armature of transmitter T', the local circuit of trans- mitter T is broken, and at its tongue and post the east line opens. The east sender, thus warned, closes his key; the sender at the distant west takes the circuit, and action similar to that just described begins with relay E, and follows a like course. Weiny-Phillips Repeater. A theoretical diagram of the Weiny-Phillips repeater is given herewith. It is in general use by one of the principal telegraph companies, and is .!!,. Fig. 5. introduced here because it involves the principle of differentiation in mag- net coils, which plays so important a part in duplex telegraphy. As in the Milliken, there are three distinct circuits in duplicate; and in the diagrams the parts performing like functions in the two types of repeaters are simi- larly lettered. The connections and functions of the main line (solid black) circuits and of local (dotted) circuits are identical with those of the Milli- ken. But instead of the extra magnets and pendent armature of the latter, we have a tubular iron shell enclosing a straight iron core and its windings, the combination of shell and straight core performing the same functions as the usual horse-shoe core. The turns of wire around the core of the extra magnetare equally divided, and the current traverses the two halves in opposite directions. Such a core is said to be differentially wound, be- cause the core is energized by the difference in strength of the currents in the coilsj but when the coils are equal in resistance, the equal currents, passing in opposite directions around the core, neutralize each other. If one of the coils is opened, the core at once becomes a magnet capable of holding the armature at the moment when, the repeater in operation, the *'east" station opens his key, opening relay E\ then transmitter T; then 1044 TELEGRAPHY. opening the "west" wire, which would open relay W, transmitter T r , and therefore the east wire; but the opening of transmitter T' is prevented by the energizing at the critical moment of core ~W\ one coil of which is opened when transmitter T opens. When the distant west breaks and sends, the action begins with the west relay instead of the east, and follows the same course. Duplex Telegraphy. That method of telegraphy by which messages can be sent and received over one wire at the same time is called duplex; and the system in general use, known as the polar duplex, is illustrated in the accompanying diagram. In single telegraphy all the relays in the circuit, including the home one, respond to the movements of the key; the duplex system implies a home relay and sounder unresponsive, but a distant relay responsive to the move- ments of the home key; and this result is effected by a differential arrange- ment of magnet coils, of which the extra magnet coils in the Weiny-Phillips repeater furnished an example. A current dividing between two coils and their connecting wires of equal resistance will divide equally, and passing round the cores, will produce no magnetic effect in them. This condition is established when the resistance of the wire marked -*r> «■— in the diagram WEST EAS- THEORETICAL DIAGRAM OF POLAR DUPLEX BALANCING SWITCH OMITTED Fig. 6. is balanced by the resistance of a set of adjustable coils in a rheostat marked R. This is called the ohmic balance (from ohm, the unit of resistance); and the static balance is effected by neutralizing the static discharge on long lines by means of an adjustable condenser C, and retardation coil r, shunt- ing the rheostat as shown. In the single line relay the movement of the armature is effected by the help of a retractile spring in combination with alternating conditions of current and no current on the line. In the polar relay the spring is dispensed with, and the backward movement of the arm- ature is effected, not by a spring, but by means of a current in a direction opposite to that which determined the forward movement. This reversal of the direction of the current is effected by means of a pole-changer, PC, whose lever, T, connected with the main and artificial lines, makes contact, by means of a local circuit and key, K. with the zinc ( — ) and copper ( -+• ) terminal of a battery alternately. The usage in practice is zinc to the line when the key is closed; copper, when open. The law for the production of magnetic poles by a current is this: When a core is looked at "end on," EEPEATERS. 1045 A current passing round it in the direction of the hands of a clock produces south-seeking magnetism, S; in the opposite direction, north-seeking mag- netism, marked N. A springless armature, permanently magnetized and pivoted, as shown in the drawing, will, if its free end is placed between S and N magnetic poles, be moved in obedience to the well-known law that like poles repel, while unlike poles attract each other. The "east" and "west" terminal is each a duplicate of the other m every respect; and a description of the operation at one terminal will answer for both. Under the conditions shown, the keys are open; and the batteries, which have the same E.M.F., oppose their copper ( + ) poles to each other, so that no current flows in the main line. But in the artificial line the current flows round the core in such direction as, according to the rule just given, to produce N and S polarities as marked, opening the sounder circuits at both terminals. If, by means of key, K', the pole-changer, PC, of "east" station is closed, the connections of battery, B', are changed; it is said to be reversed; and it now adds its E.M.F. to that of battery B, the current flowing in a direction from "west" to "east"; i.e., from copper to zinc. But the current in the main line is to that in the artificial as 2 to 1; and if the relative strength of the resultant magnetic poles is represented by small type for that produced by the current in the artificial line, and by large type for the main, the magnetic conditions can be graphically shown, as they are produced on each side of the permanently magnetized armatures marked (N) and (N'). In relay, PR', it is Sn (N') sN, causing it to remain open; in relay PR it has changed to Ns (N) nS — just the reverse of that shown in the diagram — the relay therefore closes, and the sounder also. If key, K, of the west station is closed at the same time, the batteries are again placed in opposition, but with zinc ( — ) poles to the line, instead of, as in the first instance, copper ( + ) poles. The result is no current on the main line; but the current in the artificial lines, flowing in the direction from the ground (whose potential is 0) to the zinc ( — ) of the batteries, the magnetic condi- tion at "east" station is represented by n (N') s, which closes relay, PR'; and at "west" station by n (N) s, which closes relay PR. The conditions necessary to duplex work, viz., that the movement of key, K', should have no effect on relay, PR', but should operate the distant relay, PR, are thus fulfilled, and the transmssion of messages in opposite directions at the same time is made practicable. In the case of the Wheatstone Automatic duplex this exchange goes on at high rate of speed, the maximum rate being 250 words a minute. There have already been traced out the magnetic poles formed in the inside ends of the relay cores as the result of three possible combinations of current: (1) copper to line at each end; (2) zinc at east, copper at west end; (3) zinc to line at each end. One other possible combination remains to be traced out with reference to the poles formed; it is shown in Fig. 7, where the duplex is represented in a form more nearly approaching that which obtains in practice. At the west, or Pittsburg end, zinc is to the line; at the east, or New Yprk end, it is copper; the effect on the distant relay in each case is indicated in the draw- ing. For the sake of clearness the local systems are omitted ; at each terminal the artificial circuit is represented by a dotted line; the main line by solid black; the relays with their windings are shown in a manner fitted for tracing the magnetic effects. Representing the polarity of the armatures by (N) and OS), and the magnetic condition of the cores in the manner adopted in the preceding paragraph, it must be understood that the point of view is midway between the cores. The direction of the current on the main line in this diagram is from New York to Pittsburg. At the New York end the direction of the current in the artificial line is from the battery to the ground; at the Pittsburg end the current sets in from the ground to the zinc pole of the dynamo. In the Pittsburg relay the magnetic conditions, begin- ning with the lowest core, are Ns (N) nS; the large letters are the poles produced by the main line current; the small are those resulting from the current in the artificial line whose direction is from ground to dynamo; the armature is drawn upward and the relay opens, as shown. In the New York relay, the magnetic conditions (lower core first) are Ns (S) nS; the armature is drawn down and the local points closed. Otner details of the duplex are apparent on examination of the diagram. The two boxes with disks on the top are rheostats; each contains a number of co ; ls in series for making the resistance of the artificial line equal to that 1046 TELEGRAPHY. of the main. Under the rheostats are the condensers for eliminating the effects on the relay of the static discharge of the line. At the New York end is a chemical battery with the old style of pole changer; when open, as shown, it sends copper to the line, and puts zinc to the ground; when closed it puts zinc to line and copper to ground. At the Pittsburg end is shown an entirely different arrangement; it is the one now almost universally in use. Two dynamos furnish the current; the positive pole of one is grounded; Fig. 7. the other pole is led through a safety lamp to a cut-off switch, thence to the pole changer which sends zinc to the line when closed. Of the other dynamo the negative pole is grounded; the copper current goes to the right-hand post of the pole changer, which is very much simpler in form than the old style. The balancing switches, omitted from Fig. 6, are shown marked A and F; by means of these when the lever, say F, is thrown to the right, the main line wire i r detached from the pole changer and passes through a compensating resistance to the ground. REPEATERS. 1047 Duplex XiOop System. For many years after the introduction of the duplex and quadruplex the number of lines operated by those systems was small; but with im- provements in the material for wires and in line construction the number gradually increased until now nearly one half the wires of the two leading companies are utilized for one system or the other; and of the wires thus operated the working sets, to the extent of nearly one half, are assembled in main offices, and the wires themselves are worked, by what are called loops, from branch offices located mostly in the different exchanges. The appara- tus and connections by which the service of the duplex is extended to a branch are therefore an essential part of multiplex telegraphy. Fig. 8 is a diagram of the duplex loop system; the places of polar relay, pole changer and rheostat are indicated ; the main line connections shown in Figs. 6 and 7 are omitted; and the local connections which are entirely omitted from DUPLEX LOOP Fig. 8. Fig. 7 are here inserted; so that Figs. 7 and 8 combined give a representa- tion of the working duplex. The polar relay controls the local circuit, passing through its points; the thumbscrews mark the joining of the office wires with those of the instrument; the electromagnet of the pole changer is controlled by means of two keys whose connecting wires join those of the electromagnet at the thumbscrews. A sounder, a six-point switch, a three- point switch, two lamps, and a 23 -volt dynamo complete the outfit for the main office. The current is led first to the three-point switch where it divides; one circuit, called the receiving side, may be traced (dotted line) through the points of the relay, through the sounder, to a lever in the six- point switch which, if turned to the right, conducts the current through a lamp to the ground. The other circuit, called the sending side, may be traced (solid line) through the magnet of the pole changer, through two keys, thence to a lever in the six-point switch which, turned to the right, similarly conducts the current through a lamp to the ground. There are therefore two grounded circuits, with connections as described, the current for which and for many like circuits is supplied by one dynamo. In the six-point switch are shown other two points; to one, marked M, is con- nected a wire extending to a distant branch office, through a sounder there- in, thence to the ground; to the other point, marked N, is connected a wire similarly extended through a sounder and key, thence to the ground. These connections completed, the levers of the six-point switch may be turned from right to left; the use of the duplex is then extended to the branch office; the polar relay works the sounders in both main and branch 1048 TELEGRAPHY. office; the key in the branch controls the electromagnet of the pole changer in the main office. The lamps A and B are in the main office local circuits, and compensate severally for the resistance of the two extensions when the loop is cut out. Half-Atkinson Repeater. The description of the duplex local (office and branch) system prepares the way for an interesting form of repeater by means of which the offices on a single wire of considerable length may repeat into, i.e., alternately send and receive on, a duplex wire or one side of a quadruplex. This apparatus Fio. 9. REPEATERS. 1049 is named by prefixing the word "half" to whatever form of single line re- peater is used; e.g., half-Milliken, or half-Ghegan. To present as many different forms of repeaters as possible within the limits of this article, the diagram (Fig. 9) shows a half -Atkinson. In the upper right-hand corner is represented in skeleton form the duplex local system just described, to- gether with the jack in the loop switch for the placing of the repeater wedge. The apparatus of the repeater is seen to be a transmitter in the lower left corner, a common relay of 150 ohms resistance, two sounders, two keys, lamps, and a small dynamo. In the lower right corner is a jack to which on one side is connected the single line to distant points; on the other side is the main battery. With the wedge, as indicated, inserted in the jack, the main line circuit can be traced from the battery MB through the post and tongue of the transmitter, through the key and magnet coils of the relay, thence back to the jack and main line "out." In addition to the main line circuit there are four others; two of them are extensions of the 23-volt system of the duplex; of these one has in circuit a pole changer, lamp, sounder, and the local points of the common relay, and terminates in a ground; this arrangement places the pole changer in the control of the common relay. The other circuit has within it the local points of the polar relay, lamp, the electromagnet of the transmitter, and termi- nates in a ground; this arrangement places the transmitter (and the single line which passes through its post and tongue) in the control of the local points of the polar relay. Of the local circuits of the repeater proper, one (marked dot and dash) extends from one pole of a 7-volt dynamo through the lower post and lever of the transmitter, through the coils of a repeat- ing sounder RS; thence back to the other pole of the dynamo; another circuit (dotted) runs through the lever and back stop of RS, making con- nection, as shown, with the local points of the common relay. On the base of the relay the connecting posts on the right join the coils of the relay with the main line wires; the posts on the left connect with the local points of the relay. When the transmitter is open the sounder RS is open; the lever makes contact on the back stop, and completes a circuit in which is the electromagnet of the pole changer. Suppose all the circuits closed and ready for work. When a distant office on the* single line writes, he operates the relay through whose local points passes the pole changer circuit; he controls the pole changer and, therefore, the relay at the distant end of the duplex. When the distant office on the duplex writes, he operates the polar relay whose local points control the electromagnet of transmitter T, through whose tongue and post passes the single line. He thus controls every relay on the single line cir- cuit; the response of the pole changer to his own sending (which it is the purpose of the repeater to avoid) is prevented by the bridging of the local points of the common relay through the lever and back stop of RS. The distant station on the duplex may thus communicate with any office on the single line, and conversely. The action of this repeater can be utilized to repeat from one single line into another; when so arranged it is known as the Atkinson repeater, and it is the standard of one of the leading companies. Duplex Repeater. In wires worked in the duplex or quadruplex system, the static capacity of the wire, which plays little if any part in the operation of circuits by the single method, places a limit on the length of the continuous circuit. But the distance between working stations can be greatly extended by the use of repeaters in which, by an arrangement perfectly simple, the pole changer of a second circuit is controlled by the relay points of the first. The long- est regular circuit in the United States is that worked between New York and San Francisco, with six repeaters. The work of the repeater in this and many other duplex circuits has been facilitated by the recent introduction of the J. C. Barclay pole-changing relay. It consists of a polar relay so constructed that two armatures, in- sulated one from the other, move on a common arbor; one armature con- trols the local circuits; to the other is attached the main line which makes contact on front and back stop with the poles of the battery; it is thus a polar relay and pole changer combined. 1050 TELEGRAPHY. The Stearns Duplex. The operation of differential relays like M in the diagram of the quadru- plex, by alternations of "no battery" and "battery," is the principle of the Stearns duplex, which, as the first condenser-using, and therefore static- eliminating duplex in the world, has a certain historic interest. In Febru- ary, 1868, there were in use by the Franklin Telegraph Company a duplex set New York to Philadelphia, and another to Boston; and in August, 1871, by the Western Union Telegraph Company, a duplex, New York to Albany — all without condensers. In March, 1872, the Stearns duplex, with con- denser, went into operation between New York and Chicago, but it has been superseded by the polar system. Reverting to the diagram, the pole changer with its adjuncts, and the polar relay of the quadruplex, are omitted; one pole of the battery is m s |_^VWA WEST d^H'ltm'I'b^ ^^ STEARNS DUPLEX (ONE TERMINAL) Fig. 10. grounded, and the lever of transmitter, T, is grounded through a resistance equal to that of battery, B. This grounds the line through tongue, t, and leaves the battery open at the post, P. The "east " station (not shown) is a duplicate of the "west," and the control of relay, D, by the distant trans- mitter, T' t may be traced as follows. Suppose distant transmitter, T', sends copper to the line when closed, the current dividing equally between the main and artificial lines in distant relay, D', has no effect upon it; but at the west station there is no current in the artificial line in relay, D, so that the current in the main line closes it. Open the key, K', and the line is grounded through the lever of transmitter, T'\ battery B' is open, and there being no current on the wire, relay, D, is open in response to the opening of distant key, K'. Let transmitter, T, now be closed, and trace the control of relay, D, by the distant key, K*. The current, which now flows from the ground through the lever of open transmitter, T' ', to the zinc pole of battery, B, is neutralized in relay, D, by an equal current flowing from the ground through its artificial line in the opposite direction around its cores, so that relay, D, remains open. Now close distant transmitter, T', and the current in the artificial line (i.e., through the rheostat, R) of relay D is over- powered as to its effects by a current on the main line of twice its strength, and relay D is closed. It is thus shown to be controlled by the distant key, K\ irrespective of the position of home key, K, and the conditions necessary to duplex telegraphy are met. QUADRUPLEX. 1051 The quadruplex system of telegraphy allows of two messages being sent in either direction, over the same wire, and at the same time. In theory it is an arrangement of two duplexes, so different in principle as to permit of their combination for the pin-pose designated. If the accompanying diagram of the quadruplex is examined, there will be noticed in it the ( pole changer, polar relay, and all the apparatus of the polar duplex. The polar relay, K, at the "east" station will respond to signals sent by the pole changer, at the "west" in the manner described in the paragraph on the Polar Duplex, so long as the working minimum of current is main- tained. This working minimum can be doubled, trebled, or quadrupled without appreciable difference to the polar relays. In the paragraph on Single Telegraphy, the operation of the single relay, fitted with a retractile 1052 TELEGRAPHY. spring, was effected by opening and closing the key; or, in other words, by alternating periods of "no current" and "current" on the wire. It was further stated, in anticipation of its introduction at this point, that the spring could be so adjusted that a weak current, though flowing all the time through the coils, would not close it. To effect the closing an increase of battery, and therefore of current strength, is necessary, so that the relay, instead of, as in the first instance, responding to alternating periods of "no current" and "current" could be operated by alternating periods of "weak current" and "strong." The diagram, Fig. 11, illustrating the theory of the quadruplex, will be seen on examination to be a combination of the polar and Stearns duplexes, each of which has already been described. The operation of the Stearns duplex in combination differs from that described in connection with Fig. 10, only in that there is always on the wire a minimum of current sufficient to operate the polar side of the quadruplex ; the neutral relays M and M f , identical with that marked D in Fig. 10, are operated by alternating periods of "weak" current and "strong," after the manner of the Stearns. In practice the weak current is technically called the "short end"; the strong, the "long end"; and the diagram shows how, with different methods of current production, viz., the chemical battery and the dynamo, the pro- portioning of the current in the ratio usually of 1 to 3 is effected. The clock-face pole changer operates, as already described, to send when open (see diagram) copper to line and zinc to the ground; when closed, zinc to the line and copper to the ground. If the connections of transmitter T are traced it will be seen to admit to the pole changer one third of the battery when open, and the entire battery when closed; in other words, the move- ments of the transmitter determine a "short" or "long" end to line. At the left-hand terminal transmitter D effects a like result but by different means. In connection with the transmitter are two sets of resistance coils, so proportioned that when transmitter D is closed all the current from the dynamo goes to line; when open, one third of it goes to the line and two thirds is "leaked " off to the ground. One pole of each dynamo is grounded; the other is connected through a lamp to the pole changer in such a way that the rule "zinc to the line when closed, copper when open" holds good. The main line is shown in solid black; the artificial in dotted lines; the rheo- stats and condensers with their retardation coils marked RC are identical in principle with those shown in the polar duplex. In the diagram transmit- ter D with its companion pole changer is closed; transmitter T with its pole changer is open; the effect of these conditions is respectively to close relays M' and K, and to open relays M and F; the reasons for these results have already been set forth in detail in connection with the polar and Stearns duplexes, so that it is not necessary to repeat them here. In short, there is in the quadruplex a pair of polar relays which respond to changes in the direction, not in the strength of the current; and a pair of neutral relays, which respond to changes in the strength^ not in the direction of the current. The diagram shows the apparatus in its simplest form; there are a number of details in connection with its operation, the complete connec- tions for which are rather too complicated for this book. On page 199 of Mavers's American Telegraphy will be found a diagram embodying the full scheme of connections; and Thorn and Jones' Telegraphic Connections con- tains diagrams and detailed descriptions of the systems in general use. TELEGRAPH (ODES. IVIorse, used in the United States and Canada. Continental, used in Europe and elsewhere. Phillips, used in the United States for "press" work. Dash = 2 dots. Long dash =■= 4 dots. Space between elements of a letter = 1 dot. Space between letters of a word = 2 dots. Interval in spaced letters «= 2 dots. Space between words = 3 dots. TELEGRAPH CODES. 1053 A B e D E F G H I J K L M N O P a s T U V w X Y Z & betters. Morse, Continental, Humeral*. Morse. Continental. Punctuation, etc. Morse. Continental. . Period : Colon : — Colon dash ; Semicolon , Comma ? Interrogation ! Exclamation Fraction line — Dash - Hyphen Apostrophe £ Pound Sterling / Shilling mark $ Dollar mark d Pence 1054 TELEGRAPHY. Morse, Continental Capitalized letter Colon followed ) by quotation:" ) c cents . Decimal point If Paragraph Italics or underline - [] Brackets ) m _ _ " " Quotation I marks. Quotation within ) a quotation J Phillips. Period Colon — Colon dash Semicolon , Comma ? Interrogation ! Exclamation Fraction line — Dash - Hyphen ' Apostrophe £ Pound Sterling / Shilling mark $ Dollar mark d Pence Capitalized letter Colon followed by quo- tation: M c cents . Decimal point If Paragraph Italics or underline () Parentheses [] Brackets " Quotation marks Quotation within a ) quotation "' ; " ) Abbreviations in Common "Use. Min. Minute. Bn. Been. Msgr. Messenger. Bat. Battery. Msk. Mistake. Bbl. Barrel. No. Number. Col. Collect. Ntg. Nothing. Ck. Check. N.M. No more. Co. Company. O.K. All right. D.H. Free. Ofs. Office. Ex. Express. Opr. Operator. Frt. Freight. Sig. Signature. Fr. From. Pd. Paid. G.A. Go ahead. Qk. Quick. P.O. Post Office. G.B.A. Give better address. R.R. Repeat. WIRELESS TELEGRAPHY.* Revised by Frederick K. Vreeland. In consequence of the rapid changes which the art of wireless telegraphy is undergoing, it is impracticable to give here more than an outline of the principles involved, with descriptions of a few typical forms of apparatus. For further details the reader is referred to the more complete works on the subject. Wireless Telegraphy, as it is practiced to-day, is based upon the fact that an electrical oscillating system, when suitably proportioned, may become the source of electromagnetic waves, which radiate through space like light waves, and which have the power of exciting oscillations in a conductor on which they impinge. Electrical Oscillations. — The essential elements of an oscillating system are a capacity and an inductance, and means for charging the capacity and allowing it to discharge through the in- ductance. Fig. 1 represents such a system, in which the capacity C may be a Leyden jar, and the inductance L a coil of few turns of coarse wire. A is a pair of knobs sepa- rated by an air gap, and /an induction coil. When the coil I is set in operation the jar C is charged until its potential is sufficient to break down the air gap G. When a spark occurs, the air gap becomes a good conduc- tor, and the jar discharges through the inductance L. If the ohmic resistance is not too high the discharge is oscillatory, and the current surges through the circuit with a frequency ( N 2rr V LC 4 ?1 or, if R is small, Fig. 1. Closed Oscillating Circuit Operated by an Induction Coil. where 1 2n VLC N = Frequency in cycles per second. L = Inductance in henrys. C = Capacity in farads. R = Resistance in ohms. *■' R ^ ^ \/ 7, t AT becomes imaginary, and the discharge is undirectional. The frequency is usually very high; for example, if C = .005 microfarad and L = .02 millihenry, — figures which roughly represent the case cited, — N will be 500,000 cycles per second. Xlectromag-iietic Waves. — Such a closed circuit oscillator may produce very powerful inductive effects, but it gives off little energy in radiation. It may be converted into a good radiator by separating the con- ductors of the capacity, so that the electrostatic field which lies between * Many of the illustrations for this chapter are taken from Maxwell's Theory and Wireless Telegraphy, by L. Poincare' and Frederick K. Vreeland, through the courtesy of the McGraw Publishing Company. 1055 1056 WIRELESS TELEGRAPHY, them may spread out into space instead of being concentrated in the glass of the jar. Figure 2 shows an open circuit oscillator as used by Hertz in the discovery of electromagnetic waves in space. Here the capacity between the spheres Si and S2, and the inductance of the short rod joining them, are both small, and the frequency is correspondingly high, say 50,000,000 cycles per second. Fig. 2. Open Circuit " Dumb-bell " Oscillator, showing Electrostatic Lines at the Moment Before the Air-gap Breaks Down. The high frequency combined with the open character of the circuit makes this oscillator a good radiator. The dotted lines (Fig. 2) represent the electrostatic field just before the air gap breaks down. When the spark occurs and the oscillations commence, these electrostatic lines shrink back Fig. 3. Field surrounding a dumb-bell oscillator when in operation. At the moment illustrated the spheres are discharging and the lines within the large circle show the beginning of a half wave about to be detached. Outside the circle the preceding half wave is started on its journey through space. The oscillator is shown, greatly re- duced, within the small circle. (After Hertz.) into the oscillator; but the shrinking is so sudden that portions of them are snapped off, as it were, forming closed loops (Fig. 3), which go off into space with the velocity of light (300,000 kilometers per second) expanding verti- cally as they go, and carrying energy with them. This is repeated in each hali oscillation, until all the energy is radiated or wasted in internal losses. INTRODUCTION. 1057 The rapidly moving electrostatic lines carry with them a magnetic field, whose lines of force form coaxial circles with centers in the axis of the oscil- lator, expanding continuously as ripples expand about a pebble thrown into the water. Their relation to the electrostatic lines is shown in Fig. 4. This combination of electrostatic and magnetic fields, traveling outward with the velocity of light, constitutes an electromagnetic wave. When such a wave encounters a nonconducting obstacle it passes through it without interference, but if the ob- stacle be a conductor.the mag- netic lines cutting it induce currents which absorb energy from the wave. If the ob- stacle be large, such as a sheet of metal, the wave is com- pletely cut off and reflected as from a mirror; if the ob- stacle be a wire parallel to the axis of the oscillator, it be- comes the seat of secondary oscillations, like those in the oscillator, but weaker. Any instrument capable of detect- ing these oscillations may be used as the receiver of a wireless telegraph system, of which the oscillator is the transmitter. The Antenna. — The Hertzian oscillator shown in Fig. 2 is operative only over short distances. The energy of the waves is limited by the small capacity of the oscillator, and waves of such high frequency are readily absorbed by obstacles. In actual practice the oscillator takes the form of a vertical wire or antenna, supported by a mast, and grounded at the lower end through a spark gap (Fig. 5). Fig. 4. A Portion of the Spherical Wave- front proceeding from an Oscillator. The Full Lines Indicate the Magnetic Force, the Broken Lines the Electric Force. The Direction of Propagation is Perpendicular to Both of these, and is therefore Radial. i Fig. 5. Transmitter with Simple Antenna. Fig. 6. Receiver with Simple Antenna and Coherer. This is equivalent to half of a Hertzian oscillator, the lower half being removed and replaced by the earth. The capacity and inductance are dis- tributed along the whole length of the wire, and the law of their distribution is such that the wave-length is four times the height of the antenna. Thus 1058 WIRELESS TELEGRAPHY. with a wire 50 meters high the wave-length would be 200 meters, and the 300000^000 200 = 1,500,000 frequency = velocity + wave-length, would be cycles per second. A free Hertzian oscillator emits free Hertzian waves, which travel through space like light. A grounded oscillator gives off grounded waves (Fig. 7). They are half waves, whose electrostatic lines, instead of being self-closed, terminate in the earth, to which they are inseparably bound. Instead of traveling always in straight lines, they must follow the contour of the con- ducting surface over which they slide, and so they may cross mountains or travel about the earth. In gliding over the conducting sur- face of the earth they are accompanied by alternating currents in the surface. These currents waste energy in over- coming the ohmic resistance of the surface, with the result of diminishing the intensity of the waves. For this reason the propagation is much better over water or moist ground than over dry or frozen ground whose resistance is high. #55 Fig. 7. Propagation of Grounded Waves from an Antenna over a Curved Surface. A further cause of attenuation of the waves exists in the space through which they travel. When the sun is shining the air becomes ionized, in which state it is partially opaque to the waves and they are more or less absorbed. Where the distance of signaling is great the difference be- tween the strength of signals in the day and their strength at night is some- times very marked . With a grounded transmitter, a grounded receiver is used (Fig. 6). This is another vertical antenna A , with a detector C, connected in series near the ground. B is a battery and R a relay or telephonic receiver. The Coherer, — ODe of the best known detectors of electrical oscil- lations is the coherer. A typical form is shown in Fig. 8. T is a glass tube in which are two tightly fitting silver plugs, E and E', attached to leading-in wires. The ends of the plugs are about .5 millimeter apart, Fig. 8. Coherer — Longitudinal Cross Section. and the space between them contains a mixture of silver and nickel filings, with sometimes a trace of mercury. The tube is then exhausted and sealed. Normally, the filings lie loosely together, and present a high resistance. The coherer is practically open circuited, but under the influence of the electrical oscillations the filings cohere, and the resistance falls at once to a few hundred ohms. If the coherer be connected in circuit with a battery and a sensitive relay (Fig. 9), this drop in resistance will operate the relay and give a signal. The filings continue to cohere after the cessation of the impulse that affected them, but they may be separated by a mechanical shock. Or- dinarily an automatic tapper is arranged to strike the tube whenever the relay gives a signal, and so restore it to its sensitive condition, ready for the next impulse. SYNTONIC SIGNALING, 1059 Fig. 9. Arrangement of Coherer C with Battery B and Relay R Recording Instrument, and T an Automatic Tapper. / is a A simple grounded antenna has a definite natural period of vibration, but its tendency to adhere to this period is weak, and it may execute forced vibrations over a wide range of frequencies. Thus a given receiving an- tenna will respond to the radiations of various sending antennae, with only a slight preference for radiations whose period is the same as its own. Such an antenna constitutes a simple "responsive" system, which is adapted to use on shipboard or between ships and shore, where it is desirable that any station may communicate with any other station in the vicinity. When a number of stations are so close together as to interfere with each other, a responsive system is not suitable, but the apparatus must be made selective, so that any given pair of stations may intercommunicate without interference from the others. The most usual way of securing selectivity is by applying the principle of Electrical Resonance or Syntony. An electrical oscillating circuit may be so constructed as to make it a stiff vibrator, i.e., the positiveness of its vibration period may be greatly increased, so that it will respond readily to vibrations having its own nat- ural period but will be little affected by impulses of a different period: just as a stretched string will respond to a sound to which it is tuned, but not to sounds of different pitch. Damping*. — The criterion of sharp resonance is a persistent oscillation in both transmitter and receiver. In the transmitter there is a certain initial supply of energy stored in the antenna or other charged condenser, and this energy is gradually expended in radiation or in resistance of the conductors and spark gap and other internal losses. The rate at which the stored energy is expended determines the "damping" or rate of decay of the oscillation. In the receiver, energy is received by the antenna and consumed in doing useful work in the detector, or wasted in ohmic and other losses. To secure a large resonant accumulation of energy, all these losses should be reduced to a minimum. In other words, the damping of both trans- mitter and receiver must be small. A simple antenna is a poor oscillator because its energy is radiated rapidly, and the amplitude of its oscillations decreased at a corresponding rate. The curve (Fig. 10) represents the strongly damped oscillation of a dumb-bell oscillator (Fig. 2) as determined byBjerknes. The amplitude falls to yo of its initial value after nine oscil- lations. The oscillation of a simple grounded antenna may decay even more rapidly still, and this is why sharp resonance is impossible between two such simple oscillating systems. 1060 WIRELESS TELEGRAPHY. A closed oscillating circuit (Fig. 1) may be made quite a persistent vi- brator, as little energy is lost in radiation, and the damping of the oscilla- Time-Hundred- .millionths of a second Fio. 10. Discharge Curve of Dumb-bell Oscillator. tions is due mainly to the ohmic resistance of the circuit, such a system is represented by the equation The oscillation of Q ■ Q p ^ y e ■fit cos (yt + «). where t = time, Q = initial charge of condenser, when t = o, q = charge after time t t = R_ 2L A /j_ _ JP y " \ LC 4L*' a = tan — - » y R being the resistance in ohms (or in absolute units). L being the inductance in henrys (or in absolute units). C being the capacity in farads (or in absolute units). The expression cos (yt ■+■ a ) determines the frequency of the oscillation, AT = ~- and is represented by a simple harmonic curve, while the ex- ponential factor, r e -iz' determines the damping, and is represented by the logarithmic curve shown in dotted lines in Fig. 10. For t = T, a complete period, the exponential term becomes which is the ratio of any two consecutive maxima. The exponent -r-j- T is the natural logarithm of this ratio, and is called the "logarithmic decre- ment." (According to the convention of some writers, the logarithmic SYNTONIC SIGNALING. 1061 decrement is defined as the logarithm of the ratio of two consecutive turn- ing points, and hence has half the above value.) /3 In a persistent vibrator of high frequency the ratio — is small, and the equation may be written, ~Pt cos yt. q = Qe' or ■■ Q€ cos ▼ LC 4L* u This form is more convenient than the complete equation, and is sufficiently accurate for practical purposes. Skin Cffect. — The value of R as here used is quite different from the resistance as measured by ordinary methods, owing to the fact that such rapidly oscillating currents are confined to a thin superficial layer on the outside of the conductor. The thickness in centimeters of the skin measured to the point where the current density is — of its value at the surface, is, '-Vb where f solid back devised by Mr. W. W. Dean. In this transmitter, the carbon retaining chamber is formed in the diaphragm, and, therefore, there is introduced by the vibration of the latter an additional x endency to shake up the carbon granules. In detail design and size of parts this transmitter adheres closely to the Bell "Solid- back" model. " Corn IMaster " Type. — Another type of granular transmitter con- siderably used but not so good as the preceding, is that employing a felt washer as the containing chamber for the granular carbon. . Such a trans- mitter depends upon the elasticity of the felt to permit of the relative motions of the electrodes which close the chamber at the front and rear resDectivelv. *• Packing* and "Unpacking*." — A packed transmitter maybe recog- nized by the dullness of the transmitted tone, the life being so far taken out of the tone at times as to render the words indistinguishable. To unpack a transmitter a slight jarring will at times suffice, this being best accomplished by striking the casing sharp, light blows with a hard object. gratings in the front and air ducts at the base. How to Use a OrannlarButton Transmitter. — The electrodes of the transmitter should always be in a nearly vertical plane. The lips should be placed close to the transmitter and the voice directed into the mouth- piece. As the weight of the parts to be moved is considerable, a large pro- portion of the energy of the voice must be expended upon the diaphragm. When used properly, a tone of voice, such as used in ordinary conversation, should be amply sufficient, and of this scarcely any need escape to the sur- rounding air. Induction Coil. — When the battery transmitter was first introduced it was planned to connect it directly in the line in series with the battery and receiver. In this connection the total allowable resistance change in the transmitter is very small in comparison with the total line resistance, and therefore the corresponding current changes in the receivers are small and of little effect. Furthermore, the longer the line, the less proportional part of the total resistance is the changeable part of the transmitter resist- ance, and thus the longer the line, the less the possible transmitting effect. To obviate this difficulty Edison introduced the induction coil connecting the transmitter and battery in circuit with the low-resistance primary and connecting the secondary in series with the telephone and line. With this arrangement not only is the variable transmitter resistance made a large proportion of that of its circuit and this proportion made invariable with the length of the line, but also, by making the number of turns in the secondary winding large in comparison with those of the primary, the generated secondary voltage is made quite high, and thus suitable for long lines. There is yet another effect* viz. : the variable current of the trans- mitter circuit becomes transformed into a true alternating current. Construction of Induction Coil. — The induction coil is almost invariably of the open magnetic circuit type. The core is composed of a bundle of annealed iron wire, upon which is wound the primary, usually of comparatively heavy, insulated copper wire, while the secondary of fine wire surrounds this. Desig*n of tne Induction Coil. — Thus far no general method of computing induction coils has been developed, the best design for any work being found by a "cut and try" method. Usually each manufacturer has determined by a series of experiments, more or less elaborate, that a certain induction coil will give good results when coupled with his trans- mitter and receiver. He will then use this coil until something better is happened upon. Very few comparative tests of induction coils are upon record, and such as are, give no clew to any relation whatever between good transmission and the physical dimensions and electrical constants of the coil. CALLING APPARATUS. 1075 HOOK SWITCH. After attempting in vain to use as a means of calling greatly magnified currents of the telephone type, produced by over-exciting the transmitter, there remained but two alternatives. Of these, one was to parallel the telephone line with a calling line, each line to carry currents of its own type; while the second was to use the telephone line in a double function, switching upon the ends either calling or talking apparatus as desired. This latter method was used, hand switches being adopted until the forgetfulness of users proved that such were most unreliable, a talking and a calling apparatus being frequently inadvertently left connected together in a manner to defeat the whole system. The hook or automatic switch proved a fairly satisfactory means of overcoming this difficulty, being to-day in almost universal use. In the first place the switch lever is pronged to form a support for the receiver, and it should furthermore be about the only visible means of support for the receiver. When the weight of the receiver is upon the prongs, the lever is depressed so that the calling appa- ratus alone is connected to the circuits. On the other hand, when the hook rises in response to a spring, the receiver being removed, the switch operates to connect in the talking circuits. Design of Hook Swi telle*. — Hook switches are of many designs, each manufacturer producing his preferred idea. Many are of equal effi- ciency. The main points to be considered, are: first, to have the switch springs perform exactly the functions desired; second, to be sure that they perform no accidental and detrimental functions; third, to have the motion of the springs limited by positive stops; fourth, to be sure that the weight of the receiver is ample to actuate the switch; fifth, to have a sliding motion at the points of contact which should preferably be platinum tipped; and, sixth, to have the hook prongs so shaped as not to injure the receiver. In explanation of these points, it may be said that in usual systems, the switch lever on rising must connect two contact points to a third in common, as will be seen from later circuit sketches. In the depressed position some- times it is merely necessary to break this connection, and sometimes in addition necessary to make a third connection. As to positive stops it may be said that when switch springs are allowed to come to a position of rest due to their own set, they are quite sure in time to have the position of normal set sufficiently disturbed to disarrange the apparatus. A sliding motion of the contacts over each other is desirable, as the contacts thus become largely self-cleaning. As to the hook prongs, it has probably been noted that nearly all are now provided with ring ends which cannot be forced against the receiver diaphragm. CAIIOG APPARATUS, Calling apparatus has been worked out upon several complete systems. The most obvious one, employing direct current from a battery with push buttons and vibrating bells, while still holding its own for the very short lines of some house systems and for toy lines, has proved unsuited for com- mercial telephony. This system will therefore be ignored here, but it will be mentioned in the sections on House or Interior systems. For general commercial working the polarized bell, sensitive to alternating currents, has proved to be the best. To produce the alternating currents for actuating it, a magneto generator, i. e„ a dynamo having permanent magnets for fields, was long ago adopted, and this fact has given the name to this system, viz., the " Magneto " system. Recently a calling system, a combination of battery and magneto calling has been extensively adopted. With this system, calls for the stations are made by means of the polarized bell with alternating current, while calls towards the central or interconnect- ing station are made by direct battery current operating an annunciator. The sending of the calling signal is effected by merely removing the receiver from the hook. This is the calling system employed with the now prevalent common battery" system. 1076 TELEPHONY. SERIES AUTD BRIDGING SYSTEMS DEFINED. There are two methods of connecting calling apparatus into telephone circuits. The first of these is termed "series," and is that shown in Fig. 8, where it will be seen that the generator and bell are wired in series, and if there be an extension bell as in Fig. 9, this is connected in series also. In the " bridging" system, on the other hand, the generator and bell are pEIP- -^fflp HH5H..r- , ^h M j- Fig. 9. Diagram showing Proper Connections of Extension Bell. Fig. 8. Diagram of Connections of Series Magneto Bell and Telephone Set. connected across'the line in parallel, or, in other words, they are "bridged" across the line. In case there is but one wire used for the line, the earth serving for a return circuit, the bridges are made from the line to earth. Diagrams of bridging sets are shown in Figs. 10, 27, 28, and 68. /-K /-K Fig. 10. As the requirements for operation of the calling apparatus are very different in the series and bridging systems, it will be necessary, from now on, to point out the differences in the apparatus designed for them. THE POIARIZED BE11. The working parts of a polarized bell always include an electromagnet, a permanent magnet, a pivoted armature carrying a bell clapper, and two gongs. These may be disposed with reference to each other in a variety of ways, but always with the same result. It will, therefore, be necessary to consider the most general type only, a diagrammatical view of which type of bell is shown in Fig. 11, and a side view in Fig. 12. The armature is pivoted to vibrate in front of the poles of the electro- magnet, the pivot lying in a plane parallel to the pole faces, being midway between the two poles and so placed with reference to them that the arma- ture cannot touch both poles at the same time. The permanent or polar- izing magnet, usually a very broad U, has one of its poles secured to the middle of the yoke of the electromagnet, while the other extends to a point just beyond and over the middle of, but out of contact with, the armature. The coils of the electromagnets are connected directly together and to the wiring, without movable contacts of any kind. When there is no current flowing in the coils, the electromagnet cores act merely as extensions of the permanent magnet, both poles of it becoming magnetized alike and of opposite polarity to that of the free end of the per- THE POLARIZED BELL. 1077 manent magnet. The armature also becomes magnetized, but by induction, with two free and one consequent pole, the free poles being such that there is an attraction for each by the opposed core of the electromagnet. t These attractions are not equal, except when the armature is exactly in its mid, an unstable, position. In any other position the attraction is greater for the nearer end of the armature than for the other. Thus the armature naturally comes to rest against one or the other pole, as the case may be. When alternating current is put on the line, the first impulse may do one of two things: it may be of direction such as to strengthen electro magnetically the pull of the pole upon which the armature is resting, by adding the effect of the current to that of the permanent magnet, while at the same time decreasing the effect of the other pole by a similar but sub tractive effect; or the current being in the opposite direction may weaken the pull of the poles in proximity and strengthen that of those separated. It is this latter kind of impulse which starts the bell, for the armature will rapidly tilt in response to the changed attractions, only to be tilted back immedi- ately by the succeeding current impulse of opposite sign. This action is .±v^ Fio. 11. Magneto-Generator and Bell Fig. 12. Polarized Bell with Long Core for Ringer of Bridging Bell. repeated for each reversal of the current, the armature and bell clapper making a double vibration for each cycle of the current. For bridging working it will be seen that the bells are shunted directly across the talking instruments, and they must therefore be designed with reference to this effect. It has been found that with a resistaxice of winding of 1000 ohms, using No. 33 copper wire and cores about three inches long, the shunting effect is negligible even when a considerable number of bells are placed across the line. It is essential, of course, that the resistance be all or almost all wound upon the cores Of the bell, as the telephone current being alternating the virtual resistance due to the inductive winding is far greater in effect than the ohmic resistance, and again, as the efficiency of the bells demands the greatest possible number of turns where effective in operating the armature. For series systems the very opposite condition obtains, for not only is the bell always removed from influence upon the talking circuit, but econ- omy demands that the resistance be kept low, especially where several pieces of apparatus are in series. Eighty ohms is the usual resistance for series bells, and the cores are made much shorter than for bridging bells. Recently a type of bell known as "biased" bells has come into use for certain party line systems. Such bells have in addition to the features above mentioned, an adjustable spring which serves to give the armature a bias in one direction so that it will always come to rest against the same pole piece. 1078 TELEPHONY. COarSTRUCTIOlV OF ^agmio generator. As previously noted, the magneto generator is provided with a field by permanent magnets. From two to six U magnets are used, three being the most frequent number. These magnets are usually cold bent from bar steel approximately \" X 1" in section, and after quenching in cold running water from a red heat, are magnetized by stroking. These magnets span a pole frame within which the armature turns. The armature is of the H. Siemens type, usually of cast iron, and wound full with fine wire. The number of turns of wire and the size of wire vary considerably with the use for which the apparatus is designed. The armature is driven by hand through a gear train arranged so that one will ordinarily drive the arma- ture about 1000 revolutions per minute. At this speed the proper potential for operating the bells should be delivered. This latter ranges from forty volts up, series system machines usually generating a higher voltage and less current than those for bridging systems. One terminal of the arma- ture is usually brought out through an insulated shaft pin to a brush, while the frame serves as the other terminal. Fie. 13. There are many points of design upon which considerable thought has been expended. Such is the interposition of a flexible spring coupling between the driving gear and the armature shaft to render the generator noiseless. Another is the proper proportioning of the span of the armature poles and the gap between the pole tips of the field to obtain the most effective wave form. It is generally conceded that these dimensions should be equal for best results. The automatic switch is still another feature. This is a switch so arranged that the generator is cut from circuit except when actually in use. For a series system cutting from circuit of course involves short-circuiting the generator, while for bridging systems the generator bridge must be opened. In some of the older telephone sets a push button serves in lieu of the automatic switch, and in still others the driving handle must be depressed to connect in the generator. In modern sets, some centrifugal or spring device integral with the driving mechanism automatically controls it. A prevalent type of automatic switch is shown in Fig. 13. Here the driving pin rides the sloping gear hub to move the shaft longitudinally to the lef|>. For the Common Battery system, as before mentioned, no generator is required. The bell, however, is exactly that already described. This system will be referred to more fully later, when its operation and circuits are described. FACTORS AFFECTING TELEPHONE TRANSMISSION. 1079 FACTORS AFFECTING TELEPHONE TRA1¥§MI§. SIOUT: ODIJCTA^CE, CAPACITY, »~ESi:STA]¥CE. As mentioned earlier, the telephone current is an alternating current, and is therefore subject to all the influences of inductance and capacity. These are, moreover, exceedingly potent in their effects, because of the very high frequency of the telephone current, and because this current is made up of superimposed waves of many frequencies. Inductance always tends to choke off alternating currents passing through it. While all lines have inductance, that with which we are most concerned is due to coils of wire about a core of iron. Such coils are variously called in telephony, choke coils, retardation coils, inductance coils, and, although not entirely properly, impedance coils. The inductance of a coil such as in a receiver or a bell magnet has a reducing effect equal to a long length of line; and a few small coils in series in a line, or one large one, will have the practical effect of so lengthening it, as to put the transmitting station beyond the reach of the receiving station. It is this choking effect of Fig. 14. Complete Magneto-Bell Post Pattern. Fig. 15. The Bridging Bell. inductance which renders the bridging bell practicable. Inductance has another effect, viz., it distorts and confuses transmission; the reason being that inductance chokes the higher frequency waves, i.e., the high tones, far more than the lower. Even when present in small degree, it gives the transmitted tone a "drummy" sound. The effect of capacity or condensers is also twofold. Capacity placed or bridged across a line conducts the telephone current, but affords a freer path for the higher frequencies. It thus reduces the volume of the whole transmission and distorts by shunting out the high pitches. In series with a line the distorting effect of capacity is just the opposite of this. It obstructs the low frequency and permits the passage of a disproportionate quantity of high frequency current. Capacity exists in its shunting relation, in all lines, because every pair of conductors forms a condenser. When capacity exists in series with a line it is in the form of a condenser of thin plates. It is used in this rela- tion to the line whenever it is desirable to permit the flow of alternating current and to stop the flow of direct current. Similarly it must be under- stood that inductance coils can be used to permit the flow of direct cur- rents and arrest the flow of alternating currents. Capacity and inductance are also used in conjunction, each to partially neutralize the effect of the other. An example of such a use is the shunting by a condenser of a relay 1080 TELEPHONY. the coil of which is necessarily included in series in a talking circuit for sig- naling purposes. Resistance acts just as would be expected, to attenuate the telephone current. As all component periodicities are reduced equally, however, there is no distortion. Leaving out of consideration the conduction of direct currents, the only case in which resistance is of much importance is when it is combined with distributed capacity. For a long time Lord Kelvin attempted to apply his KR (capacity resistance) law to telephone lines, but this law has been found to not fit the case. The best light upon the subject seems to show that the combined effect of distributed capacity and resistance is nearer proportional to the square root of their product, thus, a ^KR, rather than the product itself. Besides these three most important factors, there are several other, though less important, effects. Among these there are losses due to Fou- cault or eddy currents, hysteresis losses, and reflection losses. These last Sjiturns for # 8 B.W.G. 3 turns for 12 N^EL&jGU. Fig. 16. Regular and Pole Transpositions. occur when there is any abrupt and considerable change in the transmit- ting medium. Thus, for instance, where a line of almost no inductance is connected directly to a line of very high inductance, such as is used in the Pupin system of transmission. These reflections are analogous to the reflections of light and sound. In most telephone work little consideration ia given to these last mentioned losses. EARTH CURRENTS. 1081 EARTH CIRREXT§, INDUCTION, CROSS-TALK. When the telephone was first adopted, all lines were worked as "grounded n circuits. That is, but one wire was used in connection with an earth re- turn. As long as the lines were fairly short, and there was an inconsider- able use of the earth for a return for other systems, trouble was experienced due to disturbing earth currents only in times of general magnetic storms. Slight disturbances occurred, however, at all times. It has been found that the earth is subject to continual potential fluctua- tions, usually minute, but changing with great rapidity. These cause disturbing currents to flow over grounded telephone lines. When neigh- boring trolley lines also use the earth as a return, grounded circuits be- come unbearable not only from the earth potential disturbances, but also from induction. This latter effect is due to a mutual inductive action between the telephone and neighboring wires. Induction may be due to electromagnetic or electrostatic effect. The former occurs when the varying field of force about a wire carrying a disturbing current, cuts and sets up a corresponding field about a parallel telephone wire. Electro- static induction is caused by a series of rapid redistributions of the natural B C JBABC B A B C B A B C BAB Lower Cross Arm I Fig. 17. Transpositions on Twenty-Wire Lines. charges in the telephone line in the attempt to maintain a constant electro- static balance in the neighborhood of the disturbing wire. That it is this latter effect to which most line induction may be traced was proved by J. J. Carty in a series of most interesting experiments, reported in 1889 to the New York Electric Club, and in 1891 to the American Institute of Electrical Engineers. Cross-talk is the name given to induction or leakage from one telephone line to another. It is distinguished by the faint sound of voices. metallic Circuits. — With metallic circuits it is possible, though not always practicable, to do away entirely with disturbances. It is, however, almost always practicable to reduce disturbances to a point where they will not interfere materially with conversation. By metallic circuit is meant not merely a two-wire circuit without qualification, but it means an all-metallic line, both of whose limbs have the same and similarly dis- tributed resistance, the same capacity, and the same insulation resistance. Moreover, both limbs should be equally exposed to all disturbing influ- 1082 TELEPHONY. ences. With insulated wires this last condition is easily obtained by twist- ing the two wires about each other to form what is called a " twisted pair.'" With bare wires "transposition" must be resorted to. Open IFire Circuits. — Open wire circuits are carried upon poles, or in cities, sometimes upon house-top fixtures, although this latter type of construction is rapidly disappearing. The principles underlying the construction of telephone pole lines are exactly similar to those for other lines. The factor allowed for wind-pressure and for weight of ice from sleet storms must, however, be proportionally greater than for most other kinds of lines, because of the large exposed surface of conductor. Cross-arms for telephone lines are usually 10 or 6 pin, the wires adjacent to the pole being 16 inches apart and others 12 inches apart. Cross-arms are mounted two feet apart. Poles are usually set to give an average span of 130 feet, i.e., 40 poles to the mile. The requirements for metallic circuits dictate that both wires of a pair shall be of the same diameter and material, and that they shall be placed in adjacent positions on the same cross-arm. Furthermore, at intervals the two wires must change places, in a manner such that both shall have the same average distance from all disturbing influences. This interchange of wires is termed "transposition." In case of extreme exposure, such as where telephone signal-wires are run upon the same poles as high-tension transmission lines, continuous transposition may be resorted to. Under ordinary conditions of telephone practice, it is found satisfactory, however, to transpose the wires upon a system which treats each two cross-arms as a pair, i.e., 20 circuits as a group, and which provides for the transposition of each with reference to its mate and to disturbing untransposed wires, at least once each mile. This brings "transposition poles" one quarter mile, or approximately 10 poles apart. A dia- FiG. 18. English Method gram of this transposition scheme is shown in of Transposing Metallic Fig. 16. Circuit. Fig. 17 shows a diagram of this transmission system, a study of which will show that only those wires furthest apart in the group, transpose upon the same pole. For very long lines a further refinement must be introduced treating four cross- arms as a transposition group, for it has been found that cross-talk will occur between alternate arms of the two-arm system. Fig. 18 shows a method of continuous transposition. Recently > much of the transposition has been of a type known as single pin. This is much cheaper than that shown in Fig. 16. By this method a cross over of two wires is distributed over two spans of the line, the actual cross taking place at one pin of the middle pole. This pin is provided with a double groove transposition insulator, while its mate carries none. In the first span, one wire passes from its own pin position to the base of the glass in its mate's position. It then continues in this position while the mate wire passes over to the position in the second span vacated by the first wire. If both wires be tied to the same side of the insulator at the middle pole there is no danger of a short circuit. The properties of conductors need not be discussed here. Suffice it to say that for open-wire circuits, iron, steel, aluminum, bronze, and copper have been used. Hard drawn copper is undoubtedly standard. Iron and steel are less satisfactory not only because of high resistance, but be- cause of the difficulty of making good permanent joints, of deterioration, and of their highly magnetic properties with attending inductance. CABLES. Conductors laid up into cables were first brought into use to relieve congested or overcrowded pole lines. At first they were of small copper wire insulated with rubber or similar compounds. With the introduction of metallic circuits came the introduction of twisted pair cables. Such cables are of course relatively free from cross-talk so annoying with SAMPLE SPECIFICATIONS. 1083 straight away cables. Because of the very high specific inductive capacity of rubber, and the proximity of the wires of a pair, so high a mutual electrostatic capacity was introduced as to greatly reduce transmission. For aerial lines, rubber cables are yet used in some localities, especially for emergency and temporary work. General practice has, however, substi- tuted the cheaper and far better paper insulated cable for all uses. Properties of Paper Insulated Cables. — - Present day telephone cables are what are known as dry core cables, as the insulation is untreated paper, thoroughly dried. Strips of paper are loosely spiraled about the cable wire, and this is then twisted together in pairs with a lay approximating 3 inches. The pairs are then layed up in reversed layers to form a cylin- drical core which is served with paper or cotton yarn or both. The core is then thoroughly dried by baking, and it is run' directly from the kiln to the lead press which surrounds it with a moisture proof sheathing of either pure lead, or an alloy of lead with 3 per cent of tin, this alloy being tougher than pure lead. The paper used is very porous, and being loosely wrapped the insulation about each wire is largely dry air, and it is this fact to which the low electrostatic capacity and the high insulation of such cable is due. The slightest moisture will greatly impair and may ruin paper cables and the core is so dry that sufficient moisture may be absorbed from the air to injure them. To prevent this, the ends of each length of cable are usually "filled" with paraffin for a few feet, and whenever a cable is cut at an unfilled spot, it is immediately "boiled out" by pouring over it hot paraffin- wax. Probably the greatest number of cables now in use are of No. 19 B. and S. gauge wire, while of those being manufactured the greatest number are of No. 22 gauge wire. For long-distance lines cables have been used of Nos. 18, 16, 13, and 10 gauge. Cables are known, according to their use, as aerial, distributing, under- ground, and submarine. Aerial cables are made as light weight as is con- sistent with durability. The usual sizes are from 15 to 100 pairs. Distributing cables have a thicker sheath than aerial, but are made in about the same sizes. Underground cables are used in conduit beneath the streets. The usual sizes are from 100 to 300 pairs if the size of wire be No. 19, and 150 to 400 pairs if the wire be No. 22. Underground cables have been made up to 600 pairs, but such cables are not practicable at present for general use, as the allowable diameter of cable is limited, on the one hand, by the size of the conduit duct, usually 3 in. in diameter, and it is limited on the other by the electrostatic capacity. The smaller the cable of a given number of pairs the higher the capacity per pair. Until recently submarine cables were all rubber covered and of not over 10 pairs. Now paper submarine cables of far better insulation, less electro- static capacity, and a greater number of pairs of wires have been success- fully developed. These cables are of from 30 to 150 pairs size. The lead sheath is usually thicker than for underground cables, and after being served with jute is covered with an armor of steel wires. The following sample cable contract written by A. V. Abbott sets forth in tabular form the details of several types of cable. SA.UPIS SPECIFICATION EOH lELEPHONE CABLES, (A. V. Abbott.) Gentlemen: — Under the conditions hereinafter specified, please deliver the following enumerated telephone cables free on board cars at freight depot in reel, marked , containing feet of No. B. and S. gauge, pair, aerial (or underground) paper cable, capacity to m.f. per mile, inch plain lead (or with per cent tin) at quoted price of cents per foot reel, marked , containing, etc. Conductors. — Each conductor shall fully and throughout its entire length have the diameter corresponding to the gauge stated above, and not more than . . . 25 . . . not more than . . . 31 . . . not more than . . . 38 . . . not more than . . . 47 . . . not more than . ..59... not more than . . . 95 . . . 1084 TELEPHONY. shall be cylindrical and free from imperfections. The material of the con- ductors shall be soft-drawn copper. Insulation. — Each conductor shall be insulated with one (or two reversed) wrapping of dry paper; the insulation of one conductor in each pair shall be colored blue and that of the other conductor red. lumber of Pairs. — Each cable shall have the number of pairs called for above, plus at least one extra or additional pair for each one hundred (100) or fractional part of one hundred (100) pairs of conductors called for. Twisting:. — The two wires of each pair shall be twisted together with a uniform lay, not to exceed approximately three inches for No. 19 B. and S. gauge and smaller wires, and approximately six inches for larger wires in a complete twist, so as to effectively prevent cross-talk. Cabling-. — The twisted pairs shall be laid up into a cylindrical core, arranged in reversed layers, so that the length of each complete turn shall not exceed thirty inches. Sheath. — The core shall be incased in a cylindrical sheath of plain lead (or an alloy of lead and per cent tin) of the thickness specified above. The sheath shall be free from holes or other imperfections and shall be of uniform thickness and composition. Conductor Resistance. — Each conductor shall have a resistance equivalent to ohms per mile of No. 16 B. and S. gauge cable; ohms per mile of No. 17 B. and S. gauge cable; ohms per mile of No. 18 B. and S. gauge cable; ohms per mile of No. 19 B. and S. gauge cable; ohms per mile of No. 20 B. and S. gauge cable; ohms per mile of No. 22 B. and S. gauge cable. All measurements to be made at 60 deg. F. The conductivity of any wire shall be equal to at least 98 per cent of that of pure copper. Insulation Resistance. — Each wire shall have an insulation re- sistance of not less than three thousand (3000) megohms per mile at 60 deg. F., when tested at the factory in the usual manner, and shall have an insulation resistance of not less than five hundred (500) megohms per mile at 60 deg. F., when installed, spliced, and connected to office terminals; each wire being measured against all the rest and the sheath grounded. electrostatic Capacity. — The electrostatic capacity of the wires shall remain inside the limits specified above (see p. 889) . These limits to apply to measurements of each wire against all the rest and the sheath grounded and at a temperature of 60 deg. F. Packing* and Shipping-. — The cable shall be delivered on reels in lengths specified above. At least eighteen inches of the inside end of the cable shall be brought out through the side of the reel so as to be accessible for testing. This end shall be securely boxed to protect it from mechanical injury. The outside layer of cable on each reel shall be properly wrapped, and each reel shall be incased in stout lagging. Each reel to carry in plain sight the company's name, the above specified identification mark, length and size of the cable. Relive rv. — Reel marked , shall be delivered at on or shortly before 190 — . Pteel marked at on or shortty before , etc. measurements and Tests. — The company reserves the right to send an inspector to the factory to be present during the process of manu- facturing and to test the qualities of the materials used and the electrical properties of the cable before shipping. He shall have the power to reject any material or cable found defective. Such inspection, however, shall not relieve the manufacturer from furnishing perfect material and satis- factory work. Final measurements and tests are to be made after the cable is installed, spliced, and connected to office terminals. In case the cable falls so far short of the above specified requirements that the company is not willing to accept it, the manufacturer will be called upon to examine the work done by the company, and, if able, by remaking splices or repair- ing injuries to the cable received in handling and laying, to bring the cable SPECIFICATION TABLES. 1085 up to the requirements; the cost of the work shall be borne by the com- pany. If such work, however, does not bring the cable up to the require- ments, and the cable is shown to be defective in material or work done by the manufacturer, then the manufacturer shall make the cable good by replacing as many lengths as may be necessary, and shall not be entitled to pay for work done in examining and remaking splices. The company will, if the manufacturer fails to do so, perform all the work of testing and remaking splices, and charge the cost of such work to the manufacturer in case the defect is found to be due to poor material or workmanship on the part of the manufacturer. The manufacturer shall be notified as soon as the company's inspector reports any defects, and he may have a represen- tative present during such tests and work done by the company to detect or repair defects. The company reserves the right to have a representa- tive present whenever the cable is tested or work is done by the manu- facturer in repairing defects. Guarantee. — The electrostatic capacity shall not increase, nor shall the insulation resistance decrease, beyond the specified limits due to defec- tive material, manufacture of workmanship, for a period of years after the cable has been installed. Payments. — Payments for the cable shall be made within thirty (30) days from the receipt of a consignment, except that fifteen (15) per cent of the price of each consignment shall be held thirty (30) days after each separate consignment is installed and accepted by the inspector of this company, who shall make a written report accepting or rejecting the cable within twenty days after installation; in case of rejection a written notice and statement of the defects shall be sent immediately to the manufacturer, and if the manufacturer fails inside of ten days to remedy such defects they will be remedied by the company and the cost deducted from the final payments, or if the percentage is not sufficient to pay for such repairs the manufacturer must refund the difference. (Signed) Telephone Company. SPECiriCATIO]¥S FOR TELEPHONE CABLES. Table ¥. — Capacity of Aerial Telephone Cables. Revised by John A. Roebling's Sons Co. Approxi- Thick- Approxi- Approxi- mate Cost Num- B.&S. Gauge. ness of Capacity per mate mate per Foot, ber of Lead, Mile, External Weight f.o.b. Fac- Pairs. Inch Manufactured. Diameter per Foot tory, Meas. in Mils. in Pounds. in Cents (May, 1907). 10 19 fc .08 to .085 .800 .985 14.0 10 20 A .085 to .09 .760. .9 12.3 25 19 & .08 to .085 1.07 1.7 25.5 25 20 & .085 to .09 .97 1.30 20. 25 22 & .10 to .11 .76 .96 14.5 50 19 & .08 to .085 1.41 2.7 42.5 50 20 & .085 to .09 1.28 2.15 33.8 50 22 & .10 to .11 .99 1.6 25. 75 19 & .08 to .085 1.70 3.45 56.5 75 20 7 .085 to .09 1.56 3.08 48.7 75 22 .10 to .11 1.19 2.2 35.0 100 22 .10 to .11 1.35 2.68 43. 1086 TELEPHONY. Table n. — Capacity of Underground Telephone Cables. Revised by John A. Roebling's Sons Co. Approxi- Thick- Approxi- Approxi- mate Cost Num- B.&S. Gauge. ness of Capacity per mate mate per Foot, ber of Lead, Mile, External Weight f.o.b. Fac- Pairs. Inch Manufactured. Diameter per Foot, tory, Meas. in Mils. in Pounds. in Cents (May ,1907) 25 19 & .08 to .085 1.07 1.7 25.5 25 20 & .085 to .09 1. 1.54 22.5 25 22 A .10 to .11 .790 1.15 16. 50 19 & .08 to .085 1.41 2.7 42.5 50 20 & .085 to .09 1.31 2.45 37. 50 22 & .10 to .11 1.02 1.86 27.5 100 19 \ .08 to .085 1.96 4.6 74.7 100 20 .085 to .09 1.81 4.1 64.5 100 22 | .10 to .11 1.39 3. 46.5 ■ 150 19 I .08 to .085 2.33 5.8 99.9 150 20 .085 to .09 2.16 5.2 86.3 150 22 .10 to .11 1.64 3.77 61.2 200 19 1 .10 to .11 2.24 6.1 116. 200 20 | .10 to .11 2.1 5.47 99. 200 22 1 .11 to .12 1.84 4.45 75.1 250 22 1 .11 to .12 2.03 5.09 89.0 300 22 1 .11 to .12 2.21 5.7 102. 350 22 . \ .11 to .12 2.3 6.3 115. 400 22 * ,11 to .12 2.5 6.8 122. SIZES OF CABLE!. Conduits as now built readily take a 2^-inch diameter cable, and possibly one 2f-inch; so by existing construction, cables are now limited to these sizes, and design must accommodate itself thereto. It appears desir- able to have about seven varieties of cable for subscribers' lines, and three varieties of toll and trunk-line service. An appropriate set of cables is the following: Purpose. No. Pairs. Size of Wire. Capacity per Mile. Subscribers' lines, distributing cable .... Subscribers' lines, distributing cable .... Subscribers' lines, distributing cable .... Subscribers' lines, main and distributing cable Subscribers' lines, main cable Subscribers' lines, main cable Subscribers' lines, main cable Subscribers' lines, main cable Trunk line cable 10 30 50 100 200 300 400 600 75 50 10 19 19 19 19 20 20 22 24 17 14 10 .085 .085 .085 .085 .110 .115 .120 .140 .065 Toll line cable Toll line cable .050 .035 ANNUAL EXPENSES. 1087 AHHTTAJL EXPENSES OF TELEPHONE CABLES. The following has been published as a basis for computation of the annual charges to be made against cables. ' ' Even with the utmost care, and in spite of the apparent protection offered by conduit and sheath, underground cables gradually fail. In some cases life is very long, but from one cause and another, owing to extension, necessary rearrangement of plant, etc., a thousand and one causes operate to injure the cable insulation and deterioration is inevitable and must be pro- vided for, in the depreciation account. 11 For underground main cable from 5 per cent to 7 per cent is a fair annual charge, while for laterals from 8 per cent to 10 per cent is essential. Aerial cable is much more exposed to injury than underground lines, for it is a constant prey to all sorts of additional destructive forces — sleet and wind storms, lightning, crosses with high-potential wires of all kinds; the small boy with a shot-gun or rifle, and hundreds of. other influences con- stantly attack it. Moreover, aerial lines have a shorter life than under- ground ones, as being chiefly erected in districts which are growing rapidly they are soon superseded by conduit work. For these reasons an allowance of 10 to 12 per cent for depreciation for aerial cables is none too great. " The maintenance to which cable wire is subjected will depend very largely upon the rate of growth in the exchange. Where this is rapid there is a constant necessity for rearranging and remodeling cable plant. Under such circumstances maintenance charges will vary from 2 per cent to 5 per cent on the cost of installation. For where growth is slow, and there is but little change in districts, maintenance may fall as low as from H per cent to 3 per cent. With aerial cables 5 per cent for maintenance is the least charge which should be considered. Combining the charges for both depreciation and maintenance the annual expense for underground wire plant should be taken at from 5 to 10 per cent for main cables, from 10 to 15 per cent for laterals, and from 12 to 16 per cent for aerial cables." JLiu-litiiiiisf Arresters. — Many telephone lines are exposed to light- ning discharges and to accidental contact with wires carrying currents which would be destructive to the telephone apparatus and liable to cause fire. All of some lines are exposed while only short portions of others are. In both cases protection is needed although the best practice distributes it differently in the two cases. It is generally conceded that telephone cables run underground in subways wholly given up to telephone purposes are safe, per se. It has been found that three different elements are necessary for com- plete protection. These are : first, an open space cut-out for grounding momentary high-potential discharges; second, a fuse of such caliber as to amply protect the line against abnormal currents; and third, a sneak current protector or thermal cut-out, which operates with a time factor, and protects the telephone apparatus from small currents, which by a gradual heading effect might destroy it. For lines exposed throughout their length, complete protection demands all three types of safety devices on each wire, and at both ends of the line. For lines beginning in cable and with the outer end exposed, the central office end fuses are usually transferred to the outer end of the cable. It is found economical to terminate cables upon frames or strips designed to hold the various protective apparatus. At subscribers' premises the lines terminate upon a protector built up on a porcelain block, and arranged with binding posts for incoming and outgoing lines and for a ground wire. Open space cut-outs almost always consist of two carbon blocks, the one grounded and the other connected to line. These are held tightly against either side of a small sheet of mica. This mica is perforated to permit of sparking between the carbons, and it is of gauge thickness such that 350 volts difference of pressure will strike across between the carbons. Fuses are of various construction and capacity. Best practice pre- scribes a fuse between 3 and 6 amperes rating. Some prefer a fuse mounted upon a strip of mica which is provided with terminal pieces of copper, and some prefer tubular fuses. The tubular fuse has the advantage of quite effectually blowing out arcs, but it has the incidental disadvantage of at times blowing itself all to pieces upon a violent disruption. 1088 TELEPHONY. The kinds of sneak current protector are now almost legion. All depend upon the gradual heating of some substance sensitive to heat, which gives way under some mechanical strain and opens or grounds the line. The early sneak current protectors were often called heat coils, as all contained a coil of fine wire, interposed in the circuit, which became heated upon the passage of current. Later blocks of carbon served as the heat generating member. In practically all cases certain of the parts are held in proper relation by fusible metal or fusible cement, and the mounting springs tend to disturb this relation. When the solder softens, the springs overcome the adhesion and thereby move the parts to open or ground the circuits. An old form is that shown in Figs. 19 and 20, wherein the softening of the solder permits the pin to slide within the coil under the pressure of spring B, which in following grounds the circuit. Many modern heat coils, while Fig. 19. Combination Protector. A, line-post ; F, instrument post ; B, German-silver spring ; CC, carbon blocks ; M, mica sheet ; SC, sneak coil; P, releasing-pin ; G, ground- ing-strip ; D, ground wire. Fig. 20. Plan of Combination Pro- tector. different in detail, operate similarly. A disadvantage of this type lies in the necessity of reheating it for repairs. Recently several types of self- repairing protectors have been produced. One such has a star-shaped latch which, in releasing the grounding spring, resets itself while still warm. Another depends upon shearing a heat softened washer, which latter may be replaced by a new one at any time. CLASSIFICATION OF TELEPHONE LOE§. Every telephone line may be included in one of three classes, according to the extent to which it may be interconnected with other lines. Under the head "Private Lines" is included all lines which have no facility for interconnection. They may be direct, with but two stations, one at each end; or they may be provided with a considerable, number of instruments located in different places. Private lines are largely used in cities by brokers, railways, etc., and in the country upon the premises of individuals or from farm to farm. House or Hotel Systems include lines which are capable of intercon- nection, but which serve a very limited area, usually all within the premises of a single proprietor. Such systems have either one central switchboard, presided over by an attendant or of an automatic nature, or else have a switchboard at every station so that each user may perform his own switch- ing. With this latter arrangement the system is termed "intercommuni- cating." The third class includes the great bulk of telephone lines, namely those connected to an Exchange and capable of interconnection to not only all other lines of the system, but also through toll lines, to other exchange systems. Every exchange district has one or more central offices, where the switching operations necessary for interconnecting lines are performed. In each exchange system the lines are treated in groups according to the geographical location of their stations. The territory fed by each group is called a district. These vary in area according to the so-called telephone density. REQUIREMENTS FOR OPERATION. 1089 THE (ITIKIL OJb Jb ICJK. Every telephone district has its central office, from which all cables and lines in the district radiate, and where there are provided a switchboard for rapidly interconnecting lines for conversation and interconnecting frames where lines may be interchanged, or those which cross the district may be connected together from the approaching to the receding wire- route. The equipment of a central office is the result of gradual experi- ence, one feature after another having been added as the demand for it arose. For a small number of lines a switchboard of the utmost simplicity will suffice for interconnecting them. As soon, however, as the number becomes so large that it requires several operators to attend to their demands, there is difficulty in connecting together two lines appearing in front of two dif- ferent operators and special provision must be made to handle such calls. Three general systems have been developed, the multiple, the transfer, and the automatic. These will all be briefly considered. First, however, it seems best to review the general requirements of operation and the method of handling calls upon small single operator switchboards. fKEQUntEHtEEUTTS FOR SAXISrACTOHY OPERA- TION OF MVITCHBOIRD. A telephone switchboard system must be so designed that: 1. Every subscriber may signal the switchboard and give directions as to his wants. 2. Any line may be connected to any other line. 3. Any line may be signaled from the switchboard. 4. Every subscriber may signal for disconnection. The rapidity, ease, and accuracy with which these operations may be carried on largely determines the value of the system, the only qualifica- tion being that the outlay to obtain speed shall be commensurate with the' saving of operators' salaries and the advantage to the subscribers. Small Switchboard*. — The approved form for telephone switch- boards is not far different at first sight from that of an upright piano. We have running along the front at mid-height a narrow keyboard, beneath which extends the supporting frame and above which extends the appa- ratus space. A view of a small switchboard for not over 100 lines is shown in Fig. 21. In all manually "operated" switchboards the lines of the subscribers terminate in signals and in switch sockets, and there are provided flexible connecting conductors having terminals which register properly with the contacts of the socket switches. These socket switches are called "spring jacks," or, for short, "jacks," and they consist of a guiding thimble behind which are arranged contact springs of sheet metal. The flexible conductors are usually made in two lengths coupled together to form a pair of connecting cords, and there is associated with each such pair some switch by means of which the op- i erator may connect her telephone set to them at will, and also means for applying ringing current to the conducting strands of the cords. Thus far the description holds for all manually operated switchboards, but from this point a differ- entiation must be made between the various systems. For the present the magneto system only will be considered. For this system the switchboard signal for calling the attention of the operator is a "drop," a type of annunciator whose latch releases a falling shutter: hence the name. When a subscriber desires connection, he drives his magneto and throws the drop. Thereupon the Fig. 21. 1090 TELEPHONY. operator answers by selecting one of an idle pair of coils, and inserting it in the jack corresponding to the signal, and then connecting her telephone to that pair of cords. On ascertaining the number of the line desired, she takes the second cord of the pair, inserts it in the jack of the desired line, and pushes the ringing key to call the subscriber. She then disconnects herself from the cords and is ready to proceed with other connections. In all early switchboards, the operator was required to also restore the drop shutter by hand and she must still so do with many. There are, however, a number of admirable combined drops and jacks in use, where the act of answering a call by inserting a plug automatically restores the drop. There is one more piece of apparatus which has not been mentioned. This is the "clearing-out" drop, which serves as a signal for disconnection when a conversation is finished. It is to throw this signal that one turns the magneto-crank before leaving the telephone. In operation the " clear- ing-out" drop is exactly like the calling or "line drop," and indeed, the line drop may serve as a clearing-out drop. As, however, a user may not always desire disconnection when he rings up central during a connection but may desire the further attention of the operator, whenever the drop falls, instead of disconnecting immediately, the operator must first inquire "Through?" or "Waiting?" Because of this, and because the listening key through which she must respond is associated with the cords, it has been found best to associate the clearing-out signal with the cords. Just Fig. 22. Arrangement of Ringing Keys. as with bells, drops may be made with high-inductance and connected directly across the line, or they may be made of low-inductance and become cut out during conversation. For clearing-out drops the former method is always used, while line drops are made both ways. Arrangement of Hinging; Keys. — It was stated above that in calling a subscriber an operator connects alternating-current to the connecting cords. This statement must, however, for accuracy be qualified, as were the current applied to both cords of the pair simultaneously, the fact that the receiver is off the hook at one of the connected stations would not only cause the disagreeable sensation to the listening subscriber of being "rung in the ear," but in addition the call would like as not fail, the bell of the called line being shunted by the low-resistance receiver. Because of these effects, ringing keys are made not only to connect ringing- current to the cord toward the called line, but also to separate the strands of this cord from those of its mate and the listening apparatus of the oper- ator. The exact manner of accomplishing this result will be apparent from the circuit drawings. multiple Switchboard. — As soon as the number of subscribers is so large that the lines are spread out before several operators, if all of these operators are to make connection to any line, then either must two or more operators assist each other on some connections, or every operator must be given access to all lines. Both methods have been tried, and each has proved successful for a certain class of service. It is generally agreed, however, that the multiple switchboard, that in which every opera- tor has access to all lines, is the more efficient. Switchboards of this type are made up of a number of sections or independent frameworks set side THE BUSY TEST. 1091 by side as though one continuous frame. Each such section accommodates two or three operators, and the keyboard is provided with a corresponding number of equipments. Above the keyboard there are arranged sets of jacks and signals, one set for each operator. These are connected to the group of lines which the corresponding operator must answer. Beside these, there is in each section another group of jacks called the multiple. This group contains as many jacks as there are lines entering the switch- board and each line is connected in every section to that jack having a position in the group corresponding to the number of the line. That every operator may have access to every line, a full group of multiple must be within her reach, and this fact limits the practical height and length of the group, and incidentally the maximum number of lines that can be accom- modated upon a multiple switchboard. As may be inferred, the connecting cords previously described serve as the means of making connection. As before the operator answers in re- sponse to a signal using the jack in her small or "answering jack" group Line 1 Line 2 Line 3 BEpu Fig. 23. which corresponds to that signal. In calling the desired line she uses the nearest multiple jack bearing the number of that line. This may or may not be in the section before which she sits, for as the sections are placed side by side, the multiple is continuous from end to end of the switchboard, and it is often more convenient to reach into an adjacent section. Tlie Busy Test. — With a small switchboard it is at all times evi- dent to the operator just which lines are busy. On the other hand with the multiple switchboard, each line being accessible to many operators, some sort of signal must be provided to indicate when a line is busy, as it is impractical to attempt to find out by direct inquiry. The well-nigh universally adopted "busy test" is an audible one, a click being sounded in the operator's telephone if she attempts to connect one of her cords to a busy line. The guide thimbles of the jacks are expanded to expose a considerable surface upon the face of the switchboard, and all thimbles of corresponding number throughout the switchboard are wired together. A test battery becomes connected to this conductor whenever a plug is in position in any of the jacks, this being the condition with the line busy. Now if a circuit containing a telephone be connected to one of the jack 1092 TELEPHONY. thimbles in a manner to complete the test battery circuit a click will be heard in the telephone. To simplify the movements of the operators the tips of the connecting plugs usually serve as the test connection. Thus it a line is called for, the operator selects her plug and touches it against the thimble of the nearest jack of the desired line. If the line be busy the click at once announces this fact positively. If no click is heard the line is free and the connection is completed by inserting the plug. It is always a matter of perplexity to telephone users as to how operators may discover so quickly as to whether or not a line is busy, but from the above description it will be seen that the work of testing a line for busy is practically incidental to any attempt at making a connection with it, and well accounts for the quickness of the busy report. Series-Multiple Switchboard. — The series-multiple switchboard was the first developed. The fundamental circuits of this system are shown in Fig. 23. The jack thimbles serve for the terminals of one wire of the lines, while a spring in each jack serves for the other. With this system a low-resistance drop is used and it must be cut off during con- 1 , lever, S, and button, 1, at Station 4; line wire, 1, wire, d, switch, H, and bell, C, at Station 1, and back to the battery through the common talking wire. When both subscribers remove their receivers from the hooks, the circuits are completed over line wire 1 with the common talking wire as a return. At the close of the conversation the receiver is simply hung upon the hook, and the automatic mechanical device returns the lever to the home po- sition. Fig. 57 shows the application of the Ness automatic switch to an inter- communicating system, using one common and centrally located battery for supplying both the ringing and talking current. The section, TB, of the battery supplies all the microphone transmitter circuits, and the whole battery, KB, supplies the current for ringing the ordinary vibrating bells that are used in this system. In this arrangement it is evident that the number of wires passing through all the stations will in any size of system be three in excess of the number of stations. TWO- WIRE IHrTEItCOlfllflUTflCATI^O TEIEPHOIE SYSTEMS. By H. S. Webb. \*y a two-wire intercommunicating telephone system is meant one that has two wires for each telephone station in addition to the two wires used in common by all the stations in some systems for signaling purposes only. The object of using two independent wires for each telephone station is to eliminate cross-talk. In a single wire system if one wire in use by one pair of telephones over- laps the wire in use by another pair of telephones, there is very apt to be more Dr less cross-talk. Thiy can be avoided by using for each conversation two independent wires; that is, by using what is here termed a two-wire system. If the wires for an intercommunicating system are run in cables, each pair must be twisted together, as in telephone cables used in complete metallic ex- change systems. If not in cable then the different pairs must be fairly well separated, and if two pairs run parallel to each other for any distance the wires should be properly transposed in order to eliminate cross-talk. A two-wire system is shown diagrammatically in Fig. 58. A contact piece, e, is fastened to, but insulated from, the hook switch, in such a man- ner as to close th.3 circuit between d and / when the telephone receiver rests on the hook. A double switch, S, is also required. The latter maybe made in a variety of ways, but is here shown in a simple form in order that the connections may be clearly seen. The two levers to, and n, are mechanically fastened together so that moving one handle will move both levers ; but the two levers must be insulated from each other; P is a simple push- button switch. One common and centrally located battery, RB, is used for ringing the bell at any station. To call up a station the switch, S, is turned until the straps, to and n, rest on the buttons of the number of the station desired and the push-button pressed. It makes no difference whether the receiver at the station where the call originates is on or off the hook, nor is it necessary for the switch at the station called to be at the first, or home, position. However, the levers, to and n, at the station called, must rest on their home positions before any conversation can be carried on. When ringing, only one wire of a pair is used, the wire, F, serving as a common return; but when conversing, both wires of one pair are in use, and neither wire, W ', nor V, is used; thus there INTERCOMMUNICATING SYSTEMS. 121 can be no cross-talk due to induction. The ringing current may cause slight trouble from induction, as it traverses but one wire of a pair. By means of a double contact push-button switch even the ringing current can be made to flow through both wires of one pair and all danger of induction troubles be eliminated. PairS Fig. 58. Two-wire Telephone System. Such an arrangement is shown in Fig. 59. The wiring at Station 1 is so arranged that the station can be called up from any station, no matter in what position the intercommunicating switch Si may have been left. Fig. 59. Two-wire System with Automatic Switch. However, the wiring at this station has been purposely so arranged that the switch must be returned to the home position before anything can be heard in the receiver. 1122 TELEPHONY. When automatic switches are used the switch is automatically restored to the home position when the receiver is hung up. At Station 2 in Fig. 59 the connections are also suitable for use with an automatic switch. If magneto bells and generators are included in each telephone set instead of an ordinary vibrating bell, then the ringing battery and the two battery wires will not be required, and the connections would be as shown in Fig. 3=*4 P4 Pair 2 ^ Pair 3 Fig. 60. Automatic Telephone System with Magneto Bells. 60. This arrangement requires only one pair of wires for each telephone. At Station 1 the wiring is so arranged that only a simple push-button, P, is required for ringing purposes, while the wiring at Station 2 requires a double contact push, P K The wiring at this latter station may give a ' more evenly balanced system, but does not seem to possess much superiority over that at Station 1, which is the simpler. At Station 2 an insulated contact piece on the under side of the hook switch is required. USES OF ELECTRICITY IN THE UNITED STATES ARMY. Revised by Graham H. Powell. The use of electricity is prevalent in nearly every branch of the military art, being employed in the operation of searchlights, manipulation of coast defense guns, ammunition hoists; in range and position finders; for field and fortress telephones and telegraphs ; for firing guns, submarine and sub- terranean mines, and the control of dirigible torpedoes; while electrically operated chronographs are utilized in the solution of ballistic problems. In March, 1906, the President submitted to Congress the report of the National Coast Defense Board, appointed in the previous year "to recom- mend the armament, fixed and floating, mobile torpedoes, submarine mines, and all other appliances that may be necessary to complete the harbor de- fense" of the United States and its insular possessions. That board made the statement in its report that " Electricity has become of vital importance to an efficient system of gun defense." The following were the general recommendations of that board so far as electrical appliances are concerned: 1. That the electrical power for fortification and defense purposes be fur- nished by an adequate steam-driven, direct-current producing central power plant, all machinery to conform in type to approved commercial standards. 2. That each battery or group of batteries, depending upon local con- ditions, be equipped with direct- current generators, gas or oil engine driven, installed as a reserve to the central plant. 3. That searchlights, except such as are in close proximity to the central plant, be provided with and operated by self-contained units. 4. That the torpedo casemates be equipped as heretofore with independent power for submarine-mine purposes, as an integral part of the submarine- mine defense. 5. That when alternating currents are essential, they should be obtained, if practicable, from direct current by means of a suitable converter; or, if more economical, by a separate alternating unit. 6. That the current from the fortification central plants, when not needed for fortification service, may be used for garrison purposes when such distri- bution does not require too large an increase in the size and number of units. 7. That if garrison service requires alternating current, this should be supplied by the central plant, either through a suitable converter or from alternating current units specially installed for the purpose in the central station; such increase, however, and all additional cost due to post lighting being a charge against the proper appropriation. 8. That uniformity of types and accessories is desirable. 9. That all electrical power plants for the use of fortifications shall be operated by the Artillery. SEARCHLIGHTS. Searchlights are used both as offensive and defensive auxiliaries; defen- sive when used by shore fortifications to light channels or by a vessel to discover the approach of torpedo boats; offensive when used as " blinding- lights" to smother the light of an approaching vessel and confuse her pilot. The accompanying illustrations show the searchlight manufactured by Schuckert & Co. of Nurnberg, Germany. The lamp is placed on top of the two lowest longitudinal rods of the cas- ing, and is held in place by four lugs, two on each side. The carbon holders reach upward through a slit in the casing, and there is a small wheel in rear for moving the light parallel to the axis of the reflector for the purpose of focusing it. The trunnions of the casing are fastened to two longitudinal rods on each side, parallel to the axis of the cylinder, and can be moved for- ward or back so that the casing and what is carried with it will have no pre- 1123 1124 USES OF ELECTRICITY IN UNITED STATES ARMY. Fig. 1. Schuckert Searchlight. SEABCHLIGHTS. 1125 ponderance. The trunnions are supported in trunnion beds in the ends of supports which project upwards from the racer. The elevating arc is attached to another longitudinal rod beneath the cylindrical casing and is capable of adjustment on this rod. Engaging in this arc is a small gear attached to a horizontal shaft passing through the right trunnion support and carrying a small hand wheel. This small hand wheel is for the purpose of elevating or depressing the light rapidly. The light may be elevated or depressed slowly by means of a small hand wheel attached to another horizontal shaft in front of theonejust described. This shaft near its center carries a worm, engaging in a worm wheel on a vertical shaft, to which is also attached a bevel gear. This gear engages in another, which is attached to the quick-motion shaft, but is free to turn about it until it is connected with the elevating gear wheel by means of a friction clamp. The relation between the worm and worm wheel is such that a slow motion is obtained. The racer rests upon live rollers and is joined by a pintle to the base ring. Attached to the base ring is a toothed circular rack, into which on the outside a gear wheel attached to a vertical shaft engages. This vertical shaft projects upward through the racer and carries a worm wheel, which engages in a worm carried on a horizontal shaft having a hand wheel. The worm wheel is entirely independent of its vertical shaft, except when con- nected with it by means of a friction clamp. When so connected, by turn- ing the hand wheel the light is traversed by a slow motion. To traverse the light rapidly, the friction clamp is released and the light turned by hand, taking hold of the trunnion supports. One of the ends of the slow motion elevating and traversing shafts is connected with a small electric- motor, which is encased in a box on top of the racer. By means of these motors the motion of the searchlight can be controlled from a distant point. A switch is provided with contacts so arranged that the current can be passed into the armatures of the motors in either direction, so as to obtain any movement the operator may desire. The current needed for the move- ment is obtained from the lines supplying the current used in the light itself. The current is brought to the motors by means of contact points, bearing on circular contact pieces attached to the racer. The reflector is a parabolic mirror embedded in asbestos in a cast-iron frame, and is held in place by a number of brass springs. The frame of the reflector is fastened to the overhanging rear ring of the casing with studs and nuts, the overhanging part of the ring protecting the reflector from moisture. In order to enable the operator to observe the position of the carbons and the form of the crater while the apparatus is in use, small optical projectors are arranged at the side and on top of the casing, which enables images of the arc as seen from above and from the side to be observed. When the light is properly focused the positive carbon reaches a line on the glass on top of the casing. There are two screws on the positive carbon holder which enable the end of this carbon to be moved vertically or horizontally to bring it to a proper adjustment. In consequence of the ascending heat the carbons have a tendency to be consumed on top ; and to avoid this there is placed just back of the arc and concentric with the positive carbon a centering segment of iron, attached to the casing, which, becoming magnetic, so attracts the current as to equalize the upward burning of the carbons. In taking the light out of the casing this centering segment must be unfastened, and swung to the side on its hinge. An example of the method of calculating the intensity of the lignt sent out by the mirror follows : Diameter of parabolic mirror, 59.05 inches. Diameter of positive carbon, 1.5 inches. Diameter of negative carbon, 1 inch. Power consumed, 150 amperes X 59 volts. Maximum intensity of rays impinging upon the mirror, 57,000 candle- power. Average intensity of rays impinging upon mirror, 45,600 candle-power. Diameter of crater, 0.905 inch. Intensifying power of the mirror — = .. ' . =4,253. a 1 (0.y05.r 1126 USES OF ELECTRICITY IN UNITED STATES ARMY. s=s GENERATOR SAFETY FUSE Fig. 2. Diagram showing Searchlight Connections. DATA RELATIVE TO SEARCHLIGHTS. 1127 •jaAvod-aipii'BQ m jqSil jo/Lvg jo ^isna^ui •in m -in oooi %* uoi^uiurani J° PPM oaoo T* ^ lO ? " t-ooio 3 CO CO •uoisjodsid OOO CO o o o O COO} o o o , ~3 OS O* O* O* O* N ih co t>; 13 oi -t< O l-^OM nriH ^ Ol C<« •jOAiod-aipuBO piupireis at joj.iij\[ Xq pa^oona^; SxC^^ 500 ooo OOO OOO >oo oop o o; co coo ICO"^ CWO OOO^} I3\> sd CJ2 « °'§ Mtf ;* o GO 99 >>W)fl £ a •fi «H £< fl £ ^38 883 S3 §88 838 888 8 Cl CO iO iCMih O L0 O O ja^od-otpn^o pj^pu^s ooo ooi-j i-j « »o looo # © °1 "* •^•^■^5 rt USES OF ELECTRICITY IN UNITED STATES ARMY ^ffirtHft- CONTROL SWITCH When Control Switch handle is turned toward plug socket (9), connectious are as follows :- 1 with 2, 3 with-4, and 5 with 6. When in positioD shown 7 is connected ^groons P^" with 8, Point 6 on Control Switch should be connected to different ground from point marked G. Fig. 17. Field Wireless Set- Pack. Trunk Type, Wiring Diagram. ELECTRIC AMMUNITION HOIST. 1147 ELECTRIC AJIOKUltflTIOltf HOIST WITH AUTO- MATIC IAFETY STOP. As its name implies this apparatus is used for raising ammunition from the magazines to the gun positions. It is applied to two platforms, G G, Fig. 18, either of which is drawn upward, while the other descends, by a winch driven by a motor through worm or train gear. A 5-horse-power motor can raise 2,000 pounds counter- weighted by 600 pounds of the other platform at the rate of 1 foot per second. The design is simple, inexpensive, and the motor and hoist are fairly well protected. 1. M is the motor with both series and shunt fields, the latter being excited when MS is closed. RS is a three-pole reversing switch shown in position for the right-hand platform to ascend. 2. The controller has a starting rheostat, Rh; a hand lever, W ; a spring lever, V; an underload release, UL; and an overload release, OL. The Fig. 18. Ammunition Hoist. magnet UL depends for its excitation upon the voltage of the motor termi- nals and also upon the integrity of its circuit at any one of the four points OL, RS, E, or F. The main circuit from MS is through the electromag- netic brake EB, series fields OL, to the contact piece b, when the lever V is held down by UL magnet, the circuit is closed from b through d, V, W, Rh (or direct after the motor has attained full speed), to RS, M to MS. 3. The main circuit is broken either when the lever V is released (e and / taking the spark), or when W is moved to the left (k and I taking the spark). The lever V, when released by UL, is carried to the right by the spring at its axis until it strikes W. The rheostat may be designed for running the motor continuously at different speeds, or as a starting box not to be in the circuit longer than thirty seconds. 4. S is a baby switch held open by a spring. Its object is to close, if desired, the UL magnet circuit when open at E or F. 5. A and A are the devices for automatically breaking the circuit through UL, and thus the main circuit when the platform ascending strikes the lug g, which is adjustable on the bar sliding in guides h. On the lower end of 1148 USES OF ELECTRICITY IN UNITED STATES ARMY. this bar an insulate copper wedge makes, when down, contact between two copper terminals at E or F, and breaks it when up, thus making or breaking the circuit through UL. E and F are alike and adjustable vertically 6 inches. 6. The right-hand platform is at its upper level, the left-hand is at its lower; the circuit through armature M has been broken and V is up against W. If now we try to start the motor without reversing RS, the circuit through M will still be open at E. But throw RS down and the circuit through UL will be closed at F, and the left-hand platform can be raised. 7. To start the motor at all, TF must always be brought up to the left, pushing V before it until held by the underload magnet UL, then W may be moved to the right, closing the circuit first through Rh and at last with- out it. 8. When the left-hand platform, on nearly reaching its upper level, engages g and opens F, the main circuit will be opened at b and the motor will stop. 9. If it is necessary to move the platform farther up after the circuit has been broken at E or F, the switch S may be closed and the platform may then be moved by the motor. So long as S is closed V will not be released except for no voltage or overload. 10. The motor may be slowed down or even stopped by moving W to the left, provided Rh is large enough to carry the current. 11. The electromagnetic brake on the gear wheel next the motor arma- ture automatically clamps it whenever the main current ceases and the motor stops. It gives a quick stop for heavy or light loads. 12. If the electric machinery is disabled the motor is quickly thrown out and the platform can still be raised by a crank handle and gearing. XIOHT SIGHTS. Electric night sights for rapid fire guns consist of a fitting and stem which can be inserted in the front sight bracket in place of the bead sight used in daylight. This fitting receives an encased white electric light which illuminates a glass cone set under a pierced cap, so that the point of the cone only is visible as a bead to be used in aiming. The light proper is shipped into a holder and down over two plug pins to the other end of which the cable wires are soldered (Fig. 19). The rear edge of the rear (WHITE GLASS ^23 COPPER WIRES-.500 IN. DIA- -SOFT RUBBER INSULATION ^BRAIDED COTTON LENGTH OF CABLE- 3.FEET S ° FT RUBBER - Fig. 19. Front Electric Light and Plug Connections. sight ring is grooved and the groove baked full of scarlet enamel, which is illuminated by an encased red electric light, fitted similarly to the front light. Power is obtained from a battery consisting of ten O.K. dry cells, No. 4, If by 2| by 5f inches high. Four cells are connected in series through a rheostat to each lamp, a fifth cell in each case being held in reserve to put into the circuit when the four cells fail to give proper light. For use at night, range finders are equipped with lights for illuminating the cross-wires of the instrument. The illuminating device consists of two small electric lamps in sockets attached to the rear, or eye-piece, end of the telescope, the beam of light from each lamp being reflected on the cross- wires by two small mica mirrors. The lamps are approximately $ c.p., and 4 volts. Power is obtained from the main lighting circuits through Bui table resistance. FIRING MECHANISM FOR RAPID FIRE GUNS. 1149 HROG MECHAHISm IOII RAPID IIRE «£TJ]¥S. The electrical power for firing rapid fire guns is obtained from two O.K. dry batteries, each consisting of eight cells in series. These batteries are not used simultaneously, but one is kept for use in case the other should fail. Each battery is stowed in a covered box, carried in brackets bolted to the side frames of the gun carriage. A third box is similarly carried for stowing the alternative firing cable. The battery carried on the left is ordinarily used to fire the piece through the pistol connection, while the one on the right is used with the alternative firing key. One terminal of each battery is attached by a short cable to the frame of the carriage as an earth connection. The other terminal of the battery on Fig. 20. the left side of the frame is connected by a cable 4 feet long with the front nipple under the pistol (Figs. 21 and 22). When the trigger is pulled the cir- cuit is completed to the rear nipple, from which a cable, 5 feet 5 inches long, passing under the cradle and through a twisted hook to the right side connects with the contact surface plug. This is bracketed to the cradle in such position that when the piece returns into "battery" from recoil, the contact pin, pressed out by a spring in the contact-pin plug, attached to and moving with the recoil band and piece, presses upon the contact sur- face of the plug before mentioned. The connection for the next shot is thus made. From the contact-pin plug the firing-pin cable extends through a locking pin at the hinge of the breech mechanism to the firing pin, the last 10 inches being armored for protection (Fig. 22). To enable the cannoneer who fires the piece to ascertain whether the breech block is entirely closed and the connections otherwise complete, a buzzer is incased with the pistol, (Fig. 20) so that when the button over the trigger is pressed by the thumb a circuit is completed through a resistance coil, which permits just enough current to pass to sound the buzzer, but not enough to explode the primer, if kept on for an instant only. The ear must be held close to the buzzer 1150 USES OF ELECTRICITY IN UNITED STATES ARMY. lt- J 4^vMHf i ■ taH i^fefc fFitf W FIRING MECHANISM FOR RAPID FIRE GUNS. 1151 to detect the sound. When the trigger is pulled, a direct circuit is completed, permitting the full current from the battery to pass through the primer, thus firing the piece. In case the pistol or its connections become short circuited, or the insula- tion fails, the cable can be quickly disconnected from the battery and firing Fig. 22. pin and the pistol lifted out of its slot. The surface-contact plugs are then disconnected by withdrawing the locking pins which engage with bayonet studs in the contact-plug block, after which another pistol and cables may be applied or the alternative firing key and cables used. In the alternative battery, in the front box on the right side of the frame, the other terminal is directly connected with the firing pin through the TWISTED WIRE CABLE, 6. IN. LONG Fig. 23. Alternative Firing Key and Cables. alternative firing key and cables about 11.5 feet long. The length of these cables is such that the key may be taken under the piece to the left side and used by the cannoneer who is aiming. The alternative key (Fig. 23) consists of a tube into one end of which a cable end is coupled fast. The cable entering the other end is secured to a plunger which is held out by a coiled spring. When grasped in the hand 1152 USES OF ELECTRICITY IN UNITED STATES ARMY. with the thumb on the plunger end, the cable ends may be pushed together, completing the circuit. To guard against a premature discharge of the CONTACT PIN PLUG CONTACT SURFACE PLUG Fig. 24. piece, a split key is wired to this firing key to prevent forward movement of the plunger, and this is kept pushed under the plunger head until the piece is about to be fired. Figs. 21 and 22 show the connections for both night sights and firing circuits, and Fig. 24 gives details of the contact plugs. ELECTRICITY IN THE UNITED STATES NAVY. Revised by J. J. Chain. At the present time (January, 1908) the standard practice on ships of the United States Navy is to use direct current, at 125 volts, distributed on the two-wire system. Previous to 1902 the standard was 80 volts, conse- quently many vessels have apparatus of that voltage. A ship's installation is conveniently divided into dynamo room, lighting system, power system, and interior communication system. The wiring of each system is kept entirely separate from the other. The dynamo room contains the generating sets, main switchboard, and sometimes condensers for the engines. The lighting system supplies all ship's lights, searchlights, and signal lights. These are installed in two separate systems called " Battle Service " and " Lighting Service." Battle service comprises all lights necessary dur- ing action, and these lights are arranged so as to be invisible to the enemy. Lighting service comprises the additional lights necessary for ordinary hab- itation. The power system supplies the various electric auxiliary machinery which at present consists of all ammunition hoists, turret turning gear, elevating and ramming gear for the larger guns, boat cranes, deck winches, ventilat- ing fans, water-tight doors, and motors for driving line shafting in laundry and engineer's workshop. Anchor handling gear and steering gear are at present always steam driven, but electric devices are being experimented with. The auxiliaries in the engine and boiler rooms, consisting of numer- ous pumps and the forced draft fans, are all steam driven, except in a few vessels not yet finished where electric forced draft fans are being installed. The interior communication system consists of various devices for trans- mitting signals and orders from one part of the ship to another. Most of these are electric, but in some cases they are paralleled by mechanical equivalents, as, for example, voice tubes paralleling telephones. DYNAMO ROOM. The generating plant is located in a compartment called the " Dynamo Room," which is under the protective deck and adjacent to the boiler rooms (when practicable), so as to secure a direct lead of steam pipes. OIYEHiTIYGSETS. The following are the principal requirements contained in the standard specifications for reciprocating generating-sets : General Requirements. Each set to consist of an electric generator direct-coupled to a steam engine, both mounted on a common bedplate. The sets as a whole shall be as compact and light as is consistent with a due regard to strength, durability, and efficiency. The standard nizes, with their corresponding maximum allowable speeds, weights, and over-all di- mensions are : Size in Revolutions Weight in Length in Width in Height in kilowatts. per minute. pounds. inches. inches. inches. 2.5 800 560 32 20 30 5 750 1,300 50 28 40 8 550 2,500 64 34 50 16 450 5,600 78 40 60 24 400 7,300 88 48 68 32 400 10,000 101 52 .78 50 400 16,000 110 60 85 100 350 22,000 125 70 95 1153 1154 ELECTRICITY IN THE UNITED STATES NAVY. The design shall provide for accessibility to all parts requiring inspection during operation, or adjustment when under repair. Sets are to he designed to operate right-handed, i.e., counter clockwise when facing the commutator end, or left-handed, as required. The design to be preferably such that the same parts may be used in each, in order to avoid increase in number. The sets must be capable of running without undue noise, excessive wear, or heating. Must be balanced and run true at all loads, up to 33£ per cent above rating ; must be capable of running for long periods under full load and without continued attention. Cast or wrought iron shall not be used for bearing surfaces, except in cases of cylinders, valve chests, and crosshead slides. Both upper and lower halves of main bearings to be removable without removal or displace- ment of shaft. The driving shaft must be fitted with thrust collars or other suitable de- vice which will prevent a movement of the shaft in the direction of its length, as might be caused by the rolling of the ship. The combination bedplate to be a substantial casting, and provided with accurately spaced drilled holes for securing to foundation. An oil groove of ample width and depth to be cast in the upper flange of bedplate, to be continuous around the engine, and to be provided with a stopcock for drainage. The lower side of the combination bedplate to be planed perpendicular to the line of stroke of engine. Seats for all bolt heads and nuts to be faced. All nuts to be case hardened, and to be U. S. standard sizes. Where liable to work loose from vibration, nuts are to be secured by use of jam nuts and spring cotters. All bolt ends to be neatly finished. The two halves of the main coupling to be either keyed to or forged solid with the engine crank and armature shaft. The coupling to be bolted^ to- gether by well-fitted bolts, driving to be done by a cross key set in 'the faces. Adjoining portions of the machinery shall be given corresponding marks whenever this may be desirable for insuring correct assembly. Interchangeability among the different sets and their sparse parte, of the same size and make, as furnished in any one contract, is required. This to be demonstrated as part of the final test for acceptance. Engines are to be of the automatic cut-off vertical enclosed type, designed to run condensing with maximum practical efficiency at all loads, but capable of satisfactory operation when running noncondensing, to be of sufficient indicated horse-power to drive the generator for an extended time at the rated speed, when said generator is carrying a one-third overload. Sizes 2| K. W., 5 K. W., and 8 K. W. to be simple engine, single or twin cylinder at the option of the contractor. Sizes of 16 K. W. and above to be cross-compound with cranks set at 180°. The normal steam pressure under which the engine, running condensing with 25-inch vacuum, for different size sets, is to operate, and the maximum allowable water consumption per K. W. hour output of the set are : K. W. Normal steam pressure. Water consumption per K. W. hour, full load. 2.5 100 105 5 100 90 8 100 65 16 100 44 24 100 40 32 100 37 50 100 35.5 50 150 33.5 100 150 31 ENGINE. 1155 In testing, corrections shall be made by calorimeter for entrained mois- ture. Superheating shall not be used in the test. Engines must run smoothly and furnish the required power for full load at any steam pressure within 20 per cent (above or below) of those given in the above table, and exhausting to condenser at 25 inches vacuum ; to fur- nish power for 90 per cent of full load at steam pressure 20 per cent below normal, and for full load at any steam pressure between normal and 20 per cent above normal, when exhausting with the atmosphere. Must be able to bear without injury the sudden throwing on or on* of one and one-third times the rated full load of the generator, by making and breaking the generator's external circuit. To be so designed that the work done by each cylinder, as shown by indi- cator cards, will be as nearly equal as practicable under all conditions of load. Indicator motions must be provided which will accurately reproduce the motion of the pistons at all points of the stroke. This will require, for cross-compound engines, the operation of the reducing motion for each cylinder from the crosshead or other moving part belonging to that cylinder. Indicator piping to be installed in a manner to secure accuracy of indi- cator cards. Connections to be made at each end of each cylinder, and piped to a three-way cock in order that one indicator may be used for both head and crank ends of cylinder. Connections are to fit the standard indi- cators of the Bureau of Equipment. The length of stroke of the engine to be not less than the diameter of the bore of the high-pressure cylinder. The cylinders to be made of hard, close-grained charcoal iron, bored and planed true, of sufficient thickness for operation after reboring once, steam and exhaust ports to be short, of ample area and free from fins, scales, sand, etc. Cylinders to be fitted with the usual drain cocks, all drains to end in one outlet. In addition to these drains, relief valves are to be fitted to each end of each cylinder, and both high-pressure and low-pressure valves are to be free to lift from their seats to relieve the cylinder of water. The low-pressure cylinder must be fitted with a flat, balanced slide valve ; a piston valve on the low-pressure cylinder will not be accepted. The pistons to be of cast iron or steel, strongly ribbed, light and rigid, and fitted with self-adjusting rings, each piston to have two or more rings. Rings to override counterbore of cylinders, to prevent wear to a shoulder. Piston rods to be of forged steel securely fastened to pistons and cross- heads. Crossheads to be of steel with adjustable shoes. Connecting rods to be of steel with removable babbitt-lined boxes for crank pins and bronze boxes for crosshead pins. The crank shaft to be forged in one piece ; counterweights for balancing reciprocating parts to be forged with it or securely fastened thereto. Valve rods, eccentric rods, and rocker shafts, as well as all finished bolts, nuts, etc., to be of best forged steel. Lagging shall be fitted as extensively as practicable to cylinders, receiv- ers, and steam chests. This shall be done after a preliminary run of the engine in order that any defects in castings or joints may be readily found. The arrangement for securing the lagging in place shall admit of its ready removal, repair, or replacement. The steam and exhaust outlets shall be so placed as to admit of piping from either side with equal facility. Blank flanges shall be furnished com- plete when required to cover alternative outlets. Throttle and exhaust valves to be 90-degree-angle valve, looking up, un- less otherwise specified. Handwheels to be marked, indicating direction of turning for opening and closing. When so directed, larger sizes shall be furnished with by-pass valves for warming up cylinders. The governor to be of the weight and spring type, arranged to operate the high-pressure valve by a shifting eccentric, thus automatically varying the valve travel and point of cut-off. No dashpots or friction washers shall be used in its construction. The speed variation must not exceed 1\ per cent when load is varied between full load and 20 per cent of full load, gradually or in one step, engine running with normal steam pressure and vacuum. A variation of not more than 3$ per cent will be allowed when full load is suddenly thrown on or off the generator, with constant steam pressure either normal, or 20 per cent above normal; a variation of not more than Z\ per cent will be allowed when 90 per cent of full load is suddenly thrown on or off the generator, with 1156 ELECTRICITY IX THE UNITED STATES NAVY. constant steam pressure 20 per cent below normal, exhaust in both cases to be either into condenser or atmosphere. No adjustment of the governor or throttle valve during the test shall be necessary to insure proper per- formance under any of the above conditions. The engine column to be designed to enclose all moving parts as far as practicable, or where weight may be saved, by using a wrought-steel frame with an enveloping enclosure of metal. Detachable hinged doors to be pro- vided for examining moving parts while in operation. The design to elim- inate all chance of oil or water leaking or being forced through. Stuffing boxes for piston rods to be slightly longer than length of stroke, in order that no part of the rod exposed to the oil in the enclosure will enter the cylinder. Stuffing boxes for piston rods and valve rods to be accessible from the outside of the enclosing case of the engine. A guard plate to be provided to prevent oil from being thrown against the lower cylinder heads and valve chests. Engines are required to operate satisfactorily without the use of lubri- cants in the steam spaces. The lubrication for all other working surfaces shall be of the most complete character. No part shall depend on squirt- can lubrication. Forced lubrication shall be used wherever practicable, which includes engine shaft, crank pins, crosshead bearings, eccentric, etc. The engine shall be capable of satisfactory operation with a low grade of lubricating oil, and the forced lubrication shall not be a necessary factor in its cool and satisfactory running. The intent of the forced lubrication is to 'reduce friction, noise, and attention required. The pressure for such forced lubrication shall be approximately 15 pounds per square inch, and shall be between 10 and 20 pounds under all service conditions. The bedplate is to contain a reservoir and cooling chamber of ample ca- pacity, to be provided with a strainer which may be removed without inter- rupting the oil supply. The pump to be direct driven by a crank or eccentric on the engine shaft, construction to be simple and durable, and to include a proper guide or support for the plunger rod. The pump to handle clean oil only, not drawing from the top or bottom of reservoir. To allow inspection while running, the engine crank is not to dip in oil in reservoir. Fly wheel to be turned on face and sides, inner edge to be flanged to retain any oil which may drip thereon. Hub to be split and clamped to shaft by through bolts. A steel starting bar or its equivalent to be fur- nished in sizes of 16 K. AY. and over, the fly-wheel surface to have not less than six holes for starting bar. Mandrels, with collars, complete, shall be furnished for renewing white metal of all bearings so fitted., GENERATOR, To be of the direct-current, multipolar type, compound-wound long- shunt connection, designed to run at constant speed and to furnish a pressure of 125 volts at the terminals, at rated speed with load varying between no load and one and one-third times rated load. The magnet yoke or frame to be circular in form, to have inwardly pro- jecting pole pieces, and to be divided in half horizontally, in all generators above 5 K. W. capacity, the two halves being secured with bolts, to allow the upper half with its pole pieces and coils to be lifted to provide for in- spection or removal of armature. Pole pieces to be bolted to frame, bolts to be accessible in assembled machine to enable removal of field coils with- out disturbing armature or frame. Magnet frame to be provided with two feet of ample size to insure a firm footing on the foundation. Facilities for vertical adjustment of frame to be provided in sizes of 16 K. W. and above. Armature spider to be designed to avoid shrinkage strains. To be accurately fitted and keyed to shaft and to have ample bearing surface thereon. The disks or laminations to be accurately punched from the best quality thoroughly annealed electrical sheet steel, slots to be punched in periphery GENERATOR. 1157 of laminations to receive armature windings. Disks to be magnetically insulated from one another, and securely keyed to spider or held in some other suitable manner to obviate all liability of displacement due to mag- netic drag, etc. Space blocks to be inserted between laminations at certain intervals to provide ventilating ducts for cooling the core and windings. Laminations to be set up under pressure and held securely by end flanges. Bolts holding these end flanges must not pass through laminations. The commutator bars or segments to be supported on a shell, which must be either part of or directly attached to the spider, to prevent any relative motion between the windings and these segments. Bars to be of hard drawn copper finished accurately to gauge. Insulation between bars to be of carefully selected mica and not less than 0.03 inch thick, and of uni- form thickness throughout. Bars to line with shaft and run true, to be securely clamped by means of bolts and clamping rings. Bolts to be accessible for tightening and remov- able for repair. Brushes to be of carbon. In sizes over 5 K. W. there shall be not less than two brushes per stud, each brush to be separately removable and adjustable without interfering with any of the others. The point of con- tact on the commutator shall not shift by the wearing away of the brush. Brush holders to be staggered in order to even the wear over entire surface of commutator ; the generator to be provided with some device for shifting all the holders simultaneously. All insulating washers and brushes to be damp proof and unaffected by temperature up to 100° C. " Finished armature to be true and balanced both electrically and mechan- ically, that it may run smoothly and without vibration. The shaft to be provided with suitable means to prevent oil from bearings working along to armature. All copper wire to have a conductivity of not less than 98 per cent. The shunt and series field coils to be separately wound and separately mounted on the pole pieces. The shunt and series coils, respectively, of any one set to be identical in construction and dimensions and to be readily removable from the pole pieces. The shunt coils as well as the series coils are to be connected in series. In sizes of 15 K. W. and above a headboard is to be mounted on the generator containing the necessary terminals for main switchboard and equalizer connections, shunt and series field connections, pilot lamp, and, if specified, an approved type of double-pole circuit breaker whose range of adjustment shall cover from 100 to 140 per cent of rated full-load current of the generator. Field current not to be broken by the circuit breaker. The field rheostat to be of fireproof construction suitable for mounting on back of switchboard, with handle or wheel projecting through to front, either directly connected or by sprocket chain, handle to be marked indicat- ing direction of rotation for raising and for lowering voltage of generator. The total range of adjustment to be from 10 per cent above to 20 per cent below rated voltage, the variation to be not more than one-half volt per step at both full load and half load. The compounding to be such that with engine working within specified limits, field rheostat and brushes in a fixed position, and starting with normal voltage at no load or at full load, if the current be varied step by step for no load to full load or from full load to no load, and back again, the variation from normal voltage shall at no point be in excess of 2 per cent. The dielectric strength or resistance to rupture shall be determined by a continued application of an alternating E.M.F. for one minute. The testing voltage for sets under 16 K. W. shall be 1,000 volts and for sets of 16 K. W. and above shall be 1,500 volts, and the source of the alter- nating E.M.F. shall be a transformer of at least 5 K. W. capacity for sets of 50 K. W. and under, and of at least 10 K. W. capacity for sets of greater output than 50 K. W. The test for dielectric strength shall be made with the completely as- sembled apparatus and not with its individual parts, and the voltage shall be applied between the electric circuits and surrounding conducting material. The tests shall be made with a sine wave of E.M.F., or where this is not available, at a voltage giving the same striking distance between needle points in air, as a sine wave of the specified E.M.F. As needles, new sew- ing needles shall be used. During the test the apparatus being tested shall 1158 ELECTRICITY IN THE UNITED STATES NAVY. be shunted by a spark gap of needle points set for a voltage exceeding the required voltage by 10 per cent. With brushes in a fixed position there shall be no sparking when load is gradually increased or decreased between no load and full load ; no detri- mental sparking when load is varied up to one and one-third times rated load ; no flashing when one and one-third load is removed or applied in one stage. The jump in voltage must not exceed 15 per cent when full load is sud- denly thrown on and off. The temperature rise of the set after running continuously under full rated load for four hours must not exceed the following : Method of measure- ment. Maximum allowable rise in °C. Armature Electrical .... Thermometer . . . Electrical .... Electrical .... Thermometer . . . 33$ 40 33$ 75 40 Commutator Field coils . • Shunt rheostat Series shunt The rise of temperature to be referred to a standard room temperature of 25° C, and normal conditions of ventilation. Room temperature to be measured by a thermometer placed 3 feet from commutator end of the gen- erator with its bulb in line with the center of the shaft. The generator to be capable of satisfactory operation for a period of two hours carrying one and one-third times its rated full load, and no part shall heat to such a degree as to injure the insulation. Generators of the same size and manufacture to be capable of operation in parallel, the division of the load to be within 20 per cent throughout the range. The magnetic leakage at full load shall be imperceptible at a hor- izontal distance of 15 feet, measurements to be taken with a horizontal force instrument. The minimum allowable efficiencies of the generators are as follows : Loads. K. W. 11 1 i * Per cent. Per cent. Per cent. Per cent. 2.5 78 78 76 73 5 80 80 78 75 8 84 84 83 80 16 87 87 86 84 24 88 88 87 85 32 88 88 87 85 50 89 89 88 86 100 90 90 89 87 SPECIFICATIONS FOR TURBO-GENERATING SETS. 1159 Typical Results of Tests on Generating- Sets Supplied under Above Specifications. Size. 100 K.W. 50 K.W. 32 K.W. 24 K.W. Water consumption per K.W. hour ; Normal steam and vacuum lbs. 29.8 31.5 29.7 28.2 35.0 35.5 35.6 33.4 34.0 34.1 Engine regulation % Full load to no load Normal steam and vacuum 2.77 1.35 2.8 2.66 1.9 2.9 2. 2.5 2.9 1.0 Engine regulation % Full load to no load 20% above normal steam with vacuum *2.65 1.2 1.96 1.75 2.4 3.0 2.65 Engine regulation % Full load to no load 20% below normal steam with vacuum ' 2.8 2.24 3.17 3.27 2.5 2.09 2.67 3.0 3.6 5.0 Generator efficiency % Full load 91.3 91.7 89.5 89.1 89.3 88.8 89.1 88.8 88.2 88.6 88.7 Temperature rise in Armature coils By resistance, °C 32.5 33.3 22. 18. 24.8 20.8 19. 22. 25.1 20.1 23.2 Temperature rise in Field coils, shunt By resistance, °C 29. 31. 24. 24. 30.7 18.1 26.7 20.8 19.2 21.3 19.0 Temperature rise Commutator By thermometer, °C 24. 28. 24.5 23. 19. 13. 14.5 15. 17. 29. 21. SPECIFICATIONS FOW TURBO-OE]¥ERATOG SETS. Each set to consist of an electric generator driven by a steam turbine, both mounted on a common bedplate. The set as a whole shall be as compact and light as is consistent with due regard to strength, durability, and efficiency. The maximum allowable normal speed, weight, and over-all dimensions are: Size in K.W. R.P.M. Weight in lbs. Length in inches. Max. width over pipe connections. Width in inches base. Height in inches. 200 300 1700 1500 25,000 29,000 150 165 inches. 100 100 75 76 87 90 The design shall provide for accessibility to all parts requiring inspec- tion during operation, or adjustment when under repair. Sets are to be designed to operate counter-clockwise when facing the steam inlet. The design to be preferably such that the same parts may be used in each, in order to avoid increase in number. 1160 ELECTKICITY IN THE UNITED STATES NAVY. The sets must be capable of running without undue noise, excessive wear, or heating. Must be balanced and run true at all loads, up to 33$ per cent above rating; must be capable of running for long periods under full load. Cast or wrought-iron shall not be used for bearing surfaces. Both upper and lower halves of main bearings to be removable without removal or dis- placement of shaft. Suitable thrust bearings will be provided to prevent movement of the shaft in direction of its length as might be caused by rolling of the ship. Sets to be erected with shaft extending in a fore and aft direction. The combination bedplate to be a substantial casting, and provided with accurately spaced drilled holes for securing to foundation. Provision will be made to receive duct from the ship's ventilating system. Seats for all boltheads and nuts to be faced. All nuts to be case hardened and to be United States standard sizes. Where liable to work loose from vibration, nuts are to be secured by use of jam nuts and spring cotters. All bolt ends to be neatly finished. Adjoining portions of the machinery shall be given corresponding marks whenever this may be desirable for insuring correct assembling. Wrenches and lifting eyes to be furnished in sets as specified. Canvas covers to be furnished for each set, engine covers and generator covers to be separate. To be made of Navy standard G-ounce khaki cotton ravens (Specification 215) stitched together with a double seam. If required in advance of delivery of set, templates of the combination bedplate or of the shunt field rheostat shall be furnished by the contractor free of additional expense. These may be of paper, full size, with dimen- sions entered complete in order to obviate errors due to shrinkage or expan- sion. Interchangeability among the different sets and their spare parts of the same size and make as furnished in any one contract is required. This to be demonstrated as part of the final test for acceptance. Spare parts supplied to be boxed and protected in accordance with "Specification 3B2" issued by the Navy Department, September 12, 1906. The general appearance of the set resulting from design and workman- ship must be of the highest character. Any defect not caused by misuse or neglect, which may develop within the first six months of service, to be made good by and at the expense of the contractor. The works in which the construction of the contract is being carried on shall be open at all times during working hours to the inspection officer and his assistants. Every facility shall be liven such inspectors f ;r the proper execution of their work. Copies of the original shop drawings of the generating set shall be fur- nished as part of the contract as soon as possible after said contract is awarded. Before final acceptance of generating set a complete set of first- class detail and assembly drawings on tracing cloth shall be supplied. Turbine. The turbine will be of the horizontal multi-stage type. It will be de- signed to run condensing with maximum practical efficiency at all loads. It will be of sufficient power to drive the generator for an extended time at the rated speed when said generator is carrying 1£ *oad. The normal steam pressure under which the turbine will operate, and at this steam pressure the maximum steam consumption for various degrees of vacuum, is: Steam K.W . pressure, Water consumption per K.W. hour, full load. 25 in. vac. 26 in. vac. 27 in. vac. 28 in. vac. 200 300 150 200 30£ 28f 28| 26f 27 25* SPECIFICATIONS FOR TURBO-GENERATING SETS. 1161 These rates should be interpreted as dry saturated steam, steam pres- sure being measured at throttle and vacuum in exhaust casing. Super- heating shall not be used in the test. The turbine to run smoothly and furnish the required power for full load at any steam pressure within 20 per cent (above or below) of those given in the table, and exhausting to condenser at 25 inches of vacuum; to furnish power for 90 per cent of full load at steam pressure 20 per cent below normal, and for full load at any steam pressure between normal and 20 per cent above normal, when exhausting into the atmosphere. It will bear without injury the sudden throwing on or off of one and one-third times the rated load of the generator by making and breaking the gener- ator's external circuit. The steam outlets shall be so placed as to admit of piping from either side with equal facility. Blank flanges shall be furnished complete when required to cover alternative outlets, turbine to have exhaust outlet on right or left side as specified. All piping shall be firmly supported at; points close to the turbine, so that the weight of same shall not effect the alignment of the parts involved. Steam inlet valve shall be a combination throttle and emergency valve equipped with strainer intervening between valve and steam line. It will be connected to the emergency governor ia such a way that it will auto- matically close if the speed of the turbine rises more than 15 per cent above normal. Flange drilling to conform with specifications of the Bureau of Steam Engineering. The governor will be of the centrifugal type operating a series of valves. Lagging to be fitted as extensively as practicable to turbine. It shall be done after a preliminary run of the turbine in order that any defects in casting or joints may be readily found. The arrangement for securing the lagging in place shall admit of its ready removal, repair, and replacement. The speed variation will not exceed 2\ per cent when load is varied between full load to 20 per cent of full load gradually or in one step, turbine running with normal steam pressure and vacuum. A variation of not more than 3^ per cent will be allowed when full load is suddenly thrown on or off the generator with steam pressure constant between normal and 20 per cent above normal, a variation of not more than Z\ per cent when 90 per cent of full load is suddenly thrown on or off the generator with constant steam pressure at 20 per cent below normal, exhausting in both cases either into condenser or the atmosphere. No adjustment of the governor or throttle valve during the tests shall be necessary to insure proper performance under the above conditions. The turbines will operate without the use of lubricants in the steam spaces. Forced lubrication will be used on all bearings. The bedplate will contain an oil reservoir from which oil will be drawn by a pump operating directly from the main shaft, and forced through the system. To be pro- vided with a strainer which may be removed without interrupting the oil supply. The oil will be cooled by water which will pass through a coil around which the oil will circulate. Mandrels, with collars, complete, will be furnished for renewing the white metal of all bearings so fitted. The material and design of the turbine will be such as to safely withstand all strains induced by operation at the maximum steam pressure specified. Generator. To be of the direct current, multi-polar type, compound-wound long- shunt connection, designed to run at constant speed and to furnish a pres- sure of 125 volts at the terminals, at rated speed with load varying be- tween no load and one and one-third times rated load. The magnet frame will be circular in form; will have inwardly pro- jecting pole pieces and will be divided in half horizontally, the two halves being secured with bolts to allow the upper half with its pole pieces and coils to be lifted to provide for inspection or removal of armature. The pole pieces will be bolted to the frame. The magnet frame will be provided with two feet of ample size to insure a firm footing on the foundation. Facilities for vertical adjustment of the frame will be provided. 1162 ELECTRICITY IN THE UNITED STATES NAVY. The laminations for the armature will be accurately punched from the best quality, thoroughly annealed, electrical sheet steel, slots to be punched in the periphery of laminations to receive armature windings. The lami- nations will be insulated from each other and will be assembled on the spider or shaft and securely keyed. Space blocks will be inserted between laminations at certain intervals to provide ventilating ducts for cooling the core and windings. Laminations will be set up under pressure and held securely by end flanges. The commutator bars will be supported on the shell which will be keyed directly on the shaft so that no relative motion can take place between the windings and bars. The bars will be of hard drawn copper finished accu- rately to gauge. The insulation between bars will be of carefully selected mica not less than .03 inch thick. The bars will line with the shaft and run true and will be securely held in place by means of clamping rings. The brushes will be of carbon. Each brush will be separately removable and adjustable without interfering with any of the others. The point of contact on the commutator will not shift by the wearing away of the brush. Brush holders to be staggered in order to even the wear over entire sur- face of commutator; the generator to be provided with some devices for shifting all the holders simultaneously. All insulating washers and bushings to be damp proof and unaffected by temperature up to 100 degrees C. Finished armature to be true and balanced both electrically and mechan- ically, that it may run smoothly and without vibration. The shaft to be provided with suitable means to prevent oil from bearings working along to armature. All copper wire to have a conductivity of not less than 98 per cent. For sets of 100 K.W. and less the shunt and series field coils to be sepa- rately wound and separately mounted on the pole pieces. The shunt and series coils, respectively, of any one set to be identical in construction and dimensions and to be readily removable from the pole pieces. The shunt coils as well as the series coils are to be connected in series. A headboard will be mounted on the generator containing the necessary terminals for main switchboard, equalizing connections, shunt and series field connections, and pilot lamp. The field rheostat to be of fire-proof construction suitable for mounting on back of switchboard, to be provided with handle or wheel projecting through to front, either directly connected or by sprocket chain, handle to be marked indicating direction of rotation for raising and for lowering volt- age of generator. The total range of adjustment, to be from 10 per cent above to 20 per cent below rated voltage, the variation to be not more than one-half volt per step at both full load and half load. Operation of Generator. The compounding to +>e such that with turbine working within specified limits, field rheostats and brushes in a fixed position, and starting with normal voltage at no load or at full load, if the current be varied step by step from no load to full load or from full load to no load, and back again, the difference between maximum observed voltage and minimum observed voltage shall not exceed 2£ volts. The compounding and heat run (full load and overload) of the generating sets must be made with identical brvrh positions. The dielectric strength for resistance to rupture shall be determined by a continued application of alternating E.M.F. of 1500 volts for one minute. Test for dielectric strength shall be made with the completely assembled apparatus and not with the individual parts, and the voltage shall be applied between the electric circuits and surrounding conducting material. With brushes in a fixed position there shall be no sparking when load is gradually increased or decreased between no load and full load; no detri- mental sparking when load is varied up to one and one-third times rated load, no flashing when one and one-third load is removed or applied in one stage. The iump in voltage must not exceed 15 per cent when full load is sud' clenly thrown on and off. SPECIFICATIONS FOR TURBO-GENERATING SETS. 1163 The temperature rise of this set, after running continuously under full rated load with air of auxiliary ventilation at room temperature for four hours must not exceed the following: Degrees G. Armature, by thermometer 40 Commutator, by thermometer 45 Series field coils, thermometer 40 Shunt field coils, resistance method 40 Shunt rheostat, resistance method 75 Series shunt, thermometer 40 The rise in temperature to be referred to standard room temperature of 25 degrees C. Room temperature to be measured by a thermometer placed three feet from commutator end of the generator with its bulb in line with the center of shaft. A system of air ducts for the ventilation of armature and commutator shall be provided. This system shall be connected to the ship's venti- lating system. The amount of air per minute required for the various sized sets will not exceed the following: Size K.W. Cubic feet air per minute. 200 2000 300 3000 The generator to be capable of satisfactory operation for a period of two hours carrying one and one-third times its rated full load; also full load continuously in a room temperature of 30 degrees C, without auxiliary ven- tilating system, and no part shall heat to such a degree as to injure the insulation. Generators of the same size and manufacture to be capable of operation in parallel, the division of the load to be within 20 per cent throughout the range. The magnetic leakage at full load shall be imperceptible at a hori- zontal distance of 15 feet, measurements to be taken with a horizontal force instrument. The dynamo room is supplied by a special steam pipe which usually is so connected that it can take steam direct from any boiler or from the auxil- iary steam pipe, it passes into a steam separator from which branches lead to each of the genera ting-sets in the dynamo room. This separator is drained by a steam trap which sends the water back to the hot well in the main engine room. The exhaust pipe from each set joins a common exhaust which connects with the auxiliary exhaust service of the ship. If the sets are located below the level of the ship's auxiliary exhaust pipe, a separator is placed in the common exhaust pipe before it goes up and joins the ship's auxiliary exhaust. This separator is drained by a small steam pump, which is automatically started and stopped by means of a float in the body of the separator, which float starts the pump when the separator is full and stops it when empty. In the latest vessels a separate condenser is installed in the dynamo room for the generating sets. MVJTCHBOiHDS. Switchboards are divided into: (a) Generator boards. (6) Distribution boards. The generator boards are provided with two sets of bus-bars, one set for the lighting system, and the other set for the power system. The design is such that any of the generators can be operated singly or in parallel on either system. Fig. 1 shows diagrammatically the generator board used on the U. S. S. " Vermont." Current is supplied to the different appliances by means of distribution switchboards, wnich have two sets of bus-bars, one for lighting and one for power, and are connected directly to the corresponding bus-bars on the main generator board. Feeders run direct from these distribution boards, 1164 ELECTRICITY IN THE UNITED STATES NAVY. each feeder being provided with a fused switch. Distribution boards are sometimes located at various parts of the ship and sometimes made con- tinuous with the main board. On several of the first vessels using electric turret turning gears on the Ward-Leonard system of control, a separate generator was used for each turret. This required an additional set of bus-bars on the generator switch- POWER FEEDERS " kotSTAIBtmOt ^Bc/f/fo LIGHTJH6 FEEDERS TO vsTiuatnuA BOARD , - COffMW //EG. s EWUIER Generators NSA Fig. 1. Diagram of Vermont Generator Switchboard. board for each turret. Fig. 2 shows the design as used on the U. S. S. "Illinois," except there are four more generators connected on exactly like the four shown. Each generator has a headboard carrying a double-pole circuit breaker, and clips for a series field short circuiting shunt used for turret turning. The diagram shows generators Nos. 1 and 2 operating in parallel on the power system, No. 3 alone on the light system, and No. 4 operating the after turret turning motors. It is to be noted that the three generators on the power and lighting systems have the right-hand blades of their triple pole field switches closed, giving self-excitation through the field rheostat, while the machine for turret turning has the middle blades closed, jving separate field excitation from the power bus-bars and through the >eld resistance attached to the controller in the turret. t SWITCHBOARDS. 1165 •siy Auvnixnv A r(I> 1 S < »- H S uj CO Q tr > >. a: 3 1166 ELECTRICITY IN THE UNITED STATES NAVY. DOIJBIE DYHrAlWEO UOOHS. Some of the latest ships have been designed with two complete generating plants each in a separate room, one forward and the other aft, so that any k0 J J L-o* I'ofd U Turret Fig. 3. Diagram of Double Dynamo Room Distribution. accident disabling one plant will not affect the fighting ability of the ship. Each plant is of sufficient capacity to carry the entire working load. The distribution is shown diagrammatically in Fig. 3. The generators in one room are controlled by the same board. The feeders to the various WIRING. 1167 parts of the ship are supplied by the two distribution boards, one forward and one aft. Each of these distribution boards can take energy from either of the generator boards by means of transfer switches and interconnecting feeders. The circuits supplying the lights in the engine and fire rooms^ and the turret feeders are made double, one set running from each distribution board, and transfer switches provided at their ends; thus allowing these important parts to be supplied even if either dynamo room or either dis- tribution board is destroyed. 1VIKOG. Specifications. The principal requirements of the Navy standard specifications for light and power conductors are : All conductors to be of soft-annealed pure copper wire, and, unless other- wise specified, each wire to be thoroughly and evenly tinned. All single strands must show a conductivity of not less than 98 per cent and the finished cable not less than 95 per cent of that of pure copper of the same number of circular mils. All layers of pure Para rubber must contain at least 98 per cent pure Para rubber; must be concentric, of uniform thickness, elastic, tough, and free from flaws and holes. All layers of vulcanized-rubber compound shall consist of the best grade of fine unrecovered Para rubber, mixed with sulphur and dry inorganic mineral matter only. The compound shall contain from 39 to 44 per cent, by weight, of fine Para rubber, and not more than 3 per cent, by weight, of sulphur. This sulphur shall be so combined with the Para rubber that not more than two-tenths of 1 per cent shall remain in the compound as free sulphur. The rubber shall be so compounded and vulcanized, that when test pieces taken from the wire (2 inches between jaws and £ inch wide when possible) are subjected to a tensile stress, they shall show a breaking strain of not less than 1,000 pounds per square inch, and shall stretch to at least three and one-half times their original length. The jaws will be separated at the rate of 3 inches per minute. When test pieces, as described above, are subjected to a stress of 900 pounds per square inch for ten minutes, the compound shall be of such a character as to return to within 50 per cent in excess of its original length at the end of ten minutes after being released. All layers of vulcanized rubber must be concentric, continuous, and free from flaws or holes; must have a smooth surface and circular section; and must be made to a diameter in the finished conductor as tabulated. Measured dimensions "over vulcanized rubber " or "over tape" must come within 2\ per cent of tabulated values, the departure in no case to exceed ^ inch. All layers of cotton tape must be thoroughly filled with a rubber-insulating compound, the tape to be of a width best adapted to the diameter of that part of the conductor which it is intended to bind. The tape must lap about one-half its width; must be of such thickness as to make dimensions conform to tabulated values, and be so worked on as to insure a smooth surface and circular section of that part of the finished conductor which is beneath it. The tape must not adhere to the rubber. All exterior braid or braids must be closely woven, and all, except silk braid, must be thoroughly saturated with a black insulating waterproof compound which shall be neither injuriously affected by nor have injurious effect on the braid at a temperature of 95° C. (dry heat), or at any stage of the baking test, nor render the conductor less pliable. Wherever a di- ameter oyer outside braid is tabulated or specified, the outside surface must be sufficiently smooth to secure a neat working fit in a standard rubber gasket of that diameter for the purpose of making water-tight joints. Measured dimensions "over braid" must come within 5 per cent of tabulated values, the departure in no case to exceed ^ inch. All wire and cable shall be subjected to a test for continuity and for insu- lating properties, the latter by measurement of insulation resistance and by high potential test on the entire length of the cables, either or both, as per the following table: 1168 ELECTRICITY IN THE UNITED STATES NAVY. t Insulation resistance. Test voltage, 30 min- utes. Lighting wire. Up to and including: 500,000 cm., single .... 650,000 cm., single .... 800,000 cm., single .... 1,000,000 cm., single .... All twin wire: Between conductors .... From conductors to ground . Double conductor. Plain: Between conductors .... Each conductor to ground Diving: Between conductors .... Each conductor to ground Silk 1,000 megohms per knot .... 900 megohms per knot .... 800 megohms per knot .... 750 megohms per knot .... 1,000 megohms per knot .... 1,000 megohms per knot .... 1,000 megohms per 1,000 feet . 1,000 megohms per 1,000 feet . 1,000 megohms per 1,000 feet . 1,000 megohms per 1,000 feet . No test 4,500 4,500 4,500 4,500 3,500 3,500 2,500 3,500 3,500 3,500 5,000 Bell wire 500 megohms per 1,000 feet . . No test 1,500 Bell cord 5,000 Cable. Interior-communication cable: Between conductors .... Each conductor to ground Night-signal cable; Conductor for 1,000 megohms per 1,000 feet . 1,000 megohms per 1,000 feet . 1,000 megohms per 1,000 feet . 1,500 3,500 3,500 Completed cable: Between conductors .... Cable to ground 1,000 megohms per 1,000 feet . 50 megohms per length .... 3,500 3,500 Tests for insulation resistance shall be made after immersion of wire (not less than three days after manufacture, the three days to be reckoned back from the end of the immersion period) in fresh water at a tempera- ture of 22° C. for a period of twenty-four hours, the test to be made by the direct-deflection method at a potential of 500 volts after five minutes electrification. High-potential tests shall then be made with the wire still immersed, the source of power supply to be a transformer of not less than 5 K.W. capacity. For double-conductor silk and bell cord the high-potential tests will be made with the dry wire freely suspended in the air. Six-inch samples of wire, with carefully paraffined ends, shall be sub- merged in fresh water of a temperature of 22° C. for a period of twenty- four hours. The weight of the wire before and after submersion, deduct- ing weight of copper and vulcanized rubber, will give the per cent of water absorbed by the braids. This shall not be more than 10 per cent. A sample of suitable length (1 foot long for small wires) shall be exposed for several hours at a time, alternately, to a temperature of 95° C. (dry heat) and the temperature of the atmosphere, over a period of three days. The braid and insulation must then stand sharp bending to a radius of seven times the diameter without breaking or cracking. For twin conductor the minimum diameter will be used. Unless otherwise called for, all wire supplies to be delivered in lengths of not less than 500 feet. To be delivered on reels of strong construction to admit of transportation to long distance, which reels on direct purchase? will remain the property of the Government. The flanges of the reels to be WIRING. 1169 at least 8 inches longer in diameter than the diameter through the coil. The loose end of the coil to be secured to prevent damage in transit. To insure maximum flexibility, the pitch of the "standing" or "spiral lay" of all conductors so formed shall not exceed values tabulated: Length of pitch, expressed in Number of wires forming strand. diameters of indi- vidual wires. 7 30 19 60 37 90 61 120 91 150 127 180 When greater conducting area than that of 14 B. & S. G. is required, the conductor shall be stranded in a series of 7, 19, 37, 61, 91, 127, wires, or as may be specified, the strand consisting of one central wire, the remainder laid around it concentrically, each layer to be twisted in the opposite direc- tion from the preceding; and all single wires forming the strand must be of the diameter given in the American wire-gauge table as adopted by the American Institute of Electrical Engineers, October, 1893. Single Conductor. Table of Standard Dimensions: Actual C. M. ^§ %™ S.S 3 N) • S3« Diameter, inches. Diameter in 32ds of an inch. Over copper. Over Para rubber. Approxi- mate C. M. Over vul- can- ized rub- ber. Over tape. Over braid. 4,000 9,000 11,000 15,000 18,000 20,000 30,000 40,000 50,000 60,000 75,000 100,000 125,000 150,000 200,000 250,000 300,000 375,000 400,000 500,000 650,000 800,000 1,000,000 4,107 9,016 11,368 14,336 18,081 22,799 30,856 38,912 49,077 60,088 75,776 99,064 124,928 157,563 198,677 250,527 296,387 373,737 413,639 521,589 657,606 829,310 1,045,718 1 7 7 7 7 7 19 19 19 37 37 61 61 61 61 61 91 91 127 127 127 127 127 14 19 18 17 16 15 18 17 16 18 17 18 17 16 15 14 15 14 15 14 13 12 11 .06408 . 10767 .12090 . 13578 .15225 .17121 .20150 .22630 .25410 .28210 .31682 .36270 .40734 .45738 .51363 .57672 .62777 .70488 .74191 .83304 .93548 1.05053 1.17962 .0953 . 1389 .1522 .1670 .1837 .2025 .2328 .2576 .2854 .3134 .3481 .3940 .4386 .4885 .5449 .6080 .6590 .7361 .7732 .8643 .9667 1.0818 1.2109 7 10 10 10 11 12 12 13 14 15 16 18 19 20 22 24 26 29 30 34 38 42 46 9 12 12 12 13 14 14 15 16 17 18 20 21 22 24 26 28 31 32 36 40 44 48 11 14 14 14 15 16 16 17 18 19 20 22 23 24 26 28 30 33 34 38 42 46 50 1170 ELECTRICITY IN THE UNITED STATES NAVY. All single- lighting conductors shall be insulated as follows: First. A layer of pure Para rubber, not less than 5 \ inch in thicknes*, rolled on. On the larger conductors this thickness must be increased, if necessary, to meet the requirements of paragraph 2 (m). Second. A layer of vulcanized rubber. Third. A layer of cotton tape. Fourth. A close braid to be made of No. 20 two-ply cotton thread, braided with three ends, for all conductors under 60,000 circular mils, and of No. 16 three-ply cotton thread, braided with four ends, for all conductors of and above 60,000 circular mils. The outside diameter over the braid to be in conformity with that tabulated. Twin Conductor. Table of Standard Dimensions: Actual CM. 00 o u O c3 GO PQ .faO Diameter, inches. Diameter in 32ds of an inch. Ap- Over 73 Over tape. Over 1st braid. Over 2d braid. proxi- mate CM. Ob *o Over copper. Para rub- ber. > 3 One con- Two con- One con- Two con- One con- Two con- 3 fc <0 u * duc- duc- duc- duc- duc- duc- W > o tor. 6 tors. 12 tor. 8 tors. 14 tor. 10 tors. 4,000 4,107 1 14 .06408 .092 5 15 9,000 9,016 7 19 . 10767 .139 7 9 18 11 20 13 21 11,000 11,368 7 18 .12090 .156 8 10 20 12 22 14 23 15,000 14,336 7 17 .13578 .172 8 10 20 12 22 14 23 18,000 18,081 7 16 . 15225 .190 9 11 22 13 24 15 25 20,000 22,799 7 15 .17121 .209 10 12 24 14 26 16 27 30,000 30,856 19 18 .20150 .243 11 13 26 15 28 17 29 40,000 38,912 19 17 .22630 .268 12 14 28 16 30 18 31 50,000 49,077 19 16 .25410 .298 13 15 30 17 32 19 33 60,000 60,088 37 18 .28210 .327 14 16 32 18 34 20 35 All twin lighting conductors shall consist of two conductors, each one of which shall be insulated as follows: First. A layer of pure Para rubber, not less than ^ of an inch in thick- ness, rolled on. Second. A layer of vulcanized rubber. Third. A layer of cotton tape. Two such insulated conductors shall be laid together, the interstices being filled with jute, and covered with two layers of close braid. Each braid to be made of No. 20 two-ply cotton thread, braided with three ends. Methods of Installing* Conductors. Three methods of installing conductors are used. 1. Conduit ; 2. Molding ; and 3. Porcelain supports. 1. Conduit is the principal method, being used in almost all spaces below the protective deck, and wherever wiring is exposed to mechanical injury or the weather. Iron-armored conduit is used, except within 12 feet of the standard compass, where brass is used. Conduit passing through water-tight bulkheads is made water-tight by means of stuffing-boxes and hemp-packing. Water-tightness is provided at the ends of conduit by a stuffing-box and a soft-rubber gasket, through which the conductor passes. Long lines of conduit passing through several LIGHTING-SYSTEM. 1171 water-tight compartments are provided with gland couplings at proper intervals, which divide the run into water-tight sections, thus preventing an injury in a flooded compartment from allowing the water to run through the conduit into another compartment. These gland couplings are also used where conduit passes vertically through decks. 2. Wood molding is used in living spaces but has been abandoned on the latest vessels. It consists of a backing piece fastened to the iron work of the ship, to which the molding proper is secured by screws and covered with a wooden capping-piece. Where leads installed in molding pass through water-tight bulkheads, a bulkhead stuffing-box is provided for water-tightness. 3. Porcelain supports are used in dynamo rooms and for the long feeders which are run in the wing passages where there is no danger of interference. Stuffing-tubes are used where the wires pass through bulkheads, the same as with molding. Junction Boxes. All conductors are branched by being run into standard junction boxes, which are usually provided with fuses. Where conduit is used these boxes are tapped, to have the conduit screwed into them ; where molding or porcelain is used the boxes are provided with stuffing-tubes. The box covers are made water-tight with rubber gaskets ; inside the fuses and connection strips are mounted on porcelain bases. IIGHTI^f;.§l§TE^, Wiring". The maximum drop allowed on any main is 3 per cent at the farthest lamp. Mains are required to be of the same size throughout, and to be of 1,000 circular mils per ampere of normal load. fixtures. Most incandescent lamps are installed in air-tight glass globes of different shapes, depending upon position or location. Magazines are lighted by "Magazine Light Boxes," which are water-tight metal boxes set into the magazines through one of its walls, and provided with a water-tight door opening into the adjacent compartment, so that the interior of the box is accessible without entering the magazine. The sides of the boxes have glass windows, and each box is fitted with two incandescent lamps, each lamp having its own separate fused branch to the main, so that one lamp can be used as a spare. " Switch Receptacles " containing a snap switch and a plug socket are provided for attaching portable lamps. lamps, The principal requirements of the standard Navy specifications are : Unit of Candle-Power. — The unit of candle-power shall be the candle as determined by the Bureau of Standards at Washington, D. C. Photometric Measure. — The basis of comparison of all lamps shall be the same spherical candle-power. The normal candle-power referred to in these specifications shall be the mean horizontal candle-power of lamps having a mean spherical candle-power value of 82.5 per cent of the mean horizontal candle-power, which is the standard value for filaments of the oval anchored type. For lamps having filaments giving a different ratio of mean spherical to mean horizontal candle-power, the horizontal candle-power measurement will be corrected by a reduction factor determined by the Bureau of Stand- ards or other authority mutually agreed upon. Vest Quantity. — The test quantity shall consist of 10 per cent or more of any lot or package, and in no case be less than ten lamps. 1172 ELECTRICITY IN THE UNITED STATES NAVY. From each package there will be selected at random, the test quantity for the purpose of determining the mechanical and physical characteristics of the lamps, the individual limits of candle-power and watts per lamp, and finally the life and candle-power maintenance. These lamps will be known as the test lamps. All lamps shall conform to the manufacturers' standard shapes and sizes of bulbs, and to the standard forms of filament, and the standard candle- power and watts per lamp. All bulbs shall be uniform in size and shape, clear, clean, and free from flaws and blemishes. All lamps, unless otherwise specified, shall be fitted with the standard Edison scre.7 base, fitted with glass buttons, forming the insulation between contacts, and rendered impervious to moisture. The shells of the bases shall be of good quality brass, firmly and accurately fitted to the bulb with moisture-proof cement, and in length to conform to the National Electric Code of Fire Underwriters. The lamp filament must be symmetrically disposed in the bulb and shall ixOi, droop excessively during the life of the lamp when the lamp is burned on test in the one horizontal position at a voltage corresponding to an initial specific consumption of 3.76 watts per mean spherical candle and without excessive vibration. All filaments must be uniform and free from all imperfections, spots, and discolorations. Leading in wires must be fused into the glass with the joints between cop- per and platinum wires bedded well within the glass; the wires to be straight, well separated, and securely soldered to the base and cap, without excess of solder and so threads of base are free from solder. All lamps must have first-class vacuum, showing the characteristic glow of good vacuum when tested on an induction coil. A printed label, showing manufacturer's name or trade-mark, voltage, and candle-power, must be placed on each lamp near base. The lamps must be well made and free from all defects and imperfections, so as to satisfactorily meet the conditions of the lighting service. If 10 per cent of the test quantity of lamps selected from any package show any physical defects incompatible with good workmanship, good ser- vice, or v.dth any clause of these specifications, the entire lot from which these lamps were selected may be rejected without further tests when tests are made at the lamp factory. When the tests are made elsewhere, if the first test quantity prove unacceptable, 20 per cent more lamps will be selected from the package or lot of lamps, and should 10 per cent of this second lot of sample lamps be found to have any of the physical defects above mentioned, the entire lot from which these lamps were selected may be rejected without further test. When tested at rated voltage the test lamps shall not exceed the limits ,_ven in schedule. If 10 per cent of test lamps from any package is found to Jail beyond the limits stated, when tests are made at the lamp factory the entire lot from which these lamps were selected may be rejected without further test. When tests are made elsewhere, if the first test quantity prove unacceptable, 20 per cent more lamps will be selected from the package or lot of lamps, and should 10 per cent of these additional lamps be found to fall beyond the limits the entire package may be rejected without further Life tests shall be made as follows: From each accepted package of lamps two sample lamps shall be selected which approximate most closely to the average of the test quantity. One of the two lamps thus selected will be subjected to a life test and designated as the life test lamp, the second or duplicate lamp being reserved to replace this test lamp in case of acci- dental breakage or damage during the life test. The test lamps shall be operated for candle-power performance at constant potential, average variations of voltage not to exceed one-fourth of 1 per cent either side. The voltage for each lamp shall be that corresponding to an initial specific consumption of 3.76 watts per mean spherical candle, or if tested upon a diff- erent basis, the results shall be corrected to a basis of 3.76 watts per mean spherical candle. If desired, the life tests may be made at such other watts per candle as may be mutually agreed upon. ■ ; Readings for candle-power and wattage shall be taken dunng life at the marked voltage of the lamps at approximately fifty hours, and at least £ LIGHTING-SYSTEM. 1173 every one hundred hours afterwards until the candle-power shall have fallen 20 per cent below the initial candle-power, or until the lamp breaks, if within that period. The number of hours the lamp burns until the candle-power has decreased to 80 per cent of its initial value, or until the lamp breaks, is known as the useful or effective life. The average candle-power of lamps during life shall not be less than 91 per cent of their initial candle-power. In computing the results of test of a lot of lamps the average candle-power during life shall be taken as the arithmetical mean of the values for the individual lamps of the lot tested. Lamps selected for the life test, which for any reason do not start on such test, shall be replaced by others. Lamps which are accidentally broken but are burned out on test shall not be counted to diminish the average performance. In case both test and duplicate lamps are broken or damaged before the life test is completed, the average performance of all lamps of the same class previously determined under the same contract shall be assigned to the package represented. On all tests for determining average candle-power and life each package which will be affected by the results of test shall have at least one lamp on such test. Accurate recording voltmeter records will be obtained during the test on lamps to show the average variation on the circuit. When so tested the lamps shall average at least the values for useful life given in the tables on pages 1176 to 1178. (a) Values for Oval Anchored Plain Standard l-ig-liting" Stamps. Lamps of this type of voltages 105 and below, at 110, 120, and above, and also at 220, may have double the limits of variation in the initial limits specified for their respective classes. Lamps and other types of filaments to give equivalent performances. For lamps between 120 and 125 volts, the useful life values shall be 95 per cent of those given in the table, and for lamps between 126 and 130 volts the useful life values shall be 90 per cent of those given in the table. (b) Values for Round Bull*. Tubular, and other Irregular Types of Lamps. The individual limits for irregular types of lamps, such as round bulb and tubular lamps, shall be twice the individual limits given in the body of the preceding schedules for regular lamps of corresponding candle-power. The individual limits for metallized filament and round bulbs primo types of lamps shall be 15 per cent above and 15 per cent below the mean candle- power rating, and 15 per cent above and 15 per cent below the mean total watt rating. The candle-power ratings referred to are the mean horizontal candle-power ratings of clear lamps without reflectors. (c) Navy Special lamps. All lamps must conform in their general shape and form to drawing No. 7219-C, see Figs. 4 and 4a, and overall dimensions must not be exceeded. Rejections and Penalties. The failure of the lamps in any package to conform to the specifications as to mechanical and physical characteristics, or to initial limits, may cause the rejection of the entire package. The failure of the lamps to give within 90 per cent of the values of useful life given in the tables may cause the cancellation of the contract. Lamps which have not been used and are rejected under the terms of these specifications will be returned to the manufacturer at his expense, and no payment will be made tharefor. Prompt notice will be served upon the contractor of the test results on lamps that are rejected, or that fail to meet the specified requirements. 1174 ELECTRICITY IN THE UNITED STATES NAVT. t&LfjB. Diving LAMP Jf AM 3Z <;.* LAfifPf Fig. 4. Standard Incandescent Lamps as Covered by U. S. Navy Specifications. LIGHTING-SYSTEM. 1175 INSTRUMENT f^\ *• *ll!8 11-3 f J -291 •3Z y~co-C *C o^ a r seful or e tive life in h to 20 per drop in car power at watts per ca o ooooooooooooooooo o OOOMM03 0JiCiOW(NiCiCCOMOO * ■ > CO COC0C0COfOCOCO^'^rtiT^cOCO^^'*C0 3 PS "§1 3* > O , J) 03 ""go 8 C3 C £ 2 N o> o 2 o ^ a ".2 •5 £ • C © ! ^rH.^ COOOOOCC: 03 £ © o > w u u w u u u pq o ■3 £ OcoOCCCOOOO^^COOOOOOOO COO ^ V 43 (1) -^ g^OOOOOOOOO _rfM"3 T3*3'Ot3*0"C"0'0 00 NNhcDhcDhcOhiOOiOihcDhOCDCDcD rfi rji CO COCO COCO COCO COCO CO CO CO CO CO CO CO CO co »OCO 00 O © o3 fl ^ 0) c o •- S »- . O v • •II c3 £ ' o3 £ «.2 > © • ©.2 > © « M< K >~0 OOO • 03 £ .©£ > © 1 • 'if ig -St , o +» a) jlS * £ I © _Q © Sh L aofe a oooo . .X! ^HnTJ T3 -Q H3 © u > © o& ,fi Bj d d 0) CJ CJ t- a •S o §3 ^O OOO a©^ 2 - »OT3 r OT3-0 o © o o. i* 8 Et, 2 • ®x: ©-2 si 3 CD aWW -• HfH 8 88 28? 8 §88 ^r cocooo ft S m 9 ft h 73.' 73 © 1° 73 ft 73 V O a; 73 © S.H»0 O -WOO a> o 12 a> a c o &iOT3 73 73 73 73 a . a a s3 > o3 oJ C © C C $ © .a . © fc © © w a; S a> o a) o ftiO &X! OJ2 CM A rt OS- LO *0 O c3 ci >^£ * o ®J2 ° n' © % 08 -g ^ ^ G ©fl C © © . © . « ft « © w © c- a t- ►*• t- >• a> © 2 © 2 a 8 a -s °-s o o 10 CO o 22 on '-•'-• 6^J O C ^"c > > « > 5 5 o © CD CO « © s ^^ >3 o o g «g 1-H1-I MO >*S on r r 3 w • • w a 00^ . »H »HrH lO IO *C 1-1 05 03 c>> o2?5 add .?■ ss o 8 88 LIGHTING-SYSTEM. 1179 I>ivi ner- T.an tern*. Diving-lanterns consist of a glass cylinder closed at each end with a metal cap, having the joint between the glass and metal packed with a soft-rubber gasket. On the inside of one of the caps is provided a standard marine lamp-socket for 150 candle-power incandescent lamp, to which is connected 100 feet of twin conductor cable, at the other end of which is connected a double pole plug for a standard marine receptacle. When first submerged a considerable amount of moisture is deposited in the inside, which is drawn out through a small hole made water-tight by a screw with a rubber gasket. Sea re hi is; lit*. The requirements of the standard Navy specifications are : It shall, in general, consist of a fixed pedestal or base, surmounted by a turntable carrying a drum. The base shall contain the turning mechanism and the electric connections, and be so arranged that it can be bolted securely to a deck or platform. The turntable to be so designed that it can be revolved in a horizontal plane freely and indefinitely in either direction. The drum to be trunnioned on two arms bolted to the turntable, so as to have a free movement in a vertical plane, and to contain the lamp and re- flecting mirror. The drum to be rotated on its trunnions. The axis of the drum to be capable of a movement of not less than 70° above and 30° below the horizontal. The drum to be thoroughly ventilated and well-balanced ; to be fitted with peep sights for observing the arc in two planes, and with hand holes to give access to the lamp. It must be so designed that a parabolic mirror can be used, and means for balancing it must be provided. The mirror to be made of glass of the best quality, free from flaws and holes, and having its surface ground to exact dimensions, perfectly smooth and highly polished. Its back to be silvered in the most durable manner ; the silvering to be unaffected by heat. To be mounted in a separate metal frame lined with a non-conducting material, in such a manner as to allow for expansion due to heat and to prevent injury to it from concussion. The lamp to be of the horizontal carbon type, and designed for both hand and automatic feed. The feeding of the carbons must be effected by a posi- tive mechanical action, and not by spring or gravitation. It must burn quietly and steadily on a 125-volt circuit in series with a regulating rheostat, and shall be capable of burning for about six hours without renewing the carbons. The front of the drum to be provided with a glass door composed of strips of clear plate glass. The door to be so arranged that it can be put in place on the drum easily and quickly. Electrically Controlled .Projector. To be in all respects similar to the hand controlled, with the addition of two shunt motors, each with a train of gears ; one motor for giving the ver- tical and the other the horizontal movement of the projector. The motors and gears to be contained in the fixed base, and to be well protected from moisture and mechanical injury. A means to be provided for quickly throwing out or in the motor gears, so that the projector can be operated either by hand or by motor, as desired. The motors to be operated by means of a compact, light, and water-tight controller, which^can be located in any desired position away from the pro- jector. The design of the controller to be such that the movement of a single handle or lever, in the direction it is wished to cause the beam of light to move, will cause the current to flow through the proper motor in the proper direction to produce such movement. The rapidity of movement of the projector to be governed by the extent of the throw of the handle or lever. A suitable device to be included whereby the movement of the pro- jector can be instantly arrested when so desired. All projectors to be finished in a dead-black color throughout, excepting the working-parts, which shall be bright. 1180 ELECTRICITY IX THE UNITED STATES NAVY, SIGNAL LIGHTS. 1181 The lamps to be designed to produce the best results when taking current as follows : 18-inch, 30 to 35 amperes ; 24-inch, 40 to 50 amperes ; 30-inch, 70 to 80 amperes. The 18-inch projector shall project a beam of light of sufficient density to render plainly discernible, on a clear, dark night, a light-colored object 10 by 20 feet in size, at a distance of not less than 4,000 yards ; the 24-inch pro- jector, at a distance of not less than 5,000 yards ; and the 30-inch projector, at a distance of not less than 6,000 yards. The connections for the electrically controlled projectors as manufactured by the General Electric Company are shown in the diagram, Fig. 5. The fields of the two training motors are in series with each other and connected across the 125-volt circuit. Both horizontal and vertical training can be simultaneously produced. The controller-handle when released, is brought to the otf position by springs and short circuits, both motor armatures thus stopping all movement. The horizontal training motor drives through a worm gear, and the verti- cal motor through a revolving nut on a vertical screw shaft : all gearing can be easily thrown out for quick hand control. The highest speeds are 360° in 30 seconds horizontally, and 100° in 60 seconds vertically. The motors may also be operated at four lower speeds. The lamp has a striking magnet in series with the arc and feeding magnet in shunt with the arc. When the arc becomes too long, sufficient current is forced through the shunt feeding magnet to cause it to make its armature vibrate back and forth, and thus move the carbons together through a ratchet which turns the feed screws. The point at which the magnet will begin to feed is adjustable by means of a spring attached to armature. The feed screws are so proportioned that the positive and negative carbons are each fed together at the same rate that they are con- sumed, thus keeping the arc always in the focus of the mirror. Sight holes are provided through which the arc may be watched. A permanent magnet, fastened to the inside of the projector and surrounding the arc on all sides but the top, causes the arc to burn steadily near the upper edge of the carbons and in focus with the mirror. The rheostat is located near the switchboard, and after being once set for proper working does not need to be again changed. Double-pole circuit breakers are used at the switchboards for switches. §ia\AL IIOHT§. Ardois Sig-nals. The Ardois signals consist of four double lanterns, each containing a red and a white light, which are hung from the top of the mast, one under the other and several feet apart. By means of a special controller any number of lanterns may have either their red or white lamps lighted, thus produc- ing combinations by which any code can be signaled. The lamps used are clear, and the color is produced by having the upper lens which forms the body of the lantern colored red ; the lower lens is clear. The controller consists of eight semi-circular plates, with pieces of hard rubber set in the inner edges where needed, and a rotating center s£ud with eight plunger contacts rubbing on the edges of the plates. By suitably placing the pieces of hard rubber for any given position of the contacts, any desired combination of lights can be produced. The operation consists in moving the arm carrying the contacts to the position desired (as shown by a pointer on an indicating dial) and closing the operating switch, when the proper lamps will light. A later design is provided with a typewriter keyboard, the depression of any key making the proper contacts to light the lamps giving the combina- tion corresponding to the character on the key. Track lights. The truck lights are lanterns of construction similar to the Ardois lanterns," mounted, one on the top of both the fore and main masts. By means of a special controller the red or white light in either lantern can be lighted. 1182 ELECTRICITY IN THE UNITED STATES NAVY. PILOT LAMP Fig. 6. Diagram of Ardois Signal Set. POWER SYSTEM. 1183 POWER SYSTEM, Motors are kept entirely separate from lights by the use of different bus- bars on the generator switchboard and distribution boards. Each motor or group of motors is supplied by its own feeder running from the distribution board, where it has its own fused switch. A maximum drop of 5 per cent is allowed. Principal Requirements of Specifications for Motors. Motors to be wound for 120 volts, direct current, for both armature and field windings, unless otherwise specified, and to be either series, shunt or compound wound, according to work they are to perform. In sizes above 4 horse-power, motors to be multipolar; 4 horse-power or below may be bipolar. Motors to be as compact and light as possible, con- sistent with strength and efficiency. The method of running wires to motors to be in all cases by tapping conduit directly into the motor frames or into connection boxes attached to frames, as may be specified in each individual case; connection boxes for enclosed motors to be water-tight. Enclosed motors should be provided with openings of sufficient size and number to give easy access to brush rigging, commutator, and field coils; such openings to be provided with covers and fastenings of approved design. The contact surfaces between these covers and motor frame should be flat machined surfaces, provided with rubber gaskets. Rubber gaskets for all water-tight work to be in accordance with the Navy standard specifications for the same as issued by the Bureau of Supplies and Accounts. All en- closed motors to be provided with drain plugs or cocks which will thoroughly drain out any water that may enter the motor casing. The armature shaft to be of steel and strong enough to resist appreciable bending under any condition of overload, to have sufficient bearing surface and to be efficiently lubricated by grease or self-oiling bearings, or sight- feed oil cups, as occasion may require. Oil cups to be of size to afford lubrication for at least eight hours. A satisfactory arrangement to be made to prevent oil from running along the shaft or being spilled. Visual oil gauges to be provided for determining the amount of oil in pocket and drains for drawing oil prior to renewal. To prevent deterioration from rust and corrosion, bolts for end brackets, all bolts and pins one-half inch diameter or less not in the magnetic circuit and such nuts and other special fittings as the Bureau may direct, will be of noncorrosive metal, rolled bronze or its equivalent. All electrical connections to be designed with special reference to the pre- vention of their becoming loose from vibration or shock. All connections liable to become loose by vibration are to be provided with approved efficient locking devices. All connecting pieces and other current-carrying parts to be so propor- tioned that no undue heating will occur when they are worked under the severest possible conditions. All the field poles to be equally energized. In compound motors, series and shunt windings to be separate. The windings of armature and field to be well protected from mechanical injury, and to be painted with water- excluding material not soluble in oil or grease. No insulating substances to be used that can be injured by a temperature of 100 degrees C. The armature to be of the ironclad type, built up of thin laminated disks of soft iron or steel of the very best quality, having the spaces between the teeth punched out of each separate disk and not milled after assembly. The disks to be properly insulated from each other. The coils to be pref- erably of the removable type, and to be retained in slots of the armature body by maple wedges running full length of armature, or other approved method. No more than three band wires under poles will be accepted. Band wires must be of nonmagnetic material. The armature to be electri- cally and mechanically balanced. The winding at pulley end to be pro- tected from oil in an approved manner. The commutator segments to be of pure copper, hard-drawn or drop-forged and tempered. The segments to be of ample depth and insulated from each other and the shell by pure mica of such quality as to secure even wear with the copper. 1184 ELECTRICITY IN THE UNITED STATES NAVY. Brushes to be of carbon; current density in brushes must always be given and should be in accordance with the best practice. Special atten- tion must be given to the selection of brushes, that their material may be homogeneous and the quality such as to give perfect commutation without cutting, scratching, or smearing the commutator. Brush holders to be readily accessible for adjustment and renewal of brushes and springs; to be entirely of noncorrosive metal and of the sliding shunt-socket type, in which the brush slides in the holder and is provided with a flexible connection between brush and holder. The springs are to be phosphor-bronze and shall not be depended on to carry current. Brush holders on all motors to be adjustable for tension, and on motors of five-horse-power and above to be adjustable for tension without tools, and so constructed as to permit of proper staggering of brushes. Brush holders for nonreversible motors of five-horse-power and above to be simultaneously adjustable for position, Proper position of rocker arm to be plainly marked. This position for reversible motors to give same speed in either direction. Tests. Contractors are required to afford facilities for inspection of apparatus during manufacture, if required. Individual motors or small lots will be tested at the point of delivery, but all large lots of materials to be shipped to distant points will be tested at the works of the manufacturer. The contractor will provide all facilities, and have all the required tests made in the presence of an authorized inspector. The contractor will present a certified record of such tests with the deliv- ery. The tests to cover the following points: (a) Adjustment and JFit of I*arts. — The inspector to see that the materials and workmanship of all parts of the machine are of the best quality and satisfactory in every respect. (6) Mechanical Strength. — The base, bearings, shaft armature, field magnets, and other main parts should not spring with any reasonable force that may be applied to them. The strength to resist strains due to cen- trifugal force to be tested by running armature without load for 30 minutes at double its rated speed for shunt motors and four times full load speed for series motors. (c) Balance, — The perfection of balance of the armature to be tested by running the motor at its normal speed, at which speed the motor must not show the slightest vibration. (d) l¥oise. — The motor to run at its full-rated speed and load without noise. (e) Sparking*. — Open motors to run without sparking from no load to full load without shifting the brushes and under all conditions of full and weak field when field regulation is used. Enclosed motors to 25 per cent overload. (/) Variation of Speed. — For shunt- wound motors the variation in speed from no load to full load shall not be more than 12 per cent in motors of less than five-horse-power and not more than 9 per cent in motors of five- horse-power and above. Series and compound wound motors to make at rated outputs their rated speeds. The motor should be designed to obtain its rated speed when hot, with atmospheric temperature of approximately 25 degrees C, and the speed actually obtained on test at the end of the heat run must be within 4 per cent of the rated. The variation in speed due to heating shall not exceed 10 per cent- er) [Dielectric Strength. — The test for dielectric strength to be made with a pressure of 1,500 volts alternating E.M.F. for 60 seconds, tested with a generator or transformer of at least 5-kilowatt capacity. The insulation resistance between windings and frame to be at least one megohm measured with 500 volts direct current. (h) Heating*. — The rise of temperature of the field windings above the surrounding air is to be measured by the resistance method according to the rules and coefficients adopted by the American Institute of Electrical Engineers, appended. The rise of temperature of all other parts to be by thermometer. The temperature of the room is to be read from thermo- meters, conditions of ventilation being normal. POWER SYSTEM. 1185 The following are the maximum temperature rises allowed: '■}.) Open-type motors designed for continuous work, eight hours' run with a rise of — Commutator, 40 degrees C. Field winding, 40 degrees C. All other parts, 35 degrees C. (ii) Enclosed motors designed for continuous work, eight hours' run with a rise of — Commutator, 50 degrees C. Field winding, 50 degrees C. All other parts, 45 degrees C. (iii) Intermittent-running motors will have heating limit and length of heat-run separately specified for each case. The temperature rise of bearings shall in no case exceed 35 degrees C. (i) Efficiency. — Motors must have the highest commercial efficiency for their size and speed. Each contractor must state weight and efficiency of motors at one-quarter, one-half, three-quarters, and full load. Prefer- ence will be given to lightest weight and best efficiency consistent with good design and the specific requirements. When thorough reliability and freedom from danger of breakdown are the prime requisites, as for turret- turning motors, boat-crane motors, etc., the maximum efficiency will not be insisted on. (k) JLul»i*icati©n. — The inspector will see that oil cups and wells of the specified capacity are provided and that all the necessary provisions are made for the supply and drainage of oil without injury to the electrical parts. Electric brakes, solenoids, etc., to stand the same heat and insulation test as the apparatus to which they are attached. All spare parts to be subjected to the same tests as originals. Most intermittent running motors, such as boat crane, deck winch, turret turning, etc., have the following heat tests: Each motor shall be tested at the works of the maker by running for a continuous period of one hour at 120 volts at its rated output and speed, without increasing the temperature of the series field windings more than 70 degrees C, the shunt field windings 50 degrees C, the commutator 65 degrees C, the armature or any other part 60 degrees C. above the sur- rounding air. Principal Requirements for Controlling- Panels. Controlling panels for installation in locations not exposed to the action of water outside of ammunition passages, handling rooms, etc., where powder is handled, may be of the nonflame-proof type, in accordance with the following specifications: The panel to consist of a suitable insulating slate base with black polish finish, carrying a double pole main-line knife switch with enclosed indicating fuses, a starting arm with automatic no-voltage release and overload cir- cuit breaker and the necessary resistances mounted at the back. A double pole circuit breaker with independently operating arms may be substi- tuted for the line switch if desired. On panels where speed control by field resistance is required, suitable rheostat connections are to be pro- vided, giving ample number of steps to secure smooth control and accurate adjustment, and* must be a separate multipoint switch so arranged that the motor cannot be started on weak field. On panels where speed con- trol by armature resistance is required, the starting arm must be so con- structed that it will stay only on the contacts designed for continuous running. For motors requiring more than 60 amperes of current, the starting arm must not be relied upon to carry the current in the running position. The starting resistance must not be left in series with the field on the running position ; connections to be such that there shall be no disruptive discharge of the field on opening the circuit, either by opening the main-line switch, or by forcing the starting arm to the off position, and provision to be made to prevent arcing on the initial starting contact. Panel to be so connected that it shall be impossible to have full voltage on the field with the starting 1186 ELECTRICITY IN THE UNITED STATES NAVY. arm in the off position. Care should be taken in the design of the pinel to see that there is no interference between operating parts, such sa line switch, when opened, and starting arm. All magnet coils and all contact parts carrying currents must be renewable from the face of the panot with- out disturbing any of the rear connections. Panel to be mounted on a rigid box metal frame, with the top and bottom of solid sheet metal and the sides (if so desired) of perforated metal, which must extend the length and breadth of the slate and which must protect the connections and parts back of the panel; suitable lugs or extensions to be provided for support- ing the frame. Hinged doors with composition lock and duplicate keys shall be provided over the face of the panel. No part on the face of the panel is to project beyond the edge of the panel. The automatic no-voltage release must operate and either bring the starting arm to the off position or open the circuit breaker upon failure of voltage. The winding of the no-voltage release magnet must not be put in series with either the field winding or armature resistance. The automatic overload release must be of the nature of an ordinary spring operated circuit breaker, having the release mechanism operated by a posi- tive hammer blow, delivered by a core or armature moved against the action of gravity, and must have its own independent contacts for opening the armature circuit; and it should open the circuit in case of overload under any condition, i.e., during ordinary running, during the act of start- ing the motor, or if the starting arm should become struck on any starting point and the current then switched on from the outside. For motors having a rated full load current of 50 amperes or less, the overload release may be of the interlocking type, in which case it must be so interconnected with the starting arm that it cannot be closed with the starting arm in any but the off position. For motors requiring more than 50 amperes, a single or double pole circuit breaker entirely separate from the starting arm must be used. An overload device which operates by short-circuiting or opening the circuit of the retaining magnet of the no- voltage release will under no conditions be accepted. The overload device is to be provided with renewable arcing contacts of carbon, to be adjustable and provided with a scale graduated from normal current to 100 per cent overload to facilitate adjustment to the desired number of amperes, and to be able to carry a current of 50 per cent in excess of the rated full-load motor current continuously without undue heating. The tripping device must be able to withstand severe shock without opening. The insulating material used on the panel must be noncombustible, non- absorbent, and not damageable by moisture or by heating to a tempera- ture of 150 degrees C. The frame of the panel is to be insulated from the hull of the ship. All panels are to pass the same dielectric and insulation tests as the motors for which they are supplied. All windings of magnet coils are to be run through an insulating varnish and the outside or the coils to be well varnished and taped. When continu- ously in circuit, the temperature rise of these coils must not be more than 40 degrees C. above surrounding atmosphere, measured by ther- mometer placed on the coil. The main operating springs for the no-voltage release and the overload circuit breaker must be amply strong to prevent any sticking after the appliance has become worn or roughened. All flat springs are to be of phosphor-bronze and all helical springs of copper-plated steel. All con- tacts to be easily renewable from the face of the panel. The circuit is not to be opened on the rheostat contacts, and special arrangements to be made for opening the circuit and rupturing the arc independent of these contacts. All sliding brushes to be easily renewable and of the self-align- ing, self-adjusting type, and able to ride over any projections standing one-sixteenth of an inch above the contact segments. All operating parts to be strong and very substantial; thin sheet-metal stampings are not to be employed. All such operating parts which carry current to be copper or composition. Where the employment of oxidizable metal is necessary for magnetic purposes their surfaces shall be thoroughly protected against oxidation by copper-plating. Where used for other pur- poses to be very heavily coated with a nonvitreous enamel. The contact points to be of composition or copper, ample in size and well fitted on the Burface and easily renewable. Panels should be as small and light as pos- sible, consistent with other requirements. POWER SYSTEM. 1187 All resistances and all insulation used on them and their connecting wires must be noncorabustible, and the connecting wires must be capable of carrying their full current under all conditions of test and operation without becoming dangerously hot. All resistances to be of the unit type, so con- structed and installed that they may be easily replaced and the whole rheostat readily removed from the casing. The method of mounting and insulating the resistances is to be such that the result of a burn-out would be practically the same as would occur with an entirely enclosed resistance, and no resistance is to be used until a sample has been submitted to the Bureau for test and approval. The capacity of all controlling panel resistances must be obtained without placing the coils in parallel with each other, unless each is capable of carrying full-line voltage. Starting resist- ances when cold must be capable of carrying 50 per cent overload in cur- rent for one minute, and 100 per cent overload for twenty seconds. Incan- descent lamps or carbon shall not be used as resistance. Resistances must be mounted at the back of the panel upon the supporting frame, and not directly on the panel, for motors having a rated full load current over 50 amperes. For motors requiring 50 amperes or less, the resistance may be supported from the back of the panel by suitable brackets, if desired. Water-tight, flame-proof panels will be used as directed in locations greatly exposed to moisture and where powder is handled, as ammunition passages, handling rooms, etc. They will, in general, consist of a cast metal, water-tight, flame-proof case containing the necessary resistances, con- nections, and operating parts, which must be controlled from without by means of rods or levers passing through approved stuffing boxes. The panels must contain within the casing at least the following parts: Resist- ances, circuit breaker or overload release, no-voltage release, reversing switch (when required), starting arm and contacts, and the necessary field contacts when necessary for variable speed motors. They will conform to the requirements for nonflame-proof panels as regards connections, capacity of resistance, construction of overload and no-voltage release, springs, con- tacts, etc., but such deviations from these requirements as may be absolutely necessary to simplify the construction of the panel and reduce its size and weight to a minimum will be considered. The panel will be provided with suitable removable covers provided with clamping devices of approved construction, made water-tight by means of rubber gaskets, which will permit easy access to the interior. It must be strong and substantial in design, but of lightest weight and smallest dimen- sions consistent with other requirements. Suitable bosses for tapping conduit into casing to be supplied, the casing to be drilled and tapped after delivery. The casing is to be sufficiently water-tight to permit of immer- sion without leakage. Noncorrosive metal requirements will be strictly adhered to, and all operating levers passing through stuffing boxes will be of composition. Turret-Turning* Gear. The following are the requirements of turret control: First. Turrets to be able to be turned at a maximum rate of 100 degrees per minute, and at a minimum rate not exceeding one-fourth of a degree per minute, as large a number of speeds as possible (not less than 50) to be provided between the limits of one-fourth and 22 degrees per minute and a sufficient number of speeds between 22 and 100 degrees per minute to per- mit of smooth and easy acceleration. The total number of speeds to be not less than 70. Second. Turret to be capable of acceleration at such rate that it can be started from rest and brought to its full speed of 100 degrees per minute in ten seconds of time, and while turning at its full speed of 100 degrees per minute to be able to be stopped in five seconds of time. Third. At all speeds between and including one-fourth and 100 degrees the turret is to turn continuously throughout the arc of train on each con- troller position with practically no variation in speed due to increased load on the motors caused by allowable irregularities in track, gearing, etc. Fourth. Turret to be able to be started and stopped ten consecutive times without turning through a total arc of train greater than five minutes. 1188 ELECTRICITY IN THB UNITED STATES NAVY. There are four different systems in use at present: 1. "Ward-Leonard System. 2. Rotary Compensator System. 3. Differential Gear System. 4. Mechanical Speed Gear. 1. The "Ward-Leonard System was used on the first electrically operated turrets in the Navy. The actual connections and elementary diagram of the installation on the " Illinois " are shown in Fig. 2. The motors are shunt wound, and have the fields constantly separately excited from the bus-bars of the ship's power system. A separate generator is required which cannot be used for any other purpose when used with the turret. The generator is also separately excited from the power bus-bars, but a variable rheostat, located in the turret, is connected in the shunt- field circuit. The brushes of the motor are directly connected to the brushes of the generator, and the generator is kept running at constant speed by its driving-engine. It is now evident that by varying the rheostat in the turret, the field excitation, and consequently the voltage produced by the generator, will be varied; and any variation in the voltage of the generator will produce a corresponding variation in the speed of the motor, which has a constant field from separate excitation. The direction of rota- tion of the motor is reversed by reversing the leads to the armature. The actual connections for the application of the above principles are shown in the main part of the diagram. Generator No. 4 is shown connected for operating the after-turret. Closing the after-turret field switch and the center blades of the generator field switch separately excites the fields of the motors and generator from the power bus-bars. The regular field rheostat of the generator is entirely 'disconnected, and a rheostat located in the turret and operated by the tur- ret-turning controller is used instead. Closing the positive and negative single-pole switches on the after-current bus-bars connects the generator armature to the motor armatures, through a circuit breaker, the reversing contacts of the controller, and separate armature switches for each of the two motors, which are operated in parallel. The controller has one shaft, at the top of which are located the con- nections for the generator field rheostat, so arranged that as the controller is turned either way from the off position the rheostat is gradually cut out; below are located the reversing contacts, which reverse the connections between the generator armature and the motor armatures; these contacts are so arranged that at the off position the motor armatures are entirely disconnected from the generator, and are short-circuited through a low resistance called the "Brake resistance." The effect of this brake resist- ance is to bring the turret to a quick stop when the controller is brought to the off position, as the motor armatures revolving in a separately excited field generate a large current, which passes through the braking resist- ance, and thus absorbs the kinetic energy of the turret, giving a quick and smooth stop. In parallel with each of the large main fingers of the re- versing contacts is a small auxiliary finger and an auxiliary resistance connected to it. This auxiliary finger makes contact a little before and breaks it a little after the main finger, and thus reduces the sparking. The controller is also provided with a magnetic blow-out for reducing sparking at contacts. When used on this system for operating a turret the generator has its series coil short-circuited by a very low resistance shunt, so that it has very little effect on the field excitation, but this resistance is so proportioned that enough of the total current generated by the generator will pass through the series coil to give a quick and positive start of the turret; because if the series coil is absolutely short-circuited, and only the separately excited shunt coil used, the time required for the field to build up is sufficient to make the starting of the turret very sluggish and irregular, and prevents very fine training from being obtained. It is seen that the above-described arrangement requires a separate gen- erator for each turret, and while operating a turret no power can be taken from the generator for any other purpose. The first ships to use electric turning gear had only two turrets, and two generators can easily be allowed POWER SYSTEM. 1189 for turret turning; but on the latest ships six turrets are used, and it is very undesirable to allow six generators for this purpose. To overcome this objection the Ward-Leonard method of control is obtained by means of a motor generator located at each turret, all of which take power directly from the main bus-bars of the dynamo room, thus materially reducing the required generator capacity. An elementary diagram of the arrange- ment is shown in Fig. 7. It will be noted by comparison with Fig. 2, that only two instead of five wires have to be run from the dynamo room to each turret. The Ward-Leonard system will not give the large range and low speeds Two wires between Dynamo Room and Tuitet Fig. 7. Diagram of Motor Generator on Turret-Turning System. now required by the Navy Department and therefore the other above- mentioned systems have been devised. 2. The Rotary Compensator System is shown in Fig. 8. A and B are the armatures of a motor generator balance set, called a Rotary Compen- sator Set. L is a large shunt motor geared directly to the turret. S is a small shunt motor the shaft of which carries a worm, Wl, working in a worm wheel, W2, mounted on the shaft of L. This worm wheel is pro- vided with a magnetic clutch D so that it can turn freely on the shaft of L, or be held to it. C is a contact in the controller which opens one side of the armature circuit of L. R is a field rheostat for A and B and is operated by the controller. With the connections as drawn in the diagram, B has a weak field .and a low voltage, thus driving S at a low speed ; S is driving L through the magnetic clutch and worm gear and thus turning the turret at a very low speed; Cis open, so L turns freely, and does no work. As the controller is turned R is gradually inserted in the field of B, thus increasing the voltage and increasing the speed of S. When B has full field the mag- netic clutch is opened and C is closed, thus transferring the load from S to L and permitting S to run free. At this time A has weak field and supplies a low voltage to L, and further movement of the controller brings the arm of R back to the first position, thus increasing the voltage of A and the speed of L, until A has full field and the turret is turned at full speed. At the period of transition when the load is shifted from S to L it is necessary that the ratio of the speeds of S and L shall be the same as the ratio of the worm gearing by which S drives L, so that the transfer will be made smoothly and without shock or change in speed of turret. In shutting down the above actions occur in reverse order. Reversing is accomplished by rever- sing the armature loads of the two motors, and in the off position the armature of L is short-circuited to produce a braking effect; these results are accomplished by controller contacts similar to those for Ward-Leonard System as per Fig. 2. This system is made by the General Electric Company. 1190 ELECTRICITY IX THE UXITED STATES NAVY. Glaring Fig. S. Rotary Compensator Turret-Turning System. 3. The Differential Gear System is shown in Fig. 9. L and S are respect- ively large and small shunt' motors running continuously on the supply main. They are both directly geared to a differential gear which is so pro- portioned that with L running at full speed and S at weak field the shaft A will stand still, but any change in their relative speeds will cause A to Fig. 9. Differential Gear Turret-Turning System. rotate at a speed proportioned to the relative change. This change in relative speed is produced by the field rheostats Rl and R2 which are operated by the controller, and first decrease the speed of S by strengthening its field, and then increase the speed of L by weakening its field, thus giving the full speed range of the turret. The shaft A is geared to the turret through the gears Gl and G2, each of which is provided with a magnetic AMMUNITION HOISTS. 1191 •lutch CI and C2. G2 is geared direct, and Gl through a reverse gear, thus accomplishing the reversing of the turret motion. The magnetic clutches are operated by contacts on the controllers. This system is made by The Cutler-Hammer Manufacturing Company. 4. The Mechanical Speed Gear System uses a continuously running, con- stant speed, shunt motor geared to the turret through the speed gear. The speed gear consists of a variable volume oil pump and an oil motor mounted in a common casing and provided with mechanical means for varying the volume of oil delivered by the pump per revolution and its direction of flow. The speed gear is made by the Waterbury Tool Company. In all the above systems two sets of motors are usually provided and arranged so that by means of switches either set may be cut out and the turret operated by one set. Turrets carrying two 12-inch guns usually have two 25-horse-power main motors, and 8-inch turrets two 15-horse- power motors. loading* and Training* Grear for Guns. Guns of 8-inch and over are elevated and rammed by power ; smaller guns have hand gear. Three kinds of elevating gears are in use: 1. Plain rheostat control with series motor. 2. Ward-Leonard control. 3. Mechanical speed changing gear with constant speed, shunt motor. Rheostatic control with series motor as used in the first vessels does not give sufficiently close and even control. A 2^-horse-power, 300 r.p.m. motor with plain drum-reversing controller is used. Ward-Leonard control as used is similar to that used for turret turning as shown in Fig. 7. The control obtained is quite satisfactory, but the complication is objectionable and there is not suitable space available in the turrets for the motor generators. Ten horse-power elevating motors and eight K.W. motor generators are used. The latest vessels are using constant speed shunt motors and obtaining the control by means of mechanical speed gears as described above for turret turning. Rammers consist of a telescopic tube worked through spur and chain- gearing by a 5 H.P., 775 r.p.m. series motor. A friction slip clutch is inserted in the gearing to prevent damage when the shell seats itself in the breech. Ordinary rheostatic control is used. When ramming a shell but little power is required, as the shell slides along the breech, but as it is being forced to its seat at the end of the breech chamber a sudden rush of current of from two to three times the full-load current of the motor is produced. AMIttUXITIOM HOISTS. Power ammunition hoists are of two kinds: first, those in which a car or cage is hoisted up and down by a line wound on a drum on the motor counter-shaft; and second, those in which the motor runs an endless chain provided with toes or buckets on which the ammunition is placed and con- veyed up through a trunk. Hoist* for 12-inch and 13-inch Ammunition. These hoists are of the first kind. The motor frame is provided with bearings for a counter-shaft, geared by a spur-gear and pinion to the arma- ture shaft; on the counter-shaft is mounted a grooved drum for the hoist- ing-cable. On the armature shaft is mounted a solenoid band-brake. The cores of the solenoid are weighted and attached to the brake-setting lever so that when free their weight is sufficient to hold the loaded car from falling; when the solenoids are energized the cores are drawn up and the brake re- leased. The controller is constructed so that on the off position the solenoids are not energized and the brake is set; but at all other points, both hoisting and * o « .298 H — — 45.3 s^ 600 2200 1.350 1.041 .309 .0715 2072 .306 634 .135 61.1 2500 1140 1.352 1.036 .316 .0726 2075 1.25 2595 1.415 .555 39.2 76.2 2500 1140 1.383 1.075 .308 .0734 2037 1.25 2546 1.25 .555 44.3 74.4 4000 875 1.318 1.014 .304 .0721 2052 2.00 4105 2.08 .854 41. 83.6 5000 810 1 .377 1.062 .315 .0722 2084 2.50 5210 2.97 1.131 38.1 82.6 10,000 575 1.268 .971 .297 .0724 2016 5.00 10,080 5.00 2.014 40.3 84.1 12,000 525 1.275 .970 .305 .0725 2050 6.00 12,300 6.28 2.47 39.4 84.8 H'ATElI-TlCiUT DOORS. Electrically-operated water-tight doors are now being installed on most large ships. The requirements of a successful system are that all doors can be simultaneously closed from the bridge, that during or after this closing any door can be opened by a person desiring to pass through from either side, and after such passage the door to automatically close itself. The design in most general use is that of the Long- Arm System Co., of WATER-TIGHT DOORS. 1199 Cleveland, O. The doors are moved by a 1 H.P. compound-wound motor geared to the door plate through spur gears and a worm and rack. Control at the door is obtained by a small hand-operated controller, having an oper- ating handle on each side of the bulkhead. Control from a distance is obtained by an " emergency station" located on the bridge which closes the doors by means of a secondary circuit and solenoids. An indicator is also installed at the emergency station to show when each door closes. The system is shown diagrammatically in Fig. 12. The controller connects the motor directly to the line, without the use of any starting resistance, and the motors are specially designed for this. When the door reaches either the top or bottom of its motion, it actuates the " upper limit switch," or the " lower limit switch," which opens the line circuit and stops the door. These limit switches are actuated through a series of cams, springs, and. levers, attached to the driving gearing, so that a limit switch is opened whenever the door plate encounters any great resistance. In fact the operation of the limit switch, when the door opens or closes, is caused by the resistance to further motion, and not by the position of HAND CONTROLLER -QE£ £7 OPEN INDICATOR CONTACT Fig. 12. Diagram of Connections for Electric Control of Water-tight Sliding Doors. the door plate. This arrangement prevents any obstruction from burning out the motor, and at the same time, if the emergency station action is on, the door will continue its closing motion when the obstruction is re- moved. The emergency station consists of a series of contacts, one for each door, which, when closed, excite solenoids located in the hand controllers. The two contact plates of the hand controller, which are located at the right-hand side of the diagram, are free to rotate on the controller shaft ; and when the solenoid is excited it rotates them so that they make con- tact with their ringers, and thus produce the same result as moving the hand controller to the " close " position. The solenoid is so weak that it can be overpowered by the hand operation of the controller when it is moved to the "open" position, thus allowing a man at the door to make it open at any time when the emergency closing is in action. Upon releas- 1200 ELECTRICITY IN THE UNITED STATES NAVY. ing the handle of the hand controller it comes back by a spring to the 44 off " position, and if the emergency is still on, the door starts to close again. The mechanical construction of the emergency station is such that by moving a lever the contacts for the different doors are made one after the other with a slight interval of time between each, so that the sudden rush of starting current will not occur on all doors at the same instant. The indicator consists of a case containing a small incandescent lamp for each door. When the door closes it operates an indicator contact which lights the corresponding lamp. The door is powerful enough to cut through several inches of coal on the sill. In service the current required for operation of a vertical sliding door 2 ft. by 4 ft. 9 in. is about as follows : Opening : Sudden throw, start 25 amp. Running up 8 to 10 amp. Sudden throw, stop 15 amp. Closing : Sudden throw, start 20 amp. Running down 6 to 9 amp. Sudden throw, stop 11 amp. Voltage is 125 volts. STEERING-GEAR. Electrical steering-gears are not at present used in the United States Navy, but are somewhat used in foreign navies. One method used is shown in the diagram of connections (Fig. 13), in which M is a shunt motor oper- ated from the ship's mains and running continuously at constant speed ; its shaft is directly coupled to G, a shunt generator, the two forming a HIPS MAINS Fig. 13. Diagram of Steering-Gear. motor generator set and located at any desired place, most conveniently in the dynamo room. P is a shunt motor geared by suitable gearing to the rudder post, and has its field constantly excited from the ship's mains, its brushes are directly connected to the brushes of the constantly running generator G. R and R' are two equal and symmetrical rheostats, the con- tact arm of R being attached to the rudder post or any part of its gearing which has a similar rotation, and the contact arm R' being attached by suitable gearing to the steering-wheel. The ends of the field of G are con- nected to these two contact arms, and the two rheostats are connected across the ship's mains. It is now seen that the two rheostats and the field of G form a Wheat- stone's bridge, the parts of the rheostat on each side of the contact arms being the four resistances, the field of G taking the place of the galvanom- eter and the line being the battery. This bridge is in balance, and no STEERING-GEAR. 1201 current will flow through the field of G whenever the two rheostant arms occupy similar positions on their respective rheostats ; but if they do not occupy similar positions, then the bridge will be out of balance and current will flow through the Held of G. The operation is as follows : Starting with everything central as shown in the diagram, if the steering-wheel is turned, moving the arm of B/ a certain distance, the balance will be disturbed and current will flow through the field of G, causing it to generate an E.M.F. and start the motor P, which will continue to run until the arm of R has been moved a distance equal to the original movement made by the arm of R', when the balance will be restored, no current will flow through the field of G, which will then develop no E.M.F. , and the motor P will consequently stop. The gearing; between P and the contact arm of K is so arranged that the movement of the arm will be in the proper direction to restore the balance. The direction of current flow in the field of G, and consequently the polarity of G and direction of rotation of P, will depend upon the direction of movement of the arm of R'« It is thus seen that the arm of R is given an exact copying motion of the arm of R', both for distance moved and direction of rotation. Instead of actually turning the rudder, the motor P can be made, if desired, to only operate the valve of a steam-steering engine ; when this is done the device is called a " Telemotor." Another method (which has only been applied for use as a telemotor) has the first movement of the steering-wheel connect the operating motor directly to the ship's mains, and the motion of the motor causes a step by step mechanism to disconnect it when it has moved the engine valve a distance proportional to the original movement of the steering-wheel. Both connection and disconnection of the operating motor are made by a switch at the steering-wheel, the interrupter of the step-by-step mechanism is at the operating motor and the mechanism itself at the steering-wheel. The mechanical arrangements are quite complicated. Several ships of the Russian Navy have been fitted with direct acting steering-gears by the Electro-Dynamic Company, of Philadelphia, Pa., and work on the above first described bridge principle, with the addition of a small exciter for the generator mounted on the generator shaft, and the field of this exciter is connected with the bridge rheostats, instead of the main generator field itself. The motor of the motor-generator is rated at 70 H.P., the generator at 500 amperes and 100 volts, and the rudder motor at 50 H.P. ; all being easily capable of standing 50% overloads for short periods of time. The motor-generator runs at 650 r.p.m. and weighs 11,000 pounds ; the rudder motor runs at 400 r.p.m. and weighs 5,500 pounds ; the accessory appliances weigh 1,500 pounds, making a total weight of 18,000 pounds. Tests made on the Russian Cruiser " Variag" took 150 H.P. to move the rudder from hard-a-port to hard-a-starboard in 20 seconds, while going at a speed of 23 knots an hour. For ordinary steering at about 19 knots, readings taken every time the rudder was moved gave the following results : — Amperes. Volts. K.W. 250 4 1. 250 10 2.5 150 14 2.1 180 30 5.4 200 40 8. 100 50 5. 100 55 5.5 50 5 .25 50 25 1.25 60 40 2. 100 22 2.2 100 25 2.5 50 15 .75 200 26 5.2 100 18 1.8 100 20 2. 1202 ELECTRICITY IN THE UNITED STATES NAVY. Readings were taken for every movement occurring for a period of % hour, rudder was never moved more than 15 degrees. IXTEItIO.lt COMMUNICATION SYSTEM. The interior communication system of a ship consists of, as the name implies, the appliances for transmitting signals of all kinds from one part of the ship to another. Order and Position Indicators. Many devices have been tried for the electrical transmission of pre- arranged orders, or the position of a moving body, such as a rudder-head; but the most successful and the one generally installed consists at the re- ceiving end of a number of small incandescent lamps, each mounted in a small, separate, light tight cell with a glass front, and the whole enclosed in a suitable case. On the glass front of each light cell is marked an order or number, or whatever particular information the particular device is to in- dicate. This receiver is connected to the transmitter by a cable having a separate wire for each lamp, and one wire for a common return. The trans- mitter consists of a switching device, by means of which any lamp or lamps in the receiver may be lighted, the current being taken from the lighting mains. As many receivers as desired can be operated from one transmitter, the receivers being connected in parallel. Helm Angle Indicator. When the above-described device is used to indicate in different parts of the ship the angle that the helm is turned, the transmitter switch consists of an arm, as shown in diagram (Fig. 14) fastened to the rudder stock, and moving over a series of contact pieces arranged in an arc in the same manner as an ordinary field rheostat. Each of the contact pieces is connected through one wire of an interior communication cable to one side of one of the receiver lamps, which lamp has marked on its front the number of degrees that the given contact is situated from the center line of the ship; the other side of the lamp is connected to the common return wire, which goes to the source of current and then to the contact arm. As the rudder turns, the contact arm makes connection with the different contact pieces, and as it touches each piece the corresponding lamp in the receiver lights up and indicates its position within the limits shown; when it is just mid- way between any two pieces it will touch both and light both corresponding lamps, which doubles the closeness with which the position is indicated. As many receivers can be connected on as desired, all being operated in parallel. Engine Telegrraphs. When used for engine order telegraphs the contact arm is mounted in a metal case and operated by a hand lever of the same construction as the hand lever of an ordinary mechanical ship's engine telegraph, as shown in Fig. 15. The case contains indicator lamps in parallel with the lamps of the receiver at the engine room, so that the operator on the bridge has visual evidence of the order sent. A small magneto is geared to the trans- mitter handle, and rings a bell at the receiver whenever the handle is moved, thus calling attention to the change of order. Battle Order Indicators. The receiving indicators are of the same construction as above described for the Helm indicators, but the transmitter consists of single-pole snap switches, connected up exactly like the lamps of the indicator, so that by turning the proper switches any desired number of lamps can be lighted. INTERIOR COMMUNICATION SYSTEM. 1203 CONTACT ARM FASTENED TO RUDDER POST and of course any desired order can be marked in front of any lamp. Sev- eral indicators, located in different parts of the ship, are usually worked by each transmitter, all being connected in parallel. The case which contains the transmitter switches also contains an indica- tor, thus always showing what orders are being indicated on the system. 1204 ELECTRICITY IN THE UNITED STATES NAVY. Engine Telegraph Indicator Section of Transmitter .Ball on Padestat-j-flQ Transmitter Indicator Fig. 15. Diagram of Engine Telegraph. Rangre Indicators. Range indicators are exactly like the battle order indicators, except that instead of having different orders marked before each lamp, a number rep- resenting the range in yards is marked. A range indicator and a battle order indicator are usually mounted to- gether at desired stations, thus showing what kind of firing is to be done and at what range. Revolution Indicators. To show on the bridge the direction and speed of rotation of the engines several appliances have been devised. The one most generally used is shown in Fig. 16, and consists at the transmitter of a small gear E, mounted eccen- trically upon the propeller shaft S, and meshing with a pinion P, which is carried on the lower end of an arm A. The arm A is slotted and mounted on a pivot as shown, and when S is rotating, A will be turned to one side or the other, depending upon the direction of rotation of S, until it bits on the stop B, and will then remain against the stop and reciprocate up and down from the eccentric action of E ; on each up movement it will make contact with clip C or C, depending upon which side it is turned. INTERIOR COMMUNICATION SYSTEM. 1205 The receiver consists of two pivoted pointers, connected as shown to two electromagnets and marked " Astern" and " Ahead." From the connections shown, it is seen that at each rotation of the pro- peller shaft the pointer corresponding to the direction of rotation will make a movement, and at the same time the magnet armature will make a plainly audible click, thus indicating both visually and audibly the rotation. The 1206 ELECTRICITY IN THE UNITED STATES NAVY. other pointer corresponding to the direction in the opposite rotation will remain still. For twin screws a separate transmitter and receiver is in- stalled for each. A later design of the transmitter is shown in Fig. 17, which eliminates reciprocating motion and prevents wear. E is a large multiple worm mounted on the propeller shaft S. D is a worm wheel on the small shaft F, on which is mounted the insulating drum B. A is a metallic contact strip set into B and makes contact across two of the leads as shown. C is a cam which moves B along its shaft, and holds B in the position shown for astern motion, so that the contact A connects the center and left hand leads at each revolution of F. For ahead rotation C shifts B to the right n F B ,A C - Fig. 17. Diagram of Connections of Transmitter for Revolution Indicator. so that contact is made between the center and right-hand leads. For turbine vessels with fast running propellor shafts the gearing ratio of E and D is proportioned so that only one indicator is given for 3 and 4 revo- lutions of the main shaft. Telephones. On the latest vessels a central station is provided to which each set is directly connected. Fig. 18 is a diagram of the system as furnished by the Western Electric Co. The central board is known as the cordless and plugless type. Its main feature consists in having the connection circuits arranged in a series of horizontal bus-bars which are crossed vertically by the talking wires of each station, so that by putting an ordinary spring lever key at each intersection any desired combination of connections can be made. Usually the board is arranged for fifty stations and five connection circuits, so that five separate conversations between any five pairs of telephones may be carried on at the same time; also for issuing general orders any desired number of telephones may be connected together. The diagram shows only two connection circuits and two stations, but it can be extended as desired in either way. The operation is as follows: Three wires run to each station, two for talking and one for ringing. When the receiver is taken off the hook, current from the talking battery flows through the talking line wires and displays the line signal. When the operator throws one of the connection keys of the calling set the line signal is cut out and the talking wires connected direct to the horizontal connection bus which is permanently connected to the talking current supply. Throwing the connection key of the party to be called, which is on the same horizontal bus as the calling party, puts the pair in communi- cation. Ringing is accomplished by a separate ringing key for each set, taking current from a separate ringing battery and operating through the common ringing wire and the left-hand talking wire as shown. Each horizontal connection bus has a clearing-out signal which is dis- played when current is flowing from the talking supply. When both INTERIOR COMMUNICATION SYSTEM. 120? Qsignai ® Clearing-out Signal 5Li§DE Line 'Signal Connection < m — J fl M rfe Connection /Key Terminals at Sub-<$tetion + I- -I «♦ Talking Ringing Circuit. Circuit 44* 20 y WTSETNS/. HW.TS£tH*2 Fig. 18. Diagram of Western Electric General Telephone System. 1208 ELECTRICITY IN THE UNITED STATES NAVY. parties hang up the receivers the flow of the talking current ceases and the signal falls back. A night bell is provided which is operated by a relay when any line signal is displayed, also when any clearing-out signal falls back and the corre- sponding connection keys are not opened. Cross-talk is prevented by choke coils (not shown on diagram) inserted T£L.S£TJl/9l Tel.SctVZ y. TAIXW6 CUMENT 24 VOLTS Fig. 19. Diagram of Holtzer-Cabot General Telephone System in each side of the horizontal connection busses just after their connection to the talking current supply. Both talking and ringing currents are supplied either from batteries or motor generators taking power from the ship's generating plant, thus giving a reserve. Both water-tight and non-water-tight telephones are used. The non- water-tight are of the ordinary wood case wall pattern, while the water-tight sets have the mechanism enclosed in a brass box with the cover having a rubber gasket and heavy clamps. Figure 19 shows the design made by the Holtzer-Cabot Co. The general scheme of operation is the same as above described, the main difference INTERIOR COMMUNICATION SYSTEM. 1209 being that the operator's set is handled by a separate additional row of keys instead of being treated simply as a station. m Figure 20 shows the design made by Charles Corey & Son. It is gener- ally similar to the above systems, but uses a separate battery for each talking circuit instead of using talking current from a common bus sup- plied by dynamo current. Also each talking circuit consists of two sets of . .£ RUtGW* CURRENT Fig. 20. Diagram of Corey General Telephone System. horizontal busses, and the connection keys may be thrown either way to connect on the station, thus making it possible to reverse the direction of current flow through the contacts and instruments. Each set of connection keys is so grouped that one lever operates them all. In the above telephone diagrams the following notation is used: T. — Transmitter R. — Receiver. L. — Line signal. C. — Clearing out signal. I. — Choke coil. P. — Push button. 1210 ELECTRICITY IN THE UNITED STATES NAVY, IWftftftftftftftftftffftftt t/VUti&UUUUdUU RESTORING SPRING Fig. 21. Diagram of Connections of Electric Whistle. ITire Alarms. The fire alarm system consists of thermostats, located in all parts of the ship, and connected to an annunciator in the captain's office. The thermostats consist of a helioal metal coil, made of two strips of steel and brass, having a high temperature coefficient of expansion, mounted with one end free so that the tortional effect produced by a rise of temperature MISCELLANEOUS. 1211 causes a slight displacement of the free end, thus closing the circuit and operating the corresponding annunciator drop. The working parts are enclosed in a heavy brass case. Coal bunker and storeroom thermostats are set for 200 degrees Fahrenheit, and those in magazines at 100. Water- tig-lit Door Alarms. To give a general signal for the closing of all water-tight doors, a system of alarm whistles is used. The whistle consists of a solenoid which pulls its core down into an air chamber, and thus forces the air out through a small shrill whistle. The core is restored by spiral springs. All whistles are connected in parallel, and are operated by a make and break mechan- ism, which by the pulling of a lever will interrupt the circuit continuously for about 30 seconds, each interruption giving a blast from each whistle. Current from the lightning mains is used. The construction is shown in Fig. 21. The clockworks for operating the contact maker is constructed so that by rotating an operating lever it is wound up, and upon releasing the lever it vibrates the contact while running down, thus giving periodical signals. In the latest design the whistle is inverted and pulled up against gravity, thus dispensing with the restoring springs. Call Bells. An elaborate system of call bells, annunciators, electro-mechanical signal gongs, etc., is installed on all large ships. The main difference from ordi- nary commercial work is that all appliances are made water-tight. \ Rang-e -finder. The following is a brief outline of the principles employed in the instru- ment designed by Lieutenant Bradley Fiske of the United States Navy. In Fig. 22 let A represent the target and BC & known base. Then AC :BC : : sin ABC : sin BAC. sin ABC AC = BC X sin BAC The angle ABC can be readily measured. The angle BAC ■* DBE, the line BE being parallel to AC. The Fiske range-finder measures the angle DBE by the use of the Wheat- stone bridge, as follows: Suppose the two semi-circles in Fig. 22 replaced by two metallic arcs (Fig. 23). At the center of each of these arcs is pivoted a telescope, the pivot of which is connected to a battery B. The telescopes are in electrical contact with the arcs. These metallic arcs are connected at their extremities with a galvanometer, c, the whole forming a Wheatstone bridge, whose arms are aa bb. When the telescopes are pointed at the object A, it is evident that the arms of the bridge are unequal, and hence do not balance; and this fact is indicated by the deflection of the needle of the galvanometer. The arc FD is noted. By swinging the telescope at F around till the needle of the galvanometer indicates zero, the bridge balances, the telescope being parallel to the one at C, and the arc or angle DF — FE is equal to the angle at A. From this the distance AC can be calculated, or read off directly on a properly constructed scale. Generally, in using the instrument, the telescopes are mounted at a distance from the battery, where the view is uninterrupted, while the gal- vanometer is at the gun. The observers keep the telescopes constantly 1212 ELECTRICITY 1ST THE UNITED STATES NAVY. directed on the target, and the man at the gun balances the bridge by in- troducing a variable resistance into the circuit till the needle stands at Fig. 22. Fig. 23. zero. This variable resistance is graduated so as to indicate the range corresponding to the resistance introduced. This instrument is not now used. firing- Guns. Large guns are arranged to use both percussion and electric primers for firing. The electric primer is of the same external shape as the percussion primers, and is exploded by a fine platinum wire, heated by current from the cells of a dry battery mounted near the gun. A ground return is used and a safety switch is fastened to the breech plug, so that the circuit can- not be completed until the breech plug is closed. A push-button is used to complete the circuit and fire the gun. The same primer is also used for igniting the charge of powder to expel torpedoes from their directing tubes. Fig. 24 shows a section of the primer and diagram of connections for both torpedo and gun firing. In torpedo firing the opening of the sluice gate, which permits the torpedo to be dis- charged from the tube, closes the circuit and operates the signal lights at the tube and firing key. This also acts as a safety device by preventing the primer being fired before the gate is opened. Speed Recorder. An instrument called the "Weaver Speed Recorder" is somewhat used for measuring the speed of ships when run on the measured mile, and while being launched; also to measure the acceleration of turrets during test. It consists essentially of a clockworks, which drives a paper tape over a set of five pens operated by electromagnets, so that when any magnet is excited it pulls its pen against the moving paper tape, and makes a dot thereon. The connecting levers between the magnet and pen are arranged something like a piano finger action, so that no matter how long the magnet is kept excited, the pen will only make a quick, short dot. All pens are located side by side in the same line, so that if they were all operated at the same instant, the result would be a line of dots across the tape. When used for measuring mile runs, one pen is connected to a make and break chronometer, so that it makes a dot on the tape every second; an- MISCELLANEOUS. 1213 other pen is connected to a hand push-button, so that a dot can be made at the start and finish of the run, and at as many intermediate points as de- sired; the other three pens are connected to contact makers on the shafts of the main engines, so that a dot is made for every revolution of the en- gine. (If the ship has twin screws, of course only two of the remaining pens are used and if single screw, only one.) It is thus seen that by counting the number of second dots between the Btart and finish dots, the length of time to make the run is given, and by INSULATION SECTION OF PRIMER, SLUICE GATE CIRCUIT CLOSER A INDICATOR LAMP PILOT LAMP AT TUBE ^T FIRING KEY RESISTANCE PRIMER CONNECTIONS FOR TORPEDO TUBE FIRING. j? FIRING KEY **, SAFETY SWITCH "^T BATTERY PRIMER Fig. 24. Connections for Torpedo and Gun Firing. 1214 ELECTRICITY IN THE UNITED STATES NAVY. counting the number of revolution dots in any desired space, the speed of the engine is given. Fractional seconds or revolutions can easily be scaled. When used to obtain launching curves, a long steel wire wound on a drum has one end attached to the ship, and a contact maker is fastened to this drum. As the ship slides out the drum is revolved and dots made on the tape at each revolution; knowing the diameter of the drum, the speed at any instant is found by comparison of the revolution dots with the second dots. The hand-push is used to mark the start, finish, instant of pivoting, and any other desired matters. When used for acceleration runs on turrets, the same procedure as for launching curves is followed, except the contact maker is attached to some rotating part of the turret mechanism. RESONANCE. ^Revised by Lamar Lyndon. If in an alternating current circuit, an inductance be inserted, the self- induced E.M.F. will combine with the impressed E.M.F. and the resultant of the two will be the active E.M.F. which causes current flow. The current will always be exactly in phase with and proportional to the resultant E.M.F. The inductive E.M.F. is 90 degrees, or one-fourth of a cycle, behind the current, and, therefore, behind the resultant E.M.F. which is in phase with the current. The algebraic sum of the instantaneous values of the resultant and inductive E.M.F.'s will give the corresponding values of the impressed E.M.F. Fig. 1 shows this summation. v,v,v,v, is the resultant EJM.F. re- quired to send current i,i,i,i, which is in phase therewith, through a given resistance. L,L,L,L, is the curve of E.M.F. necessary to overcome the counter E.M.F. of the inductance, the curve of the inductance E.M.F. being equal and opposite to the curve L,L,L,L. This curve of inductance E.M.F., which is indicated by the dotted curve 1,1,1,1, is one-quarter period or 90 degrees behind the current and the resultant E.M.F. Combining the ordi- nates of v,v,v,v, and L,L,L,L, the curve e,e,e,e is produced. This represents in phase and magnitude the impressed E.M.F. required to send current i,i,ii, through the resistance and overcome the counter E.M.F. of the inductance* As may be seen, it is somewhat in advance of the resultant E.M.F. and, therefore, of the current. Also it is higher than the resultant E.M.F. by an amount which at each instant is equal to the counter E.M.E. of the in- ductance. If a condenser or capacity be included in a circuit, and an alternating current be sent into it, flow will take place in the condenser, the current entering and charging it. As the amount of electricity stored increases, the E.M.F. of the condenser increases also until the impressed and con- denser E.M.F.'s are equal. The condenser E.M.F. being a counter pressure, current flow ceases when the two E.M.F.'s balance. The current being aero at this point, and the condenser E.M.F. a maximum, it may be seen that the condenser E.M.F. is one-quarter period or 90 degrees in advance of the current, and, therefore, of the resultant E.M.F. In Fig. 2, F,7,F,7, is the resultant E.M.F. made up of the two E.M.F.'s acting on the circuit. i,i,hi, is the current, C,C,C,C, the condenser E .M.F., 1215 1216 RESONANCE. which is 90 degrees ahead of i,i,i,i. c,c,c,c, is the curve of F.M.F. necessary to overcome the condenser E.M.F., being equal and opposite to the condenser E.M.F. Combining V, V, F, F, and c,c,c,c, the impressed E.M.F. curve e,e,e,e, is produced, which is somewhat behind the current and resultant E.M.F., and behind the condenser E.M.F. Also, the impressed E.M.Fo is greater than the resultant E.M.F. From the foregoing it is evident that if either a capacity or inductance be inserted in an alternating current circuit, the phase of the current with respect to the impressed E.M.F. will change, and the current flow be re- duced. Since the one sets up an E.M.F. 90 degrees in advance of the cur- rent flow and the other a pressure 90 degrees behind it, the two effects tend to neutralize each other when connected in series, and when they are just equal, no E.M.F. other than the impressed is left to act on the circuit, the resultant and impressed E.M.F .'s are identical, and there is no phase dis- placement. This condition is called resonance and is shown in Fig. 3. Fig. 3. The curves L,L,L,L, and c,c,c,c, are equal and opposite at every instant and neutralize, leaving the impressed E.M.F. as the only one acting on the circuit. The conditions for resonance then are, that with a given frequency and current the capacity and inductance be so related that the counter E.M.F. 's set up by them are equal, or it may be stated another way. If in an alter- nating current circuit an inductance and a capacity be connected in series, either of which, if inserted in the circuit alono, reduces the current flow the same amount, resonance occurs and the current flow is not changed by the presence of the two in series. RESONANCE. 1217 The formula for alternating current flow in a circuit containing resistance, inductance and capacity is E (i) y/^ + (z w _J_y in whioh E = E.M.F.. (impressed volts), i= Current in amperes, jR — Resistance in ohms, L = Inductance in henrys, J= Capacity in farads, T = 0, (2) OiJ UiJ and formula (1) becomes /=-^L = -, (3) which is simply Ohm's law, showing that the current flow is opposed only by the resistance. The farad is too large a unit for practical work, capacities seldom being more than a few micro-farads (or one millionth of a farad). If / be taken in micro-farads and called Jm, then for resonance T _ 1,000,000 .-,/! LJm also w = 2 tt/. Therefore, /= 1 /iWOOO, 2 7T j LJm 1,000,000 (4) (5) which is the frequency at which resonance will occur for a capacity Jm and an inductance L. Since the opposing E.M.F. of the inductance in- creases with increase of frequency, and that of the condenser decreases, with a given inductance and capacity there is only one frequency at which they will neutralize and resonance result, and if this frequency be changed, the E.M.F. of one will increase while that of the other will decrease, thus destroying the balance between the two. As an example, assume a circuit having an inductance of 0.44 henry, and a capacity of 16 micro-farads. For resonance the frequency must be . 1 /l, 000,000 _ A . , /= 2^ V 0.44 x 16 = °y cles ' P er second - The opposing inductance and capacity E.M.F/s often set up local poten- tials very greatly in excess of the impressed. Since the voltage required at the terminals of an inductance to force a W given current through it =r B t t± wLI, and for resonance, 1=-^ the voltage at the inductance = tf,= ^. <«> 1218 RESONANCE. Also the voltage required to send a given em-rent through a condenser — -, or I X 1,000,000 _ E X 1,000,000 u>J m ~~ RaJm Assume the circuit of 0.44 henry 16 micro-farads and 5 ohms. / = 60 cycles, Impressed E. M. F. = 250 volts, the voltage at the terminals of the inductance, 250 X 0.44 X 2tt X 60 (7) * = - 5 while the volts at the condenser terminals 250 x 1,000,000 = 8290 volts, -Ec — 5x2ttx60x16" : 8290 volts, which is the same as the voltage at the terminals of the inductance. Fig. 4 shows the diagram of such a circuit and indicates the potentials between the different terminals. 0.44 HENRV From the foregoing it is obvious that the smaller the resistance, the greater will be the local voltages set up by the capacity and inductance. For instance, if in the previous example the resistance were 2\ ohms instead of 5 ohms, the current flow would be 100 amperes and the poten- tial at the terminals of the inductance and of the condenser would be 16,580 volts, the impressed E.M.F. being only 250 volts as before. In practice the capacities and inductances are seldom so related as to allow complete resonance to occur at commercial frequencies, though when- ever a capacity and inductance are in series the partial neutralization which takes place is liable to increase the E.M.F. locally to a higher value than that of the impressed. All the foregoing is based on an impressed E.M.F., which is a pure sine function. In practice, however, the E.M.F. wave differs more or less from this form, and may be considered as the resultant of several pure sine waves of varying amplitudes and frequencies. Those waves which have a higher frequency than the impressed E.M.F. wave, are termed higher or upper harmonics. Although the frequency of the impressed E.M.F. may not be sufficiently high to produce resonance, some one of the component waves or "upper harmonics" may have a frequency at which resonance will re- sult. From equations (6) and (7) it is clear that, with a given resistance in circuit, the rise in E.M.F. due to resonance is proportional to the im- pressed E.M.F., and since the voltage of the upper harmonics is usually small, the rise in E.M.F. cannot be great. When resonance occurs with one of the upper harmonics, the wave form of the current becomes greatly distorted, because while the other compo- nent waves must force the current against both the resistance and the reactance (i.e., inductance and capacity E.M.F.'s), this particular wave RESONANCE. 1219 has only to overoome the ohmio resistance and, therefore, sends a greater current through the circuit in proportion to its voltage than do the other E.M.F. waves. All these considerations apply only to circuits in which the inductance, resistance and capacity are in series. If the inductance and capacity be connected in parallel, as shown in Fig. 5, there can be no rise of voltage above the impressed even if the two be in resonance, but currents greater than those supplied by the source of impressed E.M.F. may surge back and forth through the local circuit, joining the condenser and the induct- ance, and, unless the resistance be high, the current sent through the main flip s 6 Fig. 5. circuit will be greatly reduced: indeed, if the resistance were zero, the alter- nator could not send any current whatever through the circuit, for at every value of the impressed E.M.F. there would be an equal and opposite E.M.F. either from the condenser or inductance, and the resultant or active E.M.F. becomes zero. This condition is represented in Fig. 6 in which the curve e represents the impressed E.M.F. c is the curve of condenser current, and L of current in the inductance. The condenser current is 90° in advance of the impressed E.M.F. while the inductance current is 90° behind it, Fig. 6. there being no resistance in the circuit. The sum of the two currents then is always equal to zero, as may be seen. The physical conception of this condition is that of current flowing into the condenser, charging it, while the previous stored energy in the induct- ance discharges. This discharge sets up an E.M.F. opposing the impressed E.M.F., and also furnishes the current supply to charge the condenser. On reversal of the impressed E.M.F. the condenser discharges into the inductance, at the same time setting up a counter E.M.F. to oppose the flow of current from the line. 1220 RESONANCE. Thus, although there may be heavy currents flowing in the branch cir- cuits, none will flow through the main circuit. In practice there is always a certain amount of resistance present in both of the branches, which will displace the phase relations of the two currents so that some current will flow in the main circuit, but this will often be less in amount than if one only of the two reactances were present when the resistances are very small. As an actual case, consider the branch circuit shown in Fig. 5. Branch A has an inductance of .02 henry and 5 ohms resistance. Branch B has a capacity of 50 micro-farads, and a resistance of 8 ohms. Frequency = 100 cycles per second, and impressed E.M.F. z=z 100 volts. Impedance of branch A = V(5) 2 + (6.28 X 100 x .02)2 — 13.5, Current through branch A = — — = 7.42 amperes. 13.5 Tan. angle of lag = ^ X 100 X .02 = 2.512, corresponding to an angle of 68° — 18'. Impedance of branch B= J®F.+ ( 6 J'^ X 5Q ) = 32.83. 100 Current through branch B = — — - — 3.05 amperes. oZ.oo 1,000,000 Tan. angle of lead = 628 ^ ^ = 3.98, o corresponding to an angle of 75°— 54'. Combining these two currents in their proper phase relation, the sum is the current through the main circuit. This can best be done graphi- cally after the usual manner of combining E.M.F. 's or currents vectorially. In Fig. 7 let the horizontal line OE represent the im- pressed E.M.F. and be the reference line. From O at an angle of 68° — 18" upwards lay off 7.42 amperes to any suitable scale. At an angle of 75° — 54" downward, lay off 3.05 amperes. Complete the parallelogram, as indicated by the dotted lines. The diagonal from Ogives the value of the resultant current through the main circuit as 5.24 amperes, and shows also that it is behind the impressed E.M.F. by 48° - 36". This, it will be seen, is less current than would flow through the circuit by branch A if the parallel branch B were entirely removed. If the reactance E.M.F.'s have the same value, the capacity being .00005 farad (=50 micro-farads), the inductance will be equal to .0506 henry. Assume that the resistance in branch B remains as before. If the resistance in branch A be 25 ohms the impedance will be = ^(25)2 +(31.83) 2 = 40.47 and current = ~^ =2.48 31 83 amperes. Tan. of the angle of lag = — ~- = 1.271, corresponding to 51° — 49". Combining these values with the 3.05 amperes at an angle of lead of 75° — 54" in Fig. 8, the result- RESONANCE. 1221 ant current is 2.49 amperes, and has an angle of lead = current is less than that in branch B alone. For the currents in two parallel branches to balance each other, so that the resultant current through the main circuit is brought in phase with the impressed E.M.F., the following condition must exist. The amperes flow through one branch, multiplied by the sine of the angle of lead or lag of the current (referred to the impressed E.M.F.), must be equal to the amperes through the other branch multiplied by the sin of its angle of lag or lead. That is: I± x sin =zl 2 x sin »//, in which I t and I 2 are the currents through the two branches, is angle of lag of I x and »// is angle of lead oil.. If in branch B of two parallel circuits the 4° - 24". This and Impressed E.M.F. = E, Capacity =. <7, Resistance = E, the impedance z= 1/ B?-\- i — = j , w being 6.28 X frequency. The current =: - vM^r Fig. 8. Tan of the angle of lead = -=-, from which the angle and its sine are found. In branch A, either the resist- ance or reactance must be known. Calling I 2 the current and \p the angle of lead in branch B, l x the current in branch A, and its angle of lag, I 2 sin \fj : Tan : : I± sin <£, sin tf> \/l — sin2 ^ where i?i is the known or assumed resistance in branch A. RJ X sin2 — i2 w 2 (i _ s i n 2 ), whence • ^ 1 / L2 « Z2u>2 E I x sin = Calling I 2 sin \jj = 0, and solving, Bui = ELm Rj + L2oj2 = I 2 sin v//. 20 : ■V: Ei ■ JV- (8) (9) (10) (11) ~P"i When BJ is equal to or greater than — -£ the quantity under the radical becomes zero or negative, and there is no reactance which will compensate for the effect of that in the other branch, the resistance being too high. 1222 RESONANCE. The sign before the radical being either plus or minus, there are two values of reactance with a given resistance which will compensate (if R x be not too great) . The lesser reactance will, of course, permit the greater current flow, both through branch A and the main circuit. As an example, assume a resistance of 8 ohms, a condenser capacity of 50 micro-farads in branch B ; also a frequency of 100 cycles per second, im- pressed E.M.F. = 100 volts, and a resistance of 10 ohms in branch A. What inductance must be inserted in branch A to compensate for the reactance in branch B ? Amperes through branch B =z I 2 =z 3.05. Angle of lead = 75° 54" = f. Sini// = .9"~ I 2 sin xjy = 3.05 X .96987 = 2.9581 = 9. Substituting in formula (11) iu = 10 ° +J < 100 > 2 (10P 2 X 2.9581 I V4X (2.9581)2 v ' •La = 16.902 ± 13.614=; { ^ J^? • Tan* = f!. 3 288 Taking the first value, Tan = -^— = . 3288, corresponding to an angle of 18° — 12", sin = .31233. 100 Current through A = = 9.47 amperes. VtlO) 2 + (3.288)2 J sin £ = 9.47 X .3123 = 2.9567, which (within the limits of tabulated values of functions of angles) checks with the value of / sin \\i. Fig. 9. The resultant current in the main circuit is found graphically — shown by full lines in Fig. 9 — to be 9.75 amperes, and in phase with the impressed E.M.F. If the greater value of Lia be taken, Tan ^ — SJ^l— 3 0516 — 71 o _ 50 ", sin (/> — 0.95015. 100 Current through branch A = , = 3.12 amperes. V(10) 2 + (30.516)2 /, sin = 3.12 X 0.95015 = 2.964, which checks with I 2 sin vj/ (within limits of tables of functions of angles). RESONANCE. 1223 Resultant current is found graphically, as shown by dotted lines in Fig. 9, to be 1.72 amperes, and is in phase with the impressed E.M.F. From the foregoing equations it can be seen that if Lot be known and R x is the quantity to be determined, b 1 = sJl^I^-l^Y (12) If R. and L of branch A, and R^ of branch B are known, the capacity re- quired in branch B is found from the formula, 1,000,000 Jm = ! , (13) in which B st I x sin <£. If R x , L, and Jm be known, D /i,ooo,ooo 77? 1,000,000 \ If .7 be taken in farads, formula 14 becomes, THE ELECTRIC AUTOMOBILE. Revised by Alexander Churchward. The Electric Automobile has proved itself successful for delivery service in cities and locations where the roads are good and the distance traveled per day is from fifteen to fifty miles, the distance being decreased in pro- portion to the, loads carried. See Motor World, 1909. Where the ''distance traveled per day under ordinary road conditions is less than ten miles and the speed low, the service can be performed at a lower cost with horse drawn vehicles. Where the distance traveled per day is greater than fifty miles for the lighter vehicles and twenty-five for the heaviest type, the gasolene electric gives better results than those whose source of energy is a storage battery. The above statements should be taken as applying to general conditions. Where the conditions are in any way special or severe, cost of operation by each of the three systems should be carefully computed. Owing to the cost of a horse and wagon being less than a motor driven vehicle, a certain amount of work must be performed each day before the efficiency of the automobile becomes apparent. The actual cost of gasolene is generally found to be greater per vehicle mile than the cost of charging storage batteries of automobiles for equal loads over the equal distances within the limits above given. Certain limits to daily travel will therefore be found when each type of trans- portation is cheapest. (For a more detailed discussion, see Motor World on "Improvement of the Electric Vehicle," May 14, 1908. "Commercial Vehicle Problems," — Motor World, Oct. 1, 1908. "The Horsepower of the Horse," — Motor World, 1908. Resistance Hue to Gravity, and Power Required. W. Worby Beaumont. The horse-power required to overcome weight, speed, road resistance, gravity resistance, and efficiency of transmission between armature shaft and road wheel, may be found as follows: Let R = the resistance to traction of the vehicle on the road in pounds per ton. G = the resistance due to gravity in pounds per ton. W = total weight on the wheels in tons. V = speed in feet per minute. v = speed in miles per hour. E = mechanical efficiency of transmission from armature shaft to road. P = brake horse-power. e = efficiency of motor. p = watts supplied to motor. p {R + G)WV m _ PE 375 } P= 33,000 E ' (1) V -(R + G)W' (5) (R + G)Wv PE 375 P - 375 E ' (2) W {R + G)v* W CB + ^-^trt' (3) v= 746 7* (7) (R + G)vW E ~ P375 ' (4) For a more detailed discussion of the mechanics of traction see Electric Traction. 1224 TIRES. 1225 Resistance to Traction on Common Roads. W. Worby Beaumont. Road Surface Material. Asphalt Wood, hard " soft Macadam, very hard and smooth .... good traffic rolled, wet steam rolled, new and muddy . new, flat spread Gravel Granite tramway "... Iron plate tramway Resistance in Lbs. per Ton. On On Iron-tired Solid Rubber Wheels. Tires. 22 to 28 35 to 40 22 ' 4 26 40 44 45 30 * * 38 40 * 45 35 44 40 45 4 52 52 4 58 58 ' 4 62 95 ' 4 105 100 * 4 140 12.5 4 4 15 10 * 4 12 In most cases these resistances increase slowly at higher speeds, and it must also be noted that the resistance on bad, soft, and gravel roads will probably be greater with propelling wheels than with most hauled wheels. Most of the figures relate to road resistance at walking or slow trotting pace. Tires. Solid rubber tires have a higher resistance than steel tires on asphalt roads and have less resistance on macadam and other roads. The per- fectly smooth surface of the asphalt produces a drag on the rubber tires, thus increasing their resistance. Pneumatic tires are best adapted to roads with slight inequalities, and for pleasure cars run at high speeds. For both solid and pneumatic tires, the draw-bar pull required to over- come the rolling resistance depends on the speed. This subject has been investigated by Mr. Alex Churchward, and the results of his tests* are reprinted below: Material of Road. Asphalt Macadam . . . . Macadam . . . . Belgium block . . Asphalt Macadam . . . . Asphalt and brick Asphalt Grade. Level 1.1% Level 9.5% 4.7% 3.75% 3.125% 2.25% Draw-Bar Miles Type of Pull in Lbs. per Hour. Tire. 24 12 Solid 37 12 Pneumatic 48 11 Solid 66 10.6 Pneumatic 29 12 Solid 44 12 Pneumatic 250 5 Solid 270 5 Pneumatic 132 7 Solid 150 7 Pneumatic 114 8 Solid 128 8 Pneumatic 95 8.5 Solid 119 8.5 Pneumatic 85 8.8 Solid 103 8.8 Pneumatic * See The Commercial Vehicle, April, 1906. 1226 THE ELECTRIC AUTOMOBILE. The above figures are averages of reading^ taken for a great many vehicles. The difference in the consumption of power when running on wet and dry pavements was discovered to be so small that the additional tractive effort required when the pavements are wet may be neglected. The temperature of the asphalt greatly affects the consumption of energy. In one case a difference of 40 per cent was found in the power required for operating a car on cold and on warm asphalt. Tractive efforts of 119 pounds per ton for two inches of sand and 138 pounds per ton for muddy roads were obtained. I Vo | — ■ — I II 1 ' ' 1 : : z / — r 70 - Zt -j. - z _/ Z/ Jl Z- -t __ -/ -/ z z / 4 -/ -T -/ z Z JL A J -f z -f / Z jl A -X S 7 -/ -/ : z _/ 4ST J ZZZZZZZ 50 I£? y I J& JL Z -J& z Z&C J&z : J^Jk. f£-2- z i : t "\ Z 25 V Z ^c- -/ s y \? 20 ' ' 1 1 1 1, 1 1 , 1 1 11 1 1 1 — LI — L 10 12 11 16 18 20 22 24 26 28 30 32,34 36 38 AQ 42 44 Speed in Miles per. Hour Fig. 1. Results of tests for the tractive effort at several speeds are shown by the curves in Fig. 1. It will be noticed that the draw-bar pull diminishes as the speed is reduced, to a minimum, and increases as the speed is still further reduced. The speed in miles per hour at which the minimum point occurs varies with different weights of vehicles, diameter of wheels and types of tires used. BATTERIES. 1227 Motors. The present general practice is to install one series-wound motor on all except the very largest trucks. However, under certain conditions of road bed and the type of tires used, it may be advantageous to use even a four- motor four-wheel drive. Usually, however, under fair road conditions, one large motor has proved more efficient than a number of smaller ones. A normal voltage has been adopted at 80-85 volts to correspond with the minimum discharge voltage of batteries adapted to 110-115 volt charging circuit. Some of the motors are designed for operation at increased speeds by shunting the fields with a resistance, especially on the higher speed pleas- ure vehicles. This practice is considered preferable to commutating the batteries. Controllers. In the past few years, the number of speed points has been almost doubled, combining this with the latest type of control by the continuous torque system; the handling of a vehicle is now smooth and more efficient. There is no perceptible jar or shock when going from one speed point to another and the result is that the maintenance of the entire vehicle has been con- siderably reduced. Batteries. The standard equipment for the wagons and trucks is 44 cells of the lead type of storage battery or 60 cells of the Edison type and of a suitable ampere capacity. These numbers permit of charging from the lighting companies' feeders at 110-115 volts with a minimum loss in the charging rheostat. Runabouts and other very small vehicles are equipped with 24 or 30 cells of moderate ampere capacity, as a saving in weight is thereby obtained over 44 cells of smaller ampere capacity that more than offsets the loss in the charging resistance. A battery can be supplied to meet almost any requirement of travel in miles per day, but it is generally found that the weight of battery required for distances above 50 miles per day for light commercial vehicles and 25 miles per day for the heaviest so reduces the efficiency of the automobile as a whole that the gain over other methods of transportation is not so marked as it is with the battery of standard size. The following lists of batteries may be used as a guide in selecting those for any equipment: The Electric Storage Battery Company, Type MV "Exide." Type PV "Exide." Number of plates . . . 7 9 11 13 15 49 17 56 19 63 21 70 5 12 7 18 9 24 11 Discharge in amperes for 4 hours 21 28 35 42 30 Size of plates: Width Height If 5| 8| 51 8| 51 81 51 8f 5i 8f 5i 8t 51 8f 411 81 411 81 411 81 m 8f Outside measurements of rubber jars, in inches: Length Width Height 21 12| 3* 12| 41 6i 12f 5 61 12f 5f 6i 12f el 12| 71 6i 12| 8 6£ 12f 2 5& m 2f fft 3* 5& 11* 41 fft Allow f inch above the top of jars for straps. Weight in pounds: Element . . Electrolyte Complete cell 18* 2* 22 23* 31 281 281 31 35i 34 4* 41 39 5 461 44 53* 491 61 601 541 7 10 1 141 14 2 191 18 3* 22 5 241 291 1228 THE ELECTRIC AUTOMOBILE. (Would Storag-e Battery Company. Type EP. Type TP. Type NP. Plates Plates Plates 51 X 8i 5f X 81 4f X 81. Number of plates . . 11 13 15 17 19 7 9 11 13 5 7 9 11 Discharge in amperes at four-hour rate . 42 49* 57 64* 72 21 28 35 42 12 18 24 30 Capacity at four-hour rate of discharge 168 198 228 258 288 84 112 140 168 48 72 96 120 Outside dimensions of rubber jar in inches: Length .... 5 51 6* 71 8 21 3* 41 5 2 21 3* 41 Width 6* 6* 6* 6* 6* 6* 6* 6i 6i 5& 5A 5^ 5 T 6 * Height .... 12 12 12 12 12 12 12 12 12 12 12 12 12 Weight of cell complete: Pounds .... 45 53 61 69 77 24* 31* 38* 45* 14i 191 241 291 To height of jar add * inch for straps, and 1 inch for bottom of tray. Rules for the Proper Care of Batteries. A battery must always be charged with direct current and in the right direction. Be careful to charge at the proper rates and to give the right amount of charge; do not undercharge or overcharge to an excessive degree. Do not bring a naked flame near the battery while charging or immediately afterwards. Do not overdischarge. Do not allow the battery to stand completely discharged. Voltage readings should be taken only when the battery is charging or discharging; if taken when the battery is standing idle they are of little or no value. Do not allow the battery temperature to exceed 100° F. Keep the electrolyte at the proper height above the top of the plates and at the proper specific gravity. Use only pure water to replace evaporation. Never add acid except under conditions as explained in the instructions. Keep the cells free from dirt and all foreign substances, both solid and liquid. Keep the battery and all connections clean; keep all bolted connections tight. . If there is lack of capacity in a battery, due to low cells, do not delay in locating and bringing them back to condition. Do not allow sediment to accumulate to the level of the plates. ELECTROCHEMISTRY. - ELECTRO- METALLURGY. Revised by Professors F. B. Crocker and M. Arendt, of Columbia University. jsiJECTitociiEnisriti:. Electrolysis : The separation of a chemical compound into its constit* uents by mean3 of an electric current. Faraday gave the nomenclature relating to electrolysis. He called the compound to be decomposed the Electrolyte, and the process Electrolysis. The plates or poles of the battery he called Electrodes. The plate where the greatest potential exists he called the Anode, and the other pole the Cathode. The products of decomposition he called Ions. Lord Rayleigh found that a current of one ampere will deposit 0.017253 grain, or 0.001118 gramme, of silver per second on one of the plates of a sil- ver voltameter, the liquid employed being a solution of silver nitrate con- taining from 15 per cent to 20 per cent of the salt. The weight of hydrogen similarly set free by a current of one ampere is .00001044 gramme per second. Knowing the amount of hydrogen thus set free, and the chemical equiva- lents of the constituents of other substances, we can calculate what weight of their elements will be set free or deposited in a given time by a given current. Thus the current that liberates 1 gramme of hydrogen will liberate 7.94 grammes of oxygen, or 107.11 grammes of silver, these numbers being the chemical equivalents for oxygen and silver respectively; the chemical equivalent being the atomic weight divided by the effective valency. To find the weight of metal deposited by a given current in a given time, find the weight of hydrogen liberated by the given current in the given time, and multiply by the chemical equivalent of the metal. Thus: Weight of silver deposited in 10 seconds by a current of 10 amperes = weight of hydrogen liberated per second X number seconds X current strength x 107.11 = .00001044 x 10 X 10 X 107.11 = .1118 gramme. Weight of copper deposited in 1 hour by a current of 10 amperes = .00001044 x 3600 X 10 X 31.55 = 11.86 grammes. Since 1 ampere per second liberates .00001044 gramme of hydrogen, strength of current in amperes _ weight in grammes of H. liberated per second "" .00001044 weight of element liberated per second "" .00001044 x chemical equivalent of element Resistances of Dilute Sulphuric Acid. (Jamin and Bouty.) < Ohms per c.c. at Ohms per Cu. In. at Density. u o • u u © . s-. r<& rAO Oo Do H& rA° Qo Do ^O O^* o °° o °3 ^o ©-* o °°. O ©CO 00 -tfi iH CD 1.1 1.37 1.04 .845 .737 .540 .409 .333 .290 1.2 1.33 .926 .666 .486 .524 .364 .262 .191 1.25 1.31 .896 .624 .434 .516 .353 .246 .171 1.3 1.36 .940 .662 .472 .535 .370 .260 .186 1.4 1.69 1.30 1.05 .896 .666 .512 .413 .353 1.5 2.74 2.13 1.72 1.52 1.16 .838 .677 .598 1.6 4.82 3.62 2.75 2.21 1.90 1.43 1.08 .870 1.7 9.41 6.25 4.23 3.07 3.71 2.46 1.67 1.21 1229 1230 ELECTROCHEMISTRY. £ p o CO © u 13 ap SCO 2^2 © « 2 fa © © s £ £ o ^ i © P ! ££ CR ©' a p O P^ is So w pj TO . p a © ©» © o 5T ^2 £ CD CO CN 38£ COMnnO O^H3fri t* ® lO -*• t* t^C»M^'-ioOL'5©r-ii-N05T|; 00 "«t | ^ lO CN rn 35 t^ 00 NddHodo'6o'6«do'rIdd(Nddd6iocoo'666-id6o 5«DC5 C5iOOC^«OT*<(Mt>-iC050«C>OOTf(NTtOOt>ir50CO ^OOt>'lOCOCOCi'3©i'-HCOUOC ©P p g^ •3 8 C P 1 ro to P"£ O p p p ,0 ^ APPLICATIONS OF ELECTROCHEMISTRY. 1231 Resistances of Sulphate of Copper at 10° C. or &0° F. (Ewing and MacGregor.) Density. Ohms per Density. Ohms per c.c. Cu. In. c.c. Cu. In. 1.0167 1.0216 1.0318 1.0622 1.0858 1.1174 164.4 134.8 98.7 59.0 47.3 38.1 64.8 53.1 38.8 23.2 18.6 15.0 1.1386 1.1432 1.1679 1.1829 1.2051 ) Saturated j 35.0 34.1 31.7 30.6 29.3 13.8 13.4 12.5 12.0 11.5 Resistances of Sulphate of Zinc at lO C. or 50° F. Ohms per Ohms per Density. Density. c.c. Cu. In. c.c. Cu. In. 1.0140 182.9 72.0 1.2709 28.5 11.2 1.0187 140.5 55.3 1.2891 28.3 11.1 1.0278 111.1 43.7 1.2895 28.5 11.2 1.0540 63.8 25.1 1.2987 28.7 11.3 1.0760 50.8 20.0 1.3288 29.2 11.5 1.1019 42.1 16.6 1.3530 31.0 12.2 1.1582 33.7 13.3 1.4053 32.1 12.6 1.1845 32.1 12.6 1.4174 33.4 13.2 1.2186 30.3 11.9 1.4220 ) Saturated j 33.7 13.3 1.2562 29.2 11.5 Specific resistance of fused sodium chloride (common salt) at various temperatures. Temperature Cent. 720° 740° 750° 770° 780° Ohms per cu. cm. .348 .310 .294 .265 .247 Applications of Electrochemistry. The word electrochemistry is here used to include electrometallurgy, as there is no generic term for the two subjects. Electrochemistry may be defined as that branch of science relating to the electrical production of chemical substances and chemical action or to the generation of electrical energy by chemical action. On the other hand electrometallurgy is the branch of science that relates to the electrical production and treatment of metals. The two subjects are based upon the same principles, the theory, laws and data of one being applicable to the other. Hence it is proper and now customary to combine them under the head of electrochemistry. Electrochemistry may be subdivided as follows: A. Electrolytic Chemistry, which consists in separating or produc- ing other action upon chemical substance by the decomposing effect of an electric current or vice versa. Since the electrolyte is usually in the liquid state, there are: "Wet methods" with solution. "Dry methods v with fused materials. In the latter case the materials are maintained in a state of fusion by the heat due to the electrolytic current or by external heat. 1 232 ELECTROCHEMISTRY. Electrolytic chemistry is applied to the following purposes: 1. Primary batteries, including various forms of voltaic cell in which electrical energy is generated by chemical action. 2. Secondary or storage batteries are similar to the foregoing, but the chemical action must be reversible, so that after periods of working the cell may be charged or brought back to an active condition by sending through it a current opposite in direction to that which it generates. 3. Electrotyping is the art of reproducing the form of type and other objects by electrodepositing metal on the object itself or on a mold ob- tained from it. 4. Electroplating is the art of coating articles with an adherent layer of metal by electrodeposition, as in nickel plating. 5. Electrolytic refining of metals and chemicals by the elimination of im- purities, as in the conversion of crude copper into pure metal. 6. Electrolytic production of metals and chemicals, as in the Hall process for extracting aluminum from alumina dissolved in fused cryolite, and in the Castner process for making caustic soda and chlorine from a solution of common salt. 7. Electrolytic chemical effects, such as bleaching, tanning, etc. 8. Electrolytic chemical analysis, as in copper determination. B. Electro thermal Chemistry includes those methods in which electric current raises the temperature of materials, usually to a high degree, in order to produce fusion, chemical action or other effects. Since elec- trolysis is not desired an alternating current is generally employed. 9. _ Chemical^ action with electrical heating, as in the production of calcium carbide from lime and carbon in an electric furnace. 10. Electrical smelting consists in reducing metallic compounds at a high temperature produced by an electric current, as in the reduction of iron ore in an electric furnace, or in the Cowles, process for making aluminum bronze from a mixture of alumina, carbon and granulated copper. 11. Electric fusion of chemicals, usually those that are very refractory, such as silica and alumina. It has been proposed to make bricks by melt- ing instead of baking clay; electric heat has been used in furnaces for melting glass. 12. Electrical heating and working of metals consists in treating metals mechanically with the aid of heat generated by electric currents, as in electrical welding, forging, rolling, casting, tempering, etc. Strictly speaking, the last two applications are not chemical, but some chemical actions usually occur and they are similar to the others in methods and results, so that it is customary to consider them under the head of electrochemistry. C. Chemical Action Due to Electrical IMscharg-es. 13. Chemical effects of electrical arcs to produce combinations of nitrogen and oxygen, for example. 14. Chemical effects of electric sparks. 15. Chemical effects of silent electrical discharge, as in the production of ozone. Historical Notes. — The first electrochemical apparatus was the primary battery invented by Volta in 1799. The next year Nicholson and Carlisle discovered the chemical action of the electric current in decomposing water. In 1807 Sir Humphrey Davy gave his famous lecture "On Some Chemical Agencies of Electricity," he having, the same year, discovered the metals sodium and potassium by reducing their compounds electrolytically. In 1834 Faraday established definite laws and nomenclature for electrochem- istry. From 1836 to 1839 Jacobi, Spencer, Jordan and Elkington applied these principles to practical use in the making of electrotypes. Plante began the development of the storage battery in 1859. Since that time, but mostly after 1886, the theory and applications of electrochemistry have made great progress, so that now it is one of the most important branches of science as well as of industry. Primary and Secondary Batteries. — The various forms of these batteries may be regarded as applications of electrochemistry, but they are treated as special subjects in other parts of this book. Electrotyping'.- — To reproduce an engraving, typographical composi- tion, or other object, a mold of gutta percha, wax, plaster or fusible alloy is made from the object. If it is not a conductor it is coated with graphite to start the action, connection being made to it by a wire or clamp put APPLICATIONS OF ELECTROCHEMISTRY. 1233 around it. It is used as the cathode in a bath consisting of a 20 per cent solution of copper sulphate acidulated with 2-8 per cent sulphuric acid, while the best results are obtained with a current density of .2-.25 amperes per square inch of cathode surface. The anode is a plate of copper. The ordinary thickness of deposit is .01 to .03 inch. The "shell" thus formed is separated from the mold and backed by a filling of type metal. Electroplating 1 an article with an adherent coating of metal requires the article to be thoroughly cleaned mechanically and chemically. Cleaning:. — Solutions for cleaning Gold, Silver, Copper, Brass and Zinc are prepared as follows: Water. Nitric Acid. Sulphu- ric. Hydro- chloric. For copper and brass . . Silver 100 100 100 100 100 50 10 2 3 100 10 8 12 2 Zinc Iron, wrought Iron, cast 2 3 Lead, Tin, Pewter, are cleaned in a solution of caustic soda. Objects to be plated with gold or silver must be carefully and thoroughly freed from acids before transfer to the solutions. Objects cleaned in soda or those cleaned in acid for transfer to acid coppering solutions may be rinsed in clean water, after which they should be transferred immediately to the depositing solution. Baths for plating-. —The reader is referred to the various books on electroplating for particulars, as but few, and those the most used solutions can be referred to here. Solutions should be adapted to the particular object to be plated, and must have little if any action upon it. Cyanide of gold and silver act chemi- cally upon copper to a slight extent and the objects should be connected to the electrical circuit before being immersed. Solutions are best made chemically, but can be made by passing a current through a plate of the required metal into the solvent. Copper. —A good solution for plating objects with copper is made by dissolving in a gallon of water 10 ounces potassium cyanide, 5 ounces copper carbonate, and 2 ounces potassium carbonate. The rate of deposit should be varied to suit the nature and form of the surface of the object, large smooth surfaces taking the greatest rate of deposit. Electrotype plates must be worked at a slow rate, owing to the rough and irregular surface. Non-metallic Surfaces may be plated by first providing a conducting sur- face of the best black lead or finely ground gas coke. Care is required in starting objects of this sort, to obtain an even distribution of the metal, and hollow places may be temporarily connected by the use of fine copper wire. Copper on iron or on any metal that is attacked by copper sulphate is effected by an alkaline solution. One which can be worked cold is made up of I ounce of copper sulphate to a pint of water. Dissolve the copper sulphate in a half pint of water, add ammonia until all the first formed precipitate re-dissolves, forming a deep blue solution, then add cyanide of potassium until the blue color disappears. A heavy current is required with this solution, enough to give off gas from the surface. This solution will deposit at a high rate but ordinarily leaves a rough and crystalline surface, and will not do good work on steel. A cyanide solution is the most used, takes well on steel.or brass, as well as on iron, and permits of many variations. For each gallon of water use : Copper carbonate 5 ozs. Potassium carbonate 2 ozs. Potassium cyanide, chem. pure 10 ozs. Dissolve about nine-tenths of the potassium cyanide in a portion of the water then add nearly all the copper carbonate, which has also been dis- solved in a part of the water: dissolve the carbonate of potash in water and add slowly to the above solution stirring slowly until thoroughly mixed. Test the solution with a small object, adding copper or cyanide until the deposit is uniform and strong. For coppering before nickel plating, the 1234 ELECTROCHEMISTRY. coating of copper must be made thick enough to stand hard buffing, and for this reason the coppering solution must be rich in cyanide and have just enough copper to give a free deposit. Use electrolytically deposited copper for anodes, as it gives off copper more freely. Regulate the current for the work in the tanks, and it should be rather weak for working this solution. Brass Solutions of any color may be made by adding carbonate of zinc in various quantities to the copper solution. The zinc should be dissolved in water with two parts, by weight, of potassium cyanide, and the mixture should then be added to the copper bath. A piece of work in the tank at the time will indicate the change in color of the deposit. Two parts copper to one zinc gives a yellow brass color. For the color of light brass add a little carbonate of ammonia to the brass solution. To darken the color add copper carbonate. Varying the amount of current will also change the color, a strong current depositing a greater amount of zinc, thus pro- ducing a lighter color. Silver. — The standard solution for silver plating is chloride of silver dissolved in potassium cyanide. This solution consists of 3 ounces silver chloride with 9 to 12 ounces of 98 percent potassium cyanide per gallon of water. Rub the silver chloride to a thin paste with water, dissolve 9 ounces potassium cyanide in a gallon of water and add the paste, stirring until dissolved. Add more cyanide until the solution works freely. The bath should be cleaned by filtering. Great care should be taken to keep the proper proportions between current, silver and cyanide. A weak cur- rent requires more free cyanide than a strong one, and too much cyanide prevents the work plating readily, and gives it a yellowish or brownish color. If there is not enough cyanide in the solution the resistance to the current is increased and the plating becomes irregular. The most suitable current for silver plating seems to De about one ampere for each sixty (60) inches of surface coated. Gold. — Cyanide of gold and potassium cyanide make the best solution for plating with gold. The solution is prepared in the same manner as the silver solution just described, using chloride of gold in place of chloride of silver. The electrical resistance of the bath is controlled by the quantity of cyanide, the more cyanide the less the resistance, but an excess of cyanide produces a pale color. Hot baths for hot gilding require from 11 to 20 grains of gold per quart of solution and a considerable excess of cyanide. Baths for cold gilding and for plating should have not less than 60 grains per quart and may have as much as 320 grains, this quantity being used with a dynamo current for quick dipping. Wickel. — The solution now almost universally used for nickel plating is made up from the double sulphate of nickel and ammonia, with the addition of a little boracic acid. The double salt is dissolved by boiling, using 12 to 14 ounces of the salts to a gallon of water; the bath is then diluted with water until a hydrometer shows a density of 6.5° to 7° Baume. Cast anodes are to be preferred as they give up the metal to the solution more freely. Anodes should be long enough to reach to the bottom of the work and should have a surface greater than that of the objects being plated. Current strength should be moderate, for if excessive the work is apt to be rough, soft or crystalline. Voltage may vary from 3.5 to 6 volts and the most suitable current is from .4 to .8 ampere per 15 square inches surface of the object. Zinc is the only metal requiring more current than this, and takes about double the amount named. A nickel bath should be slightly acid in order that the work may have a suitable color. An excess of alkali darkens the work, while too much acid causes " peeling." Iron. — A hard white film of iron can be deposited from the double chloride of iron and ammonia which can be prepared by the current process. It is somewhat used for coating copper plates to make them wear a long time, the covering being renewed occasionally. The Electromotive JForces suited to the different metals are: Copper in sulphate Volt, 1.5-2.5 11 cyanide , 4.-6. Silver in " 1.-2. Gold in " .5-3. Nickel in sulphate 2.5-5.5 THE ELECTROLYTIC REFINING OF COPPER. 1235 The Resistance will depend on the nature of the surface. Work is best effected with about equal surface of anode and objects, and the coating will be more even, the greater the distance between them, especially where there are projecting points or rough surfaces. Copper and silver should never show any sign of hydrogen being given off at the objects; gold may show a few bubbles if deep color is wanted. Nickel is always accompanied with evolution of hydrogen, but the bath should not be allowed to froth. The Rate of Deposit is proportional to current, as described under the head of " Electrolysis," in the proportions given in the table of electro- chemical equivalents except in the case of gold, the equivalent of which in combination with cyanogen is 195.7, but subject to modifications dependent upon the hydrogen action just described; there is also a partial solution of the metal, so that there is always a deduction to be made from the theoret- ical value. Thus : — Gold gives about 80 to 90 per cent. Nickel " 80 to 95 " Silver " 90 to 95 " Copper " 98 " An ampere of current maintained for one hour, which serves as a unit of quantity called the "ampere-hour," represents Gramme 0376 Ounce Troy . . . .00121 Grain . . . Ounce Avoir. .58 .00132 which multiplied by the chemical equivalent will furnish the weight of any substance deposited. The Electrolytic Refining* of Copper. The largest and most important of electrochemical industries is copper refining, conducted at many places in this country and abroad. The pro- cess of refining copper electrolytically consists in the transfer of copper from the anode to the cathode, by the selective action of the electric cur- rent, and in leaving the impurities behind in the anode, electrolyte or sediment. _ Theoretically the mere transference of copper should require no expendi- ture of energy, the energy needed to precipitate it from its solution being balanced by the energy set free upon its change to copper sulphate, but practically some is needed on account of the resistance of the electrolyte, and differences in mechanical structure as well as in chemical purity of the anode and cathode. The material at present subjected to profitable electrolytic refining is crude copper containing from 96 to 98 per cent pure copper and varying amounts of other elements according to the character of the ore and method of dry refining adopted. The composition of the crude material varies greatly, typical samples being given in the following table: No. I. Per Cent. No. II. Per Cent. No. III.* Per Cent. Copper .... Arsenic .... Antimony . . . Lead Tin 96.35 0.08 0.10 1.19 0.22 0.05 0.61 97.19 2.68 0.01 98.60 0.80 0.10 0.10 Bismuth . . . Iron Nickel .... 0.08 0.02 0.02 0.05 0.10 0.10 0.10 05 Sulphur . . . Silver .... 0.69 Oxygen and loss 0.71 100.00 100.00 100.00 * Chili bar. 1236 ELECTROCHEMISTRY. Besides these, the crude copper frequently contains small quantities of gold (about one-tenth to one-fifth ounce per ton). The crude material is cast in iron molds into anode plates, about three feet long, two feet wide, and one inch thick, weighing approximately 250 pounds. The cathode plates are of electrolytically refined copper practi- cally the same in length and width as the anodes but only one-twentietb inch thick. The electrolyte or bath in which the plates are suspended is a solution of 12 to 20 per cent copper sulphate, and 4 to 10 per cent sulphuric acid, the latter being added to decrease the resistance of the solution. This resistance is further reduced by keeping the electrolyte warm at about 40° C. •The containing tanks are of wood, usually lined with sheet lead or carefully coated with a pitch compound, and of such dimensions that a distance of about one inch exists between the faces of the plates. In some cases the plates are arranged in series, and in others in parallel or multiple, as illustrated. The former has the advantage of requiring electrical connections to be made at the first and last plates only, whereas Fig. 1. Series Arrangement of Plates. the parallel system requires a connection at every plate; but in the series system the leakage of current due to the short-circulating action of the sediment and sides of the tank is from 10 to 20 per cent, so that the parallel is more generally used. The connections between the various plates and the circuit in the parallel systems are made by copper rods, which are run at two different levels along the edges of the tanks, one bar for anodes and one for cathodes. In some instances these rods are of the inverted V shape, so that the edges will Fig. 2. Parallel Arrangement of Plates. cut through any corrosion that may happen to form at the points of con- tact. The drop in pressure at these points is not more than .01 volt. The vats are arranged so that each is accessible from all sides, and the circula- tion of the electrolyte is possible. This circulation may be obtained by blowing a stream of air through the electrolyte, but more frequently by arranging the vats in steps and connecting them by pipes so that the elec- trolyte may pass from the top of one vat to the bottom of the next, as shown in Figs. 3 and 4. This maintains a uniform density of the electrolyte which is necessary for the proper formation of the deposit. The electrical pressure required is from .2 to .4 volt per tank, with a current density of 10 to 15 amperes per square foot of cathode plate sur- face. The question of current density is very important, because upon this depends the rapidity and quality of the deposit. The rate of deposit, however, is limited and varies with different grades of the crude metal, depending upon the impurities present. For example, antimony, bismuth THE ELECTROLYTIC REFINING OF COPPER. 1237 and arsenic if present would prevent the use of a current density of more than 10 amperes per square foot, as they would be carried over and depos- ited, especially if present in a soluble form. The maximum current density employed in ordinary copper refineries is as above stated, 10 to 15 amperes per square foot. If the current density is too great the following difficulties will occur: a. Liberation of hydrogen at the cathode, and thus a resultant waste of energy. b. Poor character of deposit. If the current density is too low, the copper is in the tanks too long, and this results in excessive interest charges. The individual vats are connected in series with each other, so that the total voltage required may be approximately equal to that of the gener- Fig. 3. Circulating System. ator, allowing the usual drop of about 10 per cent. Standard generators are built to give 125 volts so that a working pressure of about 110 volts is obtained, which is a standard value for lighting and other purposes. In practice from 400 to 450 ampere-hours are required per pound of copper deposited, the theoretical amount according to Faraday's law being + I i- LEj Fig. 4. General Arrangement of Plant. only 386.2 ampere-hours. The loss varies from 4 to 20 per cent, according to the system employed. Anode Impurities and their Effect npon the Electrolyte. — The electrolyte when first added consists of 12 to 20 per cent copper sulphate and 4 to 10 per cent sulphuric acid. The impurities likely to exist in the crude metal anodes have been given in the sample analyses preceding, and the following reactions generally occur in an acidulated solution: 1. Silver and gold remain undissolved in the anode or fall to the bottom of the vat. 2. Lead is converted to lead sulphate and precipitates. 3. Antimony, bismuth and tin are partly dissolved as sulphates, or form unstable sulphates which precipitate as basic sulphates or oxides; they partly also remain in the anode sludge. 4. Arsenic, nickel, cobalt and iron dissolve, but are not under ordinary conditions redeposited, hence they merely contaminate the electrolyte. 1238 ELECTROCHEMISTRY. 5. Alkaline earth metals except barium and calcium dissolve readily, the latter precipitating as sulphates. In addition to contaminating the electrolyte and thus interfering with the purity of the deposit the presence of these impurities, except gold, silver and lead, is objectionable, due to the fact that the anode is consumed unevenly. The more electropositive metals such as tin, zinc, etc., being more rapidly attacked, the anode surface does not remain smooth, and frequently pieces break off and fall to the bottom of the tank. Arsenic, if present, often forms arsenates on the anodes, which results in a non-con- ducting film, decreasing the current and thus the output. Effect of tlie Electrolyte Impurities on tlie l>eposit. — The electrolyte does not accumulate all the impurities of the anode because many of them never go into solution but simply fall to the bottom of the vat as mud. In addition to the proper constituents of the electrolyte there may be present in the dissolved state^ the sulphates of iron, zinc, cadmium, alu- minum, sodium, etc., besides basic sulphates of arsenic, bismuth and anti- mony. The largest part of the impurities present consists of iron, but the most objectionable are compounds of arsenic and antimony, as these yield their metals at the cathode, with serious results, since as little as .01 per cent of either will reduce the electrical conductivity of copper from 4 to 5 per cent. Cuprous oxide and copper sulphide remain partly in the sludge and partly dissolve according to the acidity of the electrolyte. Their only evil effect is to neutralize some of the free sulphuric acid. The composition of the anode sludge (residue) will evidently vary ac- cording to the composition of the anode employed, and in practice various amounts of gold, silver and lead are obtained therefrom by subsequent treatment. The cost of refining copper by the electrolytic method is from | to f cent per pound. The following products of refining are marketed: com- mercial cathodes, which are sometimes shipped to consumers but more frequently cast into wire bars, ingots, cakes or slabs of standard dimensions and weight. They usually assay from 99.86 to 99.94 per cent of pure cop- per, a sample analysis being as follows: PER CENT. Copper* 99.938 Antimony .002 Iron 004 Oxygen and loss .056 100.00 The yield in commercial cathodes is from 97 to 99 per cent of the anodes treated, excluding the anode scrap which varies from 7 to 15 per cent of the original anode in parallel operated plants; but this scrap is not a loss, as it is collected and recast into anode plates. Besides electrolytic copper, most plants recover gold, silver and nickel from the slime as previously stated. The electrolytic copper refineries in the world are now producing copper at the rate of 322,295 tons annually, valued at $96,688,500 with copper selling at $300 per ton. In addition the by-product in recovered gold and silver is valued at $20,000,000 per annum. There are now in active opera- tion 33 electrolytic copper refineries, with a total generator capacity of 20,000 kilowatts; 10 of these are located in the United States, and supply about 86 per cent of the world's output; 6 plants are in England and Wales producing about 9 per cent, while the remaining plants are on the conti- nent of Europe. Silver is refined from copper bullion by taking anodes of the bullion £ inch thick and 14 inches square, and cathodes of sheet silver slightly oiled. The electrolyte consists of water with 1 per cenrt of nitric acid. When the current is started the copper and silver form nitrates of copper and silver and free nitric acid from which the silver is deposited, leaving the copper in solution. Trays are placed under the cathode for catching the deposited silver, and if there is any copper deposited owing to the solution contain- *This sample was obtained by refining the crude copper given in column III of the preceding table of crude copper anodes. PRODUCTION OF CAUSTIC SODA. 1239 ing too little silver or a superabundance of copper, the copper falls into the trays and is redissolved. In the Moebius process the deposit is continually removed from the cathode by means of a mechanical arrangement of brushes, and falls into the trays above mentioned. A In 111 in u in. — Practically the output of this metal for the entire world is now produced electrolytically. The only process used on a large scale is that invented independently in 1886 by Mr. Charles M. Hall in the United States, and by Paul L. V. Heroult in France. This process consists in electrolyzing alumina dissolved in a fused bath of cryolite. The alumina is obtained from the mineral bauxite which occurs abundantly in Georgia, Alabama and other regions. The natural material, being a hydrated alumina containing silica, iron oxide and titanic oxide in the following proportions: Al 2 O a .56 Fe 2 3 .03 SiOo .12 TiO" .03 H 2 .26 must be treated in order to drive off the water and eliminate the impurities. This may be accomplished by a chemical process, but it is effected more simply by heating the material mixed with a little carbon as a reducing agent in an electric furnace. The impurities are thus reduced and collect as a metallic regulus in the bottom of the mass. This leaves the alumina nearly pure and it may be tapped off while fused or easily separated by breaking it up after cooling. In practice it requires two pounds of alumina for each pound of aluminum produced. The flux or bath in which the alumina is dissolved consists of cryolite, a natural double fluoride of alu- minum and sodium (Al 2 F 6 .6NaF) found in Greenland. This is melted in a large carbon-lined, rectangular, sheet-iron tank, which constitutes the negative electrode, a group of 40 carbon cylinders, each 3 inches diameter and 18 inches long, which are suspended in the tank, forming the positive electrode. A direct current of about 65 horse-power at 5 to 6 volts is used. Only a portion of this voltage is required to decompose the alumina, the balance, amounting to about two or three volts, represents the heat pro- duced which keeps the bath at the proper temperature and fluidity neces- sary for electrolysis — 850 to 900° C. The passage of the current causes the aluminum to deposit on the bottom of the tank as a fused metal, being drawn off periodically. The oxygen set free combines with the carbon of the positive electrodes and passes off as carbonic oxide. The reaction is Al 2 03 -I- 3C = 2A1 -f 3CO. About one pound of carbon is consumed for one pound of aluminum produced. When the alumina becomes exhausted from the bath, the voltage rises and lights a lamp shunted across the electrodes, thus giving notice that more material is needed. Each elec- trical horse-power produces about one pound of aluminum per day of 24 hours. According to Faraday's law the weight of aluminum deposited by 1,000 amperes is .743 pound per hour. The actual yield of metal by the Hall process is about 85 per cent of this theoretical amount. The aluminum obtained averages 0.1 per cent iron, 0.3 per cent silicon, with traces of copper, titanium and carbon, but is guaranteed over 99 per cent pure. The metal when drawn from the tanks is cast into rough ingots which are afterwards remelted and converted into commercial shapes such as sheets, rods, wires, etc. jp»o»T)Ctj:o:w or caustic soda. Caustic soda or sodium hydrate (NaOH) is used in the manufacture of hard soaps, in the rendering of wood pulp for paper manufacture, in the purification of petroleum and petroleum residues, and also for the produc- tion of metallic sodium. Many attempts, extending over nearly a century, have been made to manufacture caustic soda (NaOH) and chlorine (Cl 2 ) from ordinary salt (NaCl), by means of electrolytic action. The fundamental reaction: 2NaCl + 2H 2 + Elect. = 2NaOH+H 2 + CI, 1240 ELECTROCHEMISTRY. is readily obtained experimentally, but is difficult to accomplish on a com- mercial basis. Salt, or sodium chloride, when electrolyzed in the presence of water will form caustic soda, but secondary reactions take place and the result is a mixture of salt, caustic and hypochlorite of soda. This diffi- culty can be avoided by separating the caustic soda solution that is formed by a porous diaphragm, or by drawing it off as soon as formed; and in some cases the metallic sodium is absorbed in mercury or molten lead. The following conditions have been found necessary for the success of this process: 1. Cost of power must be very low — not in excess of $30 per horse- power per annum (24 hours per day). 2. The process must be continuous. 3. The electrodes must be as nearly indestructible as possible. 4. The products of electrolysis must be capable of removal from the vessel or electrolyte as the process proceeds. 5. The maintenance costs must be small. 6. The plant must operate on a large scale. It is only lately that a few processes have been commercially successful. The two most prominent systems for the electrolytic production of caustic soda and chlorine from common salt are the Castner-Kellner and the Acker processes, one operating at moderate temperatures (40° C.) and the other at high temperatures (850° C). The Castner process employed in this country at Niagara Falls is as follows: The electrolytic tank consists of a slate box 4 feet long, 4 feet Fig. 5. Castner Cell. wide and 6 inches deep, the joints being made by means of a rubber cement. Two slate partitions reaching within T x ff inch of the bottom (under which are grooves) divide the cell into three compartments, each 15 inches by 4 feet, sealed from each other by a layer of mercury covering the bottom of the tank to a considerable depth. The two end compartments through which the brine is passed are provided with carbon anodes, shaped like a rail section, the broader flange being placed about a half inch above the mercury. These compartments are provided with tight covers and ex- haust pipes of rubber and lead to convey the chlorine away. The central compartment has an iron cathode composed of twenty upright strips and is supplied with pure water, which is drawn off whenever its specific gravity increases to 1.27, due to the presence of the maufactured caustic, while the liberated hydrogen is led from this chamber by means of pipes and used as a fuel for the concentration of the caustic. The tank is pivoted at one end on* a knife-blade and rests at the other on an eccentric, which raises and lowers that end of the tank about a half an inch once a minute and causes a circulation of the mercury between the outer and middle compart- ments. The current enters the outer chambers, splits up the sodium chloride (common salt, NaCl) into sodium and chlorine (Na and CI), the latter is liberated at the carbon anodes and passes through the exhaust pipe to the absorption chambers, where it combines with slacked lime to PRODUCTION OF CAUSTIC SODA. 1241 form bleaching powder (CaCl20 2 .CaCl2). The sodium combines with the mercury, forming an amalgam containing about 2 per cent of sodium, whirh by the tilting of the tank passes to the central chamber, where it serves as the anode, and combines with the water to form caustic soda (NaOH) and hydrogen (H), the latter appearing at the iron cathode. Each of these tanks uses 630 amperes at 4.3 volts; 10 per cent of this current is shunted around the inner cell, because otherwise the amalgam would fail to deliver enough sodium, and the mercury would oxidize, thus producing mercury salts and contaminating the caustic. The theoretical voltage required is but 2.3, the remainder being utilized in overcoming the ohmic resistance of the electrolyte and in keeping it warm, the limit of temperature being 40° C, as above this point chlorate is formed. The output of this process per horse-power per day is 12 pounds of caustic and 80 pounds of bleaching powder for each cell. The product contains from 97 to 99 per cent caustic, \ per cent sodium carbonate, .3 to .8 per cent of sodium chloride and traces of sodium sulphate and silicate. The Acker process, formerly used at Niagara, for obtaining caustic soda and chlorine from salt is similar to the Castner-Kellner process just described, but differs in that it employs molten lead in place of mercury as a seal, fused salt instead of brine as the electrolyte and operates at a temperature of 850° C. which is required to maintain the fused condition of the electrolyte. The containing vessel is a cast-iron tank five feet long, two feet wide and one foot deep, the sides above the molten lead being covered with magnesia so that the current must pass from the carbon anodes to the lead which acts as the cathode, the lower faces of the anode blocks being three-fourths inch above the lead. At one end of the tank is a small compartment separated from the remainder of the vessel by a partition dipping into the lead to such a depth that nothing but this fused lead can pass from one compartment to the other. The chambers are loosely closed by fire-clay slabs and the escaping chlorine drawn away through side flues by powerful exhausts. In the smaller compartment the lead is subjected to a stream of steam, which, acting upon the lead sodium alloy, forms ccustic soda and liberates hydrogen. The steam jet is introduced below the surface, but points vertically upwards, and the resulting spray strikes a curved hood which deflects it into a third chamber in which the lead and caustic separate, the latter flowing out of the furnace over a cast-iron lip, the lead sinking and passing back to the main chamber, while the evolved hydrogen is con- ducted away. The fused caustic is collected in an iron pan where it solidifies and is removed every hour. The output is 25 pounds of solid caustic per hour. This process avoids the evaporation of the water required in the Castner-Kellner process, but higher maintenance costs offset this advan- tage. The current employed per vessel in the Acker process is 2100 am- peres at from 6 to 7 volts, of which energy 54 per cent is used in chemical action and the remainder in maintaining the temperature. The same methods that have been commercially successful for the pro- duction of caustic soda and chlorine from salt are used to produce caustic potash and chlorine. Caustic potash is of value for the manufacture of soft soaps, the preparation of oxalic acid from sawdust, and for the ex- traction of metallic potassium. The raw material, potassium chloride (KC1), is more expensive than sodium chloride, costing approximately four times as much,* so it is an advantage to employ the electrochemical process which is more economical in raw material than an ordinary chemical method would be. Production of Metallic Sodium. — This metal was formerly ob- tained by the reduction of its carbonate or hydrate mixed with carbon, but at the present time all the metallic sodium employed in commerce is obtained by means of the Castner electrolytic process. The raw material is solid caustic which fuses readily at a low red heat and is obtained by the Castner caustic process already described. A diagrammatic view of the apparatus is shown in Fig. 6. The containing vessel is of steel, the electrodes are usually of cast iron. The electrical pressure employed is about 4.4 volts direct current, the action being as follows: The vessel is placed in an ordi- nary furnace flue, in which the gases are at a temperature high enough to maintain the caustic soda in a fused state. The current enters at the posi- * NaCl costs $9.00 per ton; KC1 costs $37.05 per ton. 1242 ELECTROCHEMISTRY. tive electrode, which is a hollow cylinder provided with vertical slits, so as to allow free circulation of the electrolyte. The negative electrode is placed at the bottom of the vessel, and terminates in the space in the center of the anode. A cylinder of iron wire gauze is placed between the electrodes, its function being to prevent the separated sodium from spreading over the entire surface and coming in contact with the oxygen liberated at the anode. The extreme fluidity of the fused caustic, however, allows it to pass readily through the gauze openings, while the greater surface tension of the liberated sodium will not allow it to pass through the same. The metallic sodium in its fused state has a lower specific gravity than the fused caustic, hence it remains at the surface, and is bailed out from time to time. The liberation of hydrogen at the cathode serves to protect the metal from the possible action of the oxygen. Potassium Chlorate is produced in considerable quantities both here and abroad. The Gibbs process used at Niagara Falls consists in the elec- Fig. 6. Castner Metallic Sodium Electrolytic Cell. trolysis of potassium chloride solutions, using a copper or iron cathode and a platinum anode. The cells are composed of a wooden frame, A, covered with some metal, B, such as lead, not attacked by the electrolyte. The latest form of cathode consists of a grid of vertical copper wires, C, kept in position by crossbars, D, of some insulating material, as shown in Fig. 6. The grid is placed in a vertical position against one side of the cell frame, and kept in place by the anode of the adjoining cell, from which it is insulated by the strips, F, and bars, D. The opposite side of the cell from that occupied by the cathode is par- tially closed by the anode (see dotted lines of Fig. 7). This consists of a thick lead plate, L, covered with platinum foil on the outer side, E (Fig. 8). This anode is held in position by the cathode and framework of the follow- ing cell. G is a pipe, reaching to the bottom of the cell, by which the po- tassium chloride is continuously supplied, and H is the overflow pipe to convey the mixed solution of the chloride and chlorate, as well as the lib- erated hydrogen gas away from the cell. S, S, S, S are lugs projecting PRODUCTION OF CAUSTIC SODA. 1243 from the framework, by means of which any number of cells can be bolted together to form a series of cells. Fig. 8 shows a group of three cells, the heavy plates (X and Y) being used to close the ends of the wooden frame- work, and form a fully closed series of cells with the only openings at the various supply and overflow points. The current connections are made at the points (m +) and (n — ). In normal working the cell is continuously fed by each of the supply pipes G, with a solution of potassium chloride, the rate of supply being so regulated as to maintain the temperature of the cell at 50° C. , and the amount of chlorate in the discharged solution slightly under 3 per cent. Since the plates C and L of each cell are in metallic contact, due to the lead lining, the electrolysis occurs between the anode of one cell and the cathode of the following cell (see narrow space between cells), this space Fig. 7. Gibbs Cell. Fig. 8. Gibbs Cell. being not more than one-eighth inch wide. The fact that the cathode is a grid allows the electrolyte to circulate around it, and all the solution thus passes upwards and out of the cells at H. The percentage of chlorate in the overflow solution is low, thus re- frigeration is necessary to recover it, and Fig. 9 is a representation of an electrolytic chlorate plant using this form of apparatus. £ is the supply tank, V the electrolytic cell, R the refrigerators, and P the pump by means of which the exhausted electrolyte is returned to the supply tank, while the chlorate precipitates out as crystals. The reason for using the refrigerator is that in solutions containing only 3 per cent of chlorate, the latter will not crystallize out upon natural cooling, as it would if present in large quantities. This low percentage of chlorate present is necessary to obtain quick recovery, as otherwise the presence of the hydrogen will cause secondary reactions, and cut down the efficiency of the conversion. The pressure employed is about four volts per cell, of which 1.4 is required to convert the chloride into chlorate 6KC1 + 6H 2 4- Elect. = 6KOH 4- 3H^ + 3CI^ 6KOH + 3H^ + 3Cb = 2KC10 3 + 4KC1 4- 3H^ and the remainder produces the heat that maintains the electrolyte at 50° C. which is necessary for the proper reaction. The current density is high, about 500 amperes per square foot of anode surface. At Niagara the plant consists of fifty such cells, connected up into two sets of 25 cells in series. A direct current of 10,000 amperes is supplied at 175 volts, which, allowing for line drop and losses at cell contacts, gives the proper pressure. 1244 ELECTROCHEMISTRY. Electrolytic chemical effects such as bleaching have been produced through the action of chlorine or other matter set free by an electric current. It is possible in this way to cause substances to act while in the nascent state and therefore more powerful. Disinfecting and deodorizing of sewage has also been accomplished in a similar manner, as in the Woolf process by Fig. 9. Arrangement of Gibbs Process, the electrolysis of a salt solution mixed with the sewage. The passage of the current liberates (Cl 2 ) chlorine and sodium hypochlorite (NaCIO), which act upon the refuse matter. Electrolytic chemical analysis is a special subject, the discussion of which is usually confined to books and journals relating particularly to chemical analysis; it is not ordinarily considered in connection with the general subject of electrochemistry. ELECTROTHERMAL CHEMISTRY. Electro thermal Chemistry includes those methods in which an elec- tric current raises the temperature of materials, usually to a high degree, in order to produce fusion, chemical action or other effects. Since elec- trolysis is not desired an alternating current is generally employed. The effect on the materials and the amount of product obtained is more or less proportional to the heat energy developed in the furnace. While the heat necessary to produce a certain change in a given amount of ma- terial is perfectly definite, the heat lost by radiation, conduction, etc., is variable, so that the efficiency must always be less than 100 per cent. The proportion existing between the heat energy employed in an electric furnace to produce a desired physical or chemical change and the total heat supplied is termed the efficiency of the furnace. The degree of efficiency attainable depends upon many factors: 1. The size of the furnace. Necessary temperature for the desired reaction. Protection from radiation. Arrangement of terminals. Method of recharging, continuous operation being most economical as the heat of the furnace walls is retained. 6. Method of removing the charge, it being undesirable to destroy a furnace to get at the charge. The most important of all these considerations is undoubtedly the size of the furnace, since the radiating surface of a small capacity is relatively greater than that of a large furnace. Consider two cubical furnaces: one of 1000 units' volume, the other of one unit's volume, the radiating surfaces would be 600 square units for the former, and 6 for the latter; hence the radiating surface for the smaller would be ten times larger per unit ca- pacity and the losses would be in the same ratio. Electric furnaces are divided into three general classes as follows: 'The material may be heated by passing current directly through it. The material may be heated by the heat gen- > erated in a conducting core. The material may be acted upon by heat radiated from an electric arc. „The material may be fed through an arc stream. ( Where the charge is conductive and is heated by 2. 3. 4. 5. Resistance Types. 6. Arc Types. c. Induction Type. currents induced in it. ELECTROTHERMAL CHEMISTRY. 1245 The phenomena occurring in a furnace may be subdivided as follows: a. Heating alone without fusion, as in the manufacture of graphite. b. Heating and fusion, as in the treatment of bauxite. c. Heating and chemical change without fusion, as in the manufacture of carborundum. d. Heating, fusion and chemical change, as in the manufacture of calcium carbide. Calcium Carbide. — This compound is produced by an electrothermal process invented by Willson in 1891, the total output throughout the world being about 300,000 tons in 1902. Its value lies in the fact that 1 pound of this substance mixed with water produces theoretically 5.5 and actually about 5 cubic feet of acetylene, equivalent in illuminating power to about 70 cubic feet of ordinary gas. The reaction yielding acetylene is CaC 2 + H 2 = CaO + C 2 H 2 . Various forms of electric furnace have been employed in the production of calcium carbide. One type invented by King and represented in Fig. 10 consists of an iron car, A, which holds the materials and carbide, at the same time acting as one electrode. It is run into place or removed as desired, and being provided with trunnions its contents may be tipped out. The other electrode consists of a bundle of carbon plates carried by a heavy rod, C, composed of a copper strip strengthened by iron side bars. The material fed through the channels G, F, consists of a mixture of 1 ton of burnt lime and f ton of ground coke to produce 1 ton of carbide, the reaction being CaO 4- 3C = CaC 2 + CO . An arc is first formed between the electrode, C, and the floor of the truck. The resulting high temperature con- verts the mixture into carbide, the electrode being gradually raised and more material added until the car is nearly filled with the product, when it is run out and replaced by another. At Niagara Falls a rotary form of furnace invented by C. S. Bradley is used, being operated continuously and producing about two tons in 24 hours when supplied with 3,500 amperes at 110 volts, or about 500 horse-power. Fig. 10. King Car- Since no electrolytic action is required, an alterna- bide Furnace, ting current is employed. Carborundum is a commercial name for carbon silicide (CSi) produced in large quantities according to the inventions of A. G. Acheson and his assistants. It is used as an abrasive, being hard enough to scratch ruby. Hit is formed by intensely heating in an electric furnace a mixture of 3£ tons of ground coke, 6 tons of sand and about .^ H tons of sawdust and salt, the yield being 3 or 4 tons»of P crystalline carborundum and about as much more of the 1 amorphous material. The furnaces used at Niagara Falls | consist ;of fire-brick hearths 16 feet long and 5 feet wide, loosely set together so that the liberated CO can readily escape, with solid brick walls at each end about 2 feet thick and 6 or 8 feet high as illustrated. In the middle of each of these walls there are iron frames through which the cur- rent is led to a core composed of carbon, weighing about 1000 pounds and extending the entire length of the furnace. This core is raised to a very high temperature (about 3000° C.) by passing through it for 36 hours an alternating current of about 1000 electrical horse-power at 190 decreasing to 125 volts. The heat from the core permeates the mass and converts it into carbon silicide, which is broken up after the furnace has cooled and used to make hones, wheels for grinding, etc. Manufacture of Graphite. — This application of the electric furnace depends only upon heat and was suggested to Acheson by the fact that when the temperature limit of the carborundum furnace was exceeded even slightly (250° C.) a large amount of graphite was formed around the conducting core. In fact, it has been stated that a variation of 3 per cent in the size of the carbon core one way or the other would seriously interfere with the working efficiency 1246 ELECTROCHEMISTRY. of the carborundum process — when the core is too small the heat becomes excessive and it is reduced to graphite — the silicon volatilizing. Acheson's experiments indicate that all metallic carbides are decomposed by the application of intense heat, the metal constituent volatilizing, the carbon remaining behind as practically pure graphite, and his patents are based upon this theory. The commercial work of the Acheson Company is in two lines: A. Graphitizing formed carbon objects. B. Graphitizing anthracite coal en masse. The product in every case is pure graphite. In case A. the material to be graphitized, is stacked up in a furnace be- tween the electrodes as a partial core 2 feet square and about 30 feet long, being thickly covered and the spaces between the pieces filled with a finely ground mixture of carbon and carborundum, alternating current of 3000 amperes at 220 volts is applied, and changed to 9000 amperes at 80 volts before the end of the run of about 20 hours. In case B it is found that the best results are obtained if the core con- sists of a rather impure form of carbon, one which when burned at ordinary temperatures would leave a large percentage of ash (10 to 15 per cent). This is ground to the size of rice chains and used as the furnace charge, with a conducting core of partially graphitized carbon, about 1000 H.P. of alternating current being applied for 20 hours. Alundum, the trade name for artificial corundum, is an abrasive made by a process due to C. B. Jacobs and others. Bauxite, a natural hydrated alumina, the same material as used in the Hall aluminum process, is cal- cined to drive off the water and then fed into an electric furnace, the con- struction of which is shown in the illustration. It consists of a conical 1111111 Fig. 12. Carborundum Furnace. sheet-iron shell mounted on a hydraulically operated plunger that raises and lowers it, to maintain a constant current of 2,000 amperes at 80 volts. The electrodes consist of two carbon rods that project into the shell, which is cooled by water, from the U-shaped trough, trickling down its outer surface. The time consumed for fusion is about 12 hours. The mass is allowed to cool and is then removed from the furnace by holding the sheet-iron shell in position and lowering the plunger, the product being broken up and sorted. It consists of four parts; namely, a red and blue mass in the in- terior, crystals that form in the blow holes, a porous outer portion and a by-product consisting of a metallic regulus of ferro-silicon which is used for the treatment of iron in the Bessemer and open-hearth furnaces. The porous outer part is used as a recharge, and the mass as well as the crystals, which are of the general nature of rubies and sapphires, in fact chemically identical with these gems, are ground up and used to make grinding wheels and other abrasives. Cyanides of Potassium and Sodium are produced electrochemically by the process of C. S. Bradley, C. B. Jacobs and others. A mixture of barium oxide or carbonate with carbon is heated in an electric furnace to produce barium carbide (BaC 2 ). While the mass is still hot, nitrogen (air cannot be used, as the oxygen present would oxidize the barium and carbon) is passed through it and barium cyanide forms, the complete reaction being: BaO + 3C + N 2 = BaC.N, + CO. The barium cyanide thus produced is treated with sodium carbonate, the result being a mixture of sodium cyanide and barium carbonate. The former is separated by dissolving it in water, the insoluble barium carbonate being used over again. Potassium cyanide is made in a similar ELECTROTHERMAL CHEMISTRY. 1247 manner and either salt is suitable for gold extraction and other purposes for which cyanides are employed. Electric Smelting*. — One of the earliest commercial processes in elec- trochemistry was that devised by E. H. and A. H. Cowles in 1884. A mix- ture of about 2 parts of alumina, 1 or 2 parts of granulated copper and 1 or 2 parts of carbon was introduced in a brickwork chamber. Bundles of carbon rods inserted at the ends formed the electrodes between which a current of 3000 amperes at 50 volts was maintained. At a very high temperature the alumina was reduced (A1 2 3 + 3C = AI2 + 3CO) and the resulting aluminum combined with the copper to form aluminum bronze. This process is no longer of commercial importance, since pure aluminum can be readily purchased; and when smelted with pure copper gives a better grade of aluminum bronze at a lower cost than is possible with the above method. Iron and steel can be produced by reducing iron ore with carbon in an electric furnace. For example, a mixture of magnetite and carbon can be heated by passing a current through it as in the Cowles aluminum bronze process; through a carbon core in contact with the material as in the car- borundum process; or by the action of an arc as in the carbide process. The reaction is Fe 3 4 4- 4C = 3Fe 4- 4CO. Pure (i.e., wrought) iron, cast iron or steel may be produced, depending upon the proportion of car- bon. The chief advantages are the directness of the process and the fact that the impurities in the fuel (sulphur, silicon, etc.) are not introduced. On the other hand, it is a question whether the electric furnace can com- pete in economy with the blast furnace and Bessemer converter. The field which is at present being developed is the conversion of scrap iron and pig iron into crucible steel by means of the electric furnace. This method offers reasonable chance of success, since the cost of crucible steel is high and therefore the method employed may be relatively costly. There are several distinctive types of furnaces employed, some being of the arc type, some of the resistance type, and another of the induction type. This latter method seems to be the most promising, since the pos- sibility of introducing anode impurities into the charge is absolutely done away with. X-RAYS. Revised by Edward Lyndon. The ultimate nature of X-rays is as much a matter of doubt at the present day as when Professor Roentgen presented his original papers in 1895. It is generally conceded that they are the product of cathode rays, these latter having their origin in electrical discharges through high vacua. X-rays are produced whenever cathode rays strike some solid substance, and the method employed for their production consists in exciting a vacuum tube, having electrodes sealed in its ends, by means of a static machine or from the secondary of a high potential induction coil. Under the influence of a high potential dark or cathode rays emanate from the negative terminal or cathode; these rays are repelled from the surface of the cathode, and where they impinge on a solid substance X-rays are emitted. X-rays and cathode rays are fundamentally different in that the cathode rays are subject to magnetic deflection, while X-rays are not. This fact is explained on the assumption that the cathode stream consists of particles moving at high velocity and carrying a negative charge. Such a stream is capable of being deflected by a magnetic field. When, however, the cathode stream strikes the solid substance, called the anti-cathode, the particles yield up their electric charge, and in passing from this point as X-rays show no magnetic deflection. The discharge of the cathode stream does not necessarily take place within the tube from terminal to terminal, but may be made to travel in any desired direction by altering the position and configuration of the cathode. The generally accepted idea is that these rays travel in lines normal to the surface from which they originate, and for this reason the cathode may be so shaped that the rays can be focused on the anti-cathode; that cathode rays can be focused is well known, but William Rollins holds that it is doubtful if the rays actually travel in lines normal to the cathode surface, reasoning that since the cathode stream is made up of moving particles carrying a negative charge there must exist a repelling force between all such particles; if this repelling force did not exist, the path of travel would be normal to the cathode surface, and the focus point would be found at the center of curvature of the cathode. Rollins states that the focus point lies beyond the center of curvature of the cathode and that this distance between the actual focus and the center of curvature increases with in- creasing potential across the tube terminals, due to an increased charge and consequent increased repelling force between the particles constituting the cathode stream. Where cathode rays strike upon glass or a like substance, the phenomenone of fluorescence appears. These rays are similar in many respects to X-rays, both are able to excite fluorescence, to affect sensitive films, and are sub- ject to selective absorption in passing through solid substances. The fact that reflection and refraction have not been conclusively shown by experiment to be properties of X-rays would indicate that these rays are not in the order of transverse vibrations. Quite recently, however, experiments have been made in which it was shown that X-rays are subject to polarization, and while reflection and refraction have not been absolutely proven to be properties of the rays, the generally accepted idea is that X-rays are ether vibrations of enormous frequency and short wave length. These rays, like ultra violet light, will discharge electrified bodies. This fact may be accounted for on the material theory of X-rays, on the assumption that when the charged particles making up the cathode stream strike the anti-cathode they yield up their electric charge and pass from this point as X-rays, to all purposes a stream of moving particles divested of their electric charge; these particles would then tend to become charged again in the presence of an electrified body. It 1248 X-RAYS. 1249 is more probable, however, that X-rays are ether vibrations, and that dis- charge of electrified bodies under their influence is due to ionization of the air, being similar in this respect to ultra violet light. Tubes, — Tubes for the production of X-rays are made of glass, the electrodes are sealed in the tube and the air exhausted, and upon the degree of vacuum depends the penetration of the X-rays emitted. It is desirable, and the general practice, to provide some metallic body in the tube upon which to focus the cathode rays, this being the anti- cathode, and it is from this body that X-rays are emitted. In Fig. 1, A is the anode, B the anti-cathode, and C the cathode. The relative positions of these terminals may vary, considerably with the different types, but in all cases the functions are the same. A separate electrode in the tube acting as the anti-cathode is not essen- tial in the production of X-rays; as they are emitted whenever the cathode rays strike any solid substance, they would appear if the cathode rays were focused on the glass tube itself, or the cathode rays may be focused so as to fall on the anode, making this single electrode both anode and anti- cathode. The anode and cathode are usually made of aluminum, as this mfetal undergoes very little disintegration under the action of discharge. Ov ing Fig. 1. to the difference in the expansion coefficients of glass and aluminum it is necessary to join the anode and cathode to platinum wires, sealing the platinum into the glass in order to make the external connections. Where the cathode rays strike upon a comparatively small area on the anti-cathode considerable heat is developed, consequently some metal, such as platinum, which is capable of withstanding high temperature, must be used for the anti-cathode. Under normal operating conditions the anode and the anti-cathode are connected to the positive of the source of supply, while the cathode is, of course, connected to the negative. Considerable care should be exercised in keeping the direction of current flow through the tube in the right direc- tion, for if the direction of current be reversed and continued for a length of time, blackening of the tube will result because of the disintegration of the platinum anti-cathode, and the tube becomes inoperative. The direc- tion of the current flow, per se, through the tube has nothing to do with the production of X-rays, but it is essential that the cathode stream should travel in such a direction at all times so as to strike the anti-cathode. The tube shown in Fig. 1 would emit X-rays if the exciting source were an alternating current of sufficiently high potential, but X-rays available for use, i.e., those sent out from the anti-cathode, would be emitted only half the time, or during that time in which the current would be normal in direction, while the tube would be subject to a certain amount of damage 1250 X-RAYS. during those portions of time in which the current flowed in the wrong- direction. Tubes have been made for use with alternating currents, one form of which is shown in Fig. 2. In the tube shown both terminals are so shaped as to focus the cathode rays from each terminal during the half cycle in which it is a cathode, upon a common anti-cathode. The penetration of X-rays is dependent upon the vacuum in which they originate, while the emissivity of the anti-cathode increases as the atomic weight of the substance forming it increases. Since the penetrative power of the rays is in a measure proportional to the degree of vacuum, several tubes of various degrees of exhaustion are necessary where the class of work is varied, and in all cases tubes should be selected for the particular use for which they are intended ; but one having a vacuum, the resistance of which is equivalent to a six or eight- inch spark gap, will give fairly good results for a variety of work. Fig. 2. A. W. Isenthal and H. Snowden Ward state that "there exists a condi- tion, the causes for which have not yet been sufficiently studied, when the tube emits rays of great penetration and withal yields a vigorous image, both on the fluorescent screen and on the plate. The characteristics of this stage of maximum efficiency are an incandescent anti-cathode with some traces of blue anode light in the tube. Unfortunately this state of affairs is more or less transient, and the tube soon becomes perforated." The vacuum gradually increases with the amount of use of tubes, this being ascribed to the fact that the anti-cathode and other platinum parts within the tube are subject to slow disintegration under the action of dis- charge, and the particles so separated, on cooling, occlude some of the residual gas in the tube. If the increased vacuum is due to the occlusion of the residual gas, ob- viously the original vacuum may be partially restored by the application of heat, the occluded gas being given up under the action of heat. This heat may be supplied by some external source or by sending through the tube a current of sufficient strength to appreciably warm it, the former method being preferable. In all cases it is advisable to include a spark gap in the circuit to the tube. It lessens the liability of the tube to puncture in case one of the electrodes becomes detached, and it acts as a gauge on the vacuum, discharge taking place across the gap if the vacuum and the consequent resistance of the tubes increase appreciably. X-RAYS. 1251 Regenerative Tabes. — It is impossible to prevent gradual changes in vacuum, and resulting changes in resistance and penetrative power of the ravs with continued use of a tube, but these changes from the original state may be minimized by the use of Regenerative Tubes, many types of which are on the market. There are certain substances, such as palladium, etc., which occlude gas at ordinary temperatures and yield up this occluded gas on being heated; advantage is taken of this property for maintaining the vacuum. One type of regenerative tube is shown in Fig. 3. Fig. 3. The absorbent is placed in a branch of the tube, shown at A; an auxiliary path for the current is provided through this branch, but under normal conditions no current passes via this auxiliary path. If, however, the vacuum increases beyond a predetermined spark length for which the ad- justable arm B is set, the current will then travel by way of the auxiliary path in preference to the path through the tube, with the result that the cathode rays from the auxiliary cathode in the absorbent chamber will heat the absorbent, causing it to give up its gas which lowers the vacuum in the tube. This gas, however, is reabsorbed when the tube cools. Another method of regeneration depends upon the fact that at high temperatures platinum is permeable to hydrogen. Fig. 4 shows a tube in which a platinum wire is sealed into the side neck of the tube at A and is protected by a glass cap. When the resistance of the tube increases ap- preciably the glass cap protecting the wire is removed, and as the latter is heated by means of a Bunsen Burner or a spirit lamp, hydrogen is in- troduced into the tube, lowering the vacuum. Fig. 4. The tube shown in Fig. 4 has an anti-cathode designed to obviate high temperatures at this point. This anti-cathode consists of a heavy metallic head with an oblique reflecting surface, the head forming part of a metallic tube which extends back into the comparatively cool side neck, this metallic tube being connected to the outside terminal by means of a wire. Due to the fact that the head and metallic tube have considerable mass and are good conductors of heat, exposing a large surface for radiation, the heating of the reflecting surface is not excessive. 1252 X-RAYS. Various forms of anti-cathodes have been devised to obviate high tem- peratures, generally taking the form of water cooling (not in direct contact), or by so disposing metallic bodies that the heat generated at the reflecting surface will be rapidly conducted away. Exciting* Source. — The minimum potential across the terminals of a vacuum tube for the production of X-rays has been variously estimated from 7000 to 100,000 volts. The appearance of X-rays, however, under a pres- sure of 7000 volts was due to special conditions, and, ordinarily, pressures much higher must be employed. High potentials could, of course, be obtained from specially designed transformers working on alternating current circuits, but since double focus tubes, adapted for alternating current, present difficulties in actual operation, their use has not become general, and other sources of high potential giving a uni-directional current are almost universally used. Static machines give very good results, their current being uni-directional and the potential practically constant, and therefore a steady discharge is produced through tubes excited from these machines. They are simple, and since they dispense with batteries and induction coils have much to recommend them; unfortunately, however, they behave in the most erratic fashion, the polarity being subject to reversal whenever rotation of the disks is discontinued, this, of course, being a serious disad- vantage. The most general method employed for excitation is by induction coils, giving high potentials at the terminals of the secondary winding. The induced current in the secondary winding is not, however, uni-direc- tional, but alternating in character. The wave form of the secondary current, while alternating, is not uniform, i.e., the induced E.M.F. due to rupturing the current in the primary circuit greatly exceeds the induced E.M.F. produced by closing the primary circuit, or, in other words, the in- duced E.M.F. at break is greater than E.M.F. of make. Fig. 5 shows the manner in which the current in the primary circuit varies. ^Break ■iBreak TIME Fig. 5. Because of the inductance of the coil, the current does not immediately reach its maximum value, but increases logarithmically as indicated by that portion of the curve marked "closed." The inclination of the curve, or the rapidity with which it reaches its maximum, will vary with the constants of the circuit for each particular coil, but Fig. 5 shows the general form of the current curve. The rapidity with which the current changes in a circuit is proportional to the time con- stant of the circuit or L/R, in which L is the self-induction and R the resist- ance of the circuit. When the circuit is ruptured, however, the time within which the current (alls to zero, depending upon the ratio of the inductance (L) and resistance X-RAYS. 1253 (R) of the circuit, is greatly diminished because R is increased enormously, due to opening the circuit. The ratio L/R, and consequently the time in which the current falls to zero, is very small as compared with the corre- sponding values on closing the circuit. Since the induced E.M.F. in the secondary circuit is proportional to the rate of change of magnetic lines through the turns of the secondary coil, it is evident that the induced E.M.F. of break will greatly exceed that of make, as the current of the primary circuit changes very much more rapidly in the former case than in the latter. Usually the E.M.F. due to closing the primary circuit is not of sufficient intensity to excite the tube, so, for this purpose, the current from the sec- ondary of an induction coil may be considered as uni-directional. Interrupters. — Interrupters for opening and closing the primary circuit should have the following characteristics: (1) Uniformity of inter- ruption, (2) high frequency, and (3) completeness^ of interruption. With respect to frequency of interruption there are limitations imposed by the properties of the iron core, and the disposition and number of turns of wire composing the coil. Since the primary current does not instantly reach its maximum value when the circuit is closed, a certain time must be allowed for this increase. If the speed of interrupter be such that the circuit is opened before the current has reached its maximum value, the full capabilities of the coil are not used. This condition is shown in Fig. 6, and the curves shown therein are for current in the primary with respect to time. k-c- k— C- fer-C— >- TIME Fig. 6. In the figure, the frequency of the interrupter is such that the circuit remains closedonly through the time interval indicated by the letter C, during which time the primary current has reached only a value shown by the height of the ordinate at the instant of interruption. A coil operating with an interrupter having too high a frequency may have its effectiveness increased if the E.M.F. impressed on the primary circuit be increased, thereby forcing the primary current to a higher value in the same time interval; on the other hand, the effectiveness may be increased under certain conditions by increasing the time of make and reducing the time of break, the frequency of the interrupter and the applied E.M.F. remaining the same. There are two general types of interrupters, viz., mechanical and electro- lytic. Many forms of mechanical interrupters have been devised and various designs are on the market in which provisions have been made for varying the frequency of interruption and the ratio of time of make and break. It is essential in all cases that the actual breaking oi the current should 1254 X-RAYS. (^ be as nearly instantaneous as possible, and to this end the spark forming between the breaking surfaces or points must be extinguished. In some instances sparking across contacts is obvi- ated by connecting a condenser across the interrupter, while in other designs the spark is blown out by a jet of air. The electrolytic or Wehnelt interrupter is shown in its simplest form in Fig. 7, and consists of two electrodes of widely dissimilar proportions, such as a platinum needle point and a large sheet of lead, im- mersed in a solution of dilute sulphuric acid. The platinum needle point is intro- duced into the electrolyte through a glass tube, the platinum being sealed into the glass, so that a very small area — practi- cally a point — is in direct contact with the electrolyte. If thesetwo electrodes be connected through an inductance to a source of supply, the current in the circuit will be subject to regular and rapid inter- ruptions. The platinum point electrode should be connected to the positive of the supply source. The speed of this type of interrupter is decreased by increasing the area of the positive electrode, other conditions re- maining the same, while increasing the applied E.M.F. increases the frequency and the current in the circuit. Fig. 8, shows complete diagram of connections for an X-ray outfit in which an electrolytic interrupter is made use of, the source of current supply being a storage battery. ^E>- ^^^^^ Fig. 7. Ammeter To .X.Ray s Tube i/ \ Spark Gap Inductance — 6=fo Induction Goil Fia. 8. As shown in the figure, a variable inductance is included in the circuit. This inductance is unnecessary if there is sufficient self-induction in the winding of the induction coil to properly operate the interrupter. A variable resistance is also included in circuit in order to vary the applied E.M.F. FLUORESCOPES. 1255 FLIOREiCOPES. The phenomenon of fluorescence is the emission of visible light when X- rays or cathode rays strike certain substances. In transforming the energy of X-rays into light for the examination of radioscopic images some substance must be used which fluoresces under the action of the rays. Roentgen originally used barium platino-cyanide, and this is very largely used now, although various other substances, such as potassium platino-cyanide and calcium tungstate, are in use. Since the amount of light given out by a fluorescent screen is small, it is necessary to exclude all other forms of light either by carrying out the ob- servations in a dark room or by enclosing the screen in some suitable obser- vation chamber having an opening for the eyes. The chemicals used in preparing the fluorescent screen are applied to some support, this support in turn being fastened in the observation chamber. Various supports for the chemicals, such as cardboard, vellum, blackened on one side, and rubber, have all been more or less used. ELECTRIC HEATING, COOKING AND WELDING. Revised by Max Loewenthal, E. E. For definitions of Heat, Units, Joule's Law, etc., etc., see pages 3 and 4, "Electrical Engineering Units." Various Methods of Utilizing* the Heat Generated by the electric Current. 1. Metallic Conductors (Uninterrupted Circuit). 1. Exposed coils of wire or strips. (a) Entirely surrounded by air. (b) Wound around insulating material. 2. Wire or strips of metal imbedded in enamel. (a) In the form of coils. ) Leonard, Simplex, Generj (6) In flat layers. J Crompton, and others. 3. Wire or strips of metal imbedded in asbestos and other insulating materials, (a) In the form of coils. (6) In flat layers. 4. Wire imbedded in various insulating compounds. (a) Crystallized acetate of sodium, etc. Tommasi. 5. A Film of metal. (a) Rare metal fired on enamel. It>~_~xi„ m ,. lb) Rare metal fired on mica. J ±Tometneus. (c) Silver deposited on glass. Reed. 6. Sticks of metal. (a) Crystallized silicon in tubes of glass. Le Roy. (6) Metallic powder mixed with clay and compressed. Parville\ 7. Metal in the form of powder or granules. (a) Kryptol. 8. Incandescent filaments in vacuum. (a) High wattage, low efficiency lamps. Dowsing, General Electric. II. Meat of the electric Arc (Interrupted Circuit). 1. The electric furnace. Siemens, Cowles, Parker, and others. 2. Heat of arc acting upon material, producing local fusion. Meritens, Werdemann, Bernardos, Howells, and others. 3. Welding by bringing metals in contact. Thomson. 4. Deflecting arc by magnet. Zerener. III. Hydro-electrothermic Sjstem, or Water-Pail Forg*e. Burton, Hoho and Lagrange. Referring to the above classification, Section I, the methods referred to under subheads 1 and 3 require no further explanation. The method unde* 1256 ELECTRIC HEATING, COOKING AND WELDING. 1257 subhead 2 consists in imbedding the resistance wire in some fireproof insu- lation such as enamel or glass. This insulation is of comparatively poor quality as a conductor of heat, and so thin that it affords the least possible resistance to the flow of heat from the heated resistance. The Simplex System {Carpenter Patents, subhead 2), employs high resist- ance wire imbedded in an enamel, consisting of two parts, the ground mass and the surface. The former consists of silica, crystallized borax (for flux- ing), fluorspar and magnesium carbonate, mixed in various proportions, powdered and fused. To this is added aluminum silicate and pure powdered quartz. The enamel proper consists of flint meal, also tin oxide, saltpetre, ammonia carbonate, lead sulphate, magnesium sulphate, potassium car- bonate, borax, and sometimes gypsum and arsenic. These are carefully mixed, as too much of any ingredient will make the enamel crack off, or will make the fusion point too high or too low. The insulation resistance varies from 40 megohms when cold to 1000 ohms at 400° C. Most enamels melt at about 900° C. ; The Creneral Electric quartz enamel type unit (subhead 2), consists of spirals of "Climax" resistance wire electrically insulated from the surface to be heated by quartz enamel. The quartz grains are used as an excellent binder for the enamel. The Oeneral Electric cartridge type unit (subhead 3), consists of a German silver wire flattened into a ribbon and wound edgewise in a spiral. To insulate between the turns of this spiral it is dipped in a bath of insulating cement. The mass is then squeezed together, so that a thickness of insu- lating material of .003 inch remains between the turns. The spiral, forming a solid cartridge, is slipped into a brass or German silver shell, with only .01 inch of mica between the edges of the ribbon and the shell. The heat, pass- ing through the thin thickness of mica is conducted to the outer shell and thence by direct contact to the surface to be heated. The Prometheus System (subhead 5) employs units composed of strips of mica about .004 inch thick, on which is painted a thin film of gold or plati- num, sometimes only .001 mm. thick. The metals, in the form of powders, are mixed with a flux and then painted on the mica, after which the whole is subjected to a high temperature, the finished films sometimes having a resist- ance of 100,000 ohms, each being made to consume not more than 70 watts, this giving a temperature of about 450° C. To prevent injury to the film it is covered with another strip of mica, and then together are partly enclosed in a thin metal frame. The insulation resistance of these strips varies from 50 to 300 megohms, and the increase in the resistance of the foil varies from 10 to 20 per cent during a period varying from 1 to 8 minutes. The Reed method or depositing a layer of silver on glass was described in the Electrical World, June 5, 1895. The method employed by JLelfcoy (subhead 6) consists of enclosing sticks of crystallized carbon, having a specific resistance 1333 as high as that of ordinary arc light carbon, in glass tubes. For 110 volts, rods are 100 mm. long, 10 mm. wide, and 3 mm. thick. This takes about 150 watts; and having a surface of 26 sq. cm., the dissipation of heat is at the rate of about 5 kg. calories per sq. cm. of surface, or an absorption of electrical energy of 6 watts per sq. cm. of surface. Parville (VEclairage Elec, Jan. 28, 1899) uses rods of metallic powder, mixed with fusible clay (quartz, kaolin), compressed under a pressure of 2000 kg. per sq. cm., and baked at a temperature of 1350° C. A rod 5 cm. long. 1 cm. wide, 0.3 cm. thick, has a resistance of 100 ohms, and absorbs 16500 watts per kg. One quart of water boils in 5 minutes with 15 amp. and 110 volts. Kryptol (subhead 7) is a patented German substance, consisting of a mixture of graphite, carborundum, silicate and clay in a granular form. A bed of this refractory material has an electrode of carbon at each end. The size of Kryptol granules varies according to the voltage. The current is de- termined by the thickness of the bed. Temperatures up to 3600° F. may be obtained. During a test made by H. Allen, a cube of copper weighing 8.45 grains was melted in one minute, the pressure being 240 volts and the current 15 amperes. The above methods are utilized in the construction of electric cooking and heating apparatus, while those enumerated under Sections II and III are employed for purposes of welding, smelting, and forging. 1258 ELECTRIC HEATING, COOKING AND WELDING. >d a "81 £ .M COM 00CO rH 0^3h *h y 0> © ft Si- ft .^ «^ft law ^coco CN t^ CO CN rH CN ^5 . cp CO n o ^5.!-! ^STZ! §-S » « CO as s J S3 O O '. co CN HH CN CN T* «jr PP.2 o o o XlxS 3 Sr5 < 311 H'8 ^ P •*» o*-d ® >H ^ — £ J) ftft Eg .2 fly yjd II ,0,2.3 »© g» y 5 C ©--h . I- II si*** -Still *_ -m r* y p s o.H 2 5 w^ £ 83 X p o XJ S-oil-sSs ft. SSo.ftftq &£ .^!§§§ >■§ -r^g+^^xj o P ; & Z~j © t>: ^ co «h ""*< t>. CO t- -co W 53 o . X3 s .s® .SI** * s ^ s a. y > fi »£% P . s5 co CN CO W* rH t^COCN ft yW 2 ©£ CO +2 T" cp js t-3 cs c3 co II ^ ft -FH Is* S o j £ : 5 o T3 e3 ^ ft co £ 2 > y rt y -^ Iff co co co . y * eo . 43 ^fa .CN y • ~ w -k y «2jXi o fe^cM»«Q. ftS s?i o3 CN CN • CO ■d o ft 3 O CN CO o s Jh o 4-1 (H «H • ft •d >>«? y 2 > y 63 y y a> u ^^ y ® |fc «fe CD COCN ^fO ^ sJ ^h CN ft^O 93 CN *"< yX3 . © • ftfci ' -fi CO P £ ^ fi +3 CN <-. 40. 86.1 AVERAGE WATTS % EFFICIENCY 2 4 6 8 10 12 MINUTES Fig. 1. Comparative Operating- Costs of Gas and Electric Cooking*. Report of Heating Committee, Association of Edison Illuminating Companies, September, 1905. The comparative operating cost of electric and gas cooking depends upon two questions, — the relative rates for gas and electric heat units, and the relative heat efficiencies of gas and electric apparatus. A third quantity — the effect produced by the different rates and modes of heat applications in the two classes of utensils — may effect the efficiency slightly, but the exist- ence of this effect is not yet verified. Starting with the heat of coal, which may be fairly estimated as 12,000 B.T.U. per pound, we compute the relative efficiency of the heat conversion as follows: Gas. 1 pound coal produces 5 cubic feet gas. 5 cubic feet gas contain 3000 B.T.U. Efficiency heat conversion is ^ = 25 per cent. Electricity. 1 pound coal produces 0.25 K.W. 0.25 K.W. contains 853 B.T.U. Efficiency heat conversion is 853 _ - 12000 =71perCent - Efficiency Electrical Heat Conversion _ Efficiency Gas Heat Conversion ELECTRIC COOKING. 1261 With manufacturing processes of equal cost per pound of coal converted, it is apparent, then, that an electric heat unit must cost nearly four times as much as a gas heat unit, but with present processes the relative rates are: ..1» Gas. $1 .00 per 1,000 cubic feet. 1 B.T.U. .000167 cents. 1 Electricity. ).10per K.W.H. B.T.U. 0.00293 cents. Electric B.T.U. 0.00293 Gas B.T.U. 0.000167 17.5. It is known that the efficiency of electrical apparatus is about four times that of gas, and, consequently, as the gas utensil requires four times as many B.T.U., the above figure of 17.5 is reduced to 4.4. If, then, the rate for electricity is reduced to one-quarter of that assumed, or 2 . 5 cents per K.W.H. this figure of 4.4 is changed to 1.1, and we have practically identical operating costs. Comparison between Ga» and Electric Rates. According to James I. Ayer (report for National Electric Light Associa- tion, May, 1904) electric heat at an average efficiency of seventy per cent equals .4197 K.W.H. per 1,000 effective heat units, and for 105,000 effective heat units there would be required 44.065 K.W.H. to give the same results. To compete with gas at equal rates, electricity will have to be sold at 5.67 cents per K.W.H. where gas is at $2.50 per 1,000 cubic feet, at 4.54 cents per K.W.H. where gas is at 2.00 per 1,000 cubic feet, at 3.40 cents per K.W.H. where gas is at at 2.83 cents per K.W.H. where gas is at at 2.27 cents per K.W.H. where gas is at 1.50 per 1,000 cubic feet. 1.25 per 1,000 cubic feet. 1 .00 per 1,000 cubic feet. The above is as fair a comparison as can be made where exact figures cannot well be secured. The results above quoted have been checked by records made in the same family alternately using gas and electricity each week for considerable periods in a number of cases, and from a variety of records obtained otherwise. It is assumed that suitable equipments both of electric and gas appliances are used. Cost of Operating- Electrically Heated Utensils. Article. Average Watt Hour Consump- tion. Period of Operation. Cost Dur- ing that Period at 10 cts. per K.W.H. Chafing dish Pint baby milk warmer and food heater Quart food heater Coffee percolator 400 250 500 300 500 800 1,200 60 250 500 500 500 300 300 200 1,000 1,000 50 Minutes. 20 6 6 20 15 15 15 15 30 30 30 12 20 20 30 30 30 per hour Cents. i h 1 Stove, 6 inches Stove, 8 inches li 2 Broiler 9 X 12 inch Curling iron heater Iron 3| lbs Iron 6 lbs Frying pan (7 inches diameter) . . . Waffle iron 3 2| 2| 1 Tea kettle 1 Glue pot, 1 quart Soldering iron, 2 lbs Doctor's sterilizer Bath room radiator Heating pad 1 1 5 5 * 1262 ELECTRIC HEATING, COOKING AND WELDING. S fa a — e ft a o O hJD H SuruoJi {•Bibadg o o ' paAjag o3 i JG ^ e3* " Oq i o C3^ $ 3 03 sl^gi a PQ £ O tf o3 oi ^3-c be ^ o3 O O OJ § o So I b0o3-* o3-S o3-^ i bC (- o u*Si fi o i o> a> o> ' O O O l o3 M ) o bD > Ui a) O O •paAJ9g| ^ saosja^ *C iO o iO >o »o CO 00 in 00 iO iO CO S H J C c8 s H 6t< 02 QQ w« WKW 00 coco t^OCOOS hHc ^rtHlOrH > c w » a > O 3 00?DOS o^2 03 C0rHrHT-l*< ELECTRIC HEATING. 1263 Electric Irons for Domestic and Industrial Purposes. The advantages of electric irons over irons heated by gas, coal, or other fuel are as follows: Cleanliness, continuous operation, saving time and energy by eliminating the travel between iron and source of heat, concentration of heat, so that the iron only and not the room is being heated, improved san- itary conditions and practically uniform temperature of iron face. In view of a number of these advantages, it has been found in actual practice that an average family of five persons, where the collars and cuffs are sent out to be ironed, consumes about 13.2 kilowatt hours per month for ironing, which at the 10 cent rate per K.W.H. amounts to SI .32 per month, which is about the same as if gas were used, costing $1 .00 per 1,000 cubic feet. The cost of operation varies with size of iron. For ordinary domestic requirements, without a current regulator, the iron most commonly used is one weighing about six pounds and consuming about 500 watts per hour. The regulators, whether of the switch in the handle or resistance in the stand type, effect a saving of from 15 to 20 per cent. The power consumption of the various types of irons is as follows: Watts 4 pounds Troy Polishing, diamond face 330 3i pounds Small Seaming (can be connected to lamp socket) . . 200 4 pounds Gentleman's Small Hat Iron 200 5^ pounds Light Domestic 500 b% pounds Light Domestic, round nose 500 7 pounds Domestic 600 5£ pounds Morocco Bottom 500 Morocco Bottom, round nose 500 Commercial Electric Laundry Equipment. (At Eshleman & Craig Company, Philadelphia, Pa.) Watts 7-5 pounds Sad Irons each 3.25 Amp. at 110 V. 2502 2-7 pounds Sad Irons each 3 .80 Amp. at 110 V. 836 2 Body Ironers each 41 .50 Amp. at 110 V. 9130 2-12 inch Sleeve Ironers each 12.40 Amp. at 110 V. 2728 1 Collar and Cuff Ironer each 6.50 Amp. at 110 V. 715 3 Bosom Ironers each 16.80 Amp. at 110 V. 5544 1 Rotary Collar Edger each 2.50 Amp. at 110 V. 275 1-7 pound Sad Iron each 23.00 Amp. at 24 V. 552 2-7 pound Sad Iron each 24.00 Amp. at 24 V. 1152 1 Collar Edging Machine each 6.25 Amp. at 20 V. 125 1 Heim Collar Shaper each 5.50 Amp. at 20 V. 121 Total Equipment 23.68 K.W. A full description of "A Model Electrically Operated Laundry," by H. S. Knowlton may be found in the July, 1905, issue of The Electrical Age, New York. ELECTlll'C HXATIIG. Unless electricity is produced at a very low cost, it is not commercially practicable to heat residences or large buildings. While this is true, the electric heater still has a field of application, in heating small offices, bath- rooms, cold corners of rooms, street railway waiting rooms, the summer villa on cool evenings, and in mild climates a still wider range. It has the peculiar advantage of being instantly available, and the amount of heat is regulated at will. The heaters are perfectly clean, do not vitiate the atmosphere, and are portable. Radiators and Convectors. {Prometheus Electric Company of England.) The heating of rooms and buildings can be accomplished either by radiant or convected heat. With the former method heating is effected by the agency of glow lamps, and with the latter by resistances working at com- paratively low temperatures. L264 ELECTRIC HEATING, COOKING AND WELDING. In the glow lamp type the filaments of the lamps are raised to an exceed- ingly high temperature, and the electric energy is transformed mainly into radiant heat, only a small portion being given off by conduction and con- vection — hence the name 'radiator." In the non-luminous type the resistances are either bare or embedded in ^namel and raised to a comparatively low temperature, which heats the air in contact with them, thereby setting up convection currents in the air. They are generally designated as radiators, though the term is a misnomer. They should rather be named "convectors or air warmers." The difference between these two methods of heating is a very wide one. The best method to employ depends entirely on the nature of the work for which the heaters are required, as explained below. Heating 1 l*y Radiation. — The heat from glow lamp radiators has been likened to sunshine. The analogy is excellent and has no doubt induced many non-technical people to universally employ this type of heat- ing in preference to any other, regardless of the nature of the work which they desire it to perform. It is very necessary in deciding which type of heater will give the most satisfactory results, to know the purpose for which it is to be used, and the conditions under which it will work. Radiant heat only raises the temperature of a body which is opaque to heat waves ; it passes through the air without heating it in the slightest, and only causes a rise of temperature in the air by heating any objects that offer opposition to its passage through them, these in turn heating the air in contact with them by conduction. Heat waves are unaffected by air currents and the glow lamp radiator is, therefore, suitable for warming oneself by out of doors, in balconies, etc., or for quickly warming any portion of one's body. The light emitted is also considered by some people to add greatly to the attractiveness of the heater. The heat rays are reflected forward by means of highly polished reflectors placed at the back of the lamps, and strike against any objects in their path. The zone of action is dependent on the shape of the reflectors, which for constructional reasons are made in simple shapes, confining the heating field to a small area. The temperature to which the glow lamp radiators will raise any opaque body when placed in any definite position relative to the lamps is dependent on the density of the heat rays on the surface on which they fall, from which no doubt has arisen the popular fallacy that a radiator, in front of which it is uncomfortable to hold one's hands, must be emitting more heat than a convector, in front of which they may be kept for any length of time without any sense of discomfort. The only true measure of the rate at which heat is being developed by two different heaters working under exactly similar con- ditions is the amount of air heated per unit of time multiplied by the tem- perature through which it is raised. Thus a heater constructed to work at a very low temperature may be giving out far more heat than one working at a high temperature, though the former would appear to be the more powerful of the two if gauged merely by the sensation produced on putting one's hands close to the flames. Air warming by radiant heat is an indirect method by which uniformity of temperature throughout a room or building can never be attained. It is of the utmost importance that the temperature be uniform, as freedom from draughts and consequent comfort and healthy conditions cannot other- wise be secured. Heating: oy Convection. — The heat generated in the resistance warms the body of the convector, and the air is heated by direct contact with the hot surfaces. Convection currents are consequently set up in the neigh- boring air, which quickly equalizes the temperature throughout the room in which the convector is placed. This method of heating dwelling rooms is, therefore, under normal conditions, far more efficient than that of radia- tion, provided the temperature of the resistance material is not high enough to materially affect the humidity of the air. Convectors are not, however, in virtue of the comparatively low temperature at which they work, so efficient as radiators for quickly warming one's hands or any portion of one's body, neither can they compete with radiators when very strong air currents are present, or for open air work such as balconies, band stands, etc. It has been asserted that convectors do not, like radiators, accomplish useful work as soon as they are switched in. Such broad statements are ELECTRIC CAR HEATING. 1265 not based on facts as the relative rate of air heating by a radiator or convec- tor, absorbing the same power, depends entirely on their capacity for heat. Naturally a convector with a heavy cast iron frame will absorb a large quan- tity of heat before it can work at its maximum efficiency, but all the heat that is stored in the frame is, of course, taken up by the air after the convector is switched off; such convectors, therefore, are suitable only for continuous work over long periods. Energy Consumption of Electric Heaters. According to Houston and Kennelly, one joule of work expended in producing heat will raise the temperature of a cubic foot of air about fe° F. The amount of power required for electrically heating a room depends greatly upon the amount of glass surface in the room, as well as upon the draughts and admission of cold air. An empirical rule, commonly employed, is to figure from li to 2 watts per cubic foot of space to be heated. According to an European authority if a sitting-room with a content of 100 cubic meters is to be heated to 17° C, while the temperature of the outside is 3° C, he estimates that 3,500 kilogram calories are required per hour; with electric heating this means a consumption of 4 kilowatt-hours for every hour, while with coal fuel, about 3 kilograms of coal are required per hour. Experience has shown, says the same authority, that for every degree Centigrade difference between the lowest outside temperature and the desired inside temperature and for every cubic meter of space to be heated 1 to 1.5 watts of electric power are required; as an approximate average 1 . 2 watts may be assumed. For instance, if the outside temperature is 10° C. below, and a sitting-room of 50 cubic meters is to be heated to 18° C, the difference of temperature is 28° C. Hence, 1,680 to 1,800 watts are required, while the time in which the desired temperature is obtained varies from one to three hours, varying of course, according to whether the neigh- boring rooms are heated or not. Comparison between Electric and Coal Heating*. A kilowatt-hour in heat is about 3,600 B.T.U., and costs a consumer in our large cities from 5 to 20 cents according to the conditions, or from 72,000 to 18,000 thermal units per dollar. On the other hand a short ton of ordi- nary good steam coal will contain 28,000,000 of B.T.U. and allowing a loss of 25 per cent in a boiler wall and flue, some 21,000,000 of heat units can be looked for in boiler water, such coal costing from one to three dollars per ton according to circumstances, and representing a yield of 21,000,000 to 7,- 000,000 of thermal units per dollar, or in the neighborhood of three hundred times more heat than the electric method would furnish. The comparison is in a certain sense unjust, seeing that the retail price of electric energy on a small scale is compared with manufacturing cost of fuel alone for heating water on a large scale, and a far better relative showing could be made where both methods were compared from either the manufacturer's or the pur- chaser's standpoint, whatever the scale of production might be. (Editorial Electrical World and Engineer.) ELECTRIC CAM UEATOO. At the Montreal meeting of the American Street Railway Association in 1895, Mr. J. F. McElroy read an exhaustive paper on the subject of car- heating, from which the following abstracts are taken: In practice it is found that 20,000 B.T.U. are necessary to heat an 18 to 20 foot car in zero weather. When the outside temperature is 12£° F. only 16,000 B.T.U. are required, etc., which shows the necessity of having electric heaters adjustable. The amount of heat necessary in a car to maintain a given inside tem- perature depends on: 1. The amount of artificial heat which is given to it. 2. The number of passengers carried. The average person is capable of giving out an amount of heat in 24 hours which is equal to 191 B.T.U. This is evidently an error, as Kent says that a person gives out about 400 heat units per hour; and tests by the Bureau of Standards show the same (413) for a person at rest, and about twice that for a man at hard labor (835)« 1266 ELECTRIC HEATING, COOKING AND WELDING. Fig. 2. Cost of Car Heating-. The following table was compiled by Mr. McElroy from the reply re- ceived from the Albany Railway Company : Average fuel cost on Albany Railway, per amp. hour — .241 cent. Average total cost for fuel, labor, oils, waste, and packings per amp. hour = .423 cent. Cost of fuel per hour for heating a car with electric heaters with coal at $2.00 per 2000 lbs. Position of Switch. 1st 2d. 3d. 4th. 5th. Amperes equal. 2.14 2.88 6.88 8.09 12.0 Simple high speed condensing . . Simple low speed condensing . . Compound high speed condensing Compound low speed condensing cts. .43 .40 .39 .36 cts. .58 .54 .52 .48 cts. 1.40 1.30 1.27 1.17 cts. 1.62 1.51 1.47 1.36 cts. 2.41 2.24 2.20 2.03 Average Cost Per Day for Stoves. 33 lbs. of coal at ,$4.55 per ton $.075 Repairs 005 Dumping and removing coal and ashes, coaling up and kindling fire, including cost of kindling, and part of cleaning car 100 Removing stoves for summer, installing for win- ter, repairing head linings, repainting, etc., average per day 0125 Total |Tl926 ELECTRIC CAR HEATING. 1267 Diagrams of Wiring* for "Consolidated" Heaters for Use Along- Truss Plank. TROLLEY, SWITCH 6-Heater Equipment *^*^ *Z 16-Heater Equipment . orou^q 24-Heater Equipment »C»* t« ip Lj.tC'»'CHC»'t>V -!■> - (..;:•« :-W[»." "■ ■ "'-»' IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIMIIIIIIIIIUirilliftNWMlUIIIJIIIIIMIIUIIIIlll 5- ->. «g*nwo co _^^ ~XX§C~ co ^ SOLIJA7E ^S*g-»iCfe»ttg-CoTg^^ j " ' ; s o?' "tO-'SC V :i.- ZJ t •* 'I' 1 ■ , , ;Tr »i»'"i|r --r..^.M— ^ Truss Plank Heater in position, showing wiring in moulding. Fig. 3. 1268 ELECTRIC HEATING, COOKING AND WELDING. Diagrams of Wiring* for fct Consolidated " Heaters for Cross Seats. 8«H.e&ter Equipment 12-Heater Equipment UJ ILl Cross Seat 'Consolidated" Heater in position. Fig. 4. ELECTRIC CAR HEATING. 1269 According to a paper read by J. T. McElroy before the Street Railway Association of New York, on car heating, about 10 to 20 per cent of the energy required for running is spent in the heaters, and the average of tests taken upon American cars with coal and electric heaters for 15-hour runs gave the price per day of 15 hours for coal as $2.33, and for electricity $2.20. Pointers to Purchasers of Electric Car Heaters. (Street Railway Journal, November 5, 1904.) We think it only fair to the electric heater to call attention to a very common fault on the part of companies purchasing electric car heating equipments, which fault usually results in the end in a condemnation of electric heaters. This fault lies in trying to get along with a few heaters worked at a high temperature rather than a large number worked at a lower temperature. The reason why companies attempt to do this is, of course, to reduce the first cost of heater equipment. If a car is to be heated as comfortably by electric heaters as by hot water, the nearer you can come to distributing the heat evenly throughout the length of the car and avoiding excessively hot points, the better will be the results. It is coming to be more and more established, that heating of any kind can be done more efficiently by a large radiating surface worked at low temperature than by a small radiating surface worked at high temperature. Furthermore, working electric heaters at low temperatures is conducive to a long life, while working at high temperature is not. Industrial Electric Heating 1 . Among the industries to which electrically heated apparatus has been successfully applied may be mentioned: Book binderies, printing shops, hat factories, candy and chocolate manufactories, laundries, wood-working establishments, shoe, paper box, glove, corset, dental goods factories, as well as hotels, hospitals, restaurants, laboratories, bakeries, etc. In fact, wherever gas or steam is being employed for the localized application of heat, electricity has been found, in most cases, a more sanitary, flexible, safer, cleaner, as well as equally economical source of heat. Electric Heat in Printing* Establishments. — The most ex- tensive, as well as most economical, heating equipment in a printing office, is, no doubt, that at the Government Printing Office at Washington, D. C, designed and installed by the Hadaway Electric Heating Company. The following pieces of apparatus are being electrically operated success- fully at the present time (1907) in this office: Matrix Drying Tables. Wax Stripping Tables. Wax Melting Kettles. Case Warming Cabinet. Case Warming Table. Wax Knife, Cutting down Machine. Building up Tool Heaters. Sweating-on Machines. Soldering Iron Heaters. Embossing and Stamping Press Heads. Glue Heater Equipments. Glue Cookers. Case Making Machines. Book Cover Shaping Machines. Finishers' Tool Heaters. Pamphlet Covering Machines. Sealing Wax Melters. Further details of this equipment have been published in the Washington Electrical Handbook, issued in September, 1904, by the American Institute of Electrical Engineers, and a series of articles in the Electrical World and Enqineer, Vol. 43, pages 9-14, and succeeding issues. The claims made by the government representatives in favor of elec- erically heated apparatus as compared with steam and gas, are as follows: 1270 ELECTRIC HEATING, COOKING AND WELDING. The absence of excess of heat that would be found in forms other than electrical. The ability to reduce the amount of time necessary to make impressions. The ability to bring the apparatus to a working condition in less time. The fact that in eight years of operation they have not had an instance of a burnt-out coil. electrically Heated Devices in the Printing* Shop of J?. JP. Collier & Son, Ufew York. The following list of apparatus is given here in order to show some of the details of this class of apparatus as well as the developments of this class of industry. Apparatus. Type and Size. Max Amp. Amp- Min. Volts Watts 2 glue pots . 23 glue pots 1 glue pot . 8 glue pots 2 glue pots . 2 wax heaters 5 press heads 1 press head 1 press head 1 press head 1 press head 1 press head 1 press head Simplex 20 gal Hadaway 1 qt Simplex 1 qt Hadaway 2 qt 2 gal 22 in. X 24 in. X 3| in. 22 in. X 24 in. X 3| in. 22 in. X 24 in. X 3| in. 22 in. X 24 in. X 3| in. 22 in. X 24 in. X Si in. 19 in. X 12 in. X 3f in. 12 in. X 12 in. X 3$ in. 100 2 2.5 10 22.8 100 35 36 36 36 36 30 25 22 2.5 40" 2.8 4 3.6 3.5 4.5 2.5 2.5 110 110 110 110 220 110 110 110 110 110 110 110 110 22,000 5,060 275 8,800 12,672 22,000 19,250 3,960 3,960 3,960 3,960 3,300 2,750 111,947 Forty-nine articles, consuming 112 Kilowatts. (11 Press Heads. Summary < 36 Glue Pots. ( 2 Wax Heaters. Laboratory Use. — The milk supply of New York City is governed by tests made in the Laboratory of the Board of Health, by means of electric stoves. Twenty-five 4-inch disc stoves, of 60 watts capacity, are used to boil the ether used in the tests. Fourteen times per hour these little stoves cause the ether to vaporize. The germ producer, measuring 22X22X22 inches, is heated to 130° C, by means of electricity, a maximum current of 16 amperes being employed for 15 minutes every hour, while 3 amperes keep up the desired temperature. Coffee and Cocoa Dryeri. — The cocoa and coffee trade has applied electric heat to its small desiccating or drying cabinets. A dryer 3£ feet by 5 feet, requiring a temperature of 150 degrees, requires about 74 watts per cubic foot when properly jacketed. The beans are particularly susceptible to the odors arising from combustion, hence the advantage of electric heat. For drying kilns 40 watts per cubic foot are recommended. Candy Manufacture. — Warming tables and chocolate dipping-pots have proved successful. Fifty watts produce sufficient heat to keep the chocolate in working condition. A 30-gallon tank holding caramel paste is supplied with 10 kilowatt hours to keep the paste at 285° C., and each melt- ing costs about 65 cents. The service is intermittent, hence the adapta- bility of electric heat. Soldering* and Branding* Irons. — The canning industry, as well as the makers of switchboards, and others, find the electric soldering iron a useful and economical tool. It has been found more economical to oper- ate electric soldering irons heated by current costing 5 cents per kilowatt hour than irons heated in gas furnaces, with gas at $1.00 per 1000 cubic feet. Heaters of 110- watt capacity are made, into which a soldering iron is thrust, thereby doing away with the connecting handle cord. One thousand hogs per hour are stamped, "Inspected," by the government meat inspectors in Chicago, by means of a 400-watt branding tool, which is an electric soldering iron with a die inserted in place of the copper tip. ELECTRIC WELDING AND FORGING. 1271 Thawing: Water Pipes. The following figures show the details of operation of a 44-cell storage battery outfit, mounted on an automobile truck, in comparison with those obtained by the use of a rheostat in series with a direct-current 3- wire Edison system with the neutral wire grounded. The figures represent the average amounts in each case. Am- peres. K.W. Hours . Time, Min. Pipe, Inch. Volt- age. Cost per Case. Revenue per Case. Storage battery Street supply . . 513 275 1.39 10.4 5.44 19.0 1 f 31.5 120.0 $10.85 14.43 $16.40 16.93 The street supply is used until the season has so far advanced that the number of cases will warrant the exclusive service of an automobile truck. ELECTRIC WEEDING AND FORGING. The current employed in electric welding may be theoretically either continuous or alternating, but on account of the difficulty of producing low tension continuous currents, it is only practicable to employ alternating current. All electric welding machines are fitted with an alternating cur- rent transformer as an integral part of the machine. Thomson Electric Welding* Process. The principle involved in the system of electric welding, invented by Prof. Elihu Thomson, is that of causing currents of electricity to pass through the abutting ends of the pieces of metal which are to be welded, thereby generating heat at the point of contact, which also becomes the point of greatest resistance, while at the same time mechanical pressure is applied to force the parts together. The passage of the current through the metal at the point of junction, gradually but quickly brings the temperature of the metal to a welding point. Pressure follows up simultaneously, a weld being effected at once. Horse-Power Used in Electric W r elding > . The power required for the different sizes varies nearly as the cross sec- tional area of the material at the joint where the weld is to be made. Within certain limits, the greater the power, the shorter the time; and vice versa. The following tables are based upon actual experience in various works, and from very careful electrical and mechanical tests made by reliable experts. The time given is that required for the application of the current only, and may be shortened with a corresponding increase in the amount of power applied. Round Iron or Steel. Diameter. Area. H.-P. Applied Time in to Dynamo. Seconds. i in. .05 2.0 10 1 in. .10 4.2 15 £ in. .22 6.-5 20 1 in. .30 9.0 25 fin. .45 13.3 30 1272 ELECTRIC HEATING, COOKING AND WELDING. Extra Heavy Iron JPipe. Inside Area. H.-P. applied Time in Diameter. to Dynamo. Seconds. i in. .30 8.9 33 | in. .40 10.5 40 1 in. .60 16.4 47 H in. .79 22.0 53 lh in. 1.10 32.3 70 2 in. 1.65 42.0 84 2£ in. 2.25 63.7 93 3 in. 3.00 96.2 106 General Table. Iron and Steel. Copper. Area in Time in H.-P. applied Area in Time in H.-P. applied sq. in. Seconds. to Dynamos. sq. in. Seconds. Dynamos. 0.5 33 14.4 .125 8 10.0 1.0 45 28.0 .25 11 23.4 1.5 55 39.4 .375 13 31.8 2.0 65 48.6 .5 16 42.0 2.5 70 57.0 .625 18 51.9 3.0 78 65.4 .75 21 61.2 3.5 85 73.7 .875 22 72.9 4.0 90 83.8 1.0 23 82.1 \" square " \\" round " 30 35 l£" square " 2" round " 1 40 1 75 2" square " 90 Axle l^elding*. 1" round axle requires 25 Horse-power for 45 seconds. 48 60 70 95 100 The slightly increased time and power required for welding the square axle is not only due to the extra metal in it, but in part to the care which it is best to use to secure a perfect alignment. Tire TFelding'. 1" x T 3 g tire requires 11 Horse-power for 15 seconds. \\" x §" " " 23 " " " 25 l£"xf" " " 23 " " " 30 \\" xf" " " 23 " " " 40 2" x£" " " 29 " " " 55 2" xf" " " 42 " " " 62 The time above given for welding is of course that required for the actual application of the current only, and does not include that consumed by placing the axles or tires in the machine, the removal of the upset, and other finishing processes. From the data thus submitted, the cost of welding can be readily figured for any locality where the price of fuel and cost of labor are known. ELECTRIC WELDING AND FORGING. 1273 A test on the electric welding equipment of the American Steel Frame and Band Iron Company of New York, made by the New York Edison Company, to determine the amount of energy used per weld, gave the following result. The equipment consists of a 50 horse-power 220 volt, direct current motor, belted to a 50 kilowatt 220 volt, 2 phase, 60 cycle, separately excited alternator, and three 7.5 kilowatt step-down transform- ers, with an approximate ratio of 45 to 1. When welding iron frames .0352 square inch in cross section, it takes 1 kilowatt hour, supplied to the transformer, to make 500 welds, the time required being 53 minutes. This averages 2 watt hours per weld, and taking the time the current is applied as 0.7 seconds per weld, the welding current figures out about 2000 amperes at 4.75 volts. A meter installed in the motor circuit showed 4 . 2 kilowatt hours direct-current input for 390 welds, making an average of 10.77 watt hours per weld. Electric Rail Welding-. The "Electric " joint, applied by the Lorain Steel Co., is made by welding plate? on both sides of the web of the rail. The plates shown in Fig. 6 are 1 inch by 3 inches, by 18 inches, and have three bosses, three welds DIAGRAM OF CONNECTIONS OF RAIL INCLDCR T * TROLLEY C . B'CI*0UIT BREAKE* R. R -RHEOSTATS M ■ tOOTQR e - booster R T • ROTARY TRANSFORMER W.T WELOtHG TRANSFORMER 3 W' SWITCH R C-RB ACTIVE COIL trC'nuomo ua*p Fig. 5. SKETCH OF BAH USED IN WILDING O o t • • 1 J*..... -. / 8' ■ ^ 1 i ■,..■ ?.. — *r Web Plates Fig. 6. being made at each joint. Great pressure up to 35 tons is maintained on the joint whilst making and cooling. The welding current runs as high as 25, 000 amperes. The connections are shown in Fig. 5- 1274 ELECTRIC HEATING, COOKING AND WELDING. Zerener System. In this system an arc is used in combination with a magnet which deflects the arc, making a flame similar to that of a blow-pipe, but having the tem- perature of the arc. The apparatus contains a self-regulating device which is driven by a small electric motor ; for welding iron a current of 40 to 50 amperes- at 40 volts will suffice for strips of metal three mm. thick. Her na rd <>!* System. In this system the article to be operated upon is made to constitute one pole of the electric circuit, while a carbon pencil attached to a portable insulated holder, and held by the workman, constitutes the other pole, the electric arc — which is the heating agent of the process — being struck between the two poles thus formed. This system has been used extensively in England for the repair of machinery. The Barrbeat-Strange Patent Barrel Syndicate use this system for the welding of the seams of sheet- steel barrels. Voltex Process for Welding- and Brazing* Consists in the use of an electric arc formed between two special carbon rods inclined to each other at an angle of about 90°. The whole apparatus can generally be held in one hand. With gas and coke, gas costing only 70 cents per 1000 cubic feet, it is claimed the complete cost of brazing and filling up a bicycle frame is $1.43, while with the Voltex process, at 6 cents per kilowatt hour, it is only 46 cents. Stassano Process of Electric Smelting* Consists of heating, in an arc furnace, briquettes composed of iron ore, carbon, and lime made into a paste with tar. The smelting process occurs in a blast furnace, the iron being reduced, and the siliceous matter of the ore slagged off. Annealing* of Armor Plate. The spot to be treated is brought to a temperature of about 1000 ° F. The current used is equivalent to 40,000 amperes per square inch, a density which is only possible by the use of cooling by water circulation. The operation generally takes seven minutes. HYDRO-ELECTROTHER^IIC sYsTEMS. S3«>ho and JLag-rang-e System. In this system an electrolytic bath is employed, into which an electric current of considerable E.M.F. is led, passing from the positive pole which forms the boundaries of the bath and presents a large surface to the elec- trolyte and thence to the negative pole, consisting of the metal or other material to be treated, and which is of relatively small dimensions. Through the electrolytic action hydrogen is rapidly evolved at the nega- tive pole and forms a gaseous envelope around the pole ; as the gas is a very poor conductor of electricity, a large resistance is thus introduced in the circuit, entirely surrounding the object to be treated. The current in "passing through this resistance develops thermal energy, and this is com- municated to the metal or other object which forms the negative pole. This system has been extensively used in England, and is described in The Electrical World, Dec. 7, 1895. Hurton Electric ITorg-e. In a patent granted to George D. Burton on an electrolytic forge, the portion to be heated is placed in a bath consisting of a solution of sal soda, or water, carbonate of soda, and borax. The tank is preferably made of porcelain or fire-clay. The anode plate has a contact surface with the liquid much greater than the area of contact of the article to be heated. This plate is composed of lead, copper, carbon, or other suitable conducting material. FUSE DATA. 1275 Fl'SE DATA. In a lecture on "The Rating and Behavior of Fuse Wires," before the A. I. E. E,, in October, 1895, Messrs. Stine, Gaytes, and Freeman arrived at the following conclusions: 1. Covered fuses are more sensitive than open ones. 2. Fuse wire should be rated for its carrying capacity for the ordinary lengths employed. 2(a). When fusing a circuit, the distance between the terminals should be considered. On important circuits, fuses should be frequently renewed. The inertia of a fuse for high currents must be considered when protecting special devices. • Fuses should be operated under normal conditions to ensure cer- tainty of results. Fuses up to five amperes should be at least 1£ inches long, one-half inch to be added for each increment of five amperes capacity. 7. Round fuse wire should not be employed in excess of 30 amperes capacity. For higher currents flat ribbons exceeding four inches in length should be employed. fuse Wire. The following table shows the sizes of fuse wire and the approximate current-carrying capacity of each size: Tested fuse IFire. (Chase-Shawmut Company, Boston.) 3. 4. 5.- 6. Carrying Capacity in Amperes. Standard Length in Inches. Diameter in Mils. Feet per Pound. f 1* 10 2,700 i if 17 950 l 1* 20 670 li if 23 510 2 if 25 430 3 1! 27 370 4 30 300 5 2 35 220 6 2 38 185 7 2 44 140 8 2 47 120 9 2 54 93 10 2 58 80 12 3 62 70 14 3 68 60 15 3 70 52 16 3 73 49 18 3 78 43 20 4 86 86 25 4 90 82 30 4 100 26 35 4 110 22 40 4 122 18 45 4 126 13 50 4 147 12.5 60 5 160 10.3 70 5 172 9.0 75 5 178 8.3 80 5 190 7.5 90 . 5 198 6.7 100 5 220 5.5 1276 ELECTRIC HEATING, COOKING AND WELDING. Installation of fuses. (H. C. Cushing, Jr.) Enclosed fuses of standard sizes are now on the market and are preferable to link fuses. Where the link fuses are used they should have contact sur- faces of tips of harder metal, having perfect electrical connection with the fusible part of the strip. The use of the hard metal tip is to afford a strong mechanical bearing for the screws, clamps, or other devices provided for holding the fuse. They should be stamped with about 80 per cent of the maximum current they can carry indefinitely, thus allowing about 25 per cent overload before the fuse melts. The following table shows the maximum break distance and the separation of the nearest metal parts of opposite polarity for plain open link fuses, when mounted on slate or marble bases for different voltages, and for different currents: 125 VOLTS OR LESS. Separation of Nearest Metal. Parts of Opposite Polarity. Minimum Break. Distance. 10 amperes or less 11-100 amperes 101-300 amperes f inch 1 inch 1 inch f inch 1 inch 1 inch 125 TO 250 VOLTS. 10 amperes or less 11-100 amperes . 101-300 amperes 1£ inch li inch 1J inch Fuse terminals should be stamped with the maker's name, initials, or some known trade-mark. The lengths of fuses and distances between terminals are important points to be considered in the proper installation of these electrical "safety valves." No fuse block should have its terminal screws nearer together than one inch on 50 or 100 volt circuit, and one inch additional space should always be allowed between terminals for every 100 volts in excess of this allowance. For example: 200 volt circuits should have their fuse terminals 2 inches apart, 300 volts 3 inches, and 500 volts 5 inches. This rule will prevent the burning of the terminals on all occasions of rupture from maximum current, and this current means a "short circuit." Enclosed Fuses. — The "Enclosed Fuse " or "Cartridge Fuse," con- sists of a fusible strip or wire placed inside of a tubular holding jacket, which is filled with porous or powdered insulating material through which the fuse wire is suspended from end to end. The wire, tube and filling are made into one complete self-contained device with brass or copper terminals or ferrules at each end, the fuse wire being soldered to the inside of the ferrules. When an enclosed fuse "blows" by excess current, the gases resulting are taken up by the filling, the explosive tendency is reduced and flashing and arcing are eliminated. "D. & W.," "G. E.," "Noark" and "Shawmut," enclosed fuses are approved by the National Electric Code. LIGHTNING CONDUCTORS. Views concerning the proper function and value of lightning rods, con- ductors, arresters and all protective devices have undergone considerable modification during the past ten years. There may be said to be four periods in the history of the development of the lightning protector. The first embraces the discovery of the identity of lightning with the disruptive discharge of electrical machines and Franklin's clear conception of the dual function of the rod as a conductor and the point as a discharger. The second begins with the experimental researches of Faraday and the minia- ture house some twelve feet high, which he built and lived in while testing the effects of external discharges. Maxwell's suggestion to the British Association, in 1876, embodies a plan based upon Faraday's experiments, for protecting a building from the effects of lightning by surrounding it with a cage of rods or stout wires. The third period begins with the experiments of Hertz upon the propagation of electro-magnetic waves, and finds its most brilliant expositor in Dr. Oliver J. Lodge, of University College, Liverpool, whose experiments made plain the important part which the momentum of an electric current plays, especially in discharges like those of the lightning flash, and all discharges that are of very high potential and oscilla- tory in character. The fourth period is that of the present time, when individual flashes are studied ; and protection entirely adequate for the particular exposure is devised, based upon some knowledge of the electrical energy of the flash, and the impedance offered by appropriate choke coils or other devices. For example, under actual working conditions, with ordinary commercial voltages, effective protection to electrical machinery connected to external conductors may be had with a few choke coils in series with intervening arresters. A good idea of the growth of our knowledge of the nature and behavior of the lightning flash may be obtained from the following publications : Franklin's letters. Experimental Researches. . . .Faraday. Report of the Lightning Rod Conference, 1882. Lodge's "Lightning Conductors and Lightning Guards," 1892. "Lightning and the Electricity of the Air." . . . McAdie and Henry, 1899. r— .„- -- FIQ. 1 EFFECT OF THE ACTION OF LIGHTNING UPON A ROD. That a lightning rod is called upon to carry safely to earth the discharge from a cloud was made plain by Franklin, and the effect of the passage of the current very prettily shown in the melting of the rod and the point (aigrette). Here indeed was a clew to the measurement of the energy of the lightning flash. W. Kohlrausch in 1890 estimated that a normal lightning discharge would melt a copper conductor 5 mm square, with a mean resistance of 0.01 ohm in from .03 to .001 second. Koppe in .1895 from measurements of two nails 4 mm in diameter fused by lightning, determined the current to be about 200 amperes and the voltage about 20,000 volts. The energy of the flash, if the time be considered as 0.1 second, would be about 70,000 horse power, or about 52,240 kilowatts. Statistics show plainly that buildings with conductors when struck by lightning suffer comparatively little damage compared with those not pro- vided with conductors. The same rod, however, cannot be expected to serve equally well for every flash of lightning. There is* great need of a classification of discharges based less upon the appearance of the flash than upon, its electrical energy. Dr. Oliver J. Lodge has made a beginning with 1277 1278 LIGHTNING CONDUCTORS. his study of steady strain and impulsive rush discharges. " The energy of an ordinary flash," says Lodge, " can be accounted for by the discharge of a very small portion of a charged cloud for an area of ten yards square at the height of a mile would give a discharge of over 2,000 foot-tons energy." We must get clearly in our minds then the idea that the cloud, the air, and the earth constitute together a large air condenser, and that when the strain in the dielectric exceeds a tension of ^ gramme weight per square centimeter, there will be a discharge probably of an oscillatory character. And as the electric strain varies, the character of the discharge will vary. Remember too that the air is constantly varying in density, humidity and purity. We should therefore expect to find, and in fact do, every type of discharge from the feeble brush to the sudden and terrific break. Recent experiments indicate that after the breaking-down of the air and the pas- sage of the first spark or flash, subsequent discharges are more easily ac- complished ; and this is why a very brilliant flash of lightning is often followed almost immediately by a number of similar flashes of diminishing brightness. The heated or incandescent air we call lightning, and the lines of fracture of the dielectric can be photographed ; but the electrical waves or oscillations in the ether are extremely rapid, and are beyond the limits of the most rapid shutter and most rapid plate. Dr. Lodge has calculated the rapidity of these oscillations to be several hundred thousand per second. Lodge has also demonstrated experimentally that the secondary or induced electrical surgings in any metallic train cannot be disregarded ; and, as in the case of the Hotel de Ville at Brussels which was most elaborately protected by a network, these surgings may spark at points, and ignite inflammable material close by. While therefore it cannot be said that any known system of rods, wires, or points affords complete and absolute protection, it can be said with con- fidence that we now understand Avhy " spitting-off " and " side " discharges occur ; and furthermore, to quote the words of Lord Kelvin, that " there is a very comfortable degree of security . . . when lightning conductors are made according to the present and orthodox rules." Selection and Installation of Rods. — The old belief that a copper rod an inch in diameter could carry safely any flash of lightning is perhaps true, but we now know that the core of such a rod would have little to do in carrying such a current as a lightning flash, or, for that matter, any high frequency currents. Therefore, since it is a matter of surface area rather than of cubic contents, and a problem of inductance rather than of simple conductivity, tape or cable made of twisted small wires can be used to advantage and at a diminished expense. All barns and exposed buildings should have lightning rods with the neces- sary points and earth connections. Ordinary dwelling-houses in city blocks well built up have less need for lightning conductors. Scattered or isolated houses in the country, and especially if on hillsides, should have rods. All protective trains, including terminals, rods, and earth connections, should be tested occasionally by an experienced electrician, and the total resist- ance of every hundred feet of conductor should not greatly exceed one ohm. Use a good iron or copper conductor. If copper, the conductor should weigh about six ounces per linear foot ; if iron, the weight should be about two pounds per foot. A sheet of copper, a sheet of iron, or a tin roof, if without breaks, and fully connected by well soldered joints, can be utilized to advantage. a b FIG. 2 AND 3 APPROVED CONDUCTORS AND FASTENINGS. PERSONAL SAFETY DURING THUNDER-STORMS. 1279 In a recently published* set of Rules for the Protection of Buildings from Lightning, issued by the Electro-Technical Society of Berlin, Dr. Slaby gives the results of the work of various committees for the past sixteen years studying this question. The lightning conductor is divided into three parts, — the terminal points or collectors, the rod or conductor proper attached to the building, and the earth plates or ground. All projecting metallic sur- faces should be connected with the conductors, which, if made of iron, should have a cross section of not less than 50 mm square (1.9 sq. inches) ; copper, about half of these dimensions, zinc about one and a half, and lead about three times these dimensions. All fastenings must be secure and lasting. The best ground which can be had is none too good for the light- ning conductor. Eor many flashes an ordinary ground will suffice, but there will come occasional flashes when even the small resistance of ^ ohm may count. Bury the earth plates in damp earth or running water. The plates should be of metal at least three feet square. " If the conductor at any part of the course goes near water or gas mains, it is best to connect it to them. Wherever one metal ramification ap- proaches another, connect them metallically. The neighborhood of small bore fusible gas pipes, and indoor gas pipes in general, should be avoided." — Dr. Lodge. «s* IP- ^1 :o C "a 7 ~ 4 " M FIG. 4 CONDUCTORS AND FASTENINGS*. (FROM ANDERSON, AND LIGHTNING ROD CONFERENCE.) The top of the rod and all projecting terminal points should be plated, or otherwise protected from corrosion and rust. Independent grounds are preferable to water and gas mains. Clusters of points or groups of two or three along the ridge rod are good. Chain or linked conductors should not be used. It is not true that the area protected by any one rod has a radius equal to twice the height of the conductor. Buildings are sometimes, for reasons which we understand, damaged within this area. All connections should be of clean well-scraped surfaces properly soldered. A few wrappings of wire around a dirty water or gas pipe does not make a good ground. It is not necessary to insulate the conductor from the building. H. W. Spang gives the following estimate of increase of property destri c- tion by lightning from the "Chronicle Fire Tables." Property loss. $ 8,879,745 11,315,414 21,767,185 ing five years No. of fires ending 1892 2,505 1897 5,637 1902 15,755 * Electrotechnische Zeitschrift, 1901, May 29. 1280 LIGHTNING CONDUCTORS. Much of this increase in property loss is said to be due to the great increase in the use of wire fences in the suburban districts, also to the vastly in- creased use of metal work inside of houses, such as metal lath, steam and water pipes, and all metal trimmings now used so much in exterior trim- mings. Electric wires and their containing tubes also attract lightning; in fact, all the metal work now used in modern building construction serves to attract lightning and convey it to the ground or store it up as in a con- denser, which, upon being released, is liable to cause a spark and thus set fire to adjacent inflammable material. It is said that grounded arresters as now employed in power stations in connection with outdoor overhead electrical conductors also invite light- ning discharges, which, if they take place in the interior of buildings, are liable to cause fire loss; and therefore, it is inadvisable to locate such light- ning arresters adjacent to wood-work or other inflammable material. Large electric signs on the roofs of buildings also serve to attract lightning, and being connected with the interior electrical wires, sometimes jeopardize the safety of the buildings. Electrical wires in the upper stories of our tall buildings are said to become highly electrified during a thunder storm, and lightning from these is liable to impair any underground electrical con- ductor connected therewith. Overhead network wires such as those used for electric light, telephone, telegraph and fire alarm, also attract lightning, and the discharges upon these wires seem to increase in proportion to the number of grounded lightning arresters connected therewith — so much so, that it is now com- mon to dispense with the lightning arresters in fire alarm boxes. ^ Where lead sheathings of underground circuits or conductors of all kinds are metallically connected with the track rails and return circuit of street rail- ways, lightning is also liable to be attracted, and discharges from it in some cases cause considerable damage. It is also said that the grounding of secondary transformer distributing systems at their neutral points has also resulted in lightning discharges to the impairment of lighting transformers. Mr. Spang suggests that rather than connect overhead circuits directly with grounded lightning arresters or to connect return circuits of railways with other metallic networks that are grounded, there should be employed an overhead parallel wire, which shall be thoroughly connected to earth at intervals, and which should preferably be located at the side of any over- head electrical circuit and parallel thereto; but experienced engineers who have made a thorough study of protection from lightning, show that this parallel conductor does not materially benefit the conditions. From the Underwriters' standpoint, therefore, the following rules are suggested as necessary for protection of buildings from lightning: 1. The employment of suitable metallic conductors about the ridges, chimneys or other ordinary elevations above the roof, in connection with all metal work about the roof and also with all exterior and interior metal work, pipes, etc., all metallically connected together so as to provide numer- ous vertical metallic paths from the roof to the cellar and thereby consti- tute with the underground water, gas and other metal pipes, a diffusive system of metallic conductors about the roof and building and over the earth. 2. The shunting of the gas meters by suitable wires or other metal conductors. 3. The employment of two vertical iron or copper conductors along opposite sides of a church spire or a high chimney between a metal cross, weather vane or other suitable air terminal conductor upon the top thereof and the metallic conductors upon the roof, which are metallically con- nected with the underground water, gas and other metal pipes or other suitable ground connection. 4. A system of wires or conductors with suitable air terminals above the roof of a barn, ice-house or storage warehouse and connected by at least four vertical conductors with ground connections distributed over a suitable aroa of adjacent earth, so that the atmospheric electricity will be diffused over a greater and better conducting area than that offered by the com- pactly stored hay, ice, etc. 5. The placing of lightning arresters or other grounded protection de- vires employed with electrical circuits about buildings in iron or non-com- bustible boxes, attached to brick, stone or other non-combustible material or buildings and preferably upon the outside thereof. CHIMNEY PROTECTION. 1281 CHIM.\£1 PROTECTION. The builders of chimneys have made an exhaustive study of lightning action and have developed a number of standard fittings for lightning rods. One form of lightning-rod point is shown in Fig. 5. 2 Copper Cable to Ground L 1 <5\* -» f .„ _ ^ |J H- *'~>j || Fig. 5. Detail of Lightning-rod Point. Usually four of these points are installed at the chimney top, connected together by a band, and having two or more conductors to the earth. The United States Government has investigated thoroughly the require- ments for chimney protection as summarized in the following paragraphs: 1. Chimney Protection for Power Plants. — Lightning con- ductors shall be laid up in the form of a seven-strand cable and each strand laid up with seven copper wires of No. 10 B. and S. gauge. For chimneys of 50 feet and less in height two lightning conductors shall be used. For chimneys over 50 feet up to and including 100 feet, three conductors shall be installed. For chimneys higher than 100 feet, four conductors shall be installed. All heights to be considered from ground level. All conductors or cables shall be symmetrically arranged about the chimney with one cable on the prevailing weather side of the chimney. Said lightning con- ductors or cables to be securely attached both mechanically and electrically to independent pure copper earth plates or bars. In cases where the chim- ney foundations have already been filled in, instead of earth plates, earth terminals may be used, composed of pure copper bars 3 by £ inches by 3 feet. In all cases the lightning conductor terminals shall extend to the ground water level, and in no case shall they extend to less than 15 feet from the ground surface. Earth plates shall consist of pure copper 3 by 3 feet by | inch. 2. Application of Conductors to Chimney. — Each lightning conductor shall be secured to the exterior of the chimney by means of bronze or brass anchors, without the intervention of any insulators or insu- lating material whatever. The brackets for attaching ring or conductors to chimneys to be of high grade bronze or brass, composition of same to be submitted for approval, and to be fitted with approved clamps for securely gripping said conductors and making a good electrical connection there- with. The tongues or shanks of the anchors or brackets shall enter the masonry of the chimney a distance of at least 6 inches, and shall be at least I inch in thickness by 1 inch wide, terminating in a suitable head or angle to prevent the anchor from being pulled out of the masonry. Anchors to be attached to conductors at intervals of not over 10 feet, and M sweated " to the conductors with solder at intervals of 50 feet. Conductors to terminate within 5 feet of the top of the chimney, and to be connected through the agency of a suitable brass or bronze fitting and be soldered to a H D Y h inch ring of copper attached to the periphery of the chimney by brackets spaced not over 2 feet apart. Said brackets to enter the brick work a distance of at least 6 inches and to be of approved design with a tongue at least 1£ inches in width and J inch in thickness, with a suitable angle or head to prevent pulling out. All joints in the continuity of said copper ring, as well as between the continuity of the ring and conductor or conductors running down to the ground bars or plates and including the latter, to be scraped bright and after making a secure mechanical joint to be " sweated with solder." Said solder shall consist of one-half lead and one-half tin. All joints when finished shall be thoroughly washed off with Water to remove every trace of soldering salts, acids, or other compounds 1282 LIGHTNING CONDUCTORS. used. All joints secured by bolts or screws to be lock-nutted. In applying conductors where the chimney is already constructed, holes shall be drilled in the brick and said anchor brackets and anchors grouted in, the best Portland cement being used. 3. Terminal Rods for JLig-li tiling- Conductors. — Copper ring shall be connected through the agency of clamps, insuring a good mechan- ical and electrical joint, with vertically arranged copper rods at least f inch in diameter and 10 feet in length. The joints to be " sweated with solder " as before described. Copper rods to be placed equidistant around this ring, and supported in a rigid position vertically through the agency of additional anchors set in the masonry and a copper spider resting on chimney top as shown in the drawings. Rods to be arranged with a uniform spacing of practically 4 feet. This is taken to mean, for example, that ten such vertical rods shall be provided for a chimney of 12 feet outside diameter of masonry at top. 4. Discliarg-e Points. — Each rod shall terminate in a two-point brass aigrette, each spur or point of this aigrette to be at least 3f inches long, the bases of which spurs shall be at least § inch in diameter, tapering to a sharp and well finished point ; said aigrette to be provided with approved means to secure a strong mechanical and electrical joint with the vertical rods heretofore described and to which it is attached. The joints shall be " sweated with solder " as heretofore described. 5. Chimney Rase Protection. — All lightning conductors shall be enclosed at bottom with a heavy galvanized-iron pipe of 1£ inch diameter, and extending 3 feet into the soil and 10 feet above. Said iron pipe to be provided with approved brackets to securely hold it to the chimney; brackets to be not over 3 feet apart. TE§T§ OF IICJHT.M^O RODS. All lightning rods should be tested for continuity and for resistance of ground plate each year, and the total resistance of the whole conductor and ground plate should never exceed an ohm. TESTS OF LIGHTNING RODS. THIS LEAD MUST BE SOLDFRED TO THE PIPE OR. OTHER EARTH SO AS TO HAVE NO RE8I8TANCE AT THIS JOINT. Fig. 6. Diagram of Connections for Test of Lightning Rods. ISOLATED ELECTRIC PLANTS. 1283 The continuity and resistance of the lightning rods above ground can be measured with a'Wheatstone bridge. The resistance of the ground plate to earth can be determined from three resistance measurements ; from ground plate to each of two other grounds and from one to the other of these arbitrarily chosen grounds, as follows : To make the test, first determine the resistance of the lead wire l x and call it l x . Then connect E x and E 2 as shown in the diagram, call the result R x ; then connect E x and 2£ 3 , call the result R 2 ; connect E 2 and E 3 and call the result R 3 . Now, R x = l x + E x + E 2 and E 2 = R x - l x - E x R 2 = i t + E x + E3 and E 3 = R 2 - h - E x i?3 = E 2 4- E% solving, we have R x + R 2 — i?3 » E x - 2 *■ The resistance of the ground plate to earth is E x as calculated from the above formula. UIRECTIO^ FOR PER§0]¥AI SJLFJ3TY DVRO& IHl^DER §!ORM§. Do not stand under trees or near wire fences ; neither in the doorways of barns, close to cattle, near chimneys or fireplaces. Lightning does not, as a rule, kill. If a person has been struck do not give him up as beyond recovery, even if seemingly dead. Stimulate respiration and circulation as best you can. Keep the body warm ; rub the limbs energetically, give water, wine, or warm coffee. Send for a physician. THE ECOAOMY OW ISOLATED EJLJECTMC 1»JL 4 M TJL*. (By Isaac D. Parsons.) ™ T1 \ e Rowing investigation was undertaken by the writer and Mr. Arnold von bchrenk in an attempt to ascertain which of the two meth- ods is the more economical in six classes of buildings in New York and to *S rn ^ e as nea fly as possible those conditions, either inherent in a class ot buildings or due to peculiarities of installation or management which materially influence the economy. The six classes referred to are: — Office buildings, loft buildings, department stores, apartment houses, hotels, and clubs and over two hundred and fifty buildings were visited in the effort to obtain reliable figures and to ascertain the various condi- tions ot operation. Of this number seventeen only were found where in- formation could be obtained which was reliable in every particular, and only these will be considered in detail, as the great variation in conditions even among similar buildings of the same class renders incomplete statis- tics of very doubtful value. The figures as to electrical output of each of these plants were obtained trom wattmeter readings or from hourly ammeter readings, and were veri- u . by ,P ers , onal observation of the instruments from which they were obtained, and were also checked by comparison with other buildings where similar conditions exist. In some cases, tests were made of the instru- ments to determine their accuracy. The figures recorded as the output ot a plant are in every case the total number of kilowatt hours supplied at the switchboard and used as light or power, and where a storage battery was installed its output only was considered. The expenses of the plants were divided into those of labor, gas, central-station auxiliary or break- down service, coal, water, ash cartage, oil and waste, repairs incandescent lamps and arc-lamp carbons, interest, depreciation, and sundry supplies not included in the foregoing. Figures concerning these items were ob- tained in most cases directly from the booJks of the chief engineer or owner and may be considered within very small limits absolutely accurate. Under the item labor are included the wages of all the engineers, firemen, oilers 1284 LIGHTNING CONDUCTORS. and coal passers employed in the plant, excepting in a few cases where extra employees were required by a large refrigerating machine or similar apparatus. In these cases the wages of the extra men were deducted from the total. If it were determined what employees could be dispensed with were the plant not installed, and the wages of these men only were taken, it would give the true cost of labor chargeable against the plant. To decide this, however, was in most instances a rather uncertain and difficult problem, and it was thought fairer to include in the expenses the wages of all the employees, which, with the other items, give the total cost of running the building with a plant. Then, by adding to the expenses of the central-station service the cost of the labor necessary for heating, elevator supervision, etc., the total cost of running under the conditions of central-station supply can be found. The difference between the two results is the true amount gained or lost by the installation of the plant. The item coal includes that which is burned to generate the steam used for the engines driving the generators, for the feed pumps, and in most cases that used for the house pump and whatever live steam is used in heating the building. In many buildings, refrigerating machines, steam laundries, steam cooking apparatus, or pumps, received steam from the same boilers as the engines driving the dynamos; but in such instances figures from recent tests were available by which the amount of coal used for these purposes could be determined. With the central-station supply either a boiler or a connection with the street mains is required to obtain the steam necessary for heating the building, as well as for the hot-water supply and for running the house pump, unless it is operated electrically. To determine what extra sum must be added to the actual cost of current in order to find the total ex- pense of running the building from the central station, figures were ob- tained from two large loft buildings, an office building, and six apartment houses and hotels using steam for heating and for house pumps only, from which the cost of coal, labor, and water required for these purposes could be calculated. The expenses for coal were reduced to dollars per 1,000 cubic feet heated, and showed practically constant factors, irrespective of the shape or size of the building, of $1.10 per 1,000 cubic feet for apart- ment houses, 90 cents per 1,000 cubic feet for office buildings, and 40 cents per 1,000 cubic feet for loft buildings. The cost of labor, which inc.;/.e. the wages of the firemen and the expense of elevator supervision, ha t > be determined in each particular case, but usually amounts to a sum ab^ut equal to the cost of coal. Interest was calculated in all cases at 5 per cent on the principal in- vested in the plant. Depreciation on dynamos, engines, and switchboards of 5 per cent, and on boilers, pumps, and steam piping of 8 per cent, war. considered liberal; and since, as a rule, the cost of the dynamos, engine:., and switchboards approximates two-thirds of the total cost of installation, and that of the boilers, pumps, and steam piping one-third, a uniform rate of 10/3 + 8/3, or 6, per cent was charged against the whole plant. If 5 per cent of the original capital invested in the engines is set aside each year as a sinking fund, this sum will accumulate interest at 5 per cent, and at the end of fourteen and one half years the total of the amounts reserved, with compound interest, will equal the original cost of the engines; so that 5 per cent depreciation assumes a life of but fourteen and one-half years. Similarly 8 per cent depreciation on boilers assumes a life of ten years. As a matter of fact, both of these periods are much exceeded in first class modern installations. On storage batteries where depreciation is a somewhat doubtful quantity, it was taken as 10 per cent, which assumes a life of but seven years. The load factor in every case was calculated for the hours the plant was in actual operation. As regards load and other conditions of operation, all the buildings can be divided into two classes — those used for business purposes, such as office and loft buildings and stores, and these which are used for residential purposes — such as hotels, apartment houses, and clubs. In the former class a small uniform lighting load during most of the day is succeeded at about 3 p.m. by a heavy load lasting but a few hours, which after 7 p.m. again becomes very small. In the latter class the heavy load, instead of falling off in the evening, continues to 1 or 2 a.m., giving a more uniform load and a higher load factor. We will consider the business buildings first. ISOLATED ELECTRIC PLANTS. 1285 8*- £ d3 ©CM t>. oCN^ "35 d i^hoo<.5 sis ©CN m 3©^ c3© ^fi^X^^S oojhHo _«© /£?'-' O^.S CO «tH i-H pL(© Pu© CO ©Ph co to - © a? ° g^ 6 3£ ^ w ©o O^r o PqCO ©^ (I) ^ o 0) (NhJ J5 CM-2 _.2 ^w=s a^o»o u P©o d -^X^S ou^EH© £ O ©cm yz ©^ ^ •3 ■4J O • e8"2 O CO £3 CO O fl ^.So . £ d CO fl ^ O p-ol-olf-oll 2P*^ fc_ fa!^«n fc- ei CO bO bC— .J2 s c — S ° •£8 8 8*5 CO >3 o > O >» O 03;^ Sj O PQH^O^^H!z;EH^>^m«Wo^ ■eM©O©i>00 000 i CO b © oq i> oo io t> © b g©THCM <^C3SW - © 5 ^H >>r-<CO©©CO©COC > iO > © 1 © >><* of cnT © ^m €© . 00 CM © CO CO © © H cm © «tf © co © © g©^HTtvtf ^ cm" I _J' TH m 0» H "$ CM I> CO rJH in rH ir. § 00 rH © © TF © I>CN ^ CO »o "* *-•«© SO ear 318 174 900 128 420 65 365 B CO >>©© iO cm" i-i^ CM e© . ©oo^tcooo© 2 © CO © T-l LO © "tf g©(N©>©~0%H TjT © ^^T-H CN es .©CO©i-"00©© 03 iC © iO l> CO © O t-i CO ^ © t> S ©*l> co © -.^ © €© L 00©©b-O©© ^ j- cm r^ t^ oo © ic oq gCM©t^CM© rf CO Tt< 3 e© , ©O0^'-i©CO©C © ^©tOiOOOCMOO^N g^NNHHHCOK © © >>CMCM Hr, oo ^e© €© CO • • a SR ^ o p . 3 -M 4 . m e8 c ,SP 'o .s'll crt • u • u > . 98 T3 iS 4 § "8 aste nd sui nd de ation > ■_- •£ -e3«^ Time co Labor Coal an Water S — C Lamps Repairs Interest Central 1 ^ 1286 LIGHTNING CONDUCTORS. OMOi © o o o-^co o o o OOOO^t 4 »0 CD 18 " cd i-I 0» i ■d ¥ Ot^coir s NNCD J ■Ow a C0O00C ■ tRcc 5 ( -V • I>b 3» 1 3 'm M *o «i 3 OiC^cO'- sa ■oc 5 COCOl>i- SB •T^O 1 ■S ©O00 • oqo s m \m o •*• l>cOb . o 1 © COl>OOOI^^ oco^^ • CM CNCOOOCO A co^cx 5 CO ■d 9 HTttioOCO-- oooo©"© *" i-H^CO§| idr-T d — ooc ) o "E o . ^ ,© $ sss > o ) o u CiCOfN^Oi- fa 6 9 ^©HCOO COI>t> ■ ^ * qqqo nc t rt q co ^ ooi>S d gcoV co" CO ^(NcO w 9 e^ e^ m ""* O "* 1> C5 IC Tt Tt< OcOCOtOi-i i>o«c > co | 00l>iq©OiC 2 to s coocc OirH~ O fc* s» €^ 2 e CO tn tj tn . . . sis . . ■ e w OJ3 O u 9 2 d 3 d '-3 1 1 i*if £ d 43+3 4J ®*~ C J. 15 s s 3 a ° o °3 S-l d, a q.^., k . £ o_ o ■♦2 ■*£ "tf «j C 'w fl-° efi Eh 3 3 3 § o-D. a; oJ c OOHJ o^ w ISOLATED ELECTRIC PLANTS. 1287 o © o coPQ . ."3 k J »o M o w o ■ -d •^cnco© ■°9 OiO ooHo^ c3© rH . OS © CO H(H PL, • o CmW^CmK — COiO© © S9B--SIS-S? I S-Htc dTS'S m-- 73 .-2 S «-« 2$aSflgd"S^ © ©^T5-d 3 >v3 *££3 o > o >> O 03-3 o HtOOMOOi 1 lONO^INO _ iOhioNho I>C0tH rH^-l ^cnjW »0©©COCOCOiO CMCOCO i-HtH + COCO i-H iO CO CM CO iO CO iO ^©© §co© ^co CO^ *- © iO ^ CO GO iO © • u, • iCCOCN -J S 3 u 0.3 O I I 1*88 O. Q, Q, .-, co c ■sts-s « d s ^ 3 3 o O O OOOOHPm s M 5 © fa i fa e fa 3 S8 - OOt^CO CO ooo NINrH o 9 8 OCMtJh CirHi-i $11,330 1,600 1,400 o CO CO ■<* Ci© 1-HCOCO csf 60 co' ^OO CO coo ]> o co c "o a "a s- c a C 8 c c3 3 -d *3 *S o ■"S3 2 o 38 ©d cc O ia Ho FOUNDATIONS AND STRUCTURAL. MATERIALS. Revised by W. W. Christie. POWER S1ATIOIV CONSTRUCTION. Chart. By E. P. Roberts & Co. Sta- tion Steam Plant Boilers {Foundation Setting Stack Link V En- gines Water. Fuel Source Pumps and injectors, valves and gauges Heaters Sediment ( Blow off \ Mud drum ( Steam pipe and J valve to heater ] Entrained water, ^Steam ^ separator f Placing in building I Placing in boiler j Removal of coke and ashes (^Removal of soot Air .... Supply to surface > Piping and valves Coverings Drains and drips Supports '"Foundation Steam to cylinder Oil to cylinder Steam from cylinder Water from cylinder Oil to engine Oil from engine Engine indicator f Steam to condenser A*» CEMEIT. Recommendations of Am. Soc. Civil Engineers. Sand. — To be crushed quartz only. To pass, 1st sieve, 400 meshes per square inch. 2d " 900 " " " " Sand to pass the 400 mesh, but be caught by the 900 mesh, all finer parti- cles to be rejected. Portland Cement. — For fineness, to pass, 1st sieve, 2500 meshes per square inch. 2d " 5476 " " " " 3d " 10000 " " " " Should be stored in bulk for at least 21 days to air-slake and free it from lime, as lime swells the bulk, and if not removed is apt to crack the work. I HO\ Al¥l> WTEEJL. Iron, weigrht of: cu. in. Cast, .2604 Lbs. Wrought, .2777 " a =*sectional area wrought-iron bar. x = weight per foot " " " 3x 10a Steel, weigrht of: cu. in. .2831 Lbs. cu. ft. 450 Lbs. 480 " cu. ft. 489.3 Lbs. Cast Iron. Test. Bar an inch square, supported on edges 1 foot apart, must sustain 1 ton at center. WEIGHT OF FLAT ROLLED IRON. 1295 ^OOt-«OlC-^CC»C^'-lOOiOO«OiO'*fCCiC'^CO(>J'-t005GOt^ ^0505C505CS050C50iCOOOOOC»aOC»C»»COt^t^l^t^t^r>;t^ SSoOOOt^SOSOiqiq rlHHHHrt HHHHHMCiO^lO'^C3i-HC500«OuOCO © C5 00 t>; SO UO Tfi CNJ rH © OS 00 bj 1ft "«* CO (N rH © OS t^ CD iq "<* CO t>t^oo"oior^codoor^<>icocoTiiifiot^o6 c 5l>OCOt>OCOt^OCOt-QCOt^OCOt^OCOt ; :OCOt>OCOt>OCOl>QCOI> w»S«irtOoqoiq«H©oooonHO(»»»q?5rto<»»iqcoHqM ^(NS©T*C*© ^iq:oiH©^©^o3©^iqcqT*ost>iqciqcq©» -c4coco*riHooi^i>o6odoTH(^co^^id^t^^o6ciorHr^^^^OCOOOOOC20^^^05?OCO iq i-i t- cqoo^©sqFHt>.cq©^©sDTHt>cq 'rHrH^^CO^T^ldoddt^OOOodciddiH^^CO^^T^ld N^coooocooi>05HP5io^C5H«iocoo^^ooocj^«£ooo«iON wqioq»HOH»^t>.Nt-(NoocoQOcoqTjjq^qiqqqqiqrj®i-i» "THTHCNC$cocdTiiTja6^oc»©©rH^e^coM tMNWw^TjQ6ooci«ooHHNN««^2!12 ^oqci>i>o6od©©©©©^ U0 OS «5lMai©C1CiiO(NOOiCHOO'*O^COQOCOOi!OlN«iOHOO^HN'fOt; cot^0^ooiHiqc5c^»ocoi>rHT^oo«iqosc^oocot'THTijoqc^»q« *iH^rH^^C5COCO^rfrji>dldldOO»l>l>Q6Q6oo"05aOSO*0"C>'-" Hc^coiocooooOHM^owooaoHw^iooooaiOHcc^iooaooQ co^cs^iqoqi-juqXTH^t^oco^oMooc^iqooi-jiqoo^^^Ocq^o "iHr^rtc4c4oicococo^^^idid»didw»^^^^odQoo6csoioio 5(^CO^OCOC^OO^Q»COOlf5iHt^eOC5liSrHt>CO«lOiHt-COOilCTHt»CO jiqt^ocqiqoqceq©ooiHcqoo^rtooiCcqc^iqt>;OC<>iq * th rH i-i th c4 e4 c4 in co co co co »* "^ ^ ^ id id id id so so so t^ t^ t«^ t- oo oo oo t~ mco OrHc<»co^iqsoi>ooooo50'Hc>^coco^iosot>t^ooc^Q'H^coco^iqsot> : c^^sq»oeMTjjsq»S«THcqiqi>«THcqiqt^ocN 'r^rHrHi^TH^^c^c^c^cdcocdcdcdTjJ^^^ididididid ^•^.o^^o^^^gs ^H^«U,»i2 1296 FOUNDATIONS AND STRUCTURAL MATERIALS. 02 co ioSSooqw5ioooo«qooc5oq3qq8 S n i-i r-( £"- l> 00 .2 csoocoi>«oiOTt;iqccooqcq^r-jot>-T^coot> : iocooqr^c^iqo»q'-j«q i^coidt^o6oco*T^"iat^c5^'coT^cdcd^uooo*c^«oo5cc^ ^i-^^t^^cococococococOco^^^oio 00 1.67 3.33 5.00 6.67 8.33 10.00 11.67 13.33 15.00 16.67 18.33 20.00 21.67 23.33 25.00 26.67 30.00 33.33 36.67 40.00 43.33 46.67 50.00 53.33 5 Hco^wt'ddN^idNoddi-'coiood'-i^^deood HrtHHHH^MC^(N(N«lCOM'tit^O .2' »COCOCOOiiOi-it^COOO^OCOCOOQC<>CCiOt>- T-icNTi5idt^c6oiHCOTt*cdt^Q6di^oocdoico"vdt>pcoo ^»N»q^»Nqq'toq«qqio«r;qNLOWHq rico^ws'^oodTHCo'^iddoocsi-icoidoodcodcrJcoio .s 5- l0H?OC;0^t^i-i^QOrixqc^co«qcieo^co©t^u5coc^©co i-ico'^»dd(»cftdc<*cc*^dt^Qo"di-iT^i>dcoidt^deo .2 5 ©Q^i^iHiHi-lC<»C0C?Ot>- coqocqioQOHTij^.qMqqiNoooTfoS^oO'^qq i^ccndi^coTi*iddQO*ddfodooi-Hcoda*i-i HHHHHHHHN(NMt-t-.«>«Dl-OiOlO'*"'^CO co^iqt^csi-icoiqc^cbi-icoict^c^i-iusc^cot^T-iiqaico i-icoco^idt^odosOTHM^iooi^aJi-^codaD^coicoo IIHHHHHHnNNNlNnCOMCO .2* 5 JSS5S5§8^5HS588g8c|S8S8S8e8§8S i-icieo^iddo6<^di-icocc-*d^*o6dcouot^dc4^cd .2 5 0^coeo^»q^t-;OOCftoS^w"^§Soooc5^Soo8 ^c4co^iddt^oo'ddc"coT^iddt>di-H*T^dQodc<*>/j lO ricoeo^»ddt^o6ddi-^cico^id^c»dcoid^d'-"co O Ml H a _JtD >- -J®** J»«ta ►>»_*, a»,<*»i -*e«i- "W^j- lOtS ,-, ,_ ,_, ,-, r-. rl rl n CO « ©«, WEIGHT OF BARS OF IRON. 1297 WEIGHTS Of §m IRE A\» ROMD BAJRS Of WROUGHT IHO.\ I* POVID8 PSA LOEAL FOOT. Iron weighing 480 lbs. per cubic foot. For steel add 2 per cent. © u o3 u u © ?3 u 93 u © 49 u e3 9 B -3 fl © e3 . 2* 5 ao "2 9 © g e3 . « . ©*> u PS C8 O PS o 'X M c! O S 9 X SR C3 O PS o 3 © 02 O MS .d 3^ Q © Si ps*"* .2 So §o MM «•- ,9 O «5M ® _ 9 PS © H © © 3 55 0) 3 © © 15 24.08 18.91 1 96.30 75.64 * .013 .010 1 25.21 19.80 7 TB 98.55 77.40 £ .052 .041 f 26.37 20.71 * 100.8 79.19 f .117 .092 27.55 21.64 T 9 B 103.1 81.00 .208 .164 it 28.76 22.59 1 105.5 82.83 t .326 .256 3 T 30.00 23.56 11 TB 107.8 ' 84.69 .469 .368 t 31.26 24.55 I 110.2 86.56 f .638 .501 32.55 25.57 ii 112.6 88.45 .833 .654 f 33.87 26.60 i 115.1 90.36 & 1.055 .828 35.21 27.65 ii 117.5 92.29 f 1.302 1.023 t 36.58 28.73 6 120.0 94.25 l J 1.576 1.237 37.97 29.82 | 125.1 98.22 1.875 1.473 f 39.39 30.94 130.2 102.3 f 2.201 1.728 40.83 32.07 | 135.5 106.4 2.552 2.004 j% 42.30 33.23 I 140.8 110.6 11 2.930 2.301 1 43.80 34.40 146.3 114.9 1 3.333 2.618 tt 45.33 35.60 i 151.9 119.3 A 3.763 2.955 1 46.88 36.82 I 157.6 123.7 ¥ 4.219 3.313 ii 48.45 38.05 7 163.3 128.3 f 4.701 3.692 1 50.05 39.31 I 169.2 132.9 5.208 4.091 if 51.68 40.59 j 175.2 137.6 T 5 5 5.742 4.510 4 53.33 41.89 181.3 142.4 I 6.302 4.950 * 55.01 43.21 187.5 147.3 & 6.888 5.410 56.72 44.55 1 193.8 152.2 I 7.500 5.890 * 58.45 45.91 f 200.2 157.2 t 8.138 6.392 60.21 47.29 I 206.7 162.4 8.802 6.913 & 61.99 48.69 8 213.3 167.6 ¥ 9.492 7.455 I 63.80 50.11 1 226.9 178.2 10.21 8.018 7 T5 65.64 51.55 J 240.8 189.2 if 10.95 8.601 ¥ 67.50 53.01 | 255.2 200.4 3 11.72 9.204 t 69.39 54.50 9 270.0 212.1 it 12.51 9.828 71.30 56.00 I 285.2 224.0 2 13.33 10.47 T§ 73.24 57.52 h 300.8 236.3 i 14.18 11.14 t 75.21 59.07 1 316.9 248.9 15.05 11.82 f 77.20 60.63 10 333.3 261.8 t 3 b 15.95 12.53 79.22 62.22 | 350.2 275.1 { 16.88 13.25 iS 81.26 63.82 I 367.5 288.6 h 17.83 14.00 83.33 65.45 1 385.2 302.5 I 18.80 14.77 * 85.43 67.10 11 403.3 316.8 t 19.80 15.55 87.55 68.76 1 421.9 331.3 20.83 16.36 t 89.70 70.45 i 440.8 346.2 • V 21.89 17.19 91.88 72.16 i 460.2 361.4 22.97 18.04 ft 94.08 73.89 12 480. 377. 1298 FOUNDATIONS AND STRUCTURAL MATERIALS. - qM»8eotcq«»8rtco8cooqnoq^«oi>wqNMOi>» lew OMOoooeoifltrOMioooONioooocoio ic«qt>oqo^c^^»ocqi>-coo^cocciOQOocovocoocoiftooOMio t>©CO©©CO©©C^O<»^OaO^T^t>Oed©©!^<»idi^lr^edc"cd<>ia6ifti^t^'- *-t» OiMMiOt»050rHMiCt>aOHCOU5t>aiOMt> OOioqi>;cqicio^cqc^rHoooooi>^iciqcqi^ooot>.iococONO00Hrtri;TH<»ioc^^o»oo5cot-ct>i>t-oo^ocio©SS^M^^^SoS «w 0150001001000010010008150080800000000000 3 QSJOOOONOO^MMCIHOCIOOOONOOMHOCSNW l5^©C0©«NWQ0rt^t : 5«iq00fH^N,«©§S5§00^O«0N00W0>lO N05?i^'-OWHeOlO»ONOt»C}H^doOeONNNrt« , c"u"c'T(ia«<»Nl>' V •d» QOOl^iflC^NOCONiOCONOOONiO«)MONCOgN«iOt>« 99^ N .^^ L ^^^ w .^99^ N .^'* ir ?® ?9 r !wSbMowMiflt'Ooo »ONo'Heoiot^05HMWNO(N^ , d(»ddcdttc'Mt>'Hio6T*o6Nddift go £> S H •R oco^Hq»i>»»qMNH5coi>qiomNq^iqNqi>io^ONOMOooia c4^cdQoo^wot>cirHeo>d©woc4T*wo*cot^i^o*o6c4cdo*cot^^ido6c4 INN^!NMMC0C0nc0^^^^^i0i0l0liaCi®Xt*^t»Q000O0>C5OOOH H« 20.00 21.67 23.33 25.00 26.67 28.33 30.00 31.67 33.33 35.00 36.67 38.33 40.00 41.67 43.33 45.00 46.67 48.33 50.00 53.33 56.67 60.00 63.33 66.67 70.00 73.33 76.67 80.00 83.33 86.67 90.00 93.33 96.67 100.00 J® OOC^QOJO©0^t^COOO^OCOC<100COOOt>OOOC-QOQ'--ieOlO«0000 HH(N!N^NW^(MCOC<:cOC<5MCCiCO^'*^^^Ol010« l »©NNI>t'000000 «t» "H HHHHHHHHiNlNINlMiNC^^iN^eOCOWMCOM^^^^iOiOlOlOiO©^ *+* qooqioMHqooqoMHOoocoMHSocooqwqqwqqMqqMq dd^(Nd^dddt^o6dddr'(Nd^ddo6dHdii5do6dHMddad «R 338$83a&8S3g$8S8S88S3g88SS8$8g88S£8SS H« 223S*S!£?228^^22 w ^ wo ^w»ot>aooMt>oeot~©Mc-©e'5t~0£2^© q^»««oqwcjMi>;rjiflq^(»N©qoMHOM»ioeoHqoowiaMHq uj»o»d^©i>i^t>o6o6o50Jooo^^c4c4M^oidot^odo500^c^M^o H^^HHHHHHHHHHHHNfKNfflflfflM •« OHaMP5^ifl?ON(»OOa)OH(NMM'*iONQOONM«5t-OOQ(NMiOt-aiO lONaHcciONftHMio^q^^oooqiNcqiqfflMNHiaq^oqN^Dqiq c4c4c4cococdeocorj*^^*T^idididid»d^o»t>t^i>o6o6o50>oo*o"rM ,d 9 - o a M ^tt^O©t~aO©©THC^2 ° mi &8S3 | ftfl° a^» a as s 3 3 H i o ^ ^22 32 ©.a !§3 0C.J3 & S* ft ► £ * 0000000 1-2 0.5 12.7 320 20. 9.072 97.65 215.28 000000 15-32 0.46875 11.90625 300 18.75 8.505 91.55 201.82 00000 7-16 0.4375 11.1125 280 17.50 7.938 85.44 188.37 0000 13-32 0.40625 10.31875 260 16.25 7.371 79.33 174.91 000 3-8 0.375 9.525 240 15. 6.804 73.24 161.46 00 11-32 0.34375 8.73125 220 13.75 6.237 67.13 148.00 5-16 0.3125 7.9375 200 12.50 5.67 61.03 134.55 1 9-32 0.28125 7.14375 180 11.25 5.103 54.93 121.09 2 17-64 0.265625 6.746875 170 10.625 4.819 51.88 114.37 3 1-4 0.25 6.35 160 10. 4.536 48.82 107.64 4 15-64 0.234375 5.953125 150 9.375 4.252 45.77 100.91 5 7-32 0.21875 5.55625 140 8.75 3.969 42.72 94.18 6 12-64 0.203125 5.159375 130 8.125 3.685 39.67 87.45 7 3-16 0.1875 4.7625 120 7.5 3.402 36.62 80.72 8 11-64 0.171875 4.365625 110 6.875 3.118 33.57 74.00 9 5-32 0.15625 3.96875 100 6.25 2.835 30.52 67.27 10 9-64 0.140625 3.571875 90 5.625 2.552 27.46 60.55 11 1-8 0.125 3.175 80 5. 2.268 24.41 53.82 12 7-64 0.109375 2.778125 70 4.375 1.984 21.36 47.09 13 3-32 0.09375 2.38125 60 3.75 1.701 18.31 40.36 14 5-64 0.078125 1.984375 50 3.125 1.417 15.26 33.64 15 9-128 0.0703125 1.7859375 45 2.8125 1.276 13.73 30.27 16 1-16 0.0625 1.5875 40 2.5 1.134 12.21 26.91 17 9-160 0.05625 1.42875 36 2.25 1.021 10.99 24.22 18 1-20 0.05 1.27 32 2. 0.9072 9.765 21.53 19 7-160 0.04375 1.11125 28 1.75 0.7938 8.544 18.84 20 3-80 0.0375 0.9525 24 1.50 0.6804 7.324 16.15 21 11-320 0.034375 0.873125 22 1.375 0.6237 6.713 14.80 22 1-32 0.03125 0.793750 20 1.25 0.567 6.103 13.46 23 9-320 0.028125 0.714375 18 1.125 0.5103 5.493 12.11 24 1^0 0.025 0.635 16 1. 0.4536 4.882 10.76 25 7-320 0.021875 0.555625 14 0.875 0.3969 4.272 9.42 26 3-160 0.01875 0.47625 12 0.75 0.3402 3.662 8.07 27 11-640 0.0171875 0.4365625 11 0.6875 0.3119 3.357 7.40 28 1-64 0.015625 0.396875 10 0.625 0.2835 3.052 6.73 29 9-640 0.0140625 0.3571875 9 0.5625 0.2551 2.746 6.05 30 1-80 0.0125 0.3175 8 0.5 0.2268 2.441 5.38 31 7-640 0.0109375 0.2778125 7 0.4375 0.1984 2.136 4.71 32 13-1280 0.01015625 0.25796875 6* 0.40625 0.1843 1.983 4.37 33 3-320 0.009375 0.238125 6 0.375 0.1701 1.831 4.04 34 11-1280 0.00859375 0.21828125 5* 0.34375 0.1559 1.678 3.70 35 5-640 0.0078125 0.1984375 5 0.3125 0.1417 1.526 3.36 36 9-1280 0.00703125 0.17859375 ^ 0.28125 0.1276 1.373 3.03 37 17-2560 0.006640625 0.168671875 *i 0.265625 0.1205 1.297 2.87 38 1-160 0.00625 0.15875 4 0.25 0.1134 1.221 2.69 1300 FOUNDATIONS AND STRUCTURAL MATERIALS. (OHTl\^ IMIIiltV OR STRUTS. IXodg-kinson's Formula for C'olnmm, P = crushing weight in pounds ; d = exterior diameter in inches ; d x = interior diameter in inches ; L = length in feet. Kind of Columns. Solid cylindrical col- ) umns of cast iron . j Hollow cylindrical columns of cast Solid cylindrical col- umns of wrought iron Solid square pillar of i Dantzic oak (dry) . j Both ends rounded, the length of the column exceeding 15 times its diameter. Both ends flat, the length of the column exceeding 30 times its diameter. P — 33,380 P — 29,120 P = 95,850 a" 3 - 76 I? P — 98,920 Pz=. 99,320 P — 299,600 tf3-55 ^3-55 (j 3-55 ^3.55 P — 24,540 — , These formulae apply only to cases of breakage caused by bending rather than mere crushing. Where the column is short, or say five times its diam- eter in length, then the following formula applies. Let P = value given in preceding formulae, K=z transverse section of column in square inches, C= ultimate compressive resistance of the material, W= crushing strength of the column. Then W- P CK P + ICK' Hodgkinson's experiments were made upon columns the longest of which for cast iron was 60£ inches, and for wrought iron 90f inches. The following are some of his conclusions : 1. In all long pillars of the same dimensions, when the force is applied in the direction of the axis, the strength of one which has flat ends is about three times as great as one with rounded ends. 2. The strength of a pillar with one end rounded and the other flat is an arithmetical mean between the two given in the preceding case of the same dimensions. 3. The strength of a pillar having both ends firmly fixed is the same as one of half the length with both ends rounded. 4. The strength of a pillar is not increased more than one-seventh by en- larging it at the middle. Gordon's formulae, deduced from Hodgkinson's experiments, are more generally used than Hodgkinson's own. They are : Columns with both ends fixed or flat P = — - 1 + a w Columns with one end flat, the other end round, P = fS 1 -f 1.8a Columns with both ends round or hinged, P = fS l 2J STRENGTH OF MATERIALS. 1301 S = area of cross section in inches ; P =. ultimate resistance of column in pounds ; f — crushing strength of the material in pounds per square inch ; . . , „ Moment of inertia r =. least radius of gyration, in inches, r 2 = -z — : J area of section ' I z= length of column in inches ; a = a coefficient depending upon the material ; / and a are usually taken as constants ; they are really empirical varia- bles, dependent upon the dimensions and character of the column as well as upon the material. (Burr.) For solid wrought-iron columns, values commonly taken are : /r= 36,000 to40,000 ; « = — ^to^. New York City Building Laws 1897-1898 give the following values for/: Cast iron f — 80,000 lbs. Rolled steel .... /= 48,000 lbs. Wrought or rolled iron / = 40,000 lbs. American oak . . . /= 6,000 lbs. Pitch or Georgia pine . / = 5,000 lbs. White pine and spruce /= 3,500 lbs. For solid cast-iron columns, f= 80,000, a =. -t™-. 80 000 For hollow cast-iron columns, fixed ends, p =z v 2 , / = length and 1 + 800^ d — diameter in the same unit, and p = strength in lbs. per square inch. Sir Benjamin Baker gives, For mild steel / = 67,000 lbs., a = 7^ ^j- For strong steel f= 114,000 lbs., a = . STHIXGTH OF MATEKIAIS, The terms stress and strain are generally used synonymously, authorities differing as to which is the proper use. Merriman defines st7 : ess as a force which acts in the interior of a body, and resists the external forces which tend to change its shape. A deformation is the amount of change of shape of a body caused by the stress. The word strain is often used as synony- mous with stress, and sometimes it is also used to designate the deforma- tion. Merriman gives the following general laws for simple tension or compression, as having been established by experiment. a. When a small stress is applied to a body, a small deformation is pro- duced, and on the removal of the stress the body springs back to its original form. For small stresses, then, materials may be regarded as perfectly elastic. b. Under small stresses the deformations are approximately proportional to the forces or stresses which produce them, and also approximately pro- portional to the length of the bar or body. c. When the stress is great enough, a deformation is produced which is partly permanent; that is, the body does not spring back entirely to its original form on removal of the stress. This permanent part is termed a set. In such cases the deformations are not proportional to the stress. d. When the stress is greater still, the deformation rapidly increases, and the body finally ruptures. e. A sudden shock or stress is more injurious than a steady stress, or than a stress gradually applied. 1302 FOUNDATIONS AND STRUCTURAL MATERIALS. £LA§TIC LIMIT. The elastic limit of a material under test for tensile strength is defined as the point where the rate of stretch begins to increase, or where the defor- mations cease to be proportional to the stresses, and the body loses its power to return completely to its former dimensions when the stress is re- moved. Modulus of Elasticity. The modulus or coefficient of elasticity is the term expressing the relation of the amount of extension or compression of a material under stress to the load producing that stress or deformation. It is the load per unit of section divided by the extension per unit of length. If P — applied load, k = sectional area of piece, I = length of the part extended, A. r= amount of extension, M = modulus of elasticity, M- P ' A - Pl k * l~ k\ Following are the Moduli of elasticity for various materials. Brass, cast 9,170,000 " wire 14,230,000 Copper 15,000,000 to 18,000,000. Lead 1,000,000 Tin, cast 4,600,000 Iron, cast 12,000,000 to 27,000,000 (?) Iron, wrought 22,000,000 to 29,000,000 Steel 26,000,000 to 32,000,000 Marble 25,000,000 Slate 14,500,000 Glass 8,000,000 Ash 1,600,000 Beech 1,300,000 Birch 1,250,000 to 1,500,000 Fir 869,000 to 2,191,000 Oak 974,000 to 2,283,000 Teak 2,414,000 Walnut 306,000 Pine, long-leaf (butt-logs) . . 1,119,200 to 3,117,000 Average, 1,926,00 factor of Safety. This may be defined as the factor by which the breaking strength of a material is divided to obtain a safe working-stress. The factor of safety is sometimes a rather indefinite quantity, owing to lack of information as to the strength of materials, and it is now becoming common to name a defi- nite stress which is substantially the result of dividing the average strengths by a factor. The following factors are found in the "Laws Relating to Building in New York City," 1897-1898. For beams, girders, and pieces subject to transverse strains, factor of safety = 4. For wrought-iron or rolled-steel posts, columns, or other vertical sup- ports, 4. For other materials subject to a compressive strain, 5. For tie-rods, tie-beams, and other pieces subject to tensile strain, 6. .no.TiF.vr or i\uiTii. The moment of inertia of a body about any axis, is the sum of the products of the mass of each particle of the body, into the square of its (least) dis- tance from the axis. MOMENT OF INERTIA. 1303 RADIUS Of GTRATIOJ. The radius of gyration of a section is the square root of the quotient of the moment of inertia, divided by the area of the section, or Radius of gyrations / Moment of inertia \ Area of section. The radius of gyration of a solid about an axis is equal to the s Moment of Inertia Mass of the Solid Use in the Formulae for .Strength of Girders and Columns. The strength of sections to resist strains, either as girders or as columns, depends on the form of the section and its area, and the property of the section which forms the basis of the constants used in the formula© for strength of girders and columns to express the effect of the form, is its moment of inertia about its neutral axis. Thus the moment of resistance of any section to transverse bending is its moment of inertia divided by the distance from the neutral axis to the fibers farthest removed from the axis ; or ,, „ . , Moment of inertia _'_ / Moment of resistance = -p-r— t= — ^ ^ -. M = — . Distance of extreme fiber from axis e Moment of Inertia of Compound Shapes. (Pencoyd Iron Works.) The moment of inertia of any section about any axis is equal to the I about a parallel axis passing through its center of gravity -f- (the area of the sec- tion x the square of the distance between the axes). By this rule, the moments of inertia or radii of gyration of any single sec- tions being known, corresponding values may be obtained for any combina- tion of these sections. Radius of Ojration of Compound Shapes. In the case of a pair of any shape without a web the value of R can always be found without considering the moment of inertia. The radius of gyration for any section round an axis parallel to another axis passing through its center of gravity is found as follows : Let r = radius of gyration around axis through center of gravity ; R zzz radius of gyration around another axis parallel to above; d =z distance be- tween axes : R=z V~d?~+r 2 When r is small, R may be taken as equal to d without material error. EIEUTEHTTS OF USUAIj SECTIONS. Moments refer to horizontal axis through center of gravity. This table is intended for convenient application where extreme accuracy is not impor- tant. Some of the terms are only approximate ; those marked * are cor- rect. Values for radius of gyration in flanged beams apply to standard minimum sections only. A = area of section ; b — breadth ; h =z depth ; D =. diameter. 1304 FOUNDATIONS AND STRUCTURAL MATERIALS. Shape of Section. Moment of Inertia. Moment of Resistance. Square of Least Radius of Gyration. Least Radius of Gyration. bh** 12 bh «* 6 /Least\ 2 * V Side J — * f Solid Rect- angle. Least side* 3 46 12 -6-- * i Hollow Rect- angle. bh^—b x h^ * bht—bihS* /i 2 + Z*! 2 * h-\- ht 12 6h 12 4.89 * -6— . f-&-\ Solid Circle. AD** 16 AD* 8 D** 16~ D* 4 k - D- ->j Hollow Circle A, area of large section ; a, area of small section. AD*— ad* AD*— ad* D*+d** 16 D + d 16 8D 5.64 y A — T Solid Triangle. bW 36 bh* 24 The least of the two: h 2 b* 18 ° r 24 The least of the two : h b -b— H or — 4.24 4.9 i Even Angle. Ah* 10.2 Ah 7.2 63 25 b 5 V « -*~J P Uneven Angle Ah* 9.5 Ah 6.5 (hb)* hb 13(h*+b*) 2.6 (h + b) HS Even Cross. Ah* 19 Ah 9.5 h* 22.5 h 4.74 1 ,1 Even Tee. 4h* 11.1 8 b* 22^5 b r-^ 4.74 L~J»- H 4 ^ I-Beam. Ah* 6.66 Ah 3.2 b* 21 b 4.58 Channel. Ah* 7.34 3.67 b* 12.5 i *— h I b 3M "1 te Deck Beam. Ah* 6.9 Ah 4 6 2 36.5 b 6 Distance of base from center of gravity, solid triangle, - ; even angle, — — ; uneven angle, — — ; even tee, ~; deck beam, — ; all other shapes ^, h D given in the table, - or ~k * ELEMENTS OF USUAL SECTIONS. 1305 Solid Cast-iron Columns. Table, based on Hodgkinson's formula (gross tons). The figures are one-tenth of the breaking weight in tons, for solid col- umns, ends flat and fixed. .5 es c> £5 Length of Column m Feet. 6. 8. 10. 12. 14. 16. 18. 20. 25. H .82 .50 .34 .25 .19 .15 .13 .11 .07 ii 1.43 .87 .60 .44 .34 .27 .22 .18 .13 2 2.31 1.41 .97 .71 .55 .44 .36 .30 .20 2£ 3.52 2.16 1.48 1.08 .83 .67 .54 .46 .31 2£ 5.15 3.16 2.16 1.58 1.22 .97 .80 .66 .56 2| 7.26 4.45 3.05 2.23 1.72 1.37 1.12 .94 .64 3 9.93 6.09 4.17 3.06 2.35 1.87 1.53 1.28 .88 3£ 17.29 10.60 7.26 5.32 4.10 3.26 2.67 2.23 1.53 4 27.96 17.15 11.73 8.61 6.62 5.28 4.32 3.61 2.47 ^ 42.73 26.20 17.93 13.15 10.12 8.07 6.60 5.52 3.78 5 62.44 38.29 26.20 19.22 14.79 11.79 9.65 8.06 5.52 5i 88.00 53.97 36.93 27.09 20.84 16.61 13.60 11.37 7.78 6 120.4 73.82 50.51 37.05 28.51 22.72 18.60 15.55 10.64 6* 160.6 98.47 67.38 49.43 38.03 30.31 24.81 20.74 14.19 7 209.7 128.6 87.98 64.53 49.66 39.57 32.30 27.08 18.53 7£ 268.8 164.8 112.8 82.73 63.66 50.73 41.53 34.72 23.76 8 339.1 207.9 142.3 104.4 80.31 64.00 52.39 43.80 29.97 8£ 421.8 258.6 177.0 129.8 99.90 79.61 65.16 54.48 37.28 9 518.2 317.7 217.4 159.5 122.7 97.80 80.05 66.92 45.80 9£ 629.5 3S6.0 264.2 193.8 149.1 118.8 97.25 81.70 55.64 10 757.2 464.3 317.7 233.1 179.3 142.9 117.0 97.79 66.92 10i 902.6 553.5 378.7 277.8 213.8 170.3 139.4 116.6 79.77 11 1067.1 654.4 447.8 328.5 252.7 201.4 164.9 137.8 94.31 n * ! 1252.3 767.9 525.5 385.4 296.6 236.4 193.5 161.7 110.7 12 ; 1459.6 895.1 612.5 449.3 345.7 275.5 225.5 188.5 129.0 sc Where the length is less than 30 diameters, Strength in tons of short columns =r - 7ri 10o -f- 5(7 S being the strength given in the above table, and C= 49 times the sec- tional area of the metal in inches. Hollow Columns. The strength nearly equals the difference between that of two solid col- umns, the diameters of which are equal to the external and internal diam- eters of the hollow one. More recent experiments carried out by the Building Department of New York City on full-size cast-iron columns, and other tests made at the Watertown Arsenal on cast-iron mill columns, show Gordon's formula, based on Hodgkinson's experiments, to give altogether too high results. The following table, from results of the New York Building Department tests, as published in the Engineering News, January 13-20, 1898, show actual results on columns such as are constantly used in buildings. Applying Gordon's formula to the same columns gives the following as the breaking load per square inch. For 15-inch columns, 57,000 lbs.; for 8-inch and 6-inch columns, 40,000 lbs., all of which are much too high, as shown by the table. Prof. Lanza gives the average of 11 columns in the Watertown tests as 29,600 pounds per square inch, and recommends that 5,000 pounds per square inch be used as the maximum safe load for crushing strength. 1306 FOUNDATIONS AND STRUCTURAL MATERIALS. Tests of Cast-iron Columns. Thickness. Breaking Load. Diam. Inches. Max. Min. Average. Pounds. Pounds per sq. in. 1 15 1 1 1,356,000 30,830 2 15 li B * 1 H 1,330,000 27,700 3 15 11 1 11 1,198,000 24,900 4 15| 1 7 1 1,246,000 25,200 5 15 1H 1 T5 1 m 1,632,000 32,100 6 15 1 1 if n 1t 3 6 2,082,000+ 40,400+ 7 7| to 8J § 1 651,00 31,900 8 8 i& i If 3 * 612,800 26,800 9 6& 1* n 1& 400,000 22,700 10 6 5 3 2 It 1 * 1& 455,200 26,300 Ultimate Strength of Hollow. Cylindrical Wrought and Cast-iron Columns, when fixed at the Ends. (Pottsville Iron and Steel Co.) Computed by Gordon's formula, p — - 1+Ci (J)" p =r Ultimate strength in lbs. per square inch ; I = Length of column, ) ^ rt+1% • Bt%mr% „ ni4 . . h = Diameter of column, } both m same umts i _ I 40,000 lbs. for wrought iron; ) 80,000 lbs. for cast iron; j C — 1/3000 for wrought iron, and 1/800 for cast iron. „ 80.000 For cast iron, p : / 1+ 800 V h) For wrought iron, p = - 40,000 l+— (-Y ^ 3,000 V hj If ollow Cylindrical Columns. Ratio of Maximum Load per sq. in. Safe Load per Square Inch. Length to Diameter. 1 h Cast Iron. Wrought Iron. Cast Iron, Factor of 6. Wrought Iron, Factor of 4. 8 74075 39164 12346 9791 10 71110 38710 11851 9677 12 67796 38168 11299 9542 14 64256 37546 10709 9386 16 60606 36854 10101 9213 18 56938 36100 9489 9025 20 53332 35294 8889 8823 22 49845 34442 8307 8610 24 46510 33556 7751 8389 26 43360 32642 7226 8161 28 40404 31712 6734 7928 30 37646 30768 6274 7692 ELEMENTS OF USUAL SECTIONS. 1307 Hollow Cylindrical Columns. — Continued. Ratio of Length to Maximum Load per Sq. In. Safe Load per Square Inch. Diameter. 1 h Cast Iron. Wrought Iron. Cast Iron, Factor of 6. Wrought Iron, Factor of 4. 32 35088 29820 5848 7455 34 32718 28874 5453 7218 36 30584 27932 5097 6983 38 28520 27002 4753 6750 40 26666 26086 4444 6522 42 24962 25188 4160 6297 44 23396 24310 3899 6077 46 21946 23454 3658 5863 48 20618 22620 3436 5655 50 19392 21818 3262 5454 52 18282 21036 3047 5259 54 17222 20284 2870 5071 56 16260 19556 2710 4889 58 15368 18856 2561 4714 60 14544 18180 2424 4545 Ultimate Strength of Wrous-ht-iron Columns. p = ultimate strength per square inch; Z= length of column in inches; r = least radius of gyration in inches. 40000 For square end-bearings, p = - For one pin and one square bearing, p = - ^40000Vr/ 40000 1 + 000\rj For two pin bearings, 30000 40000 — 1+ i fi\' ^20000 \r) For safe working-load on these columns use a factor of 4 when used in buildings, or when subjected to dead load only; but when used in bridges the factor should be 5. Wr ought-Iron Columns. Ultimate Strength in Lbs. Safe Strength in Lbs. per 1 per Square Inch. I r Square Inch — Factor of 5. r Square Pin and Pin Square Pin and Pin Ends. 39944 Sq. End. Ends. Ends. Sq.End. Ends. 10 39866 39800 10 7989 7973 7960 15 39776 39702 39554 15 7955 7940 7911 20 39604 39472 39214 20 7921 7894 7843 25 39384 39182 38788 25 7877 7836 7758 30 39118 38834 38278 30 7821 7767 7656 35 38810 38430 37690 35 7762 7686 7538 40 38460 37974 37036 40 7692 7595 7407 45 38072 37470 36322 45 7614 7494 7264 50 37646 36928 35525 50 7529 7386 7105 55 37186 36336 34744 55 7437 7267 6949 60 36697 35714 33898 60 7339 7148 6780 65 36182 34478 33024 65 7236 6896 6605 70 35634 34384 32128 70 7127 6877 6426 75 35076 33682 31218 75 7015 6736 6244 80 34482 32966 30288 80 6896 6593 6058 85 33883 32236 29384 85 6777 6447 5877 90 33264 31496 28470 90 6653 6299 5694 95 32636 30750 27562 95 6527 6150 5512 100 32000 30U00 26666 100 6400 6000 5333 105 31357 29250 25786 105 6271 5850 5157 1308 FOUNDATIONS AND STRUCTURAL MATERIALS. TRA^S¥£R§E STBEIGTH. Transverse strength of bars of rectangular section is found to vary di- rectly as the breadth of the specimen tested, as the square of its depth, and inversely as its length. The deflection under load varies as the cube of the length, and inversely as the breadth and as the cube of the depth. Alge- braically, if S := the strength and D the deflection, I the length, b the breadth, and d the depth, S varies as -7- and D varies as =-=5. I bd 3 To reduce the strength of pieces of various sizes to a common standard, the term modulus of rupture (B) is used. Its value is obtained by experi- ment on a bar of rectangular section supported at the ends and loaded in the middle, and substituting numerical values in the following formula : A - 2 bd* in which P = the breaking load in pounds, I = the length in inches, b the breadth, and d the depth. fundamental Formula* for flexure of Beams. (Merriman.) Resisting shear *± vertical shear ; Resisting moment = bending moment ; Sum of tensile stresses = sum of compressive stresses ; Resisting shear = algebraic sum of all the vertical components of the in- ternal stresses at any section of the beam. If A be the area of the section and S% the shearing unit stress, then resist- ing shear = ASs ; and if the vertical shear = V, then V=. ASa. The vertical shear is the algebraic sum of all the external vertical forces on one side of the section considered. It is equal to the reaction of one sup- port, considered as a force acting upward, minus the sum of all the vertical downward forces acting between the support and the section. The resisting moment =, algebraic sum of all the moments of the inter- nal horizontal stresses at any section with reference to a point in that sec- ts'/ tion, = — , in which S= the horizontal unit stress, tensile or compressive as the case may be, upon the fiber most remote from the neutral axis, c = the shortest distance from that fiber to said axis, and I— the moment of inertia of the cross-section with reference to that axis. The bending moment Mis the algebraic sum of the moment of the external forces on one side of the section with reference to a point in that section = moment of the reaction of one support minus sum of moments of loads be- tween the support and the section considered. M= — . c The bending moment is a compound quantity = product of a force by the distance of its point of application from the section considered, the distance being measured on a line drawn from the section perpendicular to the direc- tion of the action of the force. Concerning the above formula, Prof. Merriman, Eng. News, July 21, 1894, says : The formula just quoted is true when the unit-stress S on the part of the beam farthest from the neutral axis is within the elastic limit of the material. It is not true when this limit is exceeded, because then the neutral axis does not pass through the center of gravity of the cross section, and because also the different longitudinal stresses are not proportional to their distances from that axis, these two requirements being involved in the de- duction of the formula. But in all cases of design the permissible unit- stresses should not exceed the elastic limit, and hence the formula applies rationally, without regarding the ultimate strength of the material or any of the circumstances regarding rupture. Indeed, so great reliance is placed upon this formula that the practice of testing beams by rupture has been almost entirely abandoned, and the allowable unit-stresses are mainly de- rived from tensile and compressive tests, TRANSVERSE STRENGTH. 1309 C g i 8 a a 2 s i e = 3 « IS * ^'^ o © ^ © ft - !«LI«~ 1?" !" + «,|S MIMI I».r8 o jfe-ij o © HS m <*-( >* , ° ' ^ ^l^l^hSl^hSl^l^l^l^ Sfl« O eg c II II II II II II II II II ii © go d S 02 S © a? e« 002 ^ 5 H|«rt|^HI00 _j- i-HOOiH |«D_ |N icb E T-l 100 SS ' * co ioo 2* La CO s © o o 52 ft. SB*"* CO l ©73 q j PQ © o ft. o3 AH ■r-i \*r 3 &) •d 'el *e 'el 'el 'el ■ei « i % *e S o3 ,c h* "** h* h 4 * * ,J ' c r"tt -o 4a -o eg O ^| o;| t$\ f$\ e$\ c$| |p§ ^1 © iH I«OtH ICOC* ICO^ ICO"* ICO* ico^s i wL* Rectangle. 32(AD 2 —ad 2 ) 52(AD 2 —ad 2 ) Solid Cylinder. 700AD L 14MAD L 2AAD 2 38AD 2 Hollow 700 (AD — ad) L l4O0(AD—ad) L wL z wL* Cylinder. 24(AD 2 —ad 2 ) 38(AD 2 —ad 2 ) APPROXIMATE GREATEST SAFE LOAD IN LBS. 1311 Shape Greatest Safe Load in Lbs. Deflection in Inches. of Section. Load in Middle. Load Distributed. Load in Middle. Load Distributed. Even- legged Angle or Tee. 930 AD L 1860 A D L wL 3 VIAD* Channel or ZBar. 160CL4Z) L 3200JZ) L 53AD 2 wL 3 85AD 2 Deck Beam. UfyOAD L 2900.4/) L wL 5 50AD* wL 3 WAD* I-Beam. 17S0AD L ZoWAD L wL 3 58AB 2 wL 3 93^Z) 2 I II III IV V The rules for rectangular and circular sections are correct, while those for the flanged sections are approximate, and limited in their application to the standard shapes as given in the Pencoyd tables. The calculated safe loads will be approximately one-half of loads that would injure the elasticity of the materials. The rules for deflection apply to any load below the elastic limit, or less than double the greatest safe load by the rules. If the beams are long, without lateral support, reduce the loads for the ratios of width to span as follows : Length of Beam. Proportion of Calculated Load forming Greatest Safe Load. 20 times flange width. 30 40 50 60 70 Whole calculated load. 9-10 " " 8-10 " " 7-10 6-10 " " 5-10 " These rules apply to beams supported at each end. For beams supported otherwise, alter the coefficients of the table as described below, referring to the respective columns indicated by number. Changes of Coefficients for Special Form* of Beams. Kind of Beam. Fixed at one end, loaded at the other. Coefficient for Safe Load. One-fourth of the coeffi- cient of col. II. Coefficient for Deflec- tion. One-sixteenth of the co- efficient of col. IV. 1312 FOUNDATIONS AND STRUCTURAL MATERIALS. Changes of Coefficients — Continued. Kind of Beam. Coefficient for Safe Load. Coefficient of Deflec- tion. Fixed at one end, load evenly distributed. One-fourth of the coeffi- cient of col. III. Five forty-eighths of the coefficient of col. V. Both ends rigidly fixed, or a continuous beam, with a load in middle. Twice the coefficient of col. II. Four times the coeffi- cient of col. IV. Both ends rigidly fixed, or a continuous beam, with load evenly dis- tributed. One and a half times the coefficient of col. III. Five times the coeffi- cient of col. V. Let modulus of Elasticity and Elastic Resilience. P — tensile stress in pounds per square inch at the elastic limit ; e — elongation per unit of length at the elastic unit ; E = modulus of elasticity = P -f- e ; e = P -=- E. j p2 Then elasticity resilience per cubic inch = \Pe = 2 E' BEAMS OF MIFORM §TR£^OTH THROUGHOUT THEIR LENGTH. The section is supposed in all cases to be rectangular throughout. The beams shown in plan are of uniform depth throughout. Those shown in elevation are of uniform breadth throughout. B = breadth of beam. D = depth of beam. Fixed at one end, loaded at the other ; curve parabola, vertex at loaded end ; BD 2 p proportional to distance from loaded end. % The beam may be reversed so that the up- per edge is parabolic, or both edges may be parabolic. Fixed at one end, loaded at the other ; tri- angle, apex at loaded end ; BD 2 proportional to the distance from the loaded end. Fixed at one end ; load distributed ; tri- S angle, apex at unsupported end ; BD 2 pro- p portional to square of distance from unsup- 0, ported end. Fixed at one end ; load distributed ; curves two parabolas, vertices touching each other, at unsupported end ; BI) 2 proportional to dis- tance from unsupported end. Supported at both ends ; load at any one point ; two parabolas, vertices at the points of support, bases at point loaded ; BD 2 pro- portional to distance from nearest point of support. The upper edge or both edges may also be parabolic. Supported at both ends ; load at anv one point ; two triangles, apices at points of sup- port, bases at point loaded ; BD 2 propor- tional to distance from the nearest point of support. Supported at both ends ; load distributed ; curves two parabolas, vertices at the middle of the beam ; bases center line of beam ; BI) 2 proportional to product of distances from points of support. Supported at both ends ; load distributed ; _ curve semi-ellipge ; BD 2 proportional to the \ product of the distances from the points of support. TRENTON BEAMS AND CHANNELS. 1313 TRENTON BE4WS AND (HlWELIi. (Trenton Iron Works.) v To find which beam, supported at both ends, will be required to support with safety a given uniformly distributed load : Multiply the load in pounds by the span in feet, and take the beam whose " Coefficient for Strength " is nearest to and exceeds the number so found. The weight of the beam itself should be included in the load. The deflection in inches for such distributed load will be found by divid- ing the square of the span taken in feet, by seventy (70) times the depth of the beam taken in inches for iron beams, and by 52.5 times the depth for steel. Example. — Which beam will be required to support a uniformly distrib- uted load of 12 tons (= 24,000 lbs.) on a span of 15 feet ? 24,000 X 15= 360,000, which is less than the coefficient of the 12£-mch 125- lb. iron beam. The weight of the beam itself would be 625 lbs., which, added to the load and multiplied by the span, would still give a product less than the coefficient; thus, 24,625 X 15=369,375. The deflection will be : 15 x 15 70 X 12£ S c= 0.26 inch. The safe distributed load for each beam can be found by dividing the coefficient by the span in feet, and subtracting the weight of the beam. When the load is concentrated entirely at the center of the span, one-half of this amount must be taken. The beams must be secured against yielding sideways, or the safe loads will be much less. Tit EX TO X ROILED S1EEL BEAMS. Weight per Width of Flanges in Inches. Coefficient for Designation of Beam. Yard in Lbs. Thickness of Stem. Strength i n Lbs., Minimum Min. Max. Weight. 15 inch 150 190 5.75 .45 753,000 15 ' 123 160 5.5 .40 603,000 12 ■ 120 150 5.5 .39 500,000 12 « 96 125 5.25 .32 407,000 10 • 135 160 5.25 .45 461,000 10 « 99 125 5.0 .37 344,000 10 ' 76 100 4.75 .32 264,000 9 ' 81 105 4.75 .31 262,000 9 4 63 85 4.5 .27- 200,000 8 ' 66 85 4.5 .27 192,000 8 « 54 75 4.25 .25 154,000 7 4 60 80 4.25 .27 151,000 7 ' 46.5 65 4.0 .23 118,000 6 ' 50 65 3.5 .30 104,000 6 ' 40 55 3.0 .25 83,300 5 ' 39 52 3.13 .26 67,000 5 ' 30 42 3.0 .22 52,900 4 ■ 30 40 2.75 .24 41,200 4 l 22.5 32 2.62 .20 31,400 2 ' *i .75 i 2,660 U " 5i 1.50 s 2,300 1314 FOUNDATIONS AND STRUCTURAL MATERIALS. THE\TO\ IROI BEAMS A YI> CHi^YELS. bo H O.J3 . to _ CC M^

i> : c^«0'<^iot>iooioooc^'^o<»c^ 73 +a © © xce 8SS8gS8gggS88iSa8a8S © o .2 §5 rtOHO^CCOlOCONOlOJlOOOfflHOMHCCWNH o ft © ?s ©1-1 Pi fl £ v. © «e 05 cq oq c• GO CO C5 OS C5 O OOM CO CS OJ a fl Oh Ph WOOD. Tests of American Woods. 1317 In all cases a large number of tests were made of each wood. Minimum and maximum results only are given. All of the test specimens had a sec- tional area of 1.575 x 1.575 inches. The transverse test specimens were 39.37 inches between supports, and the compressive test specimens were 3 PI 12.60 inches long. Modulus of rupture calculated from formula R = - t-^-\ P = load in pounds at the middle, I = length in inches, b =z breadth, d = depth : Name of Wood. Transverse Compression Tests, Parallel to Modulus of Grain, pounds Rupture. per sq. in. Min. Max. Min. Max. 7440 12050 4560 7410 6560 11756 4150 5790 6720 11530 3810 6480 9680 20130 7460 9940 8610 13450 6010 7500 12200 21730 8330 11940 8310 16800 5830 9120 7470 11130 5630 7620 10190 14560 6250 9400 9830 14300 6240 7480 18500 10290 6650 8080 5950 15800 4520 8830 5180 10150 4050 5970 10220 13952 6980 8790 8250 15070 4960 8040 6720 11360 4960 7340 4700 11740 5480 6810 8400 16320 6940 8850 14870 20710 7650 10280 11560 19430 7460 8470 7010 18360 5810 9070 9760 18370 4960 8970 7900 18420 4540 8550 5950 12870 3680 6650 13850 18840 5770 7840 11710 17610 5770 8590 8390 13430 3790 6510 6310 9530 2660 5810 5640 15100 4400 7040 9530 10030 5060 7140 5610 11530 3750 5600 3780 10980 2580 4680 9220 21060 4010 10600 9900 11650 4150 5300 7590 14680 4500 7420 8220 17920 4880 9800 10080 16770 6810 10700 % Cucumber tree Yellow poplar, white wood . . White wood, Basswood . . . Sugar maple, Rock maple . . Red maple Locust Wild cherry Sweet gum Dogwood Sour gum, pepperidge .... Persimmon White ash Sassafras Slippery elm White elm ........ Sycamore, Buttonwood . . . Butternut, white walnut . . . Black walnut Shellbark hickory Pignut White oak Red oak Black oak Chestnut Beech Canoe birch, paper birch . . . Cottonwood White cedar Red cedar Cypress White pine Spruce pine Long-leaved pine, Southern pine White spruce Hemlock Red fir, yellow fir Tamarack 1318 FOUNDATIONS AND STRUCTURAL MATERIALS. iooqoo©io^^co©cb-t-«©©o©ioioioio lO t> lO CO CO c^oeDioic^Tij^Tjj^i?5eococococococi ^c^r-5r-H»Hr4oooodoooooooooooooooo 00 Tf CNS<1 i0^i0^iOCO^COO^^©eoe3£^iHC*©i-HOO«OlGTt r-H 00 «0 IO tJJ T* CO CO CO CN (T>L^H*^COCOOT^1C^~ir)-*rH^,^rHrHrHrH Ol -*©©©©©©©©©©©©©©©©©©,©© ©©©©©*©©©©©©©©'©©©'©©©©©© ©VOC0t>T*©© . . ©<©©©©©©©©© © # ©©©©©©©©© ©©©©*©©©©"©"©©©©©©©©©©© " " at a^ds H^CO^UO©t^OO©©^WCO^UO©^00©0-^CO^^^^gO©g wood. 1319 Rule. — To find the safe uniformly distributed load in tons for white pine or spruce beams, multiply the number given in the above table by the thick- ness of the beam in inches. For beams of other wood, multiply also by the following numbers : White Oak. Hemlock. White Cedar. Yellow Pine. Chestnut. 1.45 .99 .60 1.50 1.08 Formulae for White Pine Beams. Subject to vibration from live loads. w = safe load in pounds, less weight of beam. I = length of beam in inches. d = depth of beam in inches. b = breadth of beam in inches. For a beam fixed at one end and loaded at the other: 1000 6d 2 w =— a— For a beam fixed at one end and uniformly loaded : 1000 bd* For a beam supported at both ends and loaded at the middle : 2000 bd* For a beam supported at both ends and uniformly loaded : 4000 bd 2 W =—3l- Note. — In placing very heavy loads upon short, but deep and strong beams, care should be taken that the beams rest for a sufficient distance on their supports to prevent all danger from crushing or shearing at the ends. Ordinary timbers crush under 6,000 lbs. per square inch. To assure a safety of beam against crushing at the end, divide half of the load by 1000 ; the quotient will be the least number of square inches of base that should be allowed for each end to rest on. Table of Safe Load for moderately Seasoned White Pine Struts or Pillars. The following table, exhibiting the approximate strength of white pine struts or pillars, with flat ends, is outlined and interpolated from the rule of Rondolet, that the safe load upon a cube of the material being regarded as unity, the safe load upon a post whose height is, 12 times the side will be f 36 48 60 72 ft 700 pounds per square inch is assumed as the safe load upon a cube of white pine. The strength of each strut is considered with reference to the first-named dimension of its cross section, so that if the second dimension is less than the first, the strut must be supported in that direction, to fulfill the condi- tions of the computation. The strength of pillars, as well as of beams of timber, depends much on their degree of seasoning. Hodgkinson found that perfectly seasoned blocks 2 diameters long, required in many cases twice as great a load to crush them as when only moderately dry. This should be borne in mind when building with green timber. 1320 FOUNDATIONS AND STRUCTURAL MATERIALS. I. Safe Distributed Load§ upon Southern Pine Ream* One Inch in Width. (C. J. H. Woodbury.) (If the load is concentrated at the center of the span, the beams will sus- tain half the amount as given in the table.) 9 Depth of Beam in Inches. & 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 0Q Load in Pounds per Foot of Span. 5 38 86 154 240 346 470 614 778 960 6 27 60 107 167 240 327 427 540 667 807 7 20 44 78 122 176 240 314 397 490 593 705 828 8 15 34 60 94 135 184 240 304 375 454 540 634 735 9 27 47 74 107 145 190 240 296 359 427 501 581 667 759 10 22 38 60 86 118 154 194 240 290 346 406 470 540 614 11 32 50 71 97 127 161 198 240 286 335 389 446 508 12 27 42 60 82 107 135 167 202 240 282 327 375 474 13 36 51 70 90 115 142 172 205 240 278 320 364 14 31 44 60 78 99 123 148 176 207 240 276 314 15 27 38 52 68 86 107 129 154 180 209 240 273 16 34 46 60 76 94 113 135 158 184 211 240 17 30 41 53 67 83 101 120 140 163 187 217 18 36 47 60 74 90 107 125 145 167 190 19 43 54 66 80 96 112 130 150 170 20 38 49 60 73 86 101 118 135 154 21 44 54 66 78 92 107 122 139 22 50 60 71 84 97 112 127 23 45 55 65 77 89 102 116 24 50 60 70 82 94 107 25 46 55 65 75 86 98 Distributed Loads upon Southern Pine Beams Suf- ficient to Produce Standard Limit of Deflection. (C. J. H. Woodbury.) *3 4> Depth of Beam in Inches. o . a 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 v3 OQ in Load in Pounds per Foot of Span. ®5 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 3 2 10 7 5 4 23 16 12 9 7 6 44 31 23 17 14 11 9 77 53 39 30 24 19 16 13 11 1 .'•' 122 85 62 48 38 30 25 21 18 16 14 182 126 93 71 56 46 38 32 27 23 20 18 16 259 180 132 101 80 65 54 45 38 33 29 25 22 20 18 247 181 139 110 89 73 62 53 45 40 35 31 27 25 22 20 241 185 146 118 98 82 70 60 53 46 41 37 33 30 27 24 22 240 190 154 127 107 91 78 68 60 53 47 43 38 35 32 29 27 25 305 241 195 161 136 116 100 87 76 68 60 54 49 44 40 37 34 31 301 244 202 169 144 124 108 95 84 75 68 61 55 50 46 42 | 39 300 248 208 178 153 133 117 104 93 83 75 68 62 57 52 48 301 253 215 186 162 147 126 112 101 91 83 75 69 63 58 .0300 .0432 .0588 .0768 .0972 .1200 .1452 .1728 .2028 .2352 .2700 .3072 .3468 .3888 .4332 .4800 .5292 .5808 .6348 .6912 .7500 MASONRY. 1321 UfASOtflOf. Brick-Work. Brick work is generally measured by 1000 bricks laid in the wall. In con- sequence of variations in size of bricks, no rule for volume of laid brick can be exact. The following scale is, however, a fair average. 7 common bricks to a super 14 21 28 " " " 35 ft. 4-inch wall. 9-inch " 13-inch " 18-inch " 22-inch " Corners are not measured twice, as in stone-work. Openings over 2 feet square are deducted. Arches are counted from the spring. Fancy work counted 1£ bricks for 1. Pillars are measured on their face only. One thousand bricks, closely stacked, occupy about 56 cubic feet. One thousand old bricks, cleaned and loosely stacked, occupy about 72 cu- bic feet. One cubic foot of foundation, with one-fourth inch joints, contains 21 bricks. In some localities 24 bricks are counted as equal to a cubic foot. One superficial foot of gauged arches requires 10 bricks. Stock bricks commonly measure 8| inches by 4± inches by 2| inches, and weigh from 5 to 6 lbs. each. Paving bricks should measure 9 inches by 4A- inches by If inches, and weigh about 4^ lbs. each. One yard of paving requires 36 stock bricks, of above dimensions, laid flat, or 52 on edge; and 35 paving bricks, laid flat, or 82 on edge. The following table gives the usual dimensions of the bricks of some of the principal makers. Description. Inches. Description. Inches. Baltimore front . Philadelphia front Wilmington front Trenton front Croton .... Colabaugh . . . j, 8i X 4| X 2f 8£ X 4 X 2i 8J X 3f X 2f Maine .... Milwaukee . . North River . Trenton . . . Ordinary . . . n X 3f X 2f 8} X 4J X 2| 8 X 3* X 2i 8 X 4 X 2\ f 7f X 3f X 2i (8 X4ix 2J Fire Brick - J Valentine's (Woodbridge, N. J.) . . $ ( Downing's (Allentown, Pa.) .... 9 X 4| X 2£ inches X 4| X 2A- inches To compute the number of bricks in a square foot of wall. — To the face dimensions of the bricks used, add the thickness of one joint of mortar, and multiply these together to obtain the area. Divide 144 square inches by this area, and multiply by the number of times which the dimension of the brick, at right angles to its face, is contained in the thickness of the wall. Example. — How many Trenton bricks in a square foot of 12-inch wall, the joints being \ inch thick ? 8~+T x 2\-\-\ — 20.62 ; 144 -f 20.62 = 7 ; 7 X 3 — 21 bricks per square ft. 1322 FOUNDATIONS AND STRUCTURAL MATERIALS. Weight and Bulk of Bricks. Number of Bricks, by itself. in wall with cement. Gross Tons. Pounds. Cu. ft. C. Brick. F. Brick. C. Brick. F. Brick. 1 2240 22.4 448 416.6 381 347 0.04464 100 1 20 18.6 17 15£ 2.23 5000 50.00 1000 930 850 772 2.4 5376 53.76 1075 1000 914 834 2.62 5872 58.72 1130 1100 1000 913 2.88 6451 64.51 1240 1200 1100 1000 One perch of stone is 24.75 cubic feet. In New York City laws a cubic foot of brick-work is deemed to weigh 115 lbs. Building-stone is deemed to weigh 160 lbs. per cubic foot. The safe load for brick-work according to the New York City Laws is as folio ws : — In tons per superficial foot, For good lime mortar 8 tons. For good lime and cement mortar mixed . ll£ tons. For good cement mortar 15 tons. Average Ultimate -c5'2 t>. iq oo © c >oooo oo< ssssiasssasssssss&ssss r/S^ri2?5^^'^ <35C< 5coQQ«oooocp"^'foo"^.^'ft^ »^lO^OS«OOOI>;t>;CS^iOOTeOQO^OI>'*THC3St^»OCO 2oi^wwc4i^oso6t^«5^i0^^wcocceNc>5c4^*»HiHrH SSSig <»lClO^Tt(COCOCNCNCNCN»-iTHr-l c © C 3 "©3 i-bt>*0«XiftrtOt-OTfiMNC1rlrtri ^^cqu5Mcjw^N^ooS^o5i>§^coooN«C56 ^doorjojcoooitoooicoc" HCOo^QoaiOH(Nn^iO(ot'«)ao MISCELLANEOUS MATERIAL. 1325 Galvanized Iron IFire Rope. Charcoal Rope. For Ship's Rigging and Guys for Derricks. 5 M ®a3^ • w S3 43 S<^ 5^ OCC «h « 11 10* 10 9* 9 81 8 n eh 6 5f 5i 43 40 35 33 30 26 23 20 16 14 12 10 .5ft§i o o <*£ 3 tt) 5^ OOQ 5 I 3f 3 2* 2| 2§ If l| 11 1| Transmission or Haulage Rope. (Roeblingr.) Composed of H Strands and a Hemp Center, 7 Wires to the Strand. SWEDISH IRON. Approxi- mate Cir- cumfer- ence in Inches. Approxi- Allowable Mini- Trade Number. Diameter in Inches. Weight per Foot in Pounds. mate Breaking Strain in Tons of Working Strain in Tons of 2,000 mum Size of Drum or Sheave 2,000 Lbs. Pounds. in Feet. 11 u 4| 3.55 34 6.80 13 12 11 3 3.00 29 5.80 12 13 H 4 2.45 24 4.80 lOf 14 H 3* 2.00 20 4.00 9* 15 l 3 1.58 16 3.20 4 16 i 2| 1.20 12 2.40 7 4 17 | 2i 0.89 9.3 1.86 18 11 2i 0.75 7.9 1.58 6 19 i 2 0.62 6.6 1.32 5i 20 & If 0.50 5.3 1.06 4 21 h 1J 0.39 4.2 0.84 4 22 4 l| 0.30 3.3 0.66 3£ 23 1 ll 0.22 2.4 0.48 2| 24 T S 5 1 0.15 1.7 0.34 3 25 & i 0.125 1.4 0.28 2i CAST STEEL. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 if if i i 4| 4i 4 3* 3 3.55 3.00 2.45 2.00 1.58 1.20 0.89 0.75 0.62 0.50 0.39 0.30 0.22 0.15 0.125 68 58 48 40 32 24 18.6 15.8 13.2 10.6 8.4 6.6 4.8 3.4 2.8 13.6 11.6 9.60 8.00 6.40 4.80 3.72 3.16 2.64 2.12 1.68 1.32 0.96 0.68 0.56 1326 FOUNDATIONS AND STRUCTURAL MATERIALS. Standard Hoisting* Rope. Composed of 6 Strands and a Hemp Center, 19 Wires to the Strand, SWEDISH IRON. B O CO cumfeiv nches. Weight Ap. Breaking Allowable Working Strain in Min. Size of eg u H §•3 3 per Foot in Lbs. in Tons of 2,000 Lbs. Tons of 2,000 Lbs. Drum or Sheave in Foot. 2| 8§ 1195 114 22.8 16 2i 9.85 95 18.9 15 1 sj t| 8.00 78 15.60 13 2 2 6i 6.30 62 12.40 12 3 If 5* 4.85 48 9.60 10 4 If 5 4.15 42 8.40 8| 5 ll 4| 3.55 36 7.20 7f 51 If 4i 3.00 31 6.20 7 G H 4 2.45 25 5.00 6* 7 H 3£ 2.00 21 4.20 6 8 l 3 1.58 17 3.40 5i 9 1 2} 1.20 13 2.60 4 10 1 2i 0.89 9.7 1.94 4 m 1 2 0.62 6.8 1.36 3* ici & If 0.50 5.5 1.10 2f 10| ^ $ 0.39 4.4 0.88 2J 10a 7 0.30 3.4 0.68 2 106 I if 0.22 2.5 0.50 n 10c £$ i 0.15 1.7 0.34 i 10a* i 1 0.10 1.2 0.24 i CAST STEEL. 2| 8f 11.95 228 45.6 10 2* 7£ 9.85 190 37.9 9* 1 2i 7* 8.00 156 31.2 4 2 2 61 6.30 124 24.8 8 3 If 5* 4.85 96 19.2 n 4 \l 5 4.15 84' 16.8 6? 5 S 3.55 72 14.4 5| 5* If 3.00 62 12.4 5£ 6 li 4 2.45 50 10.0 5 7 ll 3J 2.00 42 8.40 4* 8 1 3 1.58 34 6.80 4 9 ! 2| 1.20 26 5.20 3£ 10 2i 0.89 19.4 3.88 3 10i f 2 0.62 13.6 2.72 21 10} ft If 0.50 11.0 2.20 If 10f t H 0.39 8.8 1.76 1* 10a t l| 0.30 6.8 1.36 ii 106 It 0.22 5.0 1.00 i 10c t 1 0.15 3.4 0.68 § lOd 1 0.10 2.4 0.48 * STEAM BOILERS. 1327 STEAM. STEAM BOILERS. Points to Remember in Selecting* a Boiler. (a) Suitability of furnace and boiler to kind of fuel. (b) Efficiency as to evaporative results. (c) Rapidity of steaming including (I.) "Water capacity for given power. (II.) Water surface for given power. {d) Steam keeping qualities. (e) Safety from explosion. (/) Floor space required. (g) Portability, and ease with which boiler can be removed when old, for replacement by a new boiler. (h) Amount of, ease of, and rapidity of repairs. (i) Simplicity and fewness of parts. (j) Ability to stand forcing in case of necessity. (k) Price, including cost of freight and setting. (I) Durability and reliability, (m) Ease of cleaning and inspection both inside and outside. (n) Freedom from excessive strains due to unequal expansion and ability to withstand same. (o) Efficient natural circulation of water. (p) Absence of joints or seams where flames may impinge. For central stations it is necessary to arrange for a number of boilers rather than one or two large ones. The size of unit adopted will depend to some extent on the character of the expected load diagram. With a number of boilers the cost of the reserve plant is reduced, though beyond, say six, there is less object in increasing the number on this account. Types. Horizontal Return Tubular. — More generally used in United States than any other. Fire first passes under the shell, returns to front through tubes, thence up the chimney, except in some cases gases are again returned over top of the shell. Limited as to size and pressures carried by reason of external tiring. Water-tube. — Very largely used where high steam pressures or safety from explosion are desirable. Fire passes about the exterior of tubes and in most cases under about one-half the circumference of the steam drums. Can be built for any size or pressure. Tubes are generally placed in a slanting position, from one set of headers to another, as in the Babcock & Wilcox, Heine & Co. ; or vertically, as in the Sterling and Cahall. Vertical Eire Tube. — Used considerably in New England. Spe- cial design by Captain Manning; tubes 15 feet long 2\ inches diameter, arranged in vertical shell with large combustion chamber surrounded by a water leg. Gases mingle in combustion chamber, and in passing through the long narrow tubes give up nearly all the heat, practicably leaving flue gases 450° to 500° F. By controlling height of water, steam can be superheated. Can be built for high pressures and of large size. m?™ * S-°I Marin « Boilers. — Not much used for electrical purposes, bnell of thick material, short in length and large in diameter. Furnaces internal, with return tubes from combustion chamber to uptake, i ?*wXn? S are the Winder boiler, of small diameter and considerable lengtn (20 to 35 feet). Fired externally, and gases pass under full length to cnimney. Flue boiler, has two or three large tubes running full length of snen, which is long and of small diameter. Fired externally under the shell, gases return through the flues to uptake. Neither of these types is now us«d for electrical purposes. The Horse-Power of Steam Boiler. The committee of the A. S. M. E. on "Trials of Steam Boilers in 1884" CTranSc,vol. vi. p. 265), discussed the question of the horse-power of boilers ; 1328 STEAM. The Committee) A.S.M.E. see Trans, vol. xxi.) approves the conclusions of the 1885 Code to the effect that the standard " unit of evaporation " should be one pound of water at 212° F. evaporated into dry steam of the same temperature. This unit is equivalent to 965.7 British thermal units. The committee recommends that, as far as possible, the capacity of a boiler be expressed in terms of the " number of pounds of water evaporated per hour from and at 212°." It does not seem expedient, however, to aban- don the widely recognized measure of capacity of stationary or land boilers expressed in terms of " boiler horse-power." The unit of commercial boiler horse-power adopted by the Committee of 1885 was the same as that used in the reports of the boiler tests made at the Centennial Exhibition in 1876. The Committee of 1885 reported in favor of this standard in language of which the following is an extract : " Your Committee, after due consideration, has determined to accept the Centennial standard, and to recommend that in all standard trials the com- mercial horse-power be taken as an evaporation of 30 pounds of water per hour from a feed-water temperature of 100° F. into steam at 70 pounds gauge pressure, which shall be considered to be equal to 34J units of evaporation ; that is, to 34£ pounds of water evaporated from a feed-water temper- ature of 212° F. into steam at the same temperature. This standard is equal to 33,305 thermal units per hour." The present Committee accepts the same standard, but reverses the order of two clauses in the statement, and slightly modifies them to read as follows : The unit of commercial horse-power developed by a boiler shall be taken as 34£ units of evaporation per hour ; that is, 34£ pounds of water evaporated per hour from a feed-water temperature of 212° F. into dry steam of the same temperature. This standard is equal to 33,317 British thermal units per hour. It is also practically equivalent to an evaporation of 30 pounds of water from a feed-water temperature of 100° F. into steam at 70 pounds gauge pressure.* The Committee also indorses the statement of the Committee of 1885 con- cerning the commercial rating of boilers, changing somewhat its wording, so as to read as follows : A boiler rated at any stated capacity should develop that capacity when using the best coal ordinarily sold in the market where the boiler is located, when fired by an ordinary fireman, without forcing the fires, while exhibit- ing good economy ; and, further, the boiler should develop at least one- third more than the stated capacity when using the same fuel and operated by the same fireman, the full draft being employed and the fires being crowded ; the available draft at the damper, unless otherwise understood, being not less than \ inch water column. Heating: Surface of Boilers. Although authorities disagree on what is to be considered the heating surface of boilers, it is generally taken as all surfaces that transmit heat from the flame or gases to the water. The outside surface of all tubes is used in calculations. Kent gives the following rule for finding the heating surface of Vertical Tubular Boilers. — Multiply the circumference of the fire- box (in inches) by its height above the grate. Multiply the combined circum- ference of all the tubes by their length, and to these two products add the area of the lower tube sheet ; from this sum subtract the area of all the tubes, and divide by 144 : the quotient is the area of heating surface in square feet. Horizontal Return Tubular Boilers. — (Christie). Multiply the length of that part of circumference of the shell (in inches) exposed to the fire by its length ; multiply the circumferences of the tubes by their num- ber, by their length in inches ; to the sum of these products add two-thirds of the area of both tube sheets less twice the area of tubes, and divide the remainder by 144. The result is the heating surface in square feet. Heating: Surface of Tubes. — Multiply the number of tubes by the diameter of a tube in inches, by its length in feet, and by .2618. The diam- eter used should be that of the fire side of the tube. * According to the tables in Porter's Treatise on the Richards Steam En- gine Indicator, an evaporation of 30 pounds of water from 100° F. into steam at 70 pounds pressure is equal to an evaporation of 34.488 pounds from and at 212° ; and an evaporation of 34\ pounds from and at 212° F. is equal to 30.010 pounds from 100° F. into steam at 70 pounds pressure. The " unit of evaporation" being equivalent to 965.7 thermal units t the commercial horse-power = 34.5 x 965.7 = 33 s 317 thermal units. STEAM BOILERS. 1329 Heating* Surface per Horse-power. — There is little uniformity of practice among builders as to the amount of heating surface per horse- power, but 12 square feet may be taken as a fair average. Babcock & Wil- cox ordinarily allow 10 square feet, but usually specify the number of square feet of heating surface. The Heine Boiler Company allow 1\ square feet, and the water-tube type in general will develop a horse-power for that amount of surface. Specifications for boilers should always clearly state the amount of heating surface required. Orate Surface. — The amount of grate surface per horse-power varies with the character of fuel used and the draught that is available. With good quality of coal about equal results can be obtained with strong draught and small grate surface, and with large grate surface and light draught. Pittsburg coal gives best results with strong draught and a small grate sur- face. The following table shows the usual requirements, but in general grate surface should be liberal in size, and a rate of combustion of about 10 lbs. per hour will be found good practice. Grrate Surface per Horse-Power. (Kent.) oSSoO on 3 S u u Pounds of Coal burned per square foot of Grate per hour. 8 10 12 15 20 25 30 35 40 Square Feet Grate per H.P. Good coal and boiler . . . Fair coal or boiler . . . Poor coal or boiler . . . Lignite and poor boiler . 10 9 8.61 8 7 6.9 6 5 3.45 3.45 3.83 4. 4.31 4.93 5. 5.75 6.9 10. .43 .35 .28 .23 .17 .14 .11 .10 .48 .38 .32 .25 .19 .15 .13 .11 .50 .40 .33 .26 .20 .16 .13 .12 .54 .43 .36 .29 .22 .17 .14 .13 .62 .49 .41 .33 .24 .20 .17 .14 .63 .50 .42 .34 .25 .20 .17 .15 .72 .58 .48 .38 .29 .23 .19 .17 .86 .69 .58 .46 .35 .28 .23 .22 1.25 1.00 .83 .67 .50 .40 .33 .29 .09 .10 .10 .11 .12 .13 .14 .17 .25 Area of Gas-Passages and Flues. This is commonly stated in a ratio to the grate area. Mr. Barrus says the highest efficiency for anthracite coal, when burning 10 to 12 lbs. per square foot of grate per hour, is with tube area | to T ^ of grate surface ; and for soft coal the tube area should be ^ to | of the grate surface. Other rules in common use are to make the area over bridge walls (for horizontal return tubular boilers) \ the grate surface ; tube area£ and chim- ney area £. Air-space in Orates. — Usual practice is 30% to 50% area of grate for air space. If fuel clinkers easily, use the largest air space available. With coal free from clinker smaller air space may be used. Distance between Under Side of Boiler and Top of Orate. (For Horizontal Tubular Boiler.) For anthracite coal this should be 24 inches for the larger sizes, and can be 20 inches for the smaller sizes, such as pea, buckwheat, and rice. For bituminous coals non-caking, the grate should be about 30 inches below the boiler, and for fatty or gaseous coals from 36 to 48 inches. For average bituminous coals the distance can be 36 inches. Anthracite and bituminous coals cannot be economically burned in the same furnace. Steam Boiler Efficiency. The ratio of the heat units utilized in making steam in a boiler, to the total heat units in the coal used is called the efficiency of the boiler, and is 1330 STEAM. rated in per cent. For example, the heating value of good anthracite coal is about 14,500 B. T. U., and will evaporate from and at 212° 15 lbs. water (14,500 -J- 966). If a boiler under test evaporates 12 lbs. water per pound of 12 X 100 combustible, the efficiency will be — — = 80%, a figure not often ob- tained, but possible under special conditions. The heating value of bitumi- nous coals varies so much that it is necessary to determine it by a coal calorimeter before it is possible to determine the boiler efficiency. Strength of Riveted Shell. (Abridged from Barr on " Boilers and Furnaces.") Wrought-iron "boiler-plates should average 45,000 lbs., and mild steel 55,000 lbs., tensile strength per square inch of section ; but the gross strength of plate is lessened by the amount which has been taken out of it for the inser- tion of rivets. The following tables give the calculated working pressure for double- riveted and triple-riveted lap joints, and for butt-joints triple riveted, the factor of safety being 5. The rule for calculating the safe working pressure is : Multiply together the tensile strength of the plate, the thickness of the plate in parts of an inch, and the efficiency of the joint (see Riveting) ; divide the product by one-half the diameter of the boiler multiplied by the factor of safety. Working* Pressure for Cylindrical Shells of Steam Boilers. Factor of Safety, 5. (Barr.) Lap-joints, Double-Riveted. Lap- Joints, Triple -Riveted. Thick- Diam- ness in eter 16ths of an Inch. Iron Steel Steel Iron Steel Steel Inches. Shell, Shell, Shell, Shell, Shell, Shell, Iron Iron Steel Iron Iron Steel Rivets. Rivets. Rivets. Rivets. Rivets. Rivets. 36 4 91 Ill Ill 100 121 123 5 112 128 137 124 139 151 40 4 82 100 100 90 109 110 5 101 115 123 112 125 136 44 4 74 91 91 83 99 100 5 91 105 112 101 114 124 48 5 84 96 102 93 104 114 6 99 107 121 110 118 135 52 5 77 89 95 86 96 105 6 92 99 112 102 109 124 54 5 75 85 91 83 93 101 6 88 96 108 98 105 120 56 5 72 82 88 80 89 97 6 85 92 104 95 101 116 60 5 67 77 82 74 83 91 6 79 85 97 88 95 108 62 6 77 83 94 85 92 104 7 88 92 108 98 103 120 64 6 74 81 91 83 89 101 7 86 89 105 95 100 117 66 6 72 78 88 80 86 98 7 83 87 102 93 97 113 68 6 70 76 86 78 84 95 7 81 80 99 90 94 110 70 6 68 74 83 76 81 92 7 78 82 96 87 91 107 72 7 76 79 93 85 89 104 8 85 89 104 97 98 117 STEAM BOILERS. 1331 Working* Pressure for Cylindrical Shells of Steam Hollers. (Barr.) Butt Joints, Triple Riveted. Factor of Safety , 5. Diameter Inches. Thick- ness in 16ths of an inch. Iron Shell, Iron Rivets. fir Steel Shell, Iron or Steel Rivets. Diam- eter, Inches. Thick- ness in 16ths of an inch. Iron Shell. Iron Rivets.* Steel Shell, Iron or Steel Rivets. 4 108 134 6 83 102 36 5 135 165 70 7 97 118 6 161 197 8 110 134 4 102 127 9 123 151 38 5 128 156 6 80 99 6 152 187 72 7 94r 115 4 97 120 8 107 131 40 5 121 148 9 120 147 6 145 178 7 90 110 4 93 115 75 8 102 125 42 5 116 141 9 115 141 6 138 169 10 128 157 4 89 109 7 87 106 44 5 110 135 78 8 99 121 6 132 161 9 111 135 4 85 105 10 123 151 46 5 106 129 8 92 112 6 126 154 9 103 126 5 101 124 84 10 115 140 48 6 121 148 11 126 158 7 141 172 12 137 167 5 97 119 8 86 105 50 6 116 142 9 96 117 7 135 165 90 10 107 131 5 93 114 11 117 143 52 6 111 137 12 128 156 7 130 159 8 80 98 5 90 110 9 90 110 54 6 107 132 96 10 100 123 7 125 153 11 110 134 5 87 106 12 120 146 56 6 103 127 8 75 92 7 121 148 9 85 104 5 84 102 102 10 94 115 58 . 6 100 123 11 104 127 7 117 142 12 113 138 6 97 118 8 71 87 60 7 111 138 9 80 98 8 128 157 108 10 89 109 6 93 115 11 98 120 62 7 109 133 12 107 130 8 124 152 8 68 83 6 90 111 9 76 93 64 7 106 129 114 10 84 103 8 120 147 11 93 113 9 135 165 12 101 123 6 88 108 8 64 78 66 7 102 125 9 71 88 8 117 143 120 10 80 98 9 131 160 11 88 108 6 85 105 12 96 117 68 7 99 121 8 113 138 9 127 155 1332 STEAM. Safe Working; Pressure for Shell Plate. U. S. Statutes. — d — diameter of boiler in inches. P=. safe working pressure, lbs. per square inch. t =z thickness of metal in inches. w =z tensile strength of metal. k =z factor of safety = 6 for U. S. and 4.5 for Great Britain. ^ y 2 x ic P — - — for single-riveted. For double-riveted, add 20%. d X 6 ° ' Board of Trade.— > X B X t X 2 P = d X fc X 100 where the notation is the same as in U. S. rule, and B = percentage of strength of joint as compared with solid plate. Rules Governing' Inspection of Boiler* in Philadelphia. In estimating the strength of the longitudinal seams in the cylindrical shells of boilers, the inspector shall apply two formulae, A and B : A, Pitch of rivets — diameter of holes punched to receive the rivets __ ~~ pitch of rivets ~" percentage of strength of the sheet at the seam. SArea of hole filled by rivet x No. of rows of rivets in seam x shear- ing strength of rivet _ pitch of rivets X thickness of sheet x tensile strength of sheet percentage of strength of the rivets in the seam. Take the lowest of the percentages as found by formulas A and B, and apply that percentage as the " strength of the seam" in the following for- mula, C, which determines the strength of the longitudinal seams : ! Thickness of sheet in parts of inch x strength of seam as obtained by formula A or B x ultimate strength of iron stamped on plates _ internal radius of boiler in inches X 5 as a factor of safety safe working pressure. Safe Working: Pressure for Flat Plates. U. S. Statutes. — P = safe working pressure. S =r surface supported, square inches. t — thickness of metal in sixteenths of an inch. k — constant for plates of different thickness, and for various condi- tions. p — greatest pitch in inches. P -L*±. ~ p 2 K— 112 for T Vinch plates and less, fitted with screw stay bolts and nuts, or plain bolt fitted with single nut and socket, or riveted head and K= 120 for plates more than T 7 g inch thick, under same conditions. #:= 140 for flat surfaces where the stays are fitted with nuts inside and out. K=z 200 for flat surfaces under same conditions, but with washer riveted to plate, washer to be one-half as thick as plate, and of a diameter f pitch. STEAM BOILKBS. 1333 No brace or stay on marine boilers to have a greater pitch than 10J inches on fire boxes and back connections. Plates fitted with double-angle irons riveted to plate, and with leaf at least two-thirds thickness of plate, and depth at least one-fourth of pitch, allowed the same pressure as plate with washer riveted on. Board of Trade. — Using same notation as in U. S. rules : p __ * (< + !)* S — 6 JT = 125 for plates not exposed to heat or flame, the stays fitted with nuts and washers, the latter at least three times the diameter of the stay and § the thickness of the plate ; K=. 187.5 for the same condition, but the washers | the pitch of stays in diameter, and thickness not less than plate ; K = 200 for the same condition, but doubling plates in place of washers, the width of which is f the pitch, and thickness the same as the plate ; K =: 112.5 for the same condition, but the stays with nuts only ; K = 75 when exposed to impact of heat or flame and steam in contact with the plates, and the stays fitted with nuts and washers three times the diameter of the stay, and § the plate's thickness ; 7T = 67.5 for the same condition, but stays fitted with nuts only ; K = 100 when exposed to heat or flame, and water in contact with the plates, and stays screwed into the plates, and fitted with nuts ; K = 66 for the same condition, but stays with riveted heads. Ductility of Boiler Plate. — U. S. Inspectors of Steam Vessels. In test for tensile strength, sample shall show reduction of area of cross- section not less than the following percentages : Iron. 45,000 lbs. tensile strength and under 15 per cent. For each additional 1000 t. s. up to 55,000 t. s. add . 1 " 55,000 lbs. tensile strength, and above 25 '* Steel. All steel plates £ inch thick and under 50 per cent. " " " } to | inch 45 " M " " I inch and above 40 " Boiler Head Stays. The United States Regulations on braces are : " No braces or stays here- after employed in the construction of boilers shall be allowed a greater strain than 6,000 lbs. per square inch of section. Braces must be put in suf- ficiently thick so that the area in inches which each has to support, multi- plied by the pressure per square inch, will not exceed 6,000 when divided by the cross-sectional area of the brace or stay. 11 Steel stay-bolts exceeding a diameter of \\ inches, and not exceeding a diameter of 1\ inches at the bottom of the thread may be allowed a strain not exceeding 8,000 lbs. per square inch of cross-section ; steel stay bolts exceeding a diameter of 1\ inches at bottom of thread may be allowed a strain not exceeding 9,000 lbs. per square inch of cross-section ; but no forged or welded steel stays will be allowed. "The ends of such stay may be upset to a sufficient thickness to allow for truing up, and including the depth of the thread. And all such stays after being upset, shall be thoroughly annealed.** 1334 STEAM. Direct Braces. — The following table is given by Mr. W io,;' Boilers and Furnaces," p. 122. The working strength assun mate strength of 6000 lbs. per square inch of section. M. Barr assumes an ulti. Diam- eter of Wrought Iron Stays. Inches square each B Pressures per race will Support for Square Inch. Brace Inches. Area sq. in. Working Strength Pounds. 75 Pounds. 100 Pounds. 125 Pounds. 150 Pounds. i 1 n if .60 .78 .99 1.23 1.48 1.77 3600 4712 5964 7362 8880 10620 7.0 7.9 8.9 9.9 10.7 11.9 6.0 6.9 7.7 8.6 9.5 10.4 5.4 6.1 6.9 7.7 8.5 9.2 4.9 5.6 6.4 7.0 7.7 8.5 Diagonal Braces. — (" Boilers and Furnaces," p. 129.) These must be calculated separately. Let Then A = surface to be supported in square inches. B z=z working pressure in lbs. H=z length of diagonal stay in inches. L = length of line drawn at right angles from surface, to be sup- ported to end of diagonal stay in inches. S =: working stress per square inch on stay in lbs. a = area required for direct stay in square inches. a x zr area of diagonal stay in square inches. T=z diameter of diagonal stay in inches. H — a x x L -± a. ▼ .7854 y .% X fix H. '854 $ X L .7854 XT* XSXL AxH Boiler Setting's. Water tube and special types of boilers require special settings largely controlled by local conditions, location of flues, etc., and cannot be tabulated here. The setting of horizontal return tubular boilers has become so nearly standardized that the table following, taken in connection with the cuts, will give all the general dimensions of brick-work required. For all special boiler settings, furnaces, etc., the reader is referred to the makers of each. STEAM BOILERS. 1335 Fig. 3. 1336 STEAM. BO •dox *« lie^l eSpug jo ssaujpiqx © BO yH-nrtHrtHrtiirirtriHrtrl eSpug jo ssau^oiqx fc 02 OS •loojtf jo dox uioij S[iBAV JO ^q§iOH m S mOOM'OOh^OhhhoooOH £t>t~t>l^t^akoO<»OOat>ae>GOCiOi© aaAO sn«Ai JO qjpjAl H £cit>.i>.t>t~t>.coocc5CiC5C5©© •S[[T?AY opig JOJ noijuptmo^ jo qjSiiaT; w M^I>t»t-.t>^l>C5C5H^H?5C0»O ^riCCCClOlOCWOQCNXC;C500 ^uoj^ jo ssaujpiqx b BD •aoBdg Jiy j^9H: Sjo ssonjpiqx M BO a> ^JOOQOOOOOCOOOQOOOCOOOOOOOOOOOQO yrtrlHrtHrtrtHHrtHrlHrlH Pi •ao^dg Jiy tpin-s q;i.\v 'sn^A! apjS 5 jo ssaa^oiqx H 09 •sit^AVaPTS opisni jo ss8U5[Oiqx 00 © yr1H rtrtrtrtrtHHr-HHrtHH •si^AV ©PTS apts^no jo ssatnpiqx fe ^JOOOOOOOOOOOOOOQOOOOOCOQOOOQOCO © S3 M a o Q \i8[iog jo ^uojj 0% 8UIO(X JO J8^U9Q s 0) yCOlO'tNt'l>00(»OC50300JOO i— 1 •auioa; u0 q8wv{£ jo dox o* J^uog; jo dox a BO j(n(n«o©c0o(n(n!n(n!n00c0c000 o<^c<>t»t>00000>05OOO^!N'*O *8UI1 JOOt^ GAoq'e ^OTjg noxnirioo jo 'o^ 6250 7100 8200 8750 9250 10700 11700 14450 17680 16600 17900 19000 19600 21550 22500 a OQ •apiS aq+ +« lim pai? loiiog; ii99Aij9q aoBdg N 0Q '300tOb-t>-OC5q5C3CJC2 ■^lOOiOiaiOlOiOlOlOiOlClOCDCOCO H •+UOJJ 4 « J9iiog jo 9pis japnn. 0; jooii jo dox ta OQ PI 9 C3 Pi *aovnjn l g; jo qjpiAl H 72 |88§§338S;S88888£ • 'Govnm^ jo q^Sn9n[ 0Q |83§33gg;S888888S PI M •j9iioa: jo 9pi§ J9pnfi oj S9 + ^J9 jo dox ti 00 H«C®<0OC0C0O00(»OOOOOO ^(MtNfMC^COC^XNOJCOCOCOCOCOCOCO Pi H •J139£[ 9t[+ 4^ S9+T3.I*) jo dox °4 8lI IT JOoy[ a 5C j^CCC0iOiO»O000000600000Q0t-(i-((N' '^(NiMiMfMC^iMC^iMfNlMlMtMCOCOCO PI 1— I •(jnojj: 9qj %v S9 4 bjo jo dox oj quvj joou Pk 03 -^ c£5 0$ c^ c<» != V^-f 4 mches. (2) Horse-power = 3.33 E^M. (3) The above formulae are by Kent, and are based on a consumption of 5 lbs. coal per h. p. per hour. W. W. Christie, in a paper read before the A.S.M.E., Trans., vol. xviii., p. 387, gives as his opinion that all chimneys should be compared and rated by using coal capacity as a basis, not horse- power. In the following table, coal capacity can be found by multiplying h.p. by 4. Size of Chimneys for Steam-Boilers. (W. W. Christie.) 72 78 84 90 96 102 108 114 120 132 144 Height of Chimney. 50 ft. 60 ft. 70 ft. 80 ft. 90 ft. 100 ft. 110 ft. 125 ft. 150 ft. 175 ft. 200 ft. 225 ft. 250 ft. 300 ft. Boiler Horse-power=r 3.25 jS H\ 4 lbs. of coal burned considered 1 H.P. 4 2 55 72 91 114 46 62 78 101 124 149 179 49 65 85 107 133 163 192 224 263 52 68 91 114 143 172 205 241 282 364 1 98 124 153 182 218 257 296 387 491 605 1 159 192 228 270 312 410 517 637 774 920 202 241 283 332 429 543 669 809 962 1131 1310 1 1 1 1 257 302 351 458 579 715 865 ia5i 1206 1401 1609 1830 2067 2314 390 510 647 797 965 1147 13*9 1563 1794 2041 2304 2584 2879 3191 3861 4596 683 845 1021 1215 1459 1654 1898 2161 2434 2734 3045 3374 4082 4859 1092 1300 1524 1768 2031 2311 2607 2925 3257 3611 4368 5200 1378 1619 1875 2155 2451 2766 3101 3455 3*29 46bl 5515 1706 1976 2269 2584 2915 3269 3643 4037 4882 5811 2165 2486 2831 3195 3578 3991 4420 5350 6367 CHIMNEYS, 1339 The following table* will prove useful to those having to do with electric installations, and gives the horse-power of chimneys to be used in power plants having very efficient engines, such as compound or triple expansion engines, when 2 lbs. of coal burned under the boiler produce one horse- power at the engine. Size of Chimney for Steam Boilers. (W. W. Christie.) Height of Chimney. a 50' 60' 70' 80' 90' 100 / IKK 125' 150 7 175' 200' 225' 250' 30O 7 e3 s Horse-i >ower = 6.5 A^H. When 2 lbs. coal burned per hour = 1 H.P. 18 21 24 27 84 110 144 182 f 92 124 156 202 98 130 170 214 104 136 182 228 1 196 248 1 30 33 36 39 228 248 298 358 266 326 384 448 286 344 410 482 306 364 436 514 318 384 456 540 404 482 566 514 604 42 526 564 728 592 774 982 1210 62* 820 1034 1274 662 858 1086 1338 702 916 1158 1430 780 1020 1294 1594 48 54 60 1366 1690 66 72 78 84 1548 1840 1618 1924 2262 2620 1730 2102 2412 2802 1930 2294 2698 3126 2042 2430 2918 3308 2184 2600 3048 3536 2756 3238 3750 3412 3952 4330 90 96 102 103 »• 3218 3660 4134 4628 3588 4082 4608 5168 3796 4322 4868 5468 4062 4622 5214 5850 4310 4902 5532 6202 4538 5168 5830 6538 4972 5662 6360 7156 114 120 132 144 5758 6382 7722 9192 6090 6748 8164 9718 6514 8736 6910 9262 7286 9764 7982 10700 Chimney Construction. A brick chimney shaft is made up of a series of steps, each of which is of uniform thickness, but as we ascend each succeeding step is thinner than the one it rests upon. These bed joints at which the thickness changes are the joints of least stability. The joints and the one at the ground line are the only ones to which it is necessary to apply the formulas for deter- mining the stability of the stack. The height of the different steps of uniform thickness varies greatly, ac- cording to the judgment of the engineer, but 170 feet is, approximately, the extreme height that any one section should be made. This length is seldom approached even in the tallest chimneys, as the brick-work has to bear, in addition to its weight, that due to the pressure of the wind. The steps should not exceed about 90 feet, unless the chimney stack is inside a tower which protects it from the wind. In chimneys from 90 to 120 feet high the steps vary from 17 to 25 feet, the top step being one brick thick ; in chim- * ^Chimney Design and Theory," W. W.Christie, D. VanNostrand Company. 1340 STEAM. B (2 0Q@ Perforated radial bricks used for chimneys. Bond in radial brick work. Fig. 4. CHIMNEYS. 1341 neys from 130 to 150 feet the steps vary from 25 to 35 feet; in chimneys from 150 to 200 feet the steps vary from 35 to 50 feet; in chimneys from 200 to 300 feet and over, the steps vary from 50 to 90 feet, the top step being on© and one-half bricks thick. The outside dimensions of a chimney at the base should generally not be less than one-tenth of the height of the stack for square chimneys; one-eleventh for octagonal, and one-twelfth for round. The battery may be 2£ inches for every 10 feet. The foundation of a chimney is one of the most important points to be considered. When this is upon solid rock it is only necessary to excavate to a depth sufficient to prevent the heat of the gases from materially affect- ing the natural stone, and to secure the spread of the base. In cases where chimneys are to be built upon alluvial clays or made ground, it is necessary to excavate until a good stiff clay, hard sand, or rock bottom is reached. The excavation is tilled with concrete in various ways, or filled according to the judgment of the engineer, so as to economize material without endangering the structure. Babcock and Wilcox give the following formula for the ability of brick chimneys to withstand wind pressure. w=. weight of chimney in lbs. (brickwork = 100 to 130 lbs. per cubic foot.) d — average diameter in feet, or width if square. h = height in feet. b = width of base. k =z constant, for square chimneys = 56. for round chimneys z=z 28. for octagonal chimneys = 35. c=zk — — and w = k -=— . w b Radial Brick Chimneys. Another type of chimney now much used in the East, is built of radial brick, perforated vertically with holes about 1" square, passing entirely through them. The advantage of these bricks is said to be a better bond, as the cement makes a dowel in the perforations. They are made of a special quality of clay, having greater care in the making, are burned at a greater heat than the red brick, and are said to be of a more uniform grade. Radial brick chimneys as built in the United States do not always have lining, for the brick are supposed to be capable of withstanding the heat of the gases usually met with, but in special cases a lining is built in them, and is carried by the outer shell. The less number of joints to the weather is also given as a point in favor of the radial brick chimney. In making comparisons of the costs of the several types of chimneys, if of brick, they should have the same height, inside diameter, lightning protection details, ladder equipment, quality of workmanship, .same factor of stability. 1342 STEAM. Draft Power for Combustion of Fuels. (R.H.Thurston.) Fuel. Draft of Chini ney in Inches of Water. Fuel. 'Draft in Ins. of Water. Wood Sawdust . . . Sawdust mixed small coal . . . Steam coal . . , Slack, ordinary , Slack, very small . with 0.20 to 0.25 0.50 0.35 0.60 0.40 0.60 0.75 0.75 0.75 0.90 1.25 Coal-dust Semi Anthracite coal Mixture of breeze and slack Anthracite ... . Mixture of breeze and coal-dust .... Anthracite slack . . 0.80 to 1.25 1.25 0.90 1.00 1.25 1.25 1.30 1.33 1.50 1.75 1.80 Heig-ht of Chimney for Burning' Given Amounts of Coal. Professor Wood (Trans. A. S. M. E., vol. xi.) derives a formula from which he calculates the height of chimney necessary to burn stated quan- tity of coal per square foot of grate per hour, for certain temperatures of the chimney gas. Absolute Pounds of Coal per Square Foot Grate Area. Temp. Outside Temp. Chim- 16 20 24 Air. ney Gases. Height of Chimney, Feet. © 700 67.8 157.6 250.9 49 800 55.7 115.8 172.4 ©J3 1000 48.7 100.0 149.1 1100 48.2 98.9 148.8 1200 49.1 100.9 152.0 « <5 1400 51.2 105.6 159.9 o %a 1600 53.5 110.9 168.8 2000 63.0 132.2 206.5 Rate of Combustion Due to Heig-ht of Chimney. Prof. Trowbridge (" Heat and Heat Engines," p. 153) gives the following table, showing the heights of chimneys for producing certain rates of com- bustion per square foot of area of section of the chimney. The ratio of the grate to the chimney section being 8 to 1. Lbs. Coal Lbs. Coal burned per Hour per sq. ft. of Grate. Lbs. Coal burned per burned per Lbs. Coal Height in Feet. Hour per Height in Hour per burned per sq. ft. of Feet. sq. ft. Sec- Hour per Section of tion of sq. ft. Grate. Chimney. Chimney. 25 68 8.5 70 126 15.8 30 76 9.5 75 131 16.4 35 84 10.5 80 135 16.9 40 93 11.6 85 139 17.4 45 99 12.4 90 144 18.0 50 105 13.1 95 148 18.5 55 111 13.8 100 152 19.0 60 116 14.5 105 156 19.5 65 121 15.1 110 160 20.0 CHIMNEYS. 1343 Dimensions and Cost of Brick Chimneys. (Buckley.) 6 o pj J4+1 2 2-3 t> qonj J8d -j'H •A^aqduej jod nuj jo ^^xo^d^o 9ATJ(I spunoj s*f g ;« JnoH add uoi^ejod^Ag; •^to'BdBO irej UIOIJ ^p'Bd'BO J8iiog 'j'H •J oOSS S8SBO ^oaj oiqno UI U^ JO ^IOBd'60 •;ax;no jo 8zxg •^oiui jo 'ra^id •^J8qdij8(j ye q^piM. fOiOOOQOOOOOOOO 3$ec3$88o388c$8S8 5 t- O CO ? '-L© CO -* © © l> C^rH, r^T^r^C^fcNCO** tff©00©~«OlOT^Ttt>t> 8SS88S83$388£aS arsa^gfsraSfa; 88§g;s8g£cSc£!§g|: •U«J JO 9ZTS g8g888Sg|88§§f 1346 STEAM. The effect of the temperature of the gases, on the power required to operate a fan, is shown very clearly by the following : Effect of Temperature of Oases on fan Load. Induced Draft. Draft in inches of water Temperature of gases at fan, degree F. . Speed of fan, revolution per minute . . Current required by fan motor — amperes Current generated by plant — amperes . Proportion used by fan — per cent . . . Boiler H.P. developed ........ 1 2 0.42 0.46 199.6 162.5 154. 179. 10.3 13.3 896. 1236. 1.15 1.17 521.7 600.6 0.24 330. 230. 20.4 960. 2.08 439.2 The blower used was an American Blower Co.'s centrifugal fan with 28 X 84 inch wheel. The third test, gases 130 deg. hotter than first, requires about 100 per cent more power, and yet the boiler evaporation is about 20 per cent less than in the first test. — Curtis Pub. Co., by Davis & Griggs. The cost of the above Mechanical Draft outfit (2 fans), including motors, was $5.53 per boiler H.P. All of the blower methods of draft production must be considered in con- nection with, and be planned with especial regard to, the quantity of fuel to be burned in a given time, and the amount of air needed for the complete combustion of the fuel, which air must necessarily pass through the blowers. 18 to 25 lbs. of coal per square foot of grate per hour is all the coal that should or can be burned with economy under natural draft; a greater amount necessitates forced draft. Another thing which should not be lost sight of in connection with the burning of small coals, is the unburnt coal falling through the grate, which in the case of anthracite culm has reached 58 per cent (found in the ashes). Kinds and Ingredients of fuels. The substances which we call fuel are : wood, charcoal, coal, coke, peat, certain combustible gases, and liquid hydrocarbons. Combustion or burning is a rapid chemical combination. The imperfect combustion of carbon produces carbonic oxide (CO), and carbonic acid or dioxide (C0 2 ). From certain experiments and comparisons Rankine concludes " that the total heat of combustion of any compound of hydrogen and carbon is nearly the sum of the quantities of heat which the hydrogen and carbon contained in it would produce separately by their combustion (CH 4 — marsh gas or fire-damp excepted)." In computing the total heat of combustion of a compound, it is conven- ient to substitute for the hydrogen a quantity of carbon which would give the same quantity of heat ; this is accomplished by multiplying the weight of hydrogen by 62032 ^- 14500 = 4.28. From experiments by Dulong, Despretz, and others, u when hydrogen and oxygen exist in a compound in the proper proportion to form water (by weight nearly 1 part H to 8 parts O), these constituents have no effect on the total heat of combustion. " If hydrogen exists in a greater proportion, take into the heat account only the surplus." Dulong's formula for the total heat of combustion of carbon, hydrogen, oxygen, and sulphur, where C,H,0, and S refer to the fractions of one pound of the compound, the remainder being ash, etc. Let h ±z total heat of combustion in B.T.U. per pound of compound. h = 14660 C-f- 62000 (h— ^\ + 4000 #. (A.S.M.E. Trans, vol. xxi.) Rankine says : " The ingredients of every kind of fuel commonly used may be thus classed : (1) Fixed or free carbon, which is left in the form of char- coal or coke after the volatile ingredients of the fuel have been distilled away. These ingredients burn either wholly in the solid state, or part in the solid state and part in the gaseous state, the latter part being first dissolved by previously formed carbonic acid. "(2) Hydrocarbons, such as olefiant gas, pitch, tar, naphtha, etc., all of which must pass into the gaseous state before being burned. FUEL. 1347 " If mixed on their first issuing from amongst the burning carbon with a large quantity of air, these inflammable gases are completely burned with a transparent blue flame, producing carbonic acid and steam. When raised to a red heat, or thereabouts, before being mixed with a sufficient quantity of air for perfect combustion, they disengage carbon in tine powder, and pass to the condition partly of marsh gas, and partly of free hydrogen ; and the higher the temperature, the greater is the proportion of carbon thus disengaged. " If the disengaged carbon is cooled below the temperature of ignition be- fore coming in contact with oxygen, it constitutes, while floating in the gas, smoke, and when deposited on solid bodies, soot. " But if the disengaged carbon is maintained at the temperature of ignition, and supplied with oxygen suihcient for its combustion, it burns while float- ing in the inflammable gas, and forms red, yellow, or white flame. The flame from fuel is the larger the more slowly its combustion is effected. " (3) Oxygen or hydrogen either actually forming water, or existing in com- bination with the other constituents in the proportions which form water. Such quantities of oxygen and hydrogen are to be left out of account in de- termining the heat generated by the combustion. If the quantity of water actually or virtually present in each pound of fuel is so great as to make its latent heat of evaporation worth considering, that heat is to be deducted from the total heat of combustion of the fuel. The presence of water or its constituents in fuel promotes the formation of smoke, or of the carbona- ceous flame, which is ignited smoke, as the case may be, probably by mechanically sweeping along fine particles of carbon. " (4) Nitrogen, either free or in combination with other constituents. This substance is simply inert. " (5) Sulphuret of iron, which exists in coal and is detrimental, as tending to cause spontaneous combustion. " (6) Other mineral compounds of various kinds, which are also inert, and form the ash left after complete combustion of the fuel, and also the clinker or glassy material produced by fusion of the ash, which tends to «hoke the grate." Total Heat of Combustion of Fu«l«. (D. K. Clark.) The following table gives the total heat evolved by combustibles and their equivalent evaporative power, with the weight of oxygen and volume of air chemically consumed. Combustibles. s I III US S M Quantity of Air Consumed per Pound of C ora- bustible. lbs. lbs. Cu. Ft. at 62°F. 02 X> S fl O oh o£« cS^ © ©^3 S3 0; h n S M H rA O S§1 5^2 3 ° p — Hydrogen Carbon making CO . . . . . Carbon making C0 2 Carbonic oxide . Light Carbureted Hydrogen . . Olefiant Gas Coal (adopted average desiccated) Coke(adopted average desiccated) Lignite, perfect Wood, desiccated Wood, 25 per cent moisture . . Petroleum . . Petroleum oils . . . c . . . Sulphur 8.0 1.33 2.66 0.57 4.00 3.43 2.45 2.49 2.04 1.40 1.05 3.29 4.12 1.00 34.8 5.8 11.6 2.48 17.4 15.0 10.7 10.81 8.85 6.09 4.57 14.33 17.93 4 35 457 76 152 33 229 196 140 142 116 80 60 188 235 57 62000 4452 14500 4325 23513 21343 14700 13548 13108 10974 7951 20411 27531 4000 64.20 4.61 15.00 4.48 24.34 22.09 15.22 14.0? 13.51 11.36 8.20 21.13 28.50 4.17 1348 STEAM. © c8 S a) •&9 «€ i^M •ooe;e s«9 •1JBj OS ?-J O rH l> "^ ■<* t>; © CO OS ^ l6 IQ ^ lH OS CO t> ^ ib CO CN r*J p t> OS ^ CO © oo co ^ t*5 «* © oo co co -^5 •ajqi^snquiOQ # qj x q?T^ 'o&IS 1« pa^ uiojj. p^jo -d«A8 ja^j\i J° spunoj ni 8S 8 88Si58oi§8 "<£ t-t lO lft CD lO CN © t^ t^ 1ft •aiq^snquioQ jo punoj jad x posiiij jo^-BAi J° spimoj uj ^5 °. otSlSSJ'OS coo © t>- CO CO Tl- CO © Tf p oo »o coooot-»c^o CS *-i -• 8 lft CO OS t>-00t-C©COlftlft-• 00 CO CO t>- © CO o 3 I o © £ 0) p cjdo> •oh a © .2 • s ■S.5S 4) « o O £ * FUEL. 1349 Temperature of fire. By reference to the table of combustibles, it will be seen that the temper- ature of the fire is nearly the same for all kinds of combustibles, under sim- ilar conditions. If the temperature is known, the conditions of combustion may be inferred. The following table, from M. Pouillet, will enable the temperature to be judged by the appearance of the fire : Appearance. Temp. F. Appearance. Temp. F. Red, just visible . . " dull " cherry, dull . . " " full . . " " clear . 977° 1290 1470 1650 1830 Orange, deep . . . " clear . . . White heat .... " bright . . . " dazzling . . 2010 2190 2370 2550 2730 To determine Temperature oy Fusion of Metals, etc. Substance. Tern. F. Metal. Tern. F. Metal. Tern. F. Tallow . . . Spermaceti . Wax, white . Sulphur . . Tin ... . 92° 120 154 239 455 Bismuth . Lead . . . Zinc . . . Antimony . Brass . . 518° 630 793 810 1650 Silver, pure . . Gold, coin . . Iron, cast, med. Steel .... Wrought iron . 1830° 2156 2010 2550 2910 American Woods. Kind of Wood. Hickory — Shell bark. White oak .... Hickory — Red heart Southern pine ... Red oak Beech Hard maple . . . Virginia pine . . . Spruce New Jersey pine . . Yellow pine . . . , White pine . . . . Weight per Cord. 4469 3821 3705 3375 3254 3126 2878 2680 2325 2137 1904 1868 Value in Tons Coal. Anthracite Bituminous .52 .504 .459 .443 .425 .391 .364 .316 .291 .259 .254 .563 .481 .467 .425 .41 .394 .363 .338 .293 .24 .235 1350 STEAM. American Coals. State. Coal. Kind of Coal. Pennsylvania. Anthracite Cannel . . . Connellsville . Kentucky. Illinois. Indiana. Maryland. Arkansas. Colorado. Semi-bituminous Stone's Gas . Youghiogheny Brown . . . Coking . . , Cannel . . . Lignite . . Bureau Co. Mercer Co.. Montauk . . Block . , . Coking . . „ Cannel . . . Cumberland . Lignite . Texas. " Washington Ter. " Pennsylvania. Petroleum < Per Cent of Ash. 3.49 6.13 2.90 15.02 6.50 10.70 5.00 5.60 9.50 2.75 2.00 14.80 7.00 5.20 5.60 5.50 2.50 5.66 6.00 13.88 5.00 9.25 4.50 4.50 3.40 Theoretical Value. In Heat Units. 14,199 13,535 14,221 13,143 13,368 13,155 14,021 14,265 12,324 14,391 15,198 13,360 9,326 13,025 13,123 12,659 13,588 14,146 13,097 12,226 9,215 13,562 13,866 12,962 11,551 20,746 Pounds of Water Evap. 14.70 14.01 14.72 13.60 13.84 13.62 14.51 14.76 12.75 14.89 16.76 13.84 9.65 13.48 13.58 13.10 14.38 14,64 13.56 12.65 9.54 14.04 14.35 13.41 11.96 2147 The weight of solid coal varies from 80 lbs. to 100 lbs. per cubic foot. Tbe Heating- Value of Coals. On page 1351 are given the results (Sibley, Journal of Engineering) of some experiments made at Cornell University with a coal calorimeter devised by Prof. R. C. Carpenter. It consists of two cylindrical chambers, in the inner one of which the sample of coal is burned in oxygen. The heated gases pass through a coiled copper tube about 10 feet long contained in the outer cham- ber. The coil is surrounded by water which expands, the expansion being measured in a finely graduated glass tube, thus giving the heat units in the coal. The calorimeter is calibrated by burning in it pure carbon. Follow- ing are the tables : FUEL. 1351 a O © . >■' & © all a . © © ©02 © >. 020 X — u •£ © C3 *3 i-KM; 00 IflC oJoiwddodcobjaddod • o l£> ic iodcoo5oddt>C5d(Ndio ^ co a io oo cs co co rjHt-t^CO»ACOCiOC50qOCOC5 coiocoiri^t^^ioioidioc^rH r*< CO t- b- " c8 o3 e8 o c8 K* (_ ^ ^ © ^ • © © © ^ © ^3.02 02 02^02 •frl 3^ P5 ■ncT" 1 ^ . ■HH'ti.- l H M 3 J «s <* s • r ° « ° b * w 5 i - 5^ • y-> >F o © H ^ 2 08 pq 3*» © a O 02th 00 05C40HCOONOOO 8fc>" © > © e8 Q.1H 02<3 © O E * o Hi d CDCOOOCOOr Q>-2 © o dStfO©©Oo 1352 STEAM. Proximate Analysis of Coal. (Power.) Designation of Coal. ANTHRACITE. Beaver Meadow, Penn Peach Mountain, Penn Lackawanna, Penn Lehigh, Penn Welsh, Wales SEMI- ANTHRACITE . Natural Coke, Virginia Cardiff, Wales Lycoming Creek, Penn Arkansas, No. 16 Geol. Survey .... SEMI-BITUMINOUS. Blossburg, Penn Mexican Fort Smith, Arkansas Cliff, New South Wales, Australia . . Skagit River, State of Washington . . Cumberland, Maryland Cambria County, Penn Mount Kembla, New South Wales, A us. Fire Creek, West Virginia Arkansas, No. 12 Geol. Survey .... BITUMINOUS. Wilkeson, Pierce County, Washington . Cowlitz, Washington New River, West Virginia Pictou, Nova Scotia Big Muddy, Illinois Bellingham Bay, Washington .... Midlothian, Virginia Connellsville, Penn Illinois, Average Carbon Hill, Washington Clover Hill, Virginia Wellington. Vancouver Island, B.C. . . Franklin, Washington Rocky Mountains Newcastle, England Mokihinui, Westport, New Zealand . . Brunner Mine, Greymouth, New Zealand Pittsburg, Penn Nanaimo, Vancouver Island, B.C. . . . Hocking Valley, Ohio Pleasant Valley, Utah ....... Kentucky Ellensburg, Washington Olympic Mountains, Washington . . . Scotch, Scotland Roslyn, Washington Cook's Inlet, Alaska Kootznahoo Inlet, Admiralty I., Alaska Liverpool, England Calispel, Washington Carbonado, Washington Upper Yakima, Washington Methow, Washington 0-,^ 1.5 1.9 2.12 3.01 1.2 1.12 1.25 .67 1.35 1.34 1.0 1.07 .85 1.19 .97 2.46 1.2 .74 .88 1.33 1.16 .67 2.57 7.12 3.98 2.46 1.26 8.93 2.16 1.34 2.15 3.5 7.55 1.5 3.96 1.59 1.7 2.25 6.95 5.43 2. 2. 5.1 3.01 3.1 1.25 3.74 2.39 1.8 1.2 2.5 o *3 2.38 2.96 3.91 3.28 6.25 12.44 12.85 13.84 14.93 14.78 14.86 17.2 17.7 18.8 19.87 20.52 20.93 22.42 24.66 25.88 26.12 26.64 27.83 29.5 29.54 29.86 30.10 30.14 31.73 32.21 34.15 34.27 34.65 34.7 34.94 35.68 36. 36.05 36.15 37.73 37.89 39.1 39.15 39.19 39.7 39.87 37.02 39.96 41.18 42.27 42.47 43.71 ^d © o *3 < 88.94 7.11 89.02 6.13 87.74 6.35 88.15 5.56 88. 4.55 75.08 11.38 81.9 4. 71.53 13.96 74.06 9.66 73.11 10.77 55.7 28.44 73.05 8.68 71.8 9.65 71.66 8.35 72.26 6.12 69.37 9.15 66.96 10.91 75.5 .8 58.2 16.26 66.75 6.04 61.9 10.69 70.66 1.53 56.98 13.39 54.64 8.74 59.9 6. 53.01 14.74 59.61 8.23 45.93 15. 55.8 10.31 56.83 10.13 54.85 8.85 54.23 8. 42.85 14.95 59.3 4.5 57.92 3.18 56.62 6.11 55. 7.3 51.95 9.75 51.3 5.56 49.40 7.44 56.01 4.1 54.4 3.4 47.01 7.77 48.81 9.34 52.65 4.55 49.89 7.82 45.15 14.09 54.9 4.62 42.92 13.21 52.11 3.82 52.21 4.12 49.27 4.26 FUEL. 1353 l»roxiniate Analysis of Coal— Continued, Designation of Coal. w 03 2C go u ■a ^k2 .!'.. " i 3 6 '.' & Culm ............. " jft FUEL. 1355 Relative Values of Coals and How to Burn Them. (By Jay M. Wliitham.) Given boilers and chimney operating under natural draft and having certain sizes and dimensions, the capacities measured in steam output, which can he produced therewith, when using good grades of these coals, are as follows : Semi-bituminous coal (8 to 10 per cent ash) . . . No. 1 buckwheat anthracite (18 to 22 per cent ash in use) No. 2 buckwheat anthracite, or rice (18 to 22 per cent ash in use) Per cent. 100 80 68 It is more than likely that the percentage of ash and refuse obtained in service with Nos. 1 and 2 buckwheat will exceed the 18 to 22 per cent above noted, while it is equally probable that with soft coal the percentage will not exceed from 8 to 10 per cent. It is, of course, a simple matter to increase the combustion of the small sizes of anthracite by the use of a fan or a steam blast. A fan blast uses from 2£ to 3 per cent of the steam produced in the boilers, while the steam blast, used for injecting air into a closed ash-pit, consumes from 7£ to 12 per cent of the steam produced by the boilers, and seldom operates under less than 10 per cent. Hence, in making any estimates as to the relative costs of operating with these fuels, these deductions must be made if an artificial draft must be used, in order to get net comparative results. Given semi-bituminous and small-sized anthracite coals of the ash com- positions noted above, my experience has shown that the relation between the costs of operating the plant with these coals, under natural draft, to produce a given output, are : Per Ton. Semi-bituminous coal $1.33 No. 1 buckwheat coal 1.00 No. 2 buckwheat (rice) coal .83 • Paying these prices, the costs for power under natural draft are the same, ho matter which coal is used, provided the cost of removing ashes is ignored. If the anthracite grades have to be burned with blasts, the relative prices which one can afford to pay for producing a given quantity of steam are as follows : Draft. Natural. Fan Blast. Steam Blast. Semi-bituminous .... No. 1 buckwheat ..... No. 2 buckwheat (rice) . . . $1.33 ::: $0.97 .82| $0.90 .76£ Semi-bituminous coals are burned to advantage only by exercising great care in the handling of fires, and by the firemen exerting themselves beyond what is necessary when burning buckwheat and rice anthracite grades. 1356 STEAM. Wood as Fuel. Green wood contains from 30 to 50 per cent of moisture. After about a year in open air the moisture is 20 to 25 per cent. The woods of various trees are nearly identical in chemical composition, which is practically as follows, showing the composition of perfectly dry wood, and of ordinary firewood holding hygroscopic moisture : Carbon . . Hydrogen Oxygen Nitrogen , Ash . . Hygrometric water Desiccated Wood. 50 per cent 6 per cent 41 per cent 1 per cent 2 per cent 100 per cent Ordinary Firewood. 37.5 per cent 4.5 per cent 30.75 per cent 0.75 per cent 1.5 per cent 75.0 per cent 25.0 per cent 100.0 Some of the pines and others of the coniferous family contain hydrocar- bons (turpentine). Ash varies in American woods from .03 per cent to 1.20 per cent. In steam boiler tests wood is assumed as 0.4 the value of the same weight of coal. The fuel value of the same weights of wood of all kinds is practically the same ; and it is important that the wood be dry. Weig-nt of Wood* per Cord. Weighs per Cord, Lbs. Equal in value to Coal, in Lbs. Average pine Poplar, chestnut, elm Beech, red and black oak .... White oak 2000 2350 3250 3850 4500 800 to 925 940 to 1050 1300 to 1450 1540 to 1715 Hickory and hard maple .... 1800 to 2000 A cord of wood = 4x4x8= 128 cubic feet. About 56 per cent is solid wood, and 44 per cent spaces. liquid Fuels. • Petroleum is a hydrocarbon liquid which is found in abundance in Amer- ica and Europe. According to the analysis of M. Sainte-Claire Deville, the composition of 15 petroleums from different sources was found to be practi- cally the same. The average specific gravity was .870. The extreme and the average elementary compositions were as follows : Chemical Composition of Petroleum. Carbon 82.0 to 87.1 per cent. Hydrogen 11.2 to 14.8 per cent. Oxygen 0.5 to 5.7 per cent. Average, 84.7 per cent. Average, 13.1 per cent. Average, 2.2 per cent. 100.0 The total heating and evaporative powers of one pound of petroleum hav- ing this average composition are as follows : Total heating power = 145 [84.7 + (4.28 x 13.1)] = 20411 units. Evaporative power : evaporating at 212°, water supplied at 62° = 18.29 lbs. Evaporative power : evaporating at 212°, water supplied at 212° = 21.13 lbs. Petroleum oils are obtained in great variety by distillation from petro- leum. They are compounds of carbon and hydrogen, ranging from C 10 H 24 to C 32 H 64 ; or, in weight ; FUEL. 1357 Chemical Composition of Petroleum Oils. -nw™ (71.42 Carbon ) ^ i 73.77 Carbon . . . 72.60 *rom { 28.58 Hydrogen f lo \ 26.23 Hydrogen . . 27.40 100.00 100.00 100.00 The specific gravity ranges from .628 to .792. The boiling point ranges from 86° to 495° F. The total heating power ranges from 28087 to 26975 units of heat ; equivalent t£> the evaporation, at 212°, of from 25.17 to 24.17 lbs. of water supplied at 62°, or from 29.08 lbs. to 27.92 lbs. of water supplied at 212°. Puraacei for the combustion of oil fuel need not be as large as when burning coal, as the latter, being solid matter, requires more time for de- composition, and the elimination of the products and supporters of com- bustion. Coal fuel requires a large fire chamber and the means for the introduction of air beneath the grate-bars to aid combustion. Compared with oil, the combustion of coal is tardy, and requires some aid by way of a strong draft. Oil having no ash or refuse, when properly burned, requires much less space for combustion, for the reason that, being a liquid, and the compound of gases that are highly inflammable when united in proper pro- portions, it gives off heat with the utmost rapidity, and at the point of igni- tion is all ready for consumption. Prof. J. E. Denton has made a number of boiler evaporative tests, using oil for fuel. In the following table the results of tests where various fuels were used are brought together, and interesting comparisons are made be- tween the cost of coal and cost of oil. See " Power," Feb., 1902. Gaseous !Fuels. — Mr. Emerson McMillin (Am. Gas. Lt. Asso., 1887) made an exhaustive investigation of the subject of fuel gas ; he states that the relative values of these gases, considering that of natural gas as of unit value, are: By Volume. Natural gas . Coal gas . . Water gas Producer gas The water gas rated in the above table is the gas obtained in the decom- position of steam by incandescent carbon, and does not attempt to fix the calorific value of illuminating water gas, which may be carbureted so as to exceed, when compared by volume, the value of coal gas. Composition of Gases. Hydrogen . . . Marsh gas . . Carbonic oxide defiant gas . . Carbonic acid . Nitrogen . . . Oxygen . . . Water vapor Sulphydric acid Natural Gas. 2.18 92.60 0.50 0.31 0.26 3.61 0.34 0.00 0.20 100.00 Volume. Coal Gas. 46.00 40.00 6.00 4.00 0.50 1.50 0.50 1.50 100.00 Water Gas. 45.00 2.00 45.00 0.00 4.00 2.00 0.50 1.50 100.00 Producer Gas. 6.00 3.00 23.50 0.00 1.50 65.00 0.00 1.00 "lOO.OO 1358 STEAM. - a o « a •51 2- SC3 B - CO o o t> ^5 i-i CO £ a? o © © o3«S •3-Q co 3 * © w £+» a b ^g^'cl 5 aj-j a |sfl tsisf © O >>M Cj _, C3-; © w © eS ^ © rj O s«sf o *g jh © £ © 4a B o ©©*£©£ • © 2 o B^ fe ' © u* s* 5 • > ©^ *» X O a sh o ««^-2 • & o ©£ ►fe— - 4 >© j- OQ 2 ^ * g bpg, £*!;3 o *»a o^^- 1 ci*>-i o o ^ a ©S ^2 4a d d d d q n ii th to t» GO 05 QO OS d d ~3 s d d 3 £ d d 8 S d d d O ri n 00 5h 8 CO GO © © © CO ^5 O O o o © © *"* ,H © co © ii S 8 g 8 § CN CN CO CO S S FUEL. 1359 Mechanical Stoking*. In boiler installations that can be conveniently handled by one man it is doubtful if we can improve on the best hand tiring; but where good firemen are scarce, or the installation is of considerable size, it is probable that the use of some form of mechanical stoker will result in economy, and especially in the prevention of large quantities of smoke, as the combustion is gradual and more nearly perfect. The types may perhaps be limited to three : the straight feed, as the Mur- phy, Honey, Wilkinson, and Brightman ; the under-feed of which the 11 American " is a good representative ; and the chain stoker, by Coxe and the B. & W. Co. Mechanical draught is generally used with the two last-mentioned types, and sometimes with the first. Mr. Eckley B. Coxe developed the chain stoker in the most scientific man- ner for the use of the cheap coals of the anthracite region. The advantages and disadvantages of mechanical stokers are stated by Mr. J. M. Whitham (Trans. A.S.M.E., vol. xvii. p. 558) to be as follows : Advantages. 1. Adaptability to the burning of the cheapest grades of fuel. 2. A 40 per cent labor saving in plants of 500 or more h. p., when provided with coal-handling machinery. 3. Economy in combustion, even under forced firing, with proper management. 4. Constancy and uniformity of furnace conditions, the fires being clean at all times, and responding to sudden de- mands made for power. This should result in prolonged life of boilers. 5. SmokelessnesSc Disadvantages. 1. High first cost, varying from $25 to $40 per square foot of grate area. 2. High cost of repairs per year, which, with some stokers, is as much as $5 per square foot. 3. The dependence of the power-plant upon the stoker engine's working. 4. Steam cost of run- ning the stoker engine, which is from £ to § of 1 per cent of the steam generated. This is about $50 a year on a 10-hour basis for 1000 h. p., where fuel is $2 per ton. 5. Cost of steam used for a steam blast, or for driving a fan blast, whenever either is used. This, for a steam blast, is from 5 per cent to 11 per cent of the steam generated by the boilers, and from 3 per cent to 5 per cent for a fan blast. This amounts to about $1000 per year for a steam blast, and $500 a year in fuel for a fan blast, for a 1000 h. p. plant on a 10-hour basis, when fuel is $2 per ton. 6. Skill required to operate the stoker. Careless management causes either loss of fuel in the ash, or loss due to poor combustion when the coal is too soon burned out on the grate, thus per- mitting cold air to freely pass through the ash. 7. The stoker is a machine subject to a severe service, and, like any other machine, wears out and requires constant attention. W. W. Christie, in article in the Engineering Magazine on the " Economy of Mechanical Stoking," says in part : The influence of the mechanical stoker upon boiler efficiency has been discussed, but definite information is not readily obtained, although general opinions as to the advantage of mechanical stoking are numerous. The efficiency of a boiler, and consequently of a group of boilers, depends upon several independent and distinct factors. Thus we have the furnace efficiency, a measure of the completeness of the combustion in the furnace ; this is measured by the ratio of the tem- perature in the furnace to the temperature of the escaping gases. We have also the efficiency of the boiler proper, measured by the quantity of heat transmitted to the water compared with that generated in the fur- nace. There are also two other kinds of efficiencies — one the heat efficiency, per pound of fuel, the other the so-called " investment efficiency," which takes into account the cost of building, apparatus, boilers, chimneys, wages, and fuel. It has been maintained that the most economical rate for steam-making is that of an evaporation of 4 lbs. of steam per hour per square foot of heating surface, which some tests will show is the case. Other tests, however, show that it may vary, while the steam economy referred to 1 lb. of coal may remain constant. The completeness of combustion can be told best by the temperature of the escaping gases, and by an analysis of their chemical composition. Thus, for an excellent combustion, the temperature of discharge gases should not be higher than 400-500° F. If the percentage of oxygen is 1.5 1360 STEAM. to 2 per cent, it indicates that the fires are too thick, and the rate of com- bustion too high for the draft employed. If the oxygen exceeds 8 per cent, the fires are too thin, the draft too heavy, or too much cold air is entering the furnace above the fire. If there is an excess of CO and of O, the boiler is faulty in design, and good results cannot be expected. The quantity of air fed to the fire also influences the economy of the boiler to a limited degree. Per ,J*P p p fcis •p ® O CS .% p<£ ® ft'+H S^-P J~ P*«w CD H £ w H ? w H £ w H w 32 62.42 0. 41 62.42 9. 50 62.41 18. 59 62.38 27.01 33 62.42 i. 42 62 42 10. 51 62.41 19. 60 62.37 28.01 34 62.42 2. 43 62.42 11. 52 62.40 20. 61 62.37 29.01 35 62.42 3. 44 62.42 12. 53 62.40 21.01 62 62.36 30.01 36 62.42 4. 45 62.42 13. 54 62.40 22.01 63 62.36 31.01 37 62.42 5. 46 62.42 14. 55 62.39 23,01 64 62.35 32.01 38 62.42 6. 47 62.42 15. 56 62.39 24.01 65 62.34 33.01 39 62.42 7. 48 64.41 16. 57 62.39 25.01 66 62.34 34.02 40 62.42 8. 49 62.41 17. 58 62.38 26.01 67 62.33 35.02 WATER. 1361 Weight of Water — Continued . MS m Ms CO M S CD Ms 50 S.Q *j • 5.Q *J ft£w> ~& ■^ |2& ~& -»-» s< tub ~P P 2 h -3 p a p bfliH o P as .,-H CD O eS s 3 © •H ® O ?3 •H fllO "eS £ *3 •7! X a P M P£ CD CD Q CD P U O ££ O Vp CD Q !> ftftfr CD H j> O - ftft&H S> ftft.^ ^a^i 212 59.71 280 57.90 350 55.52 420 52.86 490 50.03 220 59.64 290 57.59 360 55.16 430 52.47 500 49.61 230 59.37 300 57.26 370 54.79 440 52.07 510 49.20 240 59.10 310 56.93 380 54.41 450 51.66 520 48.78 250 58.81 320 56.58 390 54.03 460 51.26 530 48.36 260 58.52 330 56.24 400 53.64 470 50.85 540 47.94 270 58.21 340 55.88 410 53.26 480 50.44 550 47.52 1362 STEAM. Expansion of Water. (Kopp : corrected by Porter.) Cent. Fahr. Volume. Cent. Fahr. Volume. Cent. Fahr. Volume. 4° 39.2° 1.00000 35° 95° 1.00586 70° 158° 1.02241 5 41 1.00001 40 104 1.00767 75 167 1.02548 10 50 1.00025 45 113 1.00967 80 176 1.02872 15 59 1.00083 50 122 1.01186 85 185 1.03213 20 68 1.00171 55 131 1.01423 90 194 1.03570 25 77 1.00286 60 140 1.01678 95 203 1.03943 30 86 1.00425 65 149 1.01951 100 212 1.04332 Water for Boiler Feed.* (Hunt andClapp, A. I.M. E., 1888.) Water containing more than 5 parts per 100,000 of free sulphuric or nitric acid is liable to cause serious corrosion, not only of the metal of the boiler itself, but of the pipes, cylinders, pistons, and valves with which the steam comes in contact. The total residue in water used for making steam causes the interior lin- ings of boilers to become coated, and often produces a dangerous hard scale, which prevents the cooling action of the water from protecting the metal against burning. Lime and magnesia bicarbonates in water lose their excess of carbonic acid on boiling, and often, especially when the water contains sulphuric acid, produce, with the other solid residues constantly being formed by the evaporation, a very hard and insoluble scale. A larger amount than 100 parts per 100,000 of total solid residue will ordinarily cause troublesome scale, and should condemn the water for use in steam boilers, unless a bet- ter can not be obtained. The following is a tabulated form of the causes of trouble with water for steam purposes, and the proposed remedies, given by Prof. L. M. Norton. CAUSES OF INCRUSTATION. 1. Deposition of suspended matter. 2. Deposition of deposed salts from concentration. 3. Deposition of carbonates of lime and magnesia by boiling off carbonic Acid, which holds them in solution. 4. Deposition of sulphates of lime, because sulphate of fime is but slightly soluble in cold water, less soluble in hot water, insoluble above 270° F. 5 Deposition of magnesia, because magnesium salts decompose at high temperature. 6. Deposition of lime soap, iron soap, etc., formed by saponification of grease. MEANS FOR PREVENTING INCRUSTATION. 1. Filtration. 2. Blowing off. 3. Use of internal collecting apparatus or devices for directing the circu- lation. 4. Heating feed-water. * See also " Boiler Waters; Scale, Corrosion, Foaming" by W. Wallace Christie. WATER. 1363 5. Chemical or other treatment of water in boiler. 6. Introduction of zinc into boiler. 7. Chemical treatment of water outside of boiler. TABULAR VIEW. Troublesome Substance. Trouble. Sediment, mud, clay, etc. Incrustation. Readily soluble salts. " Bicarbonates of lime, magnesia, ) iron. j Sulphate of lime. ** C ^ium d6 and Sulphate ° f magne - } Corrosion. Carbonate of soda in large) T» M . wi amounts. J ?™™S- Acid (in mine waters). Corrosion. Dissolved carbonic acid and oxy- ) gen. } Grease (from condensed water). Organic matter (sewage). Organic matter. Priming. Corrosion. Remedy or Palliation. Filtration, Blowing off. Blowing off. ( Heating feed. Addition of \ caustic soda, lime, or ^ magnesia, etc. S Addition of carb. soda, ( barium chloride, etc. ( Addition of carbonate of \ soda, etc. (Addition of barium chlo- ( ride, etc. Alkali. {Heating feed. Addition of caustic soda, slacked lime, etc. ( Slacked lime and filtering, < Carbonate of soda. (^ Substitute mineral oil. (Precipitate with alum or \ ferric chloride and filter- Ditto. Solubilities of Scale-making* Materials. (" Boiler Incrustation,'* F. J. Rowan.) The salts of lime and magnesia are the most common of the impurities found in water. Carbonate of lime is held in solution in fresh water by an excess of carbonic acid. By heating the water the excess of carbonic acid is driven off and the greater part of the carbonate precipitated. At ordi- nary temperatures carbonate of lime is soluble in from 16,000 to 24,000 times its volume of water ; at* 212° F. it is but slightly soluble, and at 290° F. (43 lbs. pressure) it is insoluble. The solubility of sulphate of lime is also affected by the temperature ; according to Regnault, its greatest solubility is at 95° F., where it dissolves in 393 times its weight of water ; at 212° F. it is only soluble in 460 times its weight of water, and according to M. Coute, it is insoluble at 290° F. Carbonate of magnesia usually exists in much smaller quantity than the salts of lime. The effect of temperature on its solubility is similar to that of carbonate of lime. Prof. R. H. Thurston, in his " Manual of Steam Boilers," p. 261, states that: The temperatures at which calcareous matters are precipitated are : Carbonate of lime between 176° and 248° F. Sulphate of lime between 284° and 424° F. Chloride of magnesium between 212° and 257° F. Chloride of sodium between 324° and 364° F. 1364 STEAM. 11 Incrustation and sediment," Prof . Thurston says, " are deposited in boilers, the one by the precipitation of mineral or other salts previously held in solution in the feed-water, the other by the deposition of mineral insoluble matters, usually earths, carried into it in suspension or me- chanical admixture. Occasionally also vegetable matter of a glutinous nature is held in solution in the feed-water, and, precipitated by heat or concentration, covers the heating-surfaces with a coating almost impermea- ble to heat, and hence liable to cause an over-heating that may be very dan- gerous to the structure. A powdery mineral deposit sometimes met with is equally dangerous, and for the same reason. The animal and vegetable oils and greases carried over from the condenser or feed-water heater are also very likely to cause trouble. Only mineral oils should be permitted to be thus introduced, and that in minimum quantity. Both the efficiency and the safety of the boiler are endangered by any of these deposits. "The only positive and certain remedy for incrustation and sediment once deposited is periodical removal by mechanical means, at sufficiently frequent intervals to insure against injury by too great accumulation. Be- tween times, some good may be done by special expedients suited to the individual case. No one process and no one antidote will suffice for all cases. 11 Where carbonate of lime exists, sal-ammoniac may be used as a pre- ventive of incrustation, a double decomposition occurring, resulting in the production of ammonium carbonate and calcium chloride — both of which are soluble, and the first of which is volatile. The bicarbonate may be in part precipitated before use by heating to the boiling-point, and thus break- ing up the salt and precipitating the insoluble carbonate. Solutions of caustic lime and metallic zinc act in the same manner. Waters containing tannic acid and the acid juices of oak, sumach, logwood, hemlock, and other woods, are sometimes employed, but are apt to injure the iron of the boiler, as may acetic or other acid contained in the various saccharine matters often introduced into the boiler to prevent scale, and which also make the lime-sulphate scale more troublesome than when clean. Organic matters should never be used. 4 * The sulphate scale is sometimes attacked by the carbonate of soda, the products being a soluble sodium sulphate and a pulverulent insoluble cal- cium carbonate, which settles to the bottom like other sediments and is easily washed off the heating-surfaces. Barium chloride acts similarly, producing barium sulphate and calcium chloride. All the alkalies are used at times to reduce incrustations of calcium sulphate, as is pure crude petro- leum, the tannate of soda, and other chemicals. 44 The effect of incrustation and of deposits of various kinds is to enor- mously reduce the conducting power of heating-surfaces ; so much so, that the power, as well as the economic efficiency of a boiler, may become very greatly reduced below that for which it is rated, and the supply of steam furnished by it may become wholly inadequate to the requirements of the case. 44 It is estimated that a sixteenth of an inch thickness of hard 4 scale ' on the heating-surface of a boiler will cause a waste of nearly one-eighth its efficiency, and the waste increases as the square of its thickness. The boil- ers of steam vessels are peculiarly liable to injury from this cause where using salt water, and the introduction of the surface-condenser has been thus brought about as a remedy. Land boilers are subject to incrustation by the carbonate and other salts of lime, and by the deposit of sand or mud mechanically suspended in the feed- water." Kerosene oil ("Boiler Incrustation," Rowan) has been used to advantage in removing and preventing incrustation. From extended experiments made on a 100 h. p. water tube boiler, fed with water containing 6.5 grains of solid matter per gallon, it was found that one quart kerosene oil per day was sufficient to keep the boiler entirely free from scale. Prior to the in- troduction of the kerosene oil, the water had a corrosive action upon some of the fittings attached to the boiler ; but after the oil had been used for a few months it was found that the corrosive action had ceased. It should be stated, however, that objection has been made to the intro- duction of kerosene oil into a boiler for the purpose of preventing incrusta- WATER. 1365 tion, on account of the possibility of some of the oil passing with the steam into the cylinder of the engine, and neutralizing the effect of the lubricant in the cylinder. When oil is used to remove scale from steam-boilers, too much care can- not be exercised to make sure that it is free from grease or animal oil. Nothing but pure mineral oil should be used. Crude petroleum is one thing ; black oil, which may mean almost anything, is very likely to be something quite different. The action of grease in a boiler is peculiar. It does not dissolve in the water, nor does it decompose, neither does it remain on top of the water ; but it seems to form itself into " slugs," which at first seem to be slightly lighter than the water, so that the circulation of the water carries them about at will. After a short season of boiling, these " slugs," or suspended drops, acquire a certain degree of " stickiness," so that when they come in contact with shell and flues of the boiler, they begin to adhere thereto. Then under the action of heat they begin the process of " varnishing " the interior of the boiler. The thinnest possible coating of this varnish is suf- ficient to bring about over-heating of the plates. The time when damage is most likely to occur is after the fires are banked, for then, the formation of steam being checked, the circulation of water stops, and the grease thus has an opportunity to settle on the bottom of the boiler and prevent contact of the water with the fire-sheets. Under these circumstances, a very low degree of heat in the furnace is sufficient to over- heat the plates to such an extent that bulging is sure to occur. Zinc as a Scale Preventive. — Dr. Corbigny gives the following hypoth- esis : he says that " the two metals, iron and zinc, surrounded by water at a high temperature, form a voltaic pile with a single liquid, which slowly decomposes the water. The liberated oxygen combines with the most oxy- dizable metal, the zinc, and its hydrogen equivalent is disengaged at the surface of the iron. There is thus generated over the whole extent of the iron influenced a very feeble but continuous current of hydrogen, and the bubbles of this gas isolate at each instant the metallic surface from the scale-forming substance. If there is but little of the latter, it is penetrated by these bubbles and reduced to mud ; if there is more, coherent scale is produced, which, being kept off by the intervening stratum of hydrogen, takes the form of the iron surface without adhering to it." Zinc, in the shape of blocks, slabs, or as shavings inclosed in a perforated vessel, should be suspended throughout the water space of a boiler, care being used in getting perfect metallic contact between the zinc and the boiler. It should not be suspended directly over the furnace, as the oxide might fall upon the surface and be the cause of the plate being over-heated. The quantity placed in a boiler should vary with the hardness of the water, and the amount used, and should be measured by the surface presented. Generally one square inch of surface for every 50 lbs. water in the boiler is sufficient. The British Admiralty recommends the renewing of the blocks whenever the decay of the zinc has penetrated the slab to a depth of J inch below the surface. Purification of Feed-Water by Boiling 1 . Sulphates can be largely removed from feed-water by heating it to the tem- perature due to boiler pressure in a feed-water heater, or " live steam puri- fier" before introduction to boiler. This precipitates those salts in the heater and the water can then if necessary be pumped through a filter into the boiler. The feed-water U first heated as hot as possible in the ordinary exhaust feed-water heater in which the carbonates are precipitated, and then run through the purifier, which is most generally a receptacle containing a number of shallow pans, that can be removed for cleaning, over which the feed- water is allowed to flow from one to the other in a thin sheet. Live steam at boiler-pressure is introduced into the purifier, heating the water to a temperature high enough to precipitate the salts which form scale on the pans. This method of treating feed-water is said to largely increase the efficiency of a boiler plant by the almost complete avoidance of scale. Purification of feed-water by filtration before introduction to the system is often practised with good results. 1366 STEAM. Tattle of Water Analyses. Grains per U. S. Gallon of 231 Cubic Inches. Where From. Buffalo, N. Y., Lake Erie .... Pittsburgh, Allegheny River . . , Pittsburgh, Monongahela River . Pittsburgh, Pa., artesian well . . Milwaukee, Wisconsin River . . . Galveston, Texas, 1 Galveston, Texas, 2 Columbus, Ohio Washington, D. C, city supply . . Baltimore, Md., city supply . . . Sioux City, la., city supply .... Los Angeles, Cal., 1 Los Angeles, Cal., 2 Bay City, Michigan, Bay Bay City, Michigan, River .... Cincinnati, Ohio River Watertown, Conn Fort Wayne, Ind Wilmington, Del Wichita, Kansas Springfield, 111., 1 Springfield, 111., 2 Hillsboro, 111 Pueblo, Colo. Long Island City, L. I Mississippi River, above Missouri River Mississippi River, below mouth of Missouri River Mississippi River at St. Louis, W. W. Hudson River, above Poughkeepsie, N. Y, Croton River, above Croton Dam N. Y Croton River water from service pipes in New York City. .... Schuylkill River, above Philadelphia Pa 5.66 0.37 1.06 23.45 6.23 13.68 21.79 20.76 2.87 2.77 19.76 10.12 3.72 8.47 4.84 3.88 1.47 8.78 10.04 14.14 12.99 5.47 14.56 4.32 4.0 8.24 10.64 9.64 1.06 3.32 3.78 5.12 5.71 4.67 13.52 29.15 11.74 3.27 0.65 1.24 5.84 12.59 10.36 33.66 0.78 4.51 6.22 6.02 25.91 7.40 4.31 2.97 16.15 28.0 1.02 7.41 6.94 4.57 2.36 2.16 .16 .29 0.58 0.58 0.64 18.41 1.76 326.64 398.99 7.02 Trace Trace 1.17 3.51 20'.48 126.78 1.79 1.76 3.51 4.29 24.34 1.97 1.56 2.39 1.20 16.0 0.50 1.36 1.54 .11 .40 .49 0.37 0.78 1.04 20.14 Trace V.58 0.36 0.10 1.03 2.63 0.76 1.15 3.00 Trace 1.59 8.48 *2.'l9 4.28 1.63 1.97 1.22 1.57 10.76 1.92 1.36 1.30 0.18 1.50 3.20 0.82 6.50 Trace 4.00 6.50 2.10 3.80 4.40 4.10 6.00 8.74 10.92 Trace 1.78 10.98 6.17 2.00 8.62 5.83 Trace 5.12 1.0 5.25 15.86 9.85 .77 .67 9.74 6.60 10.80 49.43 39.30 353.84 453.93 46.60 8.60 7.30 27.60 26.20 23.07 49.20 179.20 6.73 9.52 31.08 35.00 66.39 33.17 21.45 21.55 28.76 39.0 15.01 36.49 29.54 12.70 7.72 3.72 4.24 pumps. 1367 PUlttPS. Feed-Pnmpg. These should be at least double the capacity fouud by calculation from the amount of water required for the engines, to allow for blowing off, leak- age, slip in the pumps themselves, etc., and to enable the pump to keep down steam in case of sudden stoppage of the engines when the fires hap- pen to be brisk, and in fact should be large enough to supply the boilers when run at their full capacity. In addition, for all important plants, there should be either a duplicate feed-pump or an injector to act as stand-by in case of accident. The speed of the plunger or piston may be 50 feet per minute and should never exceed 100 feet per minute, else undue wear and tear of the«valves results, and the efficiency is reduced. If the pump be re- quired to stand idle without continually working, the plunger or piston and rod should be of brass. If D — diameter of barrel in inches, S = stroke in inches, n = number of useful strokes per minute, w = cubic feet of water pumped per hour, W=: lbs. of water pumped per hour ; w = 1.7Mw. ~" 36.6 * If Sn—60, JF=:1.36Z>*, and 1.36 Rubber valves may be used for cold water, but brass, rubber composition, or other suitable material is required for hot water or oil. If a new pump will not start, it may be due to its imperfect connections or temporary stiffness of pump. Unless the suction lift and length of supply pipe be moderate, afoot-valve, a charging connection, and a vacuum chamber are desirable. The suction- pipe must be entirely free from air leakage. If the pump refuses to start lifting water with full pressure on, on account of the air in the pump-cham- ber not being dislodged, but only compressed each stroke, arrange for run- ning without pressure until the air is expelled and water flows. This is done with a check-valve in the delivery-pipe, and a waste delivery which may be closed when water flows. Pumping- Hot Water. — With a free suction-pipe, any good pump fitted with metal valves and with hot-water packing will pump water hav- ing a temperature of 212°, or higher, if so placed that the water will flow into it. Robert D. Kinney, in "Power," gives the following formula for deter- l 111 ?. 1 ]! 8 ,. what height water of temperatures below the boiling point can be lifted by suction. D — lift in feet, A = absolute pressure on surface of water ; if open to air = 14.7 lbs. B and W= constants. See table. B= l*U-^> x 0.8 = 115.2^—8 W W 1368 STEAM. Water Temp. B. W. Water Temp. B. W. Degrees F. Degrees F. 40 0.122 62.42 130 2.215 61.56 50 0.178 62.41 140 2.879 61.39 60 0.254 62.37 150 3.708 61.20 70 0.360 62.31 160 4.731 61.01 80 0.503 62.22 170 5.985 60.80 90 0.693 62.12 180 7.511 60.59 100 0.942 62.00 190 9.335 60.37 110 1.267 61.87 200 11.526 60.13 120 1.685 61.72 210 14.127 59.89 Speed of Water througrh Pump- Passages and Valves. The speed of water flowing through pipes and passages in pumps varies from 100 to 200 feet per minute. The loss from friction will be considerable if the higher speed is exceeded. The area of valves should be sufficient to permit the water to pass at a speed not exceeding 250 feet per minute. The amount of steam which an average engine will require per indicated horse-power is usually taken at 30 pounds. It varies widely, however, from about 12 pounds in the best class of triple expansion condensing engines up to considerably over 90 pounds in many direct-acting pumps. Where an engine is overloaded or underloaded more water per horse-power will be re- quired than when operated at rated capacity. Horizontal tubular boilers will evaporate on an average from 2 to 3 pounds of water per square foot heating-surface per hour, but may be forced up to 6 pounds if the grate sur- face is too large or the draught too great for economical working. Sizes of Direct-acting* Pumps. The two following tables are selected as representing the two common types of direct-acting pump, viz., the single-cylinder and the duplex. Efficiency of Small Direct-acting* Pumps. In ''Reports of Judges of Philadelphia Exhibition," 1876, Group xx., Chas. E. Emery says : " Experiments made with steam-pumps at the Amer- ican Institute Exhibition of 1867 showed that average size steam-pumps do not, on the average, utilize more than 50 per cent of the indicated power in the steam cylinders, the remainder being absorbed in the friction of the en- gine, but more particularly in the passage of the water through the pump. Again, all ordinary steam-pumps for miscellaneous use, require that the steam-cylinder shall have three to four times the area of the water-cylinder to give sufficient power when the steam is accidentally low ; hence, as such pumps usually work against the atmospheric pressure, the net or effective pressure forms a small percentage of the total pressure, which, with the large extent of radiating surface exposed and the total absence of expansion, makes the expenditure of steam very large. One pump tested required 120 pounds weight of steam per indicated horse-power per hour, and it is be- lieved that the cost will rarely fall below 60 pounds ; and as only 50 per cent of the indicated power is utilized, it may be safely stated that ordinary steam pumps rarely require less than 120 pounds of steam per hour for each horse-power utilized in raising water, equivalent to a duty of only 15,000.000 foot pounds per 100 pounds of coal. With larger steam-pumps, particularly when they are proportioned for the work to be done, the duty will be mate- rially increased. PUMPS. 1369 Single-Cylinder Direct-acting- Pump. (Standard Sizes for ordinary service.) ft I! Capacity per Minute at Given Speed. O "So x Diameter of 3^ 4 4 5 5| 7 7i 8 6 7 8 10 8 10 12 10 10 12 12 14 10 10 10 12 12 12 14 16 16 14 14 16 16 18 16 18 20 18 20 22 .14 .27 .39 .51 .72 1.64 1.91 2.17 1.47 2.00 2.61 4.08 2.61 4.08 5.87 4.08 6.12 5.87 8.80 12.00 4.08 6.12 8.16 5.87 8.80 11.75 15.99 13.92 20.88 12.00 15.99 13.92 20.88 26.43 20.88 26.43 32.64 26.43 32.64 39.50 300 300 300 275 275 250 250 250 250 250 250 250 250 250 250 250 200 250 175 175 250 175 150 250 175 150 150 175 150 175 150 175 150 125 125 125 125 125 125 125 130 130 125 125 125 110 110 110 100 100 100 100 100 18 35 49 64 90 180 210 239 147 200 261 408 261 100 408 100 587 100 70 100 70 408 428 587 616 70 840 100 408 70 50 100 70 50 50 428 408 587 616 587 800 80 1114 50 1044 70 840 50 800 80 1114 50 1044 50 1322 50 1044 50 1322 50 1632 50 1322 50 1632 50 1975 33 33 45i 45£ 45i 58 58 58 67 67 68i 68* 64 68^ 64 93 112 112 112 84 112 89 109 85 115 115 118 118 118 118 118 120 9h 15 15 15 17 17 17 20£ 20* 30" 20-t 30 30 30 24 30 28£ 28* 28* 30" 25 26 30 28i 26 34 34 38 27 34 34 34 40 38 40 40 40 40 40 1 1 1 1 1 1 1 1* 1* 1* 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 1 1 1 1* i| i* i* if 2 2 2 2* 2£ 2* 2* 2§ 2* 2| 2£ 2* 2£ 2i 2* 2* 2* 2* 2* 2£ 3* 3* 3i 3* 3j 8 10 12 12 12 8 12 12 12 14 12 14 16 14 16 18 1370 STEAM. Duplex-Cylinder Direct-acting 1 Pump. (Standard sizes for ordinary service.) & a © d d-d ~ o . .3'd'd T* c $ Sizes of Pipes for © © © M4 o^2 ~ J2 © £ O ft •3 ft«2 5.1 © Short Lengths. To be Increased as d d d © *d 2 © §*>js St3 0*3 S ©C-l 3 Length Increases. *>> o s © fe|* ©CO I- £ " S oa co CM O M • ^S © S ° 3 d © Is «M o M . © OD II l-l ©"" M o M CO © cS o ft£ .22 CO M *« 2 © * 9 QD ,«^ © 5s m £ © r> 1s£ -S d © ©^ go* o,4a £ 2 «3 ©©*> ©3£ It* «w ©»^ O bi)© .3 §5 © s a © © s OB d © d o © d CO © ft © Sh 08 ,d © 00 p ft Hi A di O ft CO m s 3 2 3 .04 100 to 250 8 to 20 % t * 11 1 4* 2f 4 .10 100 " 200 20" 40 4 1 2 1* 5| 3* 5 .20 100 " 200 40" 80 5 1 i| 2* ll 6 4 6 .33 100 " 150 70" 100 5f 1 i* 3 2 7* 4* 6 .42 100 " 150 85 " 125 6f l£ 2 4 3 7* 5 6 .51 100 " 150 100" 150 7 If 2 4 3 7* 4* 10 .69 75 " 125 100" 170 3 l£ 2 4 3 9 5i 10 .93 75 " 125 135" 230 2 2* 4 3 10 6 10 1.22 75 " 125 180" 300 8* 2 2* 5 4 10 7 10 1.66 75 " 125 245" 410 9f 2 2* 6 5 12 7 10 1.66 75 " 125 245" 410 91 2* 3 6 5 14 7 10 1.66 75 " 125 245" 410 »i 2* 3 6 5 12 1 10 2.45 75 " 125 365" 610 12 2* 3 6 5 14 10 2.45 75 " 125 365" G10 12 2* 3 6 5 16 8* 10 2.45 75 " 125 365" 610 12 2* 3 6 5 18* 8* 10 2.45 75 " 125 365" 610 12 3 3* 6 5 20 8* 10 2.45 75 " 125 365" 610 12 4 5 6 5 12 10i 10i 10 3.57 75 " 125 530" 890 14£ 14i lif 141 2* 3 8 7 14 10 3.57 75 " 125 530" 890 2* 3 8 7 16 101 10 3.57 75 " 125 530" 890 2* 3 8 7 18* ioi 10 3.57 75 " 125 530" 890 3 3* 8 7 20 m 10 3.57 75 " 125 530" 890 14i 4 5 8 7 14 12 10 4.89 75 " 125 730 " 1220 17 2* 3 10 8 16 12 10 4.89 75 " 125 730 " 1220 17 2* 3 10 8 18* 12 10 4.89 75 " 125 730 " 1220 17 3 3* 10 8 20 12 10 4.89 75 " 125 730 " 1220 17 4 5 10 8 18* 14 10 6.66 75 " 125 990 " 1660 19| 3 3* 12 10 20 14 10 6.66 75 u 122 990 " 1660 19| 4 5 12 10 17 10 15 5.10 50 " 100 510 " 1020 14 3 3* 10 8 20 12 15 7.34 50 " 100 730 " 1460 17 4 5 12 10 20 15 15 15 15 11.47 11.47 50 " 100 50 " 100 1145 " 2290 1145 " 2290 21 21 25 Let HTEdORS. liive Steam Injectors. W= water injected in pounds her hour. P — steam pressure in pounds per square inch- Z>zr diameter of throat in inches. T=z diameter of throat in millimeters. INJECTORS. 1371 Then JF=1280Z) 2 Vp The rule given by Rankine, " Steam Engine," p. 477, for finding the proper sectional area in square inches for the narrowest part of the nozzle is as follows : cubic feet per hour gross feed-water area = ,, 800 ^pressure in atmospheres The expenditure of steam is about one-fourteenth the volume of water Injected. The following table gives the water delivered for different sizes of injec- tors at different pressures ; but when the injector has to lift its water a de- duction must be made varying from 10 to 30 per cent according to the lift. Deliveries for Live Steam Injectors. Pressure of Steam. 1 S 08 O 00 ® 1-1 G*rZ M 2 30 lbs. 60 lbs. 80 lbs. 100 lbs. 120 lbs. 140 lbs. .pH 00 «mS 67h 235| 163| 91* 58* 40* 23 20 872 * 90 314 218 122 78 54 30| 25 1 L090 i >12* 392* 272* 152* 97* 67* 38* 30 '35 451 327 183 117 81 46 35 i £7* 549* 381* 213* 136* 94* 53§ 40 ( )80 628 436 244 156 108 61* 45 1] L02* 706* 490* 274* 175* 121* 69 50 785 545 305 195 135 76§ 75 . . 1177| 817J 457* 292* 202* 115 100 1090 610 380 270 153* 125 762* 487* 337* 191| 150 915 585 405 230 175 . . . 1067* 682* 4721 268* 200 1220 780 540 306§ 1374 STEAM. Table Giving' I.os* in Pressure due to Friction, in Pounds per Square Inch, for Pipe lOO Feet Long;. (By G. A. Ellis, C. E.) Gallons Dis- charged per Min. | in. lin. l£in. l£m. 2 in. 2^ in. 3 in. 4 in. 5 10 15 20 25 30 35 40 45 50 75 100 125 * 150 175 200 3.3 13.0 28.7 50.4 78.0 ( " ( r 1! 2 3 4 ).84 $.16 }.9S 2.3 ).0 '.5 -.0 5.0 0.31 1.05 2.38 4.07 6.40 9.15 12.4 16.1 20.2 24.9 56.1 » • • 0.12 0.47 0.97 1.66 2.62 3.75 5.05 6.52 8.15 10.0 22.4 39.0 6.12 6.42 6.91 1.60 2.44 5.32 9.46 14.9 21.2 28.1 37.5 6.21 6.8l' 1.80 3.20 4.89 7.0 9.46 12.47 6.10 6.35 0.74 1.31 1.99 2.85 3.85 5.02 6.09 6.33 6.69 1.22 IjOss of Head due to Bends. Bends produce a loss of head in the flow of water in pipes. Weisbach gives the following formula for this loss : H—f ~ where H= loss of head in feet, / = coefficient of friction, v =z ve- locity of flow in feet per second, g = 32.2. As the loss of head or pressure is inmost cases more conveniently stated in pounds per square inch, we may change this formula by multiplying by 0.433, which is the equivalent in pounds per square inch for one foot head. If P = loss in pressure in pounds per square inch, F~ coefficient of fric- tion. v 2 P = F -xt-*} v being the same as before. From this formula has been calculated the following table of values for F, corresponding to various exterior angles, A. A — F — 20° 0.020 40° 0.060 45° 0.079 60° 0.158 80° 0.320 90° 0.426 100° 0.546 110° 0.674 120° 0,806 130° 0.934 This applies to such short bends as are found in ordinary fittings, such as 90° and 45° Ells, Tees, etc. A globe valve will produce a loss about equal to two 90° bends, a straight- way valve about equal to one 45° bend. To use the above formula find the speed p. second, being one-sixtieth of that found in Table p. 1373 ; square this speed, and divide the result by 64.4; multiply the quotient by the tabular value ofF corresponding to the angle of the turn t A. For instance, a 400 h.p. battery of boilers is to be fed through a 2-inch pipe. Allowing for fluctuations we figure 40 gallons per minute, making 244 feet per minute speed, equal to a velocity of 4.6 per second. Suppose our pipe is in all 75 feet long ; we have from Table No. 36, for 40 gallons per minute, 1.60 pounds loss ; for 75 feet we have only 75 per cent of this = 1.20 pounds. Suppose we have 6 right-angled ells, each giving Fz=z 0.426. We have then 4.06 X 4.06 = 16.48 ; divide this by 64.4 = 0,256. Multiply this by F=z 0.42C FEED WATER HEATERS. 1375 pounds, and as there are 6 ells, multiply again by 6, and we have 6 x 0.426 x 256 = 0.654. The total friction in the pipe is therefore 1.20 -{- 0.654 = 1.854 pounds per square inch. If the boiler pressure is 100 pounds and the water level in the boiler is 8 feet higher than the pump suction level, we have first 8 X 433 = 3.464 pounds. The total pressure on the pump plunger then is 100 + 3.464 -4- 1.854= 105.32 pounds per square inch. If in place of 6 right- angled ells we had used three 45° ells, they would have cost us only 3 X 0.079 = 0.237 pounds ; 0.237 X 0.256 = 0.061. The total friction head would have been 1.20 + 0.061 = 1.261, and the total pressure on the plunger 100 -f- 3.464 + 1.261 = 104.73 pounds per square inch, a saving over the other plan of nearly 0.6 pounds. To be accurate, we ought to add a certain head in either case, " to produce the velocity." But this is very small, being for velocities of ; 2 . 3 . 4 ; 5 ; 6 ; 8 ; 10 ; 12 and 18 feet per sec. 0.027 ; 0.061 ; 0.108; 0.168; 0.244; 0.433; 0.672; 0.970 and 2.18 lbs. per sq. in. Our results should therefore have been increased by about 0.11 pounds. It is usual, however, to use larger pipes, and thus to materially reduce the frictional losses. Feed Water Heaters. (W. W. Christie.) Feed Water Heaters may be classified in this way : ( Steam tube. Closed Heaters (indirect) . . . . ° . . . { Water tube. j AtmospheriCo Open Heaters (direct) j Vacuum. The open heater is usually made of cast iron, as this material will with- stand the corrosive action of acids found in feed-waters better than any other metal. In this type of heater the exhaust steam from engines and pumps, and the feed-water broken up into drops by suitable means, are brought into immediate contact, and the steam not condensed in heating the water passes off to the atmosphere. The quantity of water that can be heated is only limited by the amount of steam and water that can be brought together. The steam condensed in heating the water is saved and utilized for boiler feed. An open heater should be provided with an effi- cient oil-separator, a large settling-chamber or hot well in which, if desired, a filtering bed of suitable material can be placed to insure the removal from the water, of all the impurities held in suspension, a device for skim- ming the surface of the water to remove the impurities floating on the water, and a large blow-off opening placed at the lowest point in the heater. The closed heater is made with a wrought-iron or steel cylindrical shell and cast- or wrought-iron heads, having iron or brass tubes inside, 6et in tube plates so as to make steam- and water-tight joints, provision being made for the expansion and contraction of the tubes. According to the particular design of the heater, the exhaust steam passes through or around the tubes, the water being on the opposite of the walls of the tubes. The steam and water are separated by metal through which the heat of the exhaust steam is transmitted to the water. As an oil-separator is very seldom attached to a closed heater, the steam condensed in heating the water is wasted. The quantity of water that can be heated is limited by the amount of heat that can be transmitted through the tubes. The efficiency of heat transmission is decreased by the coating of oil that covers the steam side, and the crust of scale that coats the water side of the tubes. No provision can be made for purifying the water in a closed heater, as the constant circulation of the water prevents the impurities from settling. The impurities that are in the water pass on into the boiler. Purification must be done by means of an auxiliary apparatus. When used with a condenser, the feed water heater will increase tne vacuum 1 to 2 inches ; when used with cold feed water, the economy is in- creased from 7 to 14 per cent ; if feed water is from a hot well, 7 to 8 per cent. Two things are very essential to the successful working of all heaters,— they must be kept clean from scale and oil deposits, and sufficient exhaust steam must be sent through them. The probability of there being much scale ingredients thrown down in a closed heater where temperature never exceeds 212° F., and in an open heater where temperature approaches more nearly to steam temperature, is shown By this table 1376 STEAM. Temperatures at which scale-forming ingredients are precipitated : Carbonate of lime 176°-248° F. Chloride of magnesium 212°-257° F. Sulphate of lime 284° F. -424° F. Chloride of sodium 324° F. -364° F. The rating of a feed-water heater of the closed type is a subject about which little has been written, but the common rule is to give | square feet of heating surface for one boiler horse-power. In designing, however, the heating surface should be made large enough or ample to transmit the maximum number of heat units per unit of time, and then the water velocity should be adjusted to suit the capacity required. For heat transmitted, one well-known manufacturer uses 350 B. T. Units per degree F. difference of temperature per square foot of heating surf ace per hour, as a maximum ; other types of heaters would use only 150 to 200 B. T. TJ.'s as the maximum. As the tubes forming the heating surface in closed heaters are made of different materials, if we take Copper as 100 "Wrought iron as . . .58 Brass as 87 Cast iron as .... 49 we can readily see that if one-third square foot surface area is right for a copper pipe, we will need Vs° of i or t?£i or about six-tenths for iron coils, per boiler horse-power. The power to transmit heat varies not only with the material, but also with the design of the heater, the velocity of the water, and water and steam capacity of the heater. The velocity of the water through the heater should be from 100 to 200 feet per minute. The proportions of open heaters depend largely upon the character of the water used in the heater, for it should have sufficient time to become thor- oughly heated and the scale-forming ingredients settled and eliminated from the feed as it passes out of the heater. B.T.U. PER DEG. F. DIF. TEMP. PER SQ. FT. SURFACE PER HOUR. ° S | § 8 S ill! < 100 S 5 S 150 H -n'l^M >• 'V, JsX 1 h'' ^0c ^> 'ftp ki s*e ->^£< £^ FT *Yv G INDICATE PLAIN TUBES H.EAT ABSORPTION CURVES • " -CORRUGATED TUBES Fig. 6. Saving* by Heating- Feed- Water. (W. W. Christie.) In converting water at 32° F. into steam at atmospheric pressure, it must be raised to 212° F., the boiling point. The specific heat of water varies somewhat with its temperature, so that to raise a pound of water from 32° to 212° F. or 180° F., requires 180.8 heat units. To convert it into steam, after it has reached 212° F., requires 965.8 heat Units, or in all 180.8 + 965.8=1146.6 units of heat, thermal units. The saving to be obtained by the use of waste heat, as exhaust steam, heating the water by transfer of some of its heat through metal walls, is calculated by this formula: PUMP EXHAUST. 1377 Gain in per cent = in which H 100 (h 2 — h j _ 100 (t 2 — t x ) very nearly, H—h x ~ tf-^ + 32 : total heat in steam at boiler pressure (above that in water at 32° F.) in B. T. U. h 2 = heat in feed-water (above 32° F.) after heating. h x =z heat in feed- water (above 32° F.) before heating. t 2 =. temperature of feed-water after heating °F. t t = temperature of feed-water before heating °F. given H= 1146.6, t 2 =: 212, ^ =: 112, or a difference of 100°; and we obtain by use of the above formula, gain in per cent =z 9.37, or for 10° approximately .937 per cent, for 11° 1.03 per cent, so we may say that for every 11° F. added to the feed-water temperature by use of the exhaust steam, 1 per cent of fuel saving results. The table which follows is taken from " Power." Percentage of Saving* in Fuel hy Heating- Feed-Water by Waste Steam, Steam at «0 Pounds Oaug-e Pressure. Temperature of Water Entering Boiler. is 5 '2 g 120° 130° 140° 150° 160° 170° 180° 190° 1 200° 1 210° 220° 250° 35° 7.24 8.09 8.95 9.89 10.66 11.52 12.38 13.24 14.09 14.95 15.81 19.40 40° 6.84 7.69 8.56 9.42 10.28 11.14 12.00 12.87 13.73 14.59 15.45 18.89 45° 6.44 7.30 8.16 9.03 9.90 10.76 11.62 12.49 13.36 14.22 15.09 18.37 50° 6.03 6.89 7.76 8.64 9.51 10.38 11.24 12.11 12.98 13.85 14.72 17.87 55° 5.63 6.49 7.37 8.24 9.11 9.99 10.85 11.73 12.60 13.48 14.35 17.38 60° 5.21 6.08 6.96 7.84 8.72 9.60 10.47 11.34 12.22 13.10 13.98 16.86 65° 4.80 5.67 6.56 7.44 8.32 9.20 10.08 10.96 11.84 12.72 13.60 16.35 70° 4.38 5.26 6.15 7.03 7.92 8.80 9.68 10.57 11.45 12.34 13.22 15.84 75° 3.96 4.84 5.73 6.62 7.51 8.40 9.28 10.17 11.06 11.95 12.84 15.33 80° 3.54 4.42 5.32 6.21 7.11 8.00 8.88 9.78 10.67 11.57 12.46 14.82 85° 3.11 4.00 4.90 5.80 6.70 7.59 8.48 9.38 10.28 11.18 12.07 14.32 90° 2.68 3.58 4.48 5.38 6.28 7.18 8.07 8.98 9.88 10.78 11.68 13.81 95° 2.25 3.15 4.05 4.96 5.86 6.77 7.66 8.57 9.47 10.38 11.29 13.31 100° 1.81 2.71 3.62 4.53 5.44 6.35 7.25 8.16 9.07 9.98 10.88 12.80 Pump Exhaust. In many plants the only available exhaust steam comes from the steam pumps used for elevator service, boiler-feeding, etc. ; or in condensing plants from the air-pumps, water-supply, and boiler feed-pumps. It should also be remembered that all direct-acting steam pumps are large consumers of steam, taking several boiler h. p. for each indicated h. p., and that the ex- haust steam from them will heat about six times the same quantity by weight of cold water, from 50° to 212° F., and that these pumps, or the independent condenser pumps, are more economical when all the exhaust from them is used for heating feed-water than the best kind of triple expansion condens- ing engines. With the pumps all the heat not used in doing work can be conserved and returned to the boiler in the feed-water, whereas even with triple expansion engines at least 80 per cent of the total heat in the steam is carried away in the condensing water. While the supply of exhaust from these pumps may not be sufficient to raise the temperature to the highest point, yet the saving is large and con- stant. These results do not take any account of the purifying action in the "open" heaters on the feed-water, the improved condition of which, by di- minishing the average deposit within the boiler, materially increases both Vhq boiler capacity and the economy ; while the more uniform temperature 1378 STEAM. accompanying the use of a hot feed reduces the repairs and lengthens the life of all boilers. If the quantity of water passing through the heater is only what is re- quired to furnish steam for the engine from which the exhaust comes, more than four-fifths of this exhaust steam will remain uncondensed, and will thus become available for other purposes, such as heating buildings, dryer systems, etc. ; in which case the returns can be sent back to the boiler by suitable means. Il'EI JECOWOIMZEMS. Performance of a Green Economizer with a Smoky Coal. (D. K. Clark, S. E., p, 286.) From tests by M. W. Grosseteste, covering a period of three weeks on a Green economizer, using a smoke-making coal, with a constant rate of com- bustion under the boilers, it is apparent that there is a great advantage in cleaning the pipes daily — the elevation of temperature having been in- creased by it from 88° to 153°. In the third week, without cleaning, the ele- vation of temperature relapsed in three days to the level of the first week ; even on the first day it was quickly reduced by as much as half the extent of relapse. By cleaning the pipes daily an increased elevation of tempera- ture of 65° F. was obtained, whilst a gain of 6% was effected in the evapora- tive efficiency. The action of Green's economizer was tested by M. W. Grosseteste for a period of three weeks. The apparatus consists of four ranges of vertical pipes, 6£ feet high, 3| inches in diameter outside, nine pipes in each range, connected at top and bottom by horizontal pipes. The water enters all the tubes from below, and leaves them from above. The system of pipes is enveloped in a brick casing, into which the gaseous products of combustion are introduced from above, and which they leave from below. The pipes are cleared of soot externally by automatic scrapers. The capacity for water is 24 cubic feet, and the total external heating-surface is 290 square feet. The apparatus is placed in connection with a boiler having 355 square feet of surface. Green's Economizer. — Results of Experiments on its Efficiency as Affected by the State of the Surface. (W. Grosseteste.) Temperature of Feed- Temperature of Gas- water. eous Products. Time. February and March. Enter- Leav- Enter- Leav- ing ing Differ- ing ing Differ- Feed- Feed- ence. Feed- Feed- ence. heater. heater. heater. heater. Fahr. Fahr. Fahr. Fahr. Fahr. Fahr. 1st Week 73.5° 161.5° 88.0° 849° 261° 588° 2d Week 77.0 230.0 153.0 882 297 585 3d Week — Monday . . 73.4 196.0 122.6 831 284 547 Tuesday . . 73.4 181.4 108.0 871 309 562 Wednesday 79.0 178.0 99.0 Thursday . 80.6 170.6 90.0 952 329 623 Friday . . 80.6 169.0 88.4 889 338 551 Saturday 79.0 172.4 93.4 901 351 550 1st Week. Coal consumed per hour 214 lbs. Water evaporated from 32° F. per hour 1424 Water per pound of coal ...... 6.65 2d Week. 3d Week. 216 lbs. 1525 7.06 213 lbs. 1428 6.70 FUEL ECONOMIZERS. 1379 The Fuel Economizer Company, Matteawan, N.Y., describe the construc- tion of Green's economizer, thus: The economizer consists of a series of sets of cast-iron tubes about 4 inches in diameter and 9 feet in length, made in sections (of various widths) and connected by " top " and " bottom headers," these again being coupled by " top " and " bottom branch pipes " running lengthwise, one at the top and the other at the bottom, on opposite sides and outside the brick chamber which encloses the apparatus. The waste gases are led to the economizer by the ordinary flue from the boilers to the chimney. The feed-water is forced into the economizer by the boiler pump or in- jector, at the lower branch pipe nearest the point of exit of gases, and emerges from the economizer at the upper branch pipe nearest the point where the gases enter. Each tube is provided with a geared scraper, which travels continuously up and down the tubes at a slow rate of speed, the object being to keep the external surface clean and free from soot, a non-conductor of heat. The mechanism for working the scrapers is placed on the top of the econ- omizer, outside the chamber, and the motive power is supplied either by a belt from some convenient shaft or small independent engine or motor. The power required for operating the gearing, however, is very small. The apparatus is fitted with blow-off and safety valves, and a space is pro- vided at the bottom of the chamber for the collection of the soot, which is removed by the scrapers. One boiler plant equipped with the Green economizer gave, under test, these results. The total area of heating surface in the plant was 3,126 square feet, and the number of tubes in the economizer 160. The results were as follows: — Particulars of Test. 1. Duration of test hours 2. Weight of dry coal consumed lbs. 3. Percentage of ash and refuse . . . per cent 4. Weight of coal consumed per hour per square foot grate surface lbs. 5. Weight of water evaporated lbs. 6. Horse-power developed on basis of 30 lbs. per h.p. fed at 100° and evaporated at 70 lbs., h.p. 7. Average boiler pressure (above atmosphere), lbs. 8. Average temperature of feed-water entering economizer deg. Fahr. 9. Average temperature of feed-water entering boilers deg. Fahr. 10. Number of degrees feed-water was heated by economizer deg. Fahr. 11. Average temperature of flue gases entering economizer deg. Fahr. 12. Average temperature of flue gases entering chimney deg. Fahr. 13. Number degrees flue gases were cooled by econ- omizer deg. Fahr. 14. Lbs. water evaporated per lb. of coal, as ob- served 15. Equivalent evaporation per lb. of coal from and at 212° 16. Percentage gained by using the economizer per cent The steam in this test contained 1.3 per cent of moisture. 1380 STEAM. W. S. Hntton gives the following results of tests of a steam boiler with and without an economizer. With Econ- omizer. Duration of test, hours Weight of coal, pounds Steam pressure, pounds Temp, water entering economizer, degrees " " " boiler, degrees . . Degrees feed-water heated by economizer Temp, gases entering economizer, degrees " " " chimney, degrees . Degrees gases cooled by economizer . . Evaporation per lb. coal, from and at 212°, pounds Saving by economizer, per cent . . . 7856 58 88 225 137 618 365 253 10.613 28.9 Without Econo- mizer. Hi 10282 57 ' 85 645 8.235 Green's I?uel Economizer. — Clark gives the following average re- sults of comparative trials of three boilers at Wigan used with and without economizers : Without With Economizers. Economizers. Coal per square foot of grate per hour . . . 21.6 21.4 Water at 100° evaporated per hour .... 73.55 79.32 Water at 212° per pound of coal 9.60 10.56 Showing that in burning equal quantities of coal per hour the rapidity of evaporation is increased 9.3% and the efficiency of evaporation 10% by the addition of the economizer. The average temperature of the gases and of the feed-water before and after passing the economizer were as follows : With 6-ft. grate. Before. After. Average temperature of gases . . • 649 340 Average temperature of feed-water . 47 157 Taking averages of the two grates, to raise the temperature of the feed- water 100°, the gases were cooled down 250°. With 4-ft. grate. Before. After. 501 312 41 137 §TEAM SEPARATORS. Carefully conducted experiments have shown that water, oil, or other liquids passing through pipes along with steam do not remain thoroughly mixed with the stea^i itself, but that the major portion of these liquids fol- lows the inner contour of the pipe, especially in the case of horizontal pipes. From this it would necessarily follow that a rightly designed separator to meet these conditions must interrupt the run of the liquid by breaking the continuity of the pipe, and offering a receptacle into which the liquid will flow freeiy, or fall by gravity — that this appliance must further otter the opportunity for the liquid to come to rest out of the current of steam, for it is not enough to simply provide a well or a tee in the pipe, since the current would jump or draw the liquid over this opening, especially if the velocity was high. It is also evident that means must be provided in this appliance for inter- rupting the progress of those particles of the liquid which are traveling in the current of the steam, and do this in such a way that these particles will STEAM SEPARATORS. 1381 also be detained and allowed to fall into the receptacle provided, which receptacle must be fully protected from the action of the current of the steam ; otherwise, the separated particles of water or oil will be picked up and carried on past the separator. To prevent the current from jumping the liquid over the well, and to interrupt the forward movement of those particles traveling in or with the current, it follows that some obstruction must be interposed in the path of the current. Steam separators should always be placed as near as possible to the steam inlet to the cylinder of the engine. Oil separators are placed in the run of the exhaust pipe from engines and pumps, for the purpose of removing the oil from the steam before it is used in any way where the presence of oil would cause trouble. Prof. R. C. Carpenter conducted a series of tests on separators of several makes in 1891. The following table shows results under various conditions of moisture : 4 Test with Steam of about 10% of Moisture. Tests with Varying Moisture. ®o* Quality of Steam Before. Quality of Steam After. Efficiency per cent. Quality of Steam Before. Quality of Steam • After. Average Efficiency. B A D C E F 87.0% 90.1 89.6 90.6 88.4 88.9 98.8% 98.0 95.8 93.7 90.2 92.1 90.8 80.0 59.6 33.0 15.5 28.8 66.1 to 97.5% 51.9 " 98 72.2 " 96.1 67.1 " 96.8 68.6 " 98.1 70.4 " 97.7 97.8 to 99 % 97.9 "99.1 95.5 " 98.2 93.7 " 98.4 79.3 " 98.5 84.1 " 97.9 87.6 76.4 71.7 63.4 36.9 28.4 Conclusions from the tests were : 1. That no relation existed between the volume of the several separators and their efficiency. 2. No marked decrease in pressure was shown by any of the separators, the most being 1.7 lbs. in E. 3. Although changed direction, reduced velocity, and perhaps centrifugal force are necessary for good separation, still some means must be provided to lead the water out of the current of the steam. A test on a different separator from those given above was made by Mr. Charles H. Parker, at the Boston Edison Company's plant, in November, 1897, and the following results obtained : Length of run 3-4 hrs. Average pressure of steam 158 lbs. per sq. in. Temperature of upper thermometer in calorimeter on outlet of separator 368.5° F. Temperature of lower thermometer in calorimeter on outlet of separator 291.7° F. Normal temperature of lower thermometer, when steam is at rest 292.9° F. Degrees cooling as shown by lower thermometer . . . 1.2° F. Moisture in steam delivered by separator as shown by cooling of lower thermometer 06 per cent. Water discharged from separator per hour 52 lbs. Steam and entrained water passing through engine, as shown by discharge from air pump of surface con- denser 7359 lbs. Steam and entrained water entering separator .... 7411 lbs. Moisture taken out by separator 72 Total moisture in steam (.06 plus .72) 78 per cent. Efficiency of separator 92.3 per cent. 1382 STEAM. SAFETY VAIVES. Calculation of Weight, etc., for liever Safety- Valve. Let W = weight of ball at end of lever, in pounds ; w =z weight of lever itself, in pounds ; V=z weight of valve and spindle, in pounds ; L = distance between fulcrum and center of ball, in inches ; I =: distance between fulcrum and center of valve, in inches ; g = distance between fulcrum and center of gravity of lever, in inches; A = area of valve, in square inches ; P — pressure of steam, in pounds per square inch at which valve will open. Then PA x I = T V X L + w x g + V X I ; D WL_±wg+Vl . whence P =: —jf ; PAl-wg-Vl . W ~ L ■ T _ P Al — wg—Vl L ~ W Example. — Diameter of valve, 4 inches ; distance from fulcrum to center of ball, 36 inches ; to center of valve, 4 inches ; to center of gravity of lever, 16 inches ; weight of valve and spindle, 6 lbs. ; weight of lever, 10 lbs. ; re- quired the weight of ball to make the blowing-olf pressure 100 lbs. per square inch ; area of 4-inch valve = 12.566 square inches. Then „ PAl — wg—Vl 100X12.566X4 — 10X16 — 6X4 W= =^ = — =5 = 134.5 lbs. Rules Governing* Safety- Valve*. (Rule of U. S. Supervising Inspectors of Steam-vessels as amended 1894.) The distance from the fulcrum to the valve-stem must in no case be less than the diameter of the valve-opening ; the length of the lever must not be more than ten times the distance from the fulcrum to the valve-stem ; the width of the bearings of the fulcrum must not be less than three-quarters of an inch ; the length of the fulcrum -link must not be less than four inches; the lever and fulcrum-link must be made of wrought iron or steel, and the knife-edged fulcrum points and the bearings for these points must be made of steel and hardened ; the valve must be guided by its spindle, both above and below the ground seat and above the lever, through supports either made of composition (gun-metal) or bushed with it ; and the spindle must fit loosely in the bearings or supports. Lever safety-valves to be attached to marine boilers shall have an area of not less than 1 square inch to 2 square feet of the grate surface in the boiler, and the seats of all such safety-valves shall have an angle of inclina- tion of 45° to the center line of their axes. Spring-loaded safety-valves shall be required to have an area of not less than 1 square inch to 3 square feet of grate surface of the boiler, except as hereinafter otherwise provided for water-tube or coil and sectional boilers, and each spring-loaded valve shall be supplied with a lever that will raise the valve from its seat a distance of not less than that equal to one-eighth the diameter of the valve-opening, and the seats of all such safety-valves shall have an angle of inclination to the center line of their axes af 45°. All spring-loaded safety-valves for water-tube or coil and sectional boilers required to carry a steam-pressure exceeding 175 lbs. per square inch shall be required to have an area of not less than 1 square inch to 6 square feet of the grate surface of the boiler. Nothing herein shall be construed so as to prohibit the use of two safety-values on one water-tube or coil and sectional boiler, provided the combined areax>f such valves is equal to that required by rule for one such valve. SAFETY VALVES. 1383 Rule on Safety-Valves in Philadelphia Ordinances.— Every boiler when fired separately, and every set or series of boilers when placed over one fire, shall have attached thereto, without the interposition of any other valve, two or more safety-valves, the aggregate area of which shall have such relations to the area of the grate and the pressure within the boiler as is expressed in schedule A. Schedule A. — Least aggregate area of safety-valve (being the least sec- tional area for the discharge of steam) to be placed upon all stationary boilers with natural or chimney draught (see note a). 22.5 G P-r-8.62' in which A is area of combined safety-valves in inches ; G is area of grate in square feet ; P is pressure of steam in pounds per square inch to be carried in the boiler above the atmosphere. The following table gives the results of the formula for one square foot of grate, as applied to boilers used at different pressures : Pressures per square inch : 10 20 30 40 50 60 70 80 90 100 110 120 150 175 Yalve area in square inches corresponding to one square foot of grate : 1.2 .79 .58 .46 .38 .33 .29 .25 .23 .21 .19 .17 .14 .12 [Note a.] — Where boilers have a forced or artificial draught, the inspec- tor must estimate the area of grate at the rate of one square foot of grate surface for each 16 lbs. of fuel burned on the average per hour. The various rules given to determine the proper area of a safety-valve do not take into account the effective discharge area of the valve. A correct rule should make the product of the diameter and lift proportional to the weight of steam to be discharged. Mr. A. G. Brown (The Indicator and its Practical Working) gives the fol- lowing as the lift of the lever safety-valve for 100 lbs. gauge pressure. Tak- ing the effective area of opening at 70 per cent of the product of the rise and the circumference 3 3£ 4 4J 5 6 .0507 .0492 .0478 .0462 .0446 .043 For " pop " safety-valves, Mr. Brown gives the following table for the rise, effective area, and quantity of steam discharged per hour, taking the effective area at 50 per cent of the actual on account of the obstruction which the lip of the valve offers to the escape of the steam. Diameter of valve, inches 2 1\ Rise of valve, inches . . .0583 .0523 Di. valve in. 1 n 2 2i 3 3^- 4 U 5 6 Lift inches. .125 .150 .175 .200 .225 .250 .275 .300 .325 .375 Area, sq. in. .196 .354 .550 .785 1.061 1.375 1.728 2.121 2.553 3.535 Gauge- press. Steam discharged per hour, lbs. 30 lbs. 474 856 1330 1897 2563 3325 4178 5128 6173 8578 50 669 1209 1878 2680 3620 4695 5901 7242 8718 12070 70 861 1556 2417 3450 4660 6144 7596 9324 11220 15535 90 1050 1897 2947 4207 5680 7370 9260 11365 13685 18945 100 1144 2065 3208 4580 6185 8322 10080 12375 14895 20625 120 1332 2405 3736 5332 7202 9342 11735 14410 17340 24015 140 1516 2738 4254 6070 8200 10635 13365 16405 19745 27340 160 1696 3064 4760 6794 9175 11900 14955 18355 22095 30595 180 1883 3400 5283 7540 10180 13250 16595 20370 24520 33950 200 2062 3724 5786 8258 11150 14465 18175 22310 26855 37185 If we also take 30 lbs. of steam per hour, at 100 lbs. gauge-pressure — h.p., we have from the above table : Diameter inches .11&2 2£3 3|4 4£&6 Horse-power . . 38 69 107 153 206 277 336 412 496 687 1384 STEAM. A boiler having ample grate surface and strong draft may generate double the quantity of steam its rating calls for ; therefore in determining the proper size of safety-valve for a boiler this fact should be taken into consideration and the effective discharge of the valve be double the rated steam-producing capacity of the boiler. The Consolidated Safety-valve Co.'s circular gives the following rated capacity of its nickel-seat " pop " safety-valves : Size, in . . 1 1* 1* 2 2* 3 3* I 4 4* 5 5* Boiler } from H.P. \ to 8 10 20 35 60 75 100 325 150 175 200 10 15 30 50 75 100 125 150 175 200 275 rxtjles for (o\im(ti\c; boiler tests. The Committee of the A. S. M. E. on Boiler-tests recommended the fol- lowing revised code of rules for conducting boiler trials. (Trans, vol. xx. See also p. 34, vol. xxi, A. S. M. E., for latest code. Code of 1897. Preliminaries to a Trial. jT. Determine at the outset the specific object of the proposed trial, whether it be to ascertain the capacity of the boiler, its efficiency as a steam gener- ator, its efficiency and its defects under usual working conditions, the econ- omy of some particular kind of fuel, or the effect of changes of design, proportion, or operation ; and prepare for the trial accordingly. II. Examine the boiler, both outside and inside ; ascertain the dimensions of grates, heating surfaces, and all important parts ; and make a full record, describing the same, and illustrating special features by sketches. The area of heating surfaces is to be computed from the outside diameter of water-tubes and the inside diameter of fire-tubes. All surfaces below the mean water level which have water on one side and products of combustion on the other are to be considered water-heating surface, and all surfaces above the mean water level which have steam on one side and products of combustion on the other are to be considered as superheating surface. III. Notice the general condition of the boiler and its equipment, and record such facts in relation thereto as bear upon the objects in view. If the object of the trial is to ascertain the maximum economy or capa- city of the boiler as a steam generator, the boiler and all its appurtenances should be put in first-class condition. Clean the heating surface inside and outside, remove clinkers from grates and from sides of the furnace. Re- move all dust, soot, and ashes from the chambers, smoke connections, and flues. Close air leaks in the masonry and poorly-fitted cleaning-doors. See that the damper will open wide and close tight. Test for air leaks by firing a few shovels of smoky fuel and immediately closing the damper, observing the escape of smoke through the crevices, or by passing the flame of a can- dle over cracks in the brickwork. IV. Determine the character of the coal to be used. For tests of the effi- ciency or capacity of the boiler for comparison with other boilers the coal should, if possible, be of some kind which is commercially regarded as a stan- dard. For New England and that portion of the country east of the Allegheny Mountains, good anthracite egg coal, containing not over 10 per cent of ash, and semi-bituminous Clearfield (Pa.), Cumberland (Md.), and Pocahontas (Va.) coals are thus regarded. West of the Allegheny Mountains, Poca- hontas (Va.), and New River (W. Va.) semi-bituminous, and Youghiogheny or Pittsburg bituminous coals are recognized as standards.* There is no special grade of coal mined in the Western States which is widely recog- nized as of superior quality or considered as a standard coal for boiler test- ing. Big Muddy Lump, an Illinois coal mined in Jackson County, 111., is * These coals are selected because they are about the only coals which con- tain the essentials of excellence of quality, adaptability to various Hinds of furnaces, grates, boilers, and methods of firing, and wide distribution and general accessibility in the markets. RULES FOR CONDUCTING BOILER TESTS. 1385 suggested as being of sufficiently high grade to answer the requirements in districts where it is more conveniently obtainable than the other coals men- tioned above. For tests made to determine the performance of a boiler with a particular kind of coal, such as may be specified in a contract for the sale of a boiler, the coal used should not be higher in ash and in moisture than that speci- fied, since increase in ash and moisture above a stated amount is apt to cause a falling off of both capacity and economy in greater proportion than the proportion of such increase. V. Establish the correctness of all apparatus used in the test for weighing and measuring. These are : 1. Scales for weighing coal, ashes, and water. 2. Tanks, or water meters for measuring water. Water meters, as a rule, should only be used as a check on other measurements. For accurate work, the water should be weighed or measured in a tank. 3. Thermometers and pyrometers for taking temperatures of air, steam, feed-water, waste gases, etc. 4. Pressure gauges, draft gauges, etc. The kind and location of the various pieces of testing apparatus must be left to the judgment of the person conducting the test; always keeping in mind the main object, i.e., to obtain authentic data. VI. See that the boiler is thoroughly heated before the trial to its usual working temperature. If the boiler is new and of a form provided with a brick setting, it should be in regular use at least a week before the trial, so as to dry and heat the walls. If it has been laid off and become cold, it should be worked before the trial until the walls are well heated. VII. The boiler and connections should be proved to be free from leaks before beginning a test, and all water connections, including blow and extra feed pipes, should be disconnected, stopped with blank flanges, or bled through special openings beyond the valves, except the particular pipe through which water is to be fed to the boiler during the trial. During the test the blow-off and feed-pipes should remain exposed. If an injector is used, it should receive steam directly through a felted pipe from the boiler being tested.* If the water is metered after it passes the injector, its temperature should be taken at the point at which it enters the boiler. If the quantity is deter- mined before it goes to the injector, the temperature should be determined on the suction side of the injector, and if no change of temperature occurs other than that due to the injector, the temperature thus determined is properly that of the feed-water. When the temperature changes between the injector and the boiler, as by the use of a heater or by radiation, the temperature at which the water enters and leaves the injector and that at which it enters the boiler should all be taken. The final temperature cor- rected for the heat received from the injector will be the true feed-water temperature. Thus if the injector receives water at 50° and delivers it at 120° into a heater which raises it to 210°, the corrected temperature is 210 — (120 — 50) =140°. See that the steam main is so arranged that water of condensation can- not run back into the boiler. VIII. Starting and Stopping a Test. — A test should last at least ten hours of continuous running, but, if the rate of combustion exceeds 25 pounds of coal per square foot of grate per hour it may be stopped when a total of 250 pounds of coal has been burned per square foot of grate surface. A longer test may be made when it is desired to ascertain the effect of widely vary- ing conditions, or the performance of a boiler under the working conditions °l a Prolonged run - The conditions of the boiler and furnace in all respects should be, as nearly as possible, the same at the end as at the beginning of the test. The steam pressure should be the same ; the water level the * In feeding a boiler undergoing test witH an injector talcing steam from another boiler, or the main steam pipe from several boilers, the evaporative results may be modified by a difference in the quality of the steam from such source compared with that supplied by the boiler being tested, and in some cases the connection to the injector may act as a drip for the main steam pipe. If it is known that the steam from the main pipe is of the same quality as that furnished by the boiler undergoing the test, the steam may be taken from such main pipe. 1386 STEAM. same ; the fire upon the grates should be the same in quantity and condi- tion ; and the walls, flues, etc., should be of the same temperature. Two methods of obtaining the desired equality of conditions of the fire may "be used, viz. : those which were called in the Code of 1885 " the standard method" and " the alternate method," the latter being employed where it is inconvenient to make use of the standard method. IX. Standard Method. — Steam being raised to the working pressure, remove rapidly all the fire from the grate, close the damper, clean the ash- pit, and as quickly as possible start a new fire with weighed wood and coaL, noting the time and the water level while the water is in a quiescent stated just before lighting the fire. k At the end of the test remove the whole fire, which has been burned low, clean the grates and ash-pit, and note the water level when the water is in a quiescent state, and record the time of hauling the fire. The water level should be as nearly as possible the same as at the beginning of the test. If it is not the same, a correction should be made by computation, and not by operating the pump after the test is completed. X. Alternate Method. — The boiler being thoroughly heated byaprelimi* nary run, the fires are to be burned low and well cleaned. Note the amount of coal left on the grate as nearly as it can be estimated ; note the pressure of steam and the water level, and note this time as the time of starting the test. Fresh coal which has been weighed should now be fired. The ash- pits should be thoroughly cleaned at once after starting. Before the end of the test the fires should be burned low, just as before the start, and the fires cleaned in such a manner as to leave the bed of coal of^the same depth, and in the same condition, on the grates, as at the start. The water level and steam pressures should previously be brought as nearly as possible to the same point as at the start, and the time of ending of the test should be noted just before fresh coal is fired. If the water level is not the same as at the start, a correction should be made by computation, and not by operating the pump after the test is completed. XI. Uniformity of Conditions. — In ail trials made to ascertain maximum economy or capacity, the conditions should be maintained uniformly con- stant. Arrangements should be made to dispose of the steam so that the rate of evaporation may be kept the same from beginning to end. This may be accomplished in a single boiler by carrying the steam through a waste steam pipe, the discharge from which can be regulated as desired. In a battery of boilers, in which only one is tested, the draft can be regu- lated on the remaining boilers, leaving the test boiler to work under a con- stant rate of production. Uniformity of conditions should prevail as to the pressure of steam, the height of water, the rate of evaporation, the thickness of fire, the times of firing and quantity of coal fired at one time, and as to the intervals between the times of cleaning the fires. XII. Keeping the Records. — Take note of every event connected with the progress of the trial, however unimportant it may appear. Record the time of every occurrence and the time of taking every weight and every observation. The coal should be weighed and delivered to the fireman in equal propor- tions, each sufficient for not more than one hour's run, and a fresh portion should not be delivered until the previous one has all been fired. The time required to consume each portion should be noted, the time being recorded at the instant of firing the last of each portion. It is desirable that at the same time the amount of water fed into the boiler should be accurately noted and recorded, including the height of the water in the boiler, and the average pressure of steam and temperature of feed during the time. By thus recording the amount of water evaporated by successive portions of coal, the test may be divided into several periods if desired, and the degree of uniformity of combustion, evaporation, and economy analyzed for each period. In addition to these 'records of the coal and the feed-water, half hourly observations should be made of the temperature of the feed-water, of the flue gases, of the external air in the boiler-room, of the temperature of the furnace when a furnace pyrometer is used, also of the pressure of steam, and of the readings of the instruments for determining the moisture in the steam. A log should be kept on properly prepared blanks containing columns for record of the various observations. When the "standard method" of starting and stopping the test is used, RULES FOR CONDUCTING BOILER TESTS. 1387 the hourly rate of combustion and of evaporation and: the horse-power may be computed from the records taken during the time when the fires are in active condition. This time is somewhat less than the actual time which elapses between the beginning and end of the run. This method of computation is necessary, owing to the loss of time due to kindling the fire at the beginning and burning it out at the end. XIII. Quality of Steam. — The percentage of moisture in the steam should be determined by the use of either a throttling or a separating steam calor- imeter. The sampling nozzle should be placed in the vertical steam pipe rising from the boiler. It should be made of £-inch pipe, and should extend across the diameter of the steam pipe to within half an inch of the opposite side, being closed at the end and perforated with not less than twenty £-inch holes equally distributed along and around its cylindrical surface, but none of these holes should be nearer than £ inch to the inner side of the steam pipe. The calorimeter and the pipe leading to it should be well covered with felting. Whenever the indications of the throttling or separating calorimeter show that the percentage of moisture is irregular, or occasion- ally in excess of three per cent, the results should be checked by a steam separator placed in the steam pipe as close to the boiler as convenient, with a calorimeter in the steam pipe just beyond the outlet from the separator. The drip from the separator should be caught and weighed, and the per- centage of moisture computed therefrom added to that shown by the calorimeter. Superheating should be determined by means of a thermometer placed in a mercury well inserted in the steam pipe. The degree of superheating should be taken as the difference between the reading of the thermometer for superheated steam and the readings of the same thermometer for satu- rated steam at the same pressure as determined by a special experiment, and not by reference to steam tables. XIV. Sampling the Coal and Determining its Moisture. — As each barrow load or fresh portion of coal is taken from the coal pile, a representative shovelful is selected from it and placed in a barrel or box in a cool place and kept until the end of the trial. The samples are then mixed and broken into pieces not exceeding one inch in diameter, and reduced by the process of repeated quartering and crushing until a final sample weighing about five pounds is obtained, and the size of the larger pieces is such that they will pass through a sieve with J-inch meshes. From this sample two one-quart, air-tight glass preserving jars, or other air-tight vessels which will prevent the escape of moisture from the sample, are to be promptly filled, and these samples are to be kept for subsequent determinations of moisture and of heating value, and for chemical analyses. During the process of quartering, when the sample has been reduced to about 100 pounds, a quarter to a half of it may be taken for an approximate determi- nation of moisture. This may be made by placing it in a shallow iron pan, not over three inches deep, carefully weighing it, and setting the pan in the hottest place that can be found on the brickwork of the boiler setting or flues, keeping it there for at least twelve hours, and then weighing it. The determination of moisture thus made is believed to be approximately accurate for anthracite and semi-bituminous coals, and also for Pittsburg or Youghiogheny coal ; but it cannot be relied upon for coals mined west of Pittsburg, or for other coals containing inherent moisture. For these latter coals it is important that a more accurate method be adopted. The method recommended by the Committee for all accurate tests, whatever the char- acter of the coal, is described as follows : Take one of the samples contained in the glass jars, and subject it to a thorough air-drying in a warm room, weighing it before and after, thereby determining the quantity of surface moisture it contains. Then crush the whole of it by running it through an ordinary coffee mill, adjusted so as to produce somewhat coarse grains (less than T Vinch), thoroughly mix the crushed sample, select from it a portion of from 10 to 50 grams, weigh it in a balance which will easilv show a variation as small as 1 part in 1,000, and dry it in an air or sand bath at a temperature between 240 and 280 degrees Fahr. for one hour. Weigh it and record the loss, then heat and weigh it again repeatedly, at intervals of an hour or less, until the minimum weight has been reached and the weight begins to increase by oxidation of a por- tion of the coal. The difference between the original and the minimum weight is taken as the moisture in the air-diaed coal. This moisture should (*-?) 1388 STEAM. preferably be made on duplicate samples, and the results should agree within 0.3 to 0.4 of one per cent, the mean of the two determinations being taken as the correct result. The sum of the percentage of moisture thus found and the percentage of surface moisture previously determined is the total moisture. XV. Treatment of Ashes and Refuse. — The ashes and refuse are to be weighed in a dry state. For elaborate trials a sample of the same should be procured and analyzed. XVI. Calorific Tests and Analysis of Coal. — The quality of the fuel should be determined either by heat test or by analysis, or by both. The rational method of determining the total heat of combustion is to burn the sample of coal in an atmosphere of oxygen gas, the coal to be sampled as directed in Article XIV. of this code. The chemical analysis of the coal should be made only by an expert chemist. The total heat of combustion computed from the results of the ultimate analysis may be obtained by the use of Dulong's formula (with constants modified by recent determinations), viz. : 14,600 C -\- 62,000 -J- 4,000 S, in which C, H, O, and S refer to the proportions of carbon, hydrogen, oxygen, and sulphur respectively, as determined by the ultimate analysis.* It is recommended that the analysis and the heat test be each made by two independent laboratories, and the mean of the two results, if there is any difference, be adopted as the correct figures. It is desirable that a proximate analysis should*also be made to determine the relative proportions of volatile matter and fixed carbon in the coal. XVII. Analysis of Flue Gases. — The analysis of the flue gases is an espe- cially valuable method of determining the relative value of different meth- ods of firing, or of different kinds of furnaces. In making these analyses, great care should be taken to procure average samples — since the compo- sition is apt to vary at different points of the flue. The composition is also apt to vary from minute to minute, and for this reason the drawings of gas should last a considerable period of time. Where complete determinations are desired, the analyses should be intrusted to an expert chemist. For approximate determinations the Orsat or the Hempel apparatus may be used by the engineer. XVIII. Smoke Observations. — It is desirable to have a uniform system of determining and recording the quantity of smoke produced where bitumi- nous coal is used. The system commonly employed is to express the degree of smokiness by means of percentages dependent upon the judgment of the observer. The Committee does not place much value upon a percentage method, because it depends so largely upon the personal element, but if this method is used, it is desirable that, so far as possible, a definition be given in explicit terms as to the basis and method employed in arriving at the percentage. XIX. Miscellaneous. — In tests for purposes of scientific research, in which the determination of all the variables entering into the test is de- sired, certain observations should be made which are in general unneces- sary for ordinary tests. These are the measurement of the air supply, the determination of its contained moisture, the determination of the amount of heat lost by radiation, of the amount of infiltration of air through the setting, and (by condensation of all the steam made by the boiler) of the total heat imparted to the water. As these determinations are not likely to be undertaken except by engi- neers of high scientific attainments, it is not deemed advisable to give directions for making them. XX. Calculations of Efficiency . — Two methods of defining and calculat- ing the efficiency of a boiler are recommended. They are : -. -ru« • * *.%_ *. •-, Heat absorbed per lb. combustible 1. Efficiency of the boiler = = — - „ . .. =- — rr^r • Heating value of 1 lb. combustible o -c^ • * x-u t, -i * Heat absorbed per lb. coal 2. Efficiency of the boiler and grate = — : - *1 _ .. .• Heating value of 1 lb. coal * Favre and Silberman give 14,544 B.T.U. per pound carbon; Berthelot 14,647 B. T. U. Favre anrt Silberman give 62,032 B. T. U. per pound hydro- gen ; Thomson 61,816 B. T. U. • RULES FOR CONDUCTING BOILER TESTS. 1389 The first of these is sometimes called the efficiency based on combustible, and the second the efficiency based on coal. The first ig recommended as a standard of comparison for all tests, and this is the one which is understood to be referred to when the word " efficiency " alone is used without qualifi- cation. The second, however, should be included in a report of a test, together with the first, whenever the object of the test is to determine the efficiency of the boiler and furnace together with the grate (or mechanical stoker), or to compare different furnaces, grates, fuels, or methods of firing. The heat absorbed per pound of combustible (or per pound coal) is to be calculated by multiplying the equivalent evaporation from and at 212° per pound combustible (or coal) by 965.7. (Appendix XXI.) XXI. The Heat Balance. — An approximate " heat balance," or statement of the distribution of the heating value of the coal among the several items of heat utilized and heat lost, may be included in the report of a test when analyses of the fuel and of the chimney gases have been made. It should be reported in the following form : Heat Balance, or Distribution of the Heating Value of the Combustible. Total Heat Value of 1 lb. of Combustible B. T. U. Per Cent. 1. Heat absorbed by the boiler = evaporation from and at 212° per pound of combustible X 965.7. Loss due to moisture in coal — per cent of moisture re- ferred to combustible -f 100 X [(212 — t) + 966 + 0.48 (T — 212)] (t = temperature of air in the boiler-room, T=i that of the flue gases). Loss due to moisture formed by the burning of hydro- gen = per cent of hydrogen to combustible ■—■ 100 X 9 X [(212 — t) + 966 -f 0.48 ( T — 212)]. 4.* Loss due to heat carried away in the dry chimney gases =: weight of gas per pound of combustible x 0.24 x (T—t). CO 5.f Loss due to incomplete combustion of carbons 2. 3. ~co 2 +co + per cent Cin combustible 100" X 10,150. 6. Loss due to unconsumed hydrogen and hydrocarbons, to heating the moisture in the air, to radiation, and un- accounted for. (Some of these losses may be sepa- rately itemized if data are obtained from which they may be calculated.) Totals 100.00 * The weight of gas per pound of carbon burned may be calculated from the gas analysis asfolloivs: Dry gas per pound carbon = U C ° 2 + * ff + 7 ^° + N) l in which C0 2 , CO, O, and Hare the percentages by volume of the several gases. As the sampling and analyses of the gases in the present state of the art are liable to considerable errors, the result of this calculation is usually only an approx- imate one. The heat balance itself is also only approximate for this reason, as well as for the fact that it is not possible to determine accurately the per- centage of unburned hydrogen or hydrocarbons in the flue gases. The weight of dry 'gas per pound of combustible is found by multiplying the dry gas per pounci, of carbon by the percentage of carbon in the combusti- ble, and dividing by 100. t C0 2 and CO are respectively the percentage by volume of carbonic acid and carbonic oxide in the flue gases. The quantity 10,150— No. heat units generated by burning to carbonic acid one pound of carbon contained in car' conic oxide. 1390 STEAM. XXII. Report of the Trial. — The data and results should be reported in the manner given in either one of the two following tables, omitting lines where the tests have not been made as elaborately as provided for in such tables. Additional lines may be added for data relating to the specific object of the test. The extra lines should be classified under the headings provided in the tables, and numbered, as per preceding line, with sub let- ters, a, b, etc. The Short Form of Report, Table No. 2, is recommended for commercial tests and as a convenient form of abridging the longer form for publication when saving of space is desirable. Table Ho. 1. Data and Results of Evaporative Test. Arranged in accordance with the complete form advised by the Boiler Test Committee of the American Society of Mechanical Engineers. Made by of boiler at to determine . Principal conditions | governing the trial Kind of fuel Kind of furnace State of the weather 1. Date of trial 2. Duration of trial hours Dimensions and Proportions. (A complete description of the boiler should be given on an annexed sheet.) 3. Grate surface . . . width . . . length . . . area . . sq. ft. 4. Water-heating surface " 5. Superheating surface . . - M 6. Ratio of water-heating surface to grate surface 7. Ratio of minimum draft area to grate surface Average Pressures. 8. Steam pressure by gauge lbs. 9. Force of draft between damper and boiler ins. of water 10. Force of draft in furnace " 11. Force of draft or blast in ash-pit " " Average Temperatures. 12. Of external air deg. 13. Of fireroom 14. Of steam 15. Of feed-water entering heater . . 16. Of feed-water entering economizer . 17. Of feed-water entering boiler . . . 18. Of escaping gases from boiler . . . 19. Of escaping gases from economizer Fuel. 20. Size and condition 21. Weight of wood used in lighting fire lbs. 22. Weight of coal as fired* " * Including equivalent of wood used in lighting the fire, not including un- burnt coal withdrawn from furnace at times of cleaning and at end of test. One pound of wood is taken to 'be equal to 0.4 pound of coal, or, in case greater accuracy is desired, as having a heat value equivalent to the evaporation of 6 pounds of water from and at 212° per pound (6 X 965.7 = 5,794 B. T.U.J. RULES FOR CONDUCTING BOILER TESTS. 1391 23. Percentage of moisture in coal * ... per cent. 24. Total weight of dry coal consumed . . lbs. 25. Total ash and refuse lbs. 26. Total combustible consumed 27. Percentage of ash and refuse in dry coal per cent Proximate Analysis of Coal. Of Coal. Of Combustible, 28. Fixed carbon per cent. per cent. 29. Volatile matter " " 30. Moisture " 31. Ash " 100 per cent 100 per cent. 32. Sulphur, separately determined " " Ultimate Analysis of Dry Coal. 33. Carbon (C) per cent. 34. Hydrogen (B) " 35. Oxygen (O) " 36. Nitrogen (X) " 37. Sulphur (S) " 100 per cent. 38. Moisture in sample of coal as received " Analysis of Ash and Refuse. 39. Carbon per cent. 40. Earthy matter " Fuel per Hour. 41. Dry coal consumed per hour lbs. 42. Combustible consumed per hour " 43. Dry coal per square foot of grate surface per hour ... " 44. Combustible per square foot of water-heating surface per hour " Calorific Value of Fuel. 45. Calorific value by oxygen calorimeter, per lb. of dry coal . B. T. U. 46. Calorific value by oxygen calorimeter, per lb. of combustible " 47. Calorific value by analysis, per lb. of dry coalt " 48. Calorific value by analysis, per lb. of combustible .... " Quality of Steam. 49. Percentage of moisture in steam per ceni, 50. Number of degrees of superheating deg. 51. Quality of steam (dry steam =z unity) , . . . Water. 52. Total weight of water fed to boiler J lbs. 53. Equivalent water fed to boiler from and at 212° .... " 54. Water actually evaporated, corrected for quality of steam 55. Factor of evaporation § 56. Equivalent water evaporated into dry steam from and at 212°. (Item 54 X Item 55) " * This is the total moisture in the coal as found by drying it artificially. t See formula for calorific value under Article XVI. of Code. % Corrected for inequality of water level and of steam pressure at begin ging and end of test. § Factor of evaporation — ~ ' in which H and h are respectively the total heat in steam of the average observed pressure, and in water of the aver- age observed temperature of the feed. 1392 STEAM. Water per Hour 57. Water evaporated per hour, corrected for quality of steam lbs. 58. Equivalent evaporation per hour from and at 212° .... " 59. Equivalent evaporation per hour from and at 212° per square foot of water-heating surface " Horse-Power. 60. Horse-power developed. (34£ lbs. of water evaporated per hour into dry steam from and at 212° equals one horse- power) * H.P. 61. Builders' rated horse-power «« 62. Percentage of builders' rated horse-power developed . . . per cent. Economic Results. 63. Water apparently evaporated per lb. of coal under actual conditions. (Item 53 -7- Item 22) lbs. 64. Equivalent evaporation from and at 212° per lb. of coal (including moisture). (Item 56 -j- Item 22) " 65. Equivalent evaporation from and at 212° per lb. of dry coal. (Item 56 -f- Item 24) " 66. Equivalent evaporation from and at 212° per lb. of combus- tible. (Item 56 ~ Item 26) « (If the equivalent evaporation, Items 64, 65, and 66, is not corrected for the quality of steam, the fact should be stated.) Efficiency. 67. Efficiency of the boiler ; heat absorbed by the boiler per lb. of combustible divided by the heat value of one lb. of combustible t per cent. 68. Efficiency of boiler, including the grate ; heat absorbed by the boiler, per lb. of dry coal fired, divided by the heat value of one lb. of dry coal t Cost of Evaporation. 69. Cost of coal per ton of 2,240 lbs. delivered in boiler room . $ 70. Cost of fuel for evaporating 1,000 lbs. of water under ob- served conditions $ 71. Cost of fuel used for evaporating 1,000 lbs. of water from and at 212° $ Smoke Observations. 72. Percentage of smoke as observed 73. Weight of soot per hour obtained from smoke meter . . . 74. Volume of soot obtained from smoke meter per hour . . Tattle !¥©. 2. Data and Results of Evaporative Test. Arranged in accordance with the Short Form advised by the Boiler Test Committee of the American Society of Mechanical Engineers. Made by on boiler, at to determine * Held to be the equivalent of 30 lbs. of water per hour evaporated from 100° Fahr. into dry steam at 70 lbs. gauge pressure. t In all cases where the word " combustible ** is used, it means the coal with- out moisture and ash, but including all other constituents. It is the same as what is called in Europe " coal dry and free from ash.** t The heat value of the coal is to be determined either by an oxygen calorim- eter or by calculation from ultimate analysis. When both methods are used the mean value is to be taken. RULES FOB CONDUCTING BOILER TESTS. 1393 Grate surface sq.ft. Water-heating surface Superheating surface ** Kind of fuel Kind of furnace Total Quantities. 1. Date of trial 2. Duration of trial . . . , hours. 3. Weight of coal as fired lbs. 4. Percentage of moisture in coal per cent. 5. Total weight of dry coal consumed lbs. 6. Total ash and refuse " 7. Percentage of ash and refuse in dry coal per cent. 8. Total weight of water fed to the boiler lbs. 9. Water actually evaporated, corrected for moisture or super- heat in steam " Hourly Quantities. 10. Dry coal consumed per hour lbs. 11. Dry coal per hour per square foot of grate surf ace ... " 12. Water fed per hour " 13. Equivalent water evaporated per hour from and at 212° corrected for quality of steam " 14. Equivalent water evaporated per square foot of water- heating hour " Average Pressures, Temperatures, etc. 15. Average boiler pressure . . lbs. per sq. in 16. Average temperature of feed-water deg. 17. Average temperature of escaping gases " 18. Average force of draft between damper and boiler . . . ins. of water 19. Percentage of moisture in steam, or number of degrees of superheating Horse-Power. 20. Horse-power developed (Item 13 -f- 34£) H.P. 21. Builders' rated horse-power " 22. Percentage of builders' rated horse-power per cent. Economic Results. 23. Water apparently evaporated per pound of coal under actual conditions. (Item 8 -7- Item 3) lbs. 24. Equivalent water actually evaporated from and at 212° per pound of coal as fired. (Item 9 ~ Item 3) " 25. Equivalent evaporation from and at 212° per pound of dry coal. (Item 9 -f- Item 5) M 26. Equivalent evaporation from and at 212° per pound of combustible. [Item 9 -f- (Item 5 — Item 6)] (If Items 23, 24, and 25 are not corrected for quality of steam, the fact should be stated.) Efficiency. 27. Heating value of the coal per pound B.T.U. 28. Efficiency of boiler (based on combustible) ** 29. Efficiency of boiler, including grate (based on coal) ... :i Cost of Evaporation. 30. Cost of coal per ton of 2,240 pounds delivered in boiler-room $ 31. Cost of coal required for evaporation of 1,000 pounds of water from and at 212° $ 1394 STEAM. IVETIIItJII^AiriOX ©1? THE MOISTURE IM §T£AM. The determination of the quality of steam supplied by a boiler is one of ;he most important items in a boiler test. The three conditions to be de- termined are : a. If the steam is saturated, i.e., contains the quantity of heat due to the pressure. b. If the steam is wet, i.e., contains less than the amount of heat due to the pressure. c. If the steam is superheated, i.e., contains more than the amount of heat due to the pressure. There are several methods of determining the quality of steam ; one being to condense all the steam evaporated by a boiler in a surface condenser, and weigh the condensing water, taking the temperature at its entrance to and exit from the condenser. Another is by use of a barrel calorimeter, in which a sample of the steam is condensed directly in a barrel partly filled with cold water, the added weight and temperature taken, and by use of a formula the quality of steam can be determined. Both the above-named methods are now practically obsolete, as their place has been taken by the throttling calorimeter, used for steam in which the moisture does not exceed 3 per cent, and the separating calorimeter, for steam containing a greater amount of moisture. Throttling- Calorimeter. In its simplest form this instrument can be made up from pipe fittings, the only special parts necessary being the throttling nozzle, which is readily made by boring out a piece of brass rod that is the same diameter as a half- inch steam pipe, leaving a small hole in one end, say T x s inch diameter. The inside end of the small hole should be tapered with the end of a drill so as not to cause eddies ; and the thermometer well, which is a small piece of brass pipe, plugged at one end, and fitted into a half-inch brushing to fit into place. The following cut shows the instrument as made up from fittings. The whole must be carefully covered with some non-conductor, as hair felt. THERMOMETER WELL INSULATING " MATERIAL Pig. 7. For more accurate work the instruments designed by George H. Barrus, M.E., and Prof. R. C. Carpenter, are to be preferred. Professor Carpenter's instrument is shown in the following cut, and differs from the primitive instrument previously described only by the addition of the manometer, DETERMINATION OF MOISTURE. 1395 A'hich determines the pressure of the steam above the atmosphere in the oody of the calorimeter. With a free exit to the air the pressure in the calorimeter may be taken as that of the atmosphere. Carpenter's Throttling- Calorimeter. (I size. Schaeffer & Budenberg.) Fig. 8. The perforated pipe for obtaining the sample of steam to be tested should preferably be inserted in a vertical pipe, and should reach nearly across its diameter. Directions tor Use. — Connect' as shown in the preceding cuts, fill the thermometer cup with cylinder oil and insert the thermometer. Turn on the Globe valve for ten minutes or more in order to bring the tempera- ture of the instrument to full heat, after which note the reading of the ther- mometer in the calorimeter, and of the attached manometer or of a barometer. The steam gauge should be carefully calebrated to see that it is correct. A barometer reading taken at the time the calorimeter is in use, gives greater accuracy in working up the results than taking the average atmospheric pressure as 14.65 pounds. Pressure in pounds may be deter- mined from the mercury column of the barometer and manometer by divid- ing the inches rise by 2.03, or taking one pound for each two inches of mercury. Following is the formula for determining the quality of steam by use of the throttling calorimeter. Hz=z total heat in a pound of steam at the pressure in the pipe. h bs total heat in a pound of steam at the pressure in the calorimeter. L = latent heat in a pound of steam at the pressure in the pipe. t = temperature in the calorimeter. b =. temperature of boiling point at calorimeter pressure (taken as 212° with the " fittings " instrument). 0.48 = specific heat of superheated steam. x sa quality of the steam. y sb percentage of moisture in the steam. H— h— .48(t — b) x = 100 — y. y = x ioo. 1396 STEAM. If h be taken as 212°, as it can be with but slight error, then H-n*A- -AS jt -212) xm L Following are tables calculated from the above formula. ^moisture in Steam. Determinations by Throttling Calorimeter. Gauge-pressures . s 1 5 10 20 30 40 50 60 70 75 80 85 90 Per Cent of Moisture in Steam. 0° 10° 20° 30° 40° 0.51 0.01 0.90 0.39 1.54 1.02 .51 .00 2.06 1.54 1.02 .50 2.50 1.97 1.45 .92 .39 2.90 2.36 1.83 1.30 .77 .24 3.24 2.71 2.17 1.64 1.10 .57 .03 3.56 3.02 2.48 1.94 1.40 .87 .33 3.71 3.17 2.63 2.09 1.55 1.01 .47 3.86 3.32 2.77 2.23 1.69 1.15 .60 .06 3.99 3.45 2.90 2.35 1.80 1.26 .72 .17 4.13 3.58 3.03 2.49 1.94 50° 1.40 60° .85 70° .31 Gauge-pressure. 1 •** 100 110 120 130 140 150 160 170 180 190 200 250 Per Cent of Moisture in Steam 0° 4.39 4.63 4.85 5.08 5.29 5.49 5.68 5.87 6.05 6.22 6.39 7.16 10° 3.84 4.08 4.29 4.52 4.73 4.93 5.12 5.30 5.48 5.65 5.82 6.58 20° 3.29 3.52 3.74 3.96 4.17 4.37 4.56 4.74 4.91 5.08 5.25 6.00 30° 2.74 2.97 3.18 3.41 3.61 3.80 3.99 4.17 4.34 4.51 4.67 5.41 40° 2.19 2.42 2.63 2.85 3.05 3.24 3.43 3.61 3.78 3.94 4.10 4.83 50° 1.64 1.87 2.08 2.29 2.49 2.68 2.87 3.04 3.21 3.37 3.53 4.25 60 J 1.09 1.32 1.52 1.74 1.93 2.12 2.30 2.48 2.64 2.80 2.96 3.67 70° .55 .77 .97 1.18 1.38 1.56 1.74 1.91 2.07 2.23 2.38 3.09 80° .00 .22 .42 .63 .82 1.00 1.18 1.34 1.50 1.66 1.81 2.51 90° .07 .26 .44 .61 .78 .94 1.09 1.24 1.93 100° .05 .21 .37 .52 .67 .10 1.34 110° .76 The easiest method of making the determinations from the observations is by use of the following diagram, prepared by Professor Carpenter. Find in the vertical column at the left the pressure observed in the main pipe -f- atmospheric pressure (the absolute pressure), then move hori- zontally to the right until over the line giving the degree of superheat (t — &), and the quality of steam will be found in a curve corresponding to one of those shown, and which may be interpolated where results do not lome on one of the lines laid down. DETERMINATION OF MOISTURE. 1397 1W i / 1 / / / 1 / / / 170 \ / 1 1 / i / f 1 / 1 i / / i I i / ( / / / 160 / L / / i 1 7 / / , / / i ! 1 / / / / i 1 150 / / / A. / 1 1 / / / / 1 / / *l / 1 / / 1 / / 140 / i 1 I / / 1 l 1 i 1 / i / / / 1 i I 130 3 / / : I J / / / / i t / / / ± 1 1 / / / / 1 1 §j 120 o CO / / / 1 1 fc\7 1 1 / / / i 1 1 1 / 1 / M j 7\ 1 1 f < 1 110 III / / / / • / 1 £1 / / / t / / / fi / 1 K i 10 ° / / / / / / f 1 1 A* ' / i 1 — / / / / / / */ i i ! ( j 0. / / J 1 / • / i i ' / / 1 ,- / kJ / / / / u! 5 80 CO CO / / / / li / ' ■J/ / / / 1 / / / / r f / & / / / / K z 70 < u / / / ' f/ f f/ / / L / / / / / / / / 4 / o N C ;ur VE^ 5 20 s ' /* * y FOR „' .' „," '' Th iROTTLING C> \LO RIIV ET ER 10 --* '* . ^ '" 'un : of ATM 03P HERI C PR ESSl JRE ,- --" 10 20 30 40 50 60 70 80 90 DEGREES OF SUPERHEAT IN THE CALORIMETER D.AGRAM GIVING RESULTS FROM THROTTLING CALORIMETER WITHOUT COMPUTATION FIG. 9 1398 STEAM. By putting a valve in the discharge pipe of the calorimeter, being careful that when open it offers no obstruction to a free passage of the steam, de- terminations may be made from temperatures without reference to a steam table, and by using the following diagram by Professor Carpenter no calcu- lation is necessary. a. Determine the boiling-point of the instrument by opening supply and discharge valves, and showering the instrument with cold water to produce moisture in the calorimeter, in which case the boiline-Doint will be 212° or thereabouts. v^^* b. Determine temperature due to the boiler pressure by closing the dis- charge-valve, leaving the supply-valve open, and obtain the full boiler pressure in the calorimeter. c. Open the discharge-valve and let the thermometer settle to the tempera- ture due to the superheat. Deduct the temperature of the boiling-point from this last temperature to obtain the degrees superheat. Suppose the boiling-point of the calorimeter to be 213°, the following dia- gram will give the result directly from the temperatures. To use the diagram when the boiling-point differs from 212°, add to the temperature of superheat the difference between the true boiling-point and 212°, if less than 212° ; and subtract the difference if the true boiling-point be greater than 212 ; use the result as before. /Separating Calorimeter. This instrument separates the moisture from the sample of steam, and the percentage is then found by the ordinary formula. amount of moisture x 100 . . r= per cent moisture. total steam discharged as sample " One of the most convenient forms of this type of calorimeter is the one designed by Professor Carpenter, and shown in Fig. 11. The sample of steam is let into the instrument through the angle valve 6, the moisture gathers in the inner chamber, its weight in pounds and hundredths being shown on the scale 12, and the dry steam flows out through the small calibrated orifice 8. By Napier's law the flow of steam through an orifice is proportional to the absolute pressure, until the back pressure equals .58 that of the supply. The gauge 9 at the right shows in the outer scale the flow of steam through the orifice 8 in a period of 10 minutes' time. After attaching the instrument to the pipe from which sample is taken through a perforated pipe as with the throttling or other instrument, it must be thoroughly wrapped with hair, felt, or other insulator. Steam is then turned on through the angle valve, and time enough allowed to thor- oughly heat the instrument. In taking an observation, first observe and record height of water on scale 12, then let the steam flow for 10 minutes, observing the average posi- tion of the pointer on the flow-gauge ; at the end of 10 minutes observe the height of water in gauge 12, and the difference between this and the first observation will be the amount of moisture in the sample ; the percent- age of moisture will then be found as follows : difference in scale 12 x 100 difference on scale 12 4- average for 10 minutes on the flow-gauge — % moisture. For tests and data on " Calorimeters," see papers in Trans. A.S.M.E., by Messrs G. H. Barrus, A. A. Goubert, and Professors Carpenter, Denton, Jacobus, and Peabody c DETERMINATION OF MOISTURE. 1399 220 230 m TEMPERATURE IN CALORIMETER 250 260 270 280 290 300 310 320 380 310 / / / / / / / / / / / JL t / / 400 .!©> / / |/ / / / / // / T / 47 *. 8 T --' 7 / / / / / / 390 V ffl / ' / / / / / / / / / 7 / / / o 7 I 7 / / 7 / 7 / / / y 340 / / 7 / / / / / / 1 / / / / / / z / / 330 / / f / / / T / / / / / / / 7 / / 320 / / / / / / / */ / / / / / / ' & 810 / / / 7 / / .*/ ) / / r / r / / / f 300 / / / / / / / / 7 ~Z / 290 / v ) / / / / / / / / / 280 / / L / / / / / / / in / / / / / / / I 7 2tiU / / / / / / / / / 250 / / / / / / 249 / CURVES OF QUALITY FOR USE WITH / / 230 y 7 CARPENT ER'S THROTTLING CALC )RIM / 220 / ?ic 230 240 250 260 270 280 290 300 310 TEMPERATURE IN CALORIMETER 330 340 DIAGRAM FOR COMPUTING RESULTS WITH THROTTLING CALORIMETER. Fig. 10. 1400 STEAM. Quality of Steam Shown by Color of Issuing* Jet. Fig. 11. Carpenter's New Evaporat- ing Calorimeter. (Schaeffer & Bu- denberg.) Prof. J. E. Denton (Trans. A. S. M. E., vol. x., p. 349) has demon- strated that jets of steam escaping from an orifice in a boiler or steam reservoir show unmistakable change of appearance to the eye when the steam varies less than one per cent from the condition of saturation either in the direction of wetness or superheating. Conse- quently if a jet of steam flow from a boiler into the atmosphere under circumstances such that very lit- tle loss of heat occurs through radiation, etc., and the jet be transparent close to the orifice, or be even a grayish white color, the steam may be assumed to be so nearly dry that no portable condensing "calorimeter will be capable of measuring the amount of water therein. If the jet be strongly white, the amount of water may be roughly judged up to about 2 per cent, but beyond this a calorimeter only can deter- mine the exact amount of moist- ure. With a little experience any one may determine by this meth- od the conditions of steam within the above limits. A common brass pet cock may be used as an orifice, but it should, if possible, be set into the steam drum of the boiler and never be placed farther away from the latter than four feet, and then only when the in- termediate reservoir or pipe is well covered, for a very short travel of dry steam through a naked pipe will cause it to become perceptibly moist. FACTORS OF EVAPO- HATI.O.Y, In order to facilitate the calcu- lation of reducing the actual rate of evaporation of water from a certain temperature into steam of a cer- tain pressure, into the rate from water at 212° F. into steam of 212° a table of factors of evaporation is made up from the formula H — h 965.7 where His the total heat of steam at the observed pressure, and h the total heat of feed-water of the observed temperature. FACTORS OF EVAPORATION. 1401 Table of Factors of Evaporation. ;(W. W. Christie, M.E.) Gauge Pressure. 10 20 30 40 45 50 52 54 Temp, of lbs. lbs. lbs. lbs. lbs. lbs. lbs. lbs. lbs. Feed. 212° F. 1.0003 1.0088 1.0149 1.0197 1.0237 1.0254 1.0271 1.0277 1.0283 209 1.0035 1.0120 1.018U 1.0228 1.0268 1.0286 1.0302 1.0309 1.0315 206 1.0066 1.0151 1.0212 1.0260 1.0299 1.0317 1.0334 1.0340 1.0346 203 1.0098 1.0183 1.0243 1.0291 1.0331 1.0349 1.0365 1.0372 1.0378 200 1.0129 1.0214 1.0275 1.0323 1.0362 1.0380 1.0397 1.0403 1.0409 197 1.0160 1.0246 1.0306 1.0354 1.0394 1.0412 1.0428 1.0434 1.0441 194 1.0192 1.0277 1.0338 1.0385 1.0425 1.0443 1.0460 1.0466 1.0472 191 1.0223 1.0308 1.0369 1.0417 1.0457 1.0474 1.0491 1.0497 1.0503 188 1.0255 1.0340 1.0400 1.0448 1.0488 1.0506 1.0522 1.0528 1.0535 185 1.0286 1.0371 1.0432 1.0480 1.0519 1.0537 1.0554 1.0560 1.0566 182 1.0317 1.0403 1.0463 1.0511 1.0551 1.0568 1.0585 1.0591 1.0598 179 1.0349 1.0434 1.0495 1.0542 1.0582 1.0600 1.0616 1.0623 1.0629 176 1.0380 1.0465 1.0526 1.0574 1.0613 1.0631 1.0648 1.0654 1.0660 173 1.0411 1.0497 1.0557 1.0605 1.0645 1.0663 1.0679 1.0685 1.0692 170 1.0443 1.0528 1.0589 1.0636 1.0676 1.0694 1.0710 1.0717 1.0723 167 1.0474 1.0559 1.0620 1.0668 1.0707 1.0725 1.0742 1.0748 1.0754 164 1.0505 1.0591 1.0651 1.0699 1.0739 1.0756 1.0773 1.0780 1.0786 161 1.0537 1.0622 1.0682 1.0730 1.0770 1.0788 1.0804 1.0811 1.0817 158 1.0568 1.0653 1.0714 1.0762 1.0801 1.0819 1.0836 1.0842 1.0848 155 1.0599 1.0684 1.0745 1.0793 1.0833 1.0850 1.0867 1.0873 1.0880 152 1.0631 1.0716 1.0776 1.0824 1.0864 1.0882 1.0898 1.0905 1.0911 149 1.0662 1.0747 1.0808 1.0855 1.0895 1.0913 1.0930 1.0936 1.0942 146 1.0693 1.0778 1.0839 1.0887 1.0926 1.0944 1.0961 1.0967 1.0973 143 1.0724 1.0810 1.0870 1.0918 1.0958 1.0975 1.0992 1.0998 1.1005 140 1.0756 1.0841 1.0901 1.0949 1.0989 1.1007 1.1023 1.1030 1.1036 137 1.0787 1.0872 1.0933 1.0980 1.1020 1.1038 1.1055 1.1061 1.1067 134 1.0818 1.0903 1.0964 1.1012 1.1051 1.1069 1.1086 1.1092 1.1098 131 1.0849 1.0934 1.0995 1.1043 1.1083 1.1100 1.1117 1.1123 1.1130 128 1.0881 1.0966 1.1026 1.1074 1.1114 1.1132 1.1148 1.1155 1.1161 125 1.0912 1.0997 1.1057 1.1105 1.1145 1.1163 1.1179 1.1186 1.1192 122 1.0943 1.1028 1.1089 1.1136 1.1176 1.1194 1.1211 1.1217 1.1223 119 1.0974 1.1059 1.1120 1.1168 1.1207 1.1225 1.1242 1.1248 1.1254 116 1.1005 1.1090 1.1151 1.1199 1.1239 1.1256 14273 1.1279 1.1286 113 1.1036 1.1122 1.1182 1.1230 1.1270 1.1288 1.1304 1.1310 1.1317 110 1.1068 1.1153 1.1213 1.1261 1.1301 1.1319 1.1335 1.1342 1.1348 107 1.1099 1.1184 1.1245 1.1292 1.1332 1.1350 1.1366 1.1373 1.1379 104 1.1130 1.1215 1.1276 1.1323 1.1363 1.1381 1.1398 1.1404 1.1410 101 1.1161 1.1246 1.1307 1.1355 1.1394 1.1412 1.1429 1.1435 1.1441 98 1.1192 1.1277 1.1338 1.1386 1.1426 1.1443 1.1460 1.1466 1.1473 95 1.1223 1.1309 1.1369 1.1417 1.1457 1.1475 1.1491 1.1497 1.1504 92 1.1255 1.1340 1.1400 1.1448 1.1488 1.1506 1.1522 1.1529 1.1535 89 1.1286 1.1371 1.1431 1.1479 1.1519 1.1537 1.1553 1.1560 1.1566 86 1.1317 1.1402 1.1463 1.1510 1.1550 1.1568 1.1584 1.1591 1.1597 83 1.1348 1.1433 1.1494 1.1541 1.1581 1.1599 1.1616 1.1622 1.1628 80 1.1379 1.1464 1.1525 1.1573 1.1612 1.1630 1.1647 1.1653 1.1659 77 1.1410 1.1495 1.1556 1.1604 1.1644 1.1661 1.1678 1.1684 1.1690 74 1.1441 1.1526 1.1587 1.1635 1.1675 1.1692 1.1709 1.1715 1.1722 71 1.1472 1.1558 1.1618 1.1666 1.1706 1.1723 1.1740 1.1746 1.1753 68 1.1504 1.1589 1.1649 1.1697 1.1737 1.1755 1.1771 1.1778 1.1784 65 1.1535 1.1620 1.1680 1.1728 1.1768 1.1786 1.1802 1.1809 1.1815 62 1.1566 1.1651 1.1711 1.1759 1.1799 1.1817 1.1833 1.1840 1.1846 59 1.1597 1.1682 1.1743 1.1790 1.1830 1.1848 1.1864 1.1871 1.1877 56 1.1628 1.1713 1.1774 1.1821 1.1861 1.1879 1.1896 1.1902 1.1908 53 1.1659 1.1744 1.1805 1.1852 1.1892 1.1910 1.1927 1.1933 1.1939 50 1.1690 1.1775 1.1836 1.1884 1.1923 1.1941 1.1958 1.1964 1.1970 47 1.1721 1.1806 1.1867 1.1915 1.1954 1.1972 1.1989 1.1995 1.2001 44 1.1752 1.1837 1.1898 1.1946 1.1986 1.2003 1.2020 1.2026 1.2032 41 1.1783 1.1868 1.1929 1.1977 1.2017 1.2034 1.2051 1.2057 1.2064 38 1.1814 1.1900 1.1960 1.2008 1.2048 1.2065 1.2082 1.2088 1.2095 35 1.1845 1.1931 1.1991 1.2039 1.2079 1.2096 1.2113 1.2119 1.2126 32 1.1876 1.1962 1.2022 1.2070 1.2110 1.2128 1.2144 1.2151 1.2157 1402 STEAM. Table of factors of Evaporation. Gauge Pressure. 56 58 60 65 70 75 80 1 85 90 95 Temp, of Feed. lbs. lbs. lbs. lbs. lbs, lbs. lbs. 1 lbs. lbs. lbs. 212° tf. 1.0290 1.02951 1.0301 1.0315 1.0329 1.0341 1.0353 1.0365 T.0376 1.0387 209 1.0321 1.0327 1.0333 1.0346 1.0360 1.0372 1.0385 1.0397 1.0408 1.0419 206 1.0352 1.0358 1 1.0364 1.0378 1.0391 1.0403 1.0416 1.0428 1.0439 1.0450 203 1.0384 1.0390 1.0396 1.0464 1.0423 1.0435 1.0448 1.0460 1.0471 1.0482 200 1.0415 1.0421 1.0427 1.0441 1.0454 1.0466 1.0479! 1.0491 1.0502 1.0513 197 1.0447 1.0453 1.0458 1.0477 1.0486 1.0498 1.0511 1 1.0522 1.0533 1.0544 194 1.0478 1.0484, 1.0490 1.0504 1.0517 1.0529 1.0542 1.0553 1.0565 1.0576 191 1.0510 1.0515 1 1.0521 1.0535 1.0549 1.0561 1.0573: 1.0585 1.0596 1.0607 188 1.0541 1.05471 1.0553 1.0566 1.0580 1.0592 1.0605 1.0616 1.0628 1.0639 185 1.0572 1.0578' 1.0584 1.0598 1.0611 1.0623 1.06361 1.0648 1.0659 1.0670 182 1.0604 1.0610 1.0615 1.0629 1.0643 1.0655 1.0668 1.0679 1.0690 1.0701 179 1.0635 1.0641 1.0647 1.0660 1.0674 1.0686 1.0699 1.0710 1.0722 1.0733 176 1.0666 1.06721 1.0678 1.0692 1.0705 1.0717 1.0730 1.0742 1.0753 1.0764 173 1.0698 1.07041 1.0709 1.0723 1.0737 1.0749 1.0762 1.0773 1.0784 1.0795 170 1.0729 1.0735J 1.0741 1.0754 1.0768 1.0780 1.0793 1.0804 1.0816 1.0827 167 1.0760 1.07661 1.0772 1.0786 1.0799 1.0811 1.0824 1.0836 1.0847 1.0858 164 1.0792 1.0798! 1.0803 1.0817 1.0831 1.0843 1.0856 1.0867 1.0878 1.0889 161 1.0823 1.0829| 1.0835 1.0848 1.0862 1.0874 1.0887 1.0898 1.0910 1.0921 158 1.0854 1.0860' 1.0866 1.0880 1.0893 1.0905 1.0918 1.0929 1.0941 1.0952 155 1.0886 1.0892 1 1.0897 1.0911 1.0925 1.0937 1.0949 1.0961 1.0972 1.0983 152 1.0917 1.0923 1.0929 1.0942 1.0956 1.0968 1.0981 1.0992 1.1004 1.1015 149 1.0948 1.0954! 1.0960 1.0974 1.0987 1.0999 1.1012 1.1023 1.1035 1.1046 146 1.0979 1.0985 1.0991 1.1005 1.1018 1.1030 1.1043 1.1055 1.1066 1.1077 143 1.1011 1.1017 1.1022 1.1036 1.1050 1.1062 1.1074 1.1086 1.1097 1.1108 140 1.1042 1.1048 1.1054 1.1067 1.1081 1.1093 1.1106 1.1117 1.1129 1.1140 137 1.1073 1.1079 1.1085 1.1099 1.1112 1.1124 1.1137 1.1148 1.1160 1.1171 134 1.1104 1.1110 1.1116 1.1130 1.1143 1.1155 1.1168 1.1180 1.1191 1.1202 131 1.1136 1.1142 1.1147 1.1161 1.1175 1.1187 1.1199 1.1210 1.1222 1.1233 128 1.1167 1.1173 1.1179 1.1192 1.1206 1.1218 1.1231 1.1242 1.1253 1.1264 125 1.1198 1.1204 1.1210 1.1223 1.1237 1.1248 1.1262 1.1273 1.1285 1.1296 122 1.1229 1.1235 1.1241 1.1255 1.1268 1.1281 1.1293 1.1294 1.1316 1.1327 119 1.1260 1.1266 1.1272 1.1286 1.1299 1.1311 1.1324 1.1336 1.1347 1.1358 116 1.1292 1.1298 1.1303 1.1317 1.1331 1.1343 1.1355 1.1366 1.1378 1.1389 113 1.1323 1.1329 1.1334 1.1348 1.1362 1.1374 1.1387 1.1398 1.1409 1.1420 110 1.1354 1.1360 1.1366 1.1374 1.1393 1.1405 1.1418 1.1429 1.1441 1.1452 107 1.1385 1.1391 11397 1.1411 1.1424 1.1436 1.1449 1.1460 1.1472 1.1483 104 1.1416 1.1422 1.1428 1.1442 1.1455 1.1467 1.1480 1.1491 1.1503 1.1514 101 1.1447 1.1453 1.1459 1.1473 1.1486 1.1498 1.1511 1.1523 1.1534 1.1545 98 1.1479 1.1485 1.1490 1.1504 1.1518 1.1530 1.1542 1.1554 1.1565 1.1576 95 1.1510 1.1516 1.1521 1.1535 1.1549 1.1561 1.1574 1.1583 1.1596 1.1607 92 1.1541 1.1547 1.1553 1.1566 1.1580 1.1592 1.1605 1.1616 1.1628 1.1639 89 1.1572 1.1578 1.1584 1.1598 1.1611 1.1623 1.1636 1.1647 1.1659 1.1670 86 1.1603 1.1609 1.1615 1.1629 1.1642 1.1654 1.1667 1.1678 1.1690 1.1701 83 1.1634 1.1640 1.1646 1.1660 1.1673 1.1685 1.1698 1.1709 1.1721 1.1732 80 1.1665 1.1671 1.1677 1.1691 1.1704 1.1716 1.1729 1.1741 1.1752 1.1763 77 1.1696 1.1702 1.1708 1.1722 1.1735 1.1747 1.1760 1.1772 1.1783 1.1794 74 1.1728 1.1734 1.1739 1.1753 1.1767 1.1779 1.1791 1.1803 1.1814 1.1825 71 1.1759 1.1765 1.1770 1.1784 1.1798 1.1810 1.1823 1.1834 1.1845 1.1856 68 1.1790 1.1796 1.1802 1.1815 1.1829 1.1841 1.1854 1.1865 1.1877 1.1888 65 1.1821 1.1827 1.1833 1.1846 1.1860 1.1872 1.1885 1.1896 1.1908 1.1919 62 1.1852 1.1858 1.1864 1.1877 1.1891 1.1903 1.1916 1.1927 1.1939 1.1.950 59 1.1883 1.1889 1.1895 1.1909 1.1922 1.1934 1.1947 1.1958 1.1970 1.2981 56 1.1914 1.1920 1.1926 1.1940 1.1953 1.1965 1.1978 1.1989 1.2001 1.2012 53 1.1945 1.1951 1.1957 1.1971 1.1984 1.1996 1.2009 1.2020 1.2032 1.2043 50 1.1976 1.1982 1.1988 1.2002 1.2015 1.2027 1.2040 1.2052 1.2063 1.2074 47 1.2007 1.2013 1.2019 1.2033 1.2046 1.2058 1.207l| 1.2083 1.2094 1.2105 44 1.2039 1.2044 1.2050 1.2064 1.2078 1.2090 1.21021 1.2114 1.2125 1.2136 41 1.2070 1.2076 1.2081 1.2095 1.2109 1.2121 1.2133 1.2145 1.2156 1.2167 38 1.2101 1.2107 1.2112 1.2126 1.2140 1.2162 1.2164 1.2176 1.2187 1.2198 35 1.2132 1.2138 1.2143 1.21571 1.2171 1.21831 1.21961 1.2207 1.2218 1.2229 32 1.2163 1.2169 1.2175 1.2188! 1.2202 1.2214| 1.22271 1.2239 1.2249 1.2260 FACTORS OF EVAPORATION. 1403 Table of facto rs of Evaporati on. Gauge Pressure. 100 105 115 125 135 145 155 165 185 Temp, of Feed. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. 212° F, 1.0397 1.0407 1.0427 1.0445 1.0462 1.0478 1.0493 1.0509 1.0536 209 1.0429 1.0438 1.0458 1.0476 1.0493 1.0509 1.0524 1.0540 1.0567 206 1.0460 1.0470 1.0489 1.0510 1.0527 1.0543 1.0558 1.0574 1.0601 203 1.0492 1.0502 1.0521 1.0540 10557 1.0573 1.0588 1.0604 1.0631 200 1.0523 1.0533 1.0552 1.0571 1.0588 1.0604 1.0619 1.0635 1.0662 197 1.0555 1.0565 1.0584 1.0602 1.0619 1.0635 1.0650 1.0666 1.0693 194 1.0586 1.0596 1.0615 1.0635 1.0652 1.0668 1.0683 1.0699 1.0726 191 1.0617 1.0627 1.0647 1.0665 1.0682 1.0698 1.0713 1.0729 1.0756 188 1.0649 1.0659 1.0678 1.0696 1.0713 1.0729 1.0744 1.0760 1.0787 185 1.0680 1.0690 1.0709 ±\0728 1.0745 1.0761 1.0776 1.0792 1.0819 182 1.0712 1.0722 1.0741 1.0759 1.0776 1.0792 1-0807 1.0823 1.0850 179 1.0743 1.0753 1.0772 1-0790 1.0807 1.0823 1.0838 1.0854 1.0881 176 1.0774 1.0784 1.0803 1.0822 1.0839 1.0855 1.0870 1.0886 1.0913 173 1.0806 1.0816 1.0835 1.0853 1.0870 1.0886 1.0901 1.0917 1.0944 170 1.0837 1.0847 1.0866 1.0884 1.0901 1.0917 1.0932 1.0948 1.0975 167 1.0868 1.0878 1.0897 1.0916 1.0933 1.0949 1.0964 1.0980 1.1007 164 1.0900 1.0910 1.0929 1.0946 1.0963 1.0979 1.0994 1.1010 1.1037 161 1.0931 1.0941 1.0960 1.0979 1.0996 1.1012 1.1027 1.1043 1.1070 158 1.0962 1.0972 1.0991 1.1010 1.1027 1.1043 1.1058 1.1074 1.1101 155 1.0993 1.1003 1.1023 1.1041 1.1058 1.1074 1.1089 1.1105 1.1132 152 1.1025 1.1035 1.1054 1.1073 1.1090 1.1107 1-1122 1.1138 1.1165 149 1.1056 1.1066 1.1085 1.1103 1.1120 1.1136 1.1151 1.1167 1.1194 146 1.1087 1.1097 1.1116 1.1135 1.1152 1.1168 1.1183 1.1199 1.1226 143 1.1118 1.1129 1.1148 1.1166 1.1183 1.1199 1.1214 1.1230 1.1257 140 1.1150 1.1160 1.1179 1.1197 1.1214 1.1230 1.1245 1.1261 1.1288 137 1.1181 1.1191 1.1210 1.1228 1.1245 1.1262 1.1277 1.1293 1.1320 134 1.1212 1.1222 1.1241 1.1260 1.1277 1.1293 1.1308 1.1324 1.1351 131 1.1243 1.1253 1.1273 1.1291 1.1308 1.1324 1.1339 1.1355 1.1382 128 1.1275 1.1285 1.1304 1.1322 1,1339 1.1355 1.1370 1.1386 1.1413 125 1.1306 1.1316 1.1335 1.1353 1.1370 1.1386 1.1401 1.1417 1.1444 122 11337 1.1347 1.1366 1.1384 1.1401 1.1417 1.1438 1.1448 1.1475 119 1.1368 1.1378 1.1397 1.1415 1.1432 1.1449 1.1464 1.1480 1.1507 116 1.1399 1.1409 1.1429 1.1447 1.1464 1.1480 1.1495 1.1511 1.1538 113 1.1431 1.1441 1.1460 1.1478 1.1495 1.1511 1.1526 1.1542 1.1569 110 1.1462 1.1472 1.1491 1.1509 1.1516 1.1542 1.1557 1.1573 1.1600 107 1.1493 1.1503 1.1522 1.1540 1.1557 1.1573 1.1588 1.1604 1.1631 104 1.1524 1.1534 1.1553 1.1571 1.1588 1.1605 1.1619 1.1635 1.1662 • 101 1.1555 1.1565 1.1584 1.1602 1.1620 1.1636 1.1652 1.1668 1.1695 98 1.1586 1.1596 1.1616 1.1634 1.1651 1.1667 1.1683 1.1699 1.1726 95 1.1618 1.1628 1.1647 1.1665 1.1682 1.1698 1.1713 1.1729 1.1756 92 1.1649 1.1660 1.1678 1.1696 1.1713 1.1729 1.1744 1.1760 1.1787 89 1.1680 1.1690 1.1709 1.1727 1.1744 1.1760 1.1775 1.1791 1.1818 86 1.1711 1.1721 1.1740 1.1758 1.1775 1.1791 1.1806 1.1822 1.1849 83 1.1742 1.1752 1.1771 1.1789 1.1806 1.1823 1.1837 1.1853 1.1880 80 1.1773 1.1783 1.1802 1.1820 1.1837 1.1854 1.1869 1.1885 1.1912 77 1.1804 1.1814 1.1834 1.1852 1.1869 1.1885 1.1900 1.1916 1.1943 74 1.1835 1.1845 1.1865 1.1883 1.1900 1.1916 1.1932 1.1948 1.1975 71 1.1867 1.1877 1.1896 1.1914 1.1931 1.1947 1.1961 1.1977 1.2004 68 1.1898 1.1908 1.1927 1.1945 1.1962 1.1978 1.1993 1.2009 1.2036 65 1.1929 1.1939 1.1958 1.1976 1.1993 1.2009 1.2024 1.2040 1.2067 62 1.1960 1.1970 1.1989 1.2007 1.2024 1.2040 1.2055 1.2071 1.2098 59 1.1991 1.2001 1.2020 1.2038 1.2055 1.2071 1.2086 1.2102 1.2129 56 1.2022 1.2032 1.2051 1.2069 1.2086 1.2102 1.2117 1.2133 1.2160 53 1.2053 1.2063 1.2082 1.2100 1.2117 1.2134 1.2148 1.2164 1.2191 50 1.2084 1.2094 1.2113 1.2131 1.2148 1.2165 1.2180 1.2196 1.2223 47 1.2115 1.2125 1.2144 1.2163 1.2180 1.2196 1.2211 1.2227 1.2254 44 1.2146 1.2156 1.2176 1.2194 1.2211 1.2227 1.2242 1.2258 1.2285 41 1.2177 1.2187 1.2207 1.2225 1.2242 1.2258 1.2273 1.2289 1.2316 38 1.2208 1.2219 1 2238 1.2256 1.2273 1.2289 1.2304 1.2320 1.2347 35 1.2240 1.2250 12269 12287 1.2304 1.2320 1.2335 1.2351 1.2378 32 1.2271 1.2281 1.2300 1.2318 1.2335 1.2351 1.2366 1 .2382 1.2409 1403a FACTORS OF EVAPORATION. Table of factors of evaporation. W. Wallace Christie. Gauge Pressure. 200 215 230 245 260 275 290 300 Temp, of lbs. lbs. lbs. lbs. I lbs. lbs. lbs. lbs. Feed. 212° F. 1:0555 1.0574 1.0591 1.0605 1.0622 1.0639 1.0653 1.0663 209 1.0586 1.0605 1.0622 1.0639 1.0654 1.0670 1.0684 1.0694 206 1.0616 1.0635 1.0653 1.0669 1.0685 1.0700 1.0715 1.0724 203 1.0648 1.0668 1.0684 1.0701 1.0717 1.0732 1.0746 1.0756 200. 1.0680 1.0699 1.0716 1.0733 1.0749 1.0764 1.0779 1.0788 197 1.0711 1.0730 1.0747 1.0764 1.0781 1.0795 1.0810 1.0819 194 1.0743 1.0762 1.0779 1.0796 1.0813 1.0827 1.0842 1.0851 191 1.0774 1.0793 1.0811 1.0827 1.0844 1.0858 1.0873 1.0882 188 1.0806 1.0825 1.0843 1.0859 1.0875 1.0890 1.0905 1.0914 185 1.0838 1.0857 1.0875 1.0891 1.0907 1.0923 1.0937 1.0946 182 1.0869 1.0888 1.0906 1.0923 1.0938 1.0954 1.0968 1.0977 179 1.0900 1.0917 1.0937 1.0954 1.0969 1.0985 1.0999 1.1009 176 1.0932 1.0950 1.0968 1.0985 1 . 1000 1.1016 1.1030 1 . 1040 173 1.0964 1.0983 1 . 1000 1.1017 1 . 1032 1.1048 1 . 1062 1.1072 170 1.0995 1.1014 1.1031 1.1048 1 . 1063 1.1079 1 . 1093 1.1103 167 1 . 1026 1 . 1045 1 . 1062 1.1079 1 . 1094 1.1110 1.1124 1.1134 164 1.1057 1.1076 1.1093 1.1110 1.1126 1.1141 1 . 1155 1.1165 161 1 . 1088 1.1107 1.1124 1.1141 1.1157 1.1172 1.1187 1.1196 158 1.1120 1.1139 1.1156 1.1173 1.1189 1.1204 1.1219 1 . 1228 155 1.1151 1.1170 1.1188 1.1204 1 . 1220 1.1235 1 . 1250 1.1259 152 1.1182 1 . 1201 1.1219 1 . 1235 1.1251 1.1266 1.1281 1 . 1290 149 1.1213 1 . 1232 1 . 1250 1 . 1266 1.1282 1.1297 1.1312 1.1321 146 1.1245 1.1264 1.1282 1.1298 1.1314 1.1329 1.1344 1.1353 143 1.1276 1.1295 1.1313 1 . 1329 1.1345 1.1361 1.1375 1.1384 140 1.1308 1 . 1326 1.1344 1.1360 1.1376 1 . 1392 1.1406 1.1415 137 1.1339 1.1357 1.1375 1.1392 1.1407 1.1424 1.1437 1.1447 134 1.1371 1.1389 1.1407 1 . 1424 1 . 1438 1.1456 1.1469 1.1479 131 1 . 1402 1.1421 1.1438 1.1455 1.1470 1.1487 1 . 1500 1.1510 128 1.1433 1 . 1452 1 . 1469 1.1486 1.1501 1.1518 1.1532 1.1541 125 1.1464 1 . 1483 1.1500 1.1517 1.1532 1.1549 1.1563 1.1572 122 1 . 1496 1.1515 1 . 1532 1.1549 1 . 1564 1 . 1580 1.1595 1.1604 119 1.1527 1.1546 1.1563 1.1580 1 . 1596 1.1611 1 . 1626 1.1635 116 1.1559 1.1577 1 . 1594 1.1611 1.1627 1 . 1642 1.1657 1.1666 113 1.1589 1.1608 1.1626 1.1642 1.1658 1.1673 1.1688 1.1697 110 1 . 1620 1.1639 1.1657 1 . 1673 1.1689 1.1704 1.1719 1.1728 107 1.1651 1.1670 1.1688 1.1704 1.1720 1.1735 1.1750 1.1760 104 1.1682 1.1701 1.1719 1.1735 1.1751 1.1766 1.1781 1.1790 101 1.1713 1.1732 1.1750 1.1766 1.1782 1.1797 1.1812 1.1821 " 98 1 . 1744 1 . 1763 1.1781 1.1797 1.1813 1.1829 1.1843 1.1853 95 1.1776 1 . 1794 1.1812 1.1829 1 . 1844 1 . 1860 1 . 1874 1.1884 92 1.1807 1.1826 1 . 1843 1.1860 1.1875 1.1891 1.1905 1.1915 89 1.1838 1 . 1857 1.1874 1.1891 1.1906 1.1922 1.1936 1.1946 86 1 . 1869 1 . 1888 1.1905 1.1922 1 . 1937 1.1953 1.1967 1.1977 83 1 . 1900 1.1919 1.1936 1.1953 1.1968 1.1984 1.1999 1.2008 80 1.1931 1 . 1950 1.1967 1.1984 1.2000 1.2015 1.2030 1.2039 77 1.1962 1:1981 1.1998 1.2015 1.2031 1.2046 1.2061 1.2070 74 1 . 1993 1.2012 1.2029 1.2046 1.2062 1.2077 1.2092 1.2101 71 1.2024 1.2043 1.2061 1.2077 1.2092 1.2108 1.2123 1.2132 68 1 . 2055 1.2074 1.2092 1.2108 1.2124 1.2139 1.2154 1.2163 65 1.2087 1.2105 1.2123 1.2139 1.2155 1.2170 1.2185 1.2194 62 1.2118 1.2136 1.2154 1.2172 1.2186 1.2201 1.2216 1.2225 59 1.2149 1.2167 1.2185 1.2202 1.2217 1.2233 1.2247 1.2256 56 1.2180 1.2198 1.2216 1.2232 1.2248 1.2264 1.2278 1.2288 53 1.2211 1.2229 1.2247 1.2264 1.2279 1.2295 1.2309 1.2319 50 1.2242 1.2261 1.2278 1.2295 1.2310 1.2326 1.2340 1.2350 47 1.2273 1.2292 1.2309 1 . 2326 1.2341 1.2357 1.2371 1.2381 44 1.2304 1 . 2323 1 . 2340 1.2357 1.2372 1.2388 1.2402 1.2412 41 1.2335 1.2354 1.2371 1 . 2388 1.2403 1.2419 1.2433 1.2443 38 1.2366 1.2385 1 . 2402 1.2419 1.2434 1.2450 1.2464 1.2474 35 1.2397 1.2416 1.2433 1.2450 1 . 2465 1.2481 1.2496 1.2505 32 1 . 2428 1.2447 1 . 2464 1.2481 1.2497 1.2512 1.2527 1.2536 FACTORS OF EVAPORATION. 1403b Table of factors of Evaporation. — Continued, y Gauge Pressure. Temp, of Feed. Lbs. 10 Lbs. 20 Lbs. 30 Lbs. 40 Lbs. 45 Lbs. 50 Lbs. 52 Lbs. 54 Lbs. 300° F. 295 290 287 284 281 278 275 272 269 266 263 260 257 254 251 248 245 242 239 236 233 230 227 224 221 218 215 0.907 0.912 0.917 0.921 0.924 0.927 0.930 0.933 0.936 0.940 0.943 0.946 0.949 0.952 0.955 0.958 0.961 0.964 0.967 0.970 0.974 0.977 0.980 0.983 0.986 0.989 0.993 0.997 0.915 0.920 0.926 0.930 0.933 0.936 0.939 0.942 0.945 0.948 0.951 0.955 0.958 0.961 0.964 0.967 0.970 0.974 0.977 0.981 0.984 0.987 0.990 0.993 0.996 0.999 1.002 1.005 0.922 0.927 0.932 0.936 0.939 0.942 0.945 0.948 0.951 0.954 0.958 0.961 0.964 0.967 0.970 0.974 0.977 0.980 0.983 0.986 0.989 0.992 0.996 0.999 1.002 1.005 1.008 1.010 0.926 0.932 0.937 0.940 0.944 0.947 0.950 0.953 0.956 0.959 0.963 0.966 0.969 0.972 0.975 0.978 0.982 0.985 0.988 0.991 0.994 0.998 1.001 1.004 1.007 1.010 1.013 1.016 0.930 0.936 0.941 0.945 0.948 .0.951 0.954 0.958 0.961 0.964 0.968 0.971 0.974 0.977 0.980 0.983 0.987 0.990 0.993 0.995 0.998 1.001 1.005 1.008 1.011 1.014 1.017 1.020 0.932 0.937 0.943 0.946 0.949 0.953 0.956 0.959 0.962 0.966 0.969 0.972 0.975 0.978 0.981 0.984 0.987 0.990 0.994 0.997 1.000 1.003 1.007 1.010 1.013 1.016 1.019 1.022 0.934 0.939 0.944 0.948 0.951 0.954 0.957 0.960 0.963 0.967 0.970 0.973 0.976 0.979 0.983 0.986 0.989 0.992 0.995 0.999 1.002 1.005 1.008 1.011 1.014 1.017 1.021 1.024 0.9347 0.9399 0.9453 0.9485 0.9517 0.9548 0.9580 0.9612 0.9642 0.9675 0.9708 0.9738 0.9770 0.9801 0.9833 0.9865 0.9897 0.9929 0.9960 0.9992 1.0024 1.0055 1.0087 1.0118 1.0149 1.0180 1.0212 1.0244 0.9353 0.9406 0.9459 0.9492 0.9524 0.9554 0.9586 0.9618 0.9648 0.9681 0.9714 0.9744 0.9776 0.9807 0.9840 0.9872 0.9904 0.9935 0.9966 1.0000 1.0030 1.0061 1.0093 1.0124 1.0155 1.0186 1.0217 1.0251 56 Lbs. 58 Lbs. 60 Lbs. 65 Lbs. 70 Lbs. 75 Lbs. 80 Lbs. 85 Lbs. 90 Lbs. 95 Lbs. 300° F. 295 290 287 284 281 278 275 272 269 266 263 260 257 254 251 248 245 242 239 236 233 230 227 224 221 218 215 0.9359 0.9412 0.9465 0.9498 0.9530 0.9561 0.9592 0.9624 0.9654 0.9687 0.9720 0.9750 0.9782 0.9814 0.9846 0.9877 0.9910 0.9941 0.9972 1.0004 1.0036 1.0067 1.0099 1.0130 1.0161 1.0193 1.0225 1.0257 0.9365 0.9418 0.9472 0.9504 0.9536 0.9567 0.9598 0.9630 0.9660 0.9693 0.9727 0.9757 0.9789 0.9820 0.9853 0.9884 0.9916 0.9948 0.9979 1.0011 1.0042 1.0073 1.0106 1.0137 1.0168 1.0199 1.0231 1.0263 0.9370 0.9423 0.9477 0.9509 0.9541 0.9572 0.9603 0.9635 0.9665 0.9699 0.9732 0.9762 0.9794 0.9825 0.9857 0.9889 0.9921 0.9953 0.9984 1.0016 1.0048 1.0089 1.0111 1.0142 1.0173 1.0204 1.0236 1.0268 O.c 0.1 0.1 0/ 0/ 0.1 0.1 0.! 0.1 0. 0. 0.1 0. 0. 0. 0. 0. 0. 0. 1. 1. 1. 1. 1 1. 1. 1. )385 )438 )492 )524 )556 )587 )618 )650 )680 )713 )746 )776 )808 )840 ?872 3904 3936 )968 3999 3030 3062 3094 3125 3156 0187 0218 0251 0283 0.93< 0.94, 0.95( 0.951 0.95( 0.96( 0.96 0.96 0.96 0.97 0.97 0.97 0.98 0.98 0.98 0.99 0.99 0.99 1.00 1.00 1.00 1.01 1.01 1.01 1.02 1.02 1.02 1.02 18 51 )5 M 39 )0 Jl 33 U 26 30 30 22 53 So 17 19 SO 11 43 76 V 39 70 01 32 64 •6 0.9411 0.9464 0.9517 0.9550 0.9582 0.9613 0.9644 0.9676 0.9706 0.9739 0.9772 0.9802 0.9834 0.9865 0.9897 0.9930 0.9962 0.9993 1.0024 1.0056 1.0088 1.0119 1.0151 1.0182 1.0213 1.0244 1.0276 1.0309 0.9423 0.9476 0.9530 0.9562 0.9594 0.9625 0.9656 0.9688 0.9718 0.9752 0.9784 0.9815 0.9847 0.9878 0.9910 0.9942 0.9974 1.0005 1.0036 1.0068 1.0100 1.0132 1.0134 1.0195 1.0226 1.0257 1.0289 1.0321 0.9435 0.9487 0.9541 0.9573 0.9605 0.9636 0.9667 0.9700 0.9730 0.9763 0.9796 0.9826 0.9858 0.9890 0.9921 0.9953 0.9985 1.0016 1.0047 1.0080 1.0112 1.0143 1.0175 1.0206 1.0237 1.0269 1.0300 1.0332 0.9446 0.9499 0.9553 0.9585 0.9617 0.9648 0.9679 0.9711 0.9741 0.9774 0.9807 0.9837 0.9869 0.9901 0.9933 0.9965 0.9997 1.0028 1.0059 1.0091 1.0123 1.0154 1.0186 1.0217 1.0248 1.0280 1.0312 1.0344 0.9456 0.9509 0.9563 0.9595 0.9627 0.9658 0.9690 0.9721 0.9751 0.9785 0.9818 0.9848 0.9880 0.9911 0.9943 0.9975 1.0007 1.0038 1.0069 1.0102 1.0134 1.0165 1.0197 1.0228 1.0259 1.0290 1.0322 1.0354 1403c STEAM. Table of Factors of Evaporation. — Continued. Gauge Pressure. 100 105 115 125 135 145 155 165 185 Temp, of Feed. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. 300° F. 0.9467 0.9477 0.9498 0.9514 0.9532 0.9548 0.9564 0.9579 0.9606 295 0.9520 0.9530 0.9551 0.9567 0.9585 0.9601 0.9617 0.9631 0.9659 290 0.9573 0.9584 0.9604 0.9621 0.9639 0.9655 0.9671 0.9685 0.9713 287 0.9605 0.9616 0.9636 0.9653 0.9671 0.9687 0.9703 0.9717 0.9745 284 0.9637 0.9648 0.9669 0.9685 0.9703 0.9719 0.9735 0.9749 0.9777 281 0.9669 0.9679 0.9700 0.9716 0.9734 0.9750 0.9766 0.9780 0.9808 278 0.9700 0.9710 0.9731 0.9747 0.9765 0.9781 0.9797 0.9812 0.9840 275 0.9732 0.9742 0.9763 0.9779 0.9797 0.9813 0.9829 0.9844 0.9872 272 0.9762 0.9772 0.9793 0.9810 0.9827 0.9844 0.9859 0.9874 0.9902 269 0.9795 0.9805 0.9826 0.9842 0.9860 0.9877 0.9892 0.9907 0.9935 266 0.9828 0.9838 0.9859 0.9876 0.9893 0.9910 0.9925 0.9940 0.9968 263 0.9858 0.9868 0.9889 0.9906 0.9923 0.9940 0.9955 0.9970 0.9998 260 0.9890 0.9901 0.9921 0.9938 0.9955 0.9972 0.9988 1.0002 1.0030 257 0.9921 0.9932 0.9952 0.9969 0.9986 1.0003 1.0019 1.0033 1.0061 254 0.9953 0.9964 0.9984 1.0001 1.0019 1.0035 1.0051 1.0065 1.0093 251 0.9985 0.9996 1.0017 1.0033 1.0051 1.0067 1.0083 1.0097 1.0125 248 1.0018 1.0028 1.0049 1.0065 1.0083 1.0099 1.0115 1.0129 1.0167 245 1.0049 1.0059 1.0080 1.0096 1.0114 1.0130 1.0146 1.0160 1.0188 242 1.0080 1.0090 1.0111 1.0127 1.0145 1.0162 1.0177 1.0192 1.0220 239 1.0112 1.0122 1.0143 1.0159 1.0177 1.0194 1.0209 1.0224 1.0252 236 1.0144 1.0154 1.0175 1.0192 1.0209 1.0226 1.0241 1.0256 1.0284 233 1.0175 1.0185 1.0206 1.0223 1.0240 1.0257 1.0272 1.0287 1.0315 230 1.0207 1.0217 1.0238 1.0255 1.0272 1.0289 1.0304 1.0319 1.0347 227 1.0238 1.0248 1.0269 1.0286 1.0303 1.0320 1 . 0335 1.0350 1.0378 224 1.0269 1.0280 1.0300 1.0317 1.0334 1.0351 1.0367 1.0381 1.0409 221 1.0300 1.0311 1.0331 1.0348 1.0365 1.0382 1.0398 1.0412 1.0440 218 1.0332 1.0343 1.0363 1.0380 1.0398 1.0414 1.0430 1.0444 1.0472 215 1.0364 1.0375 1.0395 1.0412 1.0430 1.0446 1.0462 1.0476 1.0504 FACTORS OF EVAPORATION. 1403d Table of factors of Evaporation. — Concluded. Gauge Pressure. 200 215 230 245 260 275 290 300 Temp, of Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Lbs. Feed. 300 °F. 0.9626 0.9645 0.9662 0.9679 0.9694 0.9710 0.9724 0.9734 295 0.9679 0.9697 0.9715 0.9732 0.9747 0.9763 0.9777 0.9787 290 0.9733 0.9751 0.9769 0.9786 0.9801 0.9817 0.9831 0.9840 287 0.9765 0.9783 0.9801 0.9818 0.9833 0.9849 0.9863 0.9873 284 0.9797 0.9S16 0.9833 0.9850 0.9865 0.9881 0.9895 0.9905 281 0.9828 0.9847 0.9864 0.9881 0.9896 0.9912 0.9926 0.9936 278 0.9859 0.9878 0.9895 0.9912 0.9927 0.9943 0.9957 0.9967 275 0.9891 0.9910 0.9927 0.9944 0.9959 0.9975 0.9989 0.9999 272 0.9921 0.9940 0.9958 0.9974 0.9990 1.0005 1.0020 1.0029 269 0.9954 0.9973 0.9991 1.0007 1.0023 1.0038 1.0053 1.0063 266 0.9987 1.0006 1.0024 1.0040 1.0056 1.0071 1.0086 1.0095 263 1.0017 1.0036 1.0054 1.0070 1.0086 1.0102 1.0116 1.0125 260 1.0049 1.0058 1.0086 1.0103 1.0118 1.0133 1.0148 1.0157 257 1.0081 1.0099 1.0117 1.0134 1.0149 1.0164 1.0179 1.0188 254 1.0113 1.0132 1.0149 1.0166 1.0181 1.0197 1.0211 1.0221 251 1.0145 1.0164 1.0181 1.0198 1.0213 1.0229 1.0243 1.0253 248 1.0177 1.0196 1.0213 1.0230 1.0245 1.0261 1.0275 1.0285 245 1.0208 1.0227 1.0244 1.0261 1.0276 1.0292 1.0306 1.0316 242 1.0239 1.0258 1.0275 1.0293 1.0308 1.0323 1.0337 1.0347 239 1.0271 1.0290 1.0307 1.0324 1.0340 1.0355 1.0370 1.0379 236 1.0303 1.0322 1.0340 1.0356 1.0372 1.0387 1.0402 1.0411 233 1.0334 1.0353 1.0371 1.0387 1.0403 1.0418 1.0433 1.0442 230 1.0367 1.0385 1.0403 1.0419 1.0435 1.0450 1.0465 1.0474 227 1.0398 1.0416 1.0434 1.0450 1.0466 1.0482 1.0496 1.0505 224 1.0429 1.0447 1.0465 1.0482 1.0497 1.0513 1.0527 1.0536 221 1.0460 1.0478 1.0496 1.0512 1.0528 1.0544 1.0558 1.0567 218 1.0492 1.0511 1.0528 1.0545 1.0560 1.0576 1.0590 1.0600 215 1.0524 1.0543 1.0562 1.0577 1.0592 1.0608 1.0622 1.0632 PROPERTUS OF SiTlRlTEI) ITEAM. (W. W. Christie, M.E.) e ° si Sf'oS 0) .• 6 *x* hi o g SPLH H Heat Units in one Pound above 32° F. Volume. « o 1* .a* *3 cot A v hi G 3«*« 1&A Relative s Specific oo g* go > Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft. in one Lb. of Steam. -ail JqQQ 29.74 29.72 29.71 29.70 .089 .096 .104 .112 32 34 36 38 2 4 6 1092.7 1090.37 1088.98 1087.59 1092.7 1092.37 1092.98 1093.59 208080 193180 179380 166380 3387 3138 2910 2700 .000295 318 344 370 29.68 29.65 29.63 29.61 .122 .132 .142 .152 40 42 44 46 8 10 12 14 1086.20 1084.81 1083.41 1082.02 1094.20 1094.81 1095.41 1096.02 154330 143220 133120 123840 2506 2328 2164 2013 399 429 462 496 29.59 29.56 29.54 29.51 .164 .176 .190 .205 48 50 52 54 16 18 20 22 1080.63 1079.25 1077.86 1076.47 1096.63 1097.25 1097.86 1098.47 115490 107630 100330 93680 1874 1745 1626 1516 533 573 615 659 29.47 29.44 29.40 29.37 .220 .236 .254 .273 56 58 60 62 24 26 28 30 1075.08 1073.69 1072.31 1070.92 1099.08 1099.69 1100.31 1100.92 87500 81740 76370 71330 1415 1321 1234 1153 706 757 810 867 29.33 29.28 29.24 29.19 .292 .313 .335 .359 64 66 68 70 32 34 36 38 1069.53 1068.14 1066.75 1065.35 1101.53 1102.14 1102.75 1103.35 66630 62290 58340 54660 1078 1009 944.7 885.0 927 991 .001059 1130 29.14 29.09 29.03 28.96 .385 .411 .440 .470 72 74 76 78 40 42 44 46 1063.96 1062.57 1061.18 1059.79 1103.96 1104.57 1105.18 1105.79 51210 48000 45060 42280 829.5 777.9 729.9 685.2 1205 1286 1370 1459 28 90 28.83 28.76 28.68 .502 .535 .571 .609 80 82 84 86 48 50 52 54 1058.40 1057.01 1055.62 1054.22 1106.40 1107.01 1107.62 1108.23 39690 37320 35100 33030 643.8 605.0 568.8 535.2 1553 1653 1758 1869 28.60 28.51 28.42 28.32 .650 .692 .738 .785 88 90 92 94 56 58 60 62 1052.83 1051.44 1050.05 1048.66 1108.84 1109.45 1110.06 1110.67 31100 29290 27600 26020 503.7 474.6 447.1 421.5 1985 .002107 2237 2372 28.22 28.11 28.00 27.89 .834 .887 .943 1.001 96 98 100 102 64 66 68 70 1047.27 1045.87 1044.48 1043.08 1111.28 1111.89 1112.50 1113.10 24540 23140 21830 20620 397.5 375.1 354.0 334.5 2516 2666 2824 2900 27.76 27.63 27.49 27.34 1.062 1.126 1.193 1.265 104 106 108 110 72 74 76 78 1041.69 1040.29 1038.90 1037.52 1113.71 1114.32 1114.93 1115.55 19500 18460 17470 16520 316.1 298.8 282.7 267.5 .003163 3347 3537 3738 27.19 27.03 26.86 26.68 1.341 1.421 1.504 1.591 112 114 116 118 80 82 84 86 1036.12 1034.74 1033.35 1031.94 1116.16 1116.78 1117.39 1117.99 15640 14820 14050 13320 253.3 239.9 227.3 215.5 3948 004168 4399 4640 26.49 26.30 1.682 1.779 120 122 88 90 1030.55 1029.16 1118.60 1119.21 12630 11980 204.4 193.9 1 4892 005156 1404 PROPERTIES OF SATtHATED §TE AHf . — Continued. m 5 « 02 (h ■gi* n Heat Units in one Pound above 32° F. Volume. a o o .3 .sg * • Ill's Rela- tive. Specific Cu. Ft. inlCu. Ft. of Water. Cu. Ft. in one Lb. of Steam. ■Sf§! 'SOW 26.09 25.88 25.65 25.41 1.879 1.984 2.096 2.213 124 126 128 130 92 94 96 98 1027.76 1026.37 1024.97 1023.58 1119.82 1120.43 1121.04 1121.65 11370 10800 10265 9760 184.1 174.8 166.1 157.8 .005432 5720 6020 6336 25.17 24.91 24.64 24.36 2.335 2.461 2.594 2.732 132 134 136 138 100 102 104 106 1022.18 1020.79 1019.39 1018.00 1122.26 1122.87 1123.48 1124.09 9276 8826 8401 7991 150.1 142.8 135.8 129.3 6664 7005 7361 7732 24.06 23.75 23.43 23.09 2.876 3.029 3.188 3.353 140 142 144 146 108.1 110.1 112.1 114.1 1016.60 1015.20 1013.81 1012.41 1124.70 1125.31 1125.92 1126.53 7613 7258 6920 6595 123.2 117.3 111.8 106.6 8120 8522 8942 9379 22.74 22.37 21.99 21.59 3.526 3.707 3.896 4.090 148 150 152 154 116.1 118.1 120.1 122.1 1011.01 1009.61 1008.22 1006.82 1127.14 1127.75 1128.36 1128.97 6290 6004 5734 5477 101.7 97.03 92.61 88.43 .009833 .01031 .01080 .01131 21.17 20.74 20.29 19.82 4.295 4.507 4.729 4.960 156 158 160 162 124.1 126.1 128.1 130.1 1005.42 1004.02 1002.62 1001.22 1129.85 1130.19 1130.80 1131.41 5232 5000 4779 4569 84.47 80.70 77.14 73.77 .01184 .01239 .01296 .01356 19.33 18.82 18.29 17.76 5.200 5.451 5.711 5.981 164 166 168 170 132.2 134.2 136.2 138.2 999.82 998.42 997.02 995.62 1132.02 1132.63 1133.24 1133.85 4368 4177 3996 3826 70.56 67.51 64.62 61.85 .01417 .01481 .01548 .01617 17.16 16.57 15.95 15.31 6.262 6.555 6.857 7.172 172 174 176 178 140.2 142.2 144.2 146.2 994.22 992.82 991.42 990.02 1134.46 1135.07 1135.68 1136.29 3664 3510 3365 3226 59.25 56.76 54.40 52.14 .01688 .01762 .01838 .01918 14.64 13.95 13.23 12.48 7.500 7.841 8.194 8.558 180 182 184 186 148.2 150.3 152.3 154.3 983.62 987.21 985.81 984.41 1136.90 1137.51 1138.12 1138.73 3093 2966 2846 2733 50.01 47.97 46.06 44.17 .02000 .02085 .02172 .02264 11.71 10.91 10.08 9.22 8.936 9.330 9.738 10.16 188 190 192 194 156.3 158.3 160.3 162.3 983.00 9S1.60 980.20 978.79 1139.34 1139.95 1140.56 1141.17 2624 2519 2420 2325 42.41 40.73 39.13 37.59 .02358 .02455 .02556 .C2660 8.33 7.40 6.45 5.46 10.59 11.05 11.52 12.00 196 198 200 202 164.3 166.4 168.4 170.4 977.39 975.98 974.58 973.17 1141.78 1142.39 1143.00 1143.61 2234 2147 2064 1985 36.13 34.73 33.40 32.13 .02768 .02879 .02994 .03112 4.44 3.38 2.28 1.15 12.50 13.02 13.56 14.12 204 206 208 210 172.4 174.4 176.4 178.5 971.76 970.36 968.95 967.54 1144.22 1144.83 1145.44 1146.05 1916 1844 1775 1708 30.92 29.76 28.63 27.57 .03235 .03361 .03493 .03628 0.00 14.70 212 |180.5 966.13 1146.66 1644 26.60 .03760 1405 1406 STEAM. PROPERTIEi OF SATURATED (Compiled by W. W. Christie.) STEAM. Pounds per Square Inch. Cup IS Heat Units in one Pound above 32° F. Vol nme. 6 00 6 Si at S3 2,2 3 *5 IIhWm Rela- tive Specific Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft, in one Lb. of Steam. 1 2 3 4 102. 126.2 141.6 153.0 70.1 94.4 109.8 121.4 1042.9 1026.0 1015.2 1007.2 1113.0 1120.4 1125.1 1128.6 20623 16730 7325 5588 330.4 171.9 117.3 89.51 .0030 .0058 .0085 .0112 5 6 7 8 162.3 170.1 176.9 182.9 130.7 138.5 145.4 151.4 1000.7 995.2 990.4 986.2 1131.4 1133.8 1135.8 1137.7 4530 3816 3302 2912 72.56 61.14 52.89 46.65 .0138 .0164 .0189 .0214 9 10 11 12 188.3 193.2 197.7 201.9 156.9 161.9 166.5 170.7 982.4 978.9 975.7 972.8 1139.3 1140.8 1142.2 1143.5 2607 2361 2159 1990 41,77 37.83 34.59 31.87 ,0239 .0264 .0289 .0314 .'304 1.3 13 14 15 16 205.8 209.5 213.0 216.3 174.7 178.4 181.9 185.2 970.0 967.4 964.9 962.6 1144.7 1145.8 1146.9 1147.9 1845 1721 1614 1519 29.56 27.58 25.85 24.33 .0338 .0363 .0387 .0411 2.3 3.3 4.3 5.3 17 18 19 20 219.4 222.3 225.2 227.9 188.4 191.4 194.2 197.0 960.4 958.3 956.3 954.4 1148.8 1149.7 1150.6 1151.4 1434 1359 1292 1231 22.98 21.72 20.70 19.73 .0435 .0459 .0483 .0507 6.3 7.3 8.3 9.3 21 22 23 24 230.5 233.0 235.4 237.7 199.6 202.2 204.6 207.0 952.5 950.8 949.0 947.4 1152.2 1153.0 1153.7 1154.4 1176 1126 1080 1038 18.84 18.04 17.30 16.62 .0531 .0554 .0578 .0602 10.3 11.3 12.3 13.3 25 26 27 28 240.0 242.1 244.2 246.3 209.3 211.5 213.6 215.7 945.8 944.2 942.7 941.3 1155.1 1155.8 1156.4 1157.0 998.4 962.3 928.8 897.6 16.00 15.42 14.88 14.38 .0625 .0649 .0672 .0695 14.3 15.3 16.3 17.3 29 30 31 32 248.3 250.2 252.1 253.9 217.7 219.7 221.6 223.5 939.9 938.9 937.1 935.9 1157.6 1158.2 1158.8 1159.3 868.5 841.3 815.8 791.8 13.91 13.48 13.07 12.68 .0719 .0742 .0765 .0788 18.3 19.3 20.3 21.3 33 34 35 36 255.7 257.4 259.1 260.8 225.3 227.1 228.8 230.5 934.6 933.3 932.1 931.0 1159.9 1160.4 1160.9 1161.5 769.2 748.0 727.9 708.8 12.32 11.98 11.66 11.37 .0812 .0835 .0858 .0881 22.3 23 3 24.3 25.3 37 38 39 40 262.4 264.0 265.6 267.1 232.1 233.8 235.3 236.9 929.8 928.6 927.5 926.4 1161.9 1162.4 1162.9 1163.4 690.8 673.7 657.5 642.0 11.07 10.79 10.53 10.28 .0904 .0927 .0949 .0972 1 2 6.3 7.3 41 42 268.6 270.0 238.4 239.9 925.4 924.3 1163.8 1164.3 627.3 613.3 10.05 9.826 .0995 .1018 PROPERTIES OF SATURATED STEAM. 1407 PROPERTIES OE SATURATED STEAM — Continued. Pounds per Square Inch. 4a e3 . — ** o 3 §£ Heat Units in one Pound above 32° F. Vol ume. 4-1 C MP $2 H Rela- tive Specific ° 2 8)£ §Ah 6 Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft. in one Lb. of Steam. 28.3 29.3 30.3 31.3 43 44 45 . 46 271.5 272.9 274.3 275.6 241.4 242.8 244.2 245.6 923.3 922.3 921.3 920.3 1164.7 1165.1 1165.6 1166.0 599.9 587.0 574.7 563.0 9.609 9.403 9.207 9.018 .1041 .1063 .1086 .1109 32.3 33.3 34.3 35.3 47 58 49 50 276.9 278.2 279.5 280.8 247.0 248.3 249.6 250.9 919.4 918.4 917.5 916.6 1166.4 1166.8 1167.2 1167.6 551.7 540.9 530.5 520.5 8.838 8.665 8.498 8.338 .1131 .1154 .1177 .1199 36.3 37.3 38.3 39.3 51 52 53 54 282.1 283.3 284.5 285.7 252.2 253.5 254.7 255.9 915.7 914.8 913.9 913.1 1167.9 1168.3 1168.7 1169.0 510.9 501.7 492.8 484.2 8.185 8.037 7.894 7.756 .1222 .1244 .1267 .1289 40.3 41.3 42.3 43.3 55 56 57 58 286.9 288.0 289.1 290.3 257.1 258.3 259.5 260.6 912.2 911.4 910.6 909.8 1169.4 1169.7 1170.1 1170.4 475.9 467.9 460.2 452.7 7.624 7.496 7.372 7.252 .1312 .1334 .1357 .1379 44.3 45.3 46.3 47.3 59 60 61 62 291.4 292.5 293.6 294.6 261.7 262.9 264.0 265.1 909.0 908.2 907.4 906.7 1170.8 1171.1 1171.4 1171.8 445.5 438.5 431.7 425.2 7.136 7.024 6.916 6.811 .1401 .1424 .1446 .1468 48.3 49.3 50.3 51.3 63 64 65 66 295.7 296.7 297.7 298.7 266.1 267.2 268.3 269.3 905.9 905.2 904.4 903.7 1172.1 1172.4 1172.7 1173.0 418.8 412.6 406.6 400.8 6.709 6.610 6.515 6.422 .1491 .1513 .1535 .1557 52.3 53.3 54.3 55.3 67 68 69 70 299.7 300.7 301.7 302.7 270.3 271.3 272.3 273.3 903.0 902.3 901.5 900.9 1173.3 1173.6 1173.9 1174.2 395.2 389.8 384.5 379.3 6.332 6.244 6.159 6.076 .1579 .1602 .1624 .1646 56.3 57.3 58.3 59.3 71 72 73 74 303.6 304.6 305.5 306.4 274.3 275.3 276.2 277.2 900.2 899.5 898.8 898.1 1174.5 1174.8 1175.1 1175.4 374.3 369.4 364.6 360.0 5.995 5.917 5.841 5.767 .1668 .1690 .1712 .1734 60.3 61.3 62.3 63.3 75 76 77 78 307.3 308.2 309.1 310.0 278.1 279.0 280.0 280.9 897.5 896.8 896.2 895.5 1175.6 1175.9 1176.2 1176.5 355.5 351.1 346.8 342.6 5.694 5.624 5.555 5.488 .1756 .1778 .1800 .1822 64.3 65.3 66.3 67.3 79 80 81 82 310.9 311.8 312.6 313.5 281.8 282.7 283.5 284.4 894.9 894.3 893.7 893.1 1176.7 1177.0 1177.3 1177.5 338.5 334.5 330.6 326.8 5.422 5.358 5.296 5.235 .1844 .1866 .1888 .1910 68.3 69.3 83 84 314.3 315.1 285.3 286.1 892.4 891.8 1177.8 1178.0 323.1 319.5 5.176 5.118 .1932 .1954 1408 STEAM. PROPERTIES OF SATtRATED STE Alff — Continued. Pounds per Square Inch. 43 . ^ o g • OQ ft® 2 Qj H Heat Units in one Pounds above 32° F. Volume. 6 u OQ 9 u Si 5* +3 TO ft--. Rela- tive Specific Cu. Ft. in lCu. Ft. of Water. Cu. Ft. n lLb. of Steam. 70.3 71.3 72.3 73.3 85 86 87 88 316.0 316.8 317.6 318.4 287.0 287.8 288.7 289.5 891.2 890.6 890.1 889.5 1178.3 1178.5 1178.8 1179.0 315.9 312.5 309.1 305.8 5.061 5.006 4.951 4.898 .1976 .1998 .2020 .2042 74.3 75.3 76.3 77.3 89 90 91 92 319.2 320.0 320.8 321.6 290.3 291.1 291.9 292.7 888.9 888.3 887.8 887.2 1179.3 1179.5 1179.8 1180.0 302.5 299.4 296.3 293.2 4.846 4.796 4.746 4.697 .2063 .2085 .2107 .2129 78.3 79.3 *0.3 31.3 93 94 95 96 322.3 323.1 323.8 324.6 293.5 294.3 295.1 295.9 886.6 886.1 885.5 885.0 1180.2 1180.4 1180.7 1180.9 290.2 287.3 284.5 281.7 4.650 4.603 4.557 4.513 .2151 .2173 .2194 .2216 32.3 i*3.3 84.3 85.3 97 98 99 100 325.3 326.1 326.8 327.5 296.6 297.4 298.1 298.9 884.5 883.9 883.4 882.9 1181.1 1181.4 1181.6 1181.8 279.0 276.3 273.7 271.1 4.469 4.426 4.384 4.342 .2238 .2260 .2281 .2303 U6.3 W.3 (58.3 *;9.3 101 102 103 104 328.2 329.0 329.7 330.4 299.6 300.4 301.1 301.8 882.3 881.8 881.3 880.8 1182.0 1182.2 1182.5 1182.7 268.5 266.0 263.6 261.2 4.302 4.262 4.223 4.185 .2325 .2346 .2368 .2390 £(0.3 S'1.3 112.3 93.3 105 106 107 108 331.1 331.8 332.4 333.1 302.5 303.3 304.0 304.7 880.3 879.8 879.3 878.8 1182.9 1183.1 1183.3 1183.5 258.9 256.6 254.3 252.1 4.147 4.110 4.074 4.038 .2411 .2433 .2455 .2476 34.3 95.3 96.3 97.3 109 110 111 112 333.8 334.5 335.1 335.8 305.4 306.1 306.8 307.4 878.3 877.8 877.3 876.9 1183.7 1183.9 1184.1 1184.3 249.9 247.8 245.7 243.6 4.003 3.969 3.935 3.902 .2498 .2519 .2541 .2563 S8.3 99.3 100.3 in .3 113 114 115 116 336.5 337.1 337.8 338.4 308.1 308.8 309.5 310.1 876.4 875.9 875.4 875.0 1184.5 1184.7 1184.9 1185.1 241.6 239.6 237.6 235.7 3.870 3.838 3.806 3.775 .2584 .2606 .2627 .2649 102.3 103.3 104.3 105.3 117 118 119 120 339.1 339.7 340.3 340.9 310.8 311.4 312.1 312.7 874.5 874.0 873.6 873.1 1185.3 1185.5" 1185.7 1185.9 233.8 231.9 230.1 228.3 3.745 3.715 3.685 3.656 .2670 .2692 .2713 .2735 108.3 107.3 103.3 103.3 121 122 123 124 341.6 342.2 342.8 343.4 313.4 314.0 314.7 315.3 872.7 872.5 871.8 871.3 1186.1 1186.3 1186.5 1186.6 226.5 224.7 223.0 221.3 3.628 3.600 3.572 3.545 .2757 .2778 .2800 .2821 115.3 111.3 125 126 344.0 344.6 315.9 316.6 870.9 870.4 1186.8 1187.0 219.6 218.0 3.518 3.492 .2842 .2864 PROPERTIES OF SATURATED STEAM. 1409 PROPERTIES OF SATURATED 8TE AM — Continued. Pounds per Square Inch. o o 6 Heat Units in one Pound above 32° F. Volume. o © d*a 6 u p 6 4 3 d U O H si a ° * d ±* c3 »»;h h3 a. fl • II °3® H Rela- tive TmTFtT. in 1 Cu Ft. of Water. Specific °o .SPdJS 'von Cu. Ft. inlLb. of Steam . 112.3 113.3 114.3 115.3 127 128 129 130 345.2 345.8 346.4 347.0 317.2 317.8 318.4 319.0 870.0 869.6 869.1 868.7 1187.2 1187.4 1187.6 1187.8 216.4 214.8 213.2 211.6 3.466 3.440 3.415 3.390 .2885 .2907 .2928 .2950 116.3 117.3 118.3 119.3 131 132 133 134 347.6 348.2 348.8 349.3 319.6 320.2 320.8 321.4 868.3 867.8 867.4 867.0 1187.9 1188.1 1188.3 1188.5 210.1 208.6 207.1 205.7 3.366 3.342 3.318 3.295 .2971 .2992 .3014 .3035 120.3 121.3 122.3 123.3 135 136 137 138 349.9 350.5 351.0 351.7 322.0 322.6 323.2 323.8 866.6 866.2 865.7 865.3 1188.6 1188.8 1189.0 1189.1 204.2 202.8 201.4 200.0 3.272 3.249 3.227 3.204 .3057 .3078 .3099 .3121 124.3 125.3 126.3 127.3 139 140 141 142 352.2 352.7 353.3 353.8 324.3 324.9 325.5 326.1 864.9 864.5 864.1 863.7 1189.3 1189.5 1189.7 1189.8 198.7 197.3 196.0 194.7 3.182 3.161 3.140 3.119 .3142 .3163 .3185 .3206 128.3 129.3 130.3 131.3 143 144 145 146 354.4 354.9 355.5 356.0 326.8 327.2 327.8 328.3 863.3 862.9 862.5 862.1 1190.0 1190.2 1190.3 1190.4 193.4 192.2 190.9 189.7 3.099 3.078 3.058 3.038 .3227 .3249 .3270 .3291 132.3 133.3 134.3 135.3 147 148 149 150 356.5 357.1 357.6 358.1 328.9 329.4 330.0 330.5 861.7 861.4 861.0 860.6 1190.6 1190.8 1191.0 1191.1 188.5 187.3 186.1 184.9 3.019 3.000 2.981 2.962 .3313 .3334 .3355 .3376 136.3 137.3 138.3 139.3 151 152 153 154 358.6 359.2 359.7 360.2 331.1 331.6 332.2 332.7 860.2 859.8 859.4 859.1 1191.3 1191.4 1191.6 1191.8 183.7 182,6 181.5 180.4 2.943 2.925 2.908 2.890 .3398 .3419 .3439 .3460 140.3 141.3 142.3 143.3 155 156 157 158 360.7 361.2 361.7 362.2 333.2 333.7 334.3 334.8 858.7 858.3 857.9 857.6 1191.9 1192.1 1192.2 1192.4 179.2 178.1 177.0 176.0 2.870 2.853 2.835 2.819 .3484 .3505 .3526 .3547 144.3 145.3 146.3 147.3 159 160 161 162 362.7 363.2 363.7 364.2 335.3 335.8 336.3 336.9 857.2 856.8 856.5 856.1 1192.5 1192.7 1192.8 1193.0 174.9 173.9 172.9 171.9 2.802 2.786 2.770 2.754 .3568 .3589 .3610 .3631 148.3 149.3 150.3 151.3 163 164 165 166 364.7 365.2 365.7 366.2 337.4 337.9 338.4 338.9 855.7 855.4 855.0 854.7 1193.1 1193.3 1193.5 1193.6 171.0 170.0 169.0 168.1 2.739 2.723 2.707 2.693 .3650 .3672 .3693 .3714 152.3 153.3 167 168 366.7 367.1 339.4 339.9 854.3 853.9 1193.7 1193.9 167.1 16§.2 2.677 2.662 .3735 .3756 1410 STEAM. PROPERTIES OF SIT1H4TE1) 8TEAM- Continued. Pounds per Square Inch. o fl g is H Heat Units in One Pound above 32° F. Volume. © o 6 u 9 00 .8* 2^ S3 a* < S*» o o >«5 ii o «.£ w Rela- tive Specific ° 2 «"£ -£LS ci .SPSS ® |2 Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft. in 1 Lb. of Steam. 154.3 155.3 156.3 157.3 169 170 171 172 367.6 368.1 368.6 369.1 340.4 340.9 341.4 341.9 853.6 853.2 852.9 852.6 1194.0 1194.2 1194 3 1194.5 165.3 164.3 163.4 162.5 2.648 2.631 2.617 2.603 .3777 .3799 .3820 .3842 158.3 159.3 160.3 161.3 173 174 175 176 369.5 370.0 370.5 370.9 342.4 342.8 343.3 343.8 852.2 851.9 851.5 851.2 1194.6 1194.8 1194.9 1195.0 161.6 160,7 159.8 158.9 2.588 2.574 2.560 2.545 .3863 .3885 .3906 .3928 162.3 163.3 164.3 165.3 177 178 179 180 371.4 371.9 372.3 372.8 344.3 344.8 345.3 345.7 850.8 850.5 850.2 849.8 1195.2 1195.3 1195.5 1195.6 158.1 157.2 156.4 155.6 2.533 2.518 2.505 2.493 .3949 .3970 .3991 .4012 166.3 167.3 168.3 169.3 181 182 183 184 373.2 373.7 374.1 374.6 346.2 346.7 347.1 347.6 849.5 849.2 848.8 848.5 1195.7 1195.9 1196.0 1196.2 154.8 154.0 153.2 152.4 2.480 2.467 2.454 2.441 .4033 .4054 .4075 .4096 170.3 171.3 172.3 173.3 185 186 187 188 375.0 375.5 375.9 376.4 348.1 348.6 349.0 349.5 848.2 847.8 847.5 847.2 1196.3 1196.4 1196.6 1196.7 151.6 150.8 150.0 149.2 2.428 2.416 2.403 2.390 .4118 .4140 .4162 .4183 174.3 175.3 176.3 177.3 189 190 191 192 376.8 377.2 377.7 378.1 349.9 350.4 350.8 351.3 846.9 846.5 846.2 845.9 1196.8 1197.0 1197.1 1197.2 148.5 147.8 147.0 146.3 2.379 2.367 2.355 2.344 .4204 .4225 .4246 .4267 178.3 179.3 180.3 181.3 193 194 195 196 378.5 379.0 379.4 379.9 351.7 352.2 352.6 353.1 845.6 845.3 845.0 844.6 1197.4 1197.5 1197.6 1197.8 145.6 144.9 144.2 143.5 2.332 2.321 2.310 2.299 .4287 .4308 .4329 .4350 182.3 183.3 184.3 185.3 197 198 199 200 380.3 380.7 381.1 381.5 353.5 354.0 354.4 354.8 844.3 844.0 843.7 843.4 1197.9 1198.0 1198.1 1198.3 142.8 142.1 141.4 140.8 2.287 2.276 2.265 2.255 .4372 .4393 .4414 .4435 186.3 187.3 188.3 189.3 201 202 203 204 381.9 382.4 382.8 383.2 355.3 355.7 356.1 356.6 843.1 842.8 842.5 842.2 1198.4 1198.5 1198.7 1198.8 140.1 139.5 138.8 138.1 2.244 2.235 2.223 2.212 .4456 .4477 .4498 .4520 190.3 191.3 192.3 193.3 205 206 207 208 383.6 384.0 384.4 384.8 357.0 357.4 357.9 358.3 841.8 841.5 841.2 841.0 1198.9 1199.0 1199.2 1199.3 137.5 136.9 136.3 135.7 2.203 2.193 2.183 2.174 .4540 .4560 .4580 .4600 194.3 195.3 209 210 385.2 385.6 £58.7 359.1 840.7 840.4 1199 4 1199.5 135.1 134.5 2.164 2.154 .4621 .4642 PROPERTIES OP SATURATED STEAM. 1411 PROPERTIES OT SATURATED STEA1T — Continued. Founds per Square Inch. 4> Heat Units in One Pound above 32° F. Volume. o 6 u 3 to 6 ■3g .8* < a* 03 © c3+j» lldHS Rela- tive Specific &2 Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft. inl Lb. of Steam. 196.3 197.3 198.3 199.3 211 212 213 214 386.1 386.5 386.9 387.3 359.6 360.0 360.4 360.9 840.1 839.8 839.5 839.2 1199.7 1199.8 1199.9 1200.1 133.9 133.3 132.8 132.2 2.145 2.135 2.126 2.117 .4663 .4684 .4705 .4726 200.3 201.3 202.3 203.3 215 216 217 218 387.7 388.1 388.5 388.9 361.3 361.7 362.1 362.5 838.9 838.6 838.3 838.0 1200.2 1200.3 1200.4 1200.5 131.6 131.0 130.4 129.9 2.108 2.098 2.089 2.080 .4747 .4768 .4789 .4810 204.3 205.3 206.3 207.3 219 220 221 222 389.3 389.6 390.1 390.5 362.9 363.3 363.7 364.1 837.8 837.5 837.3 837.0 1200.7 1200.8 1201.0 1201.1 129.3 128.7 128.1 127.6 2.070 2.061 2.052 2.043 .4831 .4852 .4873 .4894 208.3 209.3 210.3 211.3 223 224 225 226 390.8 391.2 391.6 392.0 364.5 364.9 365.3 365.8 836.7 836.4 836.1 835.8 1201.2 1201.3 1201.4 1201.6 127.0 126.5 126.0 125.4 2.035 2.027 2.018 2.010 .4915 .4936 .4956 .4977 212.3 213.3 214.3 215.3 227 228 229 230 392.4 392.8 393.2 393.5 366.1 366.5 366.9 367.3 835.6 835.3 835.0 834.7 1201.7 1201.8 1201.9 1202.0 124.9 124.4 123.9 123.3 2.002 1.993 1.984 1.976 .4998 .5019 .5040 .5061 216.3 217.3 218.3 219.3 231 232 233 234 393.9 394.3 394.7 395.1 367.7 368.1 368.5 368.9 834.4 834.1 833.9 833.6 1202.1 1202.2 1202.4 1202.5 122.9 122.4 121.9 121.4 1.968 1.960 1.952 1.944 .5082 .5103 .5124 .5145 220.3 221.3 222.3 223.3 235 236 237 238 395.5 395.9 396.3 396.6 369.2 369.6 370. 370.4 833.4 833.1 832.8 832.5 1202.6 1202.7 1202.8 1202.9 120.9 120.4 119.9 119.4 1.936 1.928 1.921 1.913 .5165 .5186 .5207 .5228 224.3 225.3 226.3 227.3 239 240 241 242 397.0 397.4 397.8 398.1 370.8 371.1 371.5 371.9 832.2 832.0 831.7 831.4 1203.0 1203.1 1203.2 1203.3 119.0 118.5 118.0 117.5 1.905 1.898 1.891 1.884 .5249 .5270 .5291 .5312 228.3 229.3 230.3 231.3 243 244 245 246 398.5 398.9 399.2 399.6 372.3 372.7 373.1 373.4 831.1 830.8 830.6 830.4 1203.4 1203.5 1203.7 1203.8 117.1 116.7 116.2 115.7 1.857 1.868 1.861 1.853 .5332 .5353 .5374 .5395 232.3 233.3 234.3 235.3 247 248 249 250 400.0 400.3 400.7 401.1 373.8 374.2 374.6 375.0 830.1 829.8 829.5 829.2 1203.9 1204.0 1204.1 1204.2 115.3 114.9 114.4 114.0 1.846 1.839 1.832 1.825 .5416 .5436 .5457 .5478 238.3 241.3 253 256 402.1 403.1 376.0 377.0 828.5 827.9 1204.5 1204.9 112.7 111.4 1.806 1.785 .5540 .5603 1412 STEAM. PROPERTIES OE SATURATED STEAM — Continued. Pounds per Square Inch. ° © © £ •*? 03 a © u u S3 © Heat Units in One Pound above 32° F. Volume. © 9 Cu. Ft. in 1 Cu. Ft. of Water. Cu. Ft. in 1 Lb. of Steam. -as* 244.3 247.3 250.3 253.3 259 262 265 268 404.2 405.2 406.1 407.2 378.1 379.2 380.2 381.2 827.1 826.3 825.6 824.9 1205.2 1205.5 1205.8 1206.1 110.2 109.2 107.8 106.7 1.766 1.746 1.728 1.709 .5665 .5727 .5789 .5852 256.3 259.3 262.3 265.3 271 274 277 280 408.1 409.1 410.0 411.1 382.3 383.3 384.3 385.3 824.1 823.4 822.7 822.0 1206.4 1206.7 1207.0 1207.3 105.6 104.5 103.4 102.3 1.691 1.673 1.656 1.639 .5914 .5976 .6039 .6101 268.3 271.3 274.3 277.3 283 286 289 292 412.1 413.0 414.0 415.0 386.3 387.3 388.3 389.2 821.3 820.6 819.9 819.3 1207.6 1207.9 1208.2 1208.5 101.3 100.3 99.3 98.35 1.621 1.606 1.591 1.575 .6164 .6226 .6288 .6350 280.3 283.3 285.3 290.3 295 298 300 305 415.9 416.9 417.4 418.9 390.2 391.1 391.9 394.5 818.6 818.0 817.4 815.2 1208.8 1209.1 1209.3 1209.7 97.42 96.47 95.8 94.37 1.560 1.545 1.535 1.510 .6412 .6474 .6515 .6618 295.3 300.3 305.3 310.3 310 315 320 325 420.5 421.9 423.4 424.8 396.0 397.6 399.1 400.6 814.2 813.0 812.0 810.9 1210.2 1210.6 1211.1 1211.5 92.92 91.52 90.16 88.84 1.488 1.465 1.443 1.422 .6721 .6824 .6927 .7030 315.3 320.3 325.3 330.3 330 335 340 345 426.3 427.7 429.1 430.5 402.1 403.6 404.8 406.1 809.8 808.8 808.1 807.2 1211.9 1212.4 1212.9 1213.3 87.55 86.31 85.10 83.92 1.401 1.382 1.394 1.343 .7133 .7236 .7339 .7442 335.3 385.3 435.3 485.3 350 400 450 500 431.96 444.9 456.6 467.4 407.3 420.8 433.2 444.5 806.4 796.9 788.1 780.0 1213.7 1217.7 1221.3 1224.5 82.71 72.8 65.1 58.8 1.325 1.167 1.042 .942 .7545 .8572 .9595 1.0617 535.3 585.3 635.3 685.3 550 600 650 700 477.5 486.9 495.7 504.1 455.1 465.2 474.6 483.4 772.5 765.3 758.6 752.3 1227.6 1230.5 1233.2 1235.7 53.6 49.3 45.6 42.4 .859 .790 .731 .680 1.1638 1.2659 1.3679 1.4699 735.3 785.3 835.3 885.3 750 800 850 900 512.1 519.6 526.8 533.7 491.9 499.9 507.7 515.0 746.1 740.4 734.8 729.7 1238.0 1240.3 1242.5 1244.7 39.6 37.1 34.9 33.0 .636 .597 .563 .532' 1.5720 1.6740 1.7760 1.8780 935.3 985.3 950 1000 540.3 546.8 523.3 529.3 723.4 719.4 1246.7 1248.7 31.4 30.0 .505 .480 1.9800 2.0820 SUPERHEATED STEAM. 1413 SUPERHEATED ITEA9I. Dry saturated steam, after being heated to a higher temperature than that corresponding to its pressure, is called superheated steam. The behavior of superheated steam is similar to that of gases ; it is a bad conductor of heat, and can lose some of its heat without becoming saturated or wet steam. Superheated steam has a greater volume per unit of weight than saturated steam at the same pressure. Pressure, Pounds. 70 115 170 Vol< > at 390° F. . 1.1 1.33 1.57 1.06 1.29 1.52 1.02 "Lenke" Vol at 570° F. . 1.24 Vol at 750° F. . 1.46 Saturated steam in engines condenses during admission to 20% to 25% of the quantity admitted, causing a large part of the low theoretical efficiency when it is used. Superheated steam does not condense during this period if sufficiently superheated. 600° to 700° F. is the temperature to which steam should be superheated to get its fullest benefit. Engines must be built to stand this high temperature, or its use should not be attempted. For piping to convey superheated steam, copper is not suitable, as it loses about 40% of its strength at the high temperature. Wrought iron and steel with long lengths, and few flange joints, have proved to be the best. The expansion at 100° F. is about 4£ inches in 100 ft., and must be taken care of in the design of steam lines. Superheated steam can travel at 30 to 40% higher velocity through steam ports than saturated steam. lubrication of Engines Using* Superheated Steam, A 120 I.H.P. Engine uses 4 lbs. of oil per 24 hours for lubrication. A 300 I.H.P. Corliss Comp. Engine uses 2.2 lbs. of oil per 10 hours, both cylinders. Superheaters. Superheating is accomplished by passing the steam, immediately before use, through a series of pipes placed in the path of the furnace gases, or placed over a furnace of their own, where the steam can be given the higher temperature. The manufacture of separate superheaters in the United States is at pres- ent very limited, but abroad many types are in use, and are described in Dawson's Pocket Book. Economy of Different Types of Steani Engines Using: Superheated Steam. (W. W. Christie, in Railroad Gazette, March, 1903.) The various results given herewith should not be compared with each other on the basis of water per horse-power per hour, as pressures and other con- ditions are different, but the economy arising from the use of superheated steam over the use of saturated steam in the same engine can properly be compared by one percentage diagram. The following tests (A. S. M. E., Vol. xxi, p. 788) were made by Mr. E. H. Foster, on a Worthington duplex direct acting triple expansion pumping engine, having six cylinders arranged in tandems of three on each side. The engine was fitted with the Schwoerer patented superheater. * Compared with saturated steam. 1414 STEAM. Test No 1. 2. 3. 4. 5. I.H.P 106.3 0. 21.8 106.8 0. 21.2 103. 118.6 18.9 105. 122.5 18.5 105.1 Superheat, deg. F Steam per pump H.P. per hr., lbs. 117.7 18.0 The average economy as shown by the above tests in using steam super- heated 119.6° F. is 14.1 per cent over that of saturated steam. Perry, in the " Steam Engine," gives the results of several tests on a Cor- liss compound engine with steam jacketed cylinders when developing about 500 H.P. With saturated steam at 96 lbs. pressure the steam consumption was 19.8 lbs. per indicated horse-power per hour, but when the steam was superheated 118° F. the steam consumption dropped to 15.6 lbs., a gain of 20.8 per cent. Other tests on a single expansion engine equipped with a Schmidt superheater gave, when using saturated steam, an economy of 38 lbs. per I.H.P. per hour. When using steam with 300° superheat the steam consumption was 17 lbs., showing 55.3 per cent increase in favor of the lat- ter method. In a paper read before the Society of German Engineers in 1900, Oscar Hunger reported a test of a vertical cross compound pumping engine with 23.6 in. and 37.4 in. x 31.5 in. cylinders and running at 40 r.p.m. At 75 lbs. pressure the steam consumption was 20.5 lbs. with saturated steam. With steam superheated 180.5° and a pressure of 150 lbs., the steam consumption became 12.9 lbs., or a gain of 30.7 per cent over saturated steam at the lower pressure. Again, tests of a 3,000 H.P. vertical triple expansion engine at the Berlin electric light works (Engineering Record, vol. xlii, p. 345) show that a gain of 12.5, 17.9 and 18.7 per cent results from superheating the steam 181, 235 and 264° F. respectively. Other tests in Bavaria, with a Sulzer compound engine (Engineering News, vol. xli, p. 213), give a gain of 16 per cent with steam superheated 114°, 18.5 per cent when superheated 121°, and 25.9 per cent when superheated 173° F. 100 150 200 250 300 JPERHEAT IN STEAM-DEGREES FAHR. Fig. 12. Economy of Superheated Steam. The accompanying diagram* has been obtained from the above tests by plotting the degrees F. of superheat as abscissae and the per cent of economy as ordinates. Inspection of this diagram shows that the greatest economy results in the use of superheated steam in simple engine, as might be ex- pected. On the other hand, marked economies are shown for compound and triple expansion engines, but the percentage of gain decreases as the num- ber of expansions increases. * W. W. Christie. CONDENSATION IN STEAM-PIPES. 1415 COXDEyS4TIO\ I\ STEIW.PIPES. (W. w. c.) No very satisfactory figures are found for the absolute condensation losses in steam pipes, most of reported tests being compared with hair felt. 0.012 lbs. per 24 hours per sq. ft. of pipe per degree Fahr., difference in temperature of steam and external air, which may be used in calculations, is based on the following : Sq. ft. Sur- face. Lbs. of Water. Difference in temperature Deg. F. u o Covering. Test by. in 24 per sq. ft. hrs. in 24 hrs. Bedle & Bauer. 4130 11315 2.74 262 .0104 Asbestos. Norris. 3892 9360 2.40 234 .0103 Asbestos. Brill. 308 .0105 Magnesia sect'l. Norton. 315 .0125 Magnesia. The last test by C. L. Norton (Trans. A. S.M. E., 1898) was made with the utmost care. Mr. Norton found that a pipe boxed in with charcoal 1 inch minimum thickness was 20 per cent better insulated than when magnesia was used, corroborating Mr. Reinhart's statements concerning his experi- ence using flue dust to insulate pipes. Aboard Ship. — The battleship "Shikishima" carries 25 Belleville boilers capable under full steam of developing 15,000 I.H.P. in the main engines besides working the auxiliaries, each boiler supplying steam for 150 I.H.P. When at anchor, one boiler under easy steam, i.e., evaporating from 9 lb. to 10 lbs. of water from and at 212° F., per pound of coal — was just able to work one 48 K.W. steam dynamo at about half power, together with one feed pump, and the air and circulating pumps connected with the auxiliary condenser, into which the dynamo engine exhausted, besides working a fire and bilge pump occasionally. The dynamo was about 160 ft. of pipe length away from the boiler, the total range of steam pipe length connected being 500-600 ft. Performing the first-mentioned service with only one boiler under steam, the coal burned varied from 3£ to 5 tons per day of 18 hours, for about 65 I.H.P., or about 7 lbs. per indicated horse-power "at the best to 10 lbs. at the worst, an average of 8 lbs. and over, which shows that more than half the fuel must have been expended in keeping the pipes warm. All pipes were well covered and below decks, and machinery in first-class condition. (London-Engr.) Heating* Pipes. — To determine the boiler H.P. necessary for heating, it maybe assumed that each sq. ft. of radiating surface will condense about 0.3 lbs. of steam per hour as a maximum when in active service ; thus 20,000 sq. ft. times 0.3=6000 lbs. of condensation, which divided by 30 gives 200 boiler horse-power. Condensed steam in which there is no oil may be returned to the boiler with the feed- water to be re-evaporated. 1416 STEAM. ODTFIOW OF STEAM FROM A GIVE5T OITIAL PRIIKKMIE INTO YARIOll I (HI I It PRESSIRES. (D. K. Clark.) Absolute Outside Velocity of Actual Ve- Weight Dis- Pressure in Pressure Ratio of Outflow at locity of charged per Sq. Boiler per per Sq. Expansion. Constant Outflow In. of Orifice Sq. lnoh. Inch. Density. Expanded. per Minute. Lbs. Lbs. .Ratio. Ft. per Sec. Ft. per Sec. Lbs. 75 74 1.012 227.5 230 16.68 75 72 1.037 386.7 401 28.35 75 70 1.063 490 521 35.93 75 65 1.136 660 749 48.38 75 61.62 1.198 736 876 53.97 75 60 1.219 765 933 56.12 75 50 1.434 873 1252 64. 75 45 1.575 890 1401 65.24 75 43.46, 58 % 1.624 890.6 1446.5 65.3 75 15 1.624 890.6 1446.5 65.3 75 1.624 890.6 1446.5 65.3 When, however, steam of varying initial pressure is discharged into the atmosphere — pressures of which the atmospheric pressure is not more than 58 per cent — the velocity of outflow at constant density, that is, sup- posing the initial density to be maintained, is given by the formula — V— 3.5953 yh, where P=: the velocity of outflow in feet per minute, as for steam of the initial density, h = the height in feet of a column of steam of the given absolute initial pressure of uniform density, the weight of which is equal to the pressure on the unit of base. The following table is calculated from this formula : OUTFLOW OF STEAM 1 Y TO THE ATMOSPHERE. (D. K. Clark.) Absolute • Initial Outside Ratio of Velocity of Actual Ve- Weight Dis- Pressure in Pressure Expansion Outflow at locity of Outflow, charged per Boiler in in Lbs. per in Constant Sq. Inch of Lbs. per Sq. Inch. Sq. Inch. Nozzle. Density. Expanded. Orifice per Min. Lbs. Lbs. Ratio. Ft. per Sec. Ft. per Sec. Lbs. 25.37 14.7 1.624 863 1401 22.81 30 14.7 1.624 867 1408 26.84 40 14.7 1.624 874 1419 35.18 45 14.7 1.624 877 1424 39.78 50 14.7 1.624 880 1429 44.06 60 14.7 1.624 885 1437 52.59 70 14.7 1.624 889 1444 61.07 75 14.7 1.624 891 1447 65.30 90 14.7 1.624 895 1454 77.94 100 14.7 1.624 898 1459 86.34 115 14.7 1.624 902 1466 98.76 135 14.7 1.624 906 1472 115.61 155 14.7 1.624 910 1478 132.21 165 14.7 1.624 912 1481 140.46 215 14.7 1.624 919 1493 181.58 STEAM PIPES. 1417 §TEAM PIPES. Rankine says the velocity of steam flow in pipes should not exceed 6000 feet per minute (100 feet per second). As increased size of pipe means in- creased loss by radiation, care should be taken that in order to decrease the velocity of flow, the losses by radiation do not become considerable. The quantity discharged per minute may be approximately found by Rankine's formula (" Steam Engine," p. 298), W =. 60 ap -j- 70 = 6 ap -f- 7, in which W=: weight in pounds, a == area of orifice in square inches, and p = absolute pressure. The results must be multiplied by k =: 0.93 for a short pipe, and by k = 0.63 for their openings as in a safety valve. Where steam flows into a pressure greater than two-thirds the pressure in the boiler, W~=z 1.9 ak^(p—d) d, in which d = difference in pressure in pounds per square inch between the two sides, and a,p, and k as above. Multiply the results by 2 to reduce to h.p. To determine the necessary dif- ference in pressure where a given h.p. is required to flow through a given opening, 2 T 4 HP 2 14 a*k' Flow of Steam Tlirougrh Pipes. (G. H. Babcock in " Steam.'*) The approximate weight of any fluid which will flow in a minute through any given pipe with a given head or pressure may be found by the formula Hp l -p 2 )d* ID( '(^) in which W= weight in pounds, d = diameter in inches, Z>= density or weight per cubic foot, p t = initial pressure, p 2 = pressure at the end of the pipe, and L = length in feet. The following table gives, approximately, the weight of steam per minute which will flow from various initial pressures, with one pound loss of pres- sure through straight smooth pipes, each having a length of 240 times its own diameter. For sizes below 6 inches, the flow is calculated from the actual areas of " standard " pipe of such nominal diameters. For h.p. multiply the figures in the table by two. For any other loss of pressure, multiply by the square root of the given loss. For any other length of pipe, divide 240 by the given length expressed in diameters, and multiply the figures in the table by the square root of this quotient, which will give the flow for 1 pound loss of pressure. Conversely dividing the given length by 240 will give the loss of pressure for the flow given in the table. Table of Flow of Steam Through Pipes. Initial Pres- Diameter of Pip e in Inches. Length of each == 240 Diameters. sure by Gauge. I 1 U 2 2£ 3 4 Lbs. per Sq. Inch. Weight of Stean l per Mi a. in Lbs., with 1 I ;b. Loss of Pressure. 1 1.16 2.07 5.7 10.27 15.45 25.38 46.85 10 1.44 2.57 7.1 12.72 19.15 31.45 58.05 20 1.70 3.02 8.3 14.94 22.49 36.94 68.20 30 1.91 3.40 9.4 16.84 25.35 41.63 76.84 40 2.10 3.74 10.3 18.51 27.87 45.77 84.49 50 2.27 4.04 11.2 20.01 30.13 49.48 91.34 60 2.43 4.32 11.9 21.38 32.19 52.87 97.60 70 2.57 4.58 12.6 22.65 34.10 56.00 103.37 80 2.71 4.82 13.3 23.82 35.87 58.91 108.74 90 2.83 5.04 13.9 24.92 37.52 61.62 113.74 100 2.95 5.25 14.5 25.96 39.07 64.18 118.47 120 3.16 5.63 15.5 27.85 41.93 68.87 127.12 150 3.45 6.14 17.0 30.37 45.72 75.09 138.61 1418 STEAM. Table of Flow of Steam Tlirou g"li Pipes. — Continued. Initial Pres- Diameter of Pipe in Inches. Length of Eachr= 240 Diameters. sure by Gauge. 5 6 8 10 12 15 18 Lbs. per Sq. Inch. Weight of Steam per Min. in Lbs ., with 1 Lb. Loss of Pressure. 1 77.3 115.9 211.4 341.1 502.4 804 1177 10 95.8 143.6 262.0 422.7 622.5 996 1458 20 112.6 168.7 307.8 496.5 731.3 1170 1713 30 126.9 190.1 346.8 559.5 824.1 1318 1930 40 139.5 209.0 381.3 615.3 906.0 1450 2122 50 150.8 226.0 412.2 665.0 979.5 1567 2294 60 161.1 241.5 440.5 710.6 1046.7 1675 2451 70 170.7 255.8 466.5 752.7 1108.5 1774 2596 80 179.5 269.0 490.7 791.7 1166.1 1866 2731 90 187.8 281.4 513.3 828.1 1219.8 1951 2856 100 195.6 293.1 534.6 862.6 1270.1 2032 2975 120 209.9 314.5 573.7 925.6 1363.3 2181 3193 150 228.8 343.0 625.5 1009.2 1486.5 2378 3481 The loss of head due to getting up the velocity, to the friction of the 3team entering the pipe and passing elbows and valves, will reduce the flow given in the table. The resistance at the opening and that at a globe valve are each about the same as that for a length of pipe equal to 114 diameters divided by a number represented by 1 + -j- • For the sizes of pipes given in the table these corresponding lengths are : 1 1* 20 25 34 2 2i 41 47 3 I 4 52 | 60 6 8 10 12 15 18 71 79 84 88 92 95 The resistance at an elbow is equal to § that of a globe valve. These equivalents — for opening, for elbows, and for valves — must be added in each instance to the actual length of pipe. Thus a 4-inch pipe, 120 diame- ters (40 feet) long, with a globe valve and three elbows, would be equivalent to 120 -f 60 + 60 4- (3 X 40) = 360 diameters long ; and 360 -f- 240 = 1£. It would therefore have 1£ lbs. loss of pressure at the flow given in the table, or deliver (1 -f- ViJ = .gi6), 81.6 per cent of the steam with the same (1 lb.) loss of pressure. Equation of Pipes (Steam). It is frequently desirable to know what number of one size of pipes will equal in capacity another given pipe for delivery of steam or water. At the same velocity of flow two pipes deliver as the squares of their internal diimeters, but the same head will not produce the same velocity in pipes of diiferent sizes or lengths, the difference being usually stated to vary as the square root of the fifth power of the diameter. The friction of a fluid within itself is very slight, and therefore the main resistance to flow is the friction upon the sides of the conduit. This extends to a limited distance, and is, of course, greater in proportion to the contents of a small pipe than «i,- a il rg S^ }• may be approximated in a given pipe bv a constant multi- plied by the diameter, or the ratio of flow found by dividing some power of tne diameter by the diameter increased by a constant. Careful compari- sons ota large number of experiments, by different investigators, has de- T 2 *?,?• fcne following as a close approximation to the relative flow in pipes or dilrerent sizes under similar conditions : w d * W 00 _ V d + 3.6 W being the weight of fluid delivered in a given time, and d being the internal diameter in inches. STEAM PIPES. 1419 The diameters of " standard " steam and gas pipe, however, vary from the nominal diameters, and in applying this rule it is necessary to take the true measurements, which are given in the following table : Table of Standard Sizes Steam and Cfax Pipes. m o H Diameter. Diameter. (0 Diameter. H H 1 <1> Inter- Exter- aT Inter- Exter- Inter- Exter- N nal. nal. N nal. nal. N nal. nal. w. m m 1 .27 .40 2h 2.47 2.87 9 8.94 9.62 1 .36 .54 3 3.07 3.5 10 10.02 10.75 .49 .67 3* 3.55 4 11 11 11.75 .62 .84 4 4.03 4.5 12 12 12.75 1 .82 1.05 4i 4.51 5 13 13.25 14 1 1.05 1.31 5 5.04 5.56 14 14.25 15 \l 1.38 1.66 6 6.06 6.62 15 15.43 16 1.61 1.90 7 7.02 7.62 16 16.4 17 2 2.07 2.37 8 7.98 8.62 17 17.32 18 The following table gives the number of pipes of one size required to equal in delivery other larger pipes of the same length and under the same conditions. The upper portion above the diagonal line of blanks pertains to " standard " steam and gas pipes, while the lower portion is for pipe of the actual internal diameters given. The figures given in the table opposite the intersection of any two sizes is the number of the smaller-sized pipes required to equal one of the larger. 2 6- ° 4 o i Ul 14- 13- 12- 11- 10- 9- 8- CO U |7- £ 6- a 15- Id H w 4- O u. (- 5 S l- UJ h DIAGRAM GIVING DIAMETER OF STEAM AND EXHAUST PIPES FOR ENGINE CYLINDERS FROM 5 TO 40 INCHES CHAMETER, ^L^lvVtOtX AT PISTON SPEEDS UP TO 1,000 X \\\\\j\ FEET PER MINUTE FROM "POWER" 2- S I'l-i'I'I'' 40 T, n'l'i'i'iM'i' i'HM ' i ' i'i'i'.i'i'i ' i'iM 'i 'i ii 25 2b 15 lb 5 Fig. 13. 1420 STEAM. - I © OOlO C t-CDOOC O '-l rH O i-; UC CN (N C CO ri if 00 N OO^CNCNi-ir-i r-i ^h r-< CN CN CN CO O I>CSt-iCN O « X N C r 50 H ffiMl>000(N«NCN CM Lo 00 CI CD t 5 1H lO «j - iOt* CO lOCN i^OO^COCNr-lr-l ri ii rt N CN M CO ^ t "eia »Hi— lCNCNC0Tj00C5rHi— (i— irtHrtrt ^0005^000©'*. <» «J «5 W W 05 CN lO © © CO i-J rn O TJJ CO CO r-; CN NC^O^COHrtNuriwdt^lOCOCOCNrii-jHH t-< rH CN tj< CD © *^ C5rfCN ^t 1 CN H »— ' t-< i-H l> CO .« ^ . ^^_, <-i O CO CO CO CO ■* ■* t> t- © 00 Tt< CO »-l ^^^g^g^qco/^oowooo^cjocNt^-^i-i rHW^oooqqco 1 lOt^ODOCNrHod»^'c4cO^COTj5cOCNCNi-iT-iTH' i-ii-iiicNldodi-H© 1 ^ OOCOi-t COi*CNrt ,-li-l ia _, m pn rrt -- ^ t-t^TtiTfCOCOO'* 00 t^ C5 00 CN 00 t~ Sc^t2Scoco l ^ ,> :^ a ? c :' : ^ r 1 c ^' ;,H . °5 1 °1 *-«. « iq q oo oo eo ca ** COCNO^CNJiilOiddrHt^iQ^CNCNi-ii-ii-i r* tH r^ CN CO id © CO © t>- CO i— 1 I>COHt-( riH o ^ k.,p,«.ai^a.i COlCNWOOCON^ r-l CN © CN CN CN •* ^^S^oo^'^ ^ 1 ^ 1 ^^^ 00 ^ ^ 1 ^^ cN^qqicq-HCOcqiO OC0CNC0^THTH06ldC5«0"^C0"Cl>C0O CNOOCOCOlOOOCNCN g^g^^oqqqqcqt-t^qqco cN^t^qcooq^aiqoo co ©i-iCWrHC^dcOMC^lQCOCNr-ir-ii-? r^rHi^CNCNCO^OOCOdoO »-< rfCNi-H ©lOCNrH HCNCN 1 s. •l m ■ H Of: ■ ! 5' a- n «!d I S8S^»t^co; CN CO lO 00 ©' t> t-' © © i-l CN T* lO t- l- CO i-i lO CO ©© ooo OiOCOO CJOcoOoouOH^Hcq © oo © co cot^i-HqiiooiqcNi-jO TjicNi-HT-I T-iiicNCNCOCO^iOCOOo" 00 t>-rH«^0 CN CN CO CO l6 rH CN CO lO l> CN C5 00 C5 CO t- CN ICCN t>00": 5 C OC0 05H NONHCOiONiCOlfi JCOrfO^ CO 00 CO © lO CO CN rH CN CO CO CO CO CN t-( r-i rH CN CN CO "<* O © t- GO © t © \G © CO O r 00 lO 00 r : COC5COOOlTt* r -' r rTcOiH rHCN COr*©00©CN 1O00 rH CNCOCOrfb-OOCOCOOi ' "a ' OIOC5CO nOlOCO ,_rHrHCNCO rH rH rH rHCN C ag$g O"* C coco c ^•^wNOoqqcNHOi ^^CDOOrHldoOCOOOTfci H rH iH CN OCOlC"^lC r-ir-(CNC0'*lOCOt>00"" r ^ cs, ' ^^ ©C0Ti<3)© COO®C5COCCU5QHCMiacO{OCH»OCOCO^T(*Tj< CNCO L^t-(I>. CO©0OCO"H*TjH©-Ht^00O0COU0"rf< t- © OO CN 00i-"00lO^CNlO00^©-CNt^00t-t-OiTj• 1ft 00 lft 00©< )05 1*ONi1t>.HHiH ...5tHCNt+(©CD»-iCO'HHO •rHrHrHCNCNCO"<*lft©t-00rHL-i- l ©COCC i-( tH CO -* t> © 8©IQ 00CDCN©CNt^00©CO00'<^CNi-(CNTt<<: CNCN©CO00©'*OlCN'ti00-HHC0Tti©©00CNa5C04#TH © lO CN ^00 © t- t- CO i-J p CO CO c^ -* go LC © CI t- CO CU- CN UO © CN lO wt-^Tt rtH «^^^^cococcoccocociincowcO'*H^eo CN 5 TH,HCNC0T* b8 o S X 9 X X o © A +a £ A © 9 H +a ,£ S s © 'd "H Jod q+SaoT; ui ^aa^ •ssot; jo oi^a •inoH -iad ana ^ooj; jad snufi ui sso^; '* t~CO O CO ■.raoH izd una ^ooj .iad s^nifi ui ssoq; •^so^j *d 'H J9( * n^^ua^; ui ^aa,j •ssoq; jo oii^a •anon jad una 400,3; .iad s 4 iufi ui sso r x ■}sot 'cl "H J9( * u^ua^; in ^aa,j •sso r i jo oi 4 ^a '.moH .iad una 400£ jad s 4 iuxi ui ssoq; 00 COCO C^© CO cn i-H cn t- ■<* co t-O^QOOS t- © © © 00 soococo-** 5 t>- o CO LO 5^^HOO i-H NO^N^. Tt5 t-* th to i-3 co CM 00i-h COr*< CO ©(NTjfHJfttOrJ* CO 00 00 O •* 00 CN i-H CN t* t- i-H Tt< 850000HNO T)<«lHHOO CO Oi CM C5 b- i-H "* •^soq- * t^ oq -^ oq © © \6 co oo o> i-( ©GOCOTt*Tt»r-lC "<* O CO ri Ci C CO CN -^ Oi^COCJHL^LOMOMONOOMOO^NOQt-C^^ H00Wt>N©O5l>Tj^O5COt> lCC5COCOCCTtiCOI>.^tirt«rtrn«OCDt>;CCCCOCCO^CO^CNrHrHrH rH CN rH tH rH lONNNTtiH COCOr-n.OCJTt-rHCOI>.LOCO COt-lOCOCSCOlOCiCNrH^ ^C5iftHO5t-iOC0rtO5t ©b-lCtfiCOtNCOCNrHrHr ri t*< CO OS CN IC CO t> ^ CO CO Oi "* •* CO tK CO CO rt< CO © rH t- t£ t- CN NNO^MC5«0C5t-Cfl>NNH00NQ0M!NOl''U'5«OO«N55 OHTH^MTj*?ci>qt>^OH?o«oqMOooiooMi;i>HT(|o5 *THrHCNC>icOCOTj<»OCO(»OTHeO^COI>o6dMlc5t^ t> •«* rH rj< CO CN CO CO CO r* CO b- r-( COCO © CO CO t- IQ lO CO t> CO »A0050fOOC>WWOOOOOOMOQQO«HC00riNC50500N Ot^CO^OCOr^COL^pCOlO 5COi>©'cNcorHTf<-* »rHCOlCCCCO©iOCN rH rH rH rl rHlO (Nr-fCO^lCCOCNrtHCO CO CO t> CO lO CN "«1? ^^MioooMH<»^^qiqcqco>*OTj|t>.t>.qqqNO"*rHHCO "rH^CN^CO^CNlOcS^HHlOCOeN©CO^COCOrH^^©CN riHHINM^lO^CStHiNiCNOiaHOOlO rHrHrHrHCNCNCOCO^< •sseu^oiqx CO CO rH CJ CO rj< lO ^ -H< t> CO t- CO © rH CN r^ CO lO lO lO 1TJ lO ifl lO lO lO ©©©rHrHrHrHrHrHCNCNCOOICNCNCNCOCOCOCOCOCOCOCOCOCOCOCOCO w a a •*£ S*f CO Tt< CO rH I> CO t- CO CO CO t>CO©CNCNTtlCOrHCOCOCOTticOT*ldt^©rHCO H HrtHrtHrtH(NC5 t^ u- cq co © co cq © co co o _ ic e io cq cq cq cq t~ t- "rHrHrHrHCNI^COrlHT^iOlOCO^COcJdcNCO^lOCo'c^ rHrHi-HrHrlrirHCNCNCN ©^ i j«?< ^^10W^00050H(NW*10 STEAM PIPE. 145 5 % v- © ©£fc — ©«<» &q«2 «OC500rlOOOCSOC OCCOCSt-lO^CCdT CO uO t- t- t- rt< tH CO CO tH lO © t>> :>• CCi-«©CO''*'COCOC CO S « CO O H t- C CC C C »T. Tl- o « ■■ s*^ u« ~^r wj >^' 1 UO CO Tf iH © icOOOHCOltj OS © CO tJH O CO T* liO CN rtC ^ «g CN ^(NCOOCOCOrHCO^^OlOCSCOTlJ i-J c4 c4 n* cc © cm io ->* t}5 ^TttCOCOC-C CD © CO t>- CO OS © CO © ©" O i-t IC CO ^SSriS^nJ^S^SSSfc^SS Hr-llMC^CO'1C50© CN L.0 co qc] co 10 «oi>ci 832 o»«oeor-t 1 ^ ocn HffiOHt O lO OiTjH C Tt< T*< ■<*! UO 11 si 2 •*« c3 eg © d cr"^ r 7 ■m CS O ljO lO CO ^ C DrtC^t'MOt-?! ^Tf coc4oic4r 5o jo^ococ 5 00 t- CO *Q 5 OOO t- tDli 0O* CO CO CO 03 l>C 5«0^r(l0t-ON»NI»(NOOt»t»5TiCO^i-!C*''#COCOOOOTtGOO>rH(N»daJo6cO^COodlOCOOJcO«OT^«0 iHi-li-(T-lOJCOiO»t-0>i--©rHiHCNCNC0^3lO©t>CiCNL~^^OCft©CNC30^©t>10TjCO"'*C3COl.-~ >^^^.C5t>.iqi-ji>oqc»i-jc^^io»oqc^iH flCO^^ld^^t^oda*©"©"^CN^"iOCO"i-*io"cO*i-*T*t>"©CO £ iHiHrHTHi-ii-ii-iCOiOCC>© CJ^^rHt^^C^lt^CO ■a DO 2c^i>o»»oiqioioio»q^^^Tjj^coco h i-i r-I i-* cn cn cn cn co* co* co* co* ■<*" -<** id co t> co* ci © ih cn* co" rt* ir* cd i>*- go os © 5 « 5 *-• j-J i-i "rt^'^f CN CNcfc^CO Co'co'eO'* *^iO ?0t^Q00iOi-iC5C0^i0«Dl>0005Qi-i 73 1-IHrlHHHrtHHrtClW BOILER-TUBES. 1429 Collapsing* Pressure. Bessemer Steel Tubes, Lap Welded. A. S. M. E. Trans. 1906— R. T. Stewart. P= 1000 (l—^/l--1600 f. J • „ (A) P = 86670 l - — 1386 . (B) P=z collapsing pres. lbs. per sq. in. d z=z outs. diam. of tube — inches. t = thickness of wall — inches. Use A for values of P less than 581 lbs. for values of — less than 0.023. a Use B for values greater than these. Material tested was 56000 — 60000 lbs. tensile strength. Up to 8" diam. and 20 ft. long. Resistance of Tubes to Collapse. Bulletin, No. 5, Exp. Station — Univ. 111., 1906 — A. P. Carman Where ratio -z is greater than 0.03. a. For brass : P = 93365 4 — 2474. a b. For seamless cold drawn steel : c. For lap-welded steel : Pzr'9552o4 — 2090. a P ='83270 ~ -1025. a "Where -z is less than 0.06. a For seamless cold drawn steel : p= 1,000,000 ft\ For lap-welded steel : /»= 1,250,000 f^Y 1430 STEAM. OiflfNMNOHHMi 00© t- 1- c© co co hCO COCO t-iCO riO HrtH ^MNWO^HrilOf lOb- COCOlO COCO NOH!Nt-»Hrt CO lO CO lO "^ Ttn CO CO CO Hf»*t) t >*-W H»>0<»Hx^H' 03 H"'w e, i«^H.« o Hf.fx 95 M' n H' > H ONr-C h "mc-|oo t^o iox«*o^t*H-Hi-i M H-+f ,0 !-*4xHNH^ n M«M MNHN -HOOOS 00OS CSCS00 N(S c<*0 icto f)» r 4nHfa( n H^H a W«BH»^ N H W M **w CO CO rH CO © 00 i-tt-00 t-00 00 00 b- COCO «*r H* Mfcfi1®-WHH' e, H W M' Mr*» rH M«H t ' 5 i-H i-( *• O <* IQ ""f ^ CO i-lCO -*»-*» ^ti® «l^tiXl«tX10|l» W 'rH VHMlH^CBCClx' > H-H" t >l ( V' a H« CO t- fH i-i lO •>* CO* CO* * * CO i-i r co * CO-* co co co hi- MMMWMMMMMMMMMMMMMMWMM O O ST 1 O 4 *■• h ca o ^ 0) CO.© ^ < BB^^B I I I I h o W W©| w o2 o ° fl' ON.5 *3 -82 -i O 'c^ o 5 S-. © o u, J g § g s 5Q«h.2Q 3 5 © 5 G - ^ ,d ■^ S c« £ bJD p5^ o-E •^Ofej^^ ri 0) © flQfl I I tc , -3 ^ a? c3 c3 Ah ® S -2 y cp^3 is CD o •S3 cd'£ PIPE BENDS. 1431 Tensile Strain of Bolts. Diameter Area at At 7,000 At 10,000 At 12,000 At 15,000 At 20,000 of Bolt bottom of lbs. per sq. lbs. per sq. lbs. per sq. lbs. per sq. lbs. pel in inches. Thread. inch. inch. inch. inch. sq. inch.. | .125 875 1,250 1,500 1,875 2,500 | .196 1,372 1,960 2,350 2,940 3,920 1 .3 2,100 3,000 3,600 4,500 6,000 i .42 2,940 4,200 5,040 6,300 8,400 .55 3,850 5,500 6,600 8,250 11,000 ii .69 4,830 6,900 8,280 10,350 13,800 i^ .78 5,460 7,800 9,300 11,700 15,600 if 1.06 7,420 10,600 12,720 15,900 21,200 li • 1.28 8,960 12,800 15,360 19,200 25,600 H 1.53 10,710 15,300 18,360 22,950 30,600 1.76 12,320 17,600 21,120 26,400 35,200 i$ 2.03 14,210 20,300 24,360 30,450 40,600 2 2.3 16,100 23,000 27,600 34,500 46,000 h 3.12 21,840 31,200 37,440 46,800 62,400 2* 3.7 25,900 37,000 44,400 55,500 74,000 The breaking strength of good American bolt iron is usually taken at 50,000 lbs. per sq. in., with an elongation of 15 per cent before breaking. It should not set under a strain of less than 25,000 lbs. The proof strain is 20,000 lbs. per sq. in., and beyond this amount iron should never be strained in practice. "MM3 BE.\DS, Made from Wrought Iron or Steel Pipe* (Crane Co.) Fig. 17. The radius of any bend should not be less than 5 diameters of the pipe, and a larger radius is much preferable. The length X of straight pipe at each end of bend should be not less than as follows: 5-inch Pipe X— 6 inches, 6-inch Pipe X— 7 inches, 7-inch Pipe X= 8 inches, 8-inch Pipe Xz= 9 inches, 10-inch Pipe X= 12 inches, 12-inch Pipe X= 14 inches, 14-inch Pipe X= 16 inches. 1432 STEAM. S S § tF<©0000 «£>tHi-HCO «C> 10 CN CO CO CQtNCN , «W M |» ojWus!® Ml* «|aofi®i^x«1co,Hh*' > M"|r«:|ao a V, u Vi, O^ONrtMHHIN ,0 M «o to © oo oothi-ioo oo i> cTco ajHsoHcctHoo ! H b ViHetF+)' e *M n HCAJR.D PIPE JFJLAUGKES. A. S. M. E. and Master Steam and Hot Water Fitters' Association stan- dard, adopted August, 1894. Medium pressure includes pressures ranging below 75 pounds. High pressure ranges up to 200 pounds per square inch. C|§ 65 X © © ,3 X © 1 +3 © © h- 1 © © 1? © © © 02 © 00 9 O a H O N en © & s a? ^ si h QQ © ■ 2 o c © 03 X © © G M © 5 © X o3 CO U © © s 5 © a «3 m DO © © H © be o3 © © bo E 02 ^ O © £ 03 5 a g oS OM X '© « «H O U © g h- 1 © © o3 s 3 © "S) s © 0«5 o3 e-d fi ©2 © CfH ^Oi © 5 s^ H OQ A^ 5 E cq fc « m ol».ci® •(•ui) ssauspiqx •(•ui) q*PL& i-^^MmIN!** •(•ui) 9SEa-qns jo dox JO q^pjAY i* S S 3 £ 2> 1 ( ff I •(•ui) ^ •(•Ui) S808IJ uoisuo^xg; jo i^Sua^; '3 i£> lO <£> O CO © 73 -+3 hj £ ./ o ^ •(•ui) •(•ui) y SS13X0 Suot; "SS J-i ^ la 51 CO •(•ui) sauiSug; 3[u^jo opis SCO^^lOlO ' P IO 19 O +++++++ ■^ IT5 sD t^ 00 O i •(•ui) souiSug; ^U^JQ jo^uoo c3 £ •(•ui) souiSug; 3(u«jo opis >*l06t*00OH •(•u T ) sauiSug; •e^nuiK jod suopnjoAOH '9^12 AUqi}! '^xufi jo A^ib-ed^Q ■«*! "<* T*< lO t> 00 CO CC 55 Ol £^ ? Os 1 OS OS OS 8 OS CO CM os IO CM OS OS OS CO CM OS OS OS OS CM OS CM OS CO OS OS OS CM CX w^qCD fe OS OS o OS OS co OS OS OS OS 10 OS IO OS CO OS CO OS b- cs OS OS (M OS OS OS H«q lO CM CO 00 1 5? CO bo 8 CO t- s 1 CO CO CM CO CO 8 CO CO 8 CO CO b- CO CO 8 CO 8 CO lO cm CO CO o eo CO OS CM CO b- CO OS CO OS CO & ^ b- OS 8 1 8 CO co CM CO 2? o IO °2 L0 CO LO CM CO cS b- CM co b- b- CO 8 CO CO os os CO 3 co CM co c^ co CM CO S t- R 1 3 CM "hOM CO CO CO CO CO CO CO CO b- CO CO ^. i a CD CO OS CO 8 1 I CO CO OS H*gS - 1 lo 1 OS IO CO CM CO CO CM CO •i-H 8 CO 8 2 CO %' 8 OS 8 1 CO s 10 CO 5 3 b- lO lo CO °2 lo OS 8 10 1 8 CO 00 CO eo 8 3 s 1 OS 8 **c3e* o CM CM L0 § 8 LO i CM L0 lo OS IO 10 S 10 CM b- »q CO b- uq CO OS 10 s 10 § i CO IO CO *8S CM OS © lo 1 b- iq 10 CM L0 CM 10 CO 2 CO 1 8 10 CO CO IO CM IO CO lO ->* os *ss co 8 IO s ^ OS lo O 10 b- 10 1 CO iq CO CO u q_ 1 IO 10 OS 8 to iq L0 IO CO CM 3 1 8 8 OS CO 9 5 IO CM 10 CO lO 10 b- 10 co 10 10 OS s t- i 8 CM § CO 1 OS b- t- ? IO 3 3 10 LO b- CM iO CO iq 00 CO iq 1 LO •Cgg *> *- CM 3 CO § 5 1 1 CO CO CO 3 10 1 CO lO CO CM LO OS CM »o Hccgg CO CO CO OS 55 OS s »# CO CO CO b- 1 OS iq LO CO iq LO CO 8 CO CM CO CO 2 CO 1 CO CO CO 9 i 8 OS CO" OS O LO 1 OS LO % OS b- eo # i i eo' ^3 i 1 CM 3 8 8 CO b- CO CO O cs «£g£ 3 i eo % CO OS eq CM CO 1 CM 1— ( CM CM eo 1 OS 3 00 9 CO CO 1 CM 3 OS 1 HSg- o 8 ^ 3 § 8 CO CO CO CM CO 1 eo eo 1 1 3 % 1 s 3 +3 5tt • Appar Ratio Expand o rH CM CO "* 10 CO t- CO OS rH CM eo -* »o j STEAM ENGINES. 1443 Receiver Capacity. — In compound engines with cranks at right angles the receiver capacity should be from 1 to 1.5 times that of the high-pressure cylinder (Seaton), or not less than the capacity of the low- pressure cylinder ("Practical Engineer")- When the cranks are oppo- site, the receiver capacity need not exceed that of the steam passage from the high-pressure to the low-pressure cylinder. The general effect of large receiver capacity is to cause a drop between the pressure at the end of the high-pressure expansion stroke and the beginning of the high-pressure ex- haust stroke and low-pressure admission, thus increasing the power devel- oped in the high-pressure, and decreasing the power developed in the low- pressure cylinder ; this leads to a loss of power in the engine, and one which — at any rate in engines with cranks at right angles — is greater the more the receiver capacity exceeds that necessary for free passage of the steam. Steam Ports and Passages. — The areas of these should be such that the mean linear velocity of the steam does not exceed 5,000 to 6,000 feet per minute ; hence, if D = diameter of cylinder in inches, A — area of cylinder in square inches, a = area of port or passage in square inches, S = piston-speed in feet per minute ; __ AS _ D 2 S ' a ~" 6,000 ~" 7,639 for mean velocity of steam 6,000 feet per minute ; _ AS _ &*S a ~ 5,000 — 6,366 for mean velocity of steam 5,000 feet per minute. The lengths of the steam passages between the cylinders and valves should be as small as possible, in order to minimize clearance and resist- ance to flow of steam. Condensers and Pumps. Condensers are principally of two types, viz., Jet Condensers, in which the steam and condensing water mix in a common vessel, from which both are pumped by the air-pump ; and Surface Condensers, in which the steam generally passes into a chamber containing a number of brass tubes, through which the condensing water is made to circulate. The latter form is usually adopted where water is bad, as it enables the same feed-water to be passed through the boiler over and over again. The capacity of a jet condenser should not be less than one-fourth of the low-pressure cylinder, but need not exceed one-half, unless the engines are very quick running ; one-third is a good average ratio. Large condensers require more time for forming the vacuum, while small condensers are liable to flood and overflow back to the cylinders. The amount of condens- ing water required per pound of steam condensed varies with the tempera- ture of the exhaust, of the " hot-well," and of the condensing water. (The 44 hot-well" is the receptacle into which the air-pump delivers the water from the condenser.) The feed-water is obtained from the "hot-well," which should be maintained at 110° to 120° F. Sometimes even 130° F. can be obtained with care. The amount of cooling or tube surface depends upon the difference be- tween the temperature of the exhaust steam and the average temperature of the cooling water, and on the thermal conductivity and thickness of the metal tubes. For copper and brass tubes in good condition the rate of transmission is about 1,000 units (equivalent to about 1 lb. of steam con- densed) per square foot per 1° F. difference of temperature per hour. With the hot- well at 110° and the cooling water at 60°, the average difference is 25°, and 25 lbs. of steam should be condensed per hour per square foot. In practice allowance must be made for the working conditions of the tubes, and half the above, i.e., £lb. of steam per 1° F. difference is nearer the usual allowance ; and under the above conditions about 12.5 lbs. of steam would be condensed per square foot per hour, which is considered very fair work. The tubes are generally of brass, No. 18 S.W.G. thick, and from £ to 1 in. diameter, according to the length of th= diameter of pump in inches ; C=— , D= 13.55V ~S' n* ▼ nS Circulating pump valves should be of sufficient area so that the mean velo- city of flow does not exceed 3 or 4 feet per sec. High velocities tend to wear out the valves, and cause undue resistance in the pump. In the suc- tion and delivery pipes the velocity should not exceed 500 feet per minute, or for large and easy leads 600 feet per minute. Better results, however, will be obtained by using larger pipes, so as to reduce the velocity, espe- cially if the pipes are long. For single-acting pumps the suction may be smaller than the delivery, if the pump be below the water-level. If a = minimum area through valves in square inches, d z= minimum diameter of pipe in inches, A = area of pump in square inches, D =z diameter of pump in inches, S ss mean speed (useful) of pump in feet per minute ; AS . D^S a= 180' d = -K> where K varies from 22 for small pumps to 25 for large pumps, while for the suction of single-acting pumps it may be 27. Air chambers should always be fitted, which for single-acting pumps may be twice the capacity of the pump. An air-pipe should be fitted to the CONDENSERS. 1447 highest points of the water passages for escape of air to enable the con- denser and pipes to run full. If the speed of the circulating pump cannot be varied independently, it is advisable to fit a water valve between the two ends of the pump, so that the discharge may be varied to suit the requirements. Strainers should be fitted to the inlet of the suction pipe; and the aggre- gate area of the passages should be from two to four times the area of the pipe, according to the velocity of flow in the pipe. Owing to difficulty experienced in cleaning strainers when under water, they are sometimes fixed in a cast-iron vessel near the suction entrances to the pump, with a door arranged in some convenient position for cleaning. Foot Valve.— When the water level is below that of the pump, a foot valve should be fitted just above the surface of the water. A door should be provided for examining the valve without disturbing the suction pipe, or an air ejector may be used to charge the pump. COOLLY^ IOWER TE*T. On August 2, 1898, during a run from 7 a.m. till 12 midnight, from the daily records, the following data is reported by Vail, A.S.M.E. Trans. Vol. 20. Maximum. Minimum. Temperature, atmosphere 103° 83° Temperature, condenser discharge to tower • . * . 128° 106° Temperature, condenser suction 98° 91° Degrees of heat extracted, through tower . . .32° 21° Speed of fans, revolutions per minute ...... 160 140 Vacuum at condenser 26 20 Strokes of condenser pump 50 38 Pounds, boiler feed 121 100 Temperature, boiler feed 212° 200° Engine, horse-power developed 900 H.P. 400 H J*. A continuous heavy load was carried during the entire 17 hours' run. This was not a test record, but simply daily service. Another day, November 5, 1898, from a 20 and 36 X 42 tandem compound condensing Corliss engine, the conditions were as follows : Engine revolutions 120 per min. Steam pressure 112 Vacuum at condenser 25 The area of the cards shows the work done in highpres* sure cylinder to be 311.8 H.P. And in low-pressure cylinder to be 331.5 H.P. Total 643.3 H.P. "Work done in low-pressure cylinder below atmospheric line 185.1 horse* power. Simultaneously with the engine, the pump and fan engines were indicated. Tower used was Barnard Type of Cooling Tower. The work done by the pump 13.75 H.F The work done by the fan engines 13.5 H.P Total external work 27.25 H.F, 23.6 1.H.P. of Engine per I.H.P. of Pump and Fans. 1448 GAS. GAS EtfGKOTES.* Nearly all commercially successful gas engines are those in which the cycle of operation is that proposed and patented by M. Bean de Rochas, in France in 1862. He states as necessary to economy with an explosion engine four condi- tions : 1. The greatest possible cylinder volume with the least possible cooling surface. 2. The greatest possible rapidity of expansion, or piston speed. 3. The greatest possible expansion : and 4. The greatest possible pressure at the commencement of the expansion. From the above Bean de Rochas reasoned these operations : a. Suction during an entire outstroke of the piston. b. Compression during the following instroke. c. Ignition at the dead point and expansion during the third stroke. d. Forcing out of the burned gases from the cylinder on the fourth and last return stroke. He proposed to accomplish ignition by increase of temperature due to compression. The otto engine uses the above cycle and flame ignition. Classification. Gas engines may be classified in accordance with the principles of the cycle of operations: • 1. Explosion of gases without compression. 2. Explosion of gases with compression. 3. Combustion of gases with compression. 4. Atmospheric motors. According to the gas used they may be classified thus : — A. Coal gas. B. Carburetted gas. C. Producer or Dowson gas. The methods of igniting the charge are /. Electrical arc. g. Flame. k. Incandescence. m. Chemical or catalytic action. The Otto engine is a good example of flame ignition. Diameter of gas main from meter to engine should be dia= .027 Brake H.P. + 0.79 inches. Atmospheric air is the working fluid of all gas engines and the fuel which heats it is inflammable gas. The air and gas are mixed thoroughly before passing into the cylinder itself. ( More wasteful of fuel than four-cycle engine. Back- „, , . firing, or premature explosion of gas and air mix- i wo-cycie engine. < t ure# Used in large power units, with blast furnace i ' More readily governed than two cycle. No pumps. _ i • j No inclosed crank chambers. if our-cycle engine. «< Must ^ e Du jit heavy in comparison with power pro- duced. v Heavy flywheels. There is but little difference between gas and gasoline engines, the main difference being a special fitting to supply the oil in the form of a vapor or atomized sprav. . Gasoline being richer than gas, by its use a much larger H.P. can be ob- tained from a given size of engine. The theoretical efficiency of a gas engine is about three times greater than that of a steam engine. Contrary to steam engine experience, when underloaded it Is a compara- tively efficient heat engine. • W. W. Christie. GAS ENGINES. 1449 The highest recorded efficiency is the consumption of 8000 B.T.U.'s per Brake H.P., or a thermal efficiency of 31.75 per cent. Governing is not quite as easily accomplished under quickly varying loads, as in the steam engine, although late models leave little to be desired. In general, governing is accomplished by three methods : (1) the hit-and- miss, where the gas valve is closed during one or more revolutions of the engine; (2) by varying the mixture of air and gas in the cylinder, thereby producing explosions of greater or less pressure intensity ; (3) advancing or retarding the point of ignition. The average mixture is 1 part of gas to 8 to 12 parts of air in a gas engine. Gas engines can be run successfully and with a fair degree of economy to within 3 or 4 per cent of their normal rating. B. A. Thwaite says the " lean gases of low calorific power, such as are obtainable as a by-product of the manufacture of iron, are the very ones which enable the highest efficiency to be secured in internal-combustion engines." A gas rich in thermal units enables a larger power to be derived from a given engine than can be obtained by the use of a lean gas. Less air is required to mix with lean gas, and a higher compression is reached, for the mixture has a higher ignition point than rich gas mixtures. High compression conduces to high efficiency. Compression varies inversely as the calorific value of the gas, high for a lean gas, and vice-versa. For natural gas the compression displacement is made about 30 per cent of piston displacement. "Water for cylinder jacket should flow through at a rate of 4 to 5 gallons per H.P. per hour ; best conditions are when jacket water removes 4000 B.T.U. per H.P. per hour. Best piston speed is about 600' per minute. Comparative Economy. Lbs. of Coal per Brake H.P. per Annum. Steam engine plant — simple non-condensing Steam engine plant — compound condensing Gas engine plant with producer gas . . . 11,250 6,400 3,050 Per Cent. Thermal efficiency simple non-condensing plant 5.5 Thermal efficiency compound condensing plant 9.7 Thermal efficiency gas engine plant using producer gas . . 20.3 Thermal efficiency gas engine plant using waste blast fur- nace gas 23.5 The standard gas is the natural gas of western Pennsylvania, whose calo- rific value is about 1000 B.T.U. 's per cubic foot. Ordinary illuminating gas has 750 B.T.U's. per cubic foot. Producer gas may be as low as 120-130 B.T.U. 's per cubic foot. Consumption of gas or gasoline by engines is, conservatively: Natural gas 10-12 cu. ft. per Br. H.P. hour. Illuminating gas 18-20 cu. ft. per Br. H.P. hour. Commercial 74° gasoline . . £-% gallon per H.P. hour. Gas engines operate on, say,lj lbs. of good anthracite or bituminous coal, approximately, in some cases as low as 1 lb. anthracite or bituminous coal. Gas generated from wood in Riche's retort, according to James M. N($il, has a calorific power of 3029 calories per cubic meter, or : 340.8 B.T.U. per cubic foot ) . „. „ - ^ n4 . ar . „ na 324.5 B.T.U. per cubic foot { ls ^ lven for water S as ' 590.0 B.T.U. per cubic foot is given for coal gas. 1 ton of wood produces 25,000 cu. ft. of gas and 400 lbs. charcoal, and colts 14 cents per 1000 cu. ft. with wood at $3.00 a ton, neglecting in this calcu la- tion the charcoal. 1450 GAS. Mr. T. Fairly, Leeds, England, gives the heating power of coal gas corre- sponding to lighting powers as follows: no correction being made for the condensation of the steam produced by the combustion of hydrogen. Lighting power : — C.P. 11 12 13 14 15 16 17 18 B.T.TT. 533 555 578 601 624 648 678 704 Value of Coal Gas of Different Candle Powers for Motive Power. (C. Hunt.) Consumption Relative Value Relative Value Candle Power. Cubic Feet per for Motive for I.H.P. Power. Lighting. 11.96 30.31 1.000 1.000 15.00 24.41 1.241 1.254 17.20 22.70 1.335 1.438 22.85 17.73 1.700 1.910 26.00 16.26 1.864 2.173 29.14 15.00 2.020 2.436 Gas Engine Power Plant. Lackawanna Steel Co., Buifalo, N.Y., uses Blast Furnace Gases. 8-1000 H.P. Gas Engines in place, 1903. 16-2000 H.P. Gas Engines to go in later. Electric Generating plant consists of : 5-500 K.W. 3 phase, 25 cycle, 440 volt machines. (Gen. Elec. Co.) 4-500 K.W. 250 volt, direct current machines. (Sprague.) Eight of the above are direct connected to horizontal, duplex, 2 cycle, double-acting, Korting Gas Engines. One is direct connected to a 1000 H.P. Porter-Allen steam engine. Engines use the waste gas from the furnaces. By volume : CO, 24% ; C0 2 , 12% ; N, 60% ; H, 2% ; CH 4 , 2%. Calorific Power, 90 B.T.U.'s per cubic foot. The steam boilers in this plant are 250 H.P. Vertical Cahall Boilers ; 48 have Roney Stokers, others are gas fired. They each have a two-part cylindrical monitor on the roof of the boiler house, that is easily removed, enabling rapid and easy cleaning of tubes. m Power," Dec, 1903. Gas Engine Pumping* Plant Test. OTidvale, W.J". Triplex pump driven by a 5 H.P. gasoline engine, 7th trial. Discharge 153 gallons per minute. Lift, 65 ft. total. Used 5£ gal- lons of gasoline or 0.312 gallons per H.P. hour. Oreenstmrg*, Ind. Triplex pump driven by a 6 H.P. crude oil engine (Indianapolis, Ind., Eng. Co.), 9th trial. Discharge 184 gallons per minute, total lift 81.3 feet. Montpelier Crude Oil, 2 cents a gallon — 0.47 gallons per H.P. hour. (Eng. Rec. V. 38, 508.) Cost of Lifting- Water. With gas at 22£ cents per 1000 ft. One H.P. for 3000 hours, with a gas en- gine at,— Wilmerding, Pa $9.58 Pitcairn, Pa 10.99 E. Pittsburg, Pa 12.70 — J load on during test. (Eng. Rec. V. 38, 397.) The Heat Energy from burning gas is disposed of in the Otto gas engine as follows : STEAM TURBINES. 1451 Averages of Many Teats* 1. Actual work and friction 17 per cent. 2. Hot expelled gases 15£ per cent. 3. Water jacket 52 per cent. 4. Conduction and radiation 15$ per cent. Gas ^Engine Pumping* Plant. Pittsburg Plate Glass Co., Ford City, Pa., uses natural gas of 1000 B.T.U.'s, obtained on the premises. Each pumping unit of six units (5 now in — 1903) consists of : One 11" X 12" — 3 cyl. Westinghouse Vertical Gas Engine direct geared to a 16" x 15" single acting triplex pump, Stillwell-Bierce & Smith-Vaile Co. Compressed air is used to start the engines, being tanked in 3 steel storage tanks for this purpose. A 3 H.P. electric motor sup- plies this air at 180 lbs. pressure. Total head pumped against, 215 ft. Gallons per minute, 1101. Total cost per million gallons, $7.02. Steam plant doing same work cost $1,700 per month (average) for fuel alone. Gas method cost $180 per month for fuel alone. Full test and diagram of engine efficiency in " Power," Dec, 1903, p. 708. §TEAH TURBINE!.* Steam turbines, machines in which jets of steam striking vanes or buckets at a high velocity, are used as a motive power, may be classified thus : 1. Radial flow Outward. Inward. 2. Parallel or axial flow \ %£ s °™ ( De Laval. J Parsons j Rateau. L Curtis. 3. Mixed flow. If steam at a high pressure be allowed to escape through a suitably de- signed diverging nozzle into a lower pressure, a large proportion of its heat energy will be converted into kinetic energy, and the steam will expand adiabatically to the pressure of the medium or fluid into which it is discharged. There is a wide difference between steam turbines and water turbines, for the nozzle velocity of steam is, say, 2,000 feet per second against 96 feet for water. Then again, 1 cubic foot of water gives the same amount of kinetic energy as 1 cubic foot of steam at 50 lbs. pressure. The efficiencies of all types depend very largely upon the terminal press- ure at the exhaust end, and likewise on the completeness of the vacuum, where condensers are used ; which accords with reciprocating steam-engine practice. The absence of lubrication in the internal or steam spaces, permits the use of condensation and return of all water of condensation to the boilers. Both the above factors, as well as the use of superheated steam, assist in securing the high efficiencies already obtained with this motor. Experience shows that water carried over from the boiler does no harm in them. One point which is made in their favor, is, no boiler scale when the same feed water is used continuously. In that event, boilers may suffer even more seriously from corrosion from the water being too pure, unless raw water is added from time to time to neutralize the corrosive tendency. The steam turbine has opened up a field of usefulness all its own ; for ex- * W. W. Christie. 1452 STEAM. ample, the driving of centrifugal pumps, where in reciprocating engine prac- tice great efficiency was only obtained with low heads, turbine efficiency is maintained even at very great head. While used also to drive fans, prob- ably the greatest field ope'n to steam turbines is the driving of electric gener- ators, direct-connected or direct-geared. De Laval Steam Turbine. In this type the total power of the stream is devoted to the production of velocity in an expanding nozzle. The jet so produced is driven against a set of vanes on a single wheel, ingeniously supported and run at a very high peripheral speed, the lineal velocity of teeth in this type being about 100 feet per second, and gearing ratio 10 to 1. It is limited only by attending imperfections in gearing, and the type is not especially applicable to large sizes ; thev are not at present being built larger than 300 H.P. As now designed, there is no way to reverse this machine. imam Fig. 21. Tests of a De Laval Turbine by Dean and Main showed the saving by the use of superheated steam over that of saturated steam to be : Amount of Super- heat. Load Load Steam Dry Steam Saving by No. of with with used per used per use of Nozzles Super- Satu- Brake H.P. Brake H.P. Super- in Use. heated rated Avith Sup. with Sat. heated Steam. Steam. Steam. Steam. Steam. Deg. F. H.P. H.P. Lbs. Lbs. % Eight 84 352 333 13.94 15.17 8.8 Seven 64 298 285 14.35 15.56 8.4 Other tests by the same engineers gave with superheated steam : STEAM TUKBINES. 1453 De l^aval Turbine. Number of Nozzles Open, Eight (8). Average Reading of Barometer, 30.18 in. Average Temperature of Room, 83° F. H O w Steam used per Hour. (Lbs.) Pressure above Governor Valve. (Lbs.) Pressure below Governor Valve. (Lbs.) q M a O > GO'S Sh 35 <© S3 ao j-i . ® I* si o w Steam used per Brake Horse Power per Hour. (Lbs.) A.M. May 22 8-9 4,833 208.3 200.6 27.2 81° F. 356.6 13.55 (i 9-10 4,936 207.5 199.3 27.2 86° F. 355.7 13.88 " 10-11 5,083 207.7 202.1 27.2 91° F. 357.8 14.21 it 11-12 4,976 208.3 199.4 27.2 88° F. 354.1 14.05 « M. P.M. »« 12-1 4,841 207.5 194.3 27.3 82° F. 343.5 14.09 1-2 4,768 206.9 195.6 27.2 75° F. 344.4 13.84 Governing is accomplished by regulating the steam pressure at admission in much the same way as in reciprocating steam engines. The Parsons Steam Turbine. In the Parsons turbine the steam, after leaving the governor valve, enters a steam passage and turns to the right, first passing a stationary set of blades, then the blades of a revolving cylinder ; this operation is repeated a number of times, the steam moving in an axial direction until it has reached the FIXED MOVING. FIXED f MOVING FIXED ^MOVING Fig. 22. Vanes, Westinghouse-Parsons Turbine. other end of the turbine, when it is exhausted, sometimes at as low a tem- perature as 126° F. The steam velocity is not as great in this type as it is in the De Laval. Fig. 23 shows the relative floor space occupied by Westinghouse-Parsons* turbines, vertical and horizontal steam engines. 1454 STEAM. HORIZONTAL CORLISS ■ ' VERTICAL CORLISS Hi STEAM TURBINE 1 it I H m u CD a. (H § 0) aj n QQ Ch o © E o o fe H +3 i— • ed © a fr u © o H © CQ Fig. 23. 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 9 \ V \ Horizonti CorliBB \ V«rtioal \C CorlisB v^ G Steam ( Turbine 1,000 2,000 3,000 4,000 5,000 Electrical Horse Power Fig. 24. STEAM TURBINES. 1455 Type. Rows of Rotating Bucket Rings. Steam Velo- city, Feet per Second. Revolu- tions per Minute. Peripheral Speed, Feet per Second. Buckets. Parsons . . Rateau . . Curtis . . . De Laval 35 25 8 1 400 800 2,000 4,000 3,600 2,400 1,800 20,000 200 400 400 1,200 Inserted Inserted Solid Inserted (" Dodge.") This type can be built so as to reverse by interchanging the steam and ex- haust pipe connections. The efficiency, however, is somewhat reduced re- versing. Governing is accomplished by regulating the steam at inlet as in other types of engines. A 400 K.W. Turbine gave a steam consumption of 14.47 full load to 16 lbs. at half rating, 19 lbs. at one-quarter rating. All per brake H.P. The turbine at Hartford, average load, 1800 K.W., 155 lbs. steam pressure, 27 inch vacuum, 45° F. superheat, gave 19.1 lbs. of steam per K.W. hgur ; equal to about 11.46 per I. H.P. hour. Curtis Steam Turbine. In the Curtis Steam Turbine the velocity is given to the -steam in an expanding nozzle, designed so as to convert nearly all of the steam's expansive force into velocity in itself. STEAM Oi££T NOZZLE MOVING BLADES STATIONARY BLADES MOVING- BLADES STATIONARY BLA PES MOVING BLADES yZZLE DIAPHRAGM MOVING BLADES , WmEMMwlMilMM ccc^ccceecccccecc^ MOVING BLADES NOVING BLADES \ Fig. 25. Leaving the nozzle the steam passes successively two or more lines of vanes on the moving element, placed alternately, with reversed vanes, on the fixed element. 1456 STEAM. Governing is effected by closing or opening some of the nozzle valves, thus narrowing or widening the steam belt. Speed regulation is 2 to 4%. Revolutions per minute of 600 K.W. machine is 1500. Velocity of steam leaving the jet is 2000 ft. per second. Compared with large engine outfits in Manhattan Railway Company's New York Power Plant — the weights of Curtis is to weight of Reciprocating outfit as 1 is to 8. Fig. 26. The condensing type is usually designed for 150 lbs. gauge steam pressure, and a vacuum of 28 inches of mercury at sea level. Under these conditions normal overload may be 100%. Nozzles are different for different pressures of steam. This type is now being built with a condenser in its base, thereby securing fewer joints and connections and a better vacuum, resulting in a slight increase in the height of the machine. A vertical shaft and step bearing are typical of the Curtis turbine, STEAM TURBINES. 1457 larger sizes, and though being lubricated with oil, experiments are now being carried out, having in view the use of water in place of the oil as a floating medium, then steam packing of stem will be avoided, and no oil be used in condenser proper. This table gives some idea of the proportions of the Curtis Turbines, and is taken from a paper by A. H. Kruesi. DIRECT CURRENT. Num- Speed, R.P.M. Condensing Number Horizontal ber of K.W. Volts. or Non- of or Poles. Condensing. Stages. Vertical. 2 1J 5,000 80 Non-Condensing 1 Horizontal 2 15 4,000 80 " 1 " 2 25 3,600 125 " 1 M 4 75 2,400 125 (i 2 II 4 150 2,000 125 or 250 '* 2 U 4 300 1,800 250 (< 4 i< 4 300 2,000 550 Condensing 3 ii 4 600 1,800 550 " 3 " 4 500 1,800 550 t< 2 Vertical ALTERNATING CURRENT. — 60 CYCLES. 2 100 3,600 2,300 Condensing 3 Horizontal 4 500 1,800 2,300 u 2 Vertical 8 1500 900 2,300 (( 2 " 12 3000 600 2,300 " 4 it 14 5000 514 2,300 (1 4 it 15000 30000 HORSE POWER 40000 Fig. 27. Curves Showing the Relation of Foundation Material to Horse-Power. The Rateau type is somewhat similar to the De Laval, except that the combinations of nozzles and single wheel is repeated many times, with less expansion than in the older turbines. In making comparisons with other types of engines, it is found that for steam turbines, the average weight of the machine in lbs. per Brake H.P. is 52 lbs., against 422 for gas engines. About the same ratio, or 1 to 8, applies to reciprocating steam engines. 1458 STEAM. §§3 n © C ® — 03 Js © S.s * c aj £ CO® g c$,d S,d~ « A » jd .2 °° M Tj fcn d -J SB +1 b9 fl £ +J sd P*> X £5 >V T3 o9 d (0 I- V 49 0) d an 5&2 "d fl CQ d c-d go-- ©d 33 5 if •UOi;09g isa^uig ib -ai •bg jad -jji jad ra^g -sqi oooi uorpag isajpsuig jb -iij bg jad *jh J9H»H«H(OrtiOOiOOiaOiOO-1l>l> o6o6o6o6o6o6cJo5C5ciosoicsosososa>o>oioJoso»oi 3®»^t>ooooo5C500HTHN(Ncoco^^iq»q»w >t^t>t>t>t^t^^t>o6o6o6o6odo6odo6odododododod 'aanssajj aSn^o •uopoag ^san'Buis jb -hi •bg jad -jh J od" migajg -sqi ooox ^t-C5l»NMNI>H«OHCOH«OHCOHOH®HCOH WCCrlH^iaiqo«Jt>t>.oqoOCSO>OOrHiMC^oioididuSioiocDCD.cs>cdcDcocD; t>- 00 00 o> C5 O C COCOCOCO^COCOCOCOCOCOCOCOCOCOCOCOCOCOCO"^t S88a8S8^.SSac1g85S^88g^885Sc5 •ui *bg jad -sq^ 'ajnssaj^ aSti'eo M(NN(NMCl(NC*< STEAM TABLE. 1459 A handy rule for approximately determining the outflow of the steam is the following : If the absolute steam pressure at the inlet end of the orifice is p at- mospheres s kg. steam will flow through each mm. 2 of the smallest section area of the orifice per hour. The above company have in many trials demonstrated this to be true within five per cent. 1460 WATER-POWER. WATER-POWER. In determining the feasibility of utilizing water-power to operate electri- cally the industries of any particular town or city, careful consideration must be given to the following points, viz. : 1. The amount of water-power permanently available. 2. The cost of developing this power. 3. The in- terest on this amount. 4. The total demand for power. 5. The amounts and relative locations of the various kinds of power. 6. The cost of steam plants now in operation. 7. The interest on this amount. 8. Cost of fuel for plants now in operation. 9. Cost of operating present plants. Labor. 10. Cost of maintenance of present plants. 11. The amounts and kinds of electric power already in operation. 12. The distance of transmission. 13. The estimated cost of the hydraulic machinery. 14. The guaranteed efficiency and regulation of the hydraulic machinery. 15. Estimated cost of electric machinery. 16. Estimated cost of line construction. 17. Total cost of operating hydraulic and electric machinery. 18. Total cost of mainte- nance of hydraulic and electric plants. 19. The interest on the total esti- mated cost of proposed plant. 20. The estimated gross income. Charles T. Main makes the following general statements as to the value of a water-power : " The value of an undeveloped variable power is usually nothing if its variation is great, unless it is to be supplemented by a steam- plant. It is of value then only when the cost per horse-power for the double- plant is less than the cost of steam-power under the same conditions as mentioned for a permanent power, and its value can be represented in the same manner as the value of a permanent power has been represented. " The value of a developed power is as follows : If the power can be run cheaper than steam, the value is that of the power, plus the cost of plant, less depreciation. If it cannot be run as cheaply as steam, considering its cost, etc., the value of the power itself is nothing, but the value of the plant is such as could be paid for it new, which would bring the total cost of run- ning down to the cost of steam-power, less depreciation." Mr. Samuel Webber, Iron Age, Feb. and March, 1893, criticises the state- ments of Mr. Main and others who have made comparisons of costs of steam and of water-power unfavorable to the latter. He says : " They have based their calculations on the cost of steam, on large compound engines of 1000 or more h. p. and 120 pounds pressure of steam in their boilers, and by care- ful 10-hour trials succeeded in figuring down steam to a cost of about $20 per h. p., ignoring the well-known fact that its average cost in practical use, except near the coal mines, is from $40 to $50. In many instances dams, canals, and modern turbines can be all completed at a cost of $100 per h. p.; and the interest on that, and the cost of attendance and oil, will bring water-power up to but about $10 or $12 per annum ; and with a man compe- tent to attend the dynamo in attendance, it can probably be safely estimated at not over $15 per h. p. SYIVOPSIi OE REPORT REQUIRED OJJ WATER-POWER PROPERTY. Location. Geographical, etc. Sketch of river and its tributaries. Surrounding country and physical features. Sources ; lakes, springs, etc. Water's head; area drained, nature of, whether forest, swamp, snow- covered mountains, etc. Elevation of head waters and of mouth. Length from main source to mouth. Accessibility ; how and by what routes. Reports. Reports of IT. S. Coast or Geological Survey. Reports of Engineers U. S. Army. Any other reports. Any estimate by engineers and for what purpose. When it first attracted attention and for what reason. History. REPORT ON WATER-POWER PROPERTY. 1461 Rainfall. Average for several years for the drainage area. Maximum, what month. Minimum, what month. Comparison with other similar localities. Volume of Water, Gauging of river if possible. Reports by other engineers. Cubic feet per second flow. Cubic feet per second per mile of watershed = say .2 to .3 of total rainfall and | available as water-power. Comparison with other rivers. Reservoirs* Possibility of storing water for dry time. Available Fall. Location of ; accessibility, by what routes. Can power be used locally, or would it be necessary to transmit it, and if so, where to, and distances ? Nature of country over which it would have to be carried. Volume of water in cubic feet per second. Horse-Power of River. Calculated from available fall and volume. Horse-power for each fall or dam. Location of dams, dimensions, length, and height, best method of con- struction, estimated cost. Backwater ; volume, and how far ; what interests disturbed by it ; benefits, if any. Compare power with that of similar rivers. Probable cost of power at dams and transmitted. Applications Possible. Near by ; at distance, stating when and for what. Note industries appli- cable to ; comparison with other applications. I¥ew Industries Suggested, and old industries already going to which power is applicable. Cost to these, and comparison with cost of other forms of power already in use. Property of the Company. Land, buildings, water rights, flo wage rights, franchises, lines, rights of way. Character of deeds. Probable value. Comparison with other similar properties. Other resources. liabilities. Stocks, bonds, floating debt, other. Earning* Capacity. Probable cost of power per h. p. at power-house. Probable cost of power per h. p. delivered or transmitted. Price for which it can be sold at power-house, and price transmitted or delivered. General Features. Surrounding country, its characteristics, people, cities, and towns, indus- tries, condition of finances. Facilities for transportation, water and rail. Nearness of sources of supplies and sales of products. Horse-Power of a Waterfall. The horse-power of a waterfall is expressed in the following formula : Q = quantity of water in cubic feet flowing over the fall in 1 minute. £T= total head in feet, i.e., the distance between the surface of the water at the top of the fall, and that at its foot. In a water-power the head is the distance between the surface of the water in the head-race, and that of the water in the tail-race. 1462 WATER-POWER. w = weight of water per cubic foot = 62.36 lbs. at 60° F. O X H x w Gross horse-power of waterfall = Y or .00189 QH. Loss of head at the entrance to and exit from a water-wheel, together with the friction of the water passing through, reduces the power that can be developed to about 70 per cent of the gross power of the fall. Horse-Power of a Running 1 Stream. The power is calculated by the same formula as for a fall, but in this cae« H=r. theoretical head due to the velocity of the water in the stream = -— - where 64.4 v = velocity of water in feet per second. Q = the cubic feet of water actually impinging against the bucket per minute. Gross horse-power as .00189 QH. Wheels for use in the current of a stream realize only about .4 of the gross theoretical power. Current motors are often developed to operate in strong currents, such as that of the Niagara River opposite Buffalo, but are of little use excepting for small powers. Such a small fraction of the current velocity can be made use of that a current motor is extremely inefficient. In order to realize power from a current it is necessary to reduce its velocity in taking the power, and to get the full power would necessitate the backing up of the whole stream until the actual head equaled the theoretical. Power of Water flowing- in a Pipe* if due to velocity = — = — where v = velocity in feet per second. f H x due to pressure = — , where/ = pressure in lbs. per square foot. and w =. 62.36 lbs. =r weight 1 cubic foot of water. H 2 distance above datum line in feet. 2$r ~ r w In hydraulic transmission the work or energy of a given quantity of water under pressure is the volume in cubic feet x lbs. pressure per square foot. Q = cubic feet per second. P = pressure in lbs. per square inch. Horse-power = ^p = -2618 PQ. Mill-Power. It has been customary in the past to lease water-power in units larger than the horse-power, and the term mill-power has been used to designate the unit. The term has no uniform value, but is different in all localities. Emerson gives the following values for the seven more important water- power. Holyoke, Mass. — Each mill-power at the respective falls is declared to have the right during 16 hours in a day to draw 38 cubic feet of water per second at the upper fall when the head there is 20 feet, or a quantity proportionate to the height at the falls. This is equal to 86.2 horse-power as a maximum. Lowell, Mass. — The right to draw during 15 hours in the day so much water as shall give a power equal to 25 cubic feet a second at the great fall, when the fall there is 30 feet. Equal to 85 h. p. maximum. Lawrence, Mass. — The right to draw during 16 hours in a day so much water as shall give a horse-power equal to 30 cubic feet per second when the head is 25 feet. Equal to 85 h. p. maximum. Minneapolis, Minn. — 30 cubic feet of water per second with head of 22 feet. ' Equal to 74.8 h. p. MERCURY AND WATER. 1463 Manchester , N. H. — Divide 725 by the number of feet of fall minus 1, and the quotient will be the number of cubic feet per second in thai fall. For 20 feet fall this equals 38.1 cubic feet, equal to 86.4 h. p. maximum. Cohoes, N.Y. — "Mill-power" equivalent to the power given by 6 cubic feet per second, when the fall is 20 feet. Equal to 13.6 h. p. maximum. Passaic, jV. J. — Mill-power : The right to draw 8£ cubic feet of water per second, fall of 22 feet, equal to 21.2 horse-power. Maximum rental, $700 per year for each mill-power = $33.00 per h. p. The horse-power maximum above given i6 that due theoretically to the weight of water and the height of the fall, assuming the water-wheel to have perfect efficiency. It should be multiplied by the efficiency of the wheel, say 75 per cent for good turbines, to obtain the h.p. delivered by the wheel. At Niagara power has in all cases been sold by the horse-power delivered to the wheels if of water, and to the building-line if electrical. Charges for water in Manchester, Lowell, and Lawrence, are as follows : Manchester, About $300 per year per mill-power for original purchases. $2 per day per mill-power for surplus. Lowell. About $300 per year per mill-power for original purchases. $2 per day per mill-power during " back-water." $4 per day per mill-power for surplus under 40 per cent. $10 per day per mill-power for surplus over 40 per cent and under 50 per cent. $20 per day per mill-power for surplus over 50 per cent. $75 per day per mill-power for any excess over limitation. Lawrence. About $300 per year per mill-power for original purchases. About <$1200 per year per mill-power for new leases at present. $4 per day per mill-power fur surplus up to 20 per cent. $8 per day per mill-power for surplus over 20 and under 50 per cent. $4 per day per mill-power for surplus under 50 per cent. COMPARISON OF (OLlM\o» OF WATER I]¥ FEET. IHercury in Indie*. and Pressure in litis., per Square Inch. Lbs. Water. | Merc'ry Water. Merc'ry Lbs. Merc'ry Water. Lbs. Press. Press. Press. Sq. In. Feet. Inches. Feet. Inches. Sq. In. Inches. Feet. Sq. In. 1 2.311 2.046 1 0.8853 0.4327 1 1.1295 0.4887 2 4.622 4.092 2 1.7706 0.8654 2 2.2590 0.9775 3 6.933 6.138 3 2.6560 1.2981 3 3.3885 1.4662 4 9.244 8.184 4 3.5413 1.7308 4 4,5181 1.9550 5 11.555 10.230 5 4.4266 2.1635 5 5.6476 2.4437 6 13.866 12.2276 6 5.3120 2.5962 6 6.7771 2.9325 7 16.177 14.322 7 6.1973 3.0289 7 7.9066 3.4212 8 18.488 16.368 8 7.0826 3.4616 8 9.0361 3.9100 9 20.800 18.414 9 7.9680 3.8942 9 10.165 4.3987 10 23.111 20.462 10 8.8533 4.3273 10 11.295 4.8875 11 25.422 22.508 11 9.7386 4.7600 11 12.424 5.3762 12 27.733 24.554 12 10.624 5.1927 12 13.554 5.8650 13 30.044 26.600 13 11.509 5.6255 13 14.683 6.3537 14 32.355 28.646 14 12.394 6.0582 14 15.813 6.8425 15 34.666 30.692 15 13.280 6.4909 15 16.942 7.3312 16 36.977 32.738 16 14.165 6.9236 16 18.072 7.8200 17 39.288 34.784 17 15.050 7.3563 17 19.201 8.3087 18 41.599 36.830 18 15.936 7.7890 18 20.331 8.7975 19 43,910 38.876 19 16.821 8.2217 19 21.460 9.2862 20 46.221 40.922 20 17.706 8.6544 20 22.590 9.7750 21 48.532 42.968 21 18.591 9.0871 21 23.719 10.264 22 50.843 45.014 22 19.477 9.5198 22 24,849 10.752 23 53.154 47.060 23 20.362 9.9525 23 25.978 11.241 24 55.465 49.106 24 21.247 10.385 24 27.108 11.7300 25 57.776 51.152 25 22.133 10.818 25 28.237 12.219 26 60.087 53.198 26 23.018 11.251 26 29.367 12.707 27 62.398 55.244 27 23.903 11.683 27 30.496 13.196 28 64.709 57.290 28 24.789 12 116 28 31.626 13.685 39 67.020 59.336 29 25.674 12.549 29 32.755 14.174 30 69.331 61.386 30 26.560 12.981 30 33.885 14.662 1464 WATER-POWER. | os 00 8 CO 00 s CN CO S s 3 s 00 00 IO IO o ' fc CO CO CO 00 8 B w tj5 CM OS s 5 lO o p ft .5 €£= m 00 o 8 ^ CO IO _Ztf R ' } •" -Mb T= for iron, usually 48,000 lbs. per sq. in. T— for steel, 62,000 lbs. per sq. in. P = safe working pressure, per sq. in. t = thickness of sheet in inches. B = radius of pipe in inches. c = factor of safety : 3 to 3.5 for this work. /== proportional strength of plates after riveting: Double riveting ... 0.7 Single riveting ... 0.5 The Water Power Plant at Puyallup River near Tacoma will have a steel pipe line 1700 feet long, beginning 48" diameter, reducing to 36" diam- eter at the end, built by Ridson Iron Works, San Francisco, Cal. 1468 WATER-POWER. In many cases much expense may be saved in pipe by conveying the water in a flume or ditch along the hillside, covering in this way a large part of the distance, then piping it down to the power station by a short line. This is more especially applicable to large plants, where the cos't of the pipe is an important item. DATA FOR FLinF^ 4*1> HITCHES. To give a general idea as to the capacity of flumes and ditches for carry- ing water, the following data is submitted : The greatest safe velocity for a wooden flume is about 7 or 8 feet per second. For an earth ditch this should not exceed about 2 feet per second. In Cali- fornia it is the general practice to lay a flume on a grade of about \ inch to the rod, or often 2 inches to the 100 feet, depending on the existing conditions. Assuming a rectangular flume 3 feet wide, running 18 inches deep, its velocity and capacity would be shown as below : Grade. Vel. in Ft. per Sec. Quantity Cu. Ft. Min. 4 inch to rod 2.6 702 J " " " 3.7 999 } " " " 5.3 1,431 As the velocity of a flume or ditch is dependent largely on its size and character of formation, no more specific data than the above can be given. It is not safe to run either ditch or flume more than about | or g full. WOODE.V^flVE PIPE. Wooden-stave pipe has been used to some extent on the Pacific Coast for conveying water long distances under heads not much exceeding 200 feet. Although the construction of such pipe is quite simple, yet considerable skill and care are necessary to make water-tight work. The plant of the San Gabriel Los Angeles Transmission, California, uses several miles of wooden-stave pipe, 48 ins. diameter. The pipe is laid uniformly ten feet below hydraulic grade; and the wood is of such thickness as to be always water-soaked, and will thus outlast almost any other form of construction. The staves are placed so as to break joints, the flat sides are dressed to a true circle, and the edges to radial planes. The staves are cut off square at the ends, and the ends slotted, a tight-fitting metallic tongue being used to make the joint. The pipe depends upon steel bands for its strength, and in the case above mentioned they are of round steel rod placed ten inches apart from center to center. Where the pressures vary along the line, bands can be spaced closer or wider apart to make the necessary strength. The preference is given round bands over flat ones, on account of their embedding themselves in the wood better as it swells. They also expose less surface to rust than would flat ones of the same strength. The ends of the bands are secured together through a malleable iron shoe, having an interior shoulder for the head of the bolt, and an exterior shoulder for the nut, the whole band thus being at right angles to the line of the pipe. Where curvesare not too sharp, they can easily be made in the wooden pipe ; but for short turns, sec- tions of steel-riveted pipe of somewhat larger internal diameter than that of the wooden pipe are introduced. The joints between wood and steel are made by a bell on the steel pipe that is larger than the outside diameter of the wooden pipe. After partly filling the space between bell and wood with oakum packed hard, for the remainder use neat Portland cement. Advantages claimed for this type are that it costs less than any other form, and especiallv so where transportation is over the rugged country where it is most liable to be used ; great length of lif e, and greater capacity than either casMron or steel-riveted. Compared with new riveted pipe, the carrying capacitv of stave pipe is said to be from 10 to 40 % more, and this difference increases with age as the wooden pipe gets smoother, while the friction of the metal pipe increases to a considerable degree. As compared with open flumes, the life is so much greater and repairs so much less as to considerably more than counterbalance the first cost. For detailed information on wooden-stave pipe, see papers by A. L. Adams, September, 1898, Am. Soc. C. E. Note. Mr. Arthur L. Adams writes that the pipe laid at Astoria, Ore., about which he wrote ten years ago has not proved lasting, one third of it hf ving been renewed during the first decade of its existence. RIVETED HYDRAULIC PIPE. 1469 lABIE ©* RIVETED HYDRAVIIC PIPI. (Pelton Water Wheel Co.) Showing weight, with safe head for various sizes of double-riveted pipe. ft ft . °.S 1.5 5.5 ft ft . oq -5.5 Thickness of iron by wire gauge. Head in feet the pipe will safely stand. Cu. ft. water pipe will con- vey per min. at vel. 3 ft. per sec. Weight per lineal ft. in lbs. ft ft . 5.5 ft 'ft u_, 02 *2 Thickness of iron by wire gauge. Head in feet the pipe will safely stand. Cu. ft. water pipe will con- vey per min. at vel. 3 ft. per second. ft 3 7 18 400 9 2 18 254 16 165 320 16i 4 12 18 350 16 2* 18 254 14 252 320 20| 4 12 16 525 16 3 18 18 254 254 12 11 385 424 320 320 27J 5 20 18 325 25 3 5 30 5 20 20 16 14 500 675 25 25 18 254 10 505 320 34 5 20 20 314 314 16 14 148 227 400 400 18 6 28 18 296 36 4i 22£ fi 38 16 487 36 5| 7i 20 314 12 346 400 30 6 28 14 743 36 20 20 314 344 11 10 380 456 400 400 32£ 36£ 7 7 38 18 254 50 3 8| 38 16 419 50 22 380 16 135 480 20 7 38 14 640 50 22 22 22 380 380 380 14 12 11 206 316 347 480 480 480 24f 32f 35f 8 50 16 367 63 3 8 8 50 50 14 12 560 854 63 63 if 22 380 10 415 480 40 9 9 63 63 16 14 327 499 80 80 81 lOf 24 24 ?4 452 452 45? 14 12 11 188 290 318 570 570 570 27i 39 9 63 12 761 80 14i 24 24 452 452 10 8 379 466 570 570 43£ 53 10 78 16 14 12 11 295 100 100 100 100 9J 10 10 10 78 78 78 450 687 754 llf 15| 17* 26 26 26 530 530 530 14 12 11 175 267 294 670 670 670 42 10 78 10 900 100 26 26 530 530 10 8 352 432 670 670 47 11 95 16 14 12 11 269 120 120 120 120 9f 57J 11 11 11 9b 95 95 412 626 687 13 171 18f 28 28 ?8 615 615 615 14 12 11 102 247 273 775 775 775 31i 45 11 95 10 820 120 21 28 28 615 615 10 8 327 400 775 775 St 12 113 16 246 142 142 Hi 12 113 14 377 14 30 706 12 231 890 44 12 12 12 113 113 113 12 11 10 574 630 753 142 142 142 18| 19| 22| 30 30 30 30 706 706 706 706 11 10 8 7 254 304 375 425 890 890 890 890 48 54 65 13 132 16 228 170 12 74 13 132 14 348 170 15 36 1017 11 141 1300 58 13 13 13 132 132 132 12 11 10 530 583 696 170 170 170 20 22 24£ 36 36 36 1017 1017 1017 10 8 7 155 192 210 1300 1300 1300 67 78 88 14 153 16 211 200 13 40 1256 10 141 1600 71 14 153 14 324 200 16 40 1?56 8 174 1600 86 14 153 12 494 200 21* 40 1?56 7 189 1600 97 14 153 11 543 200 23£ 40 1256 6 213 1600 108 14 153 176~ 176 10 648 200 26 40 1256 4 250 1600 126 15 15 16 14 197 302 225 225 13| 17 42 4? 1385 1385 10 8 135 165 1760 1760 74* 91 15 IV b 12 460 225 23 42 1385 7 180 1660 102 15 J. 7 b 11 507 225 24* 42 1385 6 210 1760 114 15 IV b 10 606 225 28 42 42 1385 1385 4 240 270 1760 1760 133 16 201 16 185 255 14f 137 16 201 14 283 255 m 42 1385 3 300 1760 145 16 201 12 432 255 42 1385 1 321 1760 177 16 201 11 474 255 26£ 42 1385 363 1760 216 16 201 10 567 255 29£ 1470 WATER-POWER. Cubic Feet of Water per TOLinute Discharged Through Orifice 1 Square Inch in Area. .For any other size of orifice, multiply by its area in square inches. *& 6 V, 6 V, 6 V, © "8 ® V, ® V, ® 4* S>£ -^ <» 4J +s ®g p »« 4i »^ *? £-*> +» ©-S CO 0> a> boa a> 0) f-i £ bfls fa c3.5 CO a © *H - CO 43 © ^ a — fa ^.S A fa ^-2 A fa g.9 A fa si .5 A A fa g.S — fa rt.a 3s 2.2 b 2.2 fc c3 M 2.2 s 0QS 2 JS^ CD g *Q£ g«£ $a ^Pi X S gfli gs£ S 5 ga£ 3 1.12 13 2.20 23 2.90 33 3.47 43 3.95 53 4.39 63 4.78 4 1.27 14 2.28 24 2.97 34 3.52 44 4.00 54 4.42 64 4.81 5 1.40 15 2.36 25 3.03 35 3.57 45 4.05 55 4.46 65 4.85 6 1.52 16 2.43 26 3.08 36 3.62 46 4.09 56 4.52 66 4.89 7 1.64 17 2.51 27 3.14 37 3.67 47 4.12 57 4.55 67 4.92 8 1.75 18 2.58 28 3.20 38 3.72 48 4.18 58 4.58 68 4.97 9 1.84 19 2.64 29 3.25 39 3.77 49 4.21 59 4.63 69 5.00 10 1.94 20 2.71 30 3.31 40 3.81 50 4.27 60 4.65 70 5.03 11 2.03 21 2.78 31 3.36 41 3.86 51 4.30 61 4.72 71 5.07 12 2.12 22 2.84 32 3.41 42 3.91 52 4.34 62 4.74 72 5.09 Table Snowing* the Theoretical Velocity and Discharge in Cubic feet Through an Orifice of 1 Square Inch Issu- ing Under Mead* Varying from 1 to lOO feet. Theoreti- Theoret- Theoreti- Theoret- Theoreti- Theoret- .s • cal Dis- ical cal Dis- ical a . cal Dis- ical •d ® charge in Velocity charge in Velocity •d® charge in Velocity Sfa Cu. Ft. in Feet Sfa Cu. Ft. in Feet a>fa Cu. Ft. in Feet W per Min. per Min. a per Min. per Min. W per Min. per Min. 1 3.34 481.2 35 19.77 2847.6 69 27.74 3997.1 2 4.73 680.4 36 20.05 2887.2 70 27.94 4021.1 3 5.79 833.4 37 20.33 2926.8 71 28.14 4054.5 4 6.68 962.4 38 20.60 2966.4 72 28.34 4283.0 5 7.47 1075.8 39 20.87 3004.8 73 28.53 4111.3 6 8.18 1178.4 40 21.13 3043.2 74 28.73 4139.4 7 8.84 1273.2 41 21.38 3081.1 75 28.93 4165.2 8 9.45 1360.8 42 21.64 3118.5 76 29.11 4194.9 9 10.02 1443.6 43 21.90 3156.4 77 29.30 4222.4 10 10.57 1521.6 44 22.15 3191.8 78 29.49 4249.8 11 11.08 1596.0 45 22.40 3227.8 79 29.68 4265.9 12 11.57 1666.8 46 22.65 3263.6 80 29.87 4303.6 13 12.05 1734.6 47 22.89 3298.9 81 30.06 4330.8 14 12.50 1800.6 48 23.14 3333.8 82 30.24 4357.4 15 12.94 1863.6 49 23.38 3368.4 83 30.42 4383.6 16 13.37 1924.8 50 23.61 3402.5 84 30.61 4410.2 17 13.78 1984.2 51 23.85 3436.4 85 30.79 4436.4 18 14.18 2041.8 52 24.08 3469.9 86 30.97 4462.4 19 14.57 2097.6 53 24.31 3503.1 87 31.15 4488.2 20 14.95 2152.2 54 24.54 3536.0 88 31.33 4514.0 21 15.31 2205.0 55 24.76 3568.6 89 31.50 4539.5 22 15.67 2256.6 56 24.99 3600.9 90 31.68 4565.0 23 16.02 2307.6 57 25.21 3632.9 91 31.86 4590.3 24 16.37 2357.4 58 25.43 3664.6 92 32.04 4615.4 25 16.71 2406.0 59 25.65 3696.1 93 32.20 4040.5 26 17.04 2453.4 60 25.87 3727.3 94 32.38 4665.3 27 17.36 2500.2 61 26.08 3758.2 95 32.55 4690.1 28 17.68 2545.8 62 26.29 3788.9 96 32.72 4714.7 29 17.99 2590.8 63 26.51 3819.3 97 32.89 4739.2 30 18.30 2635.8 64 26.72 3849.6 98 33.06 4763.5 31 18.60 2679.0 65 26.92 3879.5 99 33.23 4787.8 32 18.90 2722.2 66 27.13 3909.2 100 33.40 4812.0 33 19.20 2764.2 67 27.33 3938.7 34 19.49 2806.2 68 27.54 3968.4 THEORY OF ROD FLOAT GAUGING. 1471 flow of Water Throug-li an Orifice. a = area of orifice in square inches. Q = cubic feet discharged per minute. h = head in inches. #=.624 V^xa. The best form of aperture for giving the greatest flow of water is a coni- cal aperture whose greater base is the aperture, the height or length of the section of cone being half the diameter of aperture, and the area of the small opening to the area of the large opening as 10 to 16 ; there will be no contraction of the vein, and consequently the greatest attainable discharge will be the result. MEASUREMENT OF M.OW OE WATER IX A STREAM. The quantity of water flowing in a stream may be roughly estimated as follows : Find the mean depth of the stream by taking measurements at 10 or 12 or more equal dis- tances across. Multi- ply this mean depth by the width of the stream, which will give the total cross-section of the prism. Find the velocity of the flow in feet per Fig. 28. minute, by timing a float over a measured distance, several times to get a fair average. Use a thin float, such as a shingle, so that it may not be influenced by the wind. The area or cross- section of the prism multiplied by the ve- locity per minute will give the quantity per minute in cubic feet. Owing to the friction of the bed and banks the actual flow is re- duced to about 83 per cent of the calculated flow as above. Fig. 29. THEORY OE ROD FLOAT OAUCHICG. (From Report on Barge Canal, 1901, Edward A. Bond, N. Y. State Engineer.) The hydrometric rod may consist of either a plain wooden rod of uniform diameter, weighted at its lower end with iron or lead pipe of equal diam- eter, so as to make it sink vertically in the water to nearly its full length, 1472 WATER-POWER. or of a tin tube of uniform diameter, made either continuous or in sections fitting water-tightly into each other, and properly weighted with leaden shot, bullets, etc., at the bottom. If such a rod is placed carefully in the water, so as to prevent any vertical motion, and its projecting part is not acted upon by the wind, it may be assumed that in a short time it will move with the mean velocity of the water in the vertical plane in which it floats. When a straight cylindrical rod of uniform diameter is immersed verti- cally in a moving body of water and kept from sinking, it encounters therein filaments having different velocities in the direction of the stream, anil eventually acquires an intermediate velocity which is very nearly the mean of those acting upon it. Some of the fluid particles will be moving faster than the rod, while others move slower ; tne former will tend to accelerate the motion of the rod, both by direct pressure and by the lateral friction, while the latter tend to retard it. In the ensuing state of equili- brium and uniform motion, the accelerating and retarding forces acting on the rod must be equal, and will form a couple which causes the rod to assume a sMghtly inclined position in the water. Furthermore, when the channel is regular, and the rod reaches nearly to the bottom, the general law according to which the velocity of the successive filaments from the surface downwards varies, has been determined approximately by experi- ment, and it becomes possible to express the sums of the said accelerating and retarding forces in relatively simple mathematical terms. From the equality of these expressions, it is then found that the rod assumes the velocity of the water filament, which is located at a depth =0.61 X, where (L\\ denotes the immersed length of the rod. In like manner, the velocity \v x ) of the rod may also be compared with the computed or theoretical mean vel ocity (v 2 ) of all the water filaments in the vertical line or plane from the surface to the depth (X) ; and as it is found therefrom that (t' t ) is a little lesu than (v 2 ), it may be considered that (y x ) is equal to the mean velocity {vt*) for a depth a little greater than the said length (X). Under ordinary conditions in canals and rivers with regular channels and moderate veloci- ties, the immersed length (X) of the rod should be about 94% of the depth (T) of the water in the vertical plane of observation. From his extensive experiments at Lowell with such rods 2 inches in diameter and of different length (X) ranging from 87 to 99 per cent of the depth (T), the latter being made to vary from 8.1 to 9.5 feet, and with mean velocities (v m ) ranging from 0.5 to 2.8 feet per second, Francis deduced the following empirical formula for finding (vm) from the observed velocity (Vj) of the rod: , = v x ( 1.102 — 0.116 y V Commenting on the results given by this formula in comparison with the simultaneous observations of discharge over his standard weir, Mr. Francis states that taking the whole of the experiments together, the aver- age difference is about f of 1 per cent, and that the largest difference is an excess of about 3.7 per cent over the weir measurement when the velocity was only 0.5 foot per second. It is also probable that the above formula will not give trustworthy values of (vm) when the immersed length (X) of the rod is less than 75 per cent of the depth (T); hence it is desirable to make (X) as nearly equal to (T) as the character of the bed of the channel will permit. Practical Consideration. — In order that the work of gauging a water-course with rods may be prosecuted expeditiously and with fairly accurate results, certain practical considerations should be observed. The rods should be straight cylinders of uniform diameter having the smoothest practicable surface. Their diameter should be as small as is compatible with proper strength and stiffness, and the loading at the bottom should be concentrated so as to bring the center of gravity as low down as possible in the water, at the same time being rigidly attached so as to remain in place even if the rod is inverted. They should also have ample buoyancy, in order to bring them quickly to their normal depth of immersion after accidental submergence, and the projecting portion should be as short as possible con- sistent with the function of serving as a marker. In their experiments, Francis and Cunningham used tin tubes about 2 inches in diameter, while Grebenau and others used varnished wooden rods, having diameters from 1.2 to 1.5 inches. Cunningham also used such rods, but gave the preference to the tubes. HORSE-POWER OF WATER. 1473 Miners' Inch Measurements. (Pelton Water Wheel Co.) Miners' inch is a term much in use on the Pacific Coast and in the mining regions, and is described as the amount of water flowing through a hole 1 inch square in a 2-inch plank under a head of 6 inches to the top of the orifice. Fig. 28 shows the form of measuring-box ordinarily used ; and the follow- ing table gives the discharge in cubic feet per minute of a miners' inch of water, as measured under the various heads and different lengths and heights of apertures used in California. C3 Openings 2 Inches High. Openings 4 Inches High. O b£ . rj S3 OQ SO* Head to Head to Head to Head to Head to Head to Center, Center, Center, Center, Center, Center, 5 Ins. 6 Inches. 7 Inches. 5 Inches. 6 Inches. 7 Inches. Cu.Ft. Cu. Ft. Cu. Ft. Cu. Ft. Cu. Ft. Cu. Ft. 4 1.348 1.473 1.589 1.320 1.450 1.570 6 1.355 1.480 1.596 1.336 1.470 1.595 8 1.359 1.484 1.600 1.344 1.481 1.608 10 1.361 1.485 1.602 1.349 1.487 1.615 12 1.363 1.487 1.604 1.352 1.491 1.620 14 1.364 1.488 1.604 1.354 1.494 1.623 16 1.365 1.489 1.605 1.356 1.496 1.626 18 1.365 1.489 1.606 1.357 1.498 1.62S 20 1.365 1.490 1.606 1.359 1.499 1.63C 22 1.366 1.490 1.607 1.359 1.500 1.631 24 1.366 1.490 1.607 1.360 1.501 1.632 26 1.366 1.490 1.607 1.361 1.502 1.633 28 1.367 1.491 1.607 1.361 1.503 1.634 30 1.367 1.491 1.608 1.362 1.503 1.635 40 1.367 1.492 1.608 1.363 1.505 1.637 50 1.368 1.493 1.609 1.364 1.507 1.639 60 1.368 1.493 1.609 1.365 1.508 1.64C 70 1.368 1.493 1.609 1.365 1.508 1.641 80 1.368 1.493 1.609 1.366 1.509 1.641 90 1.369 1.493 1.610 1.366 1.509 1.641 100 1.369 1.494 1.610 1.366 1.509 1.642 Note. — The apertures from which the above measurements were obtained were through material 1\ inches thick, and the lower edge 2 inches above the bottom of the measuring-box, thus giving full contraction. FI.OW Of WATER OVER WEIRS. Weir Dam Measurement. (Pelton Water Wheel Co.) Place a board or plank in the stream, as shown in Fig. 29, at some point where a pond will form above. The length of the notch in the dam should be from two to four times its depth for small quantities, and longer for large quantities. The edges of the notch should be beveled toward the intake side as shown. The overfall below the notch should not be less than twice its depth, that is, 12 inches if the notch is 6 inches deep, and so on. In the pond, about 6 feet above the dam, drive a stake, and then obstruct the water until it rises precisely to the bottom of the notch, and mark the stake at this level. Then complete the dam so as to cause all the water to flow through the notch, and, after time for the water to settle, mark the stake again for this new level. If preferred, the stake can be driven with its top precisely level with the bottom of the notch, and the depth of the water be measured with a rule after the water is flowing free, but the marki 1474 WATER-POWER. are preferable in most cases. The stake can then be withdrawn ; and the distance between the marks is the theoretical depth of flow corresponding to the quantities in the table. Francis's Formulae for Weirs. Weirs with both end contractions ) suppressed j Weirs with one end contraction \ suppressed j Weirs with full contraction . . As given by- Francis. Q =a 3.33lh* Q = 3.33(1 — Ah) ti* Q — 3.33(1 — ,2h)h? As modified by Smith. 3.29 (*+£)** 3.29lh* 3.29 (-a The greatest variation of the Francis formulae from the value of c given by Smith amounts to 3£ per cent. The modified Francis formulae, says Smith, will give results sufficiently exact, when great accuracy is not required, within the limits of h, from .5 feet to 2 feet, I being not less than 3 h. Q — discharge in cubic feet per second, I ±= length of weir in feet, h = effective head in feet, measured from the level of the crest to the level of still water above the weir. If Q / = discharge in cubic feet per minute, and V and W are taken in inches, the first of the above formulae reduces to Q / = OAl'h'* ■ The values are suf- ficiently accurate for ordinary computations of water-power for weirs without end contraction, that is, for a weir the full width of the channel of approach, and are approximate also for weirs with end contraction when I — at least 107*, but about 6 per cent in excess of the truth when I == 4ft. Weir Table. Table Showing the Quantity of Water Passing over Weirs in Cubic Feet per Minute. oo fl .a '-*"* +3 © u w E ° . o Be. P5«h*S ~ >5 'ji r-H a*.-. ^ u % ° a, © 2 u a ce ©►> 4.85 5.78 6.68 7.80 8.90 10.00 11.23 12.45 13.72 15.02 16.36 17.75 19.17 20.63 22.11 23.63 25.20 26.78 28.43 30.06 31.75 33.45 35.22 36.98 38.80 40.63 42.49 44.39 46.29 48.22 « ° a J- Sh — ' III *l 5 ^ 5 T3 fa ^ £ fc, 00 O g'S £ © 3 £ ® fc> 50.20 52.18 54.22 56.25 58.33 60.42 62.55 64.68 66.86 68.98 71.27 73.45 75.77 78.04 80.36 82.63 85.04 87.43 89.82 92.16 94.67 97.11 99.50 102.10 104.63 107.13 109.74 112.31 114.91 117.51 O ° G -2 «H •a fc os a a fa .5 * w *» _d •2 ^ © ^ be i- oo fl 5 © tH Cub per pass eacl Len Wei g-gg 120.18 m 122.82 12| 125.52 13 128.14 13i 130.93 13§ 133.65 13| 136.43 14 139.18 HI 141.99 14| 144.80 147.64 15 150.47 m 153.35 15£ 156.20 15| 159.14 16 162.07 16J 164.99 16£ 167.89 16f 169.92 17 173.90 13 176.92 179.94 17| 182.99 18 186.03 18* 189.13 18J 192.20 18| 195.32 19 198.47 m 19} 201.59 207.94 19} ^ ^ EC o a © 214.32 220.76 227.30 233.92 240.54 247.22 254.03 260.83 267.77 274.70 281.72 288.82 295.93 303.10 310.36 317.69 325.03 332.42 339.91 347.45 355.02 362.77 370.34 378.12 385.87 393.66 401.63 409.58 417.48 425.68 HORSE-POWER OF WATER. 1475 X 4 BJLES lOIt CALCULATING THE HORIE-POWER Of 1 IVAXElt. (Pelton Wheel Co.) Miners ' Inch Table. Cubic feet Table. The following table gives the horse- The following table gives the power of one miners' inch of water horse-power of one cubic foot of under heads from one up to eleven water per minute under heads from hundred feet. This inch equals 1£ one up to eleven hundred feet. cubic feet per minute. .5 Horse- .9 .SB'S Horse- B Horse- ej 02 ** Horse- Power. Power. Power. IS® Power. w. w w H 1 .0024147 320 .772704 1 .0016098 320 .515136 20 .0482294 330 .796851 20 .032196 330 .531234 30 .072441 340 .820998 30 .048294 340 .547332 40 .096588 350 .845145 40 .064392 350 .563430 50 .120735 360 .869292 50 .080490 360 .579528 60 .144882 370 .893439 60 .096588 370 .595626 70 .169029 380 .917586 70 .112686 380 .611724 80 .193176 390 .941733 80 .128784 390 .627822 90 .217323 400 .965880 90 .144892 400 .643920 100 .241470 410 .990027 100 .160980 410 .660018 110 .265617 420 1.014174 110 .177078 420 .676116 120 .289764 430 1.038321 120 .193176 430 .692214 130 .313911 440 1.062468 130 .209274 440 .708312 140 .338058 450 1.086615 140 .225372 450 .724410 150 .362205 460 1.110762 150 .241470 460 .740508 160 .386352 470 1.134909 160 .257568 470 .756606 170 .410499 480 1.159056 170 .273666 480 .772704 180 .434646 490 1.183206 180 .289764 490 .788802 190 .458793 500 1.207350 190 .305862 500 .804900 200 .482940 520 1.255644 200 .321960 520 .837096 210 .507087 540 1.303938 210 .338058 540 .869292 220 .531234 560 1.352232 220 .354156 560 .901488 230 .555381 580 1.400526 230 .370254 580 .933684 240 .579528 600 1.448820 240 .386352 600 .965880 250 .603675 650 1.569555 250 .402450 650 1.046370 260 .627822 700 1.690290 260 .418548 700 1.126860 270 .651969 750 1.811025 270 .434646 750 1.207350 280 .676116 800 1.931760 280 .450744 800 1.287840 290 .700263 900 2.173230 290 .466842 900 1.448820 300 .724410 1000 2.414700 300 .482940 1000 1.609800 310 .748557 1100 2.656170 310 .499038 1100 1.770780 When the Evact Head is found in Above Table. Example.— Have 100 foot head and 50 inches of water. How many horse-power ? By reference to above table the horse-power of 1 inch under 100 feet head is .241470. The amount multiplied by the number of inches, 50, will give 12.07 horse-power. When Exact Head is not found in Table. Take the horse-power of 1 inch under 1 foot head, and multiply by the number of inches, and then by number of feet head. The product will be the required horse-power. The above formula will answer for the cubic-feet table, by substituting the equivalents therein for those of miners' inches. Note.— The above, tables are based upon an efficiency of 85 percent. 1476 WATER-POWER. WATER-WHEELS. Undershot Wheels, in which the water passes under acting by im- pulse, when constructed in the old-fashioned way with flat boards as floats, have a maximum theoretical efficiency of 50 per cent ; but with curved floats, as in Poncelet's wheel, which are arranged so that the water enters without shock and drops from the floats into the tail-race without horizontal velo- city, the maximum efficiency is as great as for overshot wheels, and the available efficiency is found to be about 60 per cent. The velocity of the periphery should be about .5 of the theoretical velocity of the water due to the head. .Breast and Overshot Wheel*. The best peripheral velocity is about 6 feet per second, and for the water supplied to it about 12 feet per second, which is the velocity due to a fall of about 2J feet ; therefore, the point at which the water strikes the wheel should be 2\ feet below the top-water level. The chief cause of loss in over- shot wheels is the velocity which the water possesses at the moment it falls from the float or bucket ; overshot wheels are good for falls of 13 feet to 20 feet ; below that breast wheels are preferable. The capacity of the buckets should be three times the volume of water held in each. The distance apart of the buckets may be 12 inches in high-breast and overshot wheels, or 18 inches in low-breast wheels, while the opening of buckets may be 6 to 8 inches in high-breast, and 9 inches to 12 inches in low-breast wheels. TVRBOES. These may be divided into two main classes, viz., pressure and impulse turbines. The former maybe again divided into the following: parallel- flow, outward-flow, and inward-flow turbines, according to the direction in which the water flows through the turbine in relation to its axis. Parallel-flow turbines, sometimes called downward-flow, are best suited for low falls, not exceeding say 30 feet. Fontaine's turbine is of this class, the wheel being placed at the bottom of the water-pipe or flume, just above the level of the tail-race. The water passes through guide blades and strikes the curved floats of the wheel. Jonval's turbine is of similar type, but is arranged to work partly by suction, and may be placed above the level of the tail-race without loss of power, which is often more convenient for working. The efficiency is from 70 to 72 per cent with well-designed wheels of this type. Fig. 30. Victor Wheel set in ordinary Flume. Outward-flow Turbines have a somewhat higher efficiency than the parallel-flow — as much as 88 per cent has been realized by Boyden's tur- bine ; Fourneyron's has given a maximum of 79 per cent. Inward-flow Turbines have been designed by Swain and others. Tests made on a Swain turbine by J. B. Francis gave a maximum effi- ciency of 84 per cent with full supply, and with the gate a quarter open 61 per cent, the circumferential velocity of the wheel ranging from 80 to 60 per cent of the theoretical velocity due to the head of water. In Swain's turbine the edges of the floats are vertical and opposite the guide blades, DIMENSIONS OF TURBINES. 1477 the edges towards the bottom of the floats being bent into a quadrant form. The Victor turbine is claimed to give 88 per cent under favorable conditions. It receives the water upon the outside, and discharges it downward and out- ward, the lines of discharge occupying the entire diameter of the lower portion of the wheel, excepting only the space tilled by the lower end of the shaft. Impulse Turbines are suitable for very high falls. The Girard and Pelton are both of this type. It is advised that pressure turbines be used on heads of 80 feet or 100 feet, but above this an impulse turbine is best. A Girard turbine is working under a fall of 650 feet. Installing: Turbines. Particular attention must be paid to the designing and construction of water-courses. The forebay leading to the flume should be of such size that the velocity of the water never exceeds 1* feet per second, and should be free from abrupt turns or other defects likely to cause eddies. The tail-race should have similar capacity and sufficient depth below the surface of the stream to allow at least 2 feet of dead water standing when the wheels are not in motion, and with large wheels, 3 feet to 4 feet ; after extending sev- eral feet beyond the flume, this may be gradually sloped up to the level of the stream. It is not uncommon to see 2 feet or 3 feet of head lost in defective races. When setting turbines some distance above the tail-race, the mouth of the draft-tube must be 2 inches to 4 inches below the lowest level of the stand- ing tail-water. Theoretically draft-tubes may be 30 feet long ; but 20 feet is as long as is desirable on account of the difficulty of keeping air-tight ; they should be made as short as possible by. placing the turbine at the bottom of the fall. Particulars of the setting recommended for Victor turbines are given below, as an example. Table of Dimensions of Victor Turbine. A. B. C. D. E. F. 1 K. "3 © © pd O © S3 t © u_. co d 4500 35 46* 53 9 59 20 SjdS 5450 40 52* 60* 10 641 5f 22 2« O © 7500 44 48 56* 60* 65* 70* 11 12 67* 74| 5 4 24 26 9380 11700 55 68 80 14 85* 7* 28 19000 63 80* 92 16 96* 7| 1 32 oq'O O t* DIMENSIONS OF TURBINES. Tables of sizes of turbine wheels vary so much under different makers, and are so extensive, as not to permit their insertion here, but through the kindness of Mr. Axel Ekstrom of the General Electric Company I am per- mitted to print the following sheets of curves for the McCormick type turbine and the Pelton impulse wheel. From them may be made deter- minations of dimensions in much shorter time than is necessary by use of tables. 1478 WATEK-POWER. g i- a- LL £ V 5 © 3 ^r-L 1 s * B. 5a o z "^ ^ ^V CO z " ^ .§. "Sw-SA -J uj i "* ^ X^V H >■ « t s ■- v^-5- z i^ ^2^"o - ~M^ . vs v5Xi ^ Z P „ UJ ^r - v s\^^^v. £- "* tr '-&'' \A VNA& r- 3 3 •» a .-J 13^ •g^S^V^S ^ £*"* # ^S£ V^_ z © © O s « S *JS \ \Vv_ I® N "■ 3^ ^M<^ -o ^sX^J, ^SSS °, © T IS^S- ;«\ S SXS vSSS "* ^ %^^- '**, ^SS> >§^Sdz go > sjp Pr ' s!>^5^^; ^55§Sv ""£ vS :^§§g|^ s^ c*^ -^^§g§|s *2 15^ ^i§i|| -£-«- ^3r ' H Irr:==S22s: ^ OdOlO^MooOi)* *» oXi m ^ i H -^-- -^VaHf'^rdj— |— »_*_o_co_c > ■*MW»M^ NN fl ON Gh"''x3" V*"'"'"^- ^^ Vg-0005 ^^N^^- ./ 1 ^^->^-. ^jgjT 1 -S^^S^' © nnLa S S^3 ^^gifSSgjL "T ^S S ^ v, * * —■•*"" £§|222222Z si ^^^^ "V '"3 e -= — ^^E*' g§5^^Z^Z^3!! N I!^ k - V S — "^^^-^'^'^-^' rNN";^ X V -, ^^ t S ^T^^-'^'^'^!>-^^ t~y / s**7 > ~7~f~/ /< 2 ~oopo N v s s, ^^.^ """"^^^ ,3 "^ ^r^^ ^r^* s*^ , ^^^zfz7z£©;t ••t X^ ^s ^^ - •^ ^ "«»*•*'* - ^^" '^"^ 't>'\ZzJ+t^~^ ■ s::; -^. .^ ****** ^^ y „ ''-*£-*/-/ L-L-L*- 2 JL ti ^v s,^^^ *^ * *r ^ . X ^ ^ -* ^ y ^-/ £- z U£ S^^ ^^S ^^^ "^"^^ -" x^ ^^ X*' ^' -*/ -/ /A 5 3 ni H- SA V ^^ ^ V CO ^ ^ ^ ^ ■/ ^ ^--/ ^S-oooo 1 '. \ N v^^ ^ "^N^, *e» ^^ ^ ^ >' / f_ * o! O 1 1 S V N w >^ ^ ; ^ ^ / / > C ' —f o ^ ' "8^ *■' ^ ? > ^, N ^ S S ^ ^^« ^ 7^ l z v^^ ^ ^^ ' *& ^ ^ -V ^ y /^ ° © 1 \ ^ ^ ^ ' -7. >^ "7* h i? oops L "■ ^ ^ UJ "s^ ^ J/ / ^3 ■g'-wjop L ^ _ S ^ / > je > o 1 a ^S ^ ^ r ^- o g^ooo i- * p^ ^ < ^^ 4- uj-O _L u / Q- Q- o 1 I- 1 ^S JrV^ ^> ^ ( =V Zl O gj „ 5 ^ peg ee» « o ; w < 6S co I Fig. 31. DATA ON HYDRAULIC TURBINES. 1479 Fig. 32. 1480 WATER-POWER. THE IMPtliE UlTEB.nUKKL. Mr. Ross E. Browne states that " The functions of a water-wheel, operated by a jet of water escaping from a nozzle, is to convert the energy of the jet, dii3 to its velocity, into useful work. In order to utilize this energy fully, the wheel bucket, after catching the jet, must bring it to rest before dis- charging it, without inducing turbulence or agitation of the particles. This cannot be fully eif ected, and unavoidable difficulties necessitate the loss of a portion of the energy. The principal losses occur as follows : " First : In sharp or angular diversion of the jet in entering, or in its course through the bucket, causing impact, or the conversion of a portion of thei energy into heat instead of useful work. "Second: In the so-called frictional resistance offered to the motion of the water by the wetted surfaces of the buckets, causing also the conver- sion of a portion of the energy into heat instead of useful work. "Third: In the velocity of the water as it leaves the bucket, represent- ing energy which has not been converted into work. 4 Hence, in seeking a high efficiency, there are presented the following con. jiderations : " Lst. The bucket surface at the entrance should be approximately paral- lel (o the relative course of the jet, and the bucket should be curved in such a m inner as to avoid sharp angular deflection of the stream. If, for exam- ple, a jet strikes a surface at an angle and is sharply deflected, a portion of the water is backed, the smoothness of the stream is disturbed, and there results considerable loss by impact and otherwise. 2d, The number of buckets should be small, and the path of the jet in the buqiV.et short ; in other words, the total wetted surface should be small, as the loss by friction will be proportional to this. 44 Jl small number of buckets is made possible by applying the jet tangen- tial lj to the periphery of the wheel. 44 3d. The discharge end of the bucket should be as nearly tangential to the irheel-periphery, as compatible with the clearance of the bucket which follows ; and great differences of velocity in the parts of the escaping waUr should be avoided. In order to bring the water to rest at the dis- charge end of the bucket, it is easily shown mathematically that the velo- city of the bucket should be one-half the velocity of the jet. 44 in ordinary curved or cup bucket will cause the heaping of more or less dea I or turbulent water in the bottom of the bucket. This dead water is sub lequently thrown from the wheel with considerable velocity, and repre- sent s a large loss of energy. 44 The introduction of the wedge in the bucket is an efficient means of av<\ (ding this loss." Wheels of this type are very efficient under high heads of water, and have been used to a great extent in the extreme western parts of the United Stages, where the fall is in hundreds of feet. It is difficult to say at what point of head the efficiency becomes such as to induce the use of some other form of wheel; but at 200 feet head the efficiencies of both impulse and tur- bine will be so much alike that selection must be governed by other factors. T3sts of one of the leading impulse wheels show efficiencies varying from 80 % to 86 % according to head and size of jet. However, many factors besides the efficiency enter into selection of water-wheels, which must be subject to local conditions, and as in most water-power plants, each is r special case by itself, and selection of apparatus best fitted in all ways must govern. SHAFTING. 1481 SHAFTING, PULLEYS, BELTING, ROPE- DRIVING. SHAFTOG. Thurston gives the following formulae for calculating power and size of shafting. H.P. = horse-power transmitted. d = diameter of shaft in inches. r =. revolutions per minute. For head shafts well For lron > HP - = 125 ; d = V r — supported against^ For cold . z, springing. r > lled iron HPt _ ^. ^_ y/^_^£j For line shafting [ For iron > ^ p - = W ; V ~ hangers 8 feet ^ For cold . 3/55 ATP a P art - r'lld iron, JT.P. = %~)d- $ °° Ur ' ^ 55 V r f^ • r7 D «**•",, ?/625~HP. For transmission For iron > H.P. = ^gJ ^ = V r simply, no pul- J F o cold- ^ " */ WlLP. le y s - r'lld iron, H.P.= — — : d = V ^ 3o " r Jones and Laughlin's use the same formulae, with the following excey tions : For line shafts, cold-rolled iron, H.P. r= — -; dz=\ : — -'. 50 ' T r For transmission and for short-counters, ™ , . „ n d 3 r _ . 3 /50 H.P. Turned iron H.P. = — — ; d =. v • 50 ' t r Cold-rolled iron H.P. = — - ; d= V 30 " r Pulleys should be placed as near to bearings as practicable, but cate should be taken that oil does not drip from the box into the pulley. The diameter of a shaft safe to carry the main pulley at the center oi a bay may be found by multiplying the fourth power of the diameter obtained by the formulae above given, by the length of the bay, and dividing the pro- duct by the distance between centers of bearings. The fourth root of tl le quotient will be the required diameter. The following table is based upon the above rule, and is substantially correct : 1482 SHAFTING, PULLEYS, BELTING, ETC. 4) bC «M,£j Is* ■2.3 ^8 in. 2 24 3 3* 4 4* 5 54 6 Diameter of Shaft necessary to carry the Load at the Center of a Bay, which is from Center to Center of Bearings. 24 ft. in. 2§ 24 3 3 ft. 2i 2f 3| 34 4 3* ft. 2| 2| 3i 3f 44 44 5 1ft. in. 2* 54 5 ft. in. 2| 3 ?. 44 5| 6| 6 ft. in. 2| 3 54 5f 6 8 ft. m. 24 3f 4 54 6 64 10 ft. in. 3 3| 4 4' & 5| 6 Should the load be placed near one end of the bay, multiply the fourth power of the diameter of shaft necessary to safely carry the load at the cen- ter of the bay (see above table) by the product of the two ends of the shaft, and divide this product by the product of the two ends of the shaft where the pulley is placed in the center. The fourth root of this quotient will be the required diameter. A shaft carrying both receiving and driving pulleys should be figured as a head-shaft. Deflection of Shafting*. (Pencoyd Iron Works.) As the deflection of steel and iron is practically alike under similar con- ditions of dimensions and loads, and as shafting is usually determined by its transverse stiffness rather than its ultimate strength, nearly the same dimensions should be used for steel as for iron. For continuous line-shafting it is considered good practice to limit the deflection to a maximum of T £ 3 of an inch per foot of length. The weight of bare shafting in pounds = 2.6 d 2 L = W, or when as fully loaded with pulleys as is customary in practice, and allowing 40 lbs. per inch of width for the vertical pull of the belts, experience shows the load in pounds to be about 13 d 2 L = W. Taking the modulus of transverse elasticity at 26,000,000 lbs., we derive from authoritative formulae the following: L=^/S73 d 2 , d = y — , for bare shafting; L = J/ 175 d 2 , d = y — , for shafting carrying pulleys, etc.; L being the maximum distance in feet between bearings for continuous shafting subjected to bending stress alone, d r= diam. in inches. The torsional stress is inversely proportional to the velocity of rotation, while the bending stress will not be reduced in the same ratio. It is there- fore impossible to write a formula covering the whole problem and suffi« ciently simple for practical application, but the following rules are correct within the range of velocities usual in practice. For continuous shafting so proportioned as to deflect not more than ^ of an inch per foot of length, allowance being made for the weakening effect of key-seats, f = \I WR.P. 1 ,L= ^/720c?2 for bare shafts ; SHAFTING. 1483 7 70 H.P. , L =y 140d 2 , for shafts carrying pulleys, etc. d = diam. in inches, L rr length in feet, r =z revols. per minute. The following table (by J. B. Francis) gives the greatest admissible dis- tances between the bearings of continuous shafts subject to no transverse strain, except from their own weight. Distance between Bearings in ft. f * \ Diam. of Shaft, Wrought-iron Steel in inches Shafts. Shafts 2 15.46 15.89 3 17.70 18.19 4 19.48 20.02 5 20.99 21.57 Distance between Bearings in ft. t % Diam. of Shaft, Wrought-iron Steel in inches. Shafts. Shafts 6 22.30 22.92 7 23.48 24.13 8 24.55 25.23 9 25.53 26.24 The writer prefers to apply a formula in all cases rather than use tables, as shafting is nearly always one-sixteenth inch less in diameter than the sizes quoted. The following tables are made up from the formulae first given in this chapter. Horse-Power Transmitted by Turned Iron Shafting-* As Prime Mover or Head Shaft well Supported by Bearings. s Revolutions per Minute. 60 80 100 125 150 175 200 225 250 275 300 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. If 2.6 3.4 4.3 5.4 6.4 7.5 8.6 9.7 10.7 11.8 12.9 2 3.8 5.1 6.4 8 9.6 11.2 12.8 14.4 16 17.6 19.2 2J 5.4 7.3 8.1 10 12 14 16 18 20 22 24 2£ 7.5 10 12.5 15 18 22 25 28 31 34 37 2f 10 13 16 20 24 28 32 36 40 44 48 3 13 17 20 25 30 35 40 45 50 55 60 3+ 16 22 27 34 40 47 54 61 67 74 81 3* 20 27 34 42 51 59 68 76 85 93 102 3f 25 33 42 52 63 73 84 94 105 115 126 4 30 41 51 64 76 89 102 115 127 140 153 4* 43 58 72 90 108 126 144 162 180 198 216 5 60 80 100 125 150 175 200 225 250 275 300 5£ 80 106 133 166 199 233 266 299 333 366 400 Approximate Centers of Bearing's for Wrought Iron Li n«' Shafts Carrying- a fair Proportion of Pulleys. Shaft, Diameter Inches . . n If 2 2* 2£ 2| 3 3* 4 4* c. to c. Bearings — Feet . . 7 n 8 8* 9 9i 10 11 12 13 Shaft, Diameter Inches . . 5 5* 6 6* 7 16 17 8 18 9 19 10 c. to c. Bearings — Feet . . 13* 14 15 15* 20 1484 SHAFTING, PULLEYS, BELTING, ETC. Line-shafting, Bearings 8 ft. Apart. i 5 Revolutions per Minute. 100 125 150 175 200 225 250 275 300 H.P. 325 H.P. 350 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. If 6 7.4 8.9 10.4 11.9 13.4 14.9 16.4 17.9 19.4 20.9 13 7.3 9.1 10.9 12.7 14.5 16.3 18.2 20 21.8 23.6 25.4 2 8.9 11.1 13.3 15.5 17.7 20 22.2 24.4 26.6 28.8 31 2i 10.6 13.2 15.9 18.5 21.2 23.8 26.5 29.1 31.8 34.4 37 2i 12.6 15.8 19 22 25 28 31 35 38 41 44 2| 15 18 22 26 29 33 37 41 44 48 52 2* 17 21 26 30 34 39 43 47 52 56 60 2| 23 29 34 40 46 52 58 64 69 75 81 3 30 37 45 52 60 67 75 82 90 97 105 3i si 38 47 57 66 76 85 95 104 114 123 133 47 59 71 83 95 107 119 131 143 155 167 58 73 88 102 117 132 146 162 176 190 205 4 71 89 107 125 142 160 178 196 213 231 249 POWER TRANSMISSION ONLY. a Revolutions per Minute. 100 125 150 175 200 233 267 300 333 367 400 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. n 6.7 8.4 10.1 11.8 13.5 15.7 17.9 20.3 22.5 24.8 27.0 if 8.6 10.7 12.8 15 17.1 20 22.8 25.8 28.6 31.5 34.3 if 10.7 13,4 16 18.7 21.5 25 28 32 36 39 43 i* 13.2 16.5 19.7 23 26.4 31 35 39 44 48 52 2 16 20 24 28 32 37 42 48 53 58 64 2* 19 24 29 33 38 44 51 57 63 70 76 2i 22 28 34 39 45 52 60 68 75 83 90 2| 27 33 40 47 53 62 70 79 88 96 105 2* 31 39 47 54 62 73 83 93 104 114 125 2f 41 52 62 73 83 97 111 125 139 153 167 3 54 67 81 94 108 126 144 162 180 198 216 »l 68 86 103 120 137 160 182 205 228 250 273 3£ 85 107 128 150 171 200 228 257 285 313 342 Horse-power Transmitted by Cold-rolled Iron Shafting-. AS PRIME MOVER OR HEAD SHAFT "WELL SUPPORTED BY BEARINGS. i Revolutions per Minute. 3 60 80 100 125 150 175 200 225 250 275 300 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H 2.7 3.6 4.5 5.6 6.7 7.9 9.0 10 11 12 13 If 4.3 5.6 7.1 8.9 10.6 12.4 14.2 16 18 19 21 2 6.4 8.5 10.7 13 16 19 21 24 26 29 32 21 9 12 15 19 23 26 30 34 38 42 46 12 17 21 26 31 36 41 47 52 57 62 2f 16 22 27 35 41 48 55 62 70 76 82 3 21 29 36 45 54 63 72 81 90 98 108 27 36 45 57 68 80 91 103 114 126 136 3£ 34 45 57 71 86 100 114 129 142 157 172 3f 42 56 70 87 105 123 140 158 174 193 210 4 51 69 85 106 128 149 170 192 212 244 256 4* 73 97 121 151 182 212 243 273 302 333 364 SHAFTING. 1485 LINE-SHAFTING, BEARINGS 8 FT, APART. i S Revolutions per Minute. 100 125 150 175 200 225 250 275 300 325 350 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. n 6.7 8.4 10.1 11.8 13.5 15.2 16.8 18.5 20.2 21.9 23.6 if 8.6 10.7 12.8 15 17.1 19.3 21.5 23.6 25.7 28.9 31 if 10.7 13.4 16 18.7 21.5 24.2 26.8 29.5 32.1 34.8 39 i* 13.2 16.5 19.7 23 26.4 29.6 32.9 36.2 39.5 42.8 46 2 16 20 24 28 32 36 40 44 48 52 56 2| 19 24 29 33 38 43 48 52 57 62 67 22 28 34 39 45 50 56 61 68 74 80 2f 27 33 40 47 53 60 67 73 80 86 94 1 31 39 47 54 62 69 78 86 93 101 109 41 52 62 73 83 93 104 114 125 135 145 3 54 67 81 94 108 121 134 148 162 175 189 3* 68 86 103 120 137 154 172 188 205 222 240 3i 85 107 128 150 171 192 214 235 257 278 300 POWER TRANSMISSION AND SHORT COUNTERS. i Revolutions pei Minute. s 100 125 150 175 200 233 267 300 333 367 400 Ins. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H.P. H 6.5 8.1 9.7 11.3 . 13 15.2 17.4 19.5 21.7 23.9 *26 H 8.5 10.7 12.8 15 17 19.8 22.7 25.5 28.4 31 34 H 11.2 14 16.8 19.6 22.5 26 30 33 37 41 45 if 14.2 17.7 21.2 24.8 28.4 33 38 42 47 52 57 if 18 22 27 31 35 41 47 53 59 65 71 n 22 27 33 38 44 51 58 65 72 79 87 2 26 33 40 46 53 62 71 80 88 97 106 2£ 2* 32 40 47 55 63 73 84 95 105 116 127 38 47 57 66 76 89 101 114 127 139 152 2§ 44 55 66 77 88 103 118 133 148 163 178 2| 52 65 78 91 104 121 138 155 172 190 207 2* 69 84 99 113 138 161 184 207 231 254 277 3 90 112 135 157 180 210 240 270 300 330 360 Hollow Shaft*. Let d be the diameter of a solid shaft, and d x d 2 the external and internal diameters of a hollow shaft of the same material. Then the shafts will be of equal torsional strength when d 3 = * 2 • A 10-inch hollow shaft with internal diameter of 4 inches will weigh 16 % less than a solid 10-inch shaft, but its strength will be only 2.56 % less. If the hole were increased to 5 inches diameter the weight would be 25% less than that of the solid shaft, and the strength 4.25 % less. Table for laying* Out Shafting*. The table on the following page is used by Wm. Sellers & Co. for the lay- ing out of shafting. 1486 SHAFTING, PULLEYS, BELTING, ETC. Double Cone- Vise Coupl'g. •sgqoiri J0}9OIVI(X oc a> O •■« c^rt^t-??©: ©> •89qoui 4 q^Su97 iSo t- oo*ci o «-* sSlteSss > s m 1 sS , S , Sacl§5| , ^< sSfsssJs&IT ii S # i $ - SstsSssjITa - *i «_h« p^«^ ?1 1 cone bored for smaller nomim §r m cTw « « "£* TrTdT 1. •*» -4n -** •m — r< c^ ci m <*■ ia * cT©*^^^ §r -*» -*» 8 a s Nominal Size of 1st Shaft, ins. 2Sl2?e* >u . >** . ft Ph Form. 1 Form. 2 Form. 3 Form. 4 Double. 3 7 2 7/ single 3 ^"S H.P. = H.P. = H.P. = H.P. = Belt Belt. c*> 3 wv wv wv wv H.P. = H(»S *-h © fl ©©£ cc ©^ 550 1100 1000 733 wv 513* Laced. Riveted. 10 600 50 1.09 .55 .60 .82 1.17 .73 1.14 30 1200 100 2.18 1.09 1.20 1.64 2.34 1.54 2.24 30 1800 150 3.27 1.64 1.80 2.46 3.51 2.25 3.31 40 2400 200 4.36 2.18 2.40 3.27 4.68 2.90 4.33 50 3000 250 5.45 2.73 3.00 4.09 5.85 3.48 5.26 60 3600 300 6.55 3,27 3.60 4.91 7.02 3.95 6.09 70 4200 350 7.63 3.82 4.20 5.73 8.19 4.29 6.78 80 4800 400 8.73 4.36 4.80 6.55 9.36 4.50 7.36 90 5400 450 9.82 4.91 5.40 7.37 10.53 4.55 7.74 100 6000 500 10.91 5.45 6.00 8.18 11.70 4.41 7.96 ilO 6600 550 . . . • • • 4.05 7.97 120 7200 600 . . . 3.49 7.75 "Width of Belt for a given Horse-Power. The width of belt required for any given horse-power may be obtained by transposing the formulae for horse-power so as to give the value of w. Thus : From formula (1), w =r From formula (2), w = From formula (3), w — From formula (4), w = From formula (5),* w = * For double belts. 550 H. P. 9.17 H. P. 2101 H. P. 275 H. P. v V ~ d X rpm ~ L X rpm 1100 H.P. 18.33 H.P. 4202 H. P. 530 H. P. v V ~ d X rpm ~ L X rpm 1000 H. P. 16.67 H. P. 3820 H. P. 500 H.P. v V ~ d x rpm ~ L X rpm 733 H. P. 12.22 H. P. 2800 H. P. 360 H.P. v V d X rpm ~~ L < X rpm 513 H. P. 8.56 H. P. 1960 H. P. 257 H. P. v V d X rpm ~ L X rpm BELTING. 1489 \ x 3.14161 -f [2 x distance length of Belt. Approximate rule ; two pulleys I ( ^ — between centers] = length of belt. Leng-th of Belt in Roll. Outside diameter roll in inches 4- diameter hole x number turns x .1309 = length of belt in inches for double belt. Weight of Belt (approximate). Length in feet x width in inches __ double belts. 13 weight of single belt. Divide by 8 for Horse-Power Transmitted by Tig-lit. Double Endless JLeather Belting 1 . (Buckley.) Width, Inches. 4 6 8 10 12 14 16 18 20 22 24 B 2000 14 22 29 36 43 50 58 65 72 80 87 H 2400 17 26 35 44 52 60 70 78 88 96 105 £ 2800 20 30 40 51 61 71 81 91 102 112 122 ft 3000 22 33 44 54 65 76 87 98 108 120 131 % 3500 25 38 50 63 76 89 101 114 127 140 153 © 4000 29 43 58 73 87 101 116 131 145 160 174 , 4500 •- 5000 32 49 65 82 98 114 131 147 163 180 196 36 55 73 91 109 127 145 163 182 200 218 % 5500 40 60 80 100 120 140 160 180 200 220 240 © 6000 ft 0Q 44 65 87 109 130 153 175 200 218 240 260 (Speed x width -^ 550 = horse-power, light, double.) (Horse-power x 550 -f- speed = width, light, double.) Horse-Power Transmitted by Heavy, Double Endless Leather Belting*. Width, Inches. & 2000 S 2400 fe 2800 ft 3000 t 3500 © 4000 a 4500 ■3 5000 % 5500 © 6000 ft 18 27 36 21 31 42 24 36 48 27 40 53 30 45 60 35 52 70 38 59 78 43 66 87 48 72 96 52 78 104 10 43 53 61 65 75 88 98 110 120 122 12 51 62 73 78 91 104 118 130 144 153 14 60 72 85 90 106 121 137 152 168 183 16 70 83 96 104 121 139 157 174 192 210 18 80 94 109 118 137 157 176 196 216 240 20 86 105 122 129 152 174 196 218 240 262 22 96 115 135 144 168 192 216 240 264 283 24 104 120 146 157 184 209 235 262 288 312 (Speed x width ~- 460 = horse-power, heavy, double.) (Horse-power x 460 -J- speed =; width, heavy, double.) 1490 SHAFTING, PULLEYS, BELTING, ETC. ROPE IMIi V IX*. Cr= Circumference of rope in inches. D — Diameter of pulley in feet. i?= Revolutions per minute. Horse-power of Rope : CxDxR _ 200 = H.P. or, Half the diameter of rope multiplied by the hundreds of feet per min- ute traveled. (L. I. Seymour.) Breaking strength of manila rope in pounds = C 2 X coefficient. The coefficient varies from 900 for £-inch to 700 for 2-inch diameter rope. The following is a reliable table prepared by T. Spencer Miller, M.E. (See En- gineering News, December 6, 1890.) Diameter. Circumference. Ultimate Strength. Coefficient. h n 2,000 900 1 2 3,250 845 1 2J 4,000 820 | 2| 6,000 790 1 3 7,000 780 ?! 3* 9,350 765 3f 10,000 760 If 1 13,500 745 ll 15,000 735 If 5 18,200 725 If 5| 21,750 712 2 6 25,000 700 This table was compiled by averaging and graduating results of tests at the Watertown Arsenal and Laboratory of Riehle Brothers, in Philadelphia. Weight of manila rope in pounds per foot =z .032 (Circumference in inches) 2 . (C. W. Hunt.) or, diameter of rope in inches squared = weight in pounds per yard ap- proximately. The coefficient of friction on a rope working on a cast-iron pulley z= 0.28 ; when working in an ungreased groove it is increased about three times, or from 0.57 to 0.84. If the pulleys are greased, the coefficient is reduced about one-half. It has been found by experiment that a rope 6 inches cir- cumference in a grooved pulley possesses four times the adhesive resistance to slipping, exhibited by a half-worn, ungreased 4-inch single belt. The length of splice should be 72 times the diameter of rope. The strength of a rope containing a properly made " long splice" was found to be 7,000 pounds per square inch of section. A mixture of molasses and plumbago makes an excellent dope for trans- mitting ropes. Grease and oils of all kinds should be kept from transmis- sion ropes, since, as a rule, they are injurious. Following is another formula for horse-power of manila rope : jrp _ (T -QV ' '~ 33000 ' in which H.P. is the horse-power transmitted by one rope, V the velocity in feet per minute, T the maximum working stress, and Cthe centrifugal tension, so that (T — C) is the net tension available for the transmission of power. Taking the total maximum stress at 200d 2 and allow 20 % of this for slack side tension, we have T t) = IGOd 2 , so that H.P. =■ (16 d 2 — C) V 33,000 A table has been calculated by this rule, giving the horse-power per rop«, transmitted at various speeds. ROPE DRIVING. 1491 C = Centrifugal Tension in Manila Ropes ~ Pounds. Velocity of Rope in ft. per Min. Nominal Diameter of Rope tn Inches. i 2" f 1 3 1 li n If n If If 2 1000 0.7 1.1 1.5 2.1 2.7 3.4 4.3 5.1 6.2 7.2 8.3 11 1500 1.5 2.4 3.4 4.7 6.2 7.6 9.7 11 13 16 18 25 2000 2.7 4.3 6.1 8.2 11 13 17 20 24 28 33 44 2500 4.3 6.7 9.6 13 17 21 27 32 38 45 52 69 3000 6.2 9.7 13 18 24 30 39 45 55 64 74 100 3500 8.4 13 . 19 25 34 42 53 63 75 89 102 136 4000 11 17 24 33 44 54 69 82 98 116 133 177 4500 14 22 31 42 55 69 87 103 125 146 168 223 5000 17 27 39 52 69 86 109 129 156 183 210 275 5500 21 33 47 63 83 104 132 156 189 221 254 332 6000 24 39 56 75 99 125 157 188 225 257 303 396 6500 39 45 65 88 116 145 183 217 261 307 353 462 Horse-Power of Manila Ropes. Velocity of Rope. Ft. per Min. Nominal Diameter of Rope in Inches. h f I 3 1 li li If H If If 2 2000 2.25 3.51 5.14 6.84 9.08 11.5 14.0 17.0 20.3 23.8 27.5 S6.1 2100 2.35 3.67 5.27 7.15 9.40 11.8 14.7 17.8 21.1 24.8 28.8 37.6 2200 2.45 3.82 5.48 7.45 9.80 12.3 15.3 18.5 22.0 25.9 30.0 39.2 2300 2.55 3.98 5.71 7.75 10.2 12.8 15.9 19.3 22.9 26.9 31.2 40.8 2400 2.62 4.10 5.89 7.98 10.5 13.2 16.4 19.8 23.6 27.7 32.2 42.0 2500 2.70 4.21 6.05 8.21 10.8 13.6 16.8 20.4 24.3 28.5 33.1 43.2 2600 2.78 4.33 6.21 8.43 11.1 14.0 17.3 21.0 25.0 29.3 34.0 44.4 2700 2.85 4.45 6.39 8.67 11.4 14.4 17.8 21.5 25.6 30.5 35.0 45.6 2800 2.94 4.59 6.59 8.93 11.75 14.8 18.3 22.2 26.4 31.0 36.0 47.0 2900 3.00 4.68 6.73 9.13 12.0 15.1 18.7 22.7 27.0 31.6 36.8 48.0 3000 3.06 4.78 6.87 9.32 12.3 15.4 19.1 23.2 27.6 32.3 37.6 49.1 3100 3.12 4.87 7.01 9.50 12.5 15.7 19.5 23.6 28.2 33.0 38.3 50.0 3200 3.18 4.97 7.14 9.70 12.7 16.0 19.9 24.0 28.7 33.7 39.0 51.0 3300 3.25 5.07 7.27 9.89 13.0 16.3 20.3 24.5 29.2 34.3 39.8 52.0 3400 3.30 5.15 7.39 10.0 13.2 16.6 20.6 25.0 29.7 34.8 40.4 52.8 3500 3.35 5.22 7.50 10.2 13.4 16.9 20.9 25.3 30.1 35.4 41.0 53.6 3600 3.40 5.30 7.61 10.3 13.6 17.1 21.2 25.7 30.6 35.9 41.6 54.4 3700 3.44 5.36 7.70 10.4 13.7 17.3 21.5 26.0 30.0 36.3 42.1 55.0 3800 3.46 5.40 7.76 10.5 13.8 17.4 21.6 26.2 31.1 36.6 42.4 55.4 3900 3.49 5.45 7.81 10.6 13.9 17.6 21.8 26.4 31.4 36.9 42.7 55.8 4000 3.51 5.49 7.86 10.6 14.0 17.7 21.9 26.5 31.6 37.1 43.0 56.1 4100 3.53 5.52 7.92 10.7 14.1 17.8 22.0 26.7 31.8 37.3 43.2 56.4 4200 3.55 5.54 7.95 10.8 14.2 17.9 22.1 26.8 31.9 37.5 43.4 56.8 4300 3.56 5.55 7.98 10.8 14.2 17.9 22.2 26.9 32.0 37.6 43.6 56.9 4400 3.57 5.56 7.99 10.8 14.2 18.0 22.2 27.0 32.1 37.6 43.6 57.0 4500 3.56 5.55 7.96 10.8 14.2 17.9 22.2 26.9 32.0 37.6 43.5 56.9 4600 3.55 5.54 7-95 10.8 14.2 17.9 22.1 26.8 31.9 37.5 43.4 56.8 4700 3.53 5.50 7.90 10.7 14.1 17.8 22.0 26.6 31.7 37.2 43.1 56.4 4800 3.51 5.48 7.86 10.7 14.0 17.7 21.9 26.5 31.6 37.1 43.0 56.2 4900 3.49 5.45 7.81 10.6 13.9 17.6 21.8 26.4 31.4 36.9 42.7 55.8 5000 3.45 5.38 7.73 10.5 13.8 17.4 21.5 26.1 31.0 36.4 42.2 55.2 5100 3.43 5.35 7.67 10.4 13.7 17.2 21.3 25.9 30.8 36.2 41.9 54.8 5200 3.38 5.26 7.56 10.2 13.5 17.0 21.0 25.5 30.4 35.6 41.3 54.0 5300 3.34 5.20 7.47 10.1 13.3 16.8 20.8 25.2 30.0 35.2 40.8 53.4 5400 3.28 5.11 7.34 9.95 13.1 16.5 20.4 24.8 29.4 34.6 40.1 52.5 5500 3.21 5.00 7.20 9.75 12.8 16.2 20.0 24.2 28.9 33.9 39.3 51.4 6000 2.78 4.33 6.21 8.43 11.1 14.0 17.3 21.0 25.0 29.3 34.0 44.4 6500 2.17 3.38 4.85 6.60 8.6 10.9 13.5 16.4 19.5 22.9 26.5 34.7 1492 SHAFTING, PULLEYS, BELTING, ETC. HORSE POWER ^f ■? 3 f A / 4 / / J / / / / / / / / / / I / / j 1 ) / l I / 1 \ I \ \ \ \ \ \ \ ROPE DRIVING HORSE POWER OF MANILLA ROPE AT VARIOUS SPEEDS \ \ V \ \ ,iN » V A % * k f V N/- rS k v \ i fe > o 8 > 5 3 OOOCO^NOOO^^OJ Fig. 36. ROPE DKIVING. 1493 Horse-Power of " Stevedore " Transmission Rope at Various Speeds. In this table the effect of the centrifugal force has been taken into con- sideration, and the strain on the fibers of the rope is the same at all speeds when transmitting the horse-power given in the table. When more than one rope is used, multiply the tabular number by the number of the ropes. At a speed of 8,400 per minute the centrifugal force absorbs all the allowable tension the rope should bear, and no power will be transmitted. Table of the Horse-Po (Hunt's '. wer of Transmission Rope. Formula.) «M O u P ft Speed of the Rope in Feet per Minute. a q 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 6,000 7,000 8,400 in h 1.45 1.9 2.3 2.7 3. 3.2 3.4 3.4 3.1 2.2 .0 .20 5 8 2.3 3.2 3.6 4.2 4.6 5.0 5.3 5.3 4.9 3.4 .0 .25 t 3.3 4.3 5.2 5.8 6.7 7.2 7.7 7.7 7.1 4.9 .0 .30 I 4.5 5.9 7.0 8.2 9.1 9.8 10.8 10.7 9.3 6.9 .0 .36 1 5.8 7.7 9.2 10.7 11.9 12.8 13.6 13.7 12.5 8.8 .0 .42 H 9.2 12.1 14.3 16.8 18.6 20.0 21.2 21.4 19.5 13.8 .0 .54 H 13.1 17.4 20.7 23.1 26.8 28.8 30.6 30.8 28.2 19.8 .0 .60 if 18. 23.7 28.2 32.8 36.4 39.2 41.5 41.8 37.4 27.6 .0 .72 2 23.2 30.8 36.8 42.8 47.6 51.2 54.4 54.8 50. 35.2 .0 .84 For a temporary installation w r hen the rope is not to be long in use, it might be advisable to increase the work to double that given in the tables. Slip of Ropes and Relts. (W. W. Christie.) Some French trials, with constant resistance, the power expended and slip in several modes of transmission was as follows : Ropes, 158.54 gross h.p., Slip, 0.33 per cent. Cotton belt, 159.67 " " 0.78 " Leather " 158.84 " " 0.96 " " " 160.23 " " 0.78 " Stated in percentage value, the results were : Ropes, 100.00 gross power, Slip, 0.100. Cotton belt, 100.87 " " 0.237. Leather " 100.37 " " 0.292. " " 101.07 " " 0.237. ( 1494 SHAFTING, PULLEYS, BELTING, ETC. Manila Cordage. Tarred Hemp. Size, Cir- Size, Weight of Feet in Breaking Strain Weight of cumfer'ce. Diameter. 100 one of New Ropes. 100 Inches. Inches. Fathoms. Pound. Pounds. Fathoms. For Ropes in use li 3 8 31 20 deduct £ from 40 1* 1 44 14 these figures, for 55 If 9 l 6 60 10 chafing, etc. 75 2 79 n 3000 100 21 1 99 6 4000 125 2§ tl 122 5 5000 155 2| i 146 4 6000 190 3 i 176 3§ 7000 225 31 H* 207 3 8500 265 3* i£ 240 2i 9500 300 3| ij 275 % 11000 355 4 1 s if 305 2 12500 405 4i 355 If 14000 455 3 If 395 16000 500 5 If 490 H 20000 630 5* If 595 l 24000 750 6 2 705 10 in. 27000 910 6£ 2J 825 8i 31500 1050 7 2i 960 n 37000 1235 n 2f 1100 ei 42500 1400 8 2f 1255 ii 4850o 1600 8* n 1415 5 54500 1820 9 3 1585 4* 61500 2050 Hawser laid will weigh £ less. Notes on the Uses of Wire Rope. (Roebling.) Two kinds of wire rope are manufactured. The most pliable variety con- tains 19 wires in the strand, and is generally used for hoisting and running rope. For safe working load allow £ or £ of the ultimate strength, according to speed, so as to get good wear from the rope. Wire rope is as pliable as new hemp rope of the same strength ; but the greater the diameter of the sheaves the longer wire rope will last. Experience has proved that the wear increases with the speed. It is, therefore, better to increase the load than the speed. Wire rope must not be coiled or uncoiled like hemp or manila — all untwisting or kinking must be avoided. In no case should galvanized rope be used for running. One day's use scrapes off the zinc coating. Table of Strains Produced lij Load* on Inclined Planes. Strain in Lbs. on Elevation in 100 Ft. Strain in Lbs. on Elevation in 100 Ft. Rope from a Load Rope from a Load of 1 Ton. of 1 Ton. Ft. Deg. Ft. Deg. 10= 5£ 212 90 = 42 1347 20=1U 30=16f 404 100 = 45 1419 586 110 = 47| 1487 40=21| 50 = 26| 754 120 = 50J 1544 905 130 = 52| 1592 60=31 1040 140 = 54* 1633 70 = 35 1156 150 = 56J 1671 80 = 38§ 1260 160 = 58 1703 WIRE ROPE. 1495 Table of Transmission of Power by Wire Ropes. Showing necessary size and speed of wheels and rope to obtain any de- sired amount of power. (Roebling.) Diam. of Wheel in Ft. No. of Rev- Diam. of Horse- Diam. of olutions. Rope. Power. Wheel in Ft. 80 1 3.3 10 . 100 1 4.1 120 1 5. 140 1 5.8 80 ft 6.9 11 100 & 8.6 120 A 10.3 140 T 7 8 12.1 80 1 10.7 12 100 £ 13.4 120 \ 16.1 140 \ 18.7 80 9 T5 16.9 13 100 ft 21.1 120 9 15 25.3 80 f 22. 14 100 f 27.5 120 1 33. 80 § 41.5 15 100 § 51.9 120 f 62.2 No. of Rev- olutions. 80 100 120 140 100 120 140 80 100 120 140 80 100 120 80 100 120 80 100 120 Diam. of Rope. Horse- Power. 58.4 73. 87.6 102.2 75.5 94.4 113.3 132.1 99.3 124.1 148.9 173.7 122.6 153.2 183.9 148. 185. 222. 217. 259. 300. Note. For list of transmission ropes, see page 1325. The drums and sheaves should be made as large as possible. The mini- mum size of drum is given in a column in table. It is better to increase the load than the speed. Wire rope is manufactured either with a wire or a hemp center. The latter is more pliable than the former, and will wear better where there is short bending. The weight of rope with wire center is about 10 per cent more than with hemp center. 1496 CHAINS. €HAIH[. The size of chain is determined by the size of the stock used in making the links. The strength of the iron always used for chains is from 41,000 to 55,000 lbs. tensile strength per square inch. Coil Chain. (John C. Schmidt & Co., York, Pa.) Size of Links Av. Weight Proof Size of Links Av. Weight Proof Iron per Foot. per 100 Load in Iron per per 100 Load in in Ins. Ft. in Lbs. Lbs. in Ins. Foot. Ft. in Lbs. Lbs. 3-16 13 45 600 1-2 8 225 7,000 1-4 12 75 1,400 9-16 7 320 9,000 5-16 11 120 2,500 5-8 6 400 11,000 3-8 10 150 4,000 3-4 5* 590 16,000 7-16 9 200 5,000 7-8 5 770 22,000 Short I*i nk Chains. Proof Tests Adopted November 11, 1896. (Jones & Laughlins, Limited.) Size. (Ins.) Proof. (Lbs.) BB Crane. (Lbs.) BBB Crane. (Lbs.) Average Weight per Foot. (Lbs.) * 700 770 900 .5 1,200 1,320 1,500 .9 t 2,500 2,750 3,200 1.22 3,500 3,850 4,425 1.6 t 4,800 5,280 6,100 2.0 6,200 6,820 7,850 2.5 A 7,800 8,580 9,870 3.2 I 9,600 10,560 12,150 4.2 ¥ 11,500 12,650 14,550 5.0 13,800 15,180 17,475 5.9 ? 16,200 17,820 20,500 6.7 18,800 20,680 23,780 7.9 if 21,500 23,650 27,200 9.0 24,600 27,100 31,200 10.2 li- 26,300 28,930 33,300 11.4 29,500 32,450 37,300 12.7 lt 33,000 36,300 41,750 14.2 36,500 40,150 46,175 15.8 it 40,000 44,000 50,600 17.2 44,000 48,400 55,660 18.8 it 48,200 53,000 60,950 20.4 52,500 57,750 66,400 22.2 it 57,000 62,700 72,100 24.0 61,700 67,870 78,050 26.7 W 66,500 73,150 84,120 28.5 71,600 78,760 90,575 31.0 Safe working load should be about one-half of proof test. The breaking strain is about double the proof test. LUBRICATION. 1497 IlBKItlTIOX. When two bodies are compelled to move, one upon the other, the resist- ance encountered is called friction, of which we have three kinds : rolling and sliding of solids, and fluid friction of liquids and gases. The reduction of friction and its consequent generation of heat is accom- plished to a large extent by the use of lubricants. Thurston says the characteristics of an efficient lubricant must be : 1. Enough " body," or combined capillarity and viscosity to keep the sur- faces between which it is interposed from coming in contact under maxi- mum pressure. 2. The greatest fluidity consistent with the preceding requirements. 3. The lowest possible co-efficient of friction under the conditions of actual use, i.e., the sum of the two components, solid and fluid friction, should be a minimum. 4. A maximum capacity for receiving, transmitting, storing, and carrying away heat. 5. Freedom from tendency to decompose, or to change in composition by gumming or otherwise, on exposure to the air while in use. 6. Entire absence of acid or other properties liable to produce injury of materials or metals with which they may be brought in contact. 7. A high temperature of evaporization and of decomposition and a low temperature of solidification. 8. Special adaptation to the conditions as to speed and pressure of rubbing surfaces under which the unguent is to be used. 9. It must be free from grit and all foreign matter. All Animal or Vegetable Oils eventually decompose, and become gummy, and retard the speed of any machine to which they may be applied. mineral Oils — which are used in steam and electrical engineering— do not absorb oxygen, and do not take fire spontaneously, as do the animal and vegetable oils. Greases have their proper place, as in railroad car axles, and in cups feeding journals that do not require lubrication until a certain predeter- mined temperature has been reached, for which the grease to be used is suited. Vegetable Oils should not be used in anyplace from which there is any prospect of their being taken to the inside of a steam boiler, as they materially encourage corrosion and pitting of boiler shells. Weigrht of Oil per Gallon. The Pennsylvania Railroad specifica- tions call for these approximate weights : Lard oil, tallow oil, neatsfoot oil, bone oil, colza oil, mustard-seed oil, rape-seed oil, paraffin oil, 500 degree fire test oil, engine oil, and cylinder lubricant, 7£ pounds per gallon. Well oil and passenger car oil 7.4 lbs. per gallon. Navy sperm oil 7.2 " " " Signal oil , , 7.1" " " 300 degree burning oil 6.9 " " " 150 degree burning oil 6.6 " " " In many of the large power plants the lubrication of a large proportion of the bearings is controlled by a system which pumps the oil through pipes to bearings, and after its use, it is drained to a central point there to be filtered, and foreign matter eliminated, and then used over again. Lubrication is more apt to be overdone than to be neglected to damage of machinery. ( 1498 LUBRICATION. Best & ub ri cants for Different Purposes. (Thurston.) Low temperatures, as in rock drij driven by compressed air . . Very great pressures, slow speed Heavy pressures, slow speed . Heavy pressures, high speed . Light pressures, high speed Ordinary machinery .... Steam cylinders Watches and other delicate mech- anisms I Light mineral lubricating oils. ( Graphite, soapstone and other solid ( lubricants. ( The above, lard and tallow and other \ greases. ( Sperm-oil, castor-oil, and heavy ( mineral oils. ( Sperm, refined, petroleum, olive, ( rape, cotton-seed oils. !Lard oil, tallow oil, heavy mineral oils, and the heavier vegetable oils. Heavy mineral oils, lard, tallow. ( Clarified sperm, neatsfoot, porpoise, < olive, and light mineral lubricat- ( ing oils. For mixture with mineral oils, sperm is best ; lard is much used ; olive and cotton-seed oils are good. After making a series of exposure tests to ascertain the efficiency of lead and zinc paints, G. R. Henderson, N. & W. Railroad, reaches the following conclusions. Tin. — The best results were obtained with the first coat white lead, and second coat, white zinc. The second coating of zinc gave generally the best results, and the second coating of lead the most. Galvanized Iron. — The same remarks apply to galvanized iron as given for tin. Sheet Iron. — The mixture of one-third white zinc and two-thirds white lead, for both coats, gave the best results on this material, and, in general, the zinc paint gave better results than the lead paints. Poplar. — The second coats of zinc showed up well on poplar, no matter whether the priming coats were Avhite lead or white zinc, or mixed lead and zinc. The lead second coating showed up the most on this material, but in each case where the second coat was of zinc, totally or partially, the paint was in a perfect condition. White Pine. — The same remarks apply to white pine as to poplar. Yellow Pine. — This material seems to be difficult to properly treat with paints ; the best results were obtained with the first coat of lead, and the second coat of lead and zinc mixed. Where the first coat was of lead and zinc mixed or entirely of zinc, the results were poor throughout, which seems to indicate that as a general thing the lead is better for priming on this material/ Conclusion. — Lead priming and zinc coating are generally good for tin, galvanized iron, poplar and white pine. Sheet iron shows up best with both coats of mixed paints. Yellow pine appeared best with the first coat of lead and the second coat of lead and zinc mixed. Comparing the materials which were painted, we find that, generally, pop- lar retains the paint better than white pine; and would therefore, be pre- ferred for siding on buildings, etc. Yellow pine seems to be the worst of all for this purpose. Black iron as a whole retains the paint better than either tin or galvanized iron. MISCELLANEOUS TABLES. 1499 MISCELLANEOUS TABLES. WEIGHTS A!¥D MEASl T RES. Measure of Capacity. Gallon. — The standard gallon measures 231 cubic inches, and contains 8.3388822 pounds avoirdupois = 58372.1757 grains Troy, of distilled water, at its maximum density 39.83° Fahrenheit, and 30 inches barometer height. Bushel. —The standard bushel measures 2150.42 cubic inches =77.627413 pounds avoirdupois of distilled water at 39.83° Fahrenheit, barometer 30 inches. Its dimensions are 18^ inches inside diameter, 19£ inches outside, and 8 inches deep ; and when heaped, the cone must not be less than 6 inches high, equal 2747.70 cubic inches for a true cone. Pound. — The standard pound avoirdupois is the weight of 27.7015 cubic inches of distilled water, at 39.83° Fahrenheit, barometer 30 inches, and weighed in the air. Measure of JLengrth. Miles. Furlongs. Chains. Rods. Yards. Feet. Inches. 1 8 80 320 1760 5280 63360 0.125 1 10 40 220 660 7920 0.0125 0.1 1 4 22 66 792 0.003125 0.025 0.25 1 5.5 16.5 198 0.00056818 0.0045454 0.045454 0.181818 1 3 36 0.00018939 0.00151515 0.01515151 0.0606060 0.33333 1 12 0.000015783 0.000126262 0.001262626 0.00505050 0.0277777 0.083333 X Measur e of Sui ■face. Sq. Miles. Acres. S. Chains Sq. Rods. Sq. Yards Sq. Feet. Sq. Inches X 640 6400 102400 3097600 27878400 4014489600 0-001562 1 10 160 4840 43560 6272640 0.0001562 0.1 1 16 484 4356 627264 0.000009764 0.00625 0.0625 1 30.25 272.25 39204 0.000000323 0.0002066 0.002066 0.0330 1 9 1296 0.0000000358 0.00002296 0.0002296 0.00367 0.1111111 1 144 0.00000000025 0.000000159 0.00000159 0.00002552 0.0007716 0.006944 1 measure of Capacity. Cub. Yard. Bushel. Cub. Feet. Pecks. Gallons. Cub. Inch. 1 0.03961 0.037037 0.009259 21.6962 1 0.803564 0.25 0.107421 27 1.24445 1 0.31114 0.133681 0.000547 100.987 4 3.21425 X 0.429684 0.001860 201.974 9.30918 7.4805 2.32729 X 0.004329 46656 2150.42 1728 537.605 231 X 1500 MISCELLANEOUS TABLES. measure of Liquid* Gallon. Quarts. Pints. Gills. Cub. Inch. 1 0.25 0.125 0.03125 0.004329 4 1 0.5 0.125 0.17315 8 2 1 0.25 0.03463 32 8 4 1 0.13858 231 57.75 28.875 7.21875 1 Measures of Weig-hts. AVOIRDUPOIS. Ton. Cwt. Pounds. Ounces. Drams. 1 20 2240 35840 573440 0.05 1 112 1792 28672 0.00044642 0.0089285 1 16 256 0.00002790 0.000558 0.0625 1 16 0.00000174 0.0000348 0.0016 0.0625 1 TROY. Pounds. Ounces. Dwt. Grains. Pound Avoir 1 0.083333 0.004166 0.0001736 1.215275 12 1 0.05000 0.002083333 14.58333 240 20 1 0.0416666 291.6666 5760 480 24 1 7000 0.822861 0.068571 0.0034285 0.00014285 1 APOTHECARIES. Pounds. Ounces. Drams. Scruples. Grains. 1 0.08333 0.01041666 0.0034722 0.00017361 12 1 0.125 0.0416666 0.0020833 96 8 1 0.3333 0.016666 288 24 3 1 0.05 5760 480 60 20 1 Equivalent* of Lineal Iff easures — Iff etrical and English. Meters. English Measures. Inches. Feet. Yards. Miles. Millimeter . . mm Centimeter . cm Decimeter . . . Meter Decameter . . Hectometer . . Kilometer . . . .001 .01 .1 1. 10. 100. 1,000. 10,000. .039370 .393701 3.937011 39.370113 .003281 .032809 .328084 3.280843 32.80843 328.0843 3280.843 .001094 .010936 .109361 1.093614 10.93614 109.3614 1093.614 .000621 .006214 .062137 .621372 6 213718 Micron = .000,001 meter zz .C01 millimeter MISCELLANEOUS TABLES. 1501 equivalents of I,iiieal Measures — Met. and Eng-. — Continued. English Measures. Meters. Reciprocals. 1 inch 12 inches = 1 foot 3 feet = 1 yard 5£ yards=16£ feet=l rod or pole 4 poles = 66 feet = 22 yards = 1 chain (Gunter's) 80 chains = 320 poles — 5280.ft.rr 1760 yds. = lmile .02539954 .3047945 .9143835 5.029109 20.11644 1609.3149 39.37079 3.280899 1.093633 .1988424 .0497106 .00062138 A Gunter's chain has 100 links. Each link — 7.92 inches =z 0.2017 meter. Equivalents of Superficial Measures — Metrical and Eng-. (METRICAL AND ENGLISH MEASURES.) Milliare . . . Centiarerrsq.met Deciare . . . Are Decare (not used) Hectare . . . Square kilometer Square meters. .1 1. 10. 100. 1000. 10000. 1000000. English Measures. Square inches. 155.01 1550.06 15500.59 155005.9 Square feet. 1.076 10.764 107.64 1076.4 10764.3 107643. Square yards. .119 1.196 11.960 119.6033 1196.033 11960.33 Acres. 2.4711431 247.11431 Square miles. .386126 English Measures. Metrical Measures. Reciprocals. 6.451367 sq. cent. .09289968 sq.mt. .8360972 " " 25.29194 " " 1 square inch 144 square inches z=z 1 square foot . 9 square feet =: 1 square yard . . 30^ sq. yds. ) 1 perch = 1 square rod 272£ sq. ft. j ~ or pole 160perches = ) ^ 10 sq. chains } ~ * acre • • • • 640 acres == 1 square mile .... 4046.711 2589894.5 .1550059 10.7642996 1.196033 .0395383 .00024711 .00000038612 Equivalents of Weigrhts — Metrical and English. Grammes English Weights. Oz. avoir. Lbs. avoir. Tons 2000 lbs. Tons 22401bs. Troy weight. Milligramme . Centigramme . Decigramme . Gramme . . . Decagramme . Hectogramme. Kilogramme . Myriagramme . Quintal . . . Millier or Tonne .001 .01 .1 1. 10. 100. 1000. 10000. 100000. 1000000 ' .0353 .3527 3.5274 35.2739 352.7394 3527.3943 ' .ooi2 .02205 .22046 2.2046 22.0462 220.4261 2204.6215 .001102 .011023 .110231 1.102311 .000984 .009842 .098421 .984206 .015 Grs. .15 " 1.543 " 15.43235'* . . . . oz. 32.150727" 321.507266" 3215.07266 " 32150.72655" English Weights — "Avoirdupois." Grammes. Reciprocals. 1 grain .06479895 1.771836 28.349375 453.592652 45359.265 50802.376 907.18524 1016.04753 .06479895 1.555175 31 msdOfi 15.43234875 564383 24.34375 grains =z 1 dram 16 drams — 1 ounce = 437.5 16 ounces = 1 pound = 7000 100 lbs. = 1 cwt. (American 112 lbs. = 1 cwt. (English) . 20 cwt. = 1 ton (Am.) in kil< 20 cwt. = 1 ton (Eng.) in kil English Weights — " T: 1 grain grains grains ) . . . . 38 . . . • • .0352739 .00220462 .000022046 .00001968 001102311 OS . . . .000984206 roy." 15.43234875 .6430146 24 grains = 1 dwl 20 dwt — 1 oz. \ 12 oz. = 1 lb 37 3.241954 .00267923 1502 MISCELLANEOUS TABLES. 8 i— i t» eg © OH «g 8*5 cScN On a •9® ,o © 51 Nli5H(^HOJ .ONNOtHIN OO(Nt*l0rt 00 >88 CO (CriONI>0 « /v. £2 & £; 2? w ~ S ooOnmqhuso^ S3°oQo5'*n^c' .Q GO - ©£ ^HC0lOO5C0rt<(MCC>e2f0 O d pG*§ <*-i ^ eg eg c3 ri O ^ to e3 *r! ." © © >^S a •g ii-2 § as ,Q ©•£ II co fcJD rd c 5 * c3 3 MISCELLANEOUS TABLES. 1503 Metrical Measures Equivalent to English Measures. Meters. Inches. Feet. l m /m 0.039 0.0033 2 0.079 0.0066 3 0.118 0.0098 4 0.157 0.0131 5 0.197 0.0164 6 0.236 0.0197 7 0.276 0.0230 8 0.315 0.0262 9 0.354 0.0295 10°>/ m = lc/ m 0.394 0.033 2 0.787 0.066 3 1.181 0.098 4 1.575 0.13-1 5 1.969 0.164 6 2.362 0.197 7 2.756 0.230 8 3.150 0.262 9 3.543 0.295 10,/ m = .l m 3.937 0.328 .2 7.874 0.656 •3 11.811 0.984 .4 15.748 1.312 .5 19.685 1.640 .6 23.622 1.969 .7 27.560 2.297 .8 31.497 2.625 .9 35.434 2.953 1*0 39.370 3.281 Table for the Conversion of Mils. Centimeters. (l-lOOO Inches) into Centi- Centi- Centi- Centi- Mils. meters. Mils. meters. Mils. meters. Mils. meters. 1 .00254 18 .04571 35 .08888 52 .1321 2 .00508 19 :04825 36 .09142 53 .1346 3 .00762 20 .05079 37 .09396 54 .1372 4 .01016 21 .05333 38 .09650 55 .1397 5 .01270 22 .05587 39 .09904 56 .1422 6 .01524 23 .05841 40 .1016 57 .1448 7 .01778 24 .06095 41 .1041 58 .1473 8 .02032 25 .06348 42 .1067 59 .1499 9 .02286 26 .06602 43 .1092 60 .1524 10 .02540 27 .06856 44 .1118 61 .1549 11 .02793 28 .07110 45 .1143 62 .1575 12 .03047 29 .07364 46 .1168 63 .1600 13 .03301 30 .07618 47 .1194 64 .1626 14 .03555 31 .07872 48 •1219 65 .1651 15 .03809 32 .08126 49 .1245 66 .1676 16 .04063 33 .08380 50 .1270 67 .1702 17 .04317 34 .08634 51 .1295 68 .1727 1504 MISCELLANEOUS TABLES. Table for the Conversion of Mils. — Continued. Centi- Centi- Centi- Centi- Mite. meters. Mils. meters. Mils. meters. Mils. meters. 69 .1752 77 .1956 85 .2159 93 ;2362 70 .1778 78 .1981 86 .2184 94 .2387 71 .1803 79 .2006 87 .2209 95 .2413 72 .1829 80 .2032 88 .2235 96 .2438 73 .1854 81 .2057 89 .2260 97 .2465 74 .1879 82 .2083 90 .2286 98 .2489 75 .1905 83 .2108 91 .2311 99 .2514 76 .1930 84 .2133 92 .2336 100 .2540 Sng-lish Meaiurei Equivalent to Metrical Measures. CO CO w QQ 00OOC lC^©^iCOriMlOlO«OCO •sunuf) UI J9»TT I JO OOHrlHHH^rtf 'I = JTV A^IAISJ*) ogioadg -ojp^H A^TA'BJf) 0TJI09dS •noi^snq -UIOQ JO sjonpojj -noaioj\[ oiuio^y ■pqui^g -no9|oj\[ 5rH0QO0O0«DCOCO >t>-COOOt-t~ >050r>oooooco §SS OS t>e CO t-( o oo oo o Tfl COli m ■<* t^ in tjj ^ *-* .OrfHiOCJid^^ ^co o g O 02 >«* OOC If t»t >ooc *00CO t> 5»Hrl CO t~ to f 05 UO-# r-l t> rH CN ISS21 3 CO rH §35 1-1 t» e, «* « eiQ^-' . 8? * o ' .rf- c q op^ q qq . «{rj ft ft W W ft W '.oodooooo . OQOQUOO o o ""a* 1 ©" b- CO t* •<* Tt< r-l Ci OS 05 'T O '0 05 00 05 (NlCt^lOt^05^i005t^t^COlO-r-i.^t^ • OO rH CO CO rl C<» CN (N CN CO t- CC » r-l . ^>> 'S'3 rH S3 • -g» •£ © © S5* oujzi g$ 6 O *H ** 10 rH g3 bS ftWgww^W^W^W^g w w Q dodod* oo sW fc • fcogo" © . o oo o a o8- • l&S-S-as.SlS'SSes si* © « o © 3 &:B o ° s s WSo«^^&Hpq«<«£ttOr£o«5<1£oooE OJP SPECIFIC OHJLVITY 4 "¥« C»IX Water at 39.1° Fahrenheit = 4° Centigrade ; 62.425 pounds to the cubic foot (authority, Kent, Haswell, and D. K. Clark). Specific Gravity. Lbs. per Lbs. per Kilos per Authority. Cubic Cubic Cubic Foot. Inch. Deem. Aluminum, pure cast 2.56 P. R. C. 159.63 .0924 2.56 " " rolled 2.68 ** 167.11 .0967 2.68 " " anne'ld 2.66 " 165.86 .0960 2.66 " nickel alloy, cast 2.85 " 178.10 .1031 2.85 M " m rolled 2.76 " 172.10 .0996 2.76 ff " " ann'ld 2.74 44 170.85 .0989 2.74 Aluminum Bronze, 10% 7.70 Riche. 480.13 .2779 7.70 5% 8.26 " 515.63 .2984 8.26 Brass, cu. 67, zn. 33 cast 8.32 Haswell. 519.36 .3006 8.32 11 cu. 60, zn. 40 " 8.405 Thurston. 524.68 .3036 8.405 Cobalt 8.50 R.-A. 530.61 .3071 8.50 Brass, plates . . . .... high yellow . ° 8.586 P.R.C*. 535.38 .'3698' 8.586 Bronze composition . cu. 90, tin 10 . ' 8.669 Thurston. 541.17 .3132' 3 8.669 Bronze composition . . , . cu. 84, tin 16 . ' 8.832 Haswell. 551.34 .3191 ' 8.832 Lithium . • • • . 0.57 R.-A. 36.83 .0213 .57 Potassium .... 0.87 " 54.31 .0314 .87 Sodium . . . . . 0.97 <( 60.55 .0350 .97 Rubidium .... 1.52 11 94.89 .0549 1.52 Calcium 1.57 u 98.01 .0567 1.57 Magnesium . . . . 1.74 <( 108.62 .0629 1.74 Caesium . . . . , 1.88 ii 117.36 .0679 1.88 Boron ...... 2.00 Haswell. 124.85 .0723 2.00 Glucinum . « • • 2.07 R.-A. 129.22 .0748 2.07 Strontium .... 2.54 " 158.56 .0918 2.54 Barium ..... 3.75 <( 234.09 .1355 3.75 Zirconium .... 4.15 ii 259.06 .1499 4.15 Selenium . ■ . . « 4.50 Haswell. 280.91 .1626 4.50 Titanium . . . . . 5.30 ii 330.85 .1915 5.30 Vanadium . . . . 5.50 R.-A. 343.34 .1987 5.50 Arsenic . . . . . 5.67 h 353.95 .2048 5.67 Columbium .... 6.00 Haswell. 374.55 .2168 6.00 Lanthanum . . . » 6.20 " 387.03 .2240 6.20 Niobium . • * « ■ 6.27 R.-A. 391.40 .2265 6.27 Didymium .... 6.54 " 408.26 .2363 6.54 Cerium ..... 6.68 ii 417.00 .2413 6.68 Antimony .... Chromium .... 6.71 6.80 ii 418.86 429.49 .2424 .2457 6.71 6.80 Zinc, cast . . . > . 6.861 Haswell. 428.30 .2479 6.861 11 pure ... o " rolled .... 7.15 R.-A. 446.43 .2583 7.15 7.191 Haswell. 448.90 .2598 7.191 Wolfram . . . <> . 7.119 ii 444.40 .2572 7.119 Tin, pure Indium ..... 7.29 R.-A. 455.08 .2634 7.29 7.42 ii 463.19 .2681 7.42 Iron, cast .... 7.218 Kent. 450.08 .2605 7.218 " wrought . . . " wire ... o 7.70 ii 480.13 .2779 7.70 7.774 Haswell. 485.29 .2808 7.774 Steel, Bessemer . . 7.852 " 479.00 .2837 7.852 " soft . . . . 7.854 Kent. 489.74 .2834 7.854 Iron, pure .... 7.86 R.-A. 490.66 .2840 7.86 1514 MISCELLANEOUS TABLES. TABLE OF SPECIFIC GRAVITY. — Continued. Specific Gravity. Authority. Lbs. per Cubic Foot. Lbs. per Cubic Inch. Kilos per Cubic Deem. Manganese .... 8.00 R.-A. 499.40 .2890 8.00 Cinnabar 8.809 Haswell. 505.52 .2925 8.098 Cadmium ..... 8.60 R.-A. 536.85 .3107 8.60 Molybdenum . . . 8.60 (< 536.85 .3107 8.60 Gun Bronze .... 8.750 Haswell. 546.22 .3161 8.750 Tobin Bronze . . . 8.379 A. C. Co. 523.06 .3021 8.379 Nickel 8.80 R.-A. 549.34 .3179 8.80 Copper, pure . . . Copperplates and sheet 8.82 " 550.59 .3186 8.82 8.93 A. of C. M. 556.83 .3222 8.93 Bismuth 9.80 R.-A. 611.76 .3540 9.80 Silver 10.53 " 657.33 ,3805 10.53 Tantalum .... 10.80 " 674.19 .3902 10.80 Thorium 11.10 « 692.93 .4010 11.10 Lead 11.37 « 709.77 .4108 11.37 Palladium .... 11.50 « 717."88 .4154 11.50 Thalium 11.85 " 739.73 .4281 11.85 Rhodium 12.10 " 755.34 .4371 12.10 Kuthenium . . . „ 12.26 " 765.33 .4429 12.26 Mercury 13.59 " 848.35 .4909 13.59 Uranium 18.70 M 1167.45 .6755 18.70 Tungsten 19.10 M 1192.31 .6900 19.10 Gold 19.32 *' 1206.05 .6979 19.32 Platinum 21.50 (< 1342.13 .7767 21.50 Iridium 22.42 (< 1399.57 .8099 22.42 Osmium . . . „ . 22.48 11 1403.31 .8121 22.48 Authorities — R.-A. — Professor Roberts-Austen. Haswell— Haswell's Engineer's Pocket Book. P. R. C — Pittsburg Reduction Co.'s tests. Kent — Kent's Mechanical Engineer's Pocket Book. Thurston — Report of Committee on Metallic Alloys of U. S. Board appointed to test iron, steel, and other metals. Thurston's Materials of Engineering. Riche — Quoted by Thurston. A. C. Co. — Ansonia Brass and Copper Co. A. of C. M. —Association of Copper Manufacturers. SPECIFIC GRAVITY AT ©2° FAHRENHEIT OF ALUMINIM l\l) AIVIHOVM AEJLOYS. Aluminum Commercially Pure, Cast . ... ........ * 2.56 Nickel Aluminum Alloy Ingots for rolling . , . . 2.72 Casting Alloy 2.85 Special Casting Alloy, Cast 3.00 Aluminum Commercially Pure, as rolled, sheets and wire . , . , . 2.68 " " " Annealed 2.66 Nickel Aluminum Alloy, as rolled, sheets and wire ........ 2.76 " " " Sheets Annealed 2.74 Weiglit. Using these specific gravities, assuming water at 62 degrees Fahrenheit, and at Standard Barometric Height, as 62.355 lbs. per cubic foot (authority, Kent and D. K. Clark). ,. mn •„ Sheet of cast aluminum, 12 inches square and 1 inch thick, weighs 13.3024 lbs. Sheet of rolled aluminum, 12 inches square and 1 inch thick,weighs 13.9259 lbs. Bar of cast aluminum, 1 inch square and 12 inches long, weighs 1.1085 lbs. Bar of rolled aluminum, 1 inch square and 12 inches long, weighs 1.1605 lbs. Bar of aluminum, cast, 1 inch round and 12 inches long, weighs .8706 lbs, Bar of rolled aluminum, 1 inch round and 12 inches long, weighs .9114 lbs. POWER REQUIRED TO DRIVE MACHINERY SHOPS, AND TO DO VARIOUS KINDS OP WORK. PROXY BIUHE d 4 Constant = ~^- = .0001904. Fig. 1. Then Horse-power = .0001904 X d x to X revolutions per minute. Horse-Power formula*. In an article by C. H. Benjamin in March, 1899, Machinery are the follow- ing formulas for computing the horse-power required to operate tools, where W — weight metal removed per hour. Experiments with several lathes give: H.P. = .035 W for cast iron. H.P. = .067 W for machinery steel. Experiments with a Gray planer give: H.P. = .032 W for cast iron. Experiments with a Hendey shaper give H.P. = .030 W for cast iron. For milling machines we have: H.P. = .14 W. for cast iron. H.P. » .10 W for bronze. H.P. = .30 W for tool steel. In each case, the power required to run the tool, light, should be added. Power Used by Machine-Tool*. (R. E. Dinsmore, from the Electrical World.) 1. Shop shafting 2 T 3 B - in. x 180 ft. at 160 revs., carrying 26 pulleys from 6 in. diam. to 36 in., and running 20 idle machine belts . 1.32 H. F 2. Lodge-Davis upright back-geared drill-press with table, 28 in. swing, drilling f in. hole in cast iron, with a feed of 1 in. per minute 0.78 H. P. 3. Morse twist-drill grinder No. 2, carrying 26 in. wheels at 3200 revs ' 0.29 H. P. 4. Pease planer 30 in. x 36 in., table 6 ft., planing cast iron, cut \ in. deep, planing 6 sq. in. per minute, at 9 reversals .... 1.06 H. P. 5. Shaping-machine 22 in. stroke, cutting steel die, 6 in. stroke, y in. deep, shaping at rate of 1.7 square inch per minute . . . 0.37 H. P. 6. Engine-lathe 17 in. swing, turning steel shaft 2| in. diam., cut T \ deep, feeding 7.92 in. per minute 0.43 H. P. 7. Engine lathe 21 in. swing, boring cast-iron hole 5 in. diam., cut r 3 5 diam., feeding 0.3 in. per minute 0.23 H. P. 8. Sturtevant No. 2, monogram blower at 1800 revs, per minute, no piping 0.8 H. P. 9. Heavy planer 28 in. X 28 in. x 14 ft. bed, stroke 8 in., cutting steel, 22 reversals per minute 3.2 H. P 1616 1516 POWER REQUIRED TO DRIVE MACHINERY, ETC. Power Required for Machine Tools — Results of Tests. Tests of Various Machine Tools. (From a paper read by F. B. Duncan before the Engineers' Society of Western Pennsylvania.) Engine Lathes. 16 in.; motor power required, approximate, 2 H.P. at maximum. 18 in. X 6 ft.; motor power required, 2.1 H.P. 36 in. X 10 ft.; motor power required, 10 H.P. Planebs. 10 X 10 X 20 ft. ; 3 tools, f X & in. cut; cutting speed, 18 ft.; planing 40-ton iron casting. H.P. required for cut, 26.5; for return, 23.6; for re- verse, 42.9. Ratio return, 3 to 1. Motor, 30 H.P., belted to countershaft. 8 X 8 X 20 ft.; 3 tools, f X £ in. cut; cutting speed, 18 ft.; planing 32-ton iron casting; H.P. for cut, 16; for return, 14-8; for reverse, 28.2. Ratio return, 3 to 1. Motor, 25 H.P., belted to countershaft. 66 X 60 in. X 12 ft.; 2 tools \ X 1-16 in. cut; cutting speed, 21 ft.; plan- ing 4 ton open hearth casting. H.P. required for cut, 10; for return, 14; for reverse, 16. Ratio return, 3^ to 1. Motor mounted on planer housing with 42-inch 1,500-pound flywheel, running at 400 revolutions, mounted on motor shaft; flywheel used as driving pulley for return of platen. 28 X 52 in. X 6 ft.; 1 cutting tool, f X i in. cut; cutting speed, 22 ft.; planing 3-ton iron casting. H.P. required for cut, 3.1; for return, 3.8; for reverse, 4.4. Ratio return, 4 to 1. Motor, 3 H.P., 800 revolutions. Aver- age load on motor, 2.48. Flywheel, 30 in. diameter, 496 pounds, 800 revo- lutions, mounted on motor shaft and used as pulley for return of platen. Miscellaneous. 28 in. Gisholt turret lathe: machining Tropenas cast steel weight, 400 pound; size cut, one tool, f X 5-16 in.; 4 tools, £X 5-64 in.; weight casting. 400 pounds; power for cut, 3.9 H.P 21 in. drill press; power required, 1 H.P. 5 ft. radial drill; maximum power required, 2.03 H.P. Motor used, 2 H.P. 600 revolutions. Double and emery wheel stand ;two 18 X 2 in. wheels, 950 rev.; 2 laborers grinding castings; maximum H.P., momentarily, 6; average, 3.5. Motor, 5 H.P., mounted on grinder shaft. 10 ft. boring and turning mill; cutting tools, 2; cut, f X 1-16 in.; cutting speed, 20 ft.; machining 3.5-ton casting; H.P. required for cut, 8.6. Motor used, 12 H.P. Slotter; cut, | X 1-16 in.; speed of tool, 20 ft.; machining open hearth steel castings; power requirea, 6.98 H.P. Flat turret lathe; \\ H.P. motor required. Gisholt tool grinder; speed, 1,600 to 1,800 rev.; power required, 7 for short periods, 4 on average. Motor used, 5 H.P. The figures given in the following table for the power required to run the planing machines empty, do not include the maximum horse-power at the instant of reversal, but represent the average forward and return of the empty table. POWER REQUIRED FOR MACHINE TOOLS. 1517 Results of tests at the Baldwin Locomotive Works, Philadelphia : Size. Material Cut. 73 "o O H *o d Horse-Power. Kind of Machines. T3 d o «j OGQ Total Cutting. Min. Max. Ave. Wheel lathe 84 in. 84 in. 84 in. 78 in. 78 in. 36 in. X 12 in. 62 in. X 35 ft. 62 in. X 35 ft. 36 in. X 12 ft. 24 in. X 13 ft. 36 in. X 18 ft. 56 in. X 35 ft. 56 in. X 24 ft. 90 in. 42 in. 4 ft. 6 in. 5 ft. 6 in. 40 in. X15 in. 19 in. str. Cast iron Cast iron Cast iron Cast iron Cast iron Wrought iron Wrought iron Wrought iron Wrought iron Steel Wrought iron Wrought iron Wrought iron Cast steel Cast steel Cast steel Cast iron Wrought iron Wrought iron 2 2 2 1 1 1 2 2 2 2 2 2 2 2 1 1 1 1 1 2.9 4.2 5.3 4.3 5.5 4.4 20.6 23.0 11.3 7.9 5.8 6.2 4.7 7.1 6.7 21.6 26.0 13.8 6.1 Wheel lathe . 5.1 Wheel lathe . Boring mill . 1.5 5.8 4.5 Boring mill . 6.5 Slotter . . . Planer . . . Planer . . . Planer . . . Planer . . . 1.5 i.4 2.7' 1.95 3.2 4.6 4.56 1.43 0.96 2.1 1.6 1.8 1.3 1.5 11.4 5.8 3.0 4.3 4.3 9.9 6.0 2.1 1.1 2.4 2.4 2.2 1.8 5.3 21.1 24.5 12.5 8 Planer . . . 16 7 Planer . . . Planer . . . Wheel lathe 13.0 16.0 13.7 17.7 13.3 16.8 6 38 Radial drill . 2.1 Boring mill . 4 6 Boring mill . Slotter . . . 4.2 4.8 4.4 7 3 Shaper . . . 4.8 9.7 7.3 Results of tests, in ten different plants by C. H. Benjamin, to determine the proportion of power absorbed by the counters, belting, line shaft, etc. Useful Friction Horse-Power. Horse- Power. Nature of Work. <£ p 8 ** 6 8^ 8*3 © 3.5 rHQQ ^ Per Bear- ing. Per Coun- ter. Per Belt. i Per Man. Pn o Pm o & fc Boiler shop 4.77 .205 .04 .550 .538 .477 .310 .877 Bridge work . . 3.28 .137 .04 .337 .606 .521 .164 .142 Heavy machinery. 5.70 .233 .038 .581 .665 .453 .707 .160 Heavy machinery. 8.55 .306 .06 .799 .600 .475 .627 .342 Average . . . 5.57 .220 .044 .567 .602 .481 .452 .380 Light machinery 2.75 .276 .034 .204 .155 .095 .790 .099 Small tools . . 8.00 .400 .09 .689 .127 .119 .109 .152 Small tools . . 2.49 .233 .03 .240 .121 .113 .881 .227 Sewing machines 4.36 .430 .05 .397 .269 .208 .180 .204 Sewing machines 5.08 .134 .034 .406 .172 .154 .181 .093 Screw machines . 6.33 .381 .05 .633 .291 .235 .296 .396 Average . . . 4.83 .309 .048 .428 .189 .154 .406 .195 For group driving determine average horse-power for each tool, add these together and use a motor with a capacity of from 40 to 70 per cent of the total thus obtained. The size of motor will depend upon the way the ma- chines are worked — i.e., cutting speed, feed, material cut, and whether mod- ern air-hardened tools are used; also to what extent machines are to operate simultaneously. The larger the group the smaller the motor relative to total power. 1518 POWER REQUIRED TO DRIVE MACHINERY, ETC. Motor Power for Machine Tools. Actual Installation*. William R. Trigg Works. Horse-power of motors used at the Wm. R. Trigg Works, Richmond, Va. (See article by Wm. Burlingham, in September, 1902, Machinery.) *"*«■* H ofMo P t°or r 18 in. Cincinnati D. H. shaper 3 10 ft. Pond boring mill 20 18 in. Newton slotter 7£ No. 6 Baush radial drill 5 5 ft. radial drill 5 14 in. Newton slotter 5 36 in. X 12 ft. Woodward & Powell planer 15 56 in. X 56 in. X 12 ft. Gray planer 20 30 in. X 30 in. X 8 ft. Woodward & Powell planer .... 10 No. 5 Mitts & Merrill keyseater 3 No. 1 Newton floor boring machine 7.5 38 in. X 44 ft. shaft lathe 7.5 Niles hor. boring machine 15 No. 4 duplex milling machine, Newton 10 7 ft. Betts boring mill 15 10-in. Betts slotter 3 5 1-in. Baush boring mill 7.5 No. 1 Acme bolt cutter 7.5 42 in. X 42 in. X 20 ft. planer 15 Dallett & Co. portable deck planer 5 62 in. X 30 ft. Putnam lathe 10 36 in. X 25 ft. Putnam lathe 7.5 22 ft. Bending rolls Driving 35 Lifting 10 12 in. straightening rolls 15 No. 3 double punch 10 Duplex planer 15 Double angle shear 10 No. 4 punch 10 No. 4 punch 10 No. 2 punch 5 No. 3 hor. punch 7.5 No. 6 Sturtevant blower 12 Hannibal Shops. Horse-power of motors used at the Hannibal shops of the St. Joseph and Hannibal Ry. {Railroad Gazette.) Machine Shop. Mo^i't^ Horse-Power Machine ' of Motor. 54 in. planer 15 42 in. planer 10 32 in. planer 7.5 Emery grinder Grindstone Double centering machine 3 90 in. driving wheel lathe 6 2 quartering ends of same 3 48 in. lathe 5 18 in. slotter 22 in. shaft lathe 5 Car wheel borer 5 Car wheel preac ..-„,,,,., 10 MOTOR POWER FOR MACHINE TOOLS. 1519 Ayf„«i,;^« Horse-Power Machme - of Motor. Journal lathe Grindstone 10 32 in. lathe 4 18 in. shaper 5 40 in. vertical drill 2 4-spindle gang drill 7.5 Milling machine 3 Grinding machine 3 32 in. lathe Flat turret lathe 4 18 in. lathe 18 in. brass turret lathe 16 in. lathe 4 16 in. lathe 16 in. lathe Drill 5 No. 5 radial drill Acme triple bolt cutter 2 in. double bolt cutter 5 No. 6 radial drill 5 No. 5 oscillating grinder 25 24 in. lathe 24 in. lathe 5 Acme nut tapper 3 16 in. tool room lathe 2 No. 2 oscillating grinder Twist drill grinder 5 Boiler Shop. No. 6 Niles power bending rolls 35 Double punch and shears 6 Flue tumblers 15 Flue cutter Flue scarfer 3.5 Small punch 2 Blacksmith Shop. Bolt header Grindstone 5 Bolt shears 5 Punch and shears 7.5 Bradley hammer 5 Forge blower 15 Forge fan 10 Wood Mill. Automatic cut-off saw 10 38 in. band resaw 8 Vertical borer 7.5 Automatic car gainer 15 Mortiser 15 Buzz planer 7.5 Single surfacer 13 Planer and matcher 25 Self-feed large rip saw 25 Small rip saw 15 Four-sided timber planer 45 Power feed railroad cutoff saw 10 Rip saw 15 Outside moulder 22 . 5 Double surfacer 17.5 Upright moulder 9.5 Large tenoner 7.5 Scroll saw , . , , 2 1520 POWER REQUIRED TO DRIVE MACHINERY, ETC. MonhinPQ Horse-Power Machines. of Motor# Sharpener and gummer Band saw, setter and filer Emery wheels Grindstone 5 Shavings exhauster 60 Elevator 7.5 Cabinet Shop. Patternmaker's lathes 5 Scroll saw 3 Tenoning machine 5 Hollow chisel mortiser 4 Universal saw bench 5 Central Railroad of Hew Jersey Shops. Horse-power of motors used at the Central Ry. of New Jersey ShopSi (Railroad Gazette.) t tmi ^ Horse-Power Lathes - of Motor. 88 in. wheel 7* 72 in. driving wheel 5 Single head axle 2 Double head axle 5 36 in. X 16 ft 4 33 in. X 18 ft 3 30 in. X 12 ft 3 24 in. X 16 ft 3 42 in. X 14 ft 3 28 in. X 12 ft 2 Planers, Slotters, Shapers. 60 in. X 60 in. X 25 ft. Pond planer 15 36 in. X 36 in. X 10 ft. Pond planer 5 36 in. X 36 in. X 10 ft. planer . . 7* 24 in. X 24 in. X 6 ft. Pond planer 5 48 in. X 54 in. X 14 ft. planer 7$ 24 in. crank planer 4 16 in. traveling head shaper 3 8 in. slotter 3 14 in. slotter 4 24 in. slotter 4 Boring and Turning Mills — Boring Machines. 80 in. boring mill 5 39 in. boring mill t 5 39 inch vertical boring machine 3 36 in. car wheel boring machine 5 8 ft. boring mill with slotter 7£ Driving wheel quartering machine . . . ; 5 Rod borer 3 Drill Presses. No. 3 Bickford radial drill 3 30 in. drill press 2 30 in. drill press . 2 40 in. drill press (floating) 40 in. drill press 3 40 in. drill press (floating) 8-spindle arch-bar drill 5 Grinders. B. & S. surface grinder 3 Water tool grinder 5 Angle cock grinder 3 MOTOR POWER FOR MACHINE TOOLS. 1521 Miscellaneous. Horse-Power of Motor. 54 in. throat single end punch . . . l 10 No. 6 bulldozer complete 7£ 3 in. heading and forging machine 10 Newton cold-saw 10 £ in. bolt heading machine 5 i in. Acme single head bolt cutter 2 Bolt shears 4 10 ft. boiler rolls 5 84 in. driving wheel press 5 42 in. car wheel press 5 36 in. car wheel press 3 An Ideal Railway Shop. Estimated motor power for various tools for a railway shop. (From a paper read before the Master Mechanics' Convention, June, 1902, by L. R. Pomeroy.) Horse-Power Lathes, of Motor. 90 in. driving wheel 7.5 80 in. driving wheel 7.5 42 in. truck wheel tire turning, heavy 5 Axle, single, heavy, for driving axles 5 Axle, double head 5 48 in. X 14 ft. engine, heavy 5 38 in. X 16 ft. engine, heavy 3 30 in. X 12 ft. engine, heavy 3 28 in. X 12 ft. engine, heavy 2 26 in. X 8 ft. engine, very heavy 2.5 20 in. X 10 ft. engine, medium . . 2 18 in. X 10 ft. engine, medium 2 16 in. X 8 ft. engine, medium 2 2 X 24 flat turret . 3 21 in. heavy screw machine 3 20 in. universal monitor, for brass 1 18 in. universal monitor, for brass 2 16 in. Fox lathe, with turret 2 12 in. speed lathe 2 Drill Presses. 72 in. radial, heavy 5 60 in. radial, heavy 48 in. radial, medium 2 40 in. upright heavy 36 in. upright heavy 2$ 30 in. upright, heavy 20 in. upright, light Cotter drilling machine 2 Sensitive drill .5 Grinding Machines. Landis grinder for piston rods, etc 3 Surface grinder Universal grinding machine (same as No. 2B.&S.). ... 2 Twist drill grinder Sellers or Disholt tool grinder 3 Two 20 in. wet tool grinders 5 Small tool grinder (B. & S. No. 1) 1 Flexible swinging, grinding, and polishing machine .... Large buffing and polishing wheel 2£ Planers. 72 in. X 72 in. X 14 ft 15 60 in. X 60 in. X 28 ft 15 54 in. X 52 in. X 14 ft 15 42 in X 42 in. X 16 ft 10 38 in. X 38 in. X 10 ft 7.5 36 in. X 36 in. X 10 ft 7.5 30 in. X 30 in. X 8 ft 5 1522 POWER REQUIRED TO DRIVE MACHINERY, ETC. Horse-Powei , , ' of Motor. 16 m traveling head shaper 2 16 in. shaper 2 14 in. shaper 2 12 in. shaper 2 Richards side planer, 20 in. X 6 in 5 Slotting Machines. 18 in. slotting machine 7.5 14 in. slotting machine 5 10 in. slotting machine 3 Colburn keyseating machine 5 Boring Mills. 84 in. boring and turning mill, two heads 7.5 62 in. boring and turning mill, two heads 5 37 in. boring and turning mill, two heads 5 30 in. horizontal boring and drilling machine 5 Cylinder boring machine 7.5 Milling Machines. Heavy vertical milling machine 10 Vertical milling machine (No. 6 Becker-Brainard) .... 7.5 Heavy slab milling machine 15 Universal milling machine (heavy) 5 Plain horizontal milling machine (same as Becker-Brainard No. 7) 4 Small, plain milling machine for brass work 2.5 Universal milling machine (same as B. & S. No. 3) . . . . 1 Bolt and Nut Machinery. 2\ in. single head bolt cutter 2 1 J in. double head bolt cutter 4 5-spindle nut-tapping machine 3 Bolt-pointing machine 3 Nut-facing machine 3 Heavy power hacksaw 2 Small power hacksaw 1 Blacksmiths' Tools. Quick-acting belt hammer 5 3 in. bolt heading and upsetting machine 3 H bolt heading and upsetting machine 3 Heavy shear to cut 4 X 4bar 7£ Shear to cut up to 5 X 1 in 5 Shear to cut up to \\ in. round iron 5 No. 3 Newton cold saw cutting-off machine 5 Boiler Tools. 16 ft. gap hyd. fixed riveter, pump, accumulator, and crane, complete 100 Heavy boiler plate punch or shear, 48 in. throat depth . . 10 Heavy boiler plate punch or shear, 30 in. throat depth . . 7.5 Tank plate punch, 30 in. throat depth 5 Tank plate shear, 24 in. throat depth 5 Boiler plate shear, 30 in. throat depth, f in. plate ... 7.5 Flange punch 5 12 ft. boiler rolls for \ in. plate Light 6 ft. rolls 35 Plate planer, 20 ft 3 Woodworking Tools. Patternmaker's lathe 6 Band saw 3 Medium-sized saw bench, crosscut and rip saw 5 Medium-sized hand planing and jointing machine f> HOBSE-POWER IN MACHINE-SHOPS. 152£ W§ nrsft'powcr in machine -shops; Friction; Men employed. (Flatl Ler.) Horse-power. 13 o 6 > H u Kind T3 . •3 >> 'u * Name of Firm. of z* O Work. 1$ "2 «M 3 ee u 3 .S % o O o V a> o p d 6 H « tf ^ K » % Lane & Bodley .... E. & W.W. 58 132 2.27 J. A. Fay & Co W. W. 100 15 85 15 300 3.00 3.53 Union Iron Works . . E.,M. M. 400 95 305 23 1600 4.00 5.24 Frontier Iron &Brass Wks M.E.,etc. 25 8 17 32 150 6.00 8.82 Taylor Mfg. Co E. 95 230 2.42 Baldwin Loco. Works L. 2500 2000 500 80 4100 1.64 8.20 W. Sellers & Co. (one de- partment) H.M. 102 41 61 40 300 2.93 4.87 Pond Machine Tool Co. . M. T. 180 75 105 41 432 2.40 4.11 Pratt & Whitney Co. . . » 120 725 6.04 Brown & Sharpe Co. . . (< 230 900 3.91 Yale & Towne Co. . . . C.&L. 135 67 68 49 700 5.11 10.25 Ferracute Machine Co. . P. &D. 35 11 24 31 90 2.57 3.75 T. B. Wood's Sons . . . P. & S. 12 30 2.50 Bridgeport Forge Co. . H. F. 150 75 75 50 130 .86 1.73 Singer Mfg. Co S.M. 1300 3500 2.69 Howe Mfg. Co 44 350 1500 4.28 Worcester Mach. Screw Co. M. S. 40 80 2.00 Hartford " " " " 400 100 300 25 250 0.62 0.83 Nicholson File Co. . . F. 350 400 1.14 Averages 346.4 38.6% 818.3 2.96 5.13 Abbreviations: E., engine; W.W., wood-working machinery; M. M., mining machinery ; M. E., marine engines ; L., locomotives ; H. M., heavy machinery ; M. T., machine-tools ; C. & L., cranes and locks ; P. & D., presses and dies; P. & S., pulleys and shafting; H. F., heavy f orgings ; S. M., sewing-machines ; M. S., machine-screws ; F., files. Verts at tlie Win, It. TrSg-g- Works. (See September, 1902, Machinery.) 62 in. X 30 ft. lathe, turning hard cast iron. Tool of Sanderson self- hardening steel. About 6 H.P. required to run the lathe light. Experi- ments: (1) Cut, & in. deep, 1-16 in. feed; 21 ft. cutting speed; 33.8 lbs. metal removed per hour; 1.15 H.P. = .034 lb. wt. metal removed per hour. (2) Cut, i in. deep, 1-16 in. feed; 33 ft. cutting speed; 54.8 lbs. metal removed per hour; 1.52 H.P. = .028 lb. wt. metal removed per hour. 36 in. X 12 ft. Woodward & Powell planer, two tools cutting on cast steel. Cuts were \ in. deep by 1-16 in. feed. First experiment, cutting speed, 17.15 ft. per minute; reverse speed, 60 ft. per minute. H.P. cutting, 2.15; return- irg, 2.22; reverse to cut, 4.77; reverse to return, 11. Second experiment, cutting speed, 21.83 ft. per minute; reverse speed, 68.6 ft. per minute. H.P. catting, 2.85; returning, 3.06; reverse to cut, 6.52; reverse to return, 11. In these experiments the reverse to cut consumed (of course for an instant only) from 2.22 to 2 29 times the power required to cut; and the reverse to return 1524 POWER REQUIRED TO DRIVE MACHINERY, ETC. from 4.95 to 3.59 the power required to return; or from 5.11 to 3.86 the power required for cutting. 36 in. X 25 ft. Putnam lathe, cutting shaft nickel steel, oil tempered and annealed, with Sanderson self-hardening tool steel. Diameter work, 9f in. Experiments: (1) Cut £ in. deep X £ in. feed, 5.76 revolutions. H.P. = 1.5. (2) Cut 3-16 X i, 4.65 revolutions, H.P. = 1.76. (3) Cut J X £, 3.28 revo- lutions, H.P. = 1.9. (4) Cut 1 X £, 2.71 revolutions, H.P. = 1.26. Another line of experiments was conducted with the same lathe cutting nickel steel shaft 9f in. diameter, cut constant at i in. deep and feed £ in. per revolution. The speed of motor was gradually increased from No. 3 notch to No. 11 notch of the controller, representing an increase of motor revolu- tions from 220 to 700 per minute, or an increase in the revolutions of the lathe from 3.03 per minute to 9.64 per minute. The H.P. required increased from 1.068 to 4.26. Cotton Machinery. Wm. O. Webber. Looms. Make. Amoskeag, Whitin Amoskeag, Whitin Lowell Shop . . . Lowell Shop . . . Lowell Shop . . . Whitin Amoskeag .... Whitin Width. 49 in. 45 in. 40 in. 36 in. 32 in. 40 in. 48 in. .40 in. Picks per Min. 142 ft. 142 ft. 160 ft. 160 ft. 170 ft. 144 ft. 144 ft. 147 ft. Picks per Inch. 68X80 68X80 72X80 64X90 64X88 80X84 80X84 84X92 Warp. Weft. 24 X31 24 X31 24 X31 24 X38 27* X38 28 X33 28 X33 28 X33 Horse- power. .254 .214 .273 .286 .311 .2315 .257 .256 Slashers. — 2,872 ends Cut in 84 seconds = 3 . 93 horse-power. Cut in 64 seconds = 4.574 horse-power. Cut in 52 seconds =5.53 horse-power. Warpers. — 359 ends, 50 yds. per min. = .313 H.P. Shears, 4 blades and fans, 1,800 R.P.M. 100 yards per min. 42 inch cloth = 6.07 H.P. Cards. Horse- power. Finisher, Lowell 36 inch cylinder 128 R. .187 Finisher, Amoskeag 36 inch cylinder 140 R. .247 Finisher, Whitim 36 inch cylinder 140 R. .19 Lowell breaker 36 inch cylinder 128 R. .225 Amoskeag breaker 36 inch cylinder 140 R. .247 Whitin breaker 36 inch cylinder 140 R. .173 Revolving top flat card . . . 40 inch cylinder 162 R. .921 POWER REQUIRED. 1525 Printing* machinery, Power Required. Wm. 0. Webber. 30 in. X 52 in. 2 rev. No. 8 Cottrell press, 19 impressions per minute 27 in. X 41 in. No. 20 Adams press, 16 impressions per minute 32 in. X 54 in. Huber perfecting press 43 in. X 64 in. Huber perfecting press, automatic feed 27 in. X 41 in. No. 4 Adams job press 26 in. X 40 in. No. 2 Adams job press 32 in. X 54 in. No. 1 Potter cylinder roller press 26 in. No. 1 Hoe perfecting press Web paper- wetting machine Horse- Power. 1.189 .68 2.44 5.55 .43 .337 .50 5.41 .52 Newspaper Printing Machinery. One 10 page web perfecting press, 12,000 per hour One 10 page' web perfecting press, 24,000 per hour One 12 page web perfecting press, 12,000 per hour One 12 page web perfecting press, 24,000 per hour One 32 page web perfecting press, 12 000 per hour Horse- power. 15.39 31. 20.45 29.56 28.73 Calico Printing Machinery — Capacity 100 yds. print goods per min. Horse- power. One 19 cylinder, soaper and dryer, full .... One cutting machine, full One set drying cans to cutting machine, full . . One back starcher, 3 wide machines, full. . . . One indigo skying machine, 5 vats, all working full One 40 in. 5 roll calender, working full .... One single color printing machine Rev. Foot- per min. pounds. 110 2,182 65 1,525 110 1,282 115 2,330 64 2,635 234 5,390 3.97 2.77 2.33 4.24 4.78 9.80 10.6 Power Required For Sewings-machines. Light-running 20 machines to 1 h.p. Heavy work on same 15 " " " Leather-sewing 12 " *' *' Button-hole machines ... 8 to 12 " " " 1526 POWER REQUIRED TO DRIVE MACHINERY, ETC. POWER C©]¥SU]fri?Ti©]¥. Character Average K.W. Hours per Month. Average Con- nected Indi- vidual Ave. No. of Connect- ed Motor Load of Installations. Motor Load, H.P. or Group Drive.* Mo- tors. Times Average Load. O r horse-power, $1.26. per POWER USED BY MACHINE TOOLS. 1529 Saving: by Electric Drive. — Fig. Nos. 2 and 3 show graphically the saving made in power by the use of electric drive over the use of shafting and belting. 60 40 LU O OL I UJ CD £20 O I TIME ESTIMATED FRICTION LOSS IN ENGINE ESTIMATED FRICTION LOSS IN ENGINE 9 10 A.M. 3 ± P.M. Fig. 2. 1895, Diagram of Losses in Power Transmission, Factory of Central Stamping Co., Brooklyn, N.Y. Crocker- Wheeler Electric Company. Fig. 3. 1895, Diagram of Losses in Power Transmission, Factory of Central Stamping Co., Newark, N.J. Crocker-Wheeler Electric Company. 1530 POWER REQUIRED TO DRIVE MACHINERY, ETC. LIST ©E TOOLI A^D SUPPLIES USJEEUJL IH IXSTAEMXO ELECTRIC MOHTS AJD DYIAMOS. 1 Tool chest. 1 Magneto and cable. 1 Speed indicator. 1 Tape line, 75 ft. 1 Rule, 2 ft. 1 Scraper, for bearings. 1 Blow lamp. 1 Clawhammer, No. 13. 1 Ball pein hammer, No. 24. 1 B. & S. pocket wrench, No. 4. 1 Monkey wrench, 10 inch. 1 Set (2) Champion screw-drivers. 1 Large screw-driver, 12-inch. 1 Off-set screw-driver. 1 Ratchet brace, No. 33. Bits, h |, h f , I, 3, 1 inch. 1 Clarke Expansive bit, ^ to 3 inch. 1 Screw-driver bit. 1 Gimlet bit. 1 Wood countersink. 1 Extension drill, § in. length, 24 in. 1 Long or extension gimlet. 1 Cold chisel, f inch. 1 Half round cold chisel. 1 Cape chisel. 1 Wood chisel, firmer paring, f inch. 1 Brick drill. Piles, one each : round, flat, half- round and three-square. 1 Saw, 20 inch. 1 Hack-saw, 10 inch. 10 Extra saw blades. 1 Plumb bob. 1 Brad awl. 1 Pair carbon tongs. 1 Soldering copper, No. 3. 1 Pound of solder. 1 Pair of climbers. 1 Come-along. 1 Splicing-clamp. 1 Strap and vise. 1 Pair line pliers, 8 inch. 1 Pair of sirle-cutting pliers, 5 inch. 1 Pair of diagonal-cutting pliers, 5 in 1 Pair of round-nose pliers, 5 inch. 1 Pair of flat-nose pliers, 5 inch. 1 Pair of burner pliers, 7 inch. 6 Sheets of emery cloth. 6 Sheets of crocus cloth. 2 Gross of assorted machine screws. 2 Gross of assorted wood screws. 150 Special screws. Taps, 6-30, 10-24, 12-24, 18-18. Drills, 34, 21, 9, 15-64. Tap wrench. The following-named tools will probably be required in constructing lines for city or commercial lighting : (Davis.) Article. Stubs' pliers, plain . . . . , Climbers and straps . . . . , Pulley-block and ecc. clamp Come-along and strap . . . , Splicing-clamps , Linemen's tool-bag and strap , Soldering-furnace . . . . , Gasoline blow-pipes . . . . Soldering coppers , Pole-hole shovels Pole-hole spoon, regular . . , Octagon digging-bars . . . Tamping-bars , Crowbar Pick-axe Carrying-hook, heavy . . . Cant-hook Pike-poles Pole-supporter Comb, pay-out reel and straps Nail-hammer Linemen's broad hatchets Drawing-knives Hand-saw Ratchet-brace, bits .... Screw-drivers . . r . . . Wrench Bastard file Size. 8 in. 'fo ' No. 3 B. &S. 21b. 8 ft. 7 ft. 8 ft. 7 ft. 10 1b. 4 ft. 16 ft. 6 ft; ' i lb! 6 in. 12 in. 26 in. 10 in. 8 in. 12 in. 12 in. Cost about $2.00 3.00 8.00 2.25 2.50 4.80 6.00 6.00 .95 1.50 1.25 3.50 2.60 .90 .75 6.00 2.00 2.40 12.00 20.00 1.00 1.50 2.10 1.50 3.00 .80 1.25 .30 THAWING WATER PIPES. 1531 APPROXIMATE LIST OF STPPIIES REQUIRED IN INSTALLING 15 CITY LAMPS AND 20 COMMERCIAL LAMPS ON A FITE-MILE CIRCUIT, SETTING POLES 132 FEET APART. (Davis.) Articles. Electric-light poles Electric-light poles Electric-light poles Cross-arms, 4-pin . Painted oak pins . Oak pins and bolts Iron break-arms . Lag-screws and washers Glass insulators, D. G. Pole steps .... Guy stranded cable . Cross-arm brace and bolts Line wire Size or Diameter. 30 ft., 6 in. 35 ft., 7 in. 40 ft., 7 in. 4 ft. H in. \ X 7 in. f X 8'in*. fin. 6BS' Price about $2.40 each 4.15 " 5.50 " .30 " .02 " .07 " .75 " .04 " .07* " .05 M .07 lb. .20 each 125.00 mi. Quantity. 180 40 * 200 800 24 25 400 850 2500 500 lbs. 40 6 miles MATERIAL REdriRED FOR COKHECTIIO IX LAMPS. (Davis.) Sleet-proof pulleys . . Street-lamp cleats, iron Arc-lamp cordage . Suspension cable . Hard-rubber tube . Soft-rubber tubing Arc cut-out . . . Porcelain insulators screws . . . Oak brackets and spikes and fin. | in -5 IXl I in. |0.75 each. 25 " 1.25 hd. ft. .02| ft. 1.50 lb. .20 ft. 3.50 each 2.40 hd. 2.50 " 30 15 25 3000 ft. 5 lbs. 200 ft. 20 400 150 THA WIXO FROZEX WATER PIPE§ ELECTRICALLY. The use of electricity for thawing out frozen underground water pipes requires a transformer say of 10 or 20 kilowatts capacity, which can be taken to the locality required, connecting the primary with the high ten- sion circuit passing the place, and then connecting the secondary through an ampere meter and rheostat to the service in trouble. Where services from the street mains to two adjacent houses are both frozen, it is only necessary to connect the secondary circuit to the kitchen faucet of both houses and thus the circuit is complete through the service of one house to the street main and back through the service of the second house. Where the service of but one house is to be thawed, one end of the sec- ondary circuit is connected to the kitchen faucet and the other end to the nearest street hydrant or other street connection. Currents varying from 20 to 500 amperes are used, obviously, varying according to the conditions; and the time taken to thaw the ice sufficiently to start the water running will be from 10 to 45 minutes or perhaps 3 to 8 hours, according to circum- stances. 1532 POWER REQUIRED TO THAW WATER PIPES. The average time for the ordinary house service will seldom exceed 45 minutes, while for a five or six inch pipe that has been frozen solid the highest amount of current and time mentioned will be required. It is very seldom necessary to melt the entire plug of ice, as the thawing of a thin sheet nearest the metal will start the water running and that will consume the ice in a short time. The following table is compiled from data that have appeared in various periodicals. It represents average conditions for last year, and shows what may be expected in the future: Size Pipe. Length. Volts. Amps. Time Required to Thaw. I" 40 ft. 50 300 8 min. 100 ft. 55 135 10 min. F 250 ft. 50 400 20 min. 1" 250 ft. 50 500 20 min. 1" 700 ft. 55 175 5 hrs. 4" 1300 ft. 55 260 3 hrs. 10" 800 ft. 70 400 2 hrs. The following notes on melting points of various substances may be of assistance in checking thermometers and showing the safe limits on elec- trical apparatus that operates in heated conditions. C. F. Pure cane sugar (granulated) melts at 160 320 Tin melts at • . 235 455 Bismuth melts at 269 518 Lead melts at 327 618 Zinc melts at . 419 788 ^ $ I I I IV ; • ^Accoai ith pa: INDEX. Abbreviations for units, 6. Absohm, value of, 7. Absolute units, 2. Absorbent for X-ray tubes, 1251. Abvolt, value of, 7. Acceleration, average rate of, 666. definition of, 3. formula for, 664. Acetic acid in electrolyte, test for, 878. Acheson process, graphite produc- tion by, 1245. Acid, conducting power of, table of, 905. Acker process, caustic soda by, 1240. Acoustic telephone call system, 294. Action of wattmeters, 1039. Active material, increase of, 873. loss of battery plates, 881. Acyclic machines, def. of, 504. Adhesion of cement, 1294. Admittance, symbol of, 8. Admixture of copper, effect of, 144. Advance wire, properties of, 202, 207. Aerial circuits, charging current per 1000 feet of A.C., 253-258. lines, res. of, 61. telephone cables, 188. capacity of, 1085. wires, capacity per 1000 feet of, table of, 252. location of orosses in, 327. A-geing of iron and steel, 455. of transformers, guarantee against serious, 498. tests, curves of, 453. A I. E. E., copper wire tables of, 146. Air-blast transformers, 449. dielectric strength of, 233. -gap ampere turns, 367. break down, 1056. Air-gap, discussion of, 363. flux, 365. pumps, 1445. resistance, effect of moving body on, 659. space in grates, 1329. spec. ind. cap. of, 35. Alarm, fire, U. S. Navy, 1210. Alcohol, spec. ind. cap. of, 37. All-day efficiency of transformers, 454. Alloys of copper, conductivity of, table of, 910. of copper, table of, 144. phys. and elec. prop, of, table of, 134-140. Alternating circuits, power in, meas of, 69. current ammeters, use of, 945. arc circuits, reactance coil for, 466. arc lamps, 568. armatures, 410. circuit breakers, design of, 952. circuits, protection against abnormal potentials on, 981. circuits, prop, of, 259. definition of, 501. distribution, pressure for, 261. electrolysis, 860. electromagnets, 127. flow, formula for, 1213. lines, table for calc, 279. meas. of, 26, 42. motor equipments, weight of, 719. motors, 421. potential regulators, 467. power curves, 70. railway motor characteristic; 713. system, 707. 1533 1534 INDEX. Alternating current railway trolleys, 640. meas. self-induction with, 66. single-phase sub-station, views of, 943. switchboard panels, 912. voltage and current in terms of D.C., 438. wiring examples, 272. Alternators, parallel running of, 419. regulation tests of, 382. regulators for, 409. revolving field type, 409. armature reaction of, 414. connected in multiple, 420. definition of, 502. E.M.F. of. 404. Aluminum and copper compared, 195. alloys, spec, gravity of, 1514. bar data, 911. conductors, calc. of, 277. fusing effect of current on, 217. phys. and elec. prop, of, 134. production of, 1238. spec. res. of, 132. temperature coef. of, 133. wire, cost of, 195. deflection in feet of, 226. for high tension lines, 199. limit of sag for, 225. properties of, 194. reactance factors for, 266. skin effect factor for, 238. stranded, dimensions of, 197. table of resistance of, 196, 198. weather-proof, 197. Alundum furnace, 1245. Amalgamating zinc, 14. Am. Inst, of Elec. Eng., rules of, 500a. copper wire tables of, 146. Ammeters, A. C. type, use of, 945. and voltmeters, meas. res. with, 78. Bristol recording, 1036. description of, 41. differential, use of, 903. jacks for, 922. scales of, figuring of, 946. shunts for, 41. soft iron, 41. Ammunition hoist, electric, 1147, 1191. Ampere, definition of, 5. -hour meter, Shallenberger, 1028, international, def. of, 9. measurement of, 10. specification for determining, 10< value of, 8. -turns for armature teeth, 367. in field magnets, 366. of A.C. armatures, 414. of air-gap, 369. of electromagnets, table of, 114. of plunger solenoids, table of, 128. Analyses of boiler feed waters, 1366. of coals, 1352. of coke, 1353. of gaseous fuels, 1357. Anchorage of trolley wires, 637. Anchored lamps, navy spec, for, 1173. Angle of lag in three-phase circuits, 406. Angular distance between brushes, table of, 344. velocity, 3, 1505. Anilin, spec. ind. cap. of, 37. Animal oils, 1497. Annealing of armor plate, electric, 1274. Annual expenses of telephone cables, 1087. Annunciator wiring, 294. Anode, definition of, 1229. impurities, effect of, 1237. Answering jacks, 1091. Antenna, 1057. Anthony bridge, diagram of, 31. Anthracite, properties of, 1351. sizing tests of, 1354. Anti-cathode, use of, 1248. Anti-coherers, 1066. Antimony, phys. and elec. prop, of 134. spec. res. of, 132. temperature coef. of, 133. Anylene, spec. ind. cap. of, 37. Apothecaries' measure, 1500. Apparent power, def. of, 50 . INDEX. 1535 Arachid oil, spec. ind. cap. of, 37. Arc, chemical effect of electric, 1232. circuits, reactance coil for A. C, 466. dynamo, efficiency curves of, 338. ext. characteristic curve of, 337. permeability curve of, 338. lamps, candle-power of, 579- classification of, 568. regulation in, 576, trimming of, 583. light carbons, tests of, 577. circuits, ins. res. of, 81. efficiency, 580. installations, table of, 598. rectifiers, G.E. mercury type, 480. station lightning arrester, 986. switchboards, 922. type furnace, 1244. Ardois signal system, 1181. Armature coils, allowable number of turns for, 374. coils, placing of, 358. trial slots for, 373. values for number of, 373. wire for, 372. conductors, carrying cap. of, 375. drag on, 351. size of, data on, 358 . commutation, 364. cores, data on, 357. disks for, 356. energy dissipation in, 107. hysteresis in, 341. magnetic density of, 357. faults, tests for, 402. ground, test for, 402. losses, formula for, 358. reaction, 350. data on, 364. in alternators, 414. resistance loss, meas. of, 509. meas. of, 79, 401. shafts, 341. slots, design of, 357. sizes of, 372. teeth, ampere turns for, 367. teeth, design of, 357. winding, 342. Armature coils, constants, 376. for converters, 441. Armatures, disk type, 341. drum type, 341. heating of, 349. of alternators, 408. copper loss in, 407. winding of, 410. ring type, 341. slotted or toothed type, 341. temperature rise in, 358. ventilation of, 350. Armored submarine cables, 189. Armor plate, annealing of, electric, 1274. Army, U. S., use of elec. in, 1123. Artificial light needed in each month, 606. Ash in American coals, 1350. A.S.M.E. boiler test rules, 1384. direct connected sets, standards of, 1435. Astatic galvanometer, Kelvin type t 23. needle system, 23. Atkinson repeater, 1048. Atmospheric discharges, 1278. electricity, effect on transformers of, 449. Auto-coherers, 1066. Automatic block signalling, 622. booster, use of, 892. exchange systems, 1105. telephone system, 1122. Automobile batteries, 1227. electrolyte for, 877. electric, 1224. motors, 1227. power required by, 1224. Auto-starter, connections of, 954. -transformer, def . of, 503. railway control, 767. use of, 429. Auxiliary armature coils, 351. bus bars, 935. control system, 767. D.C. circuits, 939. power, 867. relays, 956. trunk signals, 1096. 1536 INDEX. Average dynamo efficiencies, table of, 377. Avoirdupois measure, 1500. Axle welding, electric, 1272 Ayrton and Mather shunt, 29. and Perry secohmmeter, 69. and Perry standard of self- induction, 66. and Sumpner method, A.C. power by, 71. and Sumpner test of transformers, 496. method, location of crosses in cables by, 327. Backward lead of motor brushes, 353. Balance coil for three-wire system generator, 355. Kelvin electric, 43. method, determ. magn. values by, 91. Balanced three-phase circuit, energy in, 405. three-phase system, 73. Balancing circuits by transposition, 285. magnetic circuits in dynamos, 349. resistance for arc lamps, 581. transformers for three-wire second- aries, 472. Baldwin Locomotive Works, power tests at, 1517. Ballistic galvanometer, 25. method, determ. magn. values by, 91. B. A. Ohm, value of, 131. Bare wires, carrying capacity of, 208. Barie, value of, 7. Barn test for motor efficiency, 803. Barometric correction, 519. Barrel armature winding constants, 376. Bars, commutator, number of, 361. Baths for plating, 1233. Batteries, automobile storage, 1227. dry, descr. of, 18. E.M.F. of, meas. of, 62, 74. E.M.F., comparison of, 76. Batteries, grouping of, 19. ins. res. of, meas. of, 87. primary, action of, 14. resistance of, 60. Battery capacity for given dis- charge, 900. charging with arc rectifiers, 482. chloride of silver type, 16. equipment, installation of, 897. plates, appearance of, 874. buckling of, 881. cadmium test of, 878. dimensions of, 883. types of, 874. system, three-wire, 899. transmitters, 1071. troubles, 881. while working, res. of, 61. Battle order indicators, U. S. Navy, 1202. service, navy, 1153. Beams and channels, Trenton, safe loads on, 1313. spacing of, 1315. bending moment of, 1308. breaking load on, 1309. coefficient changes for special forms of, 1311. coefficients for special cases of,1311. deck, 1314. deflection of, 1309. flexure of, 1308. general formulae for, 1309. max. moment of stress of, 131Q. modulus of rupture of, 1308. of uniform cross-section, tra>-,d. str. on, 1309. of uniform strength, 1312. resisting moment of, 1308. safe load on steel, 1310. on southern pine, 1320. on wood, 1318. spacing of, for various loads, 1»- 5. strength of white pine, 1319. transverse strength of, 1308. Bearing friction in dynamos, 386- friction, meas. of, 509. Bearings, meter, 1009. Bell telephone receiver, 1070. Bell wiring, 293. INDEX. 1537 Bells for cable ends, des. of, 333. U. S. Navy, spec, for, 1211. Belting, leather, 1487. horse-power of double, 1489. of leather, 1489. of single, 1489. strength of leather, 1487. Belt, length in roll, 1489. length of, 1489. -off test of motor, 396. -on test of motor, 396. slip of, 1493. • eight of leather, 1489. width for given h. p., 1488. Bending moment of beams, 1308. Bends, pipe, dimensions of, 1431. Benzene, spec. ind. cap. of, 37, 227. Bernardos system of welding, 1274. Bessemer steel, phys. and elec. prop. of, 135. Biased bells, use of, 1103. Bipolar dynamos, armature wind- ings for, 345. Birmingham wire gauge, 141. Bismuth, phys. and elec. prop, of, 135. spec. res. of, 132. temperature coef. of, 133. Bituminous coal, properties of, 1351. Blacksmith shop machinery, power to run, 1519. tools, power required for, 1522. Blake transmitters, 1072. Bleaching process, 1244. Block signalling, automatic, 622. system, distributed signal, 627. Blondel oscillograph, des. of, t ?, *-54 Blowers, effect of temperature of air on load of, 1346. for forced draught, 1344. Board of Trade, boiler rules of, 1332. regulations, 781. Boat cranes, navy spec, for, 1194. Bodies of cars, weight of, 734. Body of car, preparation of, 745. Boiler feed water, 1362. purification by boiling of, 1365. flues, collapsing pressure of, 1429. head stays, 1333. plate, ductility of, 1333. Boiler rules, U. S. statutes, 1332. settings, 1334. dimensions of, 1336. shell, strength of riveted, 1330. shop machinery, power to run, 1519. strength of riveted shells of, 1330. test codes, 1384-1392. tests, A.S.M.E. code, 1384. tools, power required for, 1522. tubes, charcoal iron, sizes of, 1428. collapsing pressure of, 1429 Boilers, steam, 1327. heating surface of, 1328. horse-power of, 1327. points in selecting, 1327. safe working pressure for, 1330 . types of, 1327. working pressure of, 1330. Boker & Co.'s wire, properties of, 202. Bolt and nut machinery, power required for, 1522. Bolts, strength of, 1431. Bonded joints and rails, rel. value of, 780. rails, electrolytic action on, 855. Bonding car tracks, 771. condition of track, 800. third rail, 778. Bonds, efficiency of, 781 . requirements for, 775. resistance of, 776. testing rail, 801. tests of, 773. types of, 772. Booster calculations for railways, 810. characteristics of, 813. comparison, 897. controlling discharge by, 889. D.C. type, 435. definition of, 50 * diagram, 810. for street railways, 807. shunt and automatic types of, 892. 1538 INDEX. Booster system, constant current, diagram of, 901. temperature rise in, 814. Boring machines, power required by, 1520. Boston Edison Co., conduit constr., cuts of, 309-313. Boulenge* chronograph, 1128. Box poles, 632. Braces, boiler head, 1334. diagonal, 1334. direct, 1334. Bracket construction, 644. Brackets for trolley lines, 635. Brake controllers, list of, 755. Brakes failing to operate, 805. Braking of cars, emergency, 731. Branch terminal telephone system, 1093. Branding irons, electric, 1270. Brass, rolled, composition of, 1323. weight of sheet and bar, 1323. Brazing by Voltex process, 1274. Break-down point test of synchron- ous motors, 399. -down tests for high voltages, 233. in armature lead, test for, 402. Breaking load of beams, 1309. Breaks in cables, location of, 327. Breast water-wheels, 1476. Brick chimneys, cost of, 1343 foundations, 1292. manholes, cost of, 303. stone, mortars, crushing load of, 1322. work, 1321. Bricks, number in wall, 1321. sizes of, 1321. weight and bulk of, 1322. Bridge, Carey-Foster, 58. connection, multiple unit system, 765. for supporting trolley, 648. method, meas. cap. by, 64. meas. mutual ind. by, 68. slide-wire, 58. Wheatstone, 31. Bridging party lines, 1111. telephone system, 1076, 1093. Brill cars, dimensions of. 737. Brilliancy, intensity of, 599. Bristol recording meters, 1036. British standard candle, 530. thermal unit, 3, 1511. Brooklyn bridge, electrolytic action on, 858. Brooks Potentiometer, 49. Brown & Sharpe wire gauge, 141. wire gauge, law of, 142. Brown rail bond tester, 802. Brush contact, friction of, 362. resistance loss, meas. of, 509. discharge from wires, 235. faces, area of, 361. drop in volts at, 362. losses at, 362. friction, 384. meas. of, 50 . machine, ext. characteristic curve of, 337. Brushes, angular dist. between, table of, 344. armature, 351. carbon and copper, res. of, 362. current densities for materials for, 442. forward lead of, 350. motor, position of, 353. B.T.U., gas and elec., ratio between, 1261. Bucking of cars, 806. Buckling of battery plates, 881 . Buildings, design of power station, 866. electrolysis in steel frame, 859. Bunsen cell, 14. photometer, 535. Burglar alarm mats, wiring of, 295. Burnley dry cell, 18. Burton electric forge, 1274. Bus bars, arrangement of, 933. copper for, 910. data on, 911. design of, 933. high-tension station, 933. structure of, 935. Bushel, 1499. Busy test, 1091. Button transmitters, 1074. Buzzer, field. 1140- INDEX. 1539 <«it»i not shop, power required tor, 1520 Cable connectors, Seeley's type, 190. ends, bells for, 333. insulating, 322. flexible dynamo, table of, 172. heads, 320. joints, Dossert type, 191. loss of power in lead sheath of, 293. machinery, depreciation of, 770. three cond. white core ins., table of, 170 Cables, break-down of insulation of, 980. breaks in, location of, 327. capacity of lead sheathed, 251. conductivity of, meas. of, 330. current carrying capacity of, 208. dielectric tests of, 332. distributing, 1083. drawing in underground, 319. faults in, location of, 328. for car wiring, prop, of, 173. high-tension, insulation of. 939. G. E. rubber ins., tables of, 164- 172. gutta-percha insulated, 232. in ducts, heating of, 210. ins. res. of, tests of, 321. joints in paper insulated, 191. lead covered, carrying cap. of, 213. locating crosses in, 327. paper and lead cov., tables of, 174-178. paper insulated, properties of, 1083. rubber covered, 161. rubber insulated, carrying cap. of, 210. rubber insulated, joints in, 190. submarine, 189. testing of, 331. telegraph, 189. telephone, 188, 1082. capacity of, 1085. expenses of, 1087. size of, 1086. testing cap. of, 325. joints of, 323. Cables, three-phase, power carrying cap. of, 216 twisted pair. 1082. types of underground 320. underground and submarine, tests of. 321. varnished cambric ins , tables of. 178a- 187a. varnished cambric ins , triple cond., 185. watts lost in. 210. Cadmium cell, Weston. 19. phys. and elec. prop, of, 135. test of battery plates, 878. Calcium carbide, production of. 1245. Calcspar, spec. ind. cap. of, 36. Calculation of A.C. lines, 279. of dynamo efficiencies, 391. of transmission lines, 264. Calibrating jacks, 941. Calibration data for Westinghouse wattmeter. 1016. of A.C. instruments, 26-28. of wattmeters, checking of, 1014. Calico printing machinery, power to run, 1525. Call bells, wiring of, 293. U. S. Navy spec, for, 1211. Calling apparatus, 1075. Calorie, 3, 1511. Calorimeter, directions for use of, 1395. separating. 1398. throttling, 1394. diagram of, 1399. Cambric ins. cables, D.C., tables of, 178a- 187a. Canals, constr. of, 868. Candle-foot, 525. -hours, variation in, 546. -meter, 525. Candle-power, meas. of, 584. Navy spec, for, 1171. of arc lamps, 579. of coal gas. 1450. of lamps, 544. of searchlights, 1127. standard of, 52 . Candy manufacture, elec. in, 1270. 1540 INDEX. Canning industry, electric heat in, 1270. Caoutchouc, spec. ind. cap. of, 36, Capacitance of transmission circuits, 249. Capacity and inductance, neutrali- zation of, 292. curves of railway motors, 676. definition of, 5. distorting effect of, 1079. effect of line, 264, electrostatic, measurement of, 40. Gott method, 326 Kelvin method, 326. loss of storage batteries, 881. measurement of, 63. meas. coef . of induction by, 65. measures of, 1499. of A.C. circuits, 259. effect of, 1216. of battery for given discharge, 900. of cables, direct discharge method, 325 Gott's method, 326. meas. of, 324. Thomson's method, 325. of gases, spec, inductive, 35, of liquids, spec, ind., table of, 37. of railway motors, 673. of solids, spec, ind., table of, 36, 37. of storage batteries, 874, 883- of telephone cables, 1085. of transformers, choice of, 458. table of, 498. of transmission circuits, 248. of various overhead transmission lines, 250. per 1000 feet of aerial wires, table of, 252. reactance, 259, reactance of transmission circuits, 248. spec, ind., measurement of, 38. susceptance of transmission cir- cuits, 249. susceptance, table of, 269. symbol of, 8. tests for locating breaks in cables, 327 Capacity tests with Lord Kelvin'* dead-beat voltmeter, 326. unit of, 4 Carcel lamp, 530. Car bodies, weight of, 734 body, preparation of, 745. controllers, 753- energy consumption per, 652. input to, 657 equipments, 613, 752, heaters, cross seat type, 1268. hints to purchasers of, 1269. truss plank type, 1267. heating, cost of, 1266. electric, 770, 1265. lighting, G. E. railway system, 851. motors, installation of, 745. test of, 392. tests, interurban, 722. wiring, 746. for heaters, diagram of, 1267. special cables for, prop, of, 173. Westinghouse railway system, 846. Carbide furnace, King, 1245. Carbon brushes, current density for, 442. res. of, 362. use of, 351 Carbon dioxide, spec. ind. cap. of, 35. disulphide, spec. ind. cap. of, 35. dust, 578. effect on steel of, 826. monoxide, spec ind. cap, of, 35. spec. res. of, 132. Carbons for enclosed arc lamps, 578. for search lights, 579, 1125. resistance of, 577. sizes of, 578. test of arc light, 577. Carborundum, production of, 1245. Care of storage batteries, 1228. Carey-Foster method, meas. res. by, 58. Carhart-Clark cell, des. of, 19. Carpenter's hrottling calorimeter 1395. curves. 139^, INDEX. 1541 Carpenter's throttling Calorimeter, directions for use, 1395. Carrying capacity of armature con- ductors, 375. of fuses, 1275. galv. iron wire, 34. of lead covered cables, 213. of rubber ins. cables, 210. of wires, 208. Cars, bucking of, 806. depreciation of, 770. dimensions of electric, table of, 732. emergency braking of , 731. energy required for, 679 . lighting of, 806. power required for, 656. speed and energy curves for, 680. Cartridge fuses, 1276. Carty bridging bell, 1102. Cascade, cap. of condensers in, 324. Cast iron magnet shoes, 352. permeability of, 89. phys. and elec. prop, of, 137. test of, 1294. water main, electrolytic action on, 854. Castner metallic sodium cell, 1242. process, caustic soda by, 1240. Castor oil, spec. ind. cap. of, 37. Cast steel, permeability of, 89. rope, wire, 1325. Catenary trolley, bridge for, 648. construction, 639. material for, 643. Cathode, definition of, 1229. rays, theory of, 1248. Caustic soda, production of, 1239. Cell, Burnley type, 18. Carhart-Clark type, 19. chloride of silver type, 16. Edison-Lalande, des. of, 17. Fuller, description of, 16. Gasner type, 18 grouping, efficiency of, 21. Leclanche*, des. of, 16. standard, construction of, 11. description of, 19. spec, for, 10. Weston cadmium type, 19. Cells, Clark type, 19. closed circuit, table of, 14. grouping of, 19. open circuit, 15. Celluvert, spec. ind. cap. of, 36. Cement, adhesion to bricks of, 1294. and sand, fineness of, 1294. crushing load of, 1322. hydraulic, strength of, 1294. mortar, 1293. Portland, strength of, 1294. wt. of, 1293. Rosendale, wt. of, 1293. strength of neat, 1294. Centering of armature, 403. Center of gravity of distribution system, 277. pole line construction, 631. Centi-ampere meter, balance used as, 43. Centigrade vs. Fahrenheit scale, 1508. Centimeter, definition of, 2. Central battery system, 1096. energy system, 1096. office apparatus, 1104. offices, adv. of one vs. several, 1094. R.R. of N.J. shops, power to run tools in, 1520. station battery connections, 899. electrically operated switch- board, 928. lightning arresters in, 983. switchboard panels, 907. three wire battery system, 903, vs. isolated plant, 1286. telephone office, 1089. Centrifugal tension in Manila ropes, 1491. C.G.S. units, 2. names of, 6. Chain, 1496. coil, 1496. proof, 1496. short link, 1496. weight of, 1496. Characteristic curve, external dynamo, 337. of dynamo, plotting of, 382. 1542 INDEX, Characteristic curve of N.Y.C. loco- motive, 742. of railway motor, 664. curves of over-compounded dynamo, 340. of solenoids, 129. Characteristics of electromagnets, 129. of G.E. single-phase motor, 713c of railway booster, 813. of railway motors, 685. of transformers, 483. of two-path armature winding, 348. of Westinghouse single-phase motor, 715. Charcoal rope, wire, 1325. Charge curves, storage battery, 876, of storage battery, loss of, 884. rate for batteries, 883. Charging batteries, 482. batteries, connections for, 899. current of line wave, 249. per 1000 feet of aerial circuit, 253. of storage batteries, 880. . Chart for calculating A.C. lines, 282. of parabolic curves in wire spans, 218. Chase-Shawmut fuse wire, 1275 Chatterton's compound, specifica- tions for, 194. Checking preliminary dynamo di- mensions, 363. wattmeters, 72. Chemical action in cells, 14. equivalent of elements, 1230. properties of rubber, 229. qualities of steel for third rail, 822. Chimney construction, 1339. weight for burning given amounts of coal, 1342. protection, 1281. rate of combustion due to height of, 1342, tables, 1338. Chimneys, 1338. dimensions and cost of, 1343. draught power of, 1338. iron, dimensions and cost of, 1344. Chimneys, necessary height of, 1342 radial brick, 1341. and bond, 1340, size of, 1338. steel, foundations for, 1343. lining for, 1343. plate, 1343, Chloride of silver cell, 16. Chlorine in electrolyte, test for, 878 Choke coils, mountingof, 984. S.K.C. arrester, 991. use of, 994. Choking effect of inductance, 1079. Chord of polar arc, values of, 371. of pole face, dimensions of, 363. Chrome-bronze, phys. and elec. prop of, 135. -steel, phys. and elec. prop, of, 135. Chronographs, types of, 1128. Chronoscope, Schultz, 1130. Circuit breaker, def . of, 52 design, 952. Westinghouse oil, 969. Circuit breakers, capacity of, 955. for booster protection, 952. for motors, capacity of, 955. for protection of transmission line, 951 c for railways, use of, 789. for storage battery protection, 952. grouping of, 929. leads for, 975. mounting of, 912. oil, arrangement of, 935. polyphase motors protected by, 954. rating of, 50 , 912. specifications for, 947 table of, 949. Circuit closer, torpedo, 1139. trunks, operation of, 1095. Circuits in buildings, ins. res. of, 85. laws of electrical, 55 multiple, res. of, 55. testing drop in railway, 804. tests of street railway, 798. Circulating pumps, 1445. Cities, electrical distribution in, 261. mill power in various, 1462. INDEX. 1543 Clark cell, description of, 19. E.M.F of, 5. method, comparison of E.M.F. by, 77. testing joints of cables by, 323. Clay conduits, constr. of, 301. foundations on, 1290. Clearing-out drops 1090. Closed cars, weight of, 734. circuit cells, table of, 14. Coal, American, heating value of, 1350 and electric heating compared, 1265. anthracite, sizing tests of, 1354. approximate analysis of, 1352. consumed by isolated plant, 1286. gas, analysis of, 1510. candle power of, 1450. spec. ind. cap. of, 35. heating value of, 1349. power, data on, 869. space to store, 1353. value of in weight of woods, 1349. weight per cubic foot of, 1353. Coals, relative value and how to burn, 1355. Coast-defense board, recomm. of, 1123. guns, manipulation of, 1134. Coasting, formula for, 668. line, location of, 668. Cocoa and coffee dryers, electric, 1270. Codes, telegraph, 1052. Coefficient of induction, meas. of, 65, of induction, symbol of, 8. of self-induction, 64. def . of, 238, formula for, 405 of temperature of metals, 133. Coefficients of expansion of solids. 1508. of magnetic eakage, 376. of reflections, 593. Coercive force, def, of, 108. Coffee and cocoa dryers, electric, 1270. Coherer, 1058. receivers, 1064. Coherers, mercury auto, 1066. Coil chain, 1496. slots, design of armature, 358. trial armature, 373. surface of field magnets, 352 Coils, armature, placing of, 358. for transformers, 444 heating of, 127. values for number of armature, 373. winding of, 112. Coke, analysis of, 1353. space required for, 1353. weight per bushel of, 1353. Collier & Sons' factory, heating devices in, 1270. Columns, comparison of water, 1463, hollow, 1305. cylindrical, 1306. pillars or struts, 1300. solid cast iron, 1305. strength of white pine, 1319. solid cast iron, 1305. tests of cast iron, 1306. ultimate strength of, 1306. wrought iron, ult. strength of, 1307. Colza oil, spec ind. cap. of, 37. Combinations of railway motors, 760. Combined volt and ammeter method, meas. A.C. power by, 71. Combustibles, properties of, table of, 1348. Combustion, draught necessary for, 1342, Commercial efficiency curve for are dynamo, 338. efficiency curve for motors, 370. of dynamos, def. of, 383. lights, burning of, 611. rating of railway motors, 675. transformers, 445. Committee on Notation, table by, 6. Common battery system, 1096, 1115. signaling battery system, 1115. trunks, 1096. Commutated rotor windings, 429. Commutating machines, def. of, 504. zone, 350. 1544 INDEX. Commutation in dynamos, 364. Commutator bars, number of, 361. brushes, sparking at, 805. brush friction, meas. of, 509. diam. of, 361. rise of temperature of, 362. segments, number of, 361. type, D.C. meters, 997. Commutators, construction of, 351. Comparative cost of gas and elec. cooking, 1260. expense of operating transformers, 458. values of lighting methods, 594. Comparison of copper and aluminum wire, 195. of interurban car tests, 724. Compensated A.C. motor character- istics, 713. revolving field alternators, 409. Compensation for power factor, 1002. method, E.M.F. of batteries, 62. Compensator regulators, definition of, 503. use of starting, 918. Compensators, construction of, 463. for induction motors, 429. Composite electric balance, 43. Compound cables, design of, 331. dynamos, characteristic of, 340. des. of, 336. regulation tests of, 382. engines, cylinder ratios for, 1441. Compressive strength of woods, 1317. Concealed lighting system, 601. Concentric cable, capacity of, 251. Concrete foundations, 1292. manholes, cost of, 303. reinforced, 1292. sub-foundations, 1292. Condensation in steam pipes, 1415. in steam pipes aboard ship, 1415. Condenser capacities, ejector, 1445. current, curve of, 1219. diagr. of connections of, 39. method, res. of batteries by, 60. unit, 5. Condensers and pumps, 1443. construction of, 38. Condensers, cooling water by, 1444. design of, 35. in cascade, cap. of, 324. in parallel, 63. cap. of, 324. in series, 63. cap. of, 324. losses in, meas. of, 513. Condensing engines, number of expansions in, 1441. Conductance, definition of, 9, 55. of multiple circuits, 55. symbol of, 8. Conducting power of sulphuric acid, table of, 905. Conductivity, definition of, 9. Matthiessen's stand, of, table of, 132. millivoltmeter meas. of, 87. Northrup method of, meas., 60. of cables, meas. of, 330. of conductors, table of, 132. of copper, 518, 910. of dielectrics, specific thermal 234. percentage, form, for, 132. relative, 132. specific, 132. symbol of, 8. Conductor rail, Potter type, 830. Conductors, carrying cap. of arma- ture, 375. dimensions of, 260. economical tapering of, 279. for electric railways, overhead, 785. for high-tension, insulation of, 939. for high-tension transmission, 235. for parallel D C. system, size of, 284. for railways, dimensions of, 791. installing, U. S. Navy, spec, for, 1170. isolation of, 936 per K. W. del'd, curves showing weight of copper, 283. res. of, 61. rotation around pole of, 109. spec. res. of, table of, 132. INDEX. 1545 Conduit, Board of Trade regulations for, 783. construction, U. S. Navy, 1170. cost of estimating, 317. itemized, 316. total, 307 def. of, 301. foot, cost per manhole of, 304< cost per, table of, 306. in cities, cost per, table of, 307 laying of, 301. multiple duct, constr. of. 301. New Orleans, 308. systems, heat dissipation in, 214. railway, 835. work, usual practice of, 302. Conduits, Chicago, underground, cost of, 317. cost of, 302. monolithic, des. of, 301. multiple, adv. of, 301. single duct, adv. of, 301. Connecting transformers to rotary converters, 476. Connection of batteries, 19. Connections of polyphase meters, checking of, 1026. of transformers, 297, 472. on switchboards, 910. Connectors, Seeley's cable, 190. Consolidated car heating, wiring diag. of, 1267. Constant current booster system, diagram of, 901. current from constant potential transformers, 464. current machines, regulation of, 513. current transformer panels, equip. of, 922. current transformers, G. E. type, 464. galvanometer, 23. hysteresis, wattmeter test for, 102, potential arc lamp, 574. machines, regulation of, 513 secondary current, transformers for, 462. Constantin wire, properties of, 202. Constants for barrel armature wind- ing, 376. hysteretic, table of, 99. of meters, values of, 1029. Construction of chimneys, 1339. of manholes, cuts of, 309. power station, chart of, 1289. tools, electric work, 1530. Consumption of energy of cars, 652. of energy of elec. heaters, 1265, Contact buttons, Westinghouse railway system, 844. plates, Westinghouse railway system, 841. Contactors, multiple unit system, 762. Continental code, 1052. Control of lights from two or more points, 294. of motors, Ward Leonard's sys- tem, 354. of water-tight doors, 1198. Controller, care of, 747. combination, 760. for oil circuit breaker, 975. series-parallel, 753. Controllers, dimensions of, 757. G.E. railway system, 851. installation of, 746. Controlling desks, 941. discharge, methods of, 888. panels, Navy spec, for, 1185. pedestal, 940. switchboards, 940. Convectors and radiators, 1263. Converter armature windings, 441. definition of, 503. panels, three-phase rotary, equip, of, 919. Converters connected to transform- ers, 477, rotary type, 436. Conveyors, ammunition, U. S. Navy, 1193. Cooking apparatus, electric, effi- ciency of, 1260 electric, cost of, 1259. gas and elec. compared, 1260. record, daily electric, 1262. utensils, electric, cost of operating, 1261. 1546 INDEX. Cooling surface of field coils, table of, 352. tower test, 1447. of transformers, 448. water for condensers. 1443 Cooper-Hewitt mercury lamps, 558 Conductivity, 132. Matthiessen's Standard, 132 Copper, admixture of, effect of, 144. and aluminum compared, 195. and brass wire and plates, weight of, 1324. bar data, 911. bars on switchboards, 909. brushes, current density for, 442. res. of, 362; use of, 351. conductivity of, 518, 910. electric welding of, 1272. electrolytic refining of, 1235. for A. C. lines, table for calc. of, 279. fusing effect of current on, 217. in railway feeders, 791. loss in alternator armatures, 407. in transformers, 445. in transformers, meas. of, 487. in transformers, Sumpner's test of, 497. in transformers, table of, 498. melting point of, 143. phys. and elec. prop, of, 135. plating, 1233. res. of cables, meas. of, 330. rise in resistance of, 379. spec. res. of, 132. strands, stand, prop, of, table of, 159. temp, coefficient of, 133, 52 , weight of, 143. of round bolt, 1323. wire and plates, 1324. fuses for railway circuits, 731. Matthiessen's form, for, 133. phys. const, of, 143. res. of, table of, 148. skin effect factor for, 238. solid, G. E. Co., prop, of, table of, 162. solid, table of, 154. stranded, table of, 155. Copper wire tables, A.I.E.E , 146. tables, explan. of, 145 tensile strength of, 156. weight of, English system, table of, 157. weight of, metric system, table of, 158. Core disks for armatures, 356. insulation, armature, 341. losses, 98. in armature, 360. in transformers, 445. in transformers, comparative, 455. in transformers, curves of, 454, 456. in transformers, meas. of, 485. in transformers, table of, 498. loss test, 383. of stator and rotor, 425. of submarine cables, design of, 331 . of three-phase- transformers, 470. type transformers, coils for, 444. Cores, cross-section of, 365. field magnet, general data on, 352. magnetic densities for transformer, 447. of armatures, data on, 357. of American transformers, types of, 443. Corey telephone system, U. S. Navy, 1209. Corn plaster transmitters, 1074. Cos a, values of, 276. Cost of aluminum wire, 195. of conduit, 302. estimating, table of, 317. itemized table of, 316. total, 307. of cooking daily meal by elec, weekly record, 1262. of duct material in place, table of, 307. of electric car heating, 1266. of 5' X 5' X 7' manhole, 316. of heating water by electricity, 1259. of incandescent lamps, 556. of manhole, estimating, table of, 317. INDEX. 1547 Cost of manholes, table of, 302 of one mile of trolley system, 629. of operating electric cooking utensils, 1259, 1261. electric elevators, 1528. elec. heaters, 1265. elec. irons, 1263. lamps, 554. of paving per sq. yd., 305 of power, curves for reducing, 868. of protected third rail, 835. of sewer connections, 303, of street excavation per conduit foot, 306. of telephone plant, 1108. of tools and supplies for installing electric work, 1531. per conduit foot for manhole, 304, per conduit ft. in cities, table of, 307 per conduit foot, table of, 306. Costs, comparative, gas and elec. cooking, 1260. Cotton covered wires, linear space occupied by, tables of, 121-126, covered wire, diam. of, 163a. machinery, power to drive, 1524. Coulomb, definition of, 5. international, def. of, 9. value of, 8. Counter cells, use of, 891. e.m.f. cells, use of, 891. e.m.f. in motor armatures, 353. torque, meas. of, 396. Cove-lighting, 592. Cover for service boxes, 315. Covers for manholes, cuts of, 313- 315. Cowles furnace, 1247. Crane chain, 1496. Cranes, boat, Navy spec, for, 1194. power to run electric, 1527. Cross connections, use of, 1104. seat heaters, wiring diag. of, 1268 . section of field core, 365. of conductors, calc. of, 277. of conductor, formula for, 265. -talk, definition of, 1081. elimination by transposition of, 289. Crosses in cables, location of, Ayrton method, 327 Crossings of wires, 639 Crushing loads for brick, stone, mortar, cement, 1322. strength of woods, 13 16. Cubic feet table, water ho p., 1475. Current carrying capacity of lead covered cables, 213. carrying capacity of rubber ins, cables, 210 carrying capacity of wires, 208. curve for railway motors, 669. definition of, 50^.. densities for transformer coil, 447, for various brush materials, 442. density at brush faces, 361. for brushes, 351. for commutator segments, 361 . distribution by railway conduc- tors, 791. in cables, max. allowable, 212- in multiple circuits, 55, in three-phase circuit, meas. of, 406. maximum, A. C. windings, 127. mean, A. C. windings, 127. measurement of, 41. millivoltmeter meas. of, 78. of alternators, 405. potentiometer meas. of, 47, 63. swapping, 859. taken by induction motors, 297. by lamps, 542. transformers, descr. of, 945. unit of, 4. . variations on water main, 857. voltmeter meas. of, 77. wave form of, " 503. telegraph, field, 1140. transposition to eliminate, 285. type furnace, 1244. wattmeters, Westinghouse, 999, 1003. wattmeters, Thomson polyphase, 1005. Inductive capacity, spec, def. of, 38. capacity of gases, values of, 35. of substances, table of, 36, 37. circuits, wattmeters on, 1000. drop in trolley, 797. effect of alternating currents, 236. load, def. of, 50£ regulation of transformer for, 492. loads, testing meters on, 1018. reactance, formula for, 239. In ohms per 1000 feet, 242. Inductive reactance fn solid iron wire, table of, 248. in three-phase line, 245. representation of, 259, Inductor alternator, def. of, 502. type synchroscope, 417. Industrial electric heating, 1269. Inertia, moment of, 1302. of rotating parts of train, 683. Ingredients of rails, table of, 780. Injectors, deliveries by live steam, 1371. exhaust, 1372. lifting cold water by, 1372. hot water by, 1372. live steam, 1370. performance of, 1371. vs. pumps for boiler feeding, 1372. Installation of battery plants. 897. of car motors, 745. of fuses, 1276. of polyphase meters, 1023. of storage batteries, 885. Instantaneous relays, 956. value of E.M.F.. 404. Instrument posts, 941. scales, figuring, 946. Instruments, electrical measuring, 21. for switchboard, 940, 945. testing, description of, 13. Insulated cables, varnished cambric, triple conductor, 185. cables, varnished cambric, tables of, 179-183. copper wires and cables, table of, 160. wires and cables, rubber cov., tables of, 164-172. carrying capacity of, 209. locating faults in, Warren's method of, 330. Insulating cable ends for tests, 322. cable joints, 191. ground near power station, 862. joints in mains, 861. materials, dielectric strength of, 228. puncturing voltages for, 225. 1562 INDEX. Insulation across fuse block, meas., of, 82. distances on switchboards, 912. of armature core, 341 . of dynamos, meas. of, 86. of high-tension cables, 939, of transformer, 447. resistance betw. conductors, N. C, 85. by loss of charge method, 322. meas. of, 514. of arc light circuits, 81. of cables, 321. of circuits, meas. of, 80, 85. of dynamos, 86. of motors, 87. of railway lines, 783. of rubber, 231. of telephone cables, 1084. U. S. Navy standard, 1168. of wiring system, 82. test of cables, 332. of dynamos, 381. of rubber, 230. of transformers, 483. Insulators for third rail, 831. Metropolitan street railway, 840. on poles, arrangement of, 291. Integrating meters, action of, 997. meters, Westinghouse, D.C., 998. photometer, 539. wattmeters, data for, 1016. induction type, 999. tests of, 1013. Westinghouse, 1004. Intensity of brilliancy, 599. of current, symbol of, 8. of illumination, laws of, 528. table of, 586. of light, 530. of magnetic field, 4. force, def. of, 108. of magnetization, 4. value of, 7. of searchlights, 1125. Interaxial distances between A.C. conductors, 240. Interborough rail, 830. Intercommunicating telephone sys- tems, 1088, 1114. Inter-connected star arrangement of three-phase transformers, 477. Interior illumination, 596. wiring, carrying cap. of cond- for, 209. Interlock switches for railway con- trol, 768. Intermediate distributing frames, 1104. Internal characteristic of shunt dynamo, 339. resistance of batteries, meas. of, 87. of cells, 20. of storage batteries, 883. International ampere, def. of, 9- ampere, specification for determ., 10. coulomb, def. of, 9. electrical units, 9. farad, def. of, 9. standard, 38. henry, value of, 10. joule, value of, 10. ohm, construction of, 30. def. of, 9. value of, 131. volt, definition of, 5, 9. determ. of, 10. watt, value of, 10. Interpolar edges, design of, 363. Interrupters, Wehnelt, 1254. for X-rays, use of, 1253. Interurban booster calculation, 812. car tests, 722, 725. Intrinsic brightness of sources of light, 529. Inverse time limit relays, 957. Inverted converter, def. of, 436. Inward flow turbines, 1476. Iron ageing tests, curves of, 453. and steel, ageing of, 455. elec. welding of, 1271. magnetic fatigue of, 45^. permeability curves of, 90. wire, constants of, 199. fusing effect of current on, 217 in electrolyte, test for, 877. INDEX- 1563 &*"**> loss curves of Westinghouse motors, 674. determinations, 107. In transformer, table of, 482. In transformer cores, 453. In transformer, Sumpner's method, 496. magnetic properties of, 89. permeability of, 89. meas. of, 94. phys. and elec. prop, of, 137. pieces of, attraction between, 111. pipe, elec. welding of, 1272. plating, 1234. poles, 633, production of, 1247. spec. res. of, 132. stacks, guyed, cost of, 1344. telegraph wire, galv., properties of, 199. temperature coef. of, 133. U. S. standard gauge, weights of, 1299. weight of, 1294. flat per foot, 1295. plate, 1298. square and round, 1297. wire for water rheostats, 34. inductive reactance in, table of, 248. properties of, 199. self induction in, 240. self induction in, table of, 248. skin effect factor for, 238. use in telephony of, 1082. Irons, electric, cost of operating, 1263. soldering and branding, elec, 1270. Isolated electric plants, economy of, 1283. plant, coal consumed by, 1286. vs. central station, 1286. Isolation of conductors on switch- boards, 929, 936. Itemized cost of conduit, tableof 9 316= JTack« for ammeter connections, 922 telephone, 1089. Jamison rule for ins, res., 85 Jigger, use of, 1065. Joint effect of electrolysis, 853, Jointing gutta-percha covered wire, 193. Joints, Dossert cable, 191. in cables, testing of, 323. in mains, insulating, 861. in paper insulated cables, 191 in rubber ins. cables, 190= in Waring cables, 191. per mile of track, 618 rail, tests of, 801. insulating cable, 191. Joly's photometer, 536. Joule, definition of, 3, value of, 5, 8, Joule's equivalent, 4. Jump distance curve, 234. Jumping-point of carbons, 577, Junction boxes, U. S. Navy gpec for, 1171 = Kapp's efficiency test of two dyna- mos, 387. potential regulators, 468. Kempe rule for ins. res. ; 85. Kelvin balance, diagram of, 44. electric balance, 43. electrostatic voltmeter, 40. galvanometer, 23. Kelvin's double bridge, 59. law, 261, 787. applied to booster distribution, 810. multicellular voltmeter, cap. test with, 326. Kerosene for boilers, 1364. Kilowatt curve for railway motors, 669. Kilowatts of energy in three-phase cables, 216. on grades, 657. Kinetic energy, 3. King carbide furnace, 1245 Kirchoff's laws, 55. Knee of saturation curve, 401, Krupp's wire, properties of, 202, 206. Kryptol method, electric heating, 1257 I a lie! rating of gem lamps, 549. Laboratories, electric heat in, 1370. 1564 INDEX. Lagging current, effect of, 439. Lake electric railway, high-speed trials on, 719. Lamination of cores, reason for, 99. Laminations for transformer core, 445. Lamp indication for oil circuit breaker, 975. renewals, 547. signals, telephone, 1098. Lamps, candle-power of, drop in, 544. current taken by, table of, 542. efficiency of, 525. life of, 544. material required for instal. of, 1531. Navy spec, for, 1171. U. S. Navy standard, table of, 1176. Lande cell, 14. Lanterns, diving, 1179. , Lap-connected armature windings, 345. Lateral, def. of, 302. effect of electrolysis, 853. Lathes, power required for, 1516. Law cell, 15. of Brown & Sharpe wire gauge, 142. of induction, 64. of plunger electromagnet, 127. of traction, 110. Maxwell's, 94. Laws, Kirchoff's, 55. of circuits, elementary, 55. Laundry irons, electric, cost of operating, 1263. Layers of cotton -covered wires, space occupied by, tables of, 121-126. Laying out dynamos, procedure in, 370. Lay-overs at end of run, 676. Lead burning, 885. covered cables, carrying capacity of, 213. covered cables, tables of, 174- 178. covering of cable joints, 191. Lead burning, fusing effect of currenl on, 217. of brushes, 350. peroxide, use in batteries of, 873 phys. and elec. prop, of, 137. plates, joining of, 885. sheathed telephone cables, 188. telegraph cables, 189. sheath of cable, loss of power in, 293. spec. res. of, 132. sulphate, use of, 873. temperature coef. of, 133. Leading current, production of, 439. Leads for transformers, 499. Leakage current on railway line, 783. coefficients, magnetic, 376. drop in transformers, 497, of magnetic lines in dynamos, 365. reactance, def. of, 503. Least exciting current of syn- chronous motors, 400. Leclanche* cell, des. of, 16. Leeds & Northrup bridge, 32. Legal ohm, value of, 131. Lemon oil, spec. ind. cap. of, 37. Length, measures of, 1499. of magnet coils, corrections for, tables of, 117-120. of magnet cores, 365. of sparks, curves of, 949. Leonard's system of electric pro- pulsion, 354. of motor control, 354. Le Roy method, electric heating, 1257. Letters, Greek, 1505. Lever switches, 963. Life of carbons, 577. of lamps, 544. tests, Navy spec, for lamp, 1172. Lifting-power of electromagnets, 110. Light and power cables, 320. control from two or more points, 294. cut off by globes, 582. data on, 528. distribution of, 599. heat and power, cost in residences of. 1287. INDEX. 1565 Light, heat and power in isolated plants, cost of, 1285. standard of, 530. units of, 530. Lighting cars, G. E. railway system, 851. circuits, res. of, meas. of, 80. lines, transposition of, 285. methods, comparative values of, 594. of street cars, 806. plant, batteries for residential, 898. schedule for London, 611. schedules, 603. service, navy, 1153. system, U. S. Navy, 1171. Ughtning arresters, arc station, 985. direct current, 984. function of, 980. Garton, 990. General Electric A.C., 987. high potential circuit, 993. horn type, 995. incandescent station, 986. in central stations, 983. in power station, 867. inspection of, 984. insulation of, 984. low equivalent, 994. magnetic blow-out, 987. multiplex three-phase, 988, non-arcing D.C., 984. metal double pole, 989 railway non-arcing, 985. S.K.C., 990. spark gaps of, 991. Stanley, 990. unit, 990. use of, 1087. Wurts type, 984. Lightning flash, data on, 1277. protection, 980. Lightning rods, history of, 1277. installation of, 1278. points of, 1281. tests of, 1282. Lime mortar, 1293. Limestones, crushing load of, 1322. Limitation of voltages, 866. Limit of sag for aluminum wire, 225. Limits of telephonic transmission, 1107. Lincoln synchronizer, 416. Lineal measures, metrical equiva- lents of, 1500. Linear space occupied by d.c. cov. wire, table of, 123-126. s.c. cov. wire, table of, 121-123. Line capacity, effect of, 264. discharger of S.K.C. arrester, 991. drops, 1090. equipment, depreciation of, 770. formulae, transmission, 275. material per mile of trolley, 643. power loss in, 261. pressure, adv. of high, 260. relay for railway control, 769. switch for railway control, 767. wire, weather-proof, table of, 160. Link shoe for third rail, 832. Liquid fuels, 1356. rheostats, 33. Liquids, measures of, 1500. measures of, metrical equivalents of, 1502. specific gravity of, 1512. ind. cap. of, table of, 37. res. of, 133. Load curve, 887. diagram, fluctuating, 888. factor, def. of, 506 of railway system, 785. factors, cost of power at various, 868. hauled by motor car, 655. losses, meas. of, 509. on foundation beds, permissible, 1292. peak, batteries to carry, 886. power factors, 279. steel beams, safe, 1310. test of motors, 395. Loading gear for guns, 1191 telephone lines, 1107. Local action in storage batteries, 878. Locating breaks in cables by cap. test, 327. crosses in cables, Ayrton method, 327, 1566 INDEX. Locating faults in cables, loop method, 328. in underground cables, 331. Location of transformers, 499. Locomotives, electric, 614. electric, table of, 739, tractive coefficient of, 662. Loft building plant, economy of, 1285. London, lighting schedule for, 611. Long distance transmission, data on, 866. transformers for, 474. Loop method, locating- faults in cables by, 328. Lord Kelvin's composite balance, 43. multicellular voltmeter, cap. test with, 326. Lord Rayleigh's method, E.M.F. of batteries, 62. Loss factors, hysteresis, table of, 99. in line, power, 261. of active material in battery plates, 881. of capacity of storage batteries, 881. of charge method, ins. res. by 322. of storage batteries, 884. of head due to bends, water, 1374. of potential method, meas. cap. by, 64. of power in cable sheath, 293. of voltage in storage batteries, 882. Losses at brush faces, 362. core, 98. electrical method of supplying, 389. in armature, formula for, 358. in machines, meas. of, 509. in transformers, 445. comparative, 455. curves of, 453. Lowell mill power, table of, 1464. Low equivalent lightning arrester, 994. resistance detector, 1065. meas. of, 59. tension lamps, 569. voltage A.C. relay, 962. D.C. relay, 961. Lubricants, best for diff. purposes, 1498. Lubrication, 1497, of engines, 1413. of motors, navy spec, for, 1185. Lumen, def- of, 521;, 5C2. Luminosity of inc. lamps, 548. Luminous flux, 529. Lummer-Brodhun photometer, 536. Lumsden's method, E.M.F. of batte- ries, 62. Machine shops, friction load in, 1523. lighting of, 597. men employed in, 1523. power to run, 1518. tools, power to drive, 1515. Magazine light boxes, U.S. navy,1171. Magnesium, phys. and elec. prop, of, 137. Magnet coils, correcting length of, table of, 117-120. coils, general data on, 352. heating of, 127. cores, design of, 365. poles, determination of number of, 355. windings, field, 369. wire, res, of, table of, 112. Magnetic blow-out lightning arrester, 987. circuit in dynamos, balancing of, 349. of transformer core, equation for, 446. principle of, 109. density of transformer cores, 447. of field magnet cores, 365. of armature cores, 357. of armature teeth, 367. of pole faces, calc. of, 356. detectors, 1067. distribution, curve of, 340. fatigue of iron and steel, 455. flux, definition of, 4. formula for, 109. field, intensity of, 4. force, intensity of, 108. gradient, 130. INDEX. 1567 Magnetic induction, definition of, 4. value of, 7. leakage coefficients, 376. in dynamos, 365* moment, 4, 7. permeability, definition of, 5. properties of iron, 89. resistance, definition of, 5. specific, 5. square method, determ. magn. values by, 93. susceptibility, definition of, 5. units, definition of, 4. symbols of, 1. table of, 7. values, determination of, 91. Magnetism, residual, def. of, 108. Magnetite arc lamp, 570. Magnetization, intensity of, 4. of electromagnets, table of, 111. curve of dynamos, 336. curves of D.C. motor, 353. Magnetizing force, definition of, 4. value of, 7. Magneto-generator, constr. of, 1078. Magnetometer method, determ. magn. values by, 91. Magneto-motive force, def. of, 5, 108. value of, 7. Magneto potential regulators, def. of, 503. Magneto transmitters, 1071. Magnets, excitation of field, 365. field, design of, 364. Main distributing frames, 1104. Mains, insulating joints in, 861. Maintenance of Nernst lamps, 564. Mance method, res . of batteries by,61 . Manganese, effect on steel of, 825. steel, phys. and elec. prop, of, 137. Manganin wire, properties of, 202, 204. Manhattan rail, 830. Manhole constr., cuts of, 309. constr. for shallow trenches, 319. improved forms of, 318. of Niagara Falls Power Co., 319. objectionable types of, 318. cost of, 5' X 5' X 7', 316. Manhole covers, cuts of, 313-315. def. of, 301. estimating cost of, 317. of conduit Metropolitan Railway. 838. Manholes, brick, cost of, 303. concrete, cost of, 303. cost of, table of, 302. sizes of, 302. Manila rope, data on, 1492. Manipulation of coast defense guns, 1134. Marble, crushing load, 1322. for switchboards, 907. Market wire gauge, use of, 201. Mascart electrometer, 39. Masonry, 1321. Master controller, multiple unit system, 764. Material per mile of trolley line, 643. required for one mile of railway, 628. Materials, strength of, 1301. Mats burglar alarm, wiring of, 295. Matthiessen's copper formula, 133. standard of conductivity, 132. Maximum current, A. C. windings, 127. output of induction motors, 398. value of E.M.F. of A.C. current, 404. Maxwell, definition of, 4. law of traction, 94. value of, 7. Mean current, A.C. windings, 127. effective pressure, table of, 1442. hemispherical candle-power, def. of, 529. horizontal intensity, 529. length per turn of coil, table of, 114-116. spherical candle-power, def .of, 529. spherical candle-power of arc lamps, 580. Measure, apothecaries*, 1500. avoirdupois, 1500. of capacity, 1499. of length, 1499. of liquids, 1500. of surface, 1499. 1568 INDEX. Measure of weights, troy, 1500. Measurement of alternating currents, 26-28. of capacity, 63. of efficiency, 508. of E.M.F., 62. of ins. res. of cables, 321. of low resistance, 59. of mutual inductance, 67. of resistance, 56. of standard ampere, 10. of three-phase power. 72. Measures, metrical equivalents, 1500. Measuring instruments, elec, 21. power in six-phase circuits, 477. Mechanical and electrical units, table of, 1258. air-gap, 363. equivalent of heat, 1511. interrupters, 1253. properties of rubber, 229. stoking, 1359. symbols, 1. units, derived, 2. table of, 6. Mega-erg, value of, 12. Megohm, definition of, 5. Melting point of copper, 143. point of substances, 1532. railway bonds, 773. Mercurous sulphate for standard cell, 11, 13. Mercury and water columns, pres- sure of, 1463. arc rectifiers, 480. for battery charging, 482. auto-coherers, 1066. for standard cell, 11. phys. and elec. prop, of, 138. spec. res. of, 132. temperature coef. of, 133. vapor lamps, 558. Merrill on water rheostats, 33. Mershon's chart for calculating transmission lines, 279. Mershon's method, meas. of wave form by, 5 . Metalized carbon lamps, 549. Metal joints in cables, 190. pipes, effect of current on, 852. Metallic arc lamp, 572. circuits in telephony, 1081. sheath, capacity of two wires in, 250. sodium, production of, 1241. Metals, phys. and elec. prop, of table of, 134-140. temperature coef. of, 133. by fusion of, 1349. Meter bearings, 1009. commutator type, DC, 997. Duncan, 998. hysteresis, G. E. type, 102-104. Shallenberger, 1028. testing formula, 1027. Westinghouse, integrating, 998. Wright discount, 1008. Meters, action of, 1039. constants of, 1029. direct current, testing of, 1020. electric, accuracy of, 997. graphic recording, 1036. integrating, action of, 997. polyphase, service connections of, 1023. testing of, 1020. remedy for electrolysis in, 861. speeds of, 1029. switchboard, list of, 945. to feet or inches, 1503. types of, 504. Methods of lighting, efficiency of, 594. Metrical measures, 1500 to 1504. Metropolitan conduit railway sys- tem, 837. street railway system, 836. Mho, value of, 8. Mica for commutators, 351. puncturing voltage of, 234. spec. ind. cap. of, 36, 227. Mioanite, spec. ind. cap. of, 227. Micro-Farad, definition of, 5, 38. Micron, 1500. Miles per hour in feet per min., 660. Millihenrys of non-magnetic wire, 241. Milliken repeater, 1041. Milling machines, power required by, 1522. INDEX. 156S Mill! voltmeter, meas. of cond. with, 87. method, meas. of current by, 78, meas. small res. with, 79. Mill power, 1462. Mils to centimeters, 1503. Mineral oils, 1497. Miner's inch, 1473. water H.P. table, 1475. Mines, electric land, 1137. Minimum size of high-tension con- ductors, 235. Mirror galvanometer, 23. spec. ind. cap. of, 36. Miscellaneous tables, 1499. Modulus of elasticity, 1302, 1312. of elastic resilience, 1312. of rupture of woods, 1317. Mohawk type locomotive, 740. Moisture in steam, 1394. Molecular magnetic friction, meas. of, 50 . Moment, magnetic, 4. of inertia, 1302. compound shapes of, 1303. table of, 1304. of resistance, table of, 1304. of rupture of beams, 1309. of stress of beam, max., 1309. Momentum, definition of, 3. Monolithic conduits, des. of, 301. Moonlight schedules, 603. Moore tube, efficiency of, 566- vacuum tube light, 565. Mortar, cement, 1293. lime, 1293. Mortars, 1293. Morse code, 1052. system, description of, 1040. Motive powers, 864. Motor brushes, backward lead of, 353. capacity curves, railway, 676. car batteries, electrolyte for, 877. dimensions of, table of, 732. horse-power of, 653. characteristics, 685. combinations, 760. control, Ward Leonard's system, 354. Motor converter, def . of, 503 . definition of, 502. equipments, weights of A.C., 719. field magnets, flux in, 367. -generator, definition of, 50 . -generators, 434. -generator turret turning system, 1189. men, personal factor of, 724. operated oil break switch, G. E type, 976. panel, D.C., equipment of, 928. panels, induction, equip, of, 918. three-phase synchronous, equip, of, 919. regulation, test for, 382. retraction, gun operation, 1134. tests, 394. traversing, gun operation, 1134. work, variable speed, system for, 354. Motors and dynamos, tests of, 378. automobile, 1227. boat crane, Navy spec, for, 1194. circuit breakers for, capacity of, 955. controlling panels for Navy spec, for, 1185. counter E.M.F. in armatures of, 353. efficiency curve of, 370. of, Navy spec, for, 1185. of railway, 803. tests of, 395. electric railway, 614. G.E. railway system, 851. induction, starting of, 918. ins. res. of, meas. of, 87. lubrication of, Navy spec, for, 1185. magnetization curve of D.C., 353. Navy spec, for, 1183. railway windings of armatures for, 348. rating of railway, 523. rise of temperature in, 378. street railway, rating of, 661, 673. synchronous, tests of, 399. used as condensers, 292. temp, rise of, Navy spec, for, 1184. 1570 INDEX. Motors, test of street car, 392. torque of armatures of, 353. used to drive machine tools, 1518. ventilation fan, Navy spec, for, 1196. Moving body on air resistance, effect of, 659. -coil galvanometers, 21. des. of, 25. -needle galvanometers, 21. Multicellular voltmeter, cap. test with, 326. Multi-circuit single winding of armature, 342. -contact transmitters, 1072. -phase transformers vs. single- phase, 871. -polar machines, armature wind- ings for, 345. -speed motors, def. of, 504. Multiple circuits, current in, 55. circuits, res. of, 55. conduits, adv. of, 301. connection of alternators, 420. of batteries, 19. control, A.C. railway system, 710. unit switch system, 766. duct conduit, constr. of, 301. switchboards, 1090. telephone system, adv. of, 1094. unit control, G.E. type, 712, 761. Multiplex armature windings, 347. telephony, 1106. three-phase lightning arrester, 988. Multiplier, Y-box, Weston, 73. Multiplying power of shunt, 29. Murray's method, locating faults in cables by, 328. Mutual inductance, def. of, 236. inductance, meas. of, 67. secohmmeter method, 69. induction, transposition to elimi- nate, 285. neutralization of capacity and inductance, 292. National coast defense board, recomm. of, 1123. National Electrical Code, standard conductors, table of, 162. Electrical Contractors' Assoc, symbols adopted by, 299. Natural draft transformers, 448. Navy electric fuse, 1137. generating sets, 1153. special lamps, 1173. specifications, 1153. standard wires, table of, 174. telephone systems, 1206. U. S., electricity in, 1153. wiring specifications, 1167. Neatsfoot oil, spec. ind. cap. of, 37. Needle point spark gap curve, 233. Negative booster, 790. Nernst lamps, descr. of, 562. rating of, 540. Ness telephone switch, 1118. Neutral, grounding of, 478. unstable, 479. Neutralization of capacity and inductance, 292. Newburgh telephone system, 1103. New York central locomotives, 741. Central third rail, 834. City, electrolysis in lower, 858. lighting table for, 604. Niagara-Buffalo Line, arrangement of, 290. Falls Power Co., manhole constr. of, 319. Nickel, phys. and elec. prop, of, 138. plating, 1234. spec. res. of, 132. steel, phys. and elec. prop, of, 138. temperature, coef. of, 133. Nickeline, phys. and elec. prop, of, 138. Night sights, electric, 1148. Nitrates in electrolyte, test for, 878. Nitric acid, spec. res. of, 133. Nitrous oxide, spec. ind. cap. of, 35- Noark fuses, 1276. Non-arcing lightning arresters, 984. metal lightning arrester, 989. railway lightning arrester, 985. Non-inductive load, def. of, 50 s . INDEX. 1571 Non-inductive load, regulation of transformer for, 492. -magnetic wires, ind. reactance of, table of, 242. -magnetic wires, self-induction in, 239. -reversible booster, use of, 893. sine wave, equivalent of, 501. Northrup instrument, des. of, 26, 28. method, conductivity by, 60. meas. ins. res. by, 82. Notation, A.I.E.E., 523. committee on, table by, 6. index, 2. used in dynamo and motor section, 334. Octane, spec. ind. cap. of, 37. Octylene, spec. ind. cap. of, 37. Oersted, definition of, 5. value of, 7. Office building plant, economy of, 1285. Ohm, definition of, 5. international, construction of, 30. def.of, 9. per mil-foot, def. of, 131. value of, 7, 8. Ohmic resistance of storage cell, 883. Ohmmeter, direct reading, 57. Sage type, 58. Ohm's law, 55. Ohms, value of various standard, 131. Oil and coal, comparative costs of, 1358. break switch, General Electric motor operated, 976. circuit breaker, controller, 975. breakers, arrangement of, 935. breakers, Westinghouse, 969. -cooled constant current trans- formers, 465. transformers, 448. flash test of transformer, 500. for lubrications, 1497. for transformers, specifications for, 500. !n transformers, use of, 448 Oil switch, General Electric, 979. switches, arrangement of, 933- hand operated, electrically tripped, 979. operation of, 967. specifications for, 947. use of, 912. weight per gallon of, 1497. Olive oil, spec. ind. cap. of, 37, 227. Open cars, weight of, 736. circuit A.C. armature winding, 410. circuit cells, 15. Open circuit in armature, test for, 402. wire circuits, 1082. Operating cost of gas and elec. cooking, 1260. cost of lamps, 554. elec. cooking utensils, cost of, 1261. elec. heaters, cost of, 1265. Opposition method of testing trans- formers, 496. Order indicators, U. S. Navy, 1202. Oscillating current, definition of, 502. Oscillations, electrical, 1055. in ether, 1278. undamped, 1068. Oscillator, dumb-bell type, 1056. Oscillograph, Blondel type, Sk. Outer rail, elevation of, 617. Outflow of steam, 1416. into atmosphere, 1416. Output of dynamos, formula for, 356. of motors, test of, 395. Outward flow turbines, 1476. Over-compounded dynamo, charac- teristic of, 340. Overhead lines, drop in, 798. lines, transposition of, 285. railway conducting system, 785. trolley construction, cost of, 629. wires, capacity of, 250. Overland wires, breaks in, location of, 327, 1572 INDEX. Overload A.C. relay, 962. capacities, 521. capacity, test of, 381. circuit breakers, 899, 950. guarantees for machines, 947. relay, Westinghouse A.C, 962. Overshot water wheels, 1476. Overspeeding of rotaries, preven- tion of, 961. Over-voltage relay, Westinghouse, D.C., 962. Ozokerite, spec. ind. cap. of, 37, 227. Packing* of transmitters, 1074. Painting, 1498. exposure tests, 1498. Palladium, phys. and elec. prop, of, 138. Pan-cake form of winding, 410. Panel switchboards, design of, 906. Panels, motor controlling, navy spec, for, 1185. rotary converters, equipment of, 924. Paper insulated cables, carrying capacity of, 208. cables, joints in, 191. tables of, 174-178. telephone cables, 188. Paper, spec. ind. cap. of, 36, 227. Parabolic curves in wire spans, charts of, 218. Paraffin, spec. ind. cap. of, 36, 227. Parallel, condensers in, 63, 324. D.C. distribution, size of con- ductors for, 284. distribution, 277. -flow turbines, 1476. running of alternators, 419. Para rubber, electrical properties of, 229. Parson's steam turbine, 1453. Party lines, demand for, 1102. telephone lines, 1108. Parville method, electric heating, 1257. Passenger elevators, operating cost of, 1528. Pasted electrode battery, advan- tages of, 880 Pasted plates of storage cells, 880. Patent-nickel wire, properties of, 202. Pavement, cost of, 619. Paving, cost of, 305, 619. depreciation of, 770. Peak discharge of batteries, 888. of load, batteries to carry, 886. Peggendorff cell, 14. Penstocks, constr. of, 869. Pentane standard lamp, 530. Percentage conductivity, 132. drop, discussion of, 262. Performance diagram, train, 663, 667. Permanent magnetism, def . of, 108. magnet voltmeters, 74. Permeability curve of arc dynamo, 338. curves of iron and steel, 90. of iron and steel, 89. value of, 7. Permeameter, Drysdale's, use of, 97. Thompson's use of, 93-96. Personal factor of motormen, 724. Petroleum, chemical composition of, 1356. furnaces, 1357. oil, spec. ind. cap. of, 37, 227. oils, chemical composition of, 1357. Phase-displacing apparatus, 512. Phase difference, def. of, 501. Philadelphia Inspection Rules for boilers, 1332. Phillip's code, 1052. Phoenix rule for ins. res., 86. Phonograph in telephony, use of, 1096. Phosphor-bronze, phys. and elec. prop, of, 139. Photo-chronograph, Squire-Crehore, 1133. Photometer, Bunsen type, 535. Photometers, integrating type, 539. Physical constants of copper wire, 143. prop, of alloys, table of, 134-140. of metals, table of, 134-140. quantities, table of, 6. INDEX. 1573 Physikalische Reichsanstalt res. unit, 30. Piles, arrangement of, 1292. foundation on, 1291. safe load on, 1291. Pilot brush, use of, 340. Pipe bends, 1431. covering, relative value of, 1422. flanges and bolts, strength of, 1431. dimensions of, 1430. high pressure, screwed, 1430. high pressure, shrink, 1432. standard, 1433. iron, elec. welding of, 1272. lines, constr. of, 869. riveted hydraulic, wt., safe head, 1469. Pipes, diam. of steam and exhaust, diagram of, 1419. dimensions of riveted steel, 1467. equation of gas, 1418. of steam, 1418. formula for riveted steel, 1467. friction of water in, 1374. loss of head due to bends in, 1374. riveted steel, 1466.' sizes for feed- water, 1373. of steam and gas, 1419. standard dimensions of extra strong, 1427. standard dimensions, of wrought , iron, 1419. thawing by electricity, 1531. wooden-stave, 1468. Piping, steam, U. S. navy spec, for, 1163. Pitch, specific inductive capacity of, 227. Planers, power required for, 1516. Plante cell, advantages of, 880. Plate box poles, 632. glass, spec. ind. cap. of, 36. surface for batteries, area of, 883. Plates, appearance of battery, 874. buckling of, 881. of batteries, cadmium test of, 878. safe working prssure for flat, 1332. types of, 874. Plating baths, 1233. Platinoid, fusing effect of current on, 217. phys. and elec. prop, of, 139. wire, properties of, 202. Platinum, fusing effect of current on, 217. in electrolyte, test for, 877. phys. and elec. prop, of, 139. silver wire, properties of, 202. spec. res. of, 132. standard of light, 532. temperature coef. of, 133. wire, properties of, 202, Plow, metropolitan street railway, 839. suspension, 840. Plug tube switches, 965. Plunger electromagnet, law of, 127. electromagnets, shapes of, 128. magnets, range of, 130. Pneumatic tires, data on, 1225. Poggendorff method, comparison of E.M.F. by, 77. Polar arc, chord of, values of, 371. duplex, 1044. relay, use of, 1044. Polarity of transformer, 495. Polarization, def . of, 14. of storage cell, 879. of X-rays, 1248. Polarized bells, biased, 1103. constr. of, 1076. use of, 1114. Pole face, dimensions of, 363. faces, shape of, 356. line construction, 630. lines for high tension work, *>71* pieces, faces of, 363. transpositions, 1082. unit strength of, 4. Poles, determination of number of, 355. of induction motor, 426. plate box type, 632. use of green wooden, 806. wooden, contents of, 633. Polyphase apparatus, c ; -cuit breakers for, 953. generator, def. of, 502. induction motor, theory of 422. 1574 INDEX. Polyphase induction motor, power of, 423. starting torque of, 423. induction wattmeters, 1003. integrating wattmeters, 1004. lines, transposition of, 287. meters, connections of, 1026. constants of, 1031. installation of, 1023. service connections of, 1023. testing of, 1020. motor protected by circuit breakers, 954. Porcelain, spec. ind. cap. of, 37, 227. Portable integrating wattmeters, data for, 1016. sub-station, 819. telephone switchboard, 1141. testing battery, 16. Portland cement, wt. of, 1293. Position indicators, U.S. navy, 1202. Post-office wheatstone bridge, 31. Potassium chlorate, production of, 1242. cyanide, production of, 1246. use of, 1233. Potential betw. plates of batteries, test of, 878. drop in feeders, 788. energy, 3. measurement of, 40. regulator, three-phase induction, 469. regulators, 467. def. of, 503. rise due to transformers, 479. transformers, descr. of, 945. Potentiometer, des. of, 47. method, E.M.F. of batteries, 63. use of, 47. Pound, 1499. calorie, 1511. -degree, C. value of, 12. Power ammunition hoists, U. S. navy, 1191. and light cables, 320. Ayrton and Sumpner method for meas. A.C., 71. carrying capacity in three-phase cables, 216. Power circuits, res. of, meas. of, 80. consumption in factories, 1517. consumption of cars, 658. curve for railway motors, 669 curves, altern. current, 70. for reducing cost of, 868. for trolley cars, 652. definition of, 3. distribution, discussion of, 262. system, A.C. railway, 718. electric, def. of, 5. meas. of, 50 . factor compensation, 1002. def. of, 279, 50 . in three-phase circuits, 72. of transformers, 458. varied by use of synchr. motors, 292. for cars, 656. house, electrolytic action near, 862. in altern. circuit, meas. of, 69. in six-phase circuits, 477. international unit of, 10. light and heat, in residences, cost of, 1287. light and heat in isolated plants, cost of, 1285. lines, transposition of, 285. loss, formula for, 265. in lead sheath of cables, 293. in line, 261. mechanical, meas. of, 50 . of polyphase induction motor, 423. of water flowing in a pipe, 1462. -operated switchboards, 906. plants, chimney protection for, 1281. plants, lightning arresters in, 983. required for automobiles, 1224. for electric cranes, 1527. for machine tools, 1516. for street railways, 656. to drive machinery, 1515. station construction, chart of, 1289. depreciation of, 770. design of, 866. efficiency of machines in, 663. INDEX. 1575 Power station for railways, 815. system, U. S. Navy, 1183. three-phase, meas. of, 72. to drive machine shops, 1518. transmission, classif. of, 864. losses in, 1529. transformers for three phase, 478. voltage for, 870. used by machine tools, 1515. Preliminary dynamo dimensions, checking of, 363. Prepayment wattmeters, 1010. wattmeters, Fort Wayne, 1012. Pressure gradients, descr. of, 283. drop in parallel distribution system, 279. drop, formula for, 264. mean effective steam, table of, 1442. of water to 1000 ft. head., 1465. working, for cylindrical shells of boilers, 1330. Prevention of electrolysis, 861. Primary batteries, action of, 14. Primer for gun firing, 1213. Principle of magnetic circuit, 109. Printing machinery, power to run, 1525. plants, electric heat in, 1269-1270. Private telephone lines, 1088. Production of metals, 1232. Projectiles, velocity of, test of, 1128. Projectors, search light, 575. U.S. Navy, 1179. Prometheus system, electric heating' 1257. Prony brake, formula for, 1515. test, 395. Propagation of waves, 1058. Properties of aluminum wire, 194. of dielectrics, 227. of galv. iron wire, 34. of saturated steam, table of, 1404. above a vacuum, 1406. of wires and cables, 131. Propulsion, electric, Leonard's sys- tem of, 354. Protected third rail, cost of, 835. Protection against high potentials on A.C. circuits, 981. of buildings from lightning, 1289. of chimneys, 1281. of steam heated surfaces, 1421. of transformers against fire, 871. relays, table of, 960. Protective relays, 956. wires, use of, 982. Protectors, telephone, 1088. Puffer's modification of Kapp's dynamo test, 389. test of street car motors, 392. Pulleys, 1487. rules for, 1487. to find size of, 1487. Pull of electromagnets, curves of, 129. of electromagnets, formula for, 110. -off curve construction, hangers for, 647. on armature conductors, formula for, 351. Pulsating current, definition of, 50,. Pulsation, def. of, 505. Pump exhaust, 1377. Pumping back test of motors, 397. test of two dynamos, 388. hot water, 1367. Pumps, 1367, 1443. air, 1445. and condensers, 1443. double cylinder, sizes of, 1370. circulating, 1445. single cylinder, sizes of, 1369 sizes of direct-acting, 1368. Puncturing voltage for dielectrics, 228. voltage of mica, 234. Pupin telephone system, 1107. Quadrant electrometer, 40. Quadruplex telegraphy, 1051. Quality of light, 600. of steam by color of issuing jet, 1400. Quantity of electricity, def. of, 5. meas. of, 25. symbol of, 8. 1576 INDEX. Quantity of electricity, unit of, 4. Quartz, spec. ind. cap. of, 37. Quick break switches, 964. Radial brick, bond in, 1340. for chimneys, 1341. telephone system, 1117. Radiation, laws of, 528. of heat in ducts, 214. f Radiators and convectors, 1263. Radioscopic images, examination of, 1255. Radius of curvature, 616. of gyration, 1303. compound shapes, 1303. table of least, 1304. Rail bonds, testing of, 801. curvature, 616. joints, testing of, 801. Potter type, 830. testers 802. welding, electric, 1273. thermit system, 778. Rails and bonded joints, rel. value of, 780. electrolytic action on, 855. impedance of steel, 795. ingredients of, 780. resistance of, 821. specifications for, 830. weight of, 615. Railway booster calculations, 809. system, 807. bonds, requirements for, 775. types of, 772. circuits, drop in, 796. testing drop in, 804. tests of, 798. conductors, dimensions of, 791. conduit systems of, 835. depreciation, table of, 770. electric, system of operating, 613. energy of electric, 706 equipments compared, 719. weights of, 730. machinery, depreciation of, 770. motor characteristics, 685. combinations, 760. motors, 614. A.C. type, 707. Railway motors, armature windings of, 348. capacity of, 673. characteristic curves for, 664. efficiency of, 803. installation of, 745. rating of, 523. selection of, 52?. speed-time curve for, 669. standard sizes of, 729. test of, 397. temperature of, 675. torque of, 731. non-arcing lightning arresters, 985. overhead conductors, 785. power station, 815. service boosters, 813. shop, power required in ideal, 1521. speed and energy curves, 680. sub-stations, equipment of, 942. system, load factor of, 785. ties, durability of, 619. turnouts, 620. Rake of poles, 633. Range finder, Fiske, 1211, finders, lights for, 1148. indicators, U. S. Navy, 1204. of carbons, 577. of solenoids, 130. Rape-seed oil, spec. ind. cap. of, 37. Rapid fire guns, firing mechanism for, 1149. Rated terminal voltage, def . of, 515 . Rate of acceleration, 666. of deposit, 1235. Rates, gas and electric, comparison between, 1261. of charge of batteries, 883. of discharge of batteries, 883. of storage batteries, 874. Rating of fuse wires, 507, 1275. of generators or motors, 506. of illuminants, 540. of railway motors, 661, 673, 729. Ratio of transformers in three-phase system, 471. test of transformer, 491. INDEX. 1577 Rayleigh's method, E.M.F. of batteries, 62. Reactance coil for A.C. arc cir- cuits, 466. factors, table of, 266. of three-phase line, inductive, 245. 'of transmission circuits, 238. symbol of, 8. -voRs for A.C. lines, 280. Reaction of alternator armatures, 414. of armatures, 350, 364. Reactive coils, use of, 982. factor, def. of, 505. Reactors, def. of, 503. Reading, illumination for, 602. Receiver, Bell telephone, 1070. capacity, 1443. with detector, 1065. Receivers, coherer with jigger, 1064. wireless telegraph, 1063. Recording meters, Bristol, 1036. wattmeters, Duncan, 1000. G. E., testing, 1030. Thomson, 998. Records of temperature test, 381. Rectifying apparatus, losses in, meas. of, 512. Reduced deflection method, res. of batteries by, 60. Reed method, electric heating, 1257. Refined iron, qualities of, 824. Refineries, copper, 1238. Refining of copper, 1235. of metals, 1232. of silver, 1238. Reflecting galvanometer, Kelvin type, 23. Reflections, coefficients of table of, 593. Regulating battery, 888. devices for induction motors, 428. reactance coil, 466. relays, 956. Regulation, importance of, 545. of arc lamps, 576. of dynamos, test for, 382. of generators, 870. of electrical machines, 513. Regulation of transformers, 458. by calculation, 492. comparative, 455. table of, 498. test of, 491. of voltage of transformers, 452. Regulations of Board of Trade, 781. Regenerative X-ray tubes, 1251. Regulators for A.C. generators, 409. for separate circuits, 469. of potential, 467. three-phase induction potential, 469. Reinforced concrete, 1292. Relative conductivity, 132. efficiency of large and small trans- formers, 459. Relay, General Electric A.C. over* load, 961. low voltage A.C, 962. D.C., 961. overload, A.C, 962. over- voltage, D.C, 960. reverse-phase A.C, 962. underload D.C, 962. Westinghouse A.C. overload, 962. D.C. over- voltage, 962. time limit, 960. Relays, auxiliary, 956. classification of, 955. commonly employed, 960. definite time limit, 956. instantaneous, 956. inverse time limit, 956. protection of A.C systems by, 959. protective, 956. regulating, 956. reverse current, 961. signalling, 955. Reliability of service, switchboards built to insure, 929. Reluctance, definition of, 5. value of, 7. Reluctivity, definition of, 5. value of, 7. Remedies for electrolysis, 861. Remote control panel switchboard, 906, 928. control switches for equalizer cir- cuits, 962. 1578 INDEX. Removal from service of storage batteries, 881. Renewals of lamps, 547, 556. Repeater, Atkinson, 1048. duplex, 1049. Ghegan, 1042. Milliken, 1041. Weiny-Phillips, 1043. Repeaters, use of, 1041. Repulsion motor, def. of, 502. Reservoirs, storage, 867. Residential plant, cost of maint. of, 1287. plant, installation of, 897. Residual magnetism, def. of, 108. Resin, spec. ind. cap. of, 37. Resistance box, decade type, 32. control of battery discharge, 891. curves on air, 659. definition of unit of, 5. due to gravity, 1224. for arc lamps, 581. high voltmeter, 75. In overhead lines, 798. returns, 798. In rotor of induction motors, 428. In stator of induction motors, 429. low, meas. of, 59. magnetic, definition of, 5. meas. of, with ohmmeter, 57. with volt and ammeter, 78. measurements, 56. of A. C. circuits, 259. conductors, effective, 238. of aerial lines, 61. of aluminum wire, 194, 196. of armature, meas. of, 401. of batteries, 60. of bonds, 776. of brushes, 362. of cables, meas. of, 330. of carbons, 577. of cells, internal, 20. of conductors, 61. table of, 266. of copper wire, table of, 148. of dilute sulphuric acid, 1229. of Driver-Harris wire, 207. of field coils, meas. of, 401 Resistance of galvanometers, 60. of German silver wire, 203. of gutta-percha, 231. of house circuits, 61. of light and power circuits, meas* insulation, 80. of multiple circuits, 55. of plating bath, 1235. of rails, 821. of steel, 825. of storage batteries, 883. of stranded aluminum wire, table of, 198. of sulphate of copper, zinc, 1231. of track rails, 779. of transformer, meas. of, 486. of trolley and track, 798 of water rheostats, 33. of wiring system, insulation, 82. of working batteries, 61. practical standard of, 30. specific, 131. magnetic, 5. symbols of, 7. table of galv. iron wire, 34. temperature coefficient, 133. to traction, 1225. type furnace, 1244. unit of, 4, 131. variation with temperature of, 228. -volts for A.C. lines, 280. wires, properties of, 202. Resistances, high, meas. of, 79. small, meas. of, 79. Resisting moment of beams, 1308. Resistivity, definition of, 9. symbol of, 8. Resonance, curves of, 1216. theory of, 1215. Retardation, rate of, 668. Retentiveness, def. of, 108. Retraction motor for gun operation, 1134. Return booster system, 808. call bell system, 293. circuit, 771. current, division of, 800. drop of ground, test of, 799. Returns, drop in, 798. regulation for railway, 781. INDEX. 1579 Reverse current circuit breaker, 950. current relay, 961. -phase A.C. relay, 962. Reverser, multiple unit system, 762. Reversible booster, use of, 894. Reversing current in armatures, 351. Revolution indicators, U. S. Navy, 1204. Revolving field alternators, 409 Rheostatic controller, 754. controllers, list of, 756. Rheostats, temperature rise in, 520. water, 33. Right of way for pole lines, 871. Ring armature, windings of, 342. down trunks, 1096. method, determ. magn. values by, 91. type armatures, 341. Ringing keys, 1090. Rise and grades, 617= of potential due to transformers, 479. of temperature in armatures, 349, 358. of commutator, 362. in dynamos, test of, 378. in field coils, 352. in transformers, test of, 483, 491. in transformers, 447, 498. meas. of, 518. U. S. Navy generators, 1158. Ritchie's photometer, 536. Riveted bonds, 774. Roadbed, depreciation of, 770. Road surface material, 1225. Rock, foundations on, 1290. salt, spec. ind. cap. of, 37. Rodding of cables, 319. Rod float gauging, theory of, 1471. Rods, lightning, installation of, 1278. Roebling galv, telegraph wire, prop- erties of, 200. steel telegraph wire, properties of, 201. wire gauge, 141. Rolling stock, depreciation on, 770. Room lighting, data on, 597. Rope driving, 1490. hemp, wt. of, 1494. horse-power of transmission, 1492. manila, velocity of, table of, 1492. wt. and strength of, 1494. Ropes and belts, slip of, 1493. horse-power of manila, 1491. of manila, diagram of, 1492. strain from loads on inclined planes, 1494. Rosa curve tracer, 51. Rosendale cement, wt. of, 1293. Rosin, specific inductive capacity of, 227. Rotaries, overspeedingof , prevention of, 961. starting diagram of connections for, 920. starting of, 440. Rotary compensator turret turning system, 1189. Rotary converter circuit protection by relays, 959. def . of, 503. panel, General Electric D. C, 925. equipment of, 919, 924. sub-station, 816. Rotary converters connected to transformers, 442, 476. descr. of, 436. for six-phase system, 475. in sub-stations, 814. starting, diagram of connections for, 920. voltage between collector rings of, 439. Rotary field of induction motor, 425. induction apparatus, temp, rise in, 520. transformers, armature windings for, 441. Rotating field in wattmeters, 1000. Rotation of conductors around pole, 109. Rotor, core of, 425. definition of, 423. resistance in, 428. slots, number of, table of, 427. windings, commutated, 429. 1580 INDEX. Rowland method, determ. magn. values by, 91. Rubber covered cables, carrying capacity of, 208. wire and cables, prop, of, 161. underwriters' test of, 161. Rubber, electrical properties of, 229. insulated cables, carrying capac- ity of, 210. cables, data on, 214. telegraph cables, 189. wires and cables, tables of, 164- 172. insulation test of, 230. specific inductive capacity of, 227. tires, data on, 1225. Rules for conductingboiler tests, 1384. Rumford's photometer, 536. Ryan electrometer, 51. Ryan's method, meas. of wave form by, 51. Sad irons, elec, cost of operating, 1263. Sag and tension in wire spans, 218. for aluminum wire, limit of, 225. in wire spans, calc. of vertical, 222. Sage direct reading ohmmeter, 58. Safe load on wooden beams, chest- nut, 1319. hemlock, 1319. southern pine, 1320. spruce, 1318. white cedar, 1319. white pine, 1318. yellow pine, 1319. load on brickwork, 1322. on steel beams, 1310. temperature for field coils, 352. Safety valves, 1382. Philadelphia rules, 1383. rules for pop valves, 1383. rules governing, 1382. Saline solutions, conducting power of, 905. Salt solution for water rheostats, 34. Sand and cement, A.S.C.E. recom- mendations, 1294. and cement, fineness of, 1294. foundations on. 1290. Sandstones, crushing load, 1323. Sangamo integrating meter, 1006. wattmeters, testing of, 1035. Saturation factor, def. of, 505. test of dynamos, 400. S.B. resistance wire, 207. Scale, galvanometer, 24. solubility of, 1363. Scales, instrument, figuring of, 946. Schedule for 35-ton car, 658. Schmidt chronoscope, 1131. Schuckert searchlights, 1123. Schultz chronoscope, 1130. Scott method of connecting convert- ers and transformers, 477. Screwed contact, current density for, 442. Searchlight carbons, 579. projectors, 575. Searchlights, data on, tables of, 1127. intensity of light of, 1125. mirrors of, 1125. Schuckert type, 1123. use of, 1123. U. S. Navy, spec, for, 1179. Secohmmeter, meas. mutual ind. by, 69. Secondary current, transformers for constant, 462. standards, checking of, 1013. Second, definition of, 2. Sectional rail, Westinghouse rail- way system, 846. Sections, elements of usual, 1303. of trolley system, laying out, 785. Seeley's cable connectors, 190. Segments, commutator, number of, 361. Selective telephone systems, 1102. Selenium, spec. ind. cap. of, 37. spec. res. of, 132. Self-inductance, meas. coef. of ind. by, 65. with altern. current, meas. of, 66 Self-induction, coefficient of, 64. def. of, 238. formula for, 239. in solid iron wire, table of, 248. in stranded wires, 241. of transmission circuits, 238. INDEX. 1581 Self-induction standard, Ayrton and Perry's, 66. Separate circuit regulators, 469. Separately excited dynamo, 338. Separating calorimeter, 1398. Separators, steam, 1380. Series A.C. regulator, G.E. type, 466. boosters for railway service, 813. commutator A.C. motor, 503. condensers in, 63, 324. i connection of batteries, 19. dynamo, descr. of, 336. ext. characteristic curve of, 337. limit switch for railway control, 769. multiple switchboards, 1092. parallel controller, 753. controllers, list of, 755. party lines, 1108. telephone system, 1076, 1109. transformers, 464. Service box cover, cut of, 315. box, def. of, 301. boxes, constr. of, 302. capacity of railway motors, 675. connection of polyphase meters, 1023. meter, tests of, 1015. reliability, switchboards built to insure, 929. Sesame oil, spec. ind. cap. of, 37. Sewer connections, cost of, 303. Sewing machine, power to run, 1525. Shafting, centers of bearings of, 1483. deflection of, 1482. hollow, 1485. horse-power of iron, 1481. tables of, 1484. laying out, 1485. pulleys, belting, rope driving, 1481. Shafts, armature, 341. hollow, 1485. Shallenberger meter, testing of, 1028. Shallow trenches, manhole constr. for, 319. Shape of moving body, effect of, 659. of pole faces, 356. Shapers, power required for, 1520. Sharp-Millar's pnotometer, 539. Shawmut soldered bond, 772. Shearing strength of woods, 1316. Shear, vertical-beams, 1308. Sheathing core, formula of, 142. Sheath, metallic, capacity of wires in, 251. Sheet metal, permeability of, 89. Sheldon method, meas. low res. by, 59. Shellac, spec. ind. cap. of, 37, 227. Shell type transformers, coils for, 444. Shelves for bus-bars, 933. Ship, condensation of steam in pipes aboard, 1415. Ships, dynamos in, gyrostatic action on, 353. Shoes, cast iron magnet, 352. third rail, 832. Short circuit in armature, test for, 402. connection winding of armatures, 343. Shunt booster, use of, 892. boxes, galvanometer, 29. dynamo, external characteristic of, 339. internal characteristic of, 339. dynamos, regulation tests of, 382. winding of compound wound machine, 369. wound dynamos, des. of, 336. Shunted detector, 1065. Shunts, ammeter, 41. Shut-down of plant, provision against, 929. Side brackets for trolley line, 635. Siding suspension, 638. Siemens' electro-dynamometer, 42. ohm, value of, 131. Sights, night, electric, 1148. Signal corps wireless telegraphy, 1145. lights, U. S. navy, 1181. stranded wire, galv., properties of 200. system, requirements of, 623. Signalling, automatic Mock. 622. relays, 955. 1582 INDEX. Signalling, syntonic, 1059. Silicon-bronze, phys. and elec. prop. of, 140. Silt, effect on storage of, 869. Silver, phys. and elec. prop, of, 139. plating, 1234. refining of, 1238. spec. res. of, 132. temperature coef. of, 133. voltameter, description of, 10. Simplex system, electric heating, 1257. Sine curve, discussion of, 404. wave, def . of, 507. Single conductor cables, watts per foot lost in, 212. conductor cable cambric ins., tables of, 179-183. conductor wire table, U. S. navy, 1170. contact transmitters, 1071. duct conduit, adv. of, 301. overhead wire, capacity of, 250. Single-phase A.C. motors, 421. A.C. railway system, 707. A.C. sub-station, views of, 943. air-blast transformers, 452. armature winding, 411. circuit, charging current per 1000 feet of, 253. circuits, self induction in, 239. feeder panel, equipment for, 916. induction wattmeters, 1003. line, capacity effect in, 249. potential regulators, 467. railway, distribution system for, 718. railway motor characteristics, 713. rotary converter, 436. transformer connections, 472. transformers vs. multi-phase, 871. transmission circuit, calc. of, 280. wiring examples, 272. Single truck cars, power for, 656. Six-phase, changing three-phase to, 475. circuits, power in, 477. Size of conductors for parallel D. C. distribution, 284. of generator units, 870. Sizes of carbons, 578. of railway motors, 729. S.K.C. high voltage testing set, 461. lightning arrester, 990. Skin effect, 1061. def. of, 236. factors, table of, 237. Slate cut-outs, res. betw. terminals of, 86. for switchboards, 907. Slawson's signal block system, 627. Slide-wire bridge, 58. Sliding trolley collector, 644. Slip of induction motor, table of, 425. of ropes and belts, 1493. Slipper shoe for third rail, 833. Slot sizes, armature, values of, 372. Slots in field-frame of induction motor, 426. of armature cores, design of, 357. Slotted or toothed type armatures, 341. Slotters, power required for, 1520. Smashing point, def. of, 540. Small resistances, meas. of, 79. Smelting by Stassano process, 1274. electric, 1247. def. of, 1232. Smooth body armatures, advantages of, 341. Sneak current protector, 1088. Soapstone for switchboards, 907. Sodium, cyanide of, production of, 1246. hydrate, production of, 1239. production of, 1241. Soft iron ammeters, 41. Soldered bonds, test of, 773. types of, 772. Soldering irons, electric, 1270. Solenoids, characteristics of, curves of, 129. coefficient of self ind. of, 65. pull of iron-clad, 127. tractive effort of, 130. Solid back transmitters, 1072. copper wire, G. E. Co., prop, of, table of, 162. prop, of, table of, 154. INDEX. 1583 Solid tires, data on, 1225. Solids, spec. ind. cap. of, table of, 36, 37. Solubilities of scale-making mater- ials, 1363. Sound, propagation of, 1069. Sources of light, intrinsic brightness of, 529. Space occupied by D.C. cov. wire, table of, 123-126. occupied by S.C. cov. wire, table of, 121-123. required by turbines vs. recipro- cating engines, 1454. Spacing of beams for various loads, 1315. Span construction, 644. wire, dip in, 634. material for, 635. Spans, chart for long, 220. chart for short, 221. tension and sag in wire, 218. Spark gap curve, 233, 462. gap, meas, 517. points, consir. of, 517. Sparking at commutator, 805. at switches, 948. distance across needle points, 462. distances, table of, 526. of brushes, 805. Sparks, chemical effect of electric, 1232, length of, curves of, 949. Special cables for car wiring, prop, of, 173. lamps, navy, table of, 1178. Specifications for det. ampere, 10. for det. intern, volt, 10. for paper ins. telegraph cables, 189. for paper ins. telephone cables, 188. for submarine cables, 189. for switchboards, 947. for 30 per cent rubber compound, 229. for telephone cables, 1083. for transformer oil, 500. for transformers, 498. U.S. Navy, 1153 Specifications for wiring, U. S. Navy, 1167. Specific conductivity, 132. energy dissipation in arm. core, 107. gravity and unit weights, tables of, 1513. gravity, table of, 1512. heat, mean, of platinum, 1509. of gases and vapors at con- stant pres., 1511. of water, 1511. heats of metals, 1509. inductive capacity, 4, 38. measurement of, 38. of dielectrics, 227. of gases, table of, 35. of liquids, table of, 37. of solids, table of, 36, 37. magnetic resistance, 5, 7. resistance, 131. of conductors, table of, 132. of liquids, table of, 133. thermal conductivity of dielec- trics, 234. Speech, definition of, 1069. Speed and energy curve, 680. curves of railway motors, 686. error table for wattmeters, 1032. headway and number of cars, 660. of cars, diam. of wheels to obtain certain, 655. of dynamos, formula for, 356. of induction motors, 424. of power generators, 870. of wattmeters, 1029. recorders, U. S. Navy, 1212. run of N. Y. C. locomotive, 743. -time curve, 667. Spendersf elds line, details of, 651 . Spermaceti, spec. ind. cap. of, 37. Sperm oil, spec. ind. cap. of, 37. Spherical candle power of lamps, 540. reduction factor, 525. Spiegeleisen, phys.and elec.prop. of, 137. Spikes, table of, 618. Spitting-off discharges, 1278. Sprague multiple unit control, 761. 1584 INDEX. Spring jacks, use of, 1089. Square roots, table of, double, 45, 46 Squire-Crehore photo-chronograph, 1133. Squirrel-cage induction motors, rotor slots for, 427. Staggering trolley, 644. Standard candle, 530. cell, construction of, 11. description of, 10, 19. filling of, 13. used with potentiometer, 47. condensers, construction of, 38. conductors, N. E. C, prop, of, table of, 132. copper wire strands, prop, of, table of, 159. of resistance, construction of, 30. of self-induction, Ayrton and Perry's, 66. symbols for wiring plans, 299. Standardization rules A.I.E.E., 500a, Standards of light, 530. Stanley lightning arrester, 990. Star connected armature windings, 413. connection of transformer, three- phase, 473. of winding, 404. Starting current test of synchronous motors, 400. devices for induction motors, 428. induction motors, methods of, 918. of rotaries, 440. rotary converters, diagram of connection for, 920. torque of polyphase induction motor, 423. Stassano process for elec. welding, 1274. Static dischargers, 992. ground detectors, installation of, 942. interrupter, 993. machines, use for X-ray of, 1252. transformer, def. of, 443. wave, action of, 993. Stationary impedance of induction motors, 398. Stator. core of, 425. Stator, definition of, 423. resistance in, 429. Stays, boiler head, 1333. Steady strain discharges, 1278. Steam, 1327. Steam boiler, efficiency of, 1329. settings, 1334. measurements of, 1336. strength of riv. shell, 1330. Steam boilers, cylinders of, 1327. flues of, 1327. gas passages and flues of, 1329. grate surface per h.-p. of, 1329. heating surface of, 1328. tubes of, 1328. per h. p., 1329. hor. return tubular, 1327. setting of, 1335. horse-power of, 1327. points in selecting, 1327. scotch or marine, 1327. types of, 1327. vertical fire tube, 1327. water tube, 1327. working pressure of, 1330. Board of Trade rule, 1332. Philadelphia rule, 1332. U. S. statutes, 1332. Steam engines, 1434. and dynamos, standards of, 1435. brake horse-power of, 1440. cylinder ratios of, 1441. horse-power of, 1440. ind. hor3e-power of, 1440. mean effective pressure table of, 1442. nominal horse-power of, 1440. receiver capacity of, 1443. regulation of, 514. tests of various types of, 1413. Steam, flow through pipes of, 1417. flow to atmosphere of, 1416. to lower pressures of, 1416. heating, boiler horse-power, 1415 moisture calorimeter diagram. 1397. in, determination of, 1394. tables of, 1396. INDEX. 1585 Steam pipe covering, cost and heat loss of, 1423. electrical tests of, 1422. diagram of, 1423. heat loss in, 1423. miscellaneous substances for, 1425. relative economy of, 1424. value of, 1422. Steam pipes, 1417. condensation in, 1415. aboard ship, 1415. heating, 1415. loss of heat from, 1421. Steam piping, U. S. navy spec, for, 1163. ports and passages, 1443. properties of saturated, 1404. 1-15 lbs. abs., 1404. quality by color of issuing jet, 1400. separators, 1380. superheated, 1413. table, DeLaval turbine, 1458. total heat of, 1511. Steam turbine, 1451. Curtis, 1455. DeLaval steam flow, table of, 1458. tests of, 1452. Parsons, 1453. vanes in Westinghouse-Parsons, 1453. Steam turbines, relative floor space of, 1454. relation of foundation to h.-p. of 1457. U. S. Navy spec, for, 1160. Steam, volume of, tables of, 1404. weight of, tables of, 1404. Steel and iron, ageing of, 455. elec. welding of, 1271. magnetic fatigue of, 455. permeability curves of, 90. wire, constants of, 199. Steel chimneys, brick lining of, 1343. cost of, 1343. foundation size of, 1343. field magnet yokes, 352. for third rail, qualities of, 822 Steel frame buildings, electrolysis in, 859. magnetic qualities of, 91. permeability of, meas. of, 94. poles, 633. weight of, 633. production of, 1247. rails, 825. impedance of, 795. resistance of, 825. strand wires for trolleys, 642. telegraph wire, properties of, 201. weight of, 1294. wire, properties of, 199, 201. use in telephony of, 1082. Steering-gear, navy spec, for, 1200. Steinmetz hysteresis formula, 98. Step-by-step method, hysteresis tests by, 101. telephone systems, 1102. Step-down transformers for Y-dis- tributions, 478. Stepping-down arrangement for long distance transmission, 474. Stepping-up arrangement for long distance transmission, 474. Stern's duplex, 1050. Stillwell potential regulator, 467. Stoking, mechanical vs. hand firing, 1359. Stone, crushing load of, 1322. foundations, 1293. Stop watch, use in meter tests of, 1015. Stops of car, table of frequency of, 658. Storage batteries, automobile, 1227. capacity of, 874, 883. care of, 1228. central station, three-wire sys- tem, 903. charge and discharge rates of, 883. charging of, 880. connections for charging, 899. constant current, booster sys- tem, 901. dimensions of, 883. discharge rate of, 874. efficiency of, 879. 1586 INDEX. Storage batteries, elements of, 872. erection of, 884. installation of, 885. internal resistance of, 883. load regulation by, 888. loss of charge of, 884. polarization of, 879. removal from service of, 881. requirements of, 874. sulphation of, 881. tests of, 882. theories of, 872. three- wire system, 899. to carry load peak, 886. troubles of, 881. uses of, 886. variation of efficiency of, 884. voltage curves of, 883. weight of, 882. Storage battery booster equipment, 902. boosters, circuit breakers pro- tecting, 952. capacity, 900. discharge, control of, 891. plant, installation of, 897. plates, cadmium test of, 878. types of, 874. Storage reservoirs, 867. Stoves, car, cost of operating, 1266. Strains in ropes on inclined planes, 1494. Strain test, 381. Stranded conductor, G. E. Co., table of, 163. copper conductors, carrying cap. of, 209. wire, prop, of, table of, 155. weather-proof aluminum wire, 197. wires, self induction of, 241. Strain, 1301. Strands, standard copper, prop, of, table of, 159. table of wire, 142. Stray power of dynamo, calculation of, 391. test of motor, 396. Streams, estimating, 869. St^«t car equipments, compared, 7 lb. Street car heating, electric, 1265. cars, lighting of, 806. possible schedule for, 658. excavation per conduit foot, cost of, 306. lighting by arc lamps, 582. Street railway booster system, 807. circuits, test of, 798. material required for one mile of, 628. motor characteristics, 713. control, Leonard's system of, 354. testing, 803. motors, armature windings of, 348. capacity of, 673. characteristic curve of, 664, 686. efficiency of, 663, 803. rating of, 661. service capacity curves of, 676. speed-time curve for, 669. test of, 392, 397. power station, 815. Street railways, depreciation on, table of, 770. power required for, 656. Strength of current, meas. of, 78. of dilute sulphuric acid, table of, 904. of materials, 1301. of riveted shell, boiler, 1330. of wire ropes, 1325. Stress, 1301. Strut bars, 1314. Submarine and underground cables, tests of, 321. cables, 189, 1083. testing of, 331. Submerged rheostats, wire for, 34. Sub-station design, 814. for railways, 815. portable type, 819. rotary converter, 816. single phase A.C., views of, 943. Sub-stations, drop between, 794. equipment of, 942. Substitution method, res. meas. by, 56. INDEX. 158? Suburban cars, types of, 612. Sulphate of copper, res. of, 1231. of lead, use of, 873. of zinc, res. of, 1231. Sulphation of storage batteries, 881. Sulphur dioxide, spec. ind. cap. of, 35. spec. ind. cap. of, 37. Sulphuric acid, conducting power of, table of, 905. resistance of, 1229. spec. res. of, 133. strength of, table of, 904. Sumpner's test of copper loss in transformers, 497. of iron loss in transformers, 496. Superficial measures, metrical equiv. 1501. Superheated steam, 1413. economy of engines using, 1413. Superheaters, 1413. Supplies, approx. list of electric work, 1531. Supplying losses, electrical method of, 389. Surface contact plates, G. E. railway system, 848. railway, G. E. system, 847. railway system, 840. shoes, G. E. railway system, 850. Surface insulation against electro- lysis, 862. measures of, 1499. Susceptance, capacity, table of, 269. symbol of, 8. Susceptibility, magnetic, definition of, 5. value of, 7. Suspended wires not on same level, sag in, 223. Suspension brackets, 635. of trolley wires, 637. Swapping of current, 859. Swedish iron rope wire, 1325. Switchboard, definition of, 906. instruments, 940. list of, 945. meters, list of, 945. Switchboards, A.C. and D.C., ro- tary converter panels for, 924. A.C. panels for, 912. aluminum bars for, 911. arc, General Electric, 922. central station, electrically op- erated, 928. panels for, 907. connections on, 910. constant current transf. panels for, 922. controlling, 940. copper bars for, 909, 911. D. C. exciter, 942. feeder panel for, 928. generator panel for, 924. motor panel for, 928. direct control panel, 906. electrically operated, 929. for battery plants, 898. for hydro-electric plant, 931. for transmission plants, 870. frames for, 908. General Electric D.C., rotary con- verter panel for, 925. generator, U. S. Navy, 1163. hand-operated, 906. remote-control, 928. illuminating lamps for, 909. induction motor panels for, equip, of, 918. insulation distances on, 912. isolation of conductors on, 929. material for, 907. panel, design of, 906. power-operated, 906. reliability of service insured by, 929. remote control panel, 906. space behind, 907. specifications for, 947. single-phase panel for, equip- ment of, 916. sub-station, equipment of, 942. telephone, common battery, 1098. design of, 1089. multiple, 1090. portable, 1141. series multiple, 1092. temperature rise of devices on, 910. 1588 INDEX. Switchboards, three-phase panels for 912. rotary converter panel for, 919. synchr. motor panels for, 919. two-phase panels for, 915 Westinghouse generator panel for, 925. rotary panel for, 925. three-wire generator panel for, 926. Switches disconnecting type, 965. for equalizer circuits, remote con- trol, 962. for high potential, 957. lever type, 963. oil, operation of, 967. plug tube type, 965. quick break, 964. sparking at, 948. Switching devices, arrangement of, 935. specifications for, 948. Switch jaws, current density for, 442. Symbols, dynamo and motor section, 334. electrical engineering, 1. for wiring plans, 299. mechanical, 1. table of, 6. Synchronizers, descr. of, 416. Synchronizing of alternators, 421. Synchronous converter, def. of, 503. impedance, 400. field current, 383. machines, def. of, 504. losses in, meas. of, 511. motor panels, equip, of, 919. motors, starting of, 431. tests of, 399. theory of, 432. used as condensers, 292. phase modifier, def. of, 502. Synchroscope, definition of, 504. inductor type, 417. Synopsis of report, water power property, 1460. Syntonic apparatus, 1062. signalling, 1059. System, C. G. S., 2. Table of angular dist. betw. brushes 344. of armature slot sizes, 372. of capacity per 1000 feet of aerial wires, 252. of change of hysteresis by heating, 457. of charging current per 1000 feet of aerial circuits, 253-258. of closed circuit cells, 14. of copper wire phys. const., 143. of cost of duct material in place, 307. of paving per sq. yd., 305. of street excavation per cond ft., 306. of double square roots, 45-46. of eddy current factors, 106. of electrical and mechanical units, 1258. of energy and work units, 12. of dissipation in arm. core, 107. of hysteretic constants, 99. of inductive reactances, 242. of magnetization of electromag- nets, 111. of physical quantities, 6. of properties of galv. iron wire, 34. of open circuit cells, 15. of resistance of aluminum wire, 196. of Driver-Harris wire, 207. of magnet wire, 112. of self-induction in millihenrys, 241. of specific ind. cap. of gases, 35. of specific res. of cond., 132. of wire gauges, 141. Tables correcting length of magnet coil, 117-120. of linear space occupied by D. C. cov. wire, 123-126. S. C. cov. wire, 121-123. Tabulation of core loss tests, 384. Tan a, values of, 276. Tangent galvanometer, des. of, 22. track, hangers per mile for, 646. Tantalum lamps, 549. candle-power of, 553. INDEX. 1589 Tapering of conductors, economical, 279. of railway conductors, 793. Teaser, use of, 477. Teeth of armature cores, design of, 357. Telautograph, U. S. Army, 1141c Telegraph cables, 189. codes, 1052. field, 1140. fortress, 1140. U. S. Navy engine, 1202. wire, galv. iron, properties of, 199. hard-drawn, prop, of, 156. steel, properties of, 201. Telegraphy, American methods of, 1040. closed circuit method of, 1040. duplex, 1044. European method 9f, 1040. open circuit method of, 1040, wireless, U. S. Army, 1145. Telephone cables, 188. capacity of, 1085. expenses of, 1087. sizes of, 1086. specifications for, 1083. lines, hotel, 1088. house, 1088. private, 1088. transposition of, 285. method, meas. mutual ind. by, 68. meas. self-induction by, 66. plant, cost of, 1108. depreciation of, 1108. receiver, Bell, 1070. watch, 1070. switchboards, common battery, 1098. design of, 1089. portable, 1141. series multiple, 1092. system, branch terminal, 1093. bridging, 1093. central battery, 1096. central energy, 1096. common battery, 1096. Pupin, 1107. radial, 1117. Telephone system, three-wire, transfer, 1094. two- wire, 1101, 1120. systems, automatic exchange, 1105. bridging of, 1110. common signalling battery, 1115. four-wire selective, 1103. intercommunicating, 1114. Newburgh, 1103. selective, 1102. series party, 1109. step-by-step, 1102. transmission, 1070. two-party selective, 1102. Telephonic transmission, limits of, 1107. Telephones, field, 1140. fortress, 1140. navy, spec, for, 1206. Telephony, duplex, 1106. multiplex, 1106. Telescope for galvanometer, 24, Temperature coef. of copper, 527. of metals, 133. correction, 519. of electric arc, 581. of fire, 1349. of apparatus during test, 508. of transformer windings, 447. or intensity of heat, 1506. rise in armatures, 349, 358. in boosters, 814. in cables, 210. in commutator, 362. in field coils, 352. in generators, U.S. Navy, 1158. in magnet coils, 127. in railway motors, 675. in switchboard devices, 910. in transformers, 491, 498. in, meas. of, 518. test by rise of resistance, 379. tests of dynamos, 378. records of, 381. variation of resistance with, 228. Tensile strength of copper wire, table of, 156. of woods, 1316, 1590 INDEX. Tension and sag in wire spans, 218. Terminal anchorage, 637. Terminals for bonds, 774. Test car, diagram of, 799. lamps, Navy standard, 1172. plate, descr. of, 537. voltage, meas. of, 517. Testing-board, Herrick's, 805. batteries, chloride of silver type, 16. capacity of cables, 325. drop and resistance in trolley lines, 798. dynamo efficiency, Kapp's method, 387. electric plants, 1283. instruments, description of, 13. integrating wattmeters, 1028. joints of cables, 323. large transformer, G. E. method, 490. rail bonds, 801. service meters, 1015. set, S.K.C. high voltage, 461. storage batteries, 882. submarine cables, 331. transformers, 459, 482. Ayrton & Sumpner's method of, 496. data for, 495. Tests of American woods, 1317. of cables, dielectric, 332. of cast iron columns, 1306. of dynamos and motors, 378. of interurban cars, 722. of R. C. wire, Underwriters', 161. of street railway circuits, 798. of synchronous motors, 399. of various types of steam engines, 1439. of, with voltmeter, 74. Thallium, phys. and elec. prop, of, 140. Thawing water pipes by electricity, 1271. power required for etc., 1531. Theater run of high speed railway, 721. Theory of polyphase induction motor, 423. Theory of storage batteries, 872. of synchronous motor, 432. Thermal conductivity of dielectrics, specific, 234. unit, British, 3. Thermit rail welding, 778. Thermometers, comparison of, 1506. Third rail bonding, 778. cost per mile of, 835. insulators, 831. location of, 830. qualities of steel for, 822. shoes, 832. system, 821. Thompson permeameter, use of, 93-96. Thomson elec. welding process, 1271. induction wattmeters, 1005. polyphase induction wattmeters, 1005. recording wattmeters, 998. -Ryan dynamo, special winding of, 351. Thomson's method, res. of gal v. by, 60. testing cap. of cables by, 325. Three conductor cables, G. E. Co., table of, 170. loss of power in sheath of, 293. watts per foot lost in, 212. paper ins. cables, table of, 178. Three-phase alternators, E.M.F. of, 404. armature winding, 413. cables, power carrying cap. of, 216. circuits, arrangement of, 291. charging current per 1000 feet of, 253. energy in, 405. self induction in, 239. delta connection armatures, loss in, 408. distribution railway system, 815. feeder panel, equip, of, 917. generator panels, 912. induction motors, current taken by, 297. potential regulators, 469. lines, balancing of, 287. INDEX. 1591 Three-phase lines, capacity effect in, 249. motors, reading watts in, 398. power, meas. of, 72. transmission, transformers for, 478. rotary converter, 437. panels, equip, of, 919, 924. rotary transformers, armatures of, 441. star connection armatures, loss in, 408. station bus-bars, 933. synchronous motor panels, equip. of, 919. system, balanced, 73. protection by relays, 959. systems, ratio of transformers in, 471. to six-phase connections, 475. transformer connections, 473. transformers, 470. transmission line, ind. react, of, 245. wiring examples, 273. Three voltmeter method, A.C. power by, 71. Three- wire battery system, 899. booster system, diagram of, 902. direct current system examples, 271. Edison system, 355. generator panel, equipment of, 926. telephone system, 1099. street railway system, 807. two-phase system, formula for, 270. variable speed motor system, 354. Throttling calorimeter, 1394. Ties, bearing surface per, 618. durability of, 619. per mile per track, 618. Time-constant, formula for, 239. element mechanism, 958. limit relays, 956. relay, Westinghouse, 960. required for elec. welding, 1272. Tin, fusing effect of current on, 217. phys. and elec. prop, of, 140. spec. res. of, 132. Tin, temperature coef. of, 133. Tire welding, electric, 1272. Tires, data on, 1225. Tirrell regulator for alternators, 409. Toluene, spec. ind. cap. of, 37. Tools and supplies for installing electric work, 1530. Toothed armatures, advantages of, 341. Torpedo circuit closer, 1139. firing, electric, 1213. Torque of induction motors, calc. of, 399. of motor armatures, 353. of polyphase induction motor, 423. of railway motors, 731. Torsion dynamometer, 42. Tower, cooling, 1447. Track and trolley, resistance of, 798. bonding, condition of, 800. bonds, efficiency of, 781. requirements for, 775. data, 618. gang, tools for, 620. laying force, 619. rail, resistance of, 779. return circuit, 771, 786. Traction data, 1224. horse-power of, 653. law of, 110. method, determ. magn. values by, 93. of electromagnets, 110. table of, 111. table of, 655. Tractive coefficient, 662. effort, 661. curves of railway motors, 686. of solenoids, 130. on grades, 657. test for, 1226. force, table of, 654. Train diagram, 787. friction, 613. curve, 679. log for interurban tests, 722. performance diagram, 663, 667. resistance curve for one car train, 683. voltage drop at, 795. 1592 INDEX. Training gear for guns, 1191. Transfer telephone system, 1094. adv. of, 1094. Transformer cells for hydro-electric plant, 931. connections, 472. cores, magnetic densities for, 447. tests, data for, 495. def. of, 503. design, 447. equations, 446. house, single-phase A.C., views of, 943. loss, meas. of, 511. oil, specifications for, 500. panels, constant current, equip. of, 922. Static, def. of, 443. testing, 482. Transformers, ageing of, 498. air-blast type, 449. capacity of, table of, 498. change of hysteresis by heating in, 457. characteristics of, 483. comparative core losses in, 455. comparative expense of operating large and small, 458. connected to rotary converters, 442, 476. connections for wiring, 297. copper loss in, table of, 498. core loss in, 445. table of, 498. cores of American types of, 443. current, descr. of, 945. det. of size of, 295. duties of perfect, 445. efficiency of, 453. test of, 493. exciting current in, table of, 498. for constant current, 464. for constant secondary current, 462. for long distance transmission, arrangement of, 474. for stepping-down high potential, 478. for transmission plants, 870. heat test of, 489, 497, Transformers, hysteresis loss of, 445. improvement in, 454. insulation of, 447. insulation test of, 483, 516. in three-phase system, ratio of, 471. iron loss for, table of, 482. leakage drop in, meas. of, 497. location of, 499. natural draft type, 448. oil-cooled, 448. polarity of, 495. potential, descr. of, 945. power factor of, 458. protection by static interrupter of, 993. regulation of, 458, 491. table of, 498. resistance of, meas. of, 486. rise of temperature in, 498. series type, 464. specifications for, 498. table of capacities of, 296. temperature of windings of, 447. temperature rise in, 520. testing, 459. testing iron and copper losses of, 496. three-phase type, 470. water-cooled type, 449. wiring for, 295. Y or delta connection of, 478. Translating devices, distribution to, 262. Transmission circuits, capacity of, 249. properties of, 238. conductors for high tension, 235. line formulae, 275. inductive react, of three-phase, 245. of known constants, 274. lines, aluminum for high tension, 199. calculation of, 264. circuit breakers protecting, 951. design of, 866. efficiency of, 512. high potential strains on, 981. regulation of, 513. INDEX. 1593 Transmission of power, classif. of, 864. of speech, 1070. plants, switchboards for, 870. system, conductors for, 260. telephonic, limits of, 1107. Transmitters, battery, 1071. Blake, 1072. granular button, 1074. high-power, 1063. magneto, 1071. multi-contact, 1072. single-contact, 1071. solid back, 1072. ungrounded, 1063. wireless telegraph, 1062. Transmitting appliances, table of, 864. Transposition of lines, 285. telephone lines, 1082. Transverse strength of beams, 1308, of woods, 1317. Traversing motor for gun operation, 1134. Trenton beams and channels, 1313. iron beams and channels, 1314. rolled steel beams, 1313. Trial armature coil slots, 372. values for number of armature coils, 373. Trigg works, motors, horse-power of, 1518. Trimming arc lamps, 583. Trip contact for relays, 958. Triple cond. varnished cambric cables, 185. Triplex armature windings, 348. Trip oil switches, use of, 916. Tripping mechanism, 958. Trolley and track, resistance of, 798. cars, energy consumption of, 652. power required for, 656. wiring of, 806. construction, cost of one mile of, 629. for A. C. railways, 640. feeders, arrangement of, 789. line, drop at end of, 800. material per mile of, 643. system, laying out, 785. Trolley and track, wheels, R.P.M. of, 655. wire, dip in, 635. size of, 786. suspension, 637. Troubles of storage batteries, 881. Troy measure, 1500. Truck lights, U. S. Navy, 1181. Trucks of cars, weight of, 734. Trunking, methods of, 1095. Trunk signals, auxiliary, 1096. Truss plank heaters, wiring diag. of, 1267. Tube lighting system, 565. Tubes, collapsing pressure of, 1429, dimensions of boiler, 1428. heating surface of, 1328. regenerative X-ray, 1251. X-ray, 1249. Tubular lamps, navy spec, for, 1173 poles, iron and steel, 633. Tungsten lamps, data on, 553. steel, phys. and elec. prop, of, 140. Turbines, dimensions of hydraulic, 1477. dimensions of Victor, 1477. impulse wheels, diagram of, 1479. installing hydraulic, 1477. inward flow of, 1476. McCormack, diagram of, 1478. outward flow of, 1476. parallel flow, 1476. steam, 1451. U. S. Navy spec, for steam, 1160. water, 1476. Turbo generating sets, spec, for, 1 159. generators, operation of, 1162. Turnout suspension, 638. Turnouts, railway, 620. Turns of wire for transformers, equation for, 446. of wire in coil, calc. of, 113. per armature coil, trial calc. for, 374. Turpentine oil, spec. ind. cap. of, 37, 227. Turret turning gear, navy spec, for, 1187. system, 1165. 1594 INDEX. Twin conductor wire table, U. S. Navy, 1170. Twisted pairs, use of, 1082. wire, res. betw. terminals of, 86. Two-circuit single winding of arma- ture, 342. -conductor cables, watts per foot lost in, 212. motors vs. four motors per car, 729. overhead wires, capacity of, 250. -party selective telephone sys- tems, 1102. -path triplex armature wind., 348. -phase armatures, loss in, 408. armature windings, 412. circuits, arrangement of, 291. feeder panel, equip, of, 918. generator panel, 915. rotary converter, 436. rotary converter panels, 921. rotary transformers, armatures of, 441. systems, formula for, 270. transformer connections, 472. transmission circuit, calc. of ,280. wiring examples, 272. Two-wire direct current system, examples, 271. telephone system, 1101, 1120. Types of plates for batteries, 874. of underground cables, 320. Undamped oscillations, 1068. Underground and submarine cables, tests of, 321. cables, drawing in, 319. locating faults in, 331. types of, 320. conduits and construction, 301. in Chicago, cost of, 317. mains, current variations on, 857. metal, deterioration of, 852. telephone cables, 188. capacity of, 1086. work at New Orleans, 308. Underhill on Electromagnets, 127- 130. Underload circuit breakers, 950. use of, 899. D.C. relay, 962. Underwriters* rules for protection of buildings, 1280. test of R. C. wire, 161. Ungrounded transmitters, 1063. Uniform railway conductors. 792. Unipolar machines, def. of, 504. losses in, meas. of, 512. Uni Signal Company system, 624. Unit difference of potential, 4. electromagnetic, definition of, 5. electro-motive force, 4. lightning arrester, 990. of capacity, 4. of current, 4. of force, 3. of horse-power, 3. of quantity, 4. of resistance, 4. of resistance, definition of, 5. of strength of pole, 4. of work, 3. switch control, A.C. railway system, 710. system, 766. weights, 1513. United States Army, use of elec. in, 1123. Navy electric fuse, 1137. electricity in, 1153. engine specifications for, 1154. generator spec, for, 1156. Units, absolute, 2. C. G. S., 2. derived geometric, 2. derived mechanical, 2. electrical, 4. and mechanical, table of, 1258. engineering, 2. electrostatic, 4. fundamental, 2. geometric, 2. international electrical, 9. magnetic, definition of, 4. of heat, 3. of light, 530, 534. of resistance, 131. symbols and abbreviations for, 6. Universal shunt, Ayrton and Mather, 30. INDEX. 1595 Unstable neutral, 479. Upper harmonics, theory of, 1218. Uses of incandescent lamps, 544, 555. of light, 600. of storage batteries, 886. U. S. Navy rule for ins. res., 85. standard lamps, table of, 1176. U. S. standard gauge for sheet and plate steel and iron, 1299. sheet metal gauge, thickn in millimeters, 1299. Utensils, electric cooking, cost of ' operating, 1259. Vacuum, spec. ind. cap. of, 35. tube light, 565. tubes, exciting source for, 1252. Value of A.C. voltage and current in terms of D.C., 438. Values for numbers of armature coils, 373. for turns per armature coil, 374. Valve, foot, 1447. Vapor lamps, Cooper-Hewitt type, 558. Vapors, specific gravity of, 1512. Variable speed motor work, 354. Variation, def. of, 506. of efficiency of lamps, 547. of resistance with temperature, 228. of voltage in storage battery, 876. Varley loop test, locating faults in cables by, 329. Varnished cambric ins. cables, tables of, 178a-187a. triple cond., 185. Vaseline, spec. ind. cap. of, 37. Vegetable oils, 1497. Velocity, angular, 1505. definition of, 3. definition of, 2. Ventilation fans, navy spec, for, 1196. of armatures, 350. of transformers, 449. Vertical shear of beams, 1308. tubular boilers, 1327. Very high res., meas. of, 79. Victor turbines, dimensions of, 1477. Virtual resistance of storage cell, 883. Voltage and current of A.C. in terms of D.C., 438. curve of railway motors, 669. curves of storage batteries, 883. drop at brush faces, 362. in parallel distribution system, 279. in storage cells, table of, 879. for power transmission, 870. limitation of, 866. loss in storage batteries, 882. meas. of, 62. regulation of transformers, 452. transformers, high-tension station, 938. variation in storage battery, 876. variations, minimizing, 1002. Voltages, discussion of standard, 521. for plating, 1234. Voltaic battery, def. of, 14. Voltameter, silver, description of, 10. Volt, definition of, 5. generation of, 336. international, def. of, 9. specification for determ., 10. value of, 7, 8. Voltmeter, balance used as, 43. Bristol recording single-phase, 1038. electrostatic, Kelvin, 40. method, meas. of current by, 77. Weston type, 41. Voltmeters, description of, 40. electrostatic, use of, 945. high res. for, 75. meas. high res. with, 79. ins. res. of circuits with, 80, ins. res. of wiring system with, 82. res. with, 78. permanent magnet type, 74. tests with, 74. Voltex process for welding aad brazing, 1274. Volume of steam, tables of, 1404. Voynow joint, 778. 1596 INDEX. Vulcanized rubber, electrical prop- erties of, 229. Wagner motor, design of, 430. single-phase motor, connections of, 431. Walmsley's rail tester, 802. Ward-Leonard system of motor control, 354. turret turning gear, 1188. Waring cables, joints in, 191. Warren's method, locating faults in ins. wires by, 330. Watch receiver, 1070. Water analyses, table of, 1366. and mercury columns, pressure of, 1463. -cooled transformers, 449. cubic feet discharged per min., 1470. expansion of, 1362. flow, estimate of, 869. in a stream, 1471, over Weirs, 1473. through an orifice, 1471. through various pipes, 1469* for boiler feed, 1362. friction in pipes of, 1374. gas, 1357. heating by electricity, cost of, 1259. horse-power, tables of, 1475. lifted by suction, 1367. loss of head due to bends in pipes, 1374. mains, effect of current on, 852. meters, electrolytic effect on, 858. motors, regulation of, 514 pipes, thawing out, 1271. power, 1460. data on, 867. synopsis of report on, 1460. yearly expense per H .P. of, 1464. pressure of, 1465. pumping hot, 1367. purification of boiler feed by boiling, 1365. rheostats, 33. rod float gauging, 1471. specific heat of, 1511. Water, specific inductive capacity of, 227. res. of, 133. speed through pump-passages and valves of, 1368. theoretical velocity and discharge of, 1470. tight door alarm, U. S. Navy, 1211. doors, control of, 1198. weight per cubic foot of, 1360. wheels, 1476. racing of, 981. Watt, definition of, 3. -second, value of, 12. value of, 5, 8. Wattless current, def. of, 296. Wattmeter, balance used as, 44. hysteresis tested by, 102. power meas. by, 72. Wattmeters, action of, 1039. bearings of, 1009. Bristol recording single-phase, 1037, calibration of, 1014. Westinghouse integrating, 1016. checking, 72. constants of, 1029. D. C. Sangamo, 1007. Fort Wayne induction, 1005. testing of, 1033. G. E. recording, 1036. testing of, 1030. integrating, testing of, 1013. on inductive circuit, 1000. polyphase and D.C., testing of, 1020. installation of, 1023. integrating, 1004. prepayment, 1010. Sangamo integrating, 1006. testing of, 1035. speed error table for, 1032. speeds of, 1029. Thomson high torque, 1005. polyphase induction, 1005. recording, 998. use of, 72. Westinghouse induction, 999, 1003, INDEX. 1597 Wattmeter, Westinghouse recording, 1037. Weston type, 42. Wright discount, 1008. Watts lost in armature cores, 360. in armature windings, 359. in cables, 210. in core of transformer, 456. transformer cores, 454. per candle of arc lamps, 540. Wave-connected armature wind- ings, 347. form, determination of, 50. E.M.F., 1218. shape, standard, 507, 508. Waves, electromagnetic, 1055. propagation of, 1058. Wax, specific inductive capacity of, 227. Weather-proof aluminum wire stranded, 197. wire, carrying capacity of, 209. table of, 160, 160a-160b-160c. Weaver speed recorder, 1212. Webb, H. S. on water rheostats, 33. Weber photometer, 537. Wehnelt interrupters, 1254. Weight and bulk of bricks, 1322. of A.C. motor equipments, 719. of aluminum, 1514. of brass, sheet and bar, 1323. of car bodies and trucks, 734. of chains, 1496. of conductors, calc. of, 277. formula for, 265. table of, 270. of copper, 143. and brass wire and plates, 1324. per K.W. del'd, curves show- ing, 283. round bolt, 1323. wire, English system, table of, 157. metric system, table of, 158. of flat iron, 1295. of iron and steel, 1294. per sq. ft. in kilograms, 1299. per sq. ft. in lbs., 1299. per sq. ft. in ounces, 1299. per sq. meter in kilograms, 1299. Weight of iron and steel per sq. meter in lbs., 1299. of oil per gallon, 1497. of plate iron, 1298. of rails, 615." of railway equipments, 739. of square and round iron, 1297c of steam, tables of, 1404. of storage cells, 882. of various woods, 1316. cf water per cubic foot, 1300. above 212° F., 1361. of wood, 634. per mile-ohm, def. of, 131. Weights and measures, 1499. apothecaries, 1500. avoirdupois, 1500. metrical equivalents, 1501. troy, 1500. Weiny-Phillips repeater, 1043. Weir dam measurement, 1473. table, 1474. Weirs, Francis' formulae for, 1474 Welding, electric, 1271. H.P. used in electric, 1271. iron pipe, 1272. tires, 1272. Western Electric telephone system U. S. Navy, 1207. Westinghouse A.C. motor character- istics, 715. A.C. railway system, 707. circuit breaker, 951. economy coil, 463. electromagnetic railway, 841. generator panel, 925. induction type wattmeters, 999, 1003. integrating meters, 998. locomotives, 744. method of balancing magnetic circ. in dynamo, 349. mercury arc rectifiers, 481. oil circuit breakers, 969. railway motors, 729. characteristic curves of, 696. rating of, 673. recording meters, 1037. relay, D. C. over- voltage, 962. rotary panel, 925. 1598 INDEX. Westinghouse single-phase potential regulators, 467. switchboard panel, 907. three-wire generator panel, equip- ment of, 926. unit switch control system, 766. wattmeters, calibration data for, 1016. test formula for, 1028. Weston cadmium cell, 19. model, Wheatstone bridge, 56. voltmeter, 41. wattmeter, 42. Wheatstone bridge, 32. Y-box multiplier, 73. Wheatstone bridge, 31. Kelvin type, 59. method, res. meas. by, 56. method, E.M.F. of batteries, 62. Wheels, R.P.M. of trolley, 655. Whistle, electric, navy spec, for, 1210. White core ins. three cond. cable, table of, 170. Winches, deck, 1196. Windage test for dynamos and motors, 383. Winding of electromagnets, 112. field-magnets, 369. plunger solenoids, 128. ring armature, 342. Windings of A.C. armatures, 410. Wind velocity on wire spans, effect of, 219. Wire, aluminum, deflection in feet of, 226. properties of, 194. resistance of stranded, table of, 198. copper, properties of, 143. cotton covered, 163a. galv. iron, water rheostats, 34. gauge, U. S., and weights of iron, 1299. gauges, table of, 141. iron and steel, prop, of, 199. magnet, table of, res. of, 112. Navy standard, 174. paper insulated, 174. resistance, 202, Wire rope, galvanized iron, 1325. notes on uses of, 1494. standard hoisting, 1326. transmission of power by. 1495. transmission or haulage by, 1325. ropes, horse-power of, 1495. rubber covered, 161. sizes for armature coils, 372. solid copper, table of, 154. spans, tension and sag in, 218. steel, properties of, 201. stranded copper, table of, 155 strands, table of, 142. table, U. S. Navy, 1169. tables, copper, A.I.E.E., 146. condensed table, 154. explan. of, 145. varnished cambric ins., 179. weather proof, 160, 160a, 1606. weight of copper, table of, 157. Wireless telegraphy receivers, 1064. theory of, 1055. transmitters, 1062. U. S. Army, 1145. Wires and cables, properties of, 131. cambric ins., tables of, 179. current carrying capacity of, 208. enameled, table of, 187b. fusing effect of current on, 217. gutta-percha covered, jointing of, 193. navy standard, table of, 174. paper ins. G. E. tables of, 174-178. rubber ins. G. E. tables of, 164- 172. space occupied by cotton covered, tables of, 121-126. suspended from points not in same level, sag in, 223. U. S. Navy spec, for, 1167. Wiring bells, 293. diagrams of cars, 806. for transformers, 295. of cars, 746 for heaters, diagram of, 1267. of houses, 279. plans, standard symbols for, 299. specifications, U. S. Navy, 1167. INDEX. 1599 Wiring system, ins. res. of, 82. Wood as fuel, 1356. beams, strength of, 1318. mill, power required to run tools for, 1519. specific inductive capacity of, 227. tests of American, 1317. weight per cord of, 1356. Wood working machinery, power to run, 1519. tools, power required for, 1522. Wooden poles, contents of, 633. painting of, 806. stave pipe, 1468. Woods, American, wt. and value as fuel of, 1349. crushing strengths of, 1316. pressure to indent sV, 1316. properties of various, 1316. relative strength for cross break- ing, 1316. shearing strength with the grain of, 1316. specific gravity, table of, 1512. tensile strength of, 1316. value in tons of coal, 1349. weight of, 634. per cubic foot of, 1316. per ft. B. M., 1316. Woolf process, disinfecting by, 1244. Work done by conductors in magn. field, 109. international unit of, 10. unit of, 3. Work units compared with energy units, 12. Workshop method, res. of batteries, 61. Wright demand meter, 1008, discount meter, 1008. Wrought iron, permeability of, 89. phys. and elec. prop, of, 137. pipe, dimensions of, 1426. poles, weight of, 633. qualities of, 824. Wurts lightning arresters, 984. X-rays, polarization of, 1248. theory of, 1248. tubes for, 1249. Xylene, spec. ind. cap. of, 37. Y-box multiplier, Weston, 73. -connection of transformers, 478. Yokes, field magnet, general data on, 352. Zerener system of welding, 1274. Zero instrument, Northrup, 26. Zinc amalgam for standard cell, 11. for boiler scale, 1365. phys. and elec. prop, of, 136, 140. spec. res. of, 132. sulphate for standard cell, 11. spec. res. of, 133. temperature coef. of, 133. Zone, commutating, 350. One of the Complete Group of Weston Round Pattern A. C, Switchboard Instruments A Weston Miniature Precision Instrument. This Group is for D. O. Service, Switchboard and Portable "For Active Service" From a purely scientific standpoint this Company is proud of its contributions to the theory of Elec- trical Measurement, it is still more proud of Indicating Electrical Measuring Instruments They are thoroughly worthy to represent Weston ideas and ideals in the field of active service. As these are the instruments most frequently encoun- tered in commercial work they are likewise most desirable for educational purposes. There are Weston Instruments in great variety for portable or switch- board service on A. C. or D. C. Cir- cuits and many others for special purposes. Write for Bulletins or Catalogs describing those which interest you. Weston Electrical Instrument Co. 83 Weston Ave., Newark, N. J. 23 Branch Offices in the Larger Cities ELECTRICAL WIRES AND CABLES TELEPHONE WIRE TROLLEY WIRE POWER CABLES WEATHERPROOF WIRE MAGNET WIRE ANNUNCIATOR WIRE LAMP CORD AUTOMOBILE CABLES WIRE ROPE STRAND MADE BY John A. Roebling's Sons Co. Main Office and Works: TRENTON, N. J. Agencies and Branches: NEW YORK BOSTON CHICAGO SAN FRANCISCO CLEVELAND PHILADELPHIA ATLANTA SEATTLE LOS ANGELES PORTLAND, ORE. TIMELINESS! The big feature which distinguishes the ELECTRICAL REVIEW is the timeliness of its editorial contents. With the entrance of the United States into the war, a multitude of problems arose which necessitated a readjustment ot operating and selling conditions in every branch of the industry. THE ELECTRICAL REVIEW solved their problems by collecting information on how England, France and Canada mobilized their industries. This information enabled the industry to profit by the experiences of the foreign bel- ligerent nations and was published in the Electrical Mobilization Number May 26, 1917 Other information of great importance to the electrical industry was contained in the Shipbuilding Number July 7, 1917 Electric Lighting and Fall Trade Number September 1, 1917 Power Plant Number September 15. 1917 National Electrical Contractors' Association Pre -Convention Number October 6, 1917 National Electrical Contractors 9 Association Convention Report Number October 20, 1917 The year 1917, on account of the war, has been such an abnormal one that it will be difficult for the industry to plan ahead as custom- arily without knowledge of what has been done during the past year. The ELECTRICAL REVIEW has, therefore, set its large organiza- tion to work gathering statistics covering the activities during the past year, which will be presented together with it6 usual review of developments in the New Year's Number January 5, 1918 This editorial policy of pointing the way and analyzing con- ditions has made the ELECTRICAL REVIEW indispensable to the electrical industry and assures intense reader interest by every element - manufacturers - engineers - central stations - contract- ors - dealers. That is the reason advertising in its columns pays. THE INTERNATIONAL TRADE PRESS, INC. MONADNOCK BLOCK CHICAGO Publishers also of the Road Maker, Engineering and Cement World and various catalogs and annuals. DOSSERT CONNECTORS 2- Way Type A Showing: Details. Dossert Connectors eliminate entirely the use of solder in making electrical con- nections and splices, and are approved for use without solder by the National Board of Fire Underwriters for all classes of wiring. By their use much labor is saved and splices obtained that will withstand any over- load. Many careful tests show that a splice made by means of a Dossert Connector will not heat as much as the cable which it connects when the cable is heavily over- loaded. Type A Connectors are for use on cables, stranded, or solid wires, rods and tubing* They are simple and effective, and by their use splices can be quickly made in conduct" ors of any size. Type A Connectors, however should not be used on a cable that is to be subjected to heavy tensile strains. Part Cross-sectional View of Type B 2-Way Type B Connectors are for use on stranded wires or cables only, and are designed to make a joint which will withstand heavy tensile strains. They are not made for wires smaller than No. 0. The Cable Tap is used to connect a branch wire, rod or bleeder, to a main wire, rod or feeder. It does not splice the main, but simply clamps on to it. Branch wire is con- nected to cable tap by means of a nut and sleeve as shown in Type A cut. With Dossert devices any combination of different sizes of cables, stranded and solid wires, rods and tubing can be connected to- gether. The cable tap will tap from any size main to any size branch. Terminal and switchboard lugs, front or back connected; angle and swivel lugs. Insulated connectors; two-ways, three-ways, equalizers, cable an- chors, reducers, elbows, Y's, service box lugs Cable Tap and plugs, grounding devices and stud con- nectors for threaded rods or flat strips or blocks. Send for Tenth Year Catalogue. Dossert & Company H. B. LOGAN, President 242 West 41st Street New York, N. Y. SEND FOR IT TO-DAY Catalog of Books on ELECTRICITY Classified by subjects as follows: Alternating Currents. Armatures. Automobiles. Batteries. Bells. Biography and History. Cables and Wires. Dictionaries, Directories. Direct Currents Dynamo Electric Machines — A. C. and D. C. Care, Repair, Testing. Electricity and Magnetism. Advanced Texts. Books for Amateurs Elementary Texts. In Mines. Various Applications. Electrochemistry. Electrometallurgy. Electroplating. Induction Coils. Lighting. Lightning Protection. Measurements and Meters- Calculations. Miscellaneous. Patents. Pocket Books. Power Plants and Trans= mission. Radium. Railways. Static Electricity. Switchboards and Circuits. Telegraphy- Cables, Codes. Wireless. Telephony- Wireless. Transformers. Wireless Telegraphy. Wiring for Light and Power. X Rays and Radiography. 96 Pages. 900 Books. 32 Subjects No matter what branch of Electrical Engi- neering you want information about, you'll find a book that gives it in this catalog. We send it free and are always glad to an- swer any questions about Scientific Books. D. Van Nostrand Company Publishers and Sellers of Scientific Books 25 PARK PLACE, NEW YORK