LIBRARY OF CONGRESS. 
 
 Chap,...„.. Cop^ight N(k 
 
 UNITED STATES OF AMERICA. 
 
STEAM ENGINEERING SERIES 
 
 VOLUME I. 
 
 BOILERS AND FURNACES 
 
BY THE SAME AUTHOR 
 
 PUMPING MACHINERY 
 
 A PRACTICAL HAND-BOOK 
 
 RELATING TO THE CONSTRUCTION 
 
 AND MANAGEMENT OF 
 
 STEAM AND POWER PUMPING MACHINES 
 
 BY 
 
 WILLIAM M. BARR 
 
 Member American Society Mechanical Engineers 
 
 WITH UPWARDS OF TWO HUNDRED AND SEVENTY ENGRAVINGS, 
 COVERING EVERY ESSENTIAL DETAIL IN PUMP CONSTRUCTION 
 
 One Volume. Octavo. 450 pages 
 Price, §5.00 
 
Boilers and Furnaces 
 
 CONSIDERED IN THEIR RELATIONS TO 
 
 STEAM ENGINEERING 
 
 WILLIAM M. BARR 
 
 MEMBER AMERICAN SOCIETY MECHANICAL ENGINEERS 
 
 WITH , UPWARDS OF FOUR HUNDRED AND FIFTY ENGRAVINGS OF BOILER AND FURNACE DETAILS 
 FROM DRAWINGS EXECUTED EXPRESSLY FOR THIS WORK 
 
 ^ APR 2 21898 
 
 PHILADELPHIA 
 THE FLORENCE COMPANY^^of^g! 
 
 It 
 
 L-- 
 
 WO COPIES RECElVED^ 
 
 2nd COPY, 0. ^^ ^\ 4- 
 
 1898* 
 

 rvioi 
 
 Copyright, 1898 
 
 BY 
 
 William M. Barr 
 
 ^^ 37(^^0 
 
 ADyERTISEMENT 
 
 This is a subscription book. Price, $3.00. Copies will be sent to any 
 address, charges prepaid, upon receipt of price by the publishers 
 
 THE FLORENCE COMPANY 
 
 PHILADELPHIA, PA. 
 
 ^i> 
 
 P. O. Box 803 
 
PREFACE. 
 
 A WORD of explanation may not be out of place in regard to the 
 bearing which this volume may have upon a former treatise of mine on 
 " High-Pressure Steam Boilers" (1880), and which is now out of print. 
 Many requests have come to me from time to time to revise that book 
 and bring it up to date. But twenty years is a long interval. Boiler 
 pressures were much lower then than now. Wrought iron was the 
 material then employed in boiler construction, now it is rarely met with. 
 The conditions having wholly changed, the problem of revision is im- 
 practicable. The first book must be set aside as a thing of the past and 
 the subject taken up anew, with special reference to the exacting con- 
 ditions which obtain at this time in both design and performance. This 
 book differs somewhat from the former one in being essentially one of 
 constructive detail, in the preparation of which I have endeavored to 
 present, by means of well-chosen illustrations, the latest and best prac- 
 tice in steam-boiler design. 
 
 Mild steel as a material for steam boilers has practically displaced 
 wrought iron. In its physical qualities it leaves scarcely anything to be 
 desired. Tensile tests of mild steel plates, possessing those chemical 
 properties which have given the best physical results, are allotted a con- 
 siderable space in this book. They are made to cover as wide a range 
 as seems necessary for ordinary stationary boiler work, or sufficiently so 
 for the preparation of any specifications requiring plates not less than 
 one-fourth inch, and not more than three-fourths inch thick. 
 
 Riveted joints have always been of first importance in boiler con- 
 struction. The numerous records of tests of purely experimental joints, 
 or those not used in boiler-making, as well as those made to the actual 
 working dimensions usually employed when riveting together plates of 
 the various thicknesses, given in Chapter III., are mainly upon speci- 
 mens prepared by direction of the Bureau of Steam Engineering of the 
 Navy Department. These experimental tests were made at the Water- 
 town Arsenal, and extend over more than ten years. They are of the 
 greatest practical value to persons engaged in the designing and con- 
 struction of steam boilers for high pressures, and it is much to be re- 
 gretted that the records of these tests are not more generally accessible. 
 In the selection of examples for illustration and record in this volume 
 great care has been exercised to include all, and only, such as shall be 
 
 5 
 
6 PREFACE 
 
 useful in stationary steam boiler design and construction. Probably no 
 such similar amount of accurate and reliable data on this important 
 detail in mechanical engineering has ever before been submitted in a 
 single volume. 
 
 The limitations imposed as to the size of the completed volume pre- 
 vented the introduction of subject matter or illustrations relating to the 
 early history and development of steam boilers, as well as illustrated 
 reference to some of the more recent examples of design and construc- 
 tion. This increase in subject matter could only have been accomplished 
 by adding to the number of pages or by reduction in the size of the en- 
 gravings. No doubt many of the latter could have been slightly reduced 
 without detriment, but it was thought that a fewer number of clearly 
 executed fundamental details would be of more value to a designer or 
 student than a larger number wanting in proper mechanical execution 
 or clearness of detail. After all, it will be seen that the omission is not 
 so much in types of boilers, to which considerable space has been given, 
 as it has been in the variations of these types. 
 
 Boilers for steamships will be described in the volume on ' ' Marine 
 Engines," to be included in this series, they requiring more or less 
 special treatment in connection with other subjects beyond the scope of 
 this volume ; so also the design and construction of locomotive boilers, 
 which will receive special consideration in the volume on " Locomotive 
 Engines," also to be included in this series. 
 
 The consideration of chimneys in this volume includes only the 
 necessary dimensions of diameter and height for boiler plants from 20 
 to 1000 horse-power, no space having been given to their consideration 
 as an isolated structure, for the reason that an illustrated volume on the 
 design and construction of "Brick and Metal Chimneys" is in prepa- 
 ration for immediate publication.! 
 
 William M. Barr. 
 
 Philadelphia, March, 1898. 
 
CONTENTS. 
 
 CHAPTER PAGE 
 
 I. — Furnace Combustion 9 
 
 II. — Materials of Construction 25 
 
 III. — Riveted Joints 50 
 
 IV. — Welding and Flanging 95 
 
 V. — Details and Strength of Construction no 
 
 VI. — Externally Fired Boilers 178 
 
 VII. — Boiler Furnaces and Settings 206 
 
 VIII. — Internally Fired Boilers 256 
 
 IX. — Sectional and Water-Tube Boilers 296 
 
 X. — Boiler Mountings and Safety Apparatus 343 
 
 XI. — Chimneys 390 
 
BOILERS AND FURNACES. 
 
 CHAPTER I. 
 
 FURNACE COMBUSTION. 
 
 Combustion in steam engineering means the controlled chemical 
 combination of the elements carbon and hydrogen in the fuel with the 
 oxygen of the atmosphere, by which an evolution of heat is secured 
 and maintained in a suitably constructed furnace for the purpose of 
 g;enerating steam. 
 
 The Unit of Work in steam engineering is a gravitation unit known 
 as the foot-pound, or the amount of work required to raise one pound 
 one foot high against the force of gravity, and is entirely independent 
 of the time it takes to do it. 
 
 Horse-Power is a unit of work employed in steam engineering to 
 denote the power-rating of a steam boiler and engine, or the power 
 transmitted by a belt, shaft, etc. In computing work done it is always 
 independent of the time taken to do it, but in computing horse-power 
 time is an essential element. The unit of horse-power is a gravity 
 measurement, and represents 33,000 pounds raised one foot high in one 
 minute against the force of gravity. 
 
 Chemical and Physical Changes. — If a combustible like wood 
 is thrown upon a fire it disappears, and nothing visible remains but 
 ashes. Changes of this kind, in which a substance disappears and 
 something else is formed in its place, are known as chemical changes. 
 
 Changes which do not affect the composition of substances are called 
 physical changes. The freezing of water into ice or the evaporation of 
 water into steam is a physical change, because the composition of the 
 substance in these three states is the same. 
 
 Chemical Changes due to combustion always involve the conver- 
 sion of the substances burnt into new substances, and these final products 
 are compounds ; the constituent elements which enter into or make the 
 compound always combine according to certain definite proportions, 
 either by weight or measure. In the combustion of carbon, for ex- 
 ample, we may have either of two possible combinations : One atom 
 of carbon uniting with one atom of oxygen produces carbonic oxide, 
 CO ; the atomic weights would be. Carbon, 12 + Oxygen, 16 = Car- 
 bonic Oxide, 28. By percentages instead of the atomic weights we 
 
 2 9 
 
lO. BOILERS AND FURNACES 
 
 would have in loo parts of carbonic oxide, Carbon, 42.86 + Oxygen, 
 57.14 = 100.00. 
 
 In the case of carbonic acid gas, COg, one atom of carbon unites 
 with two atoms of oxygen, thus : Carbon, 12 + 2 Oxygen, 16 = 44 = 
 the atomic weight of COg. By percentages instead of atomic weights 
 we have in 100 parts of carbonic acid gas 27.27 parts of carbon and 
 72.73 parts of oxygen. These proportions constitute the only two 
 direct inorganic compounds of carbon and oxygen. 
 
 Hydrogen gas burnt in oxygen combines in the proportion of ii.i 
 parts of hydrogen with 88.9 parts of oxygen to form 100 parts of water. 
 The ratio of the weights of equal volumes of hydrogen and oxygen is as 
 I : 16. If, then, two volumes of hydrogen combined with one volume 
 of oxygen, the ratio between the weights is 2 : 16, or i : 8, and these 
 gases will combine in no other proportions to form water ; any excess 
 of either gas will remain unchanged. 
 
 Furnace Combustion requires a combustible such as carbon or 
 hydrogen, which is supplied by the fuel, and a supporter of combustion, 
 — oxygen, — supplied by the atmosphere ; such combustion is accompa- 
 nied by flame and incandescence, the former a result of the combustion 
 of the hydro-carbon gases of the fuel, the latter, glowing carbon from 
 which the gases have been expelled. The color and intensity of incan- 
 descence is dependent upon the temperature, red indicating the lowest 
 and a dazzling white light the highest. Most solids emit light or become 
 a dull red at about 750° Fahr. The temperature of coke burned under 
 favorable conditions approximates 3600° Fahr. , and at this temperature 
 emits an intensely brilliant white light. 
 
 Hydrogen, H. At. Wt., i. — When pure, hydrogen is colorless, 
 tasteless, and inodorous. It occurs in nature in combination with carbon 
 in varying proportions. The compound which contains it in greatest 
 abundance is marsh-gas, of which hydrogen forms four parts, the for- 
 mula being CH^. Under ordinary temperatures hydrogen has no ten- 
 dency to enter into combination with other substances. It combines 
 with eight parts of oxygen to form water, but this combination does not 
 take place spontaneously. Pure hydrogen burns in the atmosphere 
 with a pale blue light scarcely perceptible in full daylight, giving off an 
 intense heat. The heat units of one pound of hydrogen burned in 
 oxygen were ascertained by Favre & Silbermann to be 62,032. This 
 is not equalled by any other known substance. 
 
 Carbon, C. At. Wt., 12. — This is one of the most widely diffused 
 and abundant of the elements, being the central element in organic 
 nature. It exists in three different forms, as diamond, graphite, and 
 charcoal, each having its own physical properties, the chemical proper- 
 ties being the same. Carbon is an inactive element at ordinary tempera- 
 tures, but when raised to its kindling temperature, about red heat, it 
 unites with oxygen and combustion ensues. 
 
FURNACE COMBUSTION II 
 
 Sulphur, S. At. Wt., 32. — Nearly all coals contain sulphur in 
 combination with iron in the form of golden-yellow crystals, commonly 
 known as iron-pyrites. It burns with a clear blue, feebly luminous 
 flame, being converted into sulphurous oxide, SOg. The heat devel- 
 oped by the combustion of any sulphur present in coal is not taken into 
 account in steam engineering. The amount of sulphur in coal ranges 
 from 0.3 to 5 per cent. 
 
 Oxygen, O. At. Wt., 16. — When free or uncombined, oxygen is 
 known only in the gaseous state ; when pure, it is colorless, tasteless, and 
 inodorous. It is the sustaining element in all the ordinary phenomena 
 of combustion. 
 
 Ignition is simply the incandescence of a body unattended by chem- 
 ical change, and must not be confused with combustion : the ignition of 
 solids is a source of light, the combustion of solids is a source of heat. 
 Every combustible must be heated to a certain definite temperature 
 before it will combine with oxygen. This temperature is usually called 
 the point of ignition, or its kindling temperature. 
 
 Flame is simply gas burning on its exterior surface ; its color and 
 brightness depend not only on its degree of temperature, but upon the 
 presence of solid incandescent particles in the flame ; purely gaseous 
 substances do not become highly luminous when burning. The struct- 
 ure of flame from burning wood or coal consists of three parts : first, a 
 central core of unburned hydro-carbon gas ; second, an envelope of 
 burning hydrogen and carbon, the oxidizing portion of the flame ; third, 
 still another envelope consisting of aqueous vapor and other products 
 corresponding to the combustibles contained in the issuing gas. 
 
 The Air. — The oxygen needed for furnace combustion is supplied 
 by the air, which consists essentially of two gases, nitrogen and oxygen, 
 in the proportion of 79 volumes of nitrogen to 21 volumes of oxygen, 
 or, by weight, 77 per cent, of nitrogen to 23 per cent, of oxygen. Air 
 is a mechanical mixture of these two gases and not a chemical com- 
 pound ; their union in the proportion given above is distinguished by 
 no properties which may not be attributed individually to these gases. 
 One of the most important properties of the atmosphere is that of 
 weight, without which furnace combustion would be a much more com- 
 plex operation than at present, because a different mechanical agency 
 would be required for producing draught in the furnace. The mean 
 pressure of the atmosphere at the mean level of the sea is equal to 14.7 
 pounds per square inch, or 21 16.8 pounds per square foot. 
 
 Nitrogen, N. At. Wt., 14. — By volume and by weight nitrogen is 
 the principal constituent of air ; it is a colorless, tasteless, inodorous gas ; 
 its specific gravity is . 97 1 , air = i . The specific heat of nitrogen is o. 244 
 at constant pressure. It is an inert gas in the furnace, not being itself 
 combustible, nor will it support combustion, but the presence of nitrogen 
 is useful, inasmuch as it greatly modifies the intensity of combustion, 
 
12 BOILERS AND FURNACES 
 
 and at the same time does not alter the chemical relations of the oxygen 
 to the combustible substances. If it were not for the presence of nitro- 
 gen in the air, or suppose the atmosphere were wholly oxygen, com- 
 bustion would be very hazardous, — in fact, a fatal occurrence, because 
 the coals would burn so violently as to have a combustion wholly uncon- 
 trollable ; nor would the combustion stop with the fuel, but the oxygen 
 would attack the grate bars, the furnace front, the boiler, and everything 
 else made of iron, for the latter substance burns even more violently in 
 oxygen gas than coal. Nitrogen is useful, then, in the good offices which 
 it performs in lessening the avidity of oxygen during combustion, making 
 fire a moderate, useful, and easily controlled servant in the service of 
 man ; its negative qualities make nitrogen a safe substance, and, while it 
 plays no active part in combustion, it is the means by which the oxygen 
 is delivered in the body of incandescent fuel, where it parts company 
 with the oxygen and passes on through the fire, its less specific gravity 
 aiding materially in producing the draught necessary for furnace com- 
 bustion. 
 
 Fuel is a term used in steam engineering to include combustibles of 
 every sort that can be utilized to generate heat. The combustibles in 
 common use consist almost wholly of carbon and hydrogen. 
 
 ^A^ood. — Commercially, woods are distinguished as hard or soft, 
 the former including the heavy, compact woods, as oak, hickory, etc. ; 
 the latter, pine, spruce, etc. The average composition of five dry 
 woods, including beech, oak, birch, poplar, willow, by Chevandier, 
 was: Carbon, 50.00; hydrogen, 6.00; oxygen, 41.00; nitrogen, i.oo ; 
 ash, 2.00=100.00. 
 
 The hydrogen present in wood is not available as fuel, owing to the 
 presence of oxygen, these two gases uniting to form water. Carbon 
 averaging 50 per cent, is present in all woods, and is the only com- 
 bustible available for generating heat. The American Society of Me- 
 chanical Engineers, in their rules for boiler tests, assume one pound of 
 wood to equal 0.4 pound of coal. 
 
 Tan is the spent bark from which the tannic acid has been extracted 
 in the process of tanning leather ; the barks commonly used are oak and 
 hemlock. The principal drawback to tan as a fuel is its contained mois- 
 ture, and for this reason special furnaces are made for burning it. Tan 
 bark, as commonly used for fuel, will yield about 3600 heat units per 
 pound, which is one-half the value of ordinary dry wood and about one- 
 fourth the value of good bituminous coal. 
 
 Bagasse is the woody fibre of sugar-cane after the saccharine juices 
 have been expelled for sugar-making. Special furnaces have been con- 
 trived for burning it, and with fair results ; the contained water is about 
 50 per cent, of the gross weight ; the remaining fibre is not unlike wood 
 in its heat-giving power. On an average, six pounds of bagasse are 
 equivalent to one pound of good bituminous coal. 
 
BITUMINOUS COAL 1 3 
 
 Peat is organic matter undergoing a gradual carbonization, the 
 oxygen of the plants being liberated under special conditions of mois- 
 ture and heat, leaving a spongy carbonaceous mass. When dried, peat 
 consists of about 58 per cent, of carbon. Scarcely any peat is used for 
 fuel in this country because of the abundance and superior heating 
 power of coal. 
 
 Lignite occupies a position historically between peat and bituminous 
 coal ; it is of later origin than bituminous coal and in a less advanced 
 state of decomposition. It varies considerably in appearance and struct- 
 ure : the fracture is uneven, presenting a brown to a very dark brown- 
 black color, with a dull and frequently fatty lustre ; it crumbles easily in 
 handling and will not bear rough transportation, nor will it bear long- 
 continued exposure to the weather without crumbhng. It is non-caking 
 in the fire and yields but a moderate heat, below the average of bitu- 
 minous coals. In its natural state, lignite contains from 10 to 30 per 
 cent, of water. 
 
 Brown Coal is a term sometimes used, though not correctly, as 
 interchangeable with lignite ; it is in a more advanced stage of decom- 
 position, — that is, it is nearer the bituminous coal series than the lig- 
 nites. Brown coals contain less fixed carbon than the coals of the car- 
 boniferous epoch, and usually a much larger percentage of moisture 
 when freshly mined, both of which tell against the coal commercially. 
 
 Bituminous Coal. — In physical properties bituminous coals vary 
 so widely that a single description cannot include them all. The color 
 averages from brown to pitch black ; the lustre is vitreous, resinous, 
 sometimes silky ; the structure may be compact, slaty, columnar, and 
 even fibrous ; the fracture, irrespective of structural joints and cleavage, 
 is conchoidal, often flat and rectangular, and sometimes fibrous. It is 
 distinctive of these coals to burn with a more or less yellow flame and 
 smoke. 
 
 In composition, bituminous coals range in volatile matter from 15 to 
 50 per cent., including 2 to 12 per cent, of contained moisture. The 
 fixed carbon varies from 40 to 75 per cent. , the earthy matter from 2 to 20 
 per cent. Sulphur is almost always present in bituminous coals as pyrite. 
 
 Bituminous coals are broadly classed as caking and non-caking coals. 
 
 Caking Coal. — The characteristic circumstance that lumps of coal, 
 either large or small, are rendered pasty by the action of the heat, and 
 will cohere in the fire to form a spongy-looking mass which may cover 
 the whole surface of the grate, is the property called caking ; the fixed 
 carbon remaining after the expulsion of the gases is called coke. It is 
 more difficult to completely and economically burn caking coals in a 
 boiler furnace than those of non-caking variety beckuse of this tendency 
 to run together and prevent the free flow of air through the fuel ; such 
 masses of burning coal must be broken up with a slice-bar at frequent 
 intervals to facilitate its combustion. 
 
14 BOILERS AND FURNACES 
 
 Non-caki7ig coals burn free in the fire, not unlike charcoal or soft 
 coke. The action of heat does not cause the coal to fuse or run together 
 into large masses which require afterwards to be broken up to allow free 
 access of air through the body of the burning coal to secure quick com- 
 bustion. Free-burning coal is the same as non-caking coal. 
 
 Cannel Coal is a variety of bituminous coal very rich in hydrogen. 
 This coal kindles readily and burns without melting, emitting a bright 
 flame. When thrown upon an active fire the piece splits into fragments, 
 producing a crackling noise. In appearance this coal differs from all 
 bituminous coals : its structure is a compact mass and more nearly homo- 
 geneous than others ; it varies from brown to black in color, and has a 
 dull resinous lustre. When broken it does not preserve any distinct 
 order of fracture, and is liable to split in any direction. It is highly 
 esteemed as a gas coal, but is not much used as a steam coal except 
 locally near the mines. 
 
 Block Coal is a representative non-caking bituminous coal. It 
 occurs in several of the Western States, but is found at its best in Indiana. 
 It has a laminated structure, composed of alternate thin layers of vitreous 
 dull black coal and mineral charcoal. Chemically it does not appear to 
 differ from caking coal, but in burning it behaves quite differently : it 
 does not swell, shoot out jets of gas, nor form a cake by running to- 
 gether, but retains its shape until entirely consumed to a white ash which 
 contains no trace of clinker, its behavior in the furnace being quite like 
 that of hickory wood, burning with a uniform flame that spreads evenly 
 over the exposed surface. 
 
 Coke is the fixed carbon and earthy matter remaining after the dis- 
 tillation of the gases from bituminous coal. The only coke of any com- 
 mercial value is made from caking coals. The quality of the coke is 
 affected by the temperature at which it is made : the higher the tempera- 
 ture and the longer it is exposed to that temperature, the harder, more 
 dense, and less easily combustible will be the coke. 
 
 Semi-Bituminous Coal partakes somewhat of the nature of anthra- 
 cite coal. Modified by volatile matter, it forms an excellent fuel for steam- 
 boiler furnaces. In appearance it resembles anthracite coal rather than 
 bituminous ; its fracture, however, as compared with anthracite, is less 
 conchoidal ; it is not so hard, and is of less specific gravity. When 
 thrown upon the fire it kindles more readily and burns faster than anthra- 
 cite, but without the smoke and soot characteristic of bituminous coal. 
 
 Semi-Anthracite Coals are restricted to such as average from 
 6 to 8 per cent, of volatile combustible matter, but otherwise have the 
 physical characteristics of true anthracites. These coals are in high 
 estimation for steaming purposes, as they kindle easily and burn much 
 more freely than do the harder anthracite coals. 
 
 Anthracite Coal is slow to ignite ; it does not soften or swell in the 
 fire ; the flame is quite short and nearly transparent, having a yellowish 
 
ANTHRACITE COAL 1 5 
 
 tinge at first, changing to a soft blue, with occasionally a red tinge, and 
 gives ofif no smoke. When broken it presents a conchoidal appearance 
 and is quite homogeneous in structure. In general it is compact, slaty, 
 grayish-black, splendent, varying somewhat according to the locality at 
 which it is mined. It is not found in large quantities outside of Eastern 
 Pennsylvania. 
 
 Culm is fine anthracite coal. Formerly this was waste product and 
 had no commercial value. Culm heaps are now being carefully screened 
 and assorted into sizes for use in steam-boiler furnaces. The sizing must 
 be uniform or the smaller pieces will drop into and clog the passages 
 between the larger ones and obstruct the free passage of air through the 
 fire. The percentage of ash in sizes recovered from culm is much 
 greater than in the larger sizes, such as chestnut or stove coal. A very 
 strong draught is required for burning fine anthracites, and a forced 
 draught is often resorted to or required for their rapid combustion. 
 
 Petroleum is a natural hydro-carbon oil found in large quantities, 
 especially in Pennsylvania and Ohio, the weight per gallon varying from 
 6 to 7 pounds. The specific gravity of petroleum averages about o.8, 
 with variations on either side. The composition of crude Pennsylvania 
 oil averages: Carbon, 84.00; hydrogen, 13.75; water, 2.25 = 100, 
 Heat units, 20,746 per pound. Evaporative power of one pound of oil 
 from and at 212° Fahr. = 21.47 pounds of water. 
 
 One of the interesting exhibits at the World's Fair, Chicago, was 
 the boiler plant which was furnished with crude oil from the Lima, 
 Ohio, district for fuel. Never before was there such an opportunity in 
 this country for testing liquid fuel oij so large a scale. The quantity of 
 petroleum used for firing the main boiler plant amounted to upwards of 
 31,000 tons, and the work done is stated to have been 32,316,000 horse- 
 power hours, or about 2. i pounds of oil per horse-power per hour. 
 
 Natural Gas is found locally in Western Pennsylvania, Northern 
 Ohio, and Central Indiana in paying quantities ; in lesser quantities it 
 is found in many other localities. Natural gas is an ideal fuel if used 
 near the source of supply, as no labor is required in its use except to 
 regulate the supply in the furnace ; it is not difficult to regulate the 
 supply of air to insure perfect combustion ; there is no soot, ashes, or 
 other debris. The composition of natural gas at Findlay, Ohio, is, — 
 
 By weight. By volume. 
 
 Hydrogen 0.27 2.18 
 
 Marsh gas 90-38 92.60 
 
 Carbonic oxide 0.86 0.50 
 
 Olefiant gas 0.53 0.31 
 
 Carbonic acid 0.70 0.26 
 
 Nitrogen 6.18 3,61 
 
 Oxygen 0.66 0.34 
 
 Sulphydric acid 0.42 0.20 
 
 100.00 100.00 
 
l6 BOILERS AND FURNACES 
 
 The heat units in one pound of Findlay, Ohio, natural gas = 21,520. 
 
 Evaporative power of one pound of the above gas from and at 
 212° F. = 22.27 pounds of water. 
 
 Natural-gas tests under boiler for steam-making at Pittsburg, Penn- 
 sylvania, show that one pound of good bituminous coal equals from 
 7^ to 12^ cubic feet of natural gas ; other experiments show that 1000 
 cubic feet of natural gas equal from 80 to 133 pounds of bituminous 
 coal, a variation of more than 60 per cent, between the two extremes ; 
 quality of coal and manipulation of furnace accounts for much of this 
 difference. 
 
 PRODUCTS OF COMBUSTION. 
 
 The combustible elements of wood and coal are carbon, hydrogen, 
 and sulphur. The supporter of combustion is the oxygen of the air. 
 
 Carbon when burnt in oxygen yields two products, depending upon 
 the supply of oxygen in the furnace, viz. : 
 
 
 Formula. 
 
 Combustion. 
 
 Product. 
 
 Carbonic acid gas . . 
 
 . . CO2 
 
 Complete. 
 
 Incombustible 
 
 Carbonic oxide gas . 
 
 . . CO 
 
 Incomplete. 
 
 Combustible. 
 
 Carbonic Acid Gas, COg, is the first product of carbon combus- 
 tion formed in the furnace. It is a colorless gas, with a slightly acid 
 taste and smell, and is incombustible, because it holds in combination 
 all the oxygen it has the power to combine with. Its specific gravity is 
 1.529. This gas in passing through or over a bed of red-hot carbon 
 will take up additional carbon, changing the original product, COg, 
 into a lower oxide, CO, which is a combustible gas. This change, 
 if it occurs in the furnace, is very wasteful of fuel in case the lower 
 oxide escapes unburned ; for example, carbon burned to COg = 14,500 
 heat units per pound, but if burned to CO, = 4452 heat units per 
 pound, equivalent to a loss of two-thirds of the heating power of the 
 carbon. 
 
 Carbonic Oxide Gas, CO, is a product of incomplete combustion ; 
 the gas is colorless, tasteless, and inodorous. Its specific gravity is 
 0.967. It burns in the air with a blue flame, forming carbonic acid gas, 
 COg. At high temperatures, such as obtain in steam-boiler furnaces, it 
 has a very strong tendency to combine with oxygen, but at ordinary 
 temperatures this gas does not combine readily with it. Carbon burnt 
 to carbonic oxide gas = 4452 heat units per pound of carbon, which 
 is approximately 10,000 units less than if burned to carbonic acid gas. 
 To remedy this it has long been recommended that air be admitted 
 over the fire or at the bridge wall in quantity sufficient to burn the CO 
 and convert it to COg. As this combination can only occur at a high 
 temperature, no less than that of red-hot coals, efforts of this kind 
 have not always been successful ; the best practice now is to carry a 
 moderately thick fire with a strong draught and less air over the fire 
 
PRODUCTS OF COMBUSTION 1/ 
 
 than formerly, compelling the air to pass through the fire rather than 
 over it. 
 
 Aqueous Vapor. — Hydrogen unites with oxygen to form water, 
 HgO, in which the combustion is complete and the product incom- 
 bustible. The water formed in the furnace passes off as aqueous vapor, 
 condensing in the atmosphere above the chimney. Any excess of hy- 
 drogen in the furnace over that combining with oxygen as above passes 
 off uncombined. The specific heat of gaseous steam is 0.622. 
 
 Sulphurous Oxide. — Sulphur combines with oxygen to form sul- 
 phurous oxide, SOg, a colorless gas with a suffocating odor. It is a 
 non-supporter of combustion, instantly extinguishing flame when 
 brought within its influence. Sulphurous oxide in absorbing the vapor 
 of water changes to sulphurous acid, SOg + HgO, which may, and often 
 does, become a direct cause of the external corrosion of boilers, mud- 
 drums, feed-pipes, etc. 
 
 Nitrogen is a neutral element in the furnace. It is incombustible 
 and has no affinity for any of the products of combustion. It acts 
 simply as a dilutant of the gases in the furnace. Its specific gravity, 
 0.9736, being less than that of air, it performs the useful office of assist- 
 ing the draught. 
 
 Free Air. — There is always an excess of air passing through the 
 fire above that required for combustion. Chemically, about 12 pounds 
 of air are required for the combustion of i pound of coal ; practically, 
 from 18 to 24 pounds actually pass through the furnace. This excess 
 is, of course, waste product and must be regarded as a dilutant of the 
 furnace gases. 
 
 Smoke, when taken collectively, includes all the gaseous products 
 of combustion escaping from the furnace ; specifically it means the 
 colored gases accompanying combustion discharging into the atmos- 
 phere. In the combustion of anthracite coal and coke very little smoke 
 appears at the chimney top ; but in the combustion of bituminous coal 
 the products often become a veritable nuisance in the neighborhood 
 where the chimney happens to be located. The coloring matter is car- 
 bon in a finely divided state, small particles of soot, so small and of so 
 little weight that they are carried off mechanically out of the furnace, 
 up the chimney, and into the atmosphere. The number of these sooty 
 particles determines the color of the smoke, and may vary in density 
 from light gray to black. An excess of sooty particles indicates gener- 
 ally a low temperature in the furnace. 
 
 Ashes. — Whatever incombustible substances originally in the fuel 
 remain after complete combustion are called ashes, irrespective of com- 
 position. 
 
 An average analysis of ash from a number of anthracite and bitumi- 
 nous coals, the percentage of ash averaging approximately 5 per cent., 
 gave the following : 
 
1 8 BOILERS AND FURNACES 
 
 Bituminous. Anthracite. 
 
 Silica 56.22 49.68 
 
 Alumina 36.17 39-83 
 
 Iron, Oxide 2.74 7.51 
 
 Lime . 2.24 2.17 
 
 Magnesia 92 .72 
 
 Potash and Soda 1.13 
 
 Sulphur 58 .09 
 
 100.00 100.00 
 
 The specific heat of ashes may be assumed to be 0.215 without 
 sensible error in engineering calculations. 
 
 The colors of ashes are designated as red, brown, yellow, or white, 
 as they appear to the observer. Red or reddish-brown ashes indicate the 
 presence of iron in the coal, probably in the form of pyrites. 
 
 Clinker is formed by fusing together the impurities in the coal, such 
 as iron, silica, lime, potash, etc. Each of these substances being differ- 
 ently fusible and affecting differently the fusion of each other, their final 
 form will depend somewhat on the intensity of combustion, or, in other 
 words, the temperature of the fire in which they are formed. There are 
 few colored ashes that will not soften under the action of intense heat 
 and form clinker ; and this fact in itself should affect the commercial 
 value of coals. Those which burn to a nearly pure white are the best, 
 because they contain little or no alkali, lime, or oxide of iron. 
 
 Heat Developed by Combustion. — Knowing the elementary 
 constituents of coal, the heat developed may be calculated, as shown in 
 the following example of the analysis of one pound of semi-anthracite 
 coal, containing: Carbon, .83; hydrogen, .05; oxygen, .04; sulphur, 
 .02; ashes, .06 = i.oo. 
 
 Carbon, .83 X 14,500 heat units = 12,035 heat units. 
 
 Hydrogen and oxygen unite to form water ; therefore all the oxygen 
 must be deducted, together with one-eighth of the hydrogen, thus : 
 i of ToTT ^= W5" = -005 pound of hydrogen neutralized by the presence 
 of oxygen in the coal, leaving .05 — .005 = .045 pound of available 
 hydrogen ; then, proceeding as before : 
 
 Hydrogen, .045 X 62,032 = 2,791 heat units. 
 Sulphur, .02 X 4,000 = 80 heat units. 
 
 We have then : 
 
 Carbon, = 12,035 heat units. 
 Hydrogen, = 2,791 heat units. 
 Sulphur, = 80 heat units. 
 
 The theoretic calorific value of the coal = 14,906 heat units. 
 The ash, being inert, is not taken into account. 
 
HEAT DEVELOPED BY COMBUSTION 1 9 
 
 This calorific value is had on the supposition that the carbon has 
 been burnt to carbonic acid gas, COg ; if, however, the carbon has been 
 incompletely burnt, the product being carbonic oxide gas, CO, instead, 
 a less number of heat units would be had, thus : 
 
 Carbon burnt to CO, .83 X 4452 = 3695.15 heat units. 
 Hydrogen as above, = 2791.00 heat units. 
 
 Sulphur as above, = 80.00 heat units. 
 
 Calorific value of the coal burnt to CO = 6566. 15 heat units. 
 
 This result is most likely to occur, at least in part, when carrying 
 thick fires with an insufficient air supply. 
 
 The available hydrogen in coal after deducting the combining por- 
 tion of oxygen also present may conveniently be expressed thus : 
 
 Hydrogen, 62,032 heat units 
 
 =^ 4-28. 
 
 Carbon, 14,500 heat units 
 
 Hydrogen may be taken, then, as 4.28 times as valuable as carbon in 
 thermal calculations. Practically three times the value of carbon is as 
 high as should be taken, and even this is greatly in excess of what is 
 realized. 
 
 E. T. Cox, formerly State Geologist, Indiana, informed the writer 
 that the net result of his investigations gave 20,115 heat units as the 
 average thermal value of one pound of the volatile matter liberated 
 from Indiana bituminous coals by heat during the process of combus- 
 tion, or a little less than defiant gas (21,300 heat units). Mr. Cox's 
 figures are used in the following example. 
 
 A sample of bituminous coal by proximate analysis can be calcu- 
 lated with tolerable accuracy, thus : 
 
 Per cent. Per cent. 
 
 Volatile matter 41 | ^^^^^ 3 
 
 •- Gas 38 
 
 Coke 59 I ^^^^^ ^^'''^^" 50 
 
 *- Ash 9 
 
 100 100 
 
 The volatile matter requires the expenditure of heat for its libera- 
 tion, which in a large number of experiments approximated 3600 heat 
 units per pound of volatile matter. 
 
 The volatile matter is not all combustible gas ; more or less aqueous 
 vapor passes off with it. Experimentally, the calorific value of the 
 volatile matter of bituminous coals was found to approximate 20,115 
 heat units per pound. 
 
20 BOILERS AND FURNACES 
 
 The theoretical calorific value of one pound of such coal may be 
 determined thus : 
 
 Gas .38 X 20,115 = 7,643.70 heat units. 
 
 Less 38 X 3,600 = 1,368.00 heat units. 
 
 Net value of gas . . . 6,275.70 heat units. 
 
 Fixed carbon 50 X 14,500 =r 7,250.00 heat units. 
 
 Total calorific value . 13,525.70 heat units. 
 
 Dynamical Value of Combustion. — A British thermal unit is 
 that quantity of heat necessary to raise the temperature of one pound 
 of pure water from 39° to 40° Fahr., the former being the temperature 
 of its greatest density. 
 
 The mechanical equivalent of heat is equal to raising 772 pounds 
 one foot high against the action of gravity ; 33,000 pounds raised one 
 foot high per minute = i horse-power. 
 
 The combustion of one pound of carbon yields 14,500 heat units ; 
 then 14,500 X 772 = 11,194,000 foot-pounds. 
 
 Hydrogen yields 62,032 heat units for each pound burnt to water. 
 The dynamic value of hydrogen is : 62,032 X 772 = 47,882,704 foot- 
 pounds. 
 
 Coal fed to steam-boiler furnaces is usually reckoned in pounds per 
 hour. If it be required to know the dynamic value of one pound of 
 coal per hour expressed in horse-power, the coal containing say 13,680 
 thermal units, we have 
 
 13,680 X 772 , 
 
 33,000 X 60 =5-303 horse-power. 
 
 Of this amount, however, only about 10 per cent, is available for doing 
 useful work. 
 
 Temperature of Fire. — Carbon and hydrogen are the principal 
 heat-giving constituents of coal, and of these carbon is the most effec- 
 tive in steam-boiler practice, because an incandescent bed of it can be 
 maintained at all times, the direct effect of which is to prevent violent 
 fluctuations of temperature in the furnace. To illustrate the method of 
 calculating the temperature of fire, we will take the analysis of coal 
 given on page 18, the carbon of which was 83 per cent. Carbon re- 
 quires for its complete combustion 2.67 times its own weight of oxygen ; 
 then I pound carbon + 2.67 pounds oxygen = 3.67 pounds carbonic 
 acid gas. In order to get this oxygen from the air there would remain 
 in the furnace 8.94 pounds of nitrogen, which must be taken into account 
 thus : 
 
 Products. Pounds. Specific heat. Heat units. 
 
 Carbonic acid gas .... 3.67 X .2164 = .794 
 
 Nitrogen 8.94 X .244 = 2. 181 
 
 12.61 Total . . . 2.975 
 
RATE OF COMBUSTION 21 
 
 heat units absorbed in raising the temperature of one pound of carbon 
 
 1° Fahr. The combined weight of the two gases = 12.61 pounds ; 
 
 then, 
 
 Heat units, 2.975 , .^ , 
 
 = . 236, the mean specific heat. 
 
 Pounds, 12.61 
 
 Carbon yields in its perfect combustion 14,500 heat units per pound ; 
 this divided by the 2.975 heat units absorbed as above gives 14,500 ^- 
 2.975 =: 4874° Fahr. as the highest temperature attainable by the com- 
 bustion of one pound of carbon, and with the exact amount of air 
 (11. 61 pounds) needed to furnish the necessary oxygen, a much smaller 
 allowance than is possible in the actual generation of steam by ordinary 
 furnaces. 
 
 Eighteen pounds of air is, on an average, as little as passes through 
 the furnace for each pound of carbon burnt. A reduction in tempera- 
 ture follows as here shown : 
 
 Products. Pounds. Specific heat. Heat units. 
 
 Carbonic acid gas 3.67 X .2164 = .794 
 
 Nitrogen 8.94 X .244 = 2. 181 
 
 Air in excess, uncombined . 6.39 X .2377 = 1-519 
 
 Totals 19.00 4-494 
 
 heat units absorbed in the furnace, being 1519 more per pound of 
 carbon than in the previous example. We have, then, 14,500 -i- 4.494 
 = 3226° Fahr., a reduction of 1648° Fahr. from that obtained in the 
 preceding example ; but this accords more nearly with the best practice 
 and is as high a temperature as can ordinarily be expected. 
 
 Rate of Combustion. — This is commonly expressed in pounds of 
 coal burnt per square foot of grate surface per hour. The weight of coal 
 burnt will depend, other things being equal, upon the quantity of air 
 passing through the fire ; the rate of combustion varies between wide 
 limits : horizontal tubular-boiler furnaces burn from 8 to 12 pounds of 
 anthracite coal per square foot of grate surface per hour with natural 
 draught ; bituminous coals range from 1 2 to 20 pounds, and occasionally 
 more. Internally fired boilers by reason of their relatively smaller pro- 
 portion of grate to heating surface have a rate of combustion varying 
 from 1 2 to 40 pounds of coal per square foot of grate per hour, depend- 
 ing upon the design of the boiler and the ratio of the grate to heating 
 surfaces, the smaller ratio of grate requiring a higher rate of combus- 
 tion. Boilers of this type are usually supplied with a higher chimney, 
 requiring a stronger draft than boilers of the former type. The rate of 
 combustion will vary with the quality of the fuel, draft, ability, and watch- 
 ful care of the fireman. 
 
 Efficiency. — The efficiency of a steam-boiler is a percentage indi- 
 cating how nearly the actual performance attains to the theoretical 
 possibilities ; if the latter be expressed by 100, the efficiency will always 
 
22 BOILERS AND FURNACES 
 
 be a less number. For example : a coal used for generating steam by- 
 calorimeter test yields 12,500 heat units per pound ; the equivalent evapo- 
 ration from and at 212° Fahr. would be 12,500 ^- 966 = 12.95 pounds 
 of water per pound of coal, the theoretical possibility ; but by actual test 
 only 8.75 pounds of water were evaporated, then : 
 
 8.75 X 100 
 
 Efficiency = = 68.62 per cent. 
 
 12.95 
 
 The loss of heat in this case is 31.38 per cent, of the total, accounted 
 for by heat escaping by the chimney, by radiation, the contact of the hot 
 surfaces with the air, as well as imperfect combustion. 
 
 Cylinder and flue boilers, externally fired, set in brick-work, have an 
 efficiency ranging from 45 to 60 per cent. ; tubular boilers from 50 to 70 
 per cent,; internally fired boilers from 60 to 70 per cent. ; water-tube 
 boilers from 65 to 75 per cent. Good boilers properly set and well 
 managed will average nearly the same efficiency, approximating 65 per 
 cent. 
 
 Calorific Value of Coal. — Table I. gives proximate analyses and 
 calorific values of selected American coals. Only a few of the numerous 
 coals of the United States can be mentioned, but enough are given to 
 show the average analysis for localities named. In quality, the range of 
 fuels for steaming purposes covers everything from the softest lignites, 
 which sometimes contain as much as 15 per cent, of water, to Lehigh 
 anthracite, which is nearly pure carbon. Bituminous coals are most 
 abundant, and while these vary much in calorific value, they are, in 
 general, good steaming coals. 
 
 The sulphur contained in some of the softer coals is a mischievous 
 element one would gladly be rid of, inasmuch as from i^ to 4 per 
 cent, is not uncommon, and in some localities as much as 10 per cent, 
 is recorded, notably specimens analyzed from mines in Summit County, 
 Utah ; but this high percentage is quite unusual. No account is made 
 of the contained sulphur in the coal in any calculations connected with 
 this table. 
 
 The calorific values in this table are based upon the experimental 
 determination that each pound of carbon will yield 14,500 heat units 
 when burned in oxygen to carbonic acid gas. The volatile portions of 
 the coal are calculated upon the experimentally ascertained fact, by E. T, 
 Cox, that the total average calorific value of the gases obtained by the 
 destructive distillation of bituminous coal is approximately 20,000 heat 
 units per pound, and that 3600 heat units are absorbed in the process of 
 disassociation of the gases from the coal ; this leaves 16,400 heat units 
 per pound as the net calorific value for the volatile portion of bituminous 
 coals. The calorific value given in the next to the last column in the 
 table is the sum of the carbon and volatile gases as thus ascertained, 
 expressed in British thermal units. 
 
ANALYSIS OF AMERICAN COALS 
 
 23 
 
 TABLE I. 
 
 PROXIMATE ANALYSES AND CALORIFIC VALUE OF SELECTED AMERICAN COALS. 
 
 Coals and Locality. 
 
 Volatile 
 Matter, per 
 
 CENT. 
 
 Coke, 
 PER cent. 
 
 a 
 
 Water. 
 
 Gas. 
 
 Fixed 
 Carbon. 
 
 Ash. 
 
 1.74 
 3-01 
 1-59 
 
 35-48 
 42.76 
 38-33 
 
 58.96 
 48.30 
 54-64 
 
 3.82 
 5-93 
 5-44 
 
 14,368 
 14,017 
 14,209 
 
 I-I3 
 1.52 
 •93 
 
 13-21 
 
 14-73 
 15-55 
 
 81.28 
 74-49 
 77-54 
 
 4.38 
 9.26 
 5-98 
 
 13,952 
 13,217 
 13,793 
 
 18.08 
 
 39-30 
 
 35-61 
 
 7.01 
 
 11,608 
 
 12.01 
 
 35-19 
 42-43 
 36.40 
 
 46.24 
 47.16 
 53-10 
 
 6.56 
 6.48 
 9.24 
 
 12,476 
 13,797 
 13,670 
 
 1.20 
 
 23-05 
 
 60.50 
 
 15.25 
 
 12,553 
 
 10.38 
 
 1.36 
 9.42 
 
 36-38 
 34-02 
 31.20 
 27.69 
 31-38 
 31.04 
 43-70 
 
 46.10 
 53-12 
 54-80 
 35-41 
 51-74 
 51-96 
 45-37 
 
 7.14 
 4.96 
 5.60 
 35-54 
 7-46 
 7.05 
 5.15 
 
 13,063 
 9,675 
 12,648 
 12,625 
 13,746 
 
 8.50 
 
 3-50 
 7.00 
 2.00 
 5-50 
 
 32-34 
 32-50 
 48.00 
 29.50 
 38.50 
 44.00 
 
 48.78 
 56-50 
 42.00 
 63.00 
 57-50 
 46.00 
 
 5.83 
 2.50 
 6.50 
 .50 
 2.00 
 4.50 
 
 12,377 
 13,523 
 13,962 
 13,973 
 
 iM 
 
 23-31 
 35-42 
 
 66.85 
 51-32 
 
 6^60 
 
 13,517 
 13,248 
 
 5-73 
 
 1:^ 
 
 46-54 
 40.21 
 41-35 
 
 45.60 
 48^25 
 
 2.13 
 8.75 
 3.90 
 
 14,245 
 13,247 
 13,777 
 
 1.94 
 
 36-77 
 
 52-45 
 
 8.84 
 
 13,585 
 
 2.00 
 3-30 
 3-24 
 3-60 
 
 47-85 
 39.00 
 
 47-73 
 50.50 
 49-24 
 58.80 
 
 2.42 
 7.20 
 10.96 
 7.00 
 
 14,768 
 13,719 
 13,136 
 13,544 
 
 1.23 
 •59 
 
 18.52 
 
 73.57 
 74.31 
 
 lit 
 
 13,205 
 13,812 
 
 2.54 
 5.06 
 9-03 
 
 42.62 
 34-24 
 37-48 
 
 41.14 
 47.69 
 46.24 
 
 13.70 
 13.04 
 7.25 
 
 12,955 
 12,530 
 12,852 
 
 3.01 
 
 30.23 
 
 59.71 
 
 7-05 
 
 13,616 
 
 0.21 
 
 27.82 
 
 60.88 
 
 11.09 
 
 13,390 
 
 3.10 
 
 35-00 
 
 51.50 
 
 10.40 
 
 13,208 
 
 1.79 
 
 29-56 
 
 58.30 
 
 10.35 
 
 13,302 
 
 =^ Hefted 
 
 ALABAMA. 
 
 Bibb Co., Bit., Helena Vein 
 
 Jefferson Co., Bit., Birmingham . . . 
 Tuscaloosa Co., Bit 
 
 ARKANSAS. 
 Franklin Co., Bit., Falker Slope . . . 
 
 Johnson Co., Bit., Coal Hill 
 
 Sebastian Co., Bit., Huntington Slope 
 
 CALIFORNIA. 
 Alameda Co., Bit., Livermore .... 
 
 COLORADO. 
 
 Boulder Co., Bit 
 
 Tremont Co., Bit 
 
 Las Animas, Bit 
 
 GEORGIA. 
 Dade Co., Bit 
 
 ILLINOIS. 
 Macoupin Co., Bit., Mount Olive . . 
 McLean Co., Bit., Bloomington . . . 
 
 Mercer Co., Bit 
 
 Peoria Co., Bit., Elmwood 
 
 Stark Co., Bit., Lombardville .... 
 
 Trenton Co., Bit., Clinton 
 
 Vermilion Co., Bit., Danville 
 
 INDIANA. 
 
 Block Coal, Lafayette 
 
 Clay Co., Bit., McClellan&Zeller . . 
 Davies Co., Buckeye Cannel Coal Co. 
 
 Greene Co., Bit 
 
 Owen Co., Bit 
 
 Vermillion Co., Bit 
 
 INDIAN TERRITORY. 
 
 Choctaw Nation, Bit 
 
 Choctaw Nation, Bit., Atoka 
 
 IOWA. 
 Marion Co., Bit., Oscaloosa .... 
 
 Monroe Co., Bit., Albia 
 
 Wapello Co., Bit., Ottumwa 
 
 KANSAS. 
 
 Cherokee Co., Bit 
 
 KENTUCKY. 
 
 Fulton Co., Bit 
 
 Hancock Co., Bit., Hawesville . . . 
 Lawrence Co., Bit., Peach Orchard . 
 Muhlenberg Co., Bit 
 
 MARYLAND. 
 
 Cumberland, Bit 
 
 Garrett Co., Semi-Bit., George's Creek 
 
 MISSOURI. 
 
 Bates Co., Bit 
 
 Caldwell Co., Bit., Hamilton 
 
 Putnam Co., Bit., Mendota 
 
 MONTANA. 
 Cascade Co., Bit., Sandcoulee . . . . 
 NEBRASKA. 
 
 Adams Co., Bit., Hastings 
 
 NEW MEXICO. 
 
 Colfax Co., Bit., Ranton 
 
 NORTH CAROLINA. 
 Guilford Co., Bit., Deep River .... 
 
 14.87 
 14.51 
 14.71 
 
 14.44 
 13-68 
 14.28 
 
 12.92 
 14.28 
 14-15 
 
 12.99 
 
 13.10 
 13-75 
 13-52 
 10.02 
 13.09 
 13.07 
 14-23 
 
 14.00 
 14-45 
 14.46 
 9.28 
 14-37 
 
 13-99 
 13-71 
 
 14-75 
 13-71 
 14.26 
 
 14.06 
 
 15-29 
 14.20 
 13.60 
 14.02 
 
 13.67 
 14.30 
 
 13.41 
 12.97 
 13-30 
 
 14.10 
 13-86 
 13.67 
 13.77 
 
24 
 
 BOILERS AND FURNACES 
 
 TABLE \.— Continued. 
 
 Coals and Locality. 
 
 Volatile 
 Matter, per 
 
 CENT. 
 
 Coke, 
 
 PER CENT. 
 
 II 
 
 
 
 
 
 
 Water. 
 
 Gas. 
 
 Fixed 
 Carbon. 
 
 Ash. 
 
 2.32 
 8.25 
 
 2.47 
 7.09 
 5-91 
 
 39.08 
 35-88 
 42.50 
 37-82 
 31.83 
 36.61 
 35-01 
 
 52.78 
 53-15 
 
 47.27 
 
 SS 
 
 52.00 
 55-70 
 
 5.82 
 2.72 
 5-74 
 4.71 
 1.45 
 4-30 
 3-38 
 
 14,062 
 13,591 
 13,824 
 14,267 
 14,537 
 
 4-55 
 8.00 
 13.04 
 
 40.00 
 37-83 
 46.70 
 
 48.19 
 45.17 
 32.60 
 
 7.26 
 9.00 
 7.66 
 
 13,547 
 
 ^8;P^ 
 
 i.:35 
 1.50 
 3-04 
 
 ^1i 
 
 3-95 
 
 89.06 
 88.94 
 82.66 
 
 6.14 
 io'.35 
 
 13,480 
 13,286 
 12,634 
 
 1. 12 
 
 4-99 
 
 83.98 
 
 9.91 
 
 12,995 
 
 1.27 
 1. 01 
 2.97 
 4.04 
 
 2.30 
 2.99 
 
 88.15 
 87.96 
 
 88.20 
 
 10.65 
 5.56 
 6.77 
 4-77 
 
 "Si 
 
 13,131 
 13,279 
 
 1-34 
 
 6.42 
 
 76.94 
 
 15-30 
 
 12,209 
 
 1.80 
 0.96 
 1-93 
 1.04 
 1.46 
 1.23 
 
 35-34 
 38.20 
 28.71 
 37-23 
 32.00 
 32.66 
 
 54-94 
 56^61 
 
 if 
 6.10 
 5.12 
 12.75 
 7-93 
 
 13,762 
 
 ^3'8o9 
 13,881 
 14,314 
 13,048 
 13,792 
 
 0.73 
 1. 00 
 
 24.11 
 35-00 
 
 69.11 
 58.40 
 
 6.05 
 5-60 
 
 13,975 
 14,208 
 
 2.00 
 
 1.77 
 
 IS 
 
 33-77 
 25.41 
 26.50 
 31-94 
 
 60.64 
 62.00 
 67.08 
 54-81 
 
 10.82 
 3-68 
 10.09 
 
 14,331 
 13,157 
 14,073 
 13,185 
 
 3-67 
 6.67 
 10.40 
 4.60 
 
 35-51 
 40.20 
 35-94 
 34-72 
 
 41.70 
 43-54 
 49-46 
 49.27 
 
 19.12 
 9.59 
 4.20 
 
 1 1. 41 
 
 11,971 
 12,906 
 13,066 
 12,838 
 
 342 
 3-50 
 0.43 
 
 42.81 
 43-66 
 38.90 
 
 47.81 
 56^37 
 
 5-95 
 9.73 
 4.30 
 
 13,953 
 13,411 
 14,554 
 
 1.05 
 1-34 
 
 I. So 
 
 23.62 
 30.98 
 34-33 
 33-90 
 
 72.67 
 56.83 
 59.89 
 59.25 
 
 2.66 
 10.85 
 4.87 
 5.05 
 
 14,411 
 13,321 
 14,314 
 14,151 
 
 1.22 
 0.76 
 0.50 
 
 41.50 
 19-39 
 19-83 
 
 54.58 
 72.99 
 75.63 
 
 else 
 4.04 
 
 14,720 
 13,764 
 14,218 
 
 2.00 
 1. 10 
 
 l:i 
 
 39.10 
 35-10 
 41.18 
 29-54 
 
 54.40 
 54.50 
 42.92 
 59-90 
 
 4.50 
 9.30 
 
 14,300 
 13,659 
 10,977 
 13,531 
 
 11.30 
 6.04 
 4.20 
 
 42.01 
 40.60 
 
 39-69 
 35.57 
 41.50 
 
 7.00 
 16.02 
 13.70 
 
 12,645 
 12,107 
 12,676 
 
 OHIO. 
 
 Columbiana Co., Bit., Salineville 
 
 Hocking Valley, Bit 
 
 Holmes Co., Bit., Walnut Creek 
 
 Jefferson Co., Bit., Brilliant 
 
 Mahoning Co., Bit., Brier Hill 
 
 Perry Co., Bit., New Straitsville 
 
 Trumbull Co., Bit., Liberty 
 
 OREGON. 
 
 Grant Co., Bit., John Day River 
 
 Tillamook Co., Bit., Nehalem 
 
 Benton Co., Bit., Yaquina Bay 
 
 PENNSYLVANIA. 
 Anthracite, Upper and Lower Measures . . . . 
 Anthracite, Carbon Co., Beaver Meadow . . . 
 Anthracite, Carbon Co., Buck Mountain . . . . 
 Anthracite, Lackawanna Co., Scranton, 40' 
 
 Shaft 
 
 Anthracite, Lackawanna Co., Scranton, Mt. 
 
 Pleasant 
 
 Anthracite, Lehigh Co 
 
 Anthracite, Luzerne Co., Drifton 
 
 Anthracite, Luzerne Co., Jeanesville 
 
 Anthracite, Luzerne Co., Wilkes-Barre, Lehigh 
 
 Valley Buckwheat 
 
 Bituminous, Allegheny Co., Pittsburg, Average 
 
 Bituminous, Armstrong Co 
 
 Bituminous, Fayette Co., Connellsville . . . . 
 
 Bituminous, Greene Co., Main Bench 
 
 Bituminous, Indiana Co., Lower Bench . . . . 
 Bituminous, Jefferson Co., Freeport, Average . 
 Bituminous, Westmoreland Co., Loyal Hanna, 
 
 Average 
 
 Bituminous, Westmoreland Co., Youghiogheny 
 
 TENNESSEE. 
 
 Campbell Co., Bit., Newcomb 
 
 Franklin Co., Bit 
 
 Hamilton Co., Bit., Melville 
 
 Marion Co., Bit 
 
 TEXAS. 
 
 Maverick Co., Bit., Eagle Pass. 
 
 Palo Pinto Co., Bit., Strawn 
 
 Rusk Co., Bit., Stevens . 
 
 Tarrant Co., Bit., Fort Worth 
 
 UTAH. 
 
 Emery Co., Bit., Castle Dale 
 
 Iron Co., Bit., Cedar City 
 
 Summit Co., Bit., Coalville 
 
 VIRGINIA. 
 
 Halifax Co., Bit., Elmo 
 
 Rockingham Co., Bit., Clover Hill 
 
 Scott Co., Bit., Clinch Valley . . . 
 
 Wise Co., Bit., Big Stone Gap 
 
 WEST VIRGINIA. 
 
 Logan Co., Bit., Dingess 
 
 Mineral Co., Bit., Elk Garden 
 
 Pocahontas Co., Semi-Bit 
 
 WASHINGTON. 
 
 Kittitas Co., Bit., Ellensburgh 
 
 Pierce Co., Bit., Wilkeson 
 
 Stevens Co., Bit., Calispell 
 
 Whatcom Co., Bit., Bellingham Bay 
 
 WYOMING. 
 
 Carbon Co., Bit., Dana 
 
 Sheridan Co., Bit., Sheridan 
 
 Weston Co., Bit., Cambria 
 
CHAPTER II. 
 
 MATERIALS OF CONSTRUCTION. 
 
 PART I.— CAST IRON. 
 
 The principal materials entering into the construction of steam 
 boilers are limited by commercial and practical considerations to cast 
 iron, wrought iron, and steel. Copper was formerly used in boiler con- 
 struction, especially for fire-boxes in locomotive boilers and the internal 
 heating surfaces in marine boilers. Its use is practically abandoned at 
 this time because of its want of hardness and tensile strength as com- 
 pared with either wrought iron or steel, while its cost is much greater. 
 
 Cast Iron. — This product is had by a remelting together of two or 
 more kinds of pig-iron in order to secure castings having certain quali- 
 ties determined approximately in advance. This presupposes a know- 
 ledge of the constituents of the pig-iron to be used, which is, taken 
 altogether, a very complex material, because pig-iron is always com- 
 bined with extraneous substances, which, taken collectively, are called 
 impurities ; and of these the principal ones are carbon, silicon, sulphur, 
 phosphorus, and manganese. These may be combined with the iron 
 both chemically and mechanically, and this is especially true of carbon. 
 There is no good reason for regarding carbon and silicon as impurities 
 as we ordinarily use that word, for the presence of both are beneficial, 
 if not altogether necessary, in the manufacture of iron castings. The 
 amounts of sulphur, phosphorus, and manganese ordinarily present are 
 small in amount and do no particular harm. 
 
 Carbon in Cast Iron. — Carbon, by reason of its chemical and 
 physical effects, is the most important element in cast iron. It is always 
 present in pig-iron, and repeated meltings does not sensibly diminish its 
 quantity, which varies from at least 1.5 per cent, up to 4.5 per cent. 
 As carbon may be present in a combined form or may be present as 
 graphite, it is best when referring to carbon in iron to have it under- 
 stood that such reference means total carbon, the latter determining 
 also the melting-point of the iron. 
 
 Combined Carbon. — Cast iron in a fluid state will take up at least 
 3.5 per cent, of carbon and hold it in solution, part of which is expelled 
 during, the process of cooling. For small and medium castings there 
 should be very little combined carbon, because it makes the iron hard 
 and brittle and influences adversely the amount of shrinkage. 
 
 Graphitic Carbon. — The total carbon in cast iron must be either 
 combined or in the graphitic form : the latter is characteristic of gray 
 
 3 25 
 
26 BOILERS AND FURNACES 
 
 irons. Carbon in fluid iron is combined with the iron and has no ten- 
 dency to separate from it ; but during the process of cooHng marked 
 changes occur, depending somewhat on whether the cooling be rapid or 
 slow ; if the latter, the combined carbon will separate from the iron, but 
 remain in it mechanically in the form of graphite between the crystals. 
 This precipitation of carbon renders iron soft, changing the color of 
 white iron into a darker hue, — hence the name of gray iron. 
 
 Silicon in Cast Iron. — Next after carbon the substance most com- 
 monly met with in pig-iron is silicon, the quantity ranging from o. 5 to 
 more than 4 per cent. The presence of silicon in cast iron is important, 
 because of its effect upon the contained carbon, especially in the con- 
 version of the combined carbon into graphite, by which white iron is 
 changed to gray, the hardness and brittleness of the former being thus 
 greatly modified, showing marked increase in both the transverse and 
 tensile strength of iron castings. Silicon tends also towards the elimi- 
 nation of blow-holes, and thus contributes materially towards the pro- 
 duction of sounder castings than can be had in its absence. 
 
 Sulphur in Cast Iron. — Sulphur is almost always present in 
 pig-iron. During the process of remelting an additional quantity is 
 absorbed by the fluid metal from the coke fuel in the cupola. The 
 unexplainable peculiarities of cast iron have long been attributed to sul- 
 phur ; but recent investigations, combining physical and chemical tests, 
 made for the purpose of ascertaining what the precise action of sulphur 
 is upon cast iron, show that sulphur in the quantities usually found in 
 gray iron does not injuriously affect cast iron ; and, further, that if an 
 excess of sulphur should be found to exert a pernicious influence upon 
 the iron, a slight increase in the quantity of silicon pig used would 
 counteract any such effect. 
 
 Castings composed of Southern gray irons containing 0.088 to o.ioo 
 per cent, of sulphur show them to be of good quality, with no chill, no 
 blow-holes, very low shrinkage, and high strength ; but this latter 
 quality is not to be attributed to any action of the sulphur present, but 
 simply shows that it does not within the above permissible limits of per- 
 centage detract from the strength of cast iron. 
 
 Phosphorus in Cast Iron. — Phosphorus enters into chemical union 
 with iron, the effect of which is to render iron close and compact, with 
 a tendency to become cold-short at low temperatures. Phosphorus is 
 not known to influence the change of carbon in pig-iron one way or the 
 other. In producing hardness in casting, such change is to be ascribed 
 to the influence of phosphorus alone. The permissible allowance of 
 phosphorus in cast iron is confined to narrow limits ; its presence may 
 be beneficial in some mixtures of pig-iron, but from 0.5 to i per cent, is 
 the beneficial limit. In moderate quantities phosphorus lessens the 
 tendency to form blow-holes in castings ; it also lessens shrinkage and 
 prolongs the period of fluidity. 
 
CAST IRON 27 
 
 Manganese. — The physical properties of cast iron are not greatly- 
 altered by the addition of manganese if the latter does not much exceed 
 I per cent., but as much as 1.5 per cent, makes it very hard. The 
 presence of say i per cent, of manganese is beneficial in foundry prac- 
 tice, increasing the fluidity of the iron when melted ; but when the pro- 
 portion of manganese is much greater than that, it renders cast iron less 
 plastic, more hard and brittle when cold, and increases the shrinkage. 
 The effect of manganese when used alone is, as stated above, to harden 
 cast iron, but it does not turn gray iron white, nor does it increase the 
 combined carbon, nor does it increase the tendency to chill ; its harden- 
 ing effect is to be ascribed to its one influence, which is to harden iron. 
 
 Ferro-Manganese, when added in small quantities to molten metal 
 in a foundry ladle, softens and improves the iron. The probable ex- 
 planation is, that the manganese counteracts the effect of sulphur and 
 silicon, tending to eliminate the former and neutralize the latter, and so, 
 when common iron with a tendency to hardness is employed, it some- 
 times happens that ferro-manganese may be used as a softener. The 
 hardness, however, generally returns when the iron is remelted, because 
 the manganese is oxidized and more sulphur absorbed. The good 
 effects of manganese appear to be twofold : by its action it leads directly 
 to a measure of hardness and closeness of grain which is beneficial, while 
 indirectly it is useful in preventing the absorption of sulphur during 
 remelting. 
 
 Slo^v Cooling. — Gray iron machinery castings should be slowly 
 cooled, and particularly when they are small and thin. Rapid cooling 
 tends to brittleness by preventing the separation of the combined carbon 
 into graphite ; slow cooling tends to make the grain coarser, and such 
 castings can have any necessary machine-work done much more rapidly 
 and better than if the metal was hard. 
 
 Cooling Strains. — Fractures in cooling are likely to occur where 
 two portions of a casting join each other at right angles and with square 
 corners. Thick castings joining each other at right angles or nearly so 
 are almost certain to have cavities occur at their points of intersection by 
 reason of an irregular grouping of crystals ; the direction of the shrinkage 
 not being parallel will also produce distortion. The importance of slow 
 cooling after pouring a casting of considerable size, and especially when 
 of varying thickness, is known to every foundryman : gray irons, having 
 a natural tendency to hardness, are made harder by rapid cooling ; on 
 the other hand, slow cooling tends to soften such castings. 
 
 Blow-Holes, — These are serious defects in a casting, because they 
 are generally below the surface ; there is seldom any outward indication 
 as to their location or to what extent they exist. A blow-hole not 
 only lessens the area of cross-section of the casting in which it occurs, 
 but its presence in the casting may be additionally harmful in setting up 
 internal strains within the casting which might not otherwise occur. 
 
28 BOILERS AND FURNACES 
 
 The surface of a casting should be smooth, free from bits of slag, 
 scabs, and unusual roughness ; small, sharp indentations in the surface 
 of a casting, however caused, may, as far as they go, be considered in- 
 cipient fractures, and if undue stress be applied to such a casting, a frac- 
 ture may begin at any such point at a stress much lower than would 
 otherwise break the casting. 
 
 Shrinkage of Cast Iron. — Shrinkage represents the difference in 
 size between a mould and the casting made in it. When molten iron is 
 poured into a mould it expands at the moment of solidification, and if 
 properly vented the metal will take a sharp impression of the mould. 
 The cooling of a casting begins at and along its outer surface ; so, also, 
 crystallization begins at the surface and proceeds towards the centre. It 
 sometimes happens in the case of thick castings that the interior portion 
 may be in a semi-molten state, while the whole exterior surface has been 
 solidified to a considerable depth. As cooling progresses and the 
 casting parts with its heat, it diminishes in bulk ; but this contraction is 
 not uniform throughout ; thin portions cool first and take their perma- 
 nent form, the thicker portions of the casting cooling later. It is to be 
 expected, and it often actually occurs, that the cooling of the thick por- 
 tions will show lines of incipient, if not actual, fracture along the lines 
 of intersection where a thicker portion joins a thinner one. To prevent 
 this, it is customary to place quarter-round concave fillets at all such 
 intersections. Shrinkage is at best an uncertain thing to deal with, 
 depending not only on the size and shape of the pattern, but upon the 
 temperature at which the iron is poured, the quality of the iron, and 
 especially upon the quality of hardness ; some observers state that the 
 amount of shrinkage corresponds closely to that of the total quantity 
 of carbon present. 
 
 The ordinary shrinkage of gray iron castings is -|- inch per foot ; this 
 is the graduation on standard shrinkage rules. Some irons shrink in 
 the proportion of -^ inch per foot, others still less ; the standard rule 
 is, however, sufficiently accurate for all ordinary purposes. 
 
 Strength of Castings. — Combined carbon was long thought to be 
 the medium by which strength was imparted to castings, and that its 
 conversion to graphite was an occasion of weakness ; but recent chemical 
 and physical tests show that the reverse is true, and that, if anything, 
 combined carbon weakens castings and never strengthens them. This 
 is especially true of small castings, where it has been observed that as 
 the percentage of combined carbon was decreased by the addition of 
 silicon, that the hardness and brittleness of the casting was also de- 
 creased, and that the strength of the casting had been increased at the 
 same time. 
 
 The strength of cast iron increases when uninfluenced by any other 
 element than carbon by an increase of silicon up to as much as 3.5 per 
 cent, of the total weight, and this is as high as will be found in any ordi- 
 
CAST IRON 29 
 
 nary silicon pig ; with this increase of strength there is also an absence 
 of brittleness and an increase in the size of the grain. This latter must 
 not be carried too far, however, or weakness may result from this cause 
 alone. 
 
 Crushing Strength. — That of cast iron varies with the chemical 
 and physical properties of any given sample, the range being as great 
 as from 50,000 to 75,000 pounds per square inch, averaging so high as 
 to require no special calculation, — that is to say, when patterns are 
 properly dimensioned for tensile strains, any probable amount of com- 
 pression can be safely carried by cast iron, as the carrying capacity of 
 the latter is four or five times as great as the former. From experi- 
 mental data it appears that the maximum crushing strength of cast iron 
 would be obtained with about 0.75 per cent, of silicon and 2 per cent, 
 of combined carbon. In ordinary calculations, 60,000 pounds per 
 square inch may be assumed as the ultimate crushing strength of 
 medium hard gray iron machinery castings in short lengths. 
 
 Transverse Strength of Cast Iron. — The usual method of test- 
 ing cast iron is to break test-bars one inch square by one foot in length ; 
 or, if other sizes are used, the results are reduced for comparison to the 
 one inch square section. Transverse tests when applied to the centre of 
 a test-bar supported at each end combine both a crushing and a ten- 
 sile test, the transverse test being an intermediate of the two. 
 
 The breaking strength of gray iron castings of good quality, the 
 bars one inch square by twelve inches long, loaded at the centre, will 
 vary from 1600 to 3600 pounds. 
 
 Tensile Strength of Cast Iron. — Tests for tensile strength are 
 not nearly so common as are the transverse tests, and for the reason 
 that cast iron is seldom employed where its tensile strength alone is 
 brought into use. Tensile tests do not as a rule exhibit physical quali- 
 ties which transverse tests do not bring out equally well, and such tests 
 vary, of course, with the physical and chemical properties of the iron, 
 and may, for good gray iron castings, vary anywhere from 14,000 to 
 22,000 pounds per square inch of section when the test-bars approxi- 
 mate one square inch of fractured area. As the quality of the casting 
 can only be approximately arrived at by an inspection of the grain at 
 the points where the gates are knocked off, it seems unwise to accord 
 to cast iron as high an average tensile strength as would be assigned 
 wrought iron or mild steel covering the same percentage of variations 
 in tensile strength. The writer does not, therefore, recommend more 
 than 16,000 pounds per square inch of section for the best quality gray 
 iron castings when not less than Y^ inch nor more than i^ inches thick. 
 
 Elastic Limit of Cast Iron. — This quality in cast iron is seldom 
 taken into account for ordinary machine-work, consequently but few 
 tests have been made concerning it. The comparatively few tests do 
 show, however, that the elastic limit of cast iron is about Vi of its ten- 
 
30 BOILERS AND FURNACES 
 
 sile strength. If we regard cast iron as good for 16,000 pounds per 
 square inch tensile strength, the elastic limit would then be : 16,000 -*- 
 3 = 5333 pounds per square inch. 
 
 Cast Iron as a Material for Steam Boilers. — The advantages 
 of cast iron as a material for steam boilers in preference to steel or 
 wrought iron are thus set forth by the principal manufacturer of such 
 boilers in this country. 
 
 1. The spherical or globular form of parts can be manufactured at 
 reasonable prices, which would be impossible with any wrought metal. 
 
 2. Any number of such parts can be made uniform and duplicate ; 
 likewise impossible with wrought metal, unless at enormous expense. 
 
 3. Cast metal resists much better the corrosive action of acids in 
 the water or in the products of combustion. 
 
 4. Cast metal is much less affected by oxidation. 
 
 5. Cast metal cannot blister, furrow, or pocket. 
 
 6. Cast metal also transmits heat more readily than wrought iron, 
 and while it may have to be a trifle thicker than ordinary boiler tubes, 
 it is in the case of this boiler (3^ to f inch) much thinner than ordinary 
 crown sheets.* 
 
 Objections to Cast Iron. — This material has been objected to as 
 a material for steam boilers for the following reasons : 
 
 1. Because it is a crude product to begin with, it is brittle and of low 
 tensile strength, and there is no certainty that castings can be made 
 uniform in strength or in other qualities. 
 
 2. The cooling strains in castings often produce flaws or other de- 
 fects which are hidden to the eye and thus escape detection in the 
 workshop ; further, that such defects do not become apparent even 
 when under test by hydraulic pressure, but which may when under the 
 influence of heat and unequal expansion and without a moment's warn- 
 ing end in sudden and disastrous failure. 
 
 3. Its unyielding nature is thought to especially unfit it for parts of 
 a boiler subject to unequal expansion arising from differences of tem- 
 perature. 
 
 4. Cast iron when cold seldom or never exhibits indications of weak- 
 ness or fracture in advance of actual breakage. It has been further 
 objected to because it loses coherence and will crumble under moderate 
 loads, and not infrequently by its own weight, when subjected to high 
 temperatures or those approximating red heat. 
 
 Cast Iron in the Fire. — The effects of ordinary boiler-furnace 
 heat on unprotected cast iron is to change the characteristic granular 
 structure common to all gray iron castings of good quality to coarse, 
 uneven grains having scarcely any metallic lustre and little or no granular 
 
 * This was long believed to be true, but Isherwood's experiments show the 
 relative conductivity of cast iron to be less than that of wrought iron. — Ed. 
 
WROUGHT IRON 3 1 
 
 coherence. Burnt cast iron shows a change in the color of fracture from 
 gray to brown and gray mixed, the metal is usually twisted out of shape, 
 and is so extremely brittle and lifeless that it is utterly unfit for further 
 use in the foundry in the production of castings requiring strength. The 
 continued heating and reheating of any metal would in time destroy it, 
 but cast iron, by reason of its relatively coarser grain, seems to be less 
 able to withstand the effects of reheating and cooling than either wrought 
 iron or mild steel ; but no metal of which iron is the basis should be 
 subjected to the continued action of intense heat, — that is to say, tem- 
 peratures approximating red heat. Cast iron yields to the fire sooner 
 than wrought iron ; it loses strength at temperatures much below those 
 which similarly affect wrought iron. When red hot, cast iron will 
 scarcely sustain its own weight, and when so heated is liable to crumble 
 to pieces if under compression. 
 
 Use of Cast Iron. — The fact in regard to the use of cast iron as 
 a material for steam boilers is that for steam- and hot-water-heating pur- 
 poses it is more largely employed than any other material. It is not 
 more injuriously affected by heat than are other materials when the parts 
 are properly proportioned and no thick flanges are exposed to the heat 
 of the furnace. Boilers for heating purposes work under moderate 
 pressures, seldom more than 20 pounds per square inch for water and 
 still less for steam. The Harrison boiler has been continuously and suc- 
 cessfully in the market for more than thirty years, and at this time is in 
 good repute as a safety boiler, as a steam-maker, and has through all 
 these years suffered no deterioration which is not common to all boilers. 
 This is the only high-pressure cast-iron boiler known to the writer which 
 has been continuously on the market for ten years. 
 
 PART II.— WROUGHT IRON. 
 
 Wrought Iron. — This product is commercially pure iron prepared 
 from selected pig-iron by a succession of processes, such as puddling, 
 squeezing, hammering, rolling, etc., to rid the iron of its original im- 
 purities. The operation of puddling consists in stirring a mass of spongy 
 iron in the midst of a bath of cinder, which prevents the intimate ap- 
 proximation of its particles. This cinder, adhering to the iron, opposes 
 a thorough welding of the mass and favors the production of fibrous 
 texture, since during subsequent working the molecules of the iron slide 
 over each other, giving the iron a characteristic fibrous appearance in 
 fracture. If a bar of wrought iron is nicked on one side and then bent 
 over double upon itself, it will expose the longitudinal grain of the 
 metal ; and if the iron is of good quality, the fracture thus obtained will 
 be a dense mass of fibre, slightly interspersed with fine grains of iron. 
 
 Physical Properties of Wrought Iron. — Boiler-plates should 
 possess the properties of tenacity, ductility, and welding. Each of these 
 is influenced in some measure by the impurities of the iron. That qual- 
 
32 
 
 BOILERS AND FURNACES 
 
 ity of boiler-plate is judged to be the best which has the greatest tensile 
 strength combined with ductility and freedom from brittleness ; such an 
 iron will be at once strong, tough, and fibrous. When the above prop- 
 erties are well combined, wrought iron will resist strains due to unequal 
 expansion remarkably well. 
 
 Tensile Strength. — Boiler-plates may possess high tensile strength 
 at the expense of other qualities, such as homogeneousness and tough- 
 ness. Wrought-iron plates possessing all the necessary qualities suit- 
 able for steam boilers will not have a tensile strength much, if anything, 
 above 55,000 pounds per square inch ; * a higher tensile strength is 
 liable to have associated with it the undesirable qualities of hardness and 
 brittleness, either of which will more than offset any gain in tensile 
 strength above the figures given. Wrought iron having a tensile 
 strength of less than 45,000 pounds per square inch should not be em- 
 ployed in boiler construction. 
 
 Tensile tests made in the direction of the grain of fibrous iron show 
 greater strength than those made across the grain. The following data 
 was obtained from United States government tests of wrought-iron 
 plates, which it will be observed are of very high quality. These were 
 short specimens : 
 
 Thickness. 
 
 J inch with the grain . 
 I inch across the grain 
 x®6 ii^ch with the grain . 
 x% inch across the grain 
 I inch with the grain . 
 I inch across the grain 
 
 Tensile Strength. 
 
 • 58,373 pounds. 
 
 • 53,333 pounds. 
 . 62,195 pounds. 
 . 60,202 pounds. 
 . 56,270 pounds. 
 . 56,461 pounds. 
 
 Reduction of Area. 
 
 38 per cent. 
 9 per cent. 
 43 per cent. 
 10 per cent. 
 25 per cent. 
 17 per cent. 
 
 Test- Pieces. — Wrought-iron test-pieces have usually been short 
 specimens, the tests being made with reference to tensile strength only. 
 Fig. I represents the size and form prescribed by the United States 
 
 Fig. I. 
 
 TCX. 
 
 Tesi piece io ide^a/nz «-■/"-•»] 
 
 Board of Supervising Inspectors of Steam-Vessels, — viz. , 10 inches long, 
 2 inches wide, cut out in the centre as indicated. 
 
 All sample pieces of iron plate -^ inch thick and under shall be one 
 inch wide at reduced section. Plate over -^ inch thick shall be reduced 
 
 * It will be understood that wrought iron made by the usual processes is meant, 
 and not homogeneous iron or ingot iron, which have physical properties differing 
 from ordinary wrought iron. 
 
WROUGHT IRON 33 
 
 in width at centre to an aggregate area approximating 0.4 of one 
 square inch ; but such reduced area shall in no case exceed 0.45 nor be 
 less than o, 35 of an inch ; and the force at which the piece can be parted 
 in the direction of the fibre or grain (when of iron), represented in 
 pounds avoirdupois in proportion to the ratio of its area, shall be deemed 
 the tensile strength per square inch of the plate from which the sample 
 was taken. When more than one sample shall be tested from one 
 sheet, the sample showing the lowest tensile strength shall be allowed as 
 the tensile strength of the plate. 
 
 Ductility is the property which enables a material to be drawn out 
 without breaking. It is also called elongation or extension in reports 
 on the mechanical tests to which plates or bars are subjected. Elonga- 
 tion occurs when a ductile material is subjected to a tensile stress higher 
 than its elastic limit, after which a permanent change of form takes 
 place. It may be measured in a tensile testing-machine in two ways, — 
 by the actual amount of elongation in inches and parts of an inch, and 
 by reducing the amount so found to percentage extension of its original 
 length. 
 
 Wrought-iron boiler-plates under 45,000 pounds tensile strength 
 should show a reduction of area of not less than 12 per cent. ; 45,000 to 
 50,000 pounds, 15 per cent. ; 50,000 to 55,000, 25 per cent. ; 55,000 
 pounds and over should show 35 per cent, reduction of area. 
 
 Elastic Limit. — This limit may be determined in the same manner 
 and at the same time as when making a tensile test by simply applying 
 progressive loads and then removing them. After the removal of each 
 increment of load the specimen is measured to determine whether or not 
 it has returned to its original length. The weight required to give the 
 specimen its permanent set divided by the area of the specimen will ap- 
 proximate its elastic limit. From a large range of experiments made 
 upon plates and bars it appears that the elastic limit of wrought iron is 
 approximately one-half its tensile strength. 
 
 Welding. — Wrought iron possesses the property of welding when 
 the two parts to be joined are brought up to a white heat. Welded 
 joints are, when well made, scarcely inferior to the original bar ; but 
 stays, braces, etc., for boilers should be made from whole stock if pos- 
 sible, because there is always more or less uncertainty about welded 
 joints, particularly when the parts to be joined are of considerable 
 diameter or thickness. 
 
 Fibre. — The best wrought iron has a characteristic fibre resulting 
 from its method of manufacture, the fineness and uniformity of which is 
 taken as a practical indication of the quality of the iron. 
 
 A fibrous fracture in wrought iron is always taken to indicate a high 
 grade of iron, and this is sometimes associated with the act of pulling in 
 a tensile testing-machine. But stresses of this kind do not cause an 
 extension of the grain to produce fibre, — this is already present in the 
 
34 BOILERS AND FURNACES 
 
 iron ; the act of pulling simply draws these fibres out in clusters ; and 
 as they do not all break in the same time nor in the same plane, a 
 fibrous fi-acture is thus secured. 
 
 The manner in which test-pieces are broken is of importance when 
 fractures are broadly classed as fibrous or crystalline. To illustrate : 
 two samples may be cut from the same sheet, one of which can be 
 so broken as to present a fibrous fracture, if the fibre be originally in the 
 plate, by simply applying the load gradually, — such a test will give maxi- 
 mum elongation and reduction of area at the point of fracture ; on the 
 other hand, by applying the load suddenly, or so rapidly that the flow 
 of the iron will not follow the effect of the stress, the piece is liable to 
 break with a snap, the fracture in this case having little or no fibre and 
 presenting a more or less crystalline appearance, indicating a lower 
 grade of iron than was actually under test. It will be seen that the ele- 
 ment of time is important when making tensile tests, because it gives 
 the metal a chance to adjust itself to the conditions imposed by the stress 
 beyond the elastic limit. 
 
 Bending Test. — This is one of the severest tests to which wrought- 
 iron boiler-plates can be subjected. It is made by shearing oif a strip of 
 any convenient width from the plate, say one or two inches, and bending 
 it down cold upon itself without fracture on the outer curve, as shown in 
 Fig. 9. Very few irons except the better grades of flange iron will stand 
 a test of this kind ; but the latter grade of iron f inch thick and less 
 should bend double without fracture ; plates ys" to f inch thick of a similar 
 grade should bend down cold upon a plate of its own thickness. 
 
 Hot bending tests are in no respect different from cold bending tests, 
 except that the iron is heated to at least a cherry -red heat before the 
 operation of bending is begun. Any iron which when heated to redness 
 will not bend over double upon itself, either with or against the grain, 
 and without fracture on the outer curve or along the edges of the plate, 
 is not fit to enter into steam-boiler construction. 
 
 Hammer Test. — This test is sometimes resorted to for ascertaining 
 the internal defects of a plate when caused by lamination or imperfect 
 welding during manufacture, and consists in lightly tapping the entire 
 surface of the plate with a light hand-hammer. The plate is usually 
 placed on edge, its weight resting on any convenient bearing at each end 
 to keep it off the floor ; the upper edge is supported in any manner that 
 will not interfere with the vibration of the plate. If a plate thus arranged 
 be struck with a hand-hammer, it will give a clear, ringing sound, which 
 will be the tone of that particular plate ; if now the hammering be pro- 
 ceeded with on say 4-inch centres over the entire surface and on both 
 sides of the plate, and this same characteristic tone due to the size and 
 weight of the plate be observed, the plate is judged to be solid ; but if a 
 dull sound is emitted, it is reasonably certain that a defect exists in the 
 plate at that point. 
 
WROUGHT IRON 35 
 
 Chemical Properties of Wrought Iron. — The operation of pud- 
 dling does not eliminate all the impurities in cast iron during its conver- 
 sion into wrought iron, but to show how nearly such elimination occurs 
 in practice, reference is directed to the following analysis of a sample of 
 55,000 tensile strength wrought-iron plate : 
 
 Iron 99.20 
 
 Carbon .04 
 
 Manganese .17 
 
 Silicon 15 
 
 Sulphur .03 
 
 Phosphorus .21 
 
 Oxygen .20 
 
 100.00 
 
 Of the above substances, those which act most injuriously upon iron 
 are sulphur and phosphorus, the former making the iron red-short, and 
 the latter cold-short. 
 
 Red-short or hot-short iron is a defect generally attributed to the 
 sulphur present in the finished plate or bar. This sulphur may have been 
 present in the pig-metal from which the wrought iron was made, but 
 much of it comes from the coal used as fuel in puddling. Red-short 
 irons are often tenacious and otherwise good when cold, but become 
 brittle and are easily broken when hot ; such irons weld with great diffi- 
 culty. 
 
 Cold-short iron is very brittle when cold, cracking badly, or breaking 
 if bent at a sharp angle or doubled ; but such irons may be forged and 
 welded at a high heat. This defect is a much more serious one than red- 
 shortness. Irons which have an excess of phosphorus are cold-short. 
 
 Commercial Qualities of Boiler-Plate.— The quality of boiler- 
 plate is dependent upon the impurities in the crude metal from which it 
 is made, and upon the amount of working which it receives to prepare 
 it for the market. Wrought iron is improved in quality by judicious 
 heating and working, especially hammering, but it soon reaches its 
 greatest strength, after which reheating and rolling is found to reduce 
 its strength and ductility. Unless portions of plate have been actually 
 tested, or the plates are known to have been made from blooms of the 
 very best quality, it is not safe to assume a greater tensile strength than 
 45,000 pounds per square inch of section. This applies to such irons 
 only as are stamped by reputable makers C. H. No. i and higher 
 grades. These latter are usually designated by some private brand or 
 trade-mark. 
 
 C. Iron, or charcoal iron, is the common boiled or puddled iron rolled 
 into bars or plates. This grade of iron is porous, and will become very 
 brittle with repeated heating and cooling. It will not stretch much 
 before breaking, and will break suddenly. Its tensile strength ranges 
 
36 BOILERS AND FURNACES 
 
 usually from 30,000 to 40,000 pounds per square inch. It is only suited 
 for tank-work, and ought never to enter into any portion of boiler con- 
 struction. 
 
 C. No. I Iron, or C. H. iron (charcoal hammered, as it is oftener 
 known), is the same iron as the above, except that it is subjected to 
 more careful working and is hammered into suitable blooms before 
 rolling. This iron very much resembles the common iron in its general 
 qualities, having but little elasticity and breaking suddenly. Like the 
 above, it becomes very brittle by repeated heating and cooling, though 
 somewhat stronger than C. iron, its tensile strength ranging from 35,000 
 to 45,000 pounds per square inch. It is not a suitable iron for boiler 
 construction. 
 
 C. H. No. I Shell Iron is made from C. H. blooms, with the 
 addition of selected scrap, the whole being thoroughly welded under a 
 heavy steam-hammer and afterwards rolled into plates. This iron, like 
 the other two just described, is injuriously affected by repeated heating 
 and cooling, which has the effect to render it brittle. This is the quality 
 of plate generally used in the construction of land boilers using pressures 
 of steam below 100 pounds per square inch. It rarely enters into the 
 construction of boilers for river or ocean service, its principal defect 
 being a lack of homogeneity and imperfect welding. Its tensile strength 
 is from 45,000 to 50,000 pounds per square inch. 
 
 C. H. No. I Flange Iron is similar to the above, the difference 
 being that only the very best scrap iron and charcoal hammered blooms 
 are used. The greatest care is exercised in the selection of materials, 
 and the working is such as to insure thorough welding. In texture it 
 is less fibrous and more granular than any of the irons preceding it. 
 On account of its nearer approach to a homogeneous structure it is less 
 liable to blister or crack in the fire. It will stand repeated heating and 
 cooling and should have good flanging qualities. The tensile strength 
 should never fall below 50,000 pounds per square inch and rarely ex- 
 ceeds 60,000 pounds. The elastic limit will vary from 18,000 to 25,000 
 pounds per square inch, and will stretch from 25 to 30 per cent, in ordi- 
 nary 2-inch specimens. This is the highest grade of iron regularly 
 offered in the market, and was extensively used in the construction of 
 marine boilers and for the heads and other flange-plates of land boilers. 
 
 Brand. — The brand of plate iron is a commercial index to its quality. 
 Plates of a grade corresponding to C. H. No. i iron should always be 
 stamped with the maker's name and the guaranteed tensile strength, 
 thus : 
 
 SMITH, JONES & CO., 
 
 C. H. No. I Flange, 
 
 55,000 T. S. 
 
 This method of stamping was introduced in order to meet the re- 
 quirements of the United States Government regulations with reference 
 
MILD STEEL 37 
 
 to the quality of plates entering into steam boilers intended for use on 
 steam vessels in the United States. 
 
 Defects in Iron Boiler-Plates. — These are principally imperfect 
 welding, brittleness, and low ductility, all of which may be largely over- 
 come by a proper selection of materials in the first stage of manufacture 
 and by a careful manipulation during the successive operations of re- 
 heating, welding, and especially by a thorough working under a heavy 
 steam-hammer. As boiler-plates are always cut to dimensions at the 
 mill, one defect of plate iron is not known to the purchaser, — that is, 
 rough edges, which is a sign of red-shortness, the plate containing an 
 excess of sulphur, which may have been absorbed from the coal during 
 the process of puddling. Such plates do not weld readily. 
 
 Surface Defects in Plates caused by imperfections in the rolls 
 should be carefully examined to ascertain whether the quality of the 
 plate is likely to be affected thereby ; so, also, cracks in the scale should 
 be carefully examined, though usually they do not amount to much, 
 rarely entering into the iron. Scabs, occasioned by the imbedding of 
 hard pieces of cinder while the metal is at a high heat, may extend some 
 distance below the surface and prove a serious defect. 
 
 Of interior defects lamination is perhaps the most serious. This 
 defect occurs when welding is imperfect, which may have been occa- 
 sioned by the presence of cinder, sand, or other impurity in the pile, 
 preventing welding contact. This is the defect which later on, when 
 the plate is exposed to the fire, produces a blister. Lamination is diffi- 
 cult of detection, because both sides of a sheet may be faultless and give 
 no surface indication of its occurrence. The object of the hammer test 
 is to lead to the detection and location of such defects. 
 
 Blisters. — This defect seldom manifests itself until the plate has 
 been put into the boiler and subjected to the repeated action of heating 
 and cooling : it is directly traceable to lamination, already referred to. 
 Blisters are much more common in wrought-iron plates than in those of 
 mild steel. The remedy for a blister is either a patch or a new plate. 
 
 PART III.— MILD STEEL. 
 
 Steel occupies a position intermediate between cast iron, which con- 
 tains so much carbon that it is very brittle, and wrought iron, which 
 contains so little carbon that its influence is imperceptible. The differ- 
 ence between mild steel and wrought iron does not wholly depend upon 
 the larger quantity of carbon contained in the former over the latter, 
 but rather that in the preparation of steel it has acquired that property 
 which permits casting into a malleable ingot. 
 
 Steel is characterized by fine granular texture, and when the con- 
 tained carbon amounts to 0.40 to 0.50 per cent, it has the property of 
 hardening, which unfits it for use in steam-boiler construction : steel for 
 that purpose should not contain more than from o. 12 to 0.24 percent, of 
 
38 
 
 BOILERS AND FURNACES 
 
 carbon. Mild steel is recommended as a material for steam boilers 
 because of its homogeneity, high tensile strength, malleability, ductility, 
 and freedom from lamination and blisters. 
 
 Open-Hearth Steel. — This process of steel-making was introduced 
 in this country in 1868, and by 1871 its manufacture had attained a high 
 degree of perfection. The physical qualities of open-hearth steel were 
 so admirably adapted for boiler-plates that it received the prompt recog- 
 nition to which its merits fully entitled it, rapidly displacing both crucible 
 and Bessemer boiler-plates in open market. 
 
 Carbon. — Mild steel plates for boilers vary as to the total quantity 
 of carbon from 0.12 to 0.24 per cent., depending upon the thickness of 
 the plate. In preparing specifications for mild steel plates less than ^ 
 inch thick the carbon limit should not exceed 0.15 per cent., as addi- 
 tional carbon is likely to reduce the ductility of the plate. The ultimate 
 tensile strength of open-hearth flange or boiler steel should not vary 
 outside the Hmits of 52,000 to 62,000 pounds per square inch, with an 
 elongation of not less than 25 per cent. A steel having 0.15 per cent, 
 of carbon is quite as high as should be used for boiler-plates from ^ to 
 Yo, inch thick ; but for plates say ij^ inches thick the carbon limit may 
 be as high as 0.25 per cent., with intermediate percentages of carbon 
 for intermediate thicknesses of plate. 
 
 TABLE IL 
 
 SHOWING THE CARBON PROPERTIES OF OPEN-HEARTH STEEL. 
 
 
 
 
 Average 
 
 Average Ulti- 
 
 Average 
 
 Final 
 
 Elongation. 
 
 
 Carbon. 
 
 Manganese. 
 
 Phosphorus. 
 
 Original 
 
 Sectional 
 
 Area. 
 
 mate Tensile 
 Strength per 
 Square Inch. 
 
 Average 
 Final Area. 
 
 Per cent. 
 
 Per cent. 
 
 Per cent. 
 
 Square Inch. 
 
 Pounds. 
 
 Per cent. 
 
 Per cent. 
 
 .12 
 
 •351 
 
 •0505 
 
 .6180 
 
 58,226 
 
 27.18 
 
 45.29 
 
 •13 
 
 •340 
 
 .0491 
 
 •5491 
 
 58,352 
 
 26.72 
 
 46.73 
 
 .14 
 
 •375 
 
 .0486 
 
 •5942 
 
 60,569 
 
 26.90 
 
 46.18 
 
 .15 
 
 •383 
 
 .0518 
 
 .5719 
 
 61,618 
 
 26.75 
 
 48.16 
 
 .16 
 
 .393 
 
 .0528 
 
 .5641 
 
 62,517 
 
 25.65 
 
 49.46 
 
 .17 
 
 .404 
 
 .0461 
 
 •5492 
 
 63,333 
 
 25.89 
 
 50.45 
 
 .18 
 
 .416 
 
 .0487 
 
 •5719 
 
 65,169 
 
 24.81 
 
 48.71 
 
 Three percentages above and below o. 15 per cent, are given to show 
 how the physical properties of the steel change with the varying carbon. 
 
 The above steel developed through a series of more than 100 tests 
 an average increase of tensile strength per o.oi per cent, of carbon of 
 1387.5 pounds. 
 
 Average decrease of final elongation per o.oi per cent, of carbon, 
 0.425 per cent. 
 
 Average increase of final area per o.oi per cent, of carbon, 0.600 
 per cent. 
 
MILD STEEL 39 
 
 Manganese. — In the above series of tests it was found that with 
 increasing carbon there was also an increase of manganese, which may 
 probably be accounted for in diminished oxidation in the furnace. The 
 exact function of manganese in steel is not clearly understood ; the belief 
 is, however, that it deoxidizes the bath as well as removes the sulphur. 
 This is inferred from the disappearance of most of the sulphur from the 
 iron in the bath and partly from the circumstance that only about one- 
 half the metallic manganese added at the last of the charge is found in 
 the analysis of the resultant steel. 
 
 Phosphorus. — This element in steel has the property of rendering 
 it cold-short, and as boiler-plates are usually worked cold, the less there 
 is of it in the plates the better. As a hardener of steel, phosphorus is 
 generally considered to be more effective than carbon, but its secondary 
 effects are very different. Other things remaining the same, increase 
 of phosphorus, besides raising the tensile strength, notably raises the 
 elastic ratio, diminishes the elongation, and more especially diminishes 
 the reduction of area. Its effect in diminishing elongation, and prob- 
 ably also reduction of area, appears to be largely dependent on the 
 amount of other elements present, especially of silicon. For conve- 
 nience, the latter element must be supposed to vary little and be present 
 in quantity not above 0.04 per cent., as is common in open-hearth steel. 
 The effect of phosphorus is then identical in nature to that of cold-rolling 
 or finishing, and is to be taken account of in much the same manner. 
 
 Sulphur. — The presence of this element renders steel hot-short, and 
 thus affects the working in the steel- works rather than in the boiler-shop, 
 except in flange-plates. Sulphur should not exceed 0.04 per cent, in 
 steel boiler-plates, and even at this percentage the plates should contain 
 at least 0.25 per cent, of manganese in order to counteract the hot-short 
 effects of the sulphur. 
 
 Silicon. — This element, even in small quantities, renders mild steel 
 hard and decreases its ductility. It ought not to exceed 0.05 per cent, 
 in any steel intended for steam boilers. 
 
 PHYSICAL PROPERTIES OF MILD STEEL. 
 
 The qualities to be determined and for which physical tests are 
 undertaken upon samples of steel boiler-plates are mainly for elastic 
 limit, elongation, ultimate strength, reduction of area at point of frac- 
 ture, cold-bending, the drifting test, and the quenching test. The first 
 four are determined consecutively and are classed usually under tensile 
 tests ; the others are separate tests. 
 
 Tensile Tests. — The forms and dimensions of test-pieces for boiler- 
 plate in this country do not conform to any standard. The short test- 
 piece so long included in the Rules and Regulations prescribed by the 
 Board of Supervising Inspectors of Steam Vessels, shown in Fig. i, is 
 still in use. For tensile strength only this test-piece is perhaps good 
 
40 
 
 BOILERS AND FURNACES 
 
 enough, the groove simply indicating where the break shall occur. 
 Such a break will also give the fractured area of the specimen ; but it 
 is not the best form of test-piece for obtaining the ductility of the mate- 
 rial, though this may be measured by making prick-punch marks on 
 either side of the centre at any distance which will be unaffected by the 
 stretch of the piece, and comparing the final distance with the original 
 before the test began. This form of test-piece has the advantage of 
 being easily taken from small pieces, and for pieces of small ductility 
 the value of the tensile strength so obtained is practically correct. 
 
 The short filleted form, shown in Fig. 2, was quite extensively used 
 by steel-makers in this country prior to the adoption of the standard 
 
 Fig. 
 
 ■Vl 
 
 Tesi piece io bejanie 
 
 TV 
 
 specifications, August 9, 1895, and, unless otherwise stated, may be re- 
 garded as the size used. The tensile strength and fractured area for a 
 ductile material like mild steel are much nearer the results obtained 
 from long pieces than in the groove form and more uniform, but are 
 still too high. The ductility is somewhat arbitrary in that the measured 
 length includes the short end fillets, which are in a different condition 
 of stress from the straight portion. Being very largely dependent on 
 the proportions of the test-piece, the ductility results from these speci- 
 mens always read very high, especially for the softer materials. The 
 
 Fig. 3. 
 
 drawing. Fig. 3, is a full-size reproduction of a test-specimen of basic 
 open-hearth boiler-steel by the Lukens Iron and Steel Company. The 
 dotted lines indicate the original size and the full lines the final outlines 
 after the test. Elongation and reduction of area are both clearly ex- 
 hibited. 
 
MILD STEEL 
 
 41 
 
 The long filleted form, shown in Fig. 4, is one recommended by the 
 American Society of Civil Engineers as being well adapted to show all the 
 
 Fig. 4. 
 
 physical properties likely to be developed by a tensile test. The length 
 between the fillets is commonly 8 inches, though there are variations on 
 either side, such as 6 and 10 inches. The 8-inch test-pieces are used in 
 the French navy, by the English Admiralty, and by the United States 
 Navy Department. This uniformity in the dimensions of test-pieces 
 makes the comparison of American and foreign results a simple matter. 
 
 TABLE in. 
 
 tests of annealed mild steel boiler-plates made to show the effect 
 
 of length of test-piece, united states navy department. 
 
 Fig. 5. 
 
 Fig. 6. 
 
 A. 
 
 B. 
 
 C. 
 
 D. 
 
 E. 
 
 F. 
 
 G. 
 
 Original 
 Length be- 
 tween Wit- 
 ness Marks. 
 
 Average 
 Original 
 Width. 
 
 Average 
 
 Original 
 
 Thickness. 
 
 Average 
 Original . 
 Sectional 
 Area. 
 
 Average Ulti- 
 mate Tensile 
 Strength per 
 Square Inch. 
 
 Average 
 
 Final 
 
 Elongation. 
 
 Average 
 Final 
 Area. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Square Inch. 
 
 Pounds. 
 
 Per cent. 
 
 Per cent. 
 
 Groove. 
 
 0-993 
 
 .660 
 
 .6550 
 
 63,065 
 
 40.5 
 
 46.0 
 
 ^% 
 
 0.990 
 
 .665 
 
 .6583 
 
 56,660 
 
 49- 
 
 43-5 
 
 2 
 
 1. 021 
 
 •663 
 
 .6750 
 
 56,600 
 
 44.0 
 
 41-5 
 
 2X 
 
 0.968 
 
 .668 
 
 .6466 
 
 56,200 
 
 40.0 
 
 40.0 
 
 3 
 
 1. 015 
 
 .660 
 
 .6700 
 
 56,400 
 
 38.0 
 
 39-5 
 
 4 
 
 0.980 
 
 .658 
 
 .6445 
 
 56,200 
 
 34-5 
 
 38.0 
 
 5 
 
 0.970 
 
 .650 
 
 .6300 
 
 56,445 
 
 33-0 
 
 37-0 
 
 6 
 
 1.037 
 
 .648 
 
 .6725 
 
 55,700 
 
 29-5 
 
 36.9 
 
 7 
 
 1. 000 
 
 .644 
 
 •6435 
 
 56,250 
 
 28.5 
 
 39-5 
 
 8 
 
 0-933 
 
 .664 
 
 •6195 
 
 55,200 
 
 27.9 
 
 38-5 
 
 9 
 
 0.963 
 
 •654 
 
 .6298 
 
 55,955 
 
 26.6 
 
 36.4 
 
 10 
 
 1. 000 
 
 .663 
 
 .6630 
 
 54,750 
 
 27.4 
 
 36.8 
 
42 BOILERS AND FURNACES 
 
 The standard test-piece adopted by the Association of American 
 Steel Manufacturers, August 9, 1895, as shown in Fig. 7, is for sheared 
 plates of special open-hearth steel for boilers. The tests and inspections 
 are required to be made at the place of manufacture prior to shipment ; 
 
 
 
 Fig. 7. 
 
 ■ 
 
 
 
 . 
 
 ^26o£c^ »' i 
 
 1 
 
 ....M'- 
 
 iv4••/-i•V^i.v^|../a..^..i../4^^'%lS^i|S" 
 
 the tensile strength, limit of elasticity, and ductility are to be determined 
 from a standard test-piece of the above dimensions cut from the finished 
 material. 
 
 Test-pieces for plates for the new United States war-ships were em- 
 ployed as in the form shown in Fig. 4. These pieces were as nearly as 
 possible in the same condition as finished at the rolls. The length 
 A B was 8 inches, and of uniform cross-section, in which the sectional 
 area was not less than ^ or more than 0.8 of one square inch. 
 
 The test-piece was not annealed unless the finished material was to 
 be annealed, and no steel for boilers which had to be worked at a heat 
 or which had to be annealed in the boiler-shop was annealed at the steel- 
 works. 
 
 Each test-piece was submitted to a direct tensile stress until it broke. 
 The initial stress applied was 25,000 pounds per square inch, and this 
 first load was kept in continuous action for one minute. An observation 
 was then made of the corresponding elongation measured upon the 
 original length of 8 inches. 
 
 The stress was then slowly increased until the principal elastic limit 
 was determined ; then additional loads sufficient to produce an increase 
 of stress of 5000 pounds per square inch were added at intervals of half 
 a minute until the stress reached 50,000 pounds per square inch of the 
 original section, after which the increments of stress were reduced to 
 1000 pounds per square inch. Upon close approach to the probable 
 ultimate strength, the load was increased gradually and its maximum 
 value carefully noted. The final elongation was that obtained after 
 rupture. 
 
 The acceptance of any mild steel boiler-plates under the contract 
 was conditioned on an ultimate tensile strength of not less than 57,000 
 pounds, nor more than 63,000 pounds per square inch of original 
 section, and a final elongation in 8 inches of not less than 25 per cent. ; 
 but the acceptance of plates under these tests did not relieve the con- 
 tractor from the necessity of making good any material which afterwards 
 failed in working. 
 
MILD STEEL 43 
 
 Direction of Grain. — In testing wrought iron it is quite common 
 to prepare test-pieces from plates in the direction of rolling and at right 
 angles to this direction ; these are commonly spoken of as being with 
 or across the grain. Steel, being a more homogeneous material, shows 
 little difference between the two, as indicated in the following memo- 
 randa of mild-steel tests having 0.16 per cent, carbon and 0,40 per cent, 
 manganese : 
 
 Stress applied lengthwise of the plate, — 
 
 Average tensile strength in pounds per square inch .... 59,400 
 
 Average final elongation, per cent 27.00 
 
 Average final area, per cent 47.00 
 
 Stress applied crosswise of the plate, — 
 
 Average tensile strength in pounds per square inch .... 58,995 
 
 Average final elongation, per cent 23.15 
 
 Average final area, per cent 47.00 
 
 Elastic Limit. — For mild steels intended for steam boilers, the 
 elastic limit is of much importance, because it is practically its measure 
 of strength ; for boiler-plates the elastic limit will vary from 50 to 60 
 per cent, of its ultimate tensile strength. The standard specifications 
 provide that all special open-hearth plate and rivet steels for boilers shall 
 have an elastic limit not less than one-half the ultimate tensile strength. 
 
 Elastic Ratio. — This is obtained by dividing the elastic limit per 
 square inch by the ultimate strength per square inch. For example, if 
 a plate have an ultimate tensile strength per square inch of 57,000 
 pounds and an elastic limit of 26,500 pounds, the elastic ratio would 
 be 26,500 X 100 -i- 57,000 = 46.49 per cent. Examples of this kind 
 are interesting as showing any departure from the average as to physical 
 condition other than lack of homogeneity — the condition as to internal 
 strain or comparative annealing, i.e. — by the ratio which the elastic 
 limit bears to the tensile strength. 
 
 Elongation. — The average value of the extension of mild steel 
 boiler-plates at the tensile limit varies with the quality of the steel, the 
 length of the specimen, and the thickness of the plate : a final elonga- 
 tion of 44 per cent, on a 2-inch specimen would show probably 28 per 
 cent. ; on an 8-inch specimen the elongation of a ^-inch plate would 
 exceed that of a i-inch plate by at least 5 per cent., both specimens from 
 the same ingot. The standard specifications for elongation call for 28 
 per cent, in extra soft steel, 26 per cent, for fire-box steel, and 25 per 
 cent, for flange or boiler steel. 
 
 Reduction of Area. — There are two methods of recording this 
 quantity : one by giving the actual area of the fracture, irrespective of 
 the original area, — this is the ordinary method ; the other records the 
 fractured area in percentage of the original area. In order that no con- 
 
44 BOILERS AND FURNACES 
 
 fusion of terms shall arise, the latter is called in the United States navy 
 
 reports on mild steel the ' ' final area. ' ' 
 
 To measure the final area, the piece should be fitted, the least width 
 
 measured, and thickness at each ede^e and 
 Fir 8 . ■ • 
 
 in the centre in the plane of least width. 
 
 -j-—^- ■ 
 
 
 ^ On account of the hollow in the section, 
 
 'ja as shown in Fig. 8, it is necessary to in- 
 
 ■^ troduce a formula for the thickness. This 
 
 is best done by applying Simpson's rule 
 
 for three ordinates. Thus, if t^, tg, tg be the thickness at A, C, and B 
 
 respectively, and b the least width, we shall have, — 
 
 Final area = ^ (ti + i,t^ + 13). 
 o 
 
 Obtaining data from results given below, the above formula may be 
 worked out arithmetically thus : 
 
 ti = .300, 
 
 tg = .272 4t2 = 1.088, 
 tg = .308 = 1.696. 
 
 1.696 -^ 6 = .283 = mean thickness. 
 
 Fractured width b =■ .936 X .283 = .2649 = fractured area, then 
 100 X .2649 -^ .5229 original area = 50.66 = final area in per cent, 
 of the orieinal area. 
 
 TENSILE TESTS OF MILD STEEL, UNITED STATES NAVY 
 DEPARTMENT. 
 
 CARBON, 0.14 per cent. 
 Original section : 
 
 Width 1.254 inches. 
 
 Thickness 417 inch. 
 
 Area 5229 square inch. 
 
 Fractured section : 
 
 Width 936 inch. 
 
 Thickness at edges and centre, .300, .272, .308 inch. 
 
 Area 2646 square inch. 
 
 Elastic limit, pounds per sauare inch 40, 160 
 
 Ultimate tensile strength, pounds per square inch 62,700 
 
 Elastic ratio 64.06 per cent. 
 
 Final elongation 25.5 per cent. 
 
 Final area in per cent, of original area . . . .50.60 per cent. 
 
 Modulus of elasticity 30,200,000 
 
 Elongation at tensile limit 19.0 per cent. 
 
 Stress at rupture : 
 
 Pounds per square inch of original area 54,120 
 
 Pounds per square inch of fractured area 106,950 
 Time of test 17 minutes. 
 
MILD STEEL 
 
 45 
 
 Fig. 
 
 Cold- Bending Test. — Any plate which will stand the foregoing 
 tests will stand any cold bending to which it is ever likely to be sub- 
 jected. An open-hearth steel of 60,000 pounds tensile strength and 25 
 per cent, elongation in 8 inches will bend back upon itself, as shown 
 
 in Fig. 9, which is 
 drawn from a speci- 
 men of open-hearth 
 basic boiler-plate, i ^ 
 inches thick, made by 
 the Lukens Iron and 
 Steel Company. An 
 example of double- 
 bending of ^-inch 
 open - hearth basic 
 boiler-plate shown in 
 Fig. 10 is also from 
 the same company. Where T-bars are used for stiffeners or for joining 
 braces or stays in a boiler, the flanges should stand cold bending to the 
 dotted lines shown in Fig. 11. In the cold bending of any specimen 
 prepared by shearing, care should be taken that the edges are smooth 
 and free from sharp indentations, any one of which under high stress 
 may develop a frac- 
 ture. The hammer- ^^^- ^°- 
 ing should be gradual 
 and the blows deliv- 
 ered square to the 
 surface. Failure in 
 cold - bending tests 
 may occur through 
 too high carbon per- 
 centages or by the 
 chemical action of 
 hardening impurities 
 in the steel. 
 
 Drifting Test.— 
 This test consists in 
 taking a strip of steel 
 of any convenient 
 width (2 or 3 inches), drilling a hole >^ or ^ inch diameter for the in- 
 sertion of a drift-pin, and by a succession of drifts through the cold 
 plate enlarging the hole to at least twice its original diameter. Fig. 
 12 shows the effect of drifting a piece of open-hearth steel by the 
 Lukens Iron and Steel Company. The original opening was }4 inch 
 diameter, the drifted hole i^^ inch diameter. The effect upon the 
 material is shown in the cross- section. 
 
46 
 
 BOILERS AND FURNACES 
 
 Fig. II. 
 
 Quenching Test. — The object of this test is to determine whether 
 or not the specimen tested possesses any hardening qualities. The 
 
 standard quenching test consists 
 in heating the specimen to a 
 cherry-red heat and plunging 
 it in water at a temperature of 
 82° Fahr., after which the 
 quenched specimen must be 
 capable of being bent back 
 
 1 upon itself, as shown in Fig. 
 
 J 9, without signs of fracture 
 
 \ anywhere in the bend. The 
 
 ' United States navy require- 
 
 ^ ments are that a specimen thus 
 
 prepared shall bend under a 
 
 press or hammer so that it shall 
 
 be doubled round a curve of 
 
 which the diameter is not more 
 
 than one and a half times the 
 
 thickness of the piece tested 
 
 without presenting any trace of cracking ; further, these test-pieces must 
 
 not have their sheared edges rounded off ; the only treatment permitted 
 
 is taking off the sharpness of the edge with a fine file. 
 
 Fig. 12. 
 
 Bulging Test. — This test is seldom or never required in specifica- 
 tions for steam boilers. Experiments have been made, however, to 
 determine to what extent open-hearth steel plates may be subjected to 
 stresses of this kind. Fig. 13 shows a method employed under the 
 direction of the United States Navy Department upon open-hearth steel 
 
MILD STEEL 47 
 
 plates made by the Nashua Iron and Steel Company. The steel thus 
 tested analyzed as follows : 
 
 Carbon, combined .130 
 
 Manganese . 100 
 
 Sulphur 028 
 
 Phosphorus .011 
 
 Silica 071 
 
 Iron, difference 99.660 
 
 Total 100.000 
 
 A y^g-inch plate was placed over a hole 9 inches square in a piece 
 of wrought iron ; a 6-inch cast-iron shot was then driven down by a 
 2000-pound steam-hammer having a mean stroke of 24 inches until the 
 calotte, or cup, thus formed 
 
 had a depth of 4^ inches. Fig. 13. 
 
 The lower surface was then 
 thoroughly examined and 
 no signs of fracture could 
 be detected. A second 
 trial was then similarly 
 made upon a plate \^ inch 
 thick, with a view to ascer- 
 tain to what extent this 
 test could be carried with- 
 out fracture. 
 
 After breaking three 
 shot, a wrought-iron cylin- 
 der with a spherical end 
 
 was substituted, and at about the sixtieth blow disintegration took place 
 along one side of the cup at a distance of 2 inches from its bottom. 
 The thickness of the plate at the point of fracture was reduced to ^ 
 inch, the depth of the cup being ^.J/g inches. 
 
 When heated it was found that the plate could be folded over until 
 the surfaces met and then bent in the opposite direction to a similar 
 position without fracture, and after repeating this operation four times 
 only a slight fracture took place. 
 
 Annealing. — This detail in manufacture consists in slowly heating 
 the material to a cherry-red heat and allowing it to cool slowly, com- 
 monly under a bed of ashes. The object in annealing plates is to neu- 
 tralize the interior strains which may have arisen from any cause, such 
 as unequal heating or cooling in manufacture, and especially strains set 
 up in a boiler-plate when allowed to cool on a cast-iron straightening- 
 plate. Plates which undergo partial and local heating for flanging by 
 hand, or when successive portions require to be heated for machine- 
 flanging, or plates rather high in carbon which have undergone con- 
 
48 BOILERS AND FURNACES 
 
 siderable cold-working and punching ought always to be annealed. A 
 reverberatory furnace especially constructed for the purpose is the best 
 method of annealing because it insures uniform heating over the entire 
 surface, and if properly attended to there need be little or no danger of 
 local overheating. 
 
 Manufacturers sometimes resort to annealing to prevent rejection of 
 plates when undergoing rigid inspection. For example, a plate may be 
 wanting in ductility partly through chemical and partly through physical 
 causes : annealing will in all probability increase the ductility sufficiently 
 to allow it to pass ; on the other hand, a plate may be slightly above a 
 specified tensile limit : by proper precautions in annealing the plate may, 
 within narrow limits, be reduced in tensile strength and thus come 
 within the limits of acceptance. 
 
 The effect of annealing is shown in the annexed memoranda relative 
 to certain plates intended for the United States cruiser "Chicago." 
 Eight test-pieces from four plates, representing four separate heats, were 
 made from the original sheets ; these plates contained o.ii, 0.13, 0.13, 
 and o. 14 per cent, of carbon respectively. 
 
 Average ultimate tensile strength per square inch 
 
 for the four heats 60,588 pounds. 
 
 Average final elongation 29.54 per cent. 
 
 Average final area 42.44 per cent. 
 
 The annealed test-pieces from the same plates were twenty in number, 
 in which the following differences were observed : 
 
 Average decrease of tensile strength per square inch 6680 pounds. 
 Average increase in final elongation in 8 inches . . . .06 per cent. 
 Average decrease of final area in per cent, of original 
 
 area 4.75 per cent. 
 
 The experiences of officers in charge of inspection and tests was, in 
 the matter of annealing, of somewhat contradictory nature : the average 
 effect of annealing open-hearth steel plates having a tensile strength of 
 55,000 to 65,000 pounds per square inch was a reduction varying from 
 5 to 19 per cent, of the original values for tensile strength and final area 
 and an increase of 12.5 per cent, in the original value for ductility in 
 8 inches, the average original area being practically identical for both 
 conditions. This shows a diminution of tensile strength and increase ol 
 capacity for local distortion. It was observed that changes were greatest 
 in those plates which originally appeared the softest. 
 
 Certain precautions are necessary in annealing steel plates : the fuel 
 should be comparatively free from ordinary impurities, the flame should 
 be kept neutral, and the products of combustion should be kept as much 
 as possible out of contact with the material. The heating up should be 
 rather slow, the metal must not be soaked, and, indeed, should remain 
 
MILD STEEL 49 
 
 in the furnace the shortest time required to effect the desired results. 
 The temperature ordinarily applied is too high, and especially apt to be 
 so locally ; it need not be very much above the temperature of finishing 
 at the rolls, and should never be above a medium cherry. It cannot be 
 too carefully borne in mind that the temperature at which the original 
 physical structure is destroyed and replaced by a coarser and weaker 
 structure is considerably below a bright yellow, and the metal should 
 never be heated so high in annealing. 
 
 The conclusion arrived at from a consideration of results of annealing 
 plates which have not been punched or otherwise worked so as to neces- 
 sitate the removal of purely local strains is that the effect is as apt to be 
 deleterious as beneficial as the process is ordinarily carried out. Good 
 metal shows little improvement in any case ; and while inferior metal 
 may be doctored up to show somewhat better test, the improvement in 
 intrinsic quality is uncertain. But the fact that, as commonly done, 
 annealing may continually and with reasonable certainty lower the work- 
 ing quality, and sometimes excessively so, should prevent any general 
 resort to the practice for boiler or other plate at the mills. 
 
CHAPTER III. 
 
 RIVETED JOINTS. 
 
 Plates for the shells of steam boilers are commonly joined by rivets, 
 and occasionally by welding ; but the latter method of joining the edges 
 of plates is practically confined to the manufacture of tubes, flues, and 
 cylindrical fire-boxes for internally fired boilers. Riveting is at best an 
 unsatisfactory way of making a joint ; but, notwithstanding its disad- 
 vantages in the matter of cost, and inherent weaknesses by reduction of 
 sectional area and bad workmanship, it has become one of the most 
 important details in mechanical engineering. 
 
 Strength of Iron and Steel Plates.— C. H. No. i shell iron of 
 good quality has a tensile strength ranging from 40,000 to 50,000 pounds 
 per square inch ; in the absence of a specific test it may be assumed to 
 be 45,000 pounds in all calculations relating to riveted joints. Table IV. 
 gives the physical properties of wrought-iron plates referred to in this 
 chapter in certain tests made for the Navy Department. 
 
 TABLE IV. 
 
 TENSILE STRENGTH OF WROUGHT IRON USED IN RIVETED-JOINT TESTS, 
 REFERRED TO LATER IN THIS CHAPTER, WATERTOWN ARSENAL. 
 
 -n 
 
 Fig. 14. 
 
 ■m 
 
 tv 
 
 Nominal. 
 
 Actual 
 
 Sectional 
 
 Area. 
 
 Elastic 
 
 Limit per 
 
 Square Inch. 
 
 Ultimate 
 Strength per 
 Square Inch. 
 
 Elongation 
 in Ten 
 Inches. 
 
 Area of 
 Fracture. 
 
 Contraction 
 of Area. 
 
 Inch. 
 
 Square Inch. 
 
 Pounds. 
 
 Pounds. 
 
 Per cent. 
 
 Square Inch. 
 
 Per cent. 
 
 X L 
 
 •389 
 
 36,370 
 
 48,350 
 
 17.0 
 
 .300 
 
 22.9 
 
 y^^ C 
 
 •397 
 
 25,700 
 
 38,160 
 
 4^5 
 
 •365 
 
 8.1 
 
 ^L 
 
 .560 
 
 32,679 
 
 47,590 
 
 13-6 
 
 •454 
 
 18.9 
 
 ^i^ 
 
 .768 
 
 25,789 
 
 45,830 
 
 13^8 
 
 •635 
 
 17^3 
 
 % L 
 
 •935 
 
 25,450 
 
 45,240 
 
 12.4 
 
 • 787 
 
 15.8 
 
 ^L 
 
 1.047 
 
 27,890 
 
 46,750 
 
 12.8 
 
 .866 
 
 17-3 
 
 L — Lengthwise. 
 
 C — Crosswise. 
 
 50 
 
RIVETED JOINTS 
 
 5: 
 
 Mild steel for boilers, whether fire-box or flange, should have an 
 ultimate strength of 52,000 to 62,000 pounds per square inch. In all 
 riveted-joint calculations the tensile strength may be taken as 55,000 
 pounds per square inch. 
 
 As closely agreeing with the above, reference is had to Table V. of 
 physical tests of steel from ^ to f inch thick, in which the average ten- 
 sile strength is 54,885 pounds per square inch. The steel plates were 
 supplied from one heat cast in ingots of same size, the thin plates differ- 
 ing from the thicker plates only in the amount of reduction given the 
 metal by the rolls. 
 
 TABLE V. 
 
 TENSILE STRENGTH OF MILD STEEL USED IN RIVETED-JOINT TESTS, REFERRED 
 TO LATER IN THIS CHAPTER, WATERTOWN ARSENAL. 
 
 Fig. 15. 
 
 /O'^ 
 
 ■K. 
 
 Nominal 
 Thickness. 
 
 Actual 
 
 Sectional 
 
 Area. 
 
 Elastic 
 
 Limit per 
 
 Square Inch. 
 
 Ultimate 
 Strength per 
 Square Inch. 
 
 Elongation 
 in Ten 
 Inches. 
 
 Area of 
 Fracture. 
 
 Contraction 
 of Area. 
 
 Inch. 
 
 Square Inch. 
 
 Pounds. 
 
 Pounds. 
 
 Per cent. 
 
 Square Inch. 
 
 Per cent. 
 
 X 
 
 .378 
 
 39,680 
 
 58,360 
 
 19.2 
 
 .141 
 
 62.7 
 
 y. 
 
 •559 
 
 31,810 
 
 54,025 
 
 27.1 
 
 .208 
 
 62.8 
 
 X 
 
 .751 
 
 31,290 
 
 57,790 
 
 26.1 
 
 • 360 
 
 52:1 
 
 Y, 
 
 •934 
 
 32,100 
 
 52,570 
 
 29.0 
 
 .350 
 
 62.5 
 
 'A 
 
 1. 102 
 
 30,440 
 
 51,680 
 
 28.7 
 
 •394 
 
 64^3 
 
 Loss by Reduction of Area. — The total strength of a plate in a 
 riveted joint will be lessened by that area of metal taken from it to 
 supply places for the rivets. In good quality of wrought iron, punch- 
 ing, if properly done, has little or no effect upon the plates, and no other 
 loss occurs than that due to reduction of area along the line of rivet- 
 holes, so that the net strength of what remains of an iron plate, whether 
 punched or drilled, is equal to that of the original plate for a correspond- 
 ing sectional area ; but in the case of steel plates the value of the net 
 sectional area remaining between the rivet-holes is somewhat uncertain. 
 In some cases there has been observed a marked falling off" in tensile 
 strength, the amount depending upon the hardness and thickness of the 
 plate. Inasmuch as this loss of strength is partially or wholly restored by 
 reaming, or by the annealing effect produced by the hot rivets, it will be 
 near enough to assume in ordinary calculations that the net section 
 between the holes is equal to that of the original plate for the same area 
 of cross-section. 
 
52 
 
 BOILERS AND FURNACES 
 
 Rivet-Holes. — Whether the rivet-holes shall be punched or drilled 
 is, so far as the manufacturer is concerned, largely a matter of conve- 
 nience. Punching is more quickly accomplished than drilling, conse- 
 quently punched holes are cheaper than drilled ; but the former are more 
 likely to contain errors in centring than is the case with drilled holes. 
 A punch once started cannot usually be recalled, and if the plate should 
 be out of centre at the start, the punched hole will be out of centre at 
 the finish, and it is precisely such errors that the drift-pin is called into 
 requisition to correct. 
 
 In large establishments where multiple drilling-machines are em- 
 ployed, drilled holes are supplied at a low cost, quite as low as that of 
 punching, especially for thick plates, which are first punched and after- 
 wards reamed. There can be no question as to the superiority of drilled 
 over punched work, especially when two or more plates are fastened 
 together after rolling and drilled at a single operation. 
 
 Punch and Die. — Two forms of punch are in common use, — the 
 plain punch, with projecting teat or centre, and the spiral punch, as 
 shown in Fig. i6. 
 
 When plates are laid off in the boiler-shop the centres of the hole are 
 commonly marked with a prick-punch. To facilitate centring in the 
 
 punching-machine the project- 
 FiG. i6. ing centre above referred to is ot 
 
 great assistance to the workman, 
 insuring greater accuracy in the 
 spacing of punched holes than 
 would occur if this detail was 
 omitted. 
 
 It is a common practice to 
 make the die of a larger diameter 
 than the punch, say -^^ of an inch, 
 and for all sizes of rivets and all 
 thicknesses of plates above ^ inch 
 and less than -| inch a tapering 
 hole is the result, but if care be 
 taken in riveting, the rivet will 
 conform to the shape of and com- 
 pletely fill the hole. It is claimed 
 that the enlargement of the die lessens the power required to punch the 
 hole, and that the punching is performed with less strain upon the plate. 
 The power required to punch a hole in a plate may be estimated by 
 multiplying the circumference of the punch in inches by the thickness of 
 the plate in fractions of an inch, and this by the tensile strength of the 
 material to be punched. For example, to punch a if-inch hole in a 
 ^-inch plate of steel having a tensile strength of 55,000 pounds per 
 square inch would require 2.945 X -5 X 55,000, = 80,987 pounds. 
 
RIVETED JOINTS 53 
 
 Spiral Punch. — In shearing metal plates it is the universal practice 
 to set the top blade of the shear at an angle to the bottom one, the object 
 being to bring about a gradual separation of the plate ; in other words, 
 to take a little more time to do the same amount of work than if the 
 shears were parallel. Spiral punches are constructed on this principle, 
 and for plates less than one-half the diameter of the punch the operation 
 is very satisfactory in practice. For thicker plates the advantages are 
 not so marked, and when the plate is three-fourths the diameter of the 
 punch any advantage wholly disappears. 
 
 Comparison of Plain and Spiral Punches, — Professor Benja- 
 min's experiments upon Otis steel boiler-plates ^ inch thick, punched 
 with ordinary flat punches, showed a tensile strength of 7.5 per cent, 
 less, elastic limit 5 per cent, higher, and contraction 30 per cent, less 
 than a drilled plate of the same material, showing that the effect of the 
 punching is to render the metal around the hole more brittle and less 
 ductile than before. When a spiral punch was used, the ultimate 
 strength was only 3 per cent, less than in the drilled plate. 
 
 Effects of Punching Steel Plates. — The observed changes in the 
 material in the line of punched holes are increased hardness, alteration 
 of physical structure, and loss of ductility. In tests made several years 
 ago from steel containing more carbon than is now allowed for boiler- 
 plates, from specimens which had been cut from different portions of the 
 same plate and in the same line of punched holes it appeared that the 
 disturbance of the material was confined to within a very short distance 
 around the hole, extending from -^ to y^ of an inch, and that by drilling 
 and reaming out such punched holes and then testing the plate, making 
 proper allowance for the reduced area, no perceptible decrease of strength 
 was noted. Some more recent experiments on mild steel plates contain- 
 ing 0.12 carbon, 0.37 manganese, testing 64,200 pounds tensile strength, 
 with a ductility of 25. 15 per cent, in 8 inches, showed the following effect 
 in punching, with or without countersinking : 
 
 Punched ^-inch hole in 2-inch X .510-inch plate, as in Fig. 17. 
 
 Fig. 17. 
 
 Effective dimensions of test-piece i-3i3 X .510 inch. 
 
 Effective sectional area 6696 square inch. 
 
 Ultimate tensile strength 55, 000 pounds. 
 
 Reduction of strength 14-33 per cent. 
 
54 BOILERS AND FURNACES 
 
 Punched |^-inch hole and countersunk, as shown in Fig. i< 
 Fig. i8. 
 
 Effective sectional area 6215 square inch. 
 
 Ultimate tensile strength 61,400 pounds. 
 
 Reduction of strength 4.36 per cent. 
 
 Commenced to crack at part of hole not countersunk, at 58,200 pounds 
 per square inch. 
 
 Punched ^^-inch hole and countersunk, as shown in Fig. 19. 
 
 Fig. 19. 
 
 — — — ^..,„„„„„.,.^„^^ 
 
 4.30- 
 _ ~~p~L _-j 
 
 Effective 1 t n 1 i 6163 square inch. 
 
 Ultimate tensile strength 65,560 pounds. 
 
 Increase of strength 2.12 per cent. 
 
 Commenced to crack at small edge of hole, at 61,980 pounds per square 
 inch. 
 
 These results show, first, a reduction of strength of 14.33 pc cent, 
 due to ordinary punching of a -f|-inch hole in a ^-inch steel plate of a 
 quality suited for steam boilers ; second, a recovery of strength, 4.36 per 
 cent, of the original, by partial countersinking; third, a gain of 2.12 
 per cent, when a punched hole is countersunk all the way through. The 
 loss in strength in the first example may be attributed to the formation 
 of a thin ring of highly strained metal forming the walls of the hole ; in 
 general, this thickness will depend upon the carbon properties or hard- 
 ness of the plate, upon its thickness, and the relative size of punch and 
 die. The harder the steel, the thicker the plate, and the larger the die 
 relatively to the punch, the thicker is the overstrained ring of metal, 
 and the greater the amount of subsequent reaming and drilling necessary 
 to remove it. 
 
 The increase of strength shown in the last of the three examples 
 accords with a fact now generally recognized that the effect of a hole 
 produced by a cutting tool in a steel plate is to increase the ultimate 
 strength of the net section ; and while the effect of punching is always 
 to overstrain the adjacent material, yet if the damage to the material is 
 
RIVETED JOINTS 
 
 55 
 
 equal to or less than the gain due to the difference of distribution of the 
 resisting area owing to the presence of the hole, no apparent loss of 
 strength will ensue. 
 
 Countersunk Rivets. — Experiments prove that the lap in riveted 
 joints is an element of weakness irrespective of the loss of strength by 
 rivet-holes. The thicker the plate the greater is the distorting leverage 
 shown in Figs. 25 and 25 A. Clark reasons that because the absolute 
 strength of a )^-inch lap- welded joint under test was not greater than 
 that of a 0-inch joint under similar test, the principle here noticed may 
 account for the practically equal strength of joints made with counter- 
 sunk rivets compared with those having external rivet-heads, notwith- 
 
 FlG. 20. 
 
 standing the greater reduction of solid section by countersinking ; the 
 leverage is shortened, and it may be measured from the centre of the 
 cylindrical part of the rivet in the line A B, in Fig. 20, or thereabouts, 
 towards the inner side of the plate. 
 
 Size of Rivet. — The thinnest plate used in ordinary boiler- work 
 is ^ inch, and for such plates f-inch rivets are recommended, which, 
 with the customary allowance of yV-inch clearance, makes the rivet-hole 
 Y^ inch in diameter, the proportion of diameter of hole to thickness of 
 plate being 2.75 to i. Nearly all English writers give ^ inch as the 
 diameter of rivet, or a y^g-inch hole for ^-inch plates, — a ratio of 2.25 
 to I ; but these proportions have not been in favor and are not com- 
 monly used in this country. 
 
 For thick plates, say f-inch and thicker, the size of the hole is 
 governed somewhat by the ability to properly upset the rivet to fill the 
 hole and form a proper head, especially in such places as require the 
 work to be done by hand. For f-inch plates the diameter of rivet is 
 commonly i\ inches, or ly^g hole. This gives a ratio of diameter of 
 hole to thickness of plate of 1.58 to i. For intermediate thicknesses 
 the diameter of rivet may be found by interpolation, as in Table VI, 
 
 When rivets pass through three or four plates overlapping each other 
 at a joint, it is in accordance with good practice to increase the diameter 
 of such rivets yg- or |- inch, at the discretion of the designer. Inci- 
 dentally it may be remarked that if the plates are to be lap-jointed on a 
 
56 
 
 BOILERS AND FURNACES 
 
 continuous plate, as shown in Fig. 21, one of the sheets must be forged 
 wedge-shaped and the top sheet bent to fit the taper of the wedge. If, 
 however, all the four corners meet at a common point, as in Fig. 22, 
 the two inner sheets are forged wedge-shaped, right and left, the two 
 equalling a single thickness, or approximately so. The upper and lower 
 plates require no forging. Joints of this kind should be drilled after all 
 the plates are assembled in place. 
 
 TABLE VL 
 
 RIVET-HOLES. 
 
 
 Diameter. 
 
 
 Ratio of Diameter of 
 
 
 
 
 Area of Hole. 
 
 Hole to Thickness 
 of Plate. 
 
 of Plate. 
 
 
 
 
 Rivet. 
 
 Hole. 
 
 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Square Inch. 
 
 
 X 
 
 y?, 
 
 H 
 
 ■371 
 
 2.75 to I 
 
 T6 
 
 T6 
 
 Y 
 
 .442 
 
 2.40 to I 
 
 H 
 
 %■ 
 
 If 
 
 .518 
 
 2.17 to I 
 
 s 
 
 \i 
 
 H 
 
 .601 
 
 2.00 to I 
 
 H 
 
 if 
 
 .690 
 
 1.87 to I 
 
 TS 
 
 a 
 
 I 
 
 .785 
 
 1.78 to I 
 
 H 
 
 I 
 
 ItV 
 
 .887 
 
 1.70 to I 
 
 H 
 
 \t 
 
 iVs 
 
 •994 
 
 1.64 to I 
 
 Ya 
 
 lA 
 
 1. 108 
 
 1.58 to I 
 
 Pitch of Rivets. — The distance apart from the centre of one rivet 
 to the centre of the next rivet is called its pitch. As the gross strength 
 of any plate is directly affected by the removal of a portion of its sectional 
 area necessary for the rivet-holes, it follows that the fewer holes there 
 are for a given width of plate the greater will be the strength of the 
 
 Fig. 
 
 21. 
 
 1 
 i 
 1 
 
 
 1 Q 
 
 000 
 
 
 Fig. 22. 
 
 
 D~l 
 
 ; 
 I 
 
 00 
 
 ;0 ; 
 
 
 1 
 
 1 
 1 
 
 
 plate remaining. If the pitch of rivets be too close, the strength of the 
 plate is unnecessarily diminished ; if too wide, there is danger of shear- 
 ing the rivets, unless they are increased in diameter corresponding to 
 the increased pitch, which is not always practicable. 
 
 When designing a joint, the diameter and pitch of rivets should ap- 
 
RIVETED JOINTS 
 
 57 
 
 Fig. 
 
 proximate in strength that of the plate remaining- between the rivet- 
 holes, taking into account the tensile strength per square inch of the 
 latter and the shearing strength of the former ; but such a joint can only 
 be secured by the use of large rivets and wide spacing. Such joints are 
 difificult to keep tight under high pressures, and tightness of joint is a 
 detail quite as important as that of tensile strength. 
 
 Iron Plates and Iron Rivets. — Safe working dimensions for 
 single-riveted joints have been ascertained by actual tests. For ex- 
 ample, tests were made upon a ^-inch plate having a tensile strength 
 of 47,925 pounds per square inch. Specimen was 10 inches in width, 
 with 5 ^-inch rivets on 2-inch centres, punched holes, the punch 0.695 
 inch diameter, the die 0.752 inch diameter, making a hole slightly 
 tapered. The rivets were of iron of the best quality. This joint foiled 
 by shearing four of the rivets and breaking a 
 corner out of the plate at the fifth rivet, as 
 shown in Fig. 23. The elongation of holes 
 ranged from 0.04 to 0.08 inch. The effi- 
 ciency of this joint was 64. i per cent. The 
 weakness was in the rivets. The maximum 
 shearing stress on the rivets was 38,640 
 pounds per square inch, which exhibits the usual proportion between 
 tensile strength of plate and shearing strength of rivets when both are 
 made of first-quality material. 
 
 A second experiment made from the same ^-inch plate, 9^ inches 
 wide, with 6 ^-inch iron rivets from the same lot, holes on i^-inch 
 centres, punched as above, failed by fracturing one of the plates directly 
 across the line of rivet-holes. The efficiency of this joint was 64 per 
 cent. The maximum shearing stress on the rivets was 35,200 pounds. 
 In this case the pitch could have been widened to advantage. It ap- 
 pears, therefore, that the correct spacing for ^-inch iron rivets in ^-inch 
 iron plates should be within these two limits, say i^ or i^ inches. 
 The efficiency of such a joint, if the workmanship is good, may be taken 
 at 60 per cent, of the original plate. 
 
 Steel Plates and Steel Rivets. — An open-hearth steel plate }( 
 inch thick, tensile strength 55,765 pounds per square inch ; punched 
 holes, punch 0.69 inch diameter, die 0.75 inch diameter, making a 
 hole slightly conical ; pitch of rivets 2 inches. 
 In a plate 10 inches in width, containing 5 
 rivets, failed in a single-riveted joint by tear- 
 ing one plate through the metal, as shown in 
 Fig. 24. The efficiency of this joint was 69.2 
 per cent. The first fracture which appeared 
 in sight was at the end section ; the holes in 
 the under plate elongated about -^ inch each. The lap end of the plate 
 bent about 12 degrees, similar to that indicated in Fig. 25 A. 
 
 5 
 
 Fig. 24. 
 
 [Mt^^^^ 
 
58 BOILERS AND FURNACES 
 
 Fig. 25. Fig. 25 A. 
 
 Fig. 26. 
 
 Another joint was made from the same plate ; drilled holes 0.69 
 inch diameter, i^ inch pitch ; the specimen 9.75 inches wide, contain- 
 ing 6 rivet-holes ; failed by fracturing one 
 plate through the rivet-holes, as indicated in 
 Fig. 26. The efficiency of this joint was 68.8 
 per cent. Scales started on rivet-heads, other- 
 wise they appear undisturbed. After the test 
 the rivets were loose in the plate, not frac- 
 tured. Fracture began at one edge and in 
 first section between rivets. 
 It will be observed that in both of the above tests failure occurred in 
 the plate where, by preference, it ought to fail. It appears that the 
 limit of pitch may be placed at 2 inches for ^-inch steel plates, using 
 ^-inch steel rivets. 
 
 Strength of Rivet Iron. — Tests made to determine the physical 
 properties of wrought iron commonly supplied for rivets, as well as tests 
 made of rivets after manufacture, make it appear that rivet iron varies 
 in tensile strength from 45,000 to 52,000 pounds per square inch, ex- 
 cluding a few exceptionally high specimens about which there was some 
 doubt as to the fact of their not being steel. Taking everything into 
 consideration, it appears that 47,500 pounds per square inch of section 
 is as high a tensile strength as ought to be ordinarily ascribed to wrought- 
 iron rivets. 
 
 The elastic limit of rivet iron varies from 50 to 70 per cent, of the 
 tensile strength, averaging closely to 58 per cent. There is no definite 
 line upon which elastic limit can be established for wrought iron, but the 
 percentage named is near enough for all practical purposes. 
 
 The shearing strength of rivet iron approximates closely to 72 per 
 cent, of its tensile strength, with variations on either side, diminishing 
 somewhat with increasing tensile strength. 
 
 In Table VII., 47,500 pounds tensile strength per square inch is 
 used ; so also the percentages for elastic limit and shearing strength as 
 given above, calculated for each rivet from ^ to i}i inches diameter. 
 These figures are approximately correct for commercial rivets. 
 
 Strength of Rivet Steel. — The tensile strength of open-hearth 
 steel rivets for steam boilers under the standard specifications indicate 
 that an average of 50,000 pounds per square inch would be an accept- 
 able tensile strength ; as a matter of fact, ordinary steel rivets more nearly 
 
RIVETED JOINTS 
 
 59 
 
 approach 55,000 pounds. In Table X. the tensile strength is taken at 
 52,500, which, it is believed, accords very closely with the average 
 tensile strength of steel rivets. 
 
 TABLE VII. 
 
 WROUGHT-IRON RIVETS. 
 
 -0 
 
 
 Rivets. 
 
 
 Tensile 
 Strength of 
 each Rivet 
 
 Elastic 
 
 Limit of 
 
 each Rivet 
 
 Shearing Strength of 
 
 each Rivet at 34,200 
 
 Pounds per Square 
 
 Inch 
 
 V . 
 
 
 
 
 at 47,500 
 
 at 27,550 
 
 
 
 
 
 
 Pounds per 
 
 Pounds per 
 
 
 
 
 
 
 
 
 ■H 
 
 Diameter 
 
 Diameter 
 
 Area of 
 
 Square 
 Inch. 
 
 Square 
 Inch. 
 
 Single 
 
 Double 
 
 H 
 
 of Rivet. 
 
 of Hole. 
 
 Hole. 
 
 
 
 Shear. 
 
 Shear. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Sq. Inch. 
 
 Pounds. 
 
 Pounds. 
 
 Pounds. 
 
 Pounds. 
 
 X 
 
 Y^ 
 
 H 
 
 •3712 
 
 17,632 
 
 10,227 
 
 12,695 
 
 24,120 
 
 
 \\ 
 
 
 .4418 
 
 20,986 
 
 12,150 
 
 15,110 
 
 28,709 
 
 ^ 
 
 Y 
 
 tI 
 
 .5185 
 
 24,643 
 
 14,285 
 
 17,733 
 
 33,693 
 
 T6 
 
 H 
 
 Ys 
 
 .6013 
 
 28,562 
 
 16,566 
 
 20,564 
 
 39,072 
 
 }4 
 
 ^ 
 
 if 
 
 .6903 
 
 32,749 
 
 19,018 
 
 23,608 
 
 44,855 
 
 t\ 
 
 if 
 
 I 
 
 .7854 
 
 37,307 
 
 21,638 
 
 26,861 
 
 51,036 
 
 /s 
 
 I 
 
 ItV 
 
 .8866 
 
 42,114 
 
 24,426 
 
 30,322 
 
 57,612 
 
 H 
 
 ItV 
 
 lY^ 
 
 .9940 
 
 47,215 
 
 27,385 
 
 33,995 
 
 64,591 
 
 Ya 
 
 I>^ 
 
 lA 
 
 1. 1075 
 
 52,606 
 
 30,511 
 
 37,877 
 
 71,966 
 
 The elastic limit must not be less than one-half the ultimate tensile 
 strength according to the standard specifications ; but an examination 
 of a large number of tests show that the elastic limit approximates 65 
 per cent., or 34,125 pounds per square inch. When testing steel rivet- 
 bars the same observation as in wTOUght-iron rivet-bars is had, — that 
 there is no clearly defined line of elastic limit in mild steel. 
 
 The shearing strength of steel rivets as compared with the tensile 
 strength varies considerably, from 70 to 90 per cent., but from a large 
 number of carefully conducted experiments in riveted joints and shearing 
 tests the shearing strength of open-hearth steel rivets closely approxi- 
 mated 85 per cent, of the tensile strength. 
 
 TABLE VIII. 
 
 SHOWING CHEMICAL ANALYSIS AND PHYSICAL QUALITIES OF " VICTOR" STEEL 
 
 RIVETS. 
 
 Phosphorus average 20 samples, .015 per cent. 
 
 Manganese average 20 samples, .46 percent. 
 
 Sulphur average 20 samples, .033 per cent. 
 
 Silicon average 20 samples, .005 per cent. 
 
 Carbon. average 20 samples, .11 percent. 
 
 Elastic limit average 10 samples, 36,252 pounds per sq. in. 
 
 Tensile strength average 10 samples, 51,565 pounds per sq. in. 
 
 Elongation in 8 inches . . . average 10 samples, 31.9 per cent. 
 Reduction of area .... average 10 samples, 79.7 per cent. 
 Shearing strength, single . . average 8 samples, 48,277 pounds per sq. in. 
 Shearing strength, double . average 7 samples, 45,720 pounds per sq. in. 
 
6o 
 
 BOILERS AND FURNACES 
 
 The preceding may be taken as representative of good quality steel 
 rivets now offered by rivet manufacturers to the trade. Table IX. gives 
 results of tests of rivets made for the Navy Department under a higher 
 carbon percentage than is commonly employed for rivets. 
 
 TABLE IX. 
 
 TENSILE TESTS OF OPEN-HEARTH STEEL-RIVET METAL USED IN THE CON- 
 STRUCTION OF RIVETED JOINTS, NAVY DEPARTMENT. 
 
 Diameter. 
 
 Elastic Limit per 
 
 Tensile Strength 
 
 Elongation in 
 
 Contraction 
 
 Square Inch. 
 
 per Square Inch. 
 
 Ten Inches. 
 
 of Area. 
 
 Inches. 
 
 Pounds. 
 
 Pounds. 
 
 Per cent. 
 
 Per cent. 
 
 ~\ 
 
 39,040 
 
 53,230 
 
 21.5 
 
 39-3 
 
 — 1- 
 
 39,330 
 
 55,120 
 
 21.3 
 
 40.4 
 
 -| 
 
 41,240 
 
 56,470 
 
 25-1 
 
 38.4 
 
 -■I 
 
 36,060 
 
 52,450 
 
 26.2 
 
 38.8 
 
 ItV 
 
 34,810 
 
 52,850 
 
 28.1 
 
 42.9 
 
 It\ 
 
 33,860 
 
 53,600 
 
 27.4 
 
 38.1 
 
 TABLE X.« 
 
 STEEL RIVETS. 
 
 'o 
 
 
 Rivets. 
 
 
 Tensile 
 Strength of 
 each Rivet 
 
 at 52,500 
 
 Elastic 
 
 Limit of 
 
 each Rivet 
 
 at 34,125 
 
 Shearing Strength of 
 
 each Rivet at 44,625 
 
 Pounds per Square 
 
 Inch. 
 
 
 
 
 
 Pounds per 
 
 Pounds per 
 
 
 
 
 
 
 
 SE 
 
 Diameter 
 
 Diameter 
 
 Area of 
 
 Square 
 Inch 
 
 Square 
 Inch 
 
 Single 
 
 Double 
 
 H 
 
 of Rivet. 
 
 of Hole. 
 
 Hole. 
 
 
 
 Shear. 
 
 Shear. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Sq. Inch. 
 
 Pounds. 
 
 Pounds. 
 
 Pounds. 
 
 Pounds. 
 
 X 
 
 H 
 
 \\ 
 
 •3712 
 
 19,488 
 
 12,667 
 
 16,565 
 
 31,474 
 
 
 H 
 
 H 
 
 .4418 
 
 23,195 
 
 15,076 
 
 19,715 
 
 37,459 
 
 H 
 
 Ya 
 
 if 
 
 •5185 
 
 27,221 
 
 17,694 
 
 23,138 
 
 43,962 
 
 t'^ 
 
 if 
 
 'A 
 
 .6013 
 
 31,568 
 
 20,519 
 
 26,833 
 
 50,983 
 
 'A 
 
 rs 
 
 if 
 
 .6903 
 
 36,241 
 
 23,556 
 
 30,805 
 
 58,530 
 
 y\ 
 
 it 
 
 I 
 
 •7854 
 
 41,234 
 
 26,802 
 
 35,048 
 
 
 /^ 
 
 I 
 
 ItV 
 
 .8866 
 
 46,648 
 
 30,255 
 
 39,299 
 
 74,668 
 
 li 
 
 ItV 
 
 lA 
 
 .9940 
 
 52,185 
 
 33,920 
 
 44,357 
 
 84,278 
 
 
 ^y. 
 
 ItV 
 
 1.1075 
 
 58,144 
 
 37,793 
 
 49,422 
 
 93.940 
 
 Tests and Inspection of Rivets. — In large and important con- 
 tracts, sample bars of rivet-metal i8 inches long are usually required 
 and furnished for tensile test, elastic limit, elongation and contraction of 
 
 * In the computation of this table, special steels and special tests have been 
 avoided. The values per square inch as given in the four last columns are be- 
 lieved to be those which commonly obtain in mild steel rivets. — W. M. B. 
 
RIVETED JOINTS 6l 
 
 area ; but in ordinary business routine this is quite impracticable. The 
 following tests for steel rivets may be carried out in any boiler-shop : 
 
 Fig. 27. 
 
 Fig. 28. 
 
 Fig. 29. 
 
 Two rivets to be taken at random from each keg or box of 200 
 pounds, as a portion of a lot of 10 rivets out of 1000 pounds, which 
 may be subjected to the following tests, to be made in pairs : 
 
 TABLE XI. 
 
 BOILER RIVETS. 
 
 Fig. 31 
 
 
 Button-Head, 
 
 Cone-Head, Fig. t.2. 
 
 Countersunk Head, 
 
 Diameter 
 OF Rivet. 
 
 Fig. 31. 
 
 
 
 Fig. 33- 
 
 Diameter. 
 
 Thickness. 
 
 Diameter. 
 
 Diameter. 
 
 Thickness. 
 
 Diameter. 
 
 Thickness. 
 
 A. 
 
 B. 
 
 D. 
 
 B. 
 
 c. 
 
 D. 
 
 B. 
 
 D. 
 
 Vs 
 
 IxV 
 
 tV 
 
 ItV 
 
 *- 
 
 _9_ 
 
 ItV 
 
 A 
 
 ri 
 
 1% 
 
 % 
 
 r% 
 
 h 
 
 t¥ 
 
 IT6 
 
 A 
 
 K 
 
 iX 
 
 
 ^% 
 
 h 
 
 I2 
 
 iX 
 
 ^8 
 
 if 
 
 i/s 
 
 li 
 
 1/8 
 
 
 |- 
 
 lA 
 
 if 
 
 Vs 
 
 lA 
 
 i/ 
 
 ItV 
 
 \i 
 
 A 
 
 i/s 
 
 A 
 
 it 
 
 
 H 
 
 lA 
 
 Vs 
 
 -f 
 
 IK 
 
 if 
 
 I 
 
 i>^ 
 
 H 
 
 iVs 
 
 if 
 
 ""2 
 
 ^% 
 
 % 
 
 ItV 
 
 iH 
 
 
 lU 
 
 I 
 
 3I 
 
 1% 
 
 A 
 
 1% 
 
 i^X 
 
 
 iH 
 
 ItV 
 
 -1 
 
 1% 
 
 A 
 
62 
 
 BOILERS AND FURNACES 
 
 1. Two rivets to be bent cold in the form of a hook, as shown in 
 Fig. 27, without showing cracks or flaws. 
 
 2. Two rivets to be bent hot in the form of a hook with parallel sides, 
 as shown in Fig. 28, without showing cracks or flaws. 
 
 3. Flatten the heads of 2 rivets while hot, in the manner shown in 
 Fig. 29, without cracking at the edges, the head to be flattened until its 
 diameter is 2^ times the diameter of the shank. 
 
 4. Heat 2 rivets to a low cherry-red, and quench in water at 82° 
 Fahr. ; afterwards upset them cold, and forge cold to a flattened disk 
 of ij^ diameters of shank, as shown in Fig. 30, without cracks or 
 flaws. 
 
 5. The shearing-test to consist of riveting up 2 bars of steel by a 
 single rivet, as in Fig. 38, and submitting it to a tensile strain, the 
 rivet not to shear under a stress of less than 45,000 pounds per square 
 inch. 
 
 Rivet Dimensions. — Rivets, like other manufactured articles, have 
 their proportions fixed by their manufacturers, which proportions have 
 been adapted to the needs of the trade. Table XI. gives the dimen- 
 sions of three kinds of rivet- heads in general use. These dimensions 
 may not exactly agree for all makes of rivets throughout the country, 
 but they are near enough for construction purposes and for preliminary 
 drawings. 
 
 Rivet-heads should be large enough, so that no serious distortion 
 shall occur when riveting, and especially by hand. A large head lessens 
 the distortion of the plates in tension when in single shear, and thus 
 contributes to the strength of joint. 
 
 Rivet-Points. — In hand-riveting the points are usually finished 
 conical, as in Fig. 34, which makes a good joint if the rivet is upset 
 sufl&ciently to fill the hole before the spreading of the cone is begun. 
 
 Fig. 34. 
 
 Fig. 35. 
 
 Fig. 36. 
 
 In some boiler-shops the rivet is upset to fill the hole and then finished 
 with a snap-point, as in Fig. 36, by means of a button-head set having 
 a concave depending for its size upon the diameter of the rivet ; the 
 proportions may follow the dimensions given in Table XI. It is not 
 recommended that hand-riveting be finished with snap-points for rivets 
 larger than ^ inch diameter. In machine-riveting a conical point 
 similar to Fig. 35 is in common use, so also the button-head in Fig. 36. 
 
RIVETED JOINTS 
 
 63 
 
 Fig, 
 
 It is important in machine-riveting that the work be done not too 
 rapidly, or the rivet will not upset sufficiently to fill the hole. It is also 
 important that the rivet be exactly 
 central to the machine, or bad work 
 will be done, as in Fig. 37. 
 
 Countersunk heads or points should 
 be avoided as far as possible when de- 
 signing a boiler, because of the extra 
 cost of making the hole and the lesser 
 strength of the rivet. Such a joint 
 should not be used where the stress on 
 the rivet is in the direction of its length 
 instead of shearing stress. There are 
 portions of a boiler in which it is neces- 
 sary to use a countersunk hole, — for example, around the strengthening 
 ring of a manhole and other portions of a boiler where fittings must be 
 attached which intersect a riveted joint. 
 
 Length of Rivets.— The amount of rivet-shank projecting beyond 
 the plates to be joined may be, for — 
 
 Countersunk points for 2 sheets i diameter. 
 
 Countersunk points for 3 sheets i diameter + }i inch. 
 
 Snap-points i}( diameters. 
 
 Conical points, small, hand-driven i)( diameters. 
 
 Conical points, medium, large, machine-driven . i}4 diameters. 
 
 Fig. 38. 
 
 pai7c>7ecLh|ole. 
 
 Shearing Strength of Rivets. — The shearing resistance of a rivet 
 is seriously interfered with by the compression between the rivet and 
 the bearing sides of the holes, the effect being to increase the shearing 
 resistance offered per unit of area as the size of the rivet diminishes. 
 
 Difference of quality between plate 
 and rivet, proportion of thickness 
 and diameter, and the sharpness of 
 the edges of the hole will also some- 
 what affect the shearing resistance. 
 Rivets in Single Shear.— Ex- 
 periments made upon iron and steel 
 rivets in single shear is by means 
 of 2 steel plates, as shown in 
 Fig. 38. These plates are of such 
 dimensions that no distortion occurs 
 in them and all the pull of the 
 machine is exerted in shearing the rivet. 
 
 The shearing resistance of 2 steel rivets in single shear was as 
 follows : 
 
 O 
 
64 
 
 BOILERS AND FURNACES 
 
 First rivet. — y'^g-inch steel plate ; |^-inch steel rivets. 
 
 Punched holes. — 0.753 punch ; 0.754 die. 
 
 Ultimate shearing strength, 25,750 pounds = 57,220 pounds per square inch. 
 Second rivet. — Ultimate shearing strength, 26,490 pounds = 58,870 pounds per 
 square inch. 
 
 TABLE XIL 
 
 IRON IN SINGLE AND DOUBLE SHEAR, WASHINGTON NAVY YARD. 
 
 
 Single Shear. 
 
 Double Shear. 
 
 Diameter. 
 
 Lowest, 
 Pounds. 
 
 Highest, 
 Pounds. 
 
 Mean Pounds 
 per Sq. Inch. 
 
 Lowest, 
 Pounds. 
 
 Highest, 
 Pounds. 
 
 Mean Pounds 
 per Sq. Inch. 
 
 .51 inch . . 
 .64 inch . . 
 .78 inch . . 
 .91 inch 
 1.03 inch . . 
 
 8,900 
 12,650 
 18,400 
 25,500 
 32,900 
 
 9,400 
 13,300 
 19,650 
 27,600 
 35,800 
 
 44,149 
 39,253 
 39,553 
 
 41,503 
 
 40,708 
 
 16,050 
 23,600 
 36,400 
 46,200 
 61,700 
 
 17,600 
 25,650 
 39,400 
 52,000 
 64,000 
 
 82,186 
 77,348 
 79,536 
 75,789 
 75,293 
 
 
 
 
 41,033 
 
 
 
 78,030 
 
 
 
 
 Six specimens of each size were subjected to both single and double 
 shear. The smaller diameter shows larger shearing strength per square 
 inch than larger ones, but the decrease is not regular or uniform. The 
 increase of average strength per square inch of sectional area for double 
 shear over that of single shear was, for — 
 
 Per cent. 
 
 >^-inch specimen 86.2 
 
 >^-inch specimen 97.0 
 
 ^-inch specimen loi.i 
 
 %-inch specimen 82.6 
 
 i-inch specimen 85.0 
 
 Average for all sizes 90.2 
 
 TESTS OF MILD STEEL RIVETS AND RIVET-BARS IN SINGLE 
 AND DOUBLE SHEAR, NAVY DEPARTMENT. 
 
 Nominal size of rivet, | inch ; 4 tests. 
 
 Tensile strength of rivet-bars, 60,375 pounds per square inch. 
 Final elongation, 26.48 per cent. Final area, 44.2 per cent. 
 
 Shearing tests of rivets in single shear ; 6 tests. 
 
 Diameter of hole, ^f inch. Shearing area, 0.5185 square inch. 
 Average shearing strength, 46,450 pounds per square inch. 
 Ratio of shearing to tensile strength, 76.94 per cent. 
 
 Shearing tests of rivets in double shear ; 3 tests. 
 
 Diameter of hole, \\ inch. Shearing area, 1.037 square inches. 
 Average shearing strength, 44,647 pounds per square inch. 
 Ratio of shearing to tensile strength, 73.64 per cent. 
 
 Nominal size of rivet, | inch ; 2 tests. 
 
 Tensile strength of rivet-bars, 70,120 pounds per square inch. 
 Final elongation, 24.10 per cent. Final area, 55.7 per cent. 
 
RIVETED JOINTS 
 
 65 
 
 Shearing tests of rivets in double shear ; 2 tests. 
 
 Diameter of hole, ^f inch. Shearing area, 1.380 square inches. 
 Average shearing strength, 56,100 pounds per square inch. 
 Ratio of shearing to tensile strength, 80 per cent. 
 
 Rivets in Double Shear. — It may be remarked in connection with 
 the foregoing that in testing the shearing strength of rivets under the 
 United States specifications the test- 
 
 FiG. 39. 
 
 DrtLLectlloba. 
 
 ^"5tee.l- 
 
 ^"RiVet. 
 
 O 
 
 piece consists of a double lap-joint, 
 with a single rivet, arranged as 
 shown in Fig. 39. The plate steel 
 used is in no case less than half the 
 diameter of the rivet, in order to 
 insure shearing the rivet instead of 
 the plate. The distance of the 
 nearest edge of the rivet-hole to the 
 end of the plate is not less than i^ 
 diameters of rivet. Width of plate 
 not less than three times the diame- 
 ter of the rivet. The inner end of 
 
 the filling-piece between the plates not to have an open space of more 
 than 2 inches between it and the end of the single plate. The sharp 
 edges of the rivet-hole are not to be filed down. Snap-riveted points 
 will in no case be allowed in shearing tests, but invariably the point 
 must be thoroughly and carefully worked down. The rivet-holes may 
 be either punched or drilled, as desired, and the riveting may be either 
 by hand or machine. Great care is always exercised in testing the 
 sample to insure a fair stress on the rivet. 
 
 The shearing stress of 8 samples of steel rivets in private test was 
 48,277 pounds per square inch for single shear, and 45,270 pounds per 
 square inch when subjected to double shear. For chemical and physical 
 qualities of this steel see page 59. 
 
 Friction in Riveted Joints. — Friction is at best an uncertain 
 quantity in riveted joints ; it is, therefore, seldom or never made use of 
 in calculations relating to the strength of steam boilers. But it is inter- 
 esting to know how much friction there is in a joint, because to what- 
 ever extent friction exists it adds that much to the factor of safety, pro- 
 vided a very considerable portion of it is not eliminated by springing 
 the plates apart in the operation of calking the edges to make a tight 
 joint. 
 
 Four experimental joints tested at the Watertown Arsenal from de- 
 signs similar to Fig. 40 show the friction of 5 ^-inch iron rivets in a 
 ^-inch iron plate 9^ inches wide, the holes drilled 0.69 inch diameter, 
 the corners of rivet-holes being rounded about 0.05 inch, presenting a 
 shearing area of rivets of 1.87 square inches ; this joint slipped when a 
 
66 
 
 BOILERS AND FURNACES 
 
 load of 26,000 pounds had been applied. The friction per rivet was 
 
 26,000 , „ . . , , . 
 
 — - — = 5200 pounds. Continuing the test, the rivets sheared at 70,900 
 
 pounds. The shearing strength of the rivets, 37,914 pounds per square 
 inch. 
 
 In a similar experiment with a steel plate 10 inches wide, ^ inch 
 thick, 5 f^-inch iron rivets, as above, the joint made a sudden slip when 
 
 a load of 27,000 pounds was reached ; therefore -h. = 5400 pounds 
 
 per rivet of friction. The rivets sheared at 72,700 pounds total load ; 
 the shearing strength of the rivets, 38,880 pounds per square inch. 
 
 Fig. 40. 
 
 A second series of tests was made, using a ^-inch iron plate 10 
 inches wide, with 5 f|-inch iron rivets in ^-inch drilled holes, with 
 clearance in front of rivets, corners of holes rounded about 0.05 inch ; 
 shearing area of the 5 ^-inch rivets, 2.21 square inches. Rapid slipping 
 began when the load passed 31,000 pounds. The resistance to slip- 
 ping continued at about 30,000 pounds till the total slip was 0.15 inch. 
 
 -2 J 000 
 Friction per rivet, — ' = 6200 pounds. The total load applied to 
 
 the joint was 92,000 pounds, when it failed by shearing the rivets. The 
 shearing strength of the rivets, 41,630 pounds per square inch. 
 
 A ^-inch steel plate 10 inches wide, with 5 -fl-inch iron rivets in 
 ^-inch drilled holes, as above, slipped at joint at 31,000 pounds, and 
 continued slipping until the resistance fell to 28,000 pounds, the fric- 
 tion per rivet being — = 6200 pounds. The ultimate strength of 
 
 the joint was reached at 88,000 pounds, when the rivet sheared. Shear- 
 ing strength of rivets, 39,820 pounds per square inch. 
 
 RESULTS OF TESTS OF RIVETED JOINTS. 
 
 Investigations into the strength of riveted joints have been until a 
 few years past mathematical rather than experimental, as no machines 
 were in use capable of pulling wide joints or large sectional areas, as 
 
RIVETED JOINTS 
 
 67 
 
 now made possible by the installation of large testing-machines at the 
 principal steel-works, and especially at the Watertown Arsenal. The 
 tests made upon iron and steel bars, plates, and riveted joints for the 
 United States Navy Department, which extended over several years at 
 the Watertown Arsenal, are the most extended and comprehensive ever 
 undertaken in this country. As these tests were undertaken for the 
 purpose of ascertaining how riveted joints fail, a number of engravings 
 are here introduced, copied from the records of the arsenal, which, in 
 connection with the tabulated data, give much valuable information re- 
 garding this important detail in steam-boiler construction. 
 
 Fig. 41. 
 
 Fig. 42. 
 
 Fig. 43. 
 
 CJ 
 
 A riveted joint may fail by (i) the shearing of the rivets (Figs. 41, 
 42, 43), in which the rivet area and shearing resistance are too small 
 for the thickness of plate and pitch of rivets ; to remedy which a larger 
 rivet may be introduced, or, in the case of iron rivets, the substitution 
 of steel rivets, the latter having a shearing strength more than 20 per 
 
 Fig. 44. 
 
 cent, greater than iron rivets. The engraving (Fig. 44) is a typical 
 illustration of the elongation of rivet-holes where the rivets were sheared 
 by the plate. This illustration was prepared from a tracing drawn over 
 a photograph of a joint in a ^-inch steel plate with 41^ rivets on 
 
68 
 
 BOILERS AND FURNACES 
 
 3^-inch pitch. The elongations of rivet-holes were 0.64, 0.68, 0.73, 
 
 and 0.83 decimal parts of an inch respectively. 
 
 (2) By crushing and tearing out the plate in front of the rivet (Figs. 
 
 45 and 46). This method of failure depends usually upon the thickness 
 
 of plate relative to di- 
 Fi<^- 45- ameter of rivet. Such 
 
 a fracture suggests that 
 smaller rivets be used 
 if the distance from the 
 edge of the plate to the 
 side of the hole be not 
 less than the diameter 
 of the hole. The en- 
 graving (Fig. 46) was 
 made from a tracing 
 drawn over a photo- 
 graph of a steel plate 
 -inch pitch. The plate tore out 
 
 Y-z inch thick with 5 i>^ rivets on 
 in front of the rivets as shown. 
 
 (3) By tearing the plate along the line of rivet-holes (Fig. 47). 
 This fracture depends upon the thickness of the plate and the pitch ot 
 the rivets. The resistance to this mode of fracture equals the effective 
 
 Fig. 46. 
 
 sectional area of plate between holes multiplied by its tensile strength 
 per square inch. The engraving (Fig. 47) was made from a tracing 
 drawn over a photograph of a steel plate y\ inch thick with i-inch 
 drilled holes on 3^-inch pitch. The efficiency of this joint was 80.1 
 per cent. 
 
 A riveted joint should fail preferably by fracturing the plate along 
 the line of rivets from hole to hole ; if failure should occur by either of 
 the first two methods it indicates an excess of strength in the net section 
 of the plate through the line of rivet-holes which has not been made use 
 
RIVETED JOINTS 
 
 69 
 
 of ; if, on the other hand, fracture occurs along the line of rivet-holes in 
 a well-proportioned joint, it is immaterial whether there is an excess of 
 strength in other directions or not. 
 
 Fig 
 
 It was observed during the tests of riveted joints at the Watertown 
 Arsenal that the failure of a joint was generally marked by three well- 
 defined periods. In the first period the greatest rigidity was found, and 
 it was thought that the joint was then held entirely by friction of the rivet- 
 heads, and the movement of the joint was principally that due to the 
 elasticity of the metal. 
 
 The second period was distinguished by a rapid increase in the 
 stretch of the joint, attributed to the overcoming of the friction under 
 the rivet-heads and closing up any clearance about the rivets, bringing 
 them into bearing condition against the fronts of the rivet-holes. Rivets 
 which are said to fill the holes can hardly do so completely, on account 
 of the contraction of the metal of the rivet from a higher temperature 
 than that of the plate after the rivet is driven. 
 
 After a brief interval the movement of the joint was retarded, and 
 the third period was reached. The strength of the joint was then be- 
 lieved to be due to the distortion of the rivet-holes and of the rivets 
 themselves. The movement began slowly, and so continued until the 
 elastic limit of the metal about the rivet-holes was passed, and general 
 flow took place over the entire cross-section, and rupture was reached. 
 
 These stages in the test of a joint were well defined except when the 
 plates were in a warped condition initially, when abnormal micrometer 
 readings were observed. The difference in behavior of a joint and the 
 solid metal suggested the propriety of arranging tension-joints in boiler 
 construction and elsewhere as nearly in line as possible. 
 
^o 
 
 BOILERS AND FURNACES 
 
 ENGRAVINGS BELONGING TO TABLE XIII. 
 Fig. 48. Fig. 49. 
 
 Fig. 50. 
 
 Fig. 51. 
 
 C1H3=^=^=0^ 
 
 Fig. 53. 
 
 Fig. 52. 
 
 ^ c 
 
 
 Fig. 54. 
 
 -y^.fxj//--"- 
 
RIVETED JOINTS 
 
 ENGRAVINGS BELONGING TO TABLE yHW.— Continued. 
 Fig. 55. 
 
 Upper ptaie 
 
 
 Jjiwer- plate 
 
 Fig. 56. 
 
 ENGRAVINGS BELONGING TO TABLE XIV. 
 Fig. 57- 
 
 Fig. 58. 
 
 Fig. 59. 
 
 @ e © © 
 
 M.OS'kJOt-- 
 
72 
 
 BOILERS AND FURNACES 
 
 TABLE XV. 
 
 TESTS OF TRIPLE-RIVETED LAP-JOINTS, jVlNCH IRON PLATES, WATERTOWN 
 ARSENAL. 
 
 Fig. 6o. Fig. 6i. 
 
 O 
 
 O i) „Q 
 
 O Q ^ O 
 /j'fdi 
 
 ^^3 
 
 '=Q::=Q=Q:=^ 
 
 
 t-^i"-^ 
 
 
 Material of plate 
 
 Thickness of plate, nominal, inch 
 
 Width of test-specimen, inches 
 
 Diameter of rivets, inch 
 
 Material of rivets 
 
 Diameter of holes, punched or drilled, inches 
 
 Number of rivets 
 
 Pitch of rivets, C, Fig. 84, inches 
 
 Pitch of rivets, E., Fig. 84, inches 
 
 Chain or zigzag riveting 
 
 Bearing surface of rivets, square inches 
 
 Shearing area of rivets, square inches 
 
 Tensile strength of plate, pounds per square inch . . . 
 
 Gross sectional area, square inches 
 
 Net sectional area, square inches 
 
 Maximum stress on joint per square inch : 
 
 Tension on gross section of plate, pounds . . . . 
 
 Tension on net section of plate, pounds 
 
 Compression on bearing surface of rivets, pounds 
 
 Shearing on rivets, pounds 
 
 Efficiency of joint, per cent 
 
 Fracture, similar to Figures 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 t'b 
 
 /f 
 
 TS 
 
 13-74 
 
 ^^•?5 
 
 12.78 
 
 \i 
 
 ii 
 
 if 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 %D. 
 
 KD. 
 
 /sD. 
 
 15 
 
 15 
 
 12 
 
 2% 
 2% 
 
 3 
 
 i% 
 
 2% 
 
 2% 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 5.66 
 
 5-41 
 
 4-54 
 
 9.02 
 
 9.02 
 
 7.21 
 
 59,000 
 
 52,910 
 
 58,090 
 
 5-93 
 
 6.20 
 
 5-52 
 
 4.04 
 
 4.40 
 
 4.01 
 
 45.720 
 
 48,710 
 
 48,040 
 
 67,100 
 
 68,630 
 
 66,130 
 
 47,900 
 
 55,820 
 
 58,410 
 
 30,060 
 
 33,480 
 
 36,780 
 
 77-5 
 
 92.1 
 
 82.7 
 
 60 
 
 60 
 
 60 
 
 Steel. 
 
 Iron. 
 
 14 D. 
 
 12 
 
 Chain. 
 4.60 
 7.21 
 
 59,390 
 
 4-38 
 
 46,430 
 62,650 
 59,650 
 38,060 
 78.2 
 61 
 
 Butt-joints. — It was experimentally shown at the Watertown 
 Arsenal that the behavior of joints in different thicknesses of steel 
 
 plates, single riveted, 
 ^^*^- ^^- as in Fig. 62, was 
 
 substantially the same 
 whether ^^-inch or 
 ^-inch plates were 
 used. 
 
 It will not be un- 
 derstood from this, 
 however, that, as 
 a consequence, the 
 same efficiency may 
 be obtained in differ- 
 ent thicknesses of plate for single-riveted work, because certain essential 
 conditions change as we approach the stronger joints in different thick- 
 nesses of plate. 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 

 TOWN 
 
 A.RSENAL. 
 
 
 
 
 
 
 
 Iron. 
 
 Iron. 
 
 Steel. 
 
 Steel. 
 
 Iron. 
 
 steel. 
 
 Iron. 
 
 
 Material of plate 
 
 Steel. 
 
 Thickness of plate, nominal, i 
 
 ^ 
 
 Vt. 
 
 ^ 
 
 Vi 
 
 Yi 
 
 Yi 
 
 K 
 
 K 
 
 Width of test-specimen, inche 
 
 10.03 
 
 10.02 
 
 10.08 
 
 10.03 
 
 10.5 
 
 9-5 
 
 12.02 
 
 10 
 
 Diameter of rivets, inches . 
 
 y^ 
 
 % 
 
 Y^ 
 
 if 
 
 " I 
 
 I 
 
 1/8 
 
 lJ/8 
 
 Material of rivets 
 
 Iron. 
 
 Steel. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Steel. 
 
 Iron. 
 
 Steel. 
 
 Diameter of holes, punched o 
 
 IIP. 
 
 11 D. 
 
 if P. 
 
 I P. 
 
 lAP. 
 
 It's P. 
 
 iT^P. 
 
 IT^ P. 
 
 Number of rivets 
 
 5 
 
 5 
 
 5 
 
 5 
 
 4 
 
 * 
 
 4 
 
 4 
 
 Pitch of rivets, inches . . . 
 
 2 
 
 2 
 
 2 
 
 2 
 
 2% 
 
 2^3 
 
 3 
 
 ^% 
 
 Bearing surface of rivets, squs 
 
 2.18 
 
 2.06 
 
 2.13 
 
 2-55 
 
 2.65 
 
 2.72 
 
 3.44 
 
 3.64 
 
 Shearing area of rivets, squar 
 
 2.64 
 
 2.58 
 
 2.64 
 
 3-93 
 
 3-55 
 
 3-55 
 
 4.43 
 
 4-43 
 
 Tensile strength of plate, pou 
 
 44,615 
 
 44,615 
 
 57,215 
 
 57,215 
 
 44,635 
 
 52,445 
 
 46,590 
 
 51,545 
 
 Gross sectional area, square it 
 
 5-10 
 
 5-10 
 
 4-97 
 
 4-97 
 
 6.28 
 
 5.83 
 
 8.38 
 
 7.38 
 
 Net sectional area, square inct 
 
 2.91 
 
 3-04 
 
 2.84 
 
 2.42 
 
 3.62 
 
 3-II 
 
 4-93 
 
 3-73 
 
 Maximum stress on joint per i 
 
 
 
 
 
 
 
 
 
 Tension gross section of 
 
 17,760 
 
 24,200 
 
 21,830 
 
 24,145 
 
 19,750 
 
 26,480 
 
 17,230 
 
 23,940 
 
 Tension net section of p 
 
 31,100 
 
 40,590 
 
 38,204 
 
 49,590 
 
 34,230 
 
 49,650 
 
 29,290 
 
 47,370 
 
 Compression on bearing 
 
 41,500 
 
 59,900 
 
 50,940 
 
 47,060 
 
 46,790 
 
 56,760 
 
 41,980 
 
 48.540 
 
 Shearing on rivets, pounds . 
 
 34,280 
 
 47,830 
 
 41,100 
 
 30,540 
 
 34,930 
 
 43,490 
 
 32,600 
 
 39,890 
 
 Efficiency of joint, per cent. 
 
 39-8 
 
 54-2 
 
 38.2 
 
 51.2 
 
 42 
 
 50.5 
 
 37 
 
 46.4 
 
 Fracture, similar to Figures 
 
 48 
 
 55 
 
 53 
 
 56 
 
 48 
 
 53 
 
 48 
 
 48 
 
 
 TOWN 
 
 ARSENi 
 
 VL. 
 
 
 
 
 
 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Iron. 
 
 Iron. 
 
 
 
 Steel. 
 
 Material of plate 
 
 
 
 
 
 
 
 
 
 Thickness of plate, nominal. 
 
 rs 
 
 A 
 
 fs 
 
 ^ 
 
 Yz 
 
 r8 
 
 Y^ 
 
 M 
 
 Width of test-specimen, inche 
 
 14.36 
 
 15.01 
 
 13-77 
 
 10 
 
 10.5 
 
 10.5 
 
 12 
 
 9.98 
 
 Diameter of rivets, inches . . 
 
 \% 
 
 if 
 
 M 
 
 ii 
 
 I 
 
 I 
 
 rj^ 
 
 1% 
 
 Material of rivets 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 steel. 
 
 Iron. 
 
 Iron. 
 
 Steel. 
 
 Diameter of holes, punched o 
 
 iD. 
 
 rsD. 
 
 rsD. 
 
 I P. 
 
 It's P. 
 
 lA P. 
 
 ife P. 
 
 ife P. 
 
 Number of rivets 
 
 10 
 
 12 
 
 10 
 
 10 
 
 8 
 
 8 
 
 8 
 
 8 
 
 Pitch of rivets, C, Fig. 82, in 
 
 2/8 
 
 ^% 
 
 ^% 
 
 2 
 
 w^ 
 
 m 
 
 3 
 
 2% 
 
 Pitch of rivets, E., Fig. 82, in 
 
 ^% 
 
 2% 
 
 2^8 
 
 2 
 
 ^% 
 
 2% 
 
 2% 
 
 iM 
 
 Chain or zigzag riveting . . 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Bearing surface of rivets, squ 
 
 3-7° 
 
 4.29 
 
 3-69 
 
 5.26 
 
 5-43 
 
 5.48 
 
 6.92 
 
 7-30 
 
 Shearing area of rivets, squa 
 
 7.85 
 
 7.21 
 
 6.01 
 
 7-85 
 
 7.09 
 
 7.09 
 
 8.86 
 
 8.86 
 
 Tensile strength of plate, poi 
 
 56,670 
 
 52,910 
 
 59,000 
 
 57,215 
 
 52,445 
 
 44,635 
 
 46,500 
 
 51,545 
 
 Gross sectional area, square i 
 
 5-31 
 
 6.14 
 
 5.81 
 
 5.02 
 
 5.81 
 
 6.48 
 
 8.41 
 
 7-39 
 
 Net sectional area, square inc 
 
 346 
 
 3-99 
 
 3-96 
 
 2.39 
 
 3.10 
 
 3-74 
 
 4-95 
 
 3-74 
 
 Maximum stress on joint per 
 
 
 
 
 
 
 
 
 
 Tension gross section 
 
 43,050 
 
 43,530 
 
 38,850 
 
 30,757 
 
 35,800 
 
 25,150 
 
 24,720 
 
 24,050 
 
 Tension net section of j 
 
 66,070 
 
 66,990 
 
 56,990 
 
 64,602 
 
 67,100 
 
 43,580 
 
 42,000 
 
 47,510 
 
 Compression on bearing 
 
 61,780 
 
 62,310 
 
 61,170 
 
 29,35+ 
 
 38,300 
 
 29,740 
 
 30,040 
 
 24,340 
 
 Shearing on rivets, pounds 
 
 29,120 
 
 37,070 
 
 37,550 
 
 19,670 
 
 29,340 
 
 22,990 
 
 23,460 
 
 20,060 
 
 Efficiency of joint, per cent. 
 
 76.0 
 
 82.3 
 
 65.8 
 
 53-8 
 
 68.3 
 
 56.3 
 
 53-1 
 
 46.7 
 
 Fracture, similar to Figures 
 
 59 
 
 58 
 
 58 
 
 59 
 
 59 
 
 59 
 
 59 
 
 59 
 
 
 
 
 
 
 
 
 
 
 
 
TABLE XIIL 
 
 TESTS OF SINGLE-RIVETED LAP-JOINTS, ^-INCH TO ^-INCH IRON AND STEEL PLATES, WATERTOWN ARSENAL. 
 
 Material of plate 
 
 Thickness of plate, nominal, inch 
 
 Width of test-specimen, inches 
 
 Diameter of rivets, inches 
 
 Material of rivets 
 
 Diameter of holes, puncher! or drilled, inches 
 
 Number of rivets 
 
 Pitch of rivets, inches 
 
 Bearing surface of rivets, square inches 
 
 Shearing area of rivets, square inches 
 
 Tensile strength of plate, pounds per square inch . . . 
 
 Gross sectional area, square inches 
 
 Net sectional area, square inches 
 
 Maximum stress on joint per square inch : 
 
 Tension gross section of plate, pounds 
 
 Tension net section of plate, pounds 
 
 Compression on bearing surface of rivets, pounds 
 
 Shearing on rivets, pounds 
 
 Efficiency of joint, per cent 
 
 Fracture, similar to Figures 
 
 Iron. 
 
 9-75 
 
 % 
 
 Iron. 
 
 .695 P. 
 
 6 
 
 47,925 
 2.65 
 
 Iron. 
 
 .695 p. 
 
 5 
 2 
 
 .875 
 1.92 
 47,925 
 2.52 
 1.63 
 
 29,444 
 45.520 
 82,910 
 38,640 
 64.1 
 49 
 
 Iron. 
 
 .69 D. 
 5 
 
 1.87 
 47,925 
 2.64 
 
 41,790 
 79,360 
 38,660 
 
 Steel. 
 K 
 9-75 
 
 Steel. 
 .69 D. 
 
 2.24 
 55,765 
 2.44 
 1.40 
 
 38,360 
 66,860 
 90,000 
 41,790 
 
 Steel. 
 
 Steel. 
 .69 P. 
 
 107,260 
 49,270 
 69.2 
 
 Steel. 
 
 U 
 
 10 
 
 Vs 
 Steel. 
 .69 D. 
 
 .873 
 1.87 
 55,765 
 2.53 
 
 41 ,460 
 63,190 
 120,160 
 56,100 
 74-3 
 
 Steel. 
 K 
 14.5 
 
 lA 
 
 Iron. 
 iK D. 
 
 4 
 
 3^8 
 
 1.24 
 4.9X 
 61 ,470 
 
 63,140 
 
 120,180 
 
 30,350 
 
 67.3 
 
 Iron. 
 
 3/8 
 10 
 ih 
 
 Iron. 
 
 .75 P. 
 
 5 
 
 1.44 
 2.27 
 47,180 
 3.76 
 2.32 
 
 23,110 
 37,460 
 60,340 
 
 Iron. 
 
 H 
 10.33 
 
 lAP. 
 
 5 
 
 2,^ 
 2.19 
 4-43 
 47,180 
 4.07 
 
 61,700 
 52,970 
 
 Steel. 
 
 n 
 9.98 
 
 Iron. 
 
 .75 P. 
 
 5 
 
 23,240 
 37,700 
 60,760 
 
 Steel. 
 
 H 
 10.5 
 ih 
 
 Steel. 
 
 .77 D. 
 
 6 
 
 1% 
 
 53,330 
 3-97 
 2.22 
 
 35„590 
 63,650 
 80,930 
 50,650 
 66.7 
 
 Steel. 
 
 3/8 
 11.84 
 
 a 
 
 Iron. 
 I D. 
 5 
 
 2H 
 
 1.92 
 
 3-93 
 58,340 
 
 33,050 
 57,190 
 78,330 
 38,270 
 56.6 
 
 53 
 
 Steel. 
 
 Iron. 
 
 iiP. 
 
 5 
 
 2.64 
 44,615 
 5.10 
 2.91 
 
 17,760 
 31,100 
 41,500 
 
 Iron. 
 
 % 
 10.02 
 
 steel. 
 iSD. 
 
 44,615 
 5.10 
 3.04 
 
 47,830 
 54-2 
 55 
 
 Steel. 
 
 10.08 
 
 Iron. 
 iiP. 
 
 2.13 
 
 2.64 
 57,215 
 4.97 
 2.84 
 
 21,830 
 38,204 
 50,940 
 41,100 
 38.2 
 53 
 
 Steel. 
 
 % 
 10.03 
 
 it 
 Iron. 
 
 2.55 
 3.93 
 57,215 
 4.97 
 2.42 
 
 24,145 
 49,590 
 
 Iron. 
 10.5 
 
 2K8 
 2.65 
 3.55 
 44,635 
 6.28 
 3.62 
 
 19,750 
 34,230 
 46,790 
 34,930 
 42 
 
 Steel. 
 'A 
 
 Steel. 
 
 It's P. 
 4 
 
 2/8 
 2.72 
 3.55 
 
 52,445 
 5.83 
 3." 
 
 26,480 
 49,650 
 56,760 
 43,490 
 
 50.5 
 
 53 
 
 Iron. 
 
 12.02 
 i/a 
 Iron. 
 ■A P. 
 
 3.44 
 4.43 
 46,590 
 
 17,230 
 29,290 
 
 32,600 
 
 37 
 
 TABLE XIV. 
 
 TESTS OF DOUBLE-RIVETED LAP-JOINTS, ^-INCH TO ^-INCH IRON AND STEEL PLATES, WATERTOWN ARSENAL. 
 
 Material of plate 
 
 Thickness of plate, nominal, inch 
 
 Width of test-specimen, inches 
 
 Diameter of rivets, inches 
 
 Material of rivets 
 
 Diameter of holes, punched or drilled, inches 
 
 Number of rivets 
 
 Pitch of rivets, C, Fig. 82, inches 
 
 Pitch of rivets, E., Fig. 82, inches 
 
 Chain or zigzag riveting 
 
 Bearing surface of rivets, square inches 
 
 Shearing area of rivets, square inches 
 
 Tensile strength of plate, pounds per square inch . . . 
 
 Gross sectional area, square inches 
 
 Net sectional area, square inches 
 
 Maximum stress on joint per square inch : 
 
 Tension gross section of plate, pounds 
 
 Tension net section of plate, pounds 
 
 Compression on bearing surface of rivets, pound; 
 
 Shearing on rivets, pounds 
 
 Efliciency of joint, per cent 
 
 Fracture, similar to Figures . ; 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Iron. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Iron. 
 
 Iron. 
 
 K 
 
 H 
 
 y* 
 
 y 
 
 A 
 
 A 
 
 Vs 
 
 Vs 
 
 /8 
 
 /a 
 
 }i 
 
 n 
 
 A 
 
 A 
 
 y 
 
 H 
 
 H 
 
 M 
 
 10 
 
 10 
 
 10.68 
 
 14.40 
 
 14.02 
 
 14.02 
 
 12.98 
 
 9-97 
 
 10 
 
 12.97 
 
 11.83 
 
 14.36 
 
 15.01 
 
 13-77 
 
 10 
 
 10.5 
 
 10.5 
 
 12 
 
 Vs 
 
 Vs 
 
 a 
 
 a 
 
 i§ 
 
 1§ 
 
 a 
 
 U 
 
 ik 
 
 « 
 
 a 
 
 a 
 
 il 
 
 li 
 
 n 
 
 I 
 
 I 
 
 i/a 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Steel. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Iron. 
 
 Steel. 
 
 Iron. 
 
 Iron. 
 
 .69 P. 
 
 .69 P. 
 
 /sD. 
 
 /aD. 
 
 I P. 
 
 I D. 
 
 KD. 
 
 %P- 
 
 KP. 
 
 KD. 
 
 iD. 
 
 iD. 
 
 /sD. 
 
 VsD. 
 
 I P. 
 
 lAP. 
 
 lAP. 
 
 lAP. 
 
 10 
 
 9 
 
 10 
 
 10 
 
 7 
 
 8 
 
 7 
 
 10 
 
 9 
 
 7 
 
 10 
 
 10 
 
 12 
 
 10 
 
 10 
 
 8 
 
 8 
 
 8 
 
 2 
 
 2 
 
 2/8 
 
 2/s 
 
 3% 
 
 3y 
 
 sy 
 
 2 
 
 2 
 
 3y 
 
 2% 
 
 2^8 
 
 2% 
 
 2% 
 
 2 
 
 2/8 
 
 2% 
 
 3 
 
 iK 
 
 1% 
 
 2/3 
 
 2}i 
 
 iK 
 
 2y 
 
 lys 
 
 iK 
 
 iK 
 
 1^8 
 
 2K 
 
 2% 
 
 2% 
 
 2/8 
 
 2 
 
 2y 
 
 2y 
 
 2% 
 
 Chain. 
 
 Zigzag. 
 
 Chain. 
 
 Chain. 
 
 Zigzag. 
 
 Chain. 
 
 Zigzag. 
 
 Chain. 
 
 Zigzag. 
 
 Zigzag. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 Chain. 
 
 1.79 
 
 X.63 
 
 2.25 
 
 2.17 
 
 2.20 
 
 2.44 
 
 2.06 
 
 2.85 
 
 2.49 
 
 1-93 
 
 3.85 
 
 3.70 
 
 4.29 
 
 3.69 
 
 5.26 
 
 5.43 
 
 5.48 
 
 6.92 
 
 3.85 
 
 3.46 
 
 6.01 
 
 6.01 
 
 5.50 
 
 6.28 
 
 3.08 
 
 4.54 
 
 4.08 
 
 3.08 
 
 7.85 
 
 7-85 
 
 7.21 
 
 6.01 
 
 7.85 
 
 7.09 
 
 7.09 
 
 8.86 
 
 55,765 
 
 55,765 
 
 61,000 
 
 61,470 
 
 59,300 
 
 56,760 
 
 47,180 
 
 53,330 
 
 53,330 
 
 53,330 
 
 53,730 
 
 56,670 
 
 52,910 
 
 59,000 
 
 57,215 
 
 52,445 
 
 44,635 
 
 46,500 
 
 2.49 
 
 2.52 
 
 2.74 
 
 3.57 
 
 4.28 
 
 4.28 
 
 5.09 
 
 3-73 
 
 363 
 
 4.75 
 
 4-55 
 
 5.31 
 
 6.14 
 
 5-81 
 
 5.02 
 
 5.81 
 
 6.48 
 
 8.41 
 
 1.59 
 
 1.61 
 
 1.62 
 
 2.49 
 
 3.02 
 
 3.06 
 
 3.91 
 
 2.31 
 
 2.24 
 
 3.65 
 
 2.63 
 
 3.46 
 
 3-99 
 
 3.96 
 
 2.39 
 
 3.10 
 
 3-74 
 
 4.95 
 
 39,280 
 
 41,785 
 
 42,770 
 
 47,870 
 
 40,630 
 
 46,070 
 
 23,750 
 
 40,080 
 
 39,010 
 
 34,316 
 
 38,790 
 
 43,050 
 
 43,530 
 
 38,850 
 
 30,757 
 
 35,800 
 
 25,150 
 
 24,720 
 
 61,510 
 
 65,400 
 
 72,350 
 
 68,630 
 
 57,580 
 
 64,440 
 
 30,920 
 
 64,720 
 
 63,210 
 
 44,660 
 
 67,100 
 
 66,070 
 
 66,990 
 
 56,990 
 
 64,602 
 
 67,100 
 
 43,580 
 
 42,000 
 
 54,640 
 
 64,600 
 
 52,090 
 
 78,760 
 
 79,050 
 
 80,820 
 
 58,700 
 
 52,450 
 
 56,860 
 
 84,460 
 
 45,840 
 
 61,780 
 
 62,310 
 
 61,170 
 
 29,354 
 
 38,300 
 
 29,740 
 
 30,040 
 
 25,400 
 
 30,430 
 
 19,500 
 
 28,440 
 
 31,620 
 
 31,400 
 
 39,130 
 
 32,930 
 
 34,710 
 
 52,750 
 
 22,480 
 
 29,120 
 
 37,070 
 
 37,550 
 
 19,670 
 
 29,340 
 
 22,990 
 
 23,460 
 
 70.4 
 
 74-9 
 
 70.1 
 
 77.9 
 
 68.5 
 
 80.1 
 
 S0.4 
 
 75-2 
 
 73.2 
 
 64.4 
 
 72.2 
 
 75.0 
 
 82.3 
 
 65.8 
 
 53.8 
 
 68.3 
 
 56.3 
 
 53-1 
 
 59 
 
 59 
 
 59 
 
 59 
 
 57 
 
 59 
 
 58 
 
 59 
 
 59 
 
 58 
 
 59 
 
 59 
 
 58 
 
 58 
 
 59 
 
 59 
 
 59 
 
 59 
 
RIVETED JOINTS 
 
 73 
 
 If the strength per unit of metal of the net section was constant, it 
 would be a very simple matter to compute the efficiency of any joint, 
 as it would merely be the ratio of the net to the gross area of the 
 plates. 
 
 The tenacity of the net section was observed to vary, and this 
 variation extended over wide limits, as shown in the accompanying 
 tables. 
 
 The efficiencies shown in Table XVII. are obtained by dividing the 
 tensile stress on the gross area of plate by the tensile strength as repre- 
 sented by the tensile test-strip, stating the values in per cent, of the 
 latter. 
 
 This table is valuable as showing at once the efficiencies of different 
 joints wherein the pitch of the rivets and their diameter vary. 
 
 It is seen there is considerable latitude allowed in the choice of rivets 
 and pitch without materially changing the efficiency of the joint ; thus 
 in ^-inch plate, — 
 
 )^-inch rivets (driven), i^-inch pitch, 72.4 per cent, efficiency,* 
 ^-inch rivets (driven), 2X-inch pitch, 73.3 per cent, efficiency, 
 ^-inch rivets (driven), 2^-inch pitch, 71.5 per cent, efficiency, 
 i-inch rivets (driven), 2^-inch pitch, 70.3 per cent, efficiency, 
 i-inch rivets (driven), 2^-inch pitch, 73.8 per cent, efficiency, 
 
 gave nearly the same results. 
 
 In these examples the ratios of net to gross areas of plate range from 
 60 to 67 per cent., while the rivet areas range from 0.3067 square inch 
 to 0.7854 square inch. The actual areas of net sections of plate and 
 rivets are as follows : 
 
 
 fs-inch Rivets. 
 
 %-inch Rivets. 
 
 %-inch Rivets. 
 
 I-inch Rivets. 
 
 Rivets 
 
 Plate 
 
 Square Inch. 
 
 .3067 
 
 1.486 
 
 Square Inch. 
 .4418 
 2.207 
 
 Square Inch. 
 .6013 
 2.232 
 
 Square Inch. 
 
 •7854 
 12.259 
 
 L 2.319 
 
 The areas of the rivets stand to each other as the following numbers 
 100 144 196 256 
 
 and the net areas of the plate to each other as 
 
 149 
 
 150 
 
 {; 
 
 From these illustrations it appears that to obtain the same degree of 
 efficiency in this quality of metal, although that efficiency is probably 
 
 * The difference in per cent, shown in this summary over that of Table XVII. 
 is explained by the fact that this summary covers a larger number of joints result- 
 ing as above. 
 
 6 
 
74 
 
 BOILERS AND FURNACES 
 
 not the highest attainable, a fixed ratio between rivet-metal and net 
 section of plate is not essential. 
 
 In j^-inch plate with ^-inch rivets the efficiencies of the joints are 
 nearly constant over the ranges of pitches tested. 
 
 The efficiencies and the ratio of net to gross areas of plate are as 
 follows : 
 
 PITCH OF RIVETS. 
 
 
 1% inches. 
 
 2 inches. 
 
 2j4 inches. 
 
 2% inches. 
 
 Efficiency 
 
 Ratio of areas . . . 
 
 Per cent. 
 64-5 
 53.4 
 
 Per cent. 
 66.3 
 56.3 
 
 Per cent. 
 66.3 
 58.9 
 
 Per cent. 
 66.4 
 61.I 
 
 In this we have illustrated a case which, in passing from the widest 
 pitch, having 61. i per cent, of the solid plate left, to the narrowest pitch, 
 which had 53.4 per cent, of the solid plate, the gain or excess in strength 
 in the net section almost exactly compensated for the loss of metal. 
 
 Table XVIII. exhibits the differences between the efficiencies of the 
 joints and the ratios of net to gross areas of plate. If the tenacity of 
 net section remained constant per unit of area, the efficiencies of Table 
 XVII. would, as above explained, be identical with the ratios of net to 
 gross areas of plate, and the values in this table reduced to zero. 
 
 Table XIX. shows the excess in strength of the net section of the 
 joint over the strength of the tensile test-strip in per cent, of the latter. 
 
 In this table the average of all the joints shows the highest per cent, 
 of excess of strength in the narrowest pitch, and a tendency to lose 
 this excess as the pitch increases. 
 
 The maximum gain in strength on the net section was 21.2 per cent., 
 the minimum value 2.5 per cent, of the tensile test-strip. On other 
 forms of joints and with punched holes in both iron and steel plate, 
 illustrations are numerous in which there have been large deficiencies, 
 the metal of the net section falling far below the strength of the plate. 
 
 It is believed to have been amply shown that increasing the net width 
 diminishes the apparent tenacity of the plate, although other influences 
 may tend to counteract this tendency in some joints. 
 
 In order to compare the excess of strength of one thickness of plate 
 with another having the same net widths, we have Table XVI. , reject- 
 ing those joints that failed otherwise than along the line of riveting in 
 making the averages. 
 
 The excess in strength is generally well maintained in each of the 
 several thicknesses, and were it possible to retain the same ratio of net 
 to gross areas of plate, and at the same time equal net widths between 
 rivets, it would seem from this point of view feasible to obtain the same 
 degree of efficiency in thick as in thin plates. 
 
RIVETED JOINTS 
 TABLE XVI. 
 
 75 
 
 SHOWING EXCESS OF STRENGTH OF PLATE OF DIFFERENT THICKNESSES AND 
 PITCH OF RIVETS. 
 
 
 Width of Plate between Rivet-Holes. 
 
 Plate. 
 
 i" 
 
 ^V^" 
 
 I IX" 
 
 1^8" 
 
 iVJ' 
 
 i^s" 
 
 iJi" 
 
 ^y^" 
 
 2" 
 
 Inch. 
 
 i 
 
 Per 
 cent. 
 
 16.7 
 18.4 
 16,7 
 17.7 
 II.4 
 
 Per 
 cent. 
 
 12.6 
 
 13.7 
 
 14-3 
 16.3 
 15- 1 
 
 Per 
 cent. 
 
 II.4 
 12.7 
 
 9-3 
 
 14.2 
 13-8 
 
 Per 
 cent. 
 12.0 
 
 13-5 
 10.7 
 
 14-5 
 14. 1 
 
 Per 
 cent. 
 
 13-4 
 14.6 
 
 14.6 
 7.6 
 
 Per 
 cent. 
 
 8.9 
 12.9 
 
 8.8 
 12.7 
 II. 8 
 
 Per 
 cent. 
 
 11-5 
 9.0 
 8.2 
 9-9 
 
 lO.O 
 
 Per 
 cent. 
 
 I3-I 
 13-6 
 12.2 
 9.8 
 10. 1 
 
 Per 
 cent. 
 10.6 
 
 3-5 
 
 Average of all 
 thicknesses . . 
 
 16.2 
 
 14.4 
 
 12.3 
 
 12.9 
 
 11.9 
 
 II. 
 
 9-7 
 
 11.8 
 
 7.0 
 
 The following causes, however, tend to prevent this consummation : 
 
 For equal net widths thick plates require larger rivets to avoid shear- 
 ing than thin ones, the diameters of the rivets being somewhat increased 
 for this cause ; and, again, it has become necessary to increase the metal 
 of the net section in order to retain a suitable ratio of net to gross areas 
 of plate. There results from these considerations such an increase in 
 net width of plate that the excess in strength displayed by narrower 
 sections is lost, and consequently the result is a joint of lower efficiency. 
 
 Hot Tests. — Single-riveted butt-joints, steel plate, and wrought- 
 iron rivets, tested at the Watertown Arsenal at temperatures ranging 
 between 200° to 700 Fahr. , showed an increase in tensile strength when 
 heated over the duplicate cold joints at each temperature except 200°. 
 From 200° there was a gain in strength up to 300°, when the resistance 
 fell off some at 350°, increased again at 400°, and reached the maximum 
 effect observed at 500° ; from this point the strength fell very rapidly, 
 especially at 600° and 700° Fahr. 
 
 In per cent, of the cold joint there was a loss at 200° of 3.2 per cent., 
 the average of 3 joints ; at 500° the gain was 22.6 per cent., the average 
 of 4 joints. The maximum and minimum joints at this temperature 
 showed gains of 27.6 per cent, and 18.3 per cent, respectively. The 
 highest tensile strength of the net section of plate was found in the joint 
 tested at 500° Fahr., where 81,050 pounds was reached against a 
 strength of 58,000 pounds per square inch in the cold tensile strip. 
 
 The hot joints showed less ductility than the cold ones ; those tested 
 at 200° Fahr. not being exempt from this behavior, although there was 
 no near approach to brittleness in any. 
 
 The shearing strength of iron rivets was also increased by an eleva- 
 tion of temperature. The rivets in one joint of 350° sheared at 43,060 
 pounds per square inch, while in the duplicate cold joint they sheared 
 at 38,530 pounds per square inch. 
 
76 
 
 BOILERS AND FURNACES 
 
 > ;^ 
 
 X M 
 O 
 
 Sh . 
 
 
 
 
 
 
 
 
 
 
 
 ■i 
 
 x^^ ^^^ ^^ ^^ 
 
 :?^ ^x 
 
 
 
 
 Qo 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 k 
 
 ^^ 
 
 
 
 
 
 
 
 
 ■ - • • >s- 
 
 S: 
 
 
 
 
 
 
 
 
 
 . t^ 
 
 
 ^ 
 
 CL, g 
 
 
 
 
 
 
 
 
 • ■ ■ • ^ 
 
 
 
 
 
 
 
 
 
 * 
 
 * * 
 
 
 t 
 
 
 
 
 
 
 
 
 . M 
 
 . ON . . vo 
 
 
 
 ^S 
 
 
 
 
 
 
 
 00 
 • vo 
 
 ■ ^ ■ ■ ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 >- -4-: 
 
 
 
 '^ 
 
 
 
 
 . o . 
 
 . vo . .00 
 
 
 « 
 
 £, 
 
 
 
 
 
 
 
 • R • 
 
 00 lO 
 • vo • • vo 
 
 *^co 
 
 
 
 
 „ 
 
 
 o 
 
 
 . o . 
 
 . vo . 00 t^ 
 
 
 ^ 
 
 ^fe 
 
 
 
 Tl- 
 
 
 lO 
 
 
 t^ 
 
 vo 00 M 
 
 
 M 
 
 
 
 
 
 
 
 
 
 t 
 
 1-. -^ 
 
 
 
 o 
 
 
 00 
 
 
 vo lO . 
 
 . lO . M O 
 
 
 "^ 
 
 f^S 
 
 
 
 ID 
 
 
 ^ 
 
 
 vo vo 
 
 • vo • vo lO 
 
 
 
 . .OOrJ-. . .CO. .MM . 
 
 
 12 
 
 X 
 
 ^^ 
 
 • • vg^ S • • • S • • ^ vS • 
 
 vo >0 • vo ^ 
 
 . 
 
 
 .0010. .VOID. . a\ <r) . 
 
 vo M . -^ \r> 
 
 1 
 
 o 
 
 k 
 
 ^s 
 
 . ^ <X> . . ^ (J\ . .VOM 
 
 vo M . M t^ 
 
 vo vo vo lO 
 
 %o 
 
 
 .t^■<:^. .-;tm.t^"0O 
 
 00 t^ vo vo vo 
 
 u 
 
 ^ 
 
 n'^. 
 
 
 :ss vS-vg i;? 
 
 
 
 
 ' t^ vo r^ vo vo vo lo 
 
 a; 
 
 
 
 
 
 5 
 
 u^ 
 
 .OMTi-. .MM .'^iniDvo 
 
 vo O CO M to 
 
 
 ^ 
 
 111 § 
 
 . ^ d fO . . OO" ^ . vo M VD 00 
 
 fo 00 a\ t- ON 
 
 
 
 " 
 
 
 <D tn in ir> -^ 
 
 
 
 .fomvo .u:>'^lo.coMloo^ 
 
 O lO CO M 
 
 
 :?( 
 
 n^v; 
 
 • a^vg ■ ^ ^ ^ ■ ^ ^^^ 
 
 vg ^% S5 • 
 
 
 
 
 
 
 .v£)'*r^ .rOMCOt^rOM .'o 
 
 <0 . M . . 
 
 
 >, 
 
 ^g 
 
 'VOVDI/J ■j>.VOiOVOVOVO *VO 
 
 
 
 
 1/5 in 
 
 
 
 * 
 
 
 
 S? 
 
 ^s 
 
 ci t>. M* . d ro . 00 -^ . d 
 
 
 
 
 t^vOvo t-vo vovo vo 
 
 
 
 
 
 
 
 
 
 
 
 
 -S 
 
 ^§ 
 
 ^ ^ ■ • ss 
 
 00 
 
 
 • iS • • • • 
 
 
 
 
 
 
 
 is 
 
 
 
 
 
 
 
 ^ ■ • ' t^ 
 
 
 
 
 
 
 
 
 
 
 s" 
 
 
 
 
 fln^ 
 
 •g 
 
 ^xxxxxx^^x^x^^^^^^ 1 
 
 
 
 
 
 Bl (I, 
 
 
 
 
 E- 
 
 o 
 
 
 
 
 
 
 
 
 
 
 i 
 
RIVETED JOINTS 
 
 77 
 
 TABLE XVIII. 
 
 TABLE OF DIFFERENCES BETWEEN THE EFFICIENCIES AND RATIOS OF NET TO 
 GROSS AREAS. SINGLE-RIVETED BUTT-JOINTS, STEEL PLATE, WROUGHT- 
 IRON RIVETS, WATERTOWN ARSENAL. 
 
 
 
 Width of Plate Between Rivet-Holes. 
 
 
 
 
 
 
 
 
 
 
 
 
 Diameter 
 
 
 
 
 
 
 
 
 
 
 
 Plate. 
 
 
 
 
 
 
 
 
 
 
 
 
 i" 
 
 iVi" 
 
 1 1/" 
 
 i^s" 
 
 i%" 
 
 iVi" 
 
 i%" 
 
 i%" 
 
 2" 
 
 
 Inch 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 nc es. 
 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 
 % 
 
 6.0 
 
 4.6 
 
 6.1 
 
 
 
 
 
 . . 
 
 
 n 
 
 % 
 
 7-7 
 
 7.0 
 
 6.1 
 
 5.6 
 
 8.8 
 
 
 
 
 
 u 
 
 X 
 
 8.0 
 
 8.1 
 
 8.6 
 
 II. 7 
 
 7.6 
 
 II.8 
 
 3-1 
 
 
 
 % 
 
 
 18.8 
 13- 1 
 
 7.6 
 9.1* 
 
 7.8 
 
 10.5 
 
 8.5 
 
 6.6 
 
 13-3 
 
 8.8 
 
 6.7* 
 
 I 
 
 Y% 
 
 II. 2 
 
 10. 1 
 
 7.8 
 
 3.8- 
 
 
 
 
 . . 
 
 
 
 % 
 
 9-7 
 
 8.9 
 
 8.6 
 
 7-1 
 
 7.2 
 
 10.6 
 
 
 
 
 % 
 
 H 
 
 7-3 
 
 6.6 
 
 5-9 
 
 10.4 
 
 9-5 
 
 6.4 
 
 6.2 
 
 9.8 
 
 
 I 
 
 % 
 
 II. I 
 
 8.6 
 
 6.4 
 
 
 
 
 
 
 
 H 
 
 % 
 
 II. I 
 
 lO.O 
 
 17.2 
 
 5-3 
 
 12.3 
 
 
 
 
 
 % 
 
 y^ 
 
 12.8 
 
 6.2 
 
 4.7 
 
 14.9 
 
 17.2 
 
 5-2 
 
 17.8 
 
 
 
 I 
 
 % 
 
 4.6 
 
 6.5 
 
 5-3 
 
 21.7 
 
 4.1 
 
 5-4 
 
 6.1 
 
 7-5 
 
 4.2- 
 
 iV& 
 
 % 
 
 6.4 
 
 9.7* 
 
 17.4 
 
 7-5 
 
 
 
 
 
 
 'A 
 
 % 
 
 8.^ 
 
 7-1 
 
 8.0 
 
 8.1 
 
 6.6* 
 
 3.1* 
 
 
 
 
 I 
 
 H 
 
 9-4 
 
 8.0 
 
 8.0 
 
 7-1 
 
 7-6 
 
 7-4 
 
 5.8 
 
 6.1 
 
 5.9* 
 
 ^y% 
 
 H 
 
 4.1 
 
 5.3 
 
 3.7- 
 
 6.7 
 
 
 
 
 
 
 I 
 
 H 
 
 7-1 
 
 7.2 
 
 7-9 
 
 7-3 
 
 4-7 
 
 3.1- 
 
 7.9 
 
 
 
 ^A 
 
 H 
 
 4.8 
 
 8.2 
 
 7-4 
 
 6.8 
 
 2.4 
 
 6.2 
 
 7.6 
 
 3.5* 
 
 2.1 
 
 ix 
 
 * Denote that joints did not fracture along line of riveting. 
 TABLE XIX. 
 
 EXCESS IN STRENGTH OF NET SECTION IN JOINT OVER STRENGTH OF TENSILE 
 TEST-STRIP. SINGLE-RIVETED BUTT-JOINTS, STEEL PLATE, WROUGHT-IRON 
 RIVETS, WATERTOWN ARSENAL. 
 
 
 
 
 
 
 
 Thick- 
 
 
 
 
 
 
 
 
 
 Diameter 
 
 
 
 
 
 
 
 
 
 
 
 Plate. 
 
 
 
 
 
 
 
 
 
 
 Holes. 
 
 
 1" 
 
 ■»' 
 
 15^" 
 
 m" 
 
 i%" 
 
 1^8" 
 
 i%" 
 
 i%" 
 
 2" 
 
 
 Inch. 
 
 Per 
 
 Pe, 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 cent. 
 
 
 % 
 
 9.8 
 
 7.2 
 
 9-1 
 
 
 
 
 . . 
 
 
 
 A 
 
 X 
 
 13-4 
 
 II.7 
 
 9-7 
 
 8.5 
 
 II. 
 
 
 
 
 
 H 
 
 ^ 
 
 17.4 
 
 14.4 
 
 14.7 
 
 14.7 
 
 II.9 
 
 16. 1 
 
 4-7 
 
 
 
 % 
 
 % 
 
 17.6 
 
 14-5 
 
 14.2 
 
 18.0 
 
 14. 1 
 
 10.7 
 
 17.7 
 
 13-5 
 
 10. I* 
 
 I 
 
 H 
 
 21.2 
 
 14. 1- 
 
 
 
 
 
 • • 
 
 
 
 % 
 
 n 
 
 19-5 
 
 16.8 
 
 12.5 
 
 5.8* 
 
 
 . . 
 
 
 . . 
 
 
 Ya 
 
 V, 
 
 18.2 
 
 15-9 
 
 13-5 
 
 II. 7 
 
 II.4 
 
 ib.5 
 
 • • 
 
 
 
 % 
 
 n 
 
 14.6 
 
 12.5 
 
 10.6 
 
 17.9 
 
 15-9 
 
 10.3 
 
 9.8 
 
 I5-I 
 
 
 I 
 
 % 
 
 15.8 
 
 14.4- 
 
 10.3 
 
 
 
 
 
 
 
 Ya 
 
 % 
 
 20.9 
 
 17.8 
 
 28.9 
 
 8.7 
 
 19-5* 
 
 
 
 
 
 % 
 
 y^ 
 
 25-5 
 
 II.8 
 
 5-5 
 
 25.8 
 
 28.7 
 
 8.6 
 
 28.1 
 
 
 . . 
 
 I 
 
 % 
 
 9.8 
 
 13-2 
 
 10. 1 
 
 39-7 
 
 7-3 
 
 9-1 
 
 10. 1 
 
 12.2 
 
 6.5- 
 
 lA 
 
 % 
 
 I5-I 
 
 18.8 
 
 29.6 
 
 12.2 
 
 
 
 
 
 
 n 
 
 Vs 
 
 16.6 
 
 17.2 
 
 14.6 
 
 14.0 
 
 ii.i* 
 
 5-2* 
 
 
 
 
 I 
 
 % 
 
 20.0 
 
 18.4 
 
 15-3 
 
 13.0 
 
 13-5 
 
 12.7 
 
 9-7 
 
 9.8 
 
 9.2 
 
 iVs 
 
 H 
 
 8.3 
 
 10. 1 
 
 6.8* 
 
 II-5 
 
 
 
 
 
 
 I 
 
 H 
 
 15-2 
 
 14-5 
 
 15-2 
 
 1.3.6 
 
 8.5 
 
 5..3* 
 
 II-5 
 
 
 
 iVs 
 
 % 
 
 10.8 
 
 17.4 
 
 15.0 
 
 13.2 
 
 4-7 
 
 II.O 
 
 13.0 
 
 5.7* 
 
 3-5 
 
 iX 
 
 Denote that joints did not fracture along line of riveting. 
 
78 
 
 BOILERS AND FURNACES 
 
 
 H 
 
 
 <n 
 
 
 hj 
 
 
 Oh 
 
 !x! 
 
 iJ 
 
 X 
 
 
 W 
 
 ^r 
 
 hJ 
 
 
 CQ 
 
 W 
 
 < 
 
 ^ 
 
 H 
 
 
 :^ 
 
 ^ 
 
 ^ o ,5? 
 
 o o o 
 
 v8 & 
 
 o\ K 00 lo 
 
 id X 
 
 „ I 5 § 5^ S 
 
 
 
 >S o SC ^ 
 
 lO O >0 U-) 
 
 " LO ^* to 
 
 ^ ^ 
 
 •S s 
 
 ? .s- 
 
 
 y r, ^ ^. -^ 
 
 « "C .5: 
 
 § H ^ (5 :s 5 2; E 
 
 i t; .2 s 
 
 S 2 a; rt 
 H O is § 
 
 ^ iE '5 
 
 W W (X, 
 
RIVETED JOINTS 
 
 79 
 
 ENGRAVINGS BELONGING TO TABLE XXI. 
 
 Fig. 63. 
 
 Fig. 64. 
 
 Fig. 65. 
 
 Fig. 66. 
 
 ^.J> 
 
 <i> i> 
 
 p o o o 
 
 '^dr-iZZeeiJ'ol&S. 
 
 Fig. 67. 
 
 Fig. 67 A. 
 
 j<o/z,eet7 
 
 
 ^3f- 
 
 6 
 
 
 Fig. 
 
 Fig. 69. 
 
 00000 
 O CD O O 
 
 00000 
 
 
 \U- 
 
 o © cp o 
 
 O (]) 9 o 
 
 " i5/icisre<yt- 
 
8o 
 
 BOILERS AND FURNACES 
 
 ENGRAVINGS BELONGING TO TABLE Y.yJ..— Continued. 
 Fig. 70. 
 
 ^S?iejCbns^. 
 
 (^^/"— ^ o o 
 
 Fig. 71. 
 
 Fig. 72. 
 
 o (^3f4^ o 
 
 e^ ^ J'riiled.T/di&s 
 
 
 000000 
 00000 
 
 O C^^i^ o o 
 
 00000 
 
 =^ 
 
 
 Fig. 73. 
 
 Fig. 74. 
 
 00000 
 
 
RIVETED JOINTS 
 
 8l 
 
 ENGRAVINGS BELONGING TO TABLE y^y^l.— Continued. 
 
 Fig. 76. 
 
 
 000c 
 
 0000000 
 
 O O O O O , O I o 
 
 O O O O ^,0 
 
 
 o 
 
 2Pl.cut.&^, /^~\ 
 
 Fig. 77- 
 
 Fig. 78. 
 
 ;^~ 
 
 a 
 
 — 
 
 A 
 
 -- 
 
 "^ 
 
 ? 
 
 
 
 1 
 
 ^ 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 -O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 --- 1 
 
 Fig. 79- 
 
 O H-^^'^ 
 
 > o o o (p''qj < 
 000000 
 
 000000 
 ) o o o o o c 
 
 O/^/Td^c^r^^y^^fejQ 
 
82 
 
 BOILERS AND FURNACES 
 
 Fig. 8o. 
 
 <> 
 
 
 PROPORTIONING RIVETED JOINTS, 
 
 Single-Riveted Lap-joint. — Such a joint consists of two plates 
 overlapping each other and secured by a single row of rivets, as shown 
 
 in Fig. 80. This is the simplest 
 and weakest form of riveted joint. 
 In estimating the value of a rivet 
 under shearing stress the diameter 
 of the hole should be taken, because 
 the rivet is upset to fill the hole. 
 The shearing strength of rivets is 
 given in Tables VII. and X. 
 
 The strength of a single-riveted 
 lap-joint using -^-inch steel of 55,000 
 pounds tensile strength, steel rivets 
 f inch diameter in -^-inch holes, 2 inches pitch, may be calculated thus : 
 
 Gross section of plate, 2 inches X .25 inch 5000 square inch. 
 
 Section removed by hole, .6875 inch X .25 inch . . . .1719 square inch. 
 
 Net section of plate between holes 3281 square inch. 
 
 Area of upset rivet, \^ inch 3712 square inch. 
 
 Shearing strength of rivet at 44,625 pounds per square 
 
 inch 16,565 pounds. 
 
 Plate (gross), .5 square inch X 55,000 pounds 27,500 pounds. 
 
 Plate (net), .3281 square inch X 55,000 pounds .... 18,045 pounds. 
 
 Strength of joint in per cent, of solid plate would be 
 
 18,045 X n 
 27,500 
 
 65.6 per cent. 
 
 if it were not for the fact that the shearing strength of the rivet is less 
 than the strength of the plate ; we then have 
 16,565 X 100 
 
 27,500 
 
 60.02 per cent. 
 
 An examination of the above figures shows that the net strength of 
 the plate is 18,045 pounds and the shearing strength of the rivets 16,565 
 pounds. This proportion of strength of plate to that of rivet is permis- 
 sible in single-riveted joints, because shearing of rivets or failure of any 
 kind at the joint seldom occurs except by corrosion. The excess 
 strength ought therefore to go into the plate rather than into the rivet, 
 because the plate is subject to corrosion or other deterioration which 
 affects the rivets to a less degree. 
 
 Single-riveted lap-joints when both plates and rivets are of iron are 
 weaker than similar plates and rivets of steel of the same dimensions, 
 because the latter material will resist tensile and shearing stresses to a 
 greater degree than iron. Single riveting is little used in longitudinal 
 seams, except for small boilers and low pressures ; but single-riveted 
 
Material of plate 
 
 Thickness of plate, nominal 
 Width of test-specimen, inche 
 Thickness of welt-strips, inch 
 Width of upper welt, inches 
 Width of lower welt, inches . 
 Diameter of rivets, inches . . 
 
 Material of rivets 
 
 Diameter of holes, punched o: 
 
 WATERTOWN ARSENAL. 
 
 Steel. 
 
 V, 
 
 20 
 
 ¥2 
 
 7.375 
 
 7-375 
 
 Vx 
 Steel. 
 
 St 
 
 Number of rivets in one-half i ^3 
 
 Pitch of rivets, middle row, ir 
 For details of riveting, see Fi: 
 Bearing surface of rivets, squ; 
 Shearing area of rivets, squar 12.46 
 Tensile strength of plate, pout 53,7io 
 Gross sectional area, square ir 12.41 
 Net sectional area, square inc 9-5° 
 Maximum stress on joint per \ 
 
 Tension on gross sectiori 42,720 
 
 Tension on net section 55.8oo 
 
 Compression on bearing 84,140 
 
 Shearing on rivets, pounds . | 42,54° 
 
 3 
 
 74 
 6.30 
 
 Efficiency of joint, per cent. . 
 Fracture, similar to Figures 
 
 79-5 
 74 
 
 Steel. 
 
 10.8 
 10.8 
 
 Y.{ 
 
 16 
 
 3is 
 
 75 
 
 7-73 
 15-34 
 53,710 
 12.36 
 
 9-94 
 
 43,600 
 54,220 
 
 69,720 
 35,130 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 Vb 
 
 % 
 
 % 
 
 20.12 
 
 17 
 
 16.5 
 
 Up. .6, L. .43 
 
 % 
 
 Y, 
 
 8.8 
 
 9-8 
 
 13K 
 
 23.8 
 
 9.8 
 
 13% 
 
 3 pi. = r 
 2 pi. = 1/8 
 
 }■ 
 
 M 
 
 Iron. 
 
 Steel. 
 
 Steel. 
 
 iJsandiiX 
 
 ifs D. 
 
 irVD. 
 
 19 
 
 9 
 
 II 
 
 2% 
 
 zVx 
 
 4i% 
 
 76 
 
 77 
 
 78 
 
 13-58 
 
 8.23 
 
 10.23 
 
 30.41 
 
 15-96 
 
 19-51 
 
 53-710 
 
 51,190 
 
 51,190 
 
 12.81 
 
 14.646 
 
 14.438 
 
 10.19 
 
 10.07 
 
 10.72 
 
 46,070 
 
 36,190 
 
 38,400 
 
 57,940 
 
 52,640 
 
 51,720 
 
 43,470 
 
 64,410 
 
 54,190 
 
 19,410 
 
 33,210 
 
 28,420 
 
 85.8 
 
 70.7 
 
 75-0 
 
 76 
 
 77 
 
 78 
 
 Steel. 
 % 
 i5'/i 
 
 m 
 
 2 pi. = ijl 
 Iron. 
 
 lYa and 1^ 
 
 16 
 
 2% 
 
 79 
 
 16.38 
 28.00 
 51,190 
 13-852 
 10.99 
 
 41,740 
 52,640 
 35,300 
 20,650 
 81.5 
 79 
 
TABLE XXI. 
 
 TESTS OF DOUBLE-, TRIPLE-, AND QUADRUPLE-RIVETED BUTT-JOINTS, STEEL PLATES, IRON AND STEEL RIVETS, WATERTOVVN ARSENAL. 
 
 Material of plate 
 
 Thickness of plate, nominal, inch 
 
 Width of test-specimen, inches 
 
 Thickness of welt-strips, inch 
 
 Width of upper welt, inches 
 
 Width of lower welt, inches 
 
 Diameter of rivets, inches 
 
 Material of rivets 
 
 Diameter of holes, punched or drilled, inches . . . . 
 
 Number of rivets in one-half of joint 
 
 Pitch of rivets, middle row, inches 
 
 For details of riveting, see Figures 
 
 Bearing surface of rivets, square inches 
 
 Shearing area of rivets, square inches 
 
 Tensile strength of plate, pounds per square inch . . . 
 
 Gross sectional area, square inches 
 
 Net sectional area, square inches 
 
 Ma.\inmm stress on joint per square inch : 
 
 Tension on gross section of plate, pounds .... 
 
 Tension on net section of plate, pounds 
 
 Compression on bearing surface of rivets, pound 
 
 Shearing on rivets, pounds 
 
 Efficiency of joint, per cent 
 
 Fracture, similar to Figures 
 
 Steel. 
 
 a 
 
 Iron. 
 
 %D. 
 
 9 
 
 2% 
 63 
 
 1-97 
 10.82 
 58,170 
 
 2.50 
 
 48,150 
 69,140 
 87,740 
 15.980 
 
 Steel. 
 14-32 
 
 Iron. 
 
 10 
 
 2% 
 64 
 
 61,470 
 3-54 
 2.46 
 
 46,810 
 67,370 
 76,720 
 13.790 
 76.1 
 
 Steel. 
 14.06 
 
 A 
 
 Iron. 
 
 3% 
 
 65 
 
 2.44 
 12-57 
 56,760 
 
 4.29 
 
 3-07 
 
 45.490 
 63.570 
 
 Steel. 
 1% 
 
 2.46 
 12.57 
 56,760 
 4-34 
 3-10 
 
 45.530 
 63.740 
 80,330 
 15,720 
 
 Steel. 
 
 2.30 
 7-07 
 58,340 
 4-79 
 3-64 
 
 48,610 
 63,970 
 101,250 
 32,940 
 
 Steel. 
 14-51 
 
 7.07 
 56.670 
 5-35 
 4-25 
 
 48,500 
 61,060 
 
 36,700 
 85-5 
 67 
 
 Steel. 
 14.52 
 
 7.07 
 56,670 
 5-36 
 4-25 
 
 47,700 
 60,160 
 115,700 
 36,170 
 
 Steel. 
 14.41 
 
 a 
 
 Iron. 
 I D. 
 
 5-89 
 3-85 
 
 43,360 
 66,340 
 62,440 
 16,260 
 70.9 
 
 Steel. 
 12.48 
 
 a 
 
 Iron. 
 
 %o. 
 
 8 
 3% 
 69 
 2.56 
 7.07 
 59,000 
 5-33 
 4-05 
 
 48,120 
 63,300 
 100,190 
 36,280 
 83-3 
 69 
 
 Steel. 
 27.00 
 5% 
 
 10^3 
 
 % 
 
 steel. 
 HD. 
 
 12 
 
 3% 
 
 70 
 
 6.02 
 13-80 
 59,680 
 14.44 
 12.44 
 
 39,450 
 81,530 
 35,560 
 
 57 
 
 70 
 
 Steel. 
 13-515 
 
 15J4 
 
 % 
 
 Steel. 
 
 11.04 
 60,550 
 6-73 
 5-80 
 
 Steel. 
 14-985 
 
 10.15 
 16.50 
 
 K 
 Steel. 
 i§D. 
 
 14 
 
 2j^ 
 72 
 6.17 
 17-25 
 
 48,620 ; 51,205 
 
 56,410 , 63,000 
 
 77,900 ^ 58,510 
 
 29,640 ; 20,930 
 80.3 85.5 
 
 71 I 72 
 
 Steel. 
 
 Steel. 
 
 Steel. 
 
 H 
 
 K 
 
 Vs 
 
 20 
 
 20 
 
 20 
 
 ^ 
 
 K 
 
 'A 
 
 7-375 
 
 7-375 
 
 10.8 
 
 7-375 
 
 7-375 
 
 10.8 
 
 Vx 
 
 K 
 
 K{ 
 
 Steel. 
 
 Steel. 
 
 Iron. 
 
 IS 
 
 S5 
 
 M 
 
 13 
 
 13 
 
 16 
 
 3 
 
 3 
 
 3i% 
 
 73 
 
 74 
 
 75 
 
 6.72 
 
 6.30 
 
 7-73 
 
 12.46 
 
 12.46 
 
 15-34 
 
 53,710 
 
 53,710 
 
 53.710 
 
 13.22 
 
 12.41 
 
 12.36 
 
 10,12 
 
 9-50 
 
 9-94 
 
 42,860 
 
 42,720 
 
 43.600 
 
 55,990 
 
 55,800 
 
 54,220 
 
 84,320 
 
 84,140 
 
 69,720 
 
 45,470 
 
 42,540 
 
 35,130 
 
 79-8 
 
 79-5 
 
 81.2 
 
 73 
 
 74 
 
 75 
 
 Steel. 
 Va 
 
 20.12 
 
 Up..6,L. 
 
 19 
 
 2% 
 
 76 
 13-58 
 
 30.41 
 53.710 
 
 46,070 
 
 57,940 
 43.470 
 19,410 
 
 85.8 
 
 76 
 
 steel. 
 
 Steel. 
 
 % 
 
 % 
 
 •7 
 
 16.5 
 
 % 
 
 % 
 
 9-8 
 
 13K 
 
 9-8 
 
 13K 
 
 }■ 
 
 M 
 
 Steel. 
 
 Steel. 
 
 lAD. 
 
 lAD. 
 
 9 
 
 " 
 
 3H 
 
 4i». 
 
 77 
 
 78 
 
 8.23 
 
 10.23 
 
 15-96 
 
 19-51 
 
 51.190 
 
 51,190 
 
 14.646 
 
 14-438 
 
 10.07 
 
 10.72 
 
 36.190 
 
 38,400 
 
 52,640 
 
 51,720 
 
 64,410 
 
 54,190 
 
 33,210 
 
 28,420 
 
 70.7 
 
 75-0 
 
 77 
 
 78 
 
 Steel. 
 % 
 
 15M 
 Up.r8,L-K 
 
 m 
 
 2I« 
 
 51.190 
 13-852 
 10.99 
 
 41.740 
 52,640 
 35,300 
 20,650 
 81.5 
 79 
 
RIVETED JOINTS 
 
 83 
 
 joints are in common use for the circumferential seams in steam boilers, 
 of which an example is here given : 
 
 Should the circumferential seams of a 72-inch boiler be single or 
 double riveted to withstand a pressure of 120 pounds? 
 
 A boiler 72 inches in diameter has an area of 4071.5 square inches. 
 The steam, 120 pounds per square inch, would make total pressure on 
 the head of 4071.5 X 120 = 488,580 pounds. 
 
 If the shell be of steel -^ inch thick, the diameter of steel rivets may 
 be ^1 inch in -J-inch holes. The area of each hole is 0.601 square inch. 
 The shearing strength of each rivet is 26,833 pounds, as per Table X. 
 If the pitch of the rivets be 2 inches, there will be required half as many 
 rivets as there are inches in circumference of the boiler : 
 
 226 
 
 113 rivets. 
 
 Then 26,833 X 113 = 3,038,129 pounds. If a factor of safety of 6 
 be chosen, the safe working pressure would be 
 
 3.03 
 
 ^ = 506,355 pounds, 
 
 which is 17,775 pounds additional to the required strength of 488,580 
 pounds, — showing that the single-riveted joint has ample strength for 
 the conditions given. 
 
 A series of experiments to determine the efficiencies of single-riveted 
 lap-joints for pitches from i^ to 3^ inches, with holes ^ to i^ inches 
 diameter, upon iron and steel plates ){ inch thick, resulted as follows : 
 
 TABLE XXII. 
 
 EFFICIENCIES OF SINGLE-RIVETED LAP-JOINTS IN PER CENT. OF THE 
 
 STRENGTH OF SOLID PLATE. 
 
 I, Iron ; S, Steel ; D, Drilled holes ; P, Punched holes. 
 
 Plate. 
 
 
 Diameter of Rivet-Holes. Rivets y'g Inch Less. 
 
 
 OF 
 
 Rivets. 
 
 
 
 Thick. 
 
 Tensile 
 Strength. 
 
 \h 
 
 Vs 
 
 ^ 
 
 1% 
 
 rV^ 
 
 3^1 
 
 47,925 
 
 i/s 
 
 64.0 
 
 
 
 
 
 XS 
 
 55,765 
 
 iH 
 
 68.8 
 
 
 
 . . 
 
 . . 
 
 XI 
 
 47,925 
 
 2 
 
 liilp^ 
 
 
 
 
 ■ ■ 
 
 XS 
 
 55,765 
 
 2 
 
 {fAi 
 
 
 
 
 
 XS 
 
 61 ,000 
 
 2^ 
 
 
 65.1 
 
 
 
 
 3^s 
 
 58.150 
 
 2^ 
 
 
 
 69.7 
 
 
 
 XS 
 
 61,000 
 
 27/s 
 
 . . . 
 
 
 
 70.9 
 
 
 XS 
 
 58,150 
 
 274 
 
 
 
 
 
 70.7 
 
 XS 
 
 58,150 
 
 iVz 
 
 
 . . 
 
 
 
 69.1 
 
 XS 
 
 55,740 
 
 3^8 
 
 . . . 
 
 . . 
 
 
 . . 
 
 75-7 
 
 XS 
 
 55,740 
 
 3% 
 
 
 
 
 
 67.3 
 
 The above results show that no practical advantage is had in usin^ 
 rivets larger than ^ inch in ^-inch plates. 
 
84 
 
 BOILERS AND FURNACES 
 
 TABLE XXin. 
 
 SINGLE-RIVETED LAP-JOINTS, IRON PLATES AND IRON RIVETS. 
 
 Table of least distance between rivet-holes for iron plates of 45,000 pounds 
 tensile strength, and iron rivets of 34,200 pounds shearing strength, that the 
 plates and rivets shall approximate each other in strength. 
 
 Proportionate 
 Strength of 
 
 Plate at 45,000 
 Pounds per 
 
 Square Inch. 
 
 Rivets 
 
 
 Width of 
 
 Plate to equal 
 
 Shearing 
 
 Strength of 
 
 Rivet. 
 
 Pitch op 
 
 Rivets. 
 
 
 Diameter of 
 Hole. 
 
 Shearing 
 Strength. 
 
 Decimal. 
 
 Nearest 
 Working 
 Fraction. 
 
 age of 
 Joint. 
 
 Inch. Pounds. 
 
 X = 11,250 
 
 j^e = 14,063 
 
 H = 16,875 
 ^V = 19,688 
 ^ = 22,500 
 
 T6 = 24,313 
 
 >^ = 28,125 
 H = 30,938 
 X = 33,750 
 
 Inches. 
 
 tI= .9375 
 I = 1. 000 
 
 ItV = 1.0625 
 
 Ij/g = 1. 125 
 
 It6 =1.1875 
 
 Pounds. 
 12,695 
 15,110 
 
 17,733 
 20,564 
 23,608 
 26,861 
 30,322 
 
 33,995 
 
 37,877 
 
 Inches. 
 I. Ill 
 1.074 
 1. 051 
 1.044 
 1.049 
 1. 105 
 1.078 
 1.099 
 1. 122 
 
 Inches. 
 
 1.799 = It! 
 1.824 = lif 
 1.864 = i^ 
 1.919 = lif 
 1.986 = 2 
 
 2.105 = ^H 
 
 2. 141 = 2>^ 
 2.224 = 2^ 
 2.310 = 23^6 
 
 61.76 
 58.88 
 56.38 
 54.40 
 52.82 
 52.49 
 50.35 
 49.41 
 48.57 
 
 TABLE XXIV. 
 
 SINGLE-RIVETED LAP-JOINTS, STEEL PLATES AND STEEL RIVETS. 
 
 Table of least distance between rivet-holes for steel plates of 55,000 pounds 
 tensile strength and steel rivets of 44,625 pounds shearing strength, that the 
 plates and rivets shall approximate each other in strength. 
 
 Proportionate 
 Strength of 
 
 Plates at 55,000 
 Pounds per 
 Square Inch. 
 
 Rivets 
 
 
 Width of 
 
 Plate to equal 
 
 Shearing 
 
 Strength of 
 
 Rivet. 
 
 Pitch of Rivets. 
 
 
 Diameter of 
 Hole. 
 
 Shearing 
 Strength. 
 
 Decimal. 
 
 Nearest 
 Working 
 Fraction. 
 
 age of 
 Joint. 
 
 Inch. Pounds. 
 
 Inches. 
 
 Pounds. 
 
 Inches. 
 
 Inches. 
 
 
 X = 13,750 
 
 H= .6875 
 
 16,565 
 
 1.205 
 
 1.892 = l7/s 
 
 63.69 P. 
 
 ^ = 20^625 
 
 tV = 24,063 
 
 H= .75 
 if= .8125 
 
 rs= .875 
 
 19,715 
 23,138 
 26,833 
 
 1. 147 
 1. 121 
 1. 115 
 
 1.897 = I^ 
 
 1.934 = ItI 
 
 1.990 = 2 
 
 60.46 P. 
 57.97 P. 
 
 56.03 R. 
 
 >^ = 27,500 
 
 il= -9375 
 
 30,805 
 
 1. 120 
 
 2.058 = 2xV 
 
 54.42 P. 
 
 tV = 30,938 
 >^ = 34,375 
 
 I = 1. 000 
 
 35,048 
 
 I.I33 
 
 2.133 = ^'A 
 
 ^H^ I- 
 
 lT6 = 1.0625 
 
 39,299 
 
 1. 143 
 
 2.205 = 21^^ 
 
 51.84 P. 
 
 if = 37,813 
 X = 41,250 
 
 I>^ = 1.125 
 lA = 1.1875 
 
 44,357 
 49,422 
 
 1. 173 
 1. 198 
 
 2.298 = 2i§g 
 2.386 = 2^ 
 
 51.04 p. 
 50.21 p. 
 
 P — Joint will probably fail by tearing the plate between rivet-holes. 
 R — Joint will probably fail by shearing the rivets. 
 
RIVETED JOINTS 85 
 
 Double-Riveted Lap-joint. — In this form of joint the plates lap 
 over each other far enough to admit two rows of rivets, as shown in 
 Fig. 81. A double-riveted joint is stronger 
 than a single-riveted joint, because of wider Fig. 
 
 spacing or pitch of rivets giving a larger net 
 section of metal, larger surfaces of frictional 
 contact, and a larger rivet area under shear- 
 ing stress. Longitudinal seams in steam boil- 
 ers should be at least double riveted, no plates 
 
 less than ^ inch thick, and no rivets less than ^ inch diameter, how- 
 ever small the diameter of the boiler. 
 
 The strength of a double-riveted joint may be calculated thus : Let 
 us assume that steel plates of 55,000 pounds tensile strength are to be 
 joined by iron rivets of 34, 200 pounds per square inch shearing strength, 
 the riveting and spacing to be as given below : 
 
 Thickness of plate, ^ inch 25 inch. 
 
 Diameter of rivet-hole, \^ inch 6875 inch. 
 
 Area of rivet-hole .3712 square inch. 
 
 Pitch of rivets 2.5 inches. 
 
 The strength of the whole plate would be 2,5 X .25 X 55,000 = 
 34,375 pounds. 
 
 The strength of net section of plate would be (2.5 — -6875) X .25 
 X 55.000 = 24,915 pounds. 
 
 The strength of two rivets in single shear would be .3712 X 2 X 
 34,200 = 25,390 pounds, showing that the plate is slightly the weaker 
 of the two. 
 
 The percentage of strength of the joint is 
 
 24,915 X 100 
 
 = 72.48 per cent. 
 
 34,375 
 
 Zigzag Riveting is that arrangement of rivets in which one row 
 is placed over the centre of the intervening space, as shown in Fig. 82. 
 This is the style of riveting in almost universal use, having two good 
 qualities, — strength and tightness underpressure. Correctly made, zig- 
 zag joints are equal in shearing strength to the net section of the punched 
 plates ; that is to say, the value of a joint of this kind approximates 70 
 per cent, of the whole plate for ^-inch to ^-inch iron plates with iron 
 rivets or steel plates with steel rivets. 
 
 Chain Riveting is that arrangement of rivets in which one row is 
 placed exactly above the other, as in Fig. 83. Experiments conducted 
 with a view to ascertaining the comparative strength of chain and zig- 
 zag riveting showed that, for the same spacing of the rivets from centre 
 to centre across the sheet, the chain riveting was the strongest. Chain 
 riveting requires a broader lap than zig-zag riveting, and no doubt the 
 
86 BOILERS AND FURNACES 
 
 friction of this wider joint contributes towards the observed increase in 
 strength. The commonly accepted notion is that the second row of 
 rivets counts for little or nothing in adding to the strength of the joint 
 over that of single riveting for the same pitch ; but this has been proven 
 experimentally not to be true, and the fact is that the arrangement of 
 rivets as in a chain-riveted joint is actually stronger than a zigzag joint 
 of the same relative proportions of rivet to plate area. 
 
 Strength of Double-Riveted Joints.— In Table XXV. are given 
 the percentages for joints made up of iron plates with iron rivets, steel 
 plates with iron rivets, and steel plates with steel rivets. It will be noted 
 that most of the percentages fall below 70. As this latter figure is com- 
 monly assumed to be the strength of a double-riveted lap-joint, it may 
 be said in explanation that percentages of strength are controlled by the 
 pitch and diameter of rivets, the wider the spacing and the smaller the 
 diameter of the rivet the greater will be the percentage of strength of 
 joint when relative areas of solid and perforated plates are concerned ; 
 but a due regard must be had for shearing strength of rivets which enter 
 into the construction of a joint : narrow pitches and large diameter of 
 rivets means larger resistance against shearing, but the tendency to 
 crush thin plates before large rivets must not be disregarded. Taken 
 altogether, the spacing and dimensioning of rivets in riveted joints is 
 a matter upon which a large amount of care and judgment has been be- 
 stowed, resulting in the adoption of practically the dimensions as given 
 in the table, which is fairly representative of the best boiler practice 
 at this time. 
 
 Referring to Table XXV. , in the column ' ' iron plate and iron 
 rivets," the riveting is shown to be stronger than the plate. The next 
 column, "steel plate and iron rivets," shows that by reason of the 
 higher tensile strength of the steel plates the first joint in the table is 
 weakest through the line of rivet-holes, while the remaining figures with 
 reference letter show that the rivets are weaker than the plates. By 
 referring to the last column it will be seen that by using steel rivets 
 having a higher shearing strength than iron the joints are strengthened 
 against the shearing action of the plates throughout the whole series. 
 An examination of these figures in connection with the horizontal dis- 
 tance from centre to centre of rivet would seem to indicate that a re- 
 vision of the pitch of rivets would be advisable, that a higher percentage 
 of joint might be secured by increasing the pitch of the rivets for the 
 thinner plates and slightly decreasing the pitch for the thicker ones ; 
 but the practical consideration of securing a tight joint is one of great 
 importance, and it is not recommended that there be any considerable 
 deviation from centre to centre of the horizontal rows of rivets as given. 
 
 Triple- Riveted Lap-joints. — In any case where a double-riveted 
 joint is deficient in rivet area, an increase in strength is had by simply 
 extending the lap of the joint sufficiently to admit another row of rivets ; 
 
RIVETED JOINTS 
 
 87 
 
 ^s:|:^sl-;^3-^>x 
 
 
 
 
 
 
 ^J"^s^s!«X^i^^ 
 
 > 
 
 2 
 
 1 
 
 • 
 
 ffi|"^oi|" o>|c;i^o>|wP\oi|i-' 
 
 > 
 
 
 
 ^K^j^"" 
 
 w 
 
 ►.gas 
 
 ■ 2.00 
 
 
 p 
 
 HI 
 
 1 
 
 n 
 
 'V 
 
 
 
 £5i«J»^ 
 
 w 
 
 < 
 
 1 
 
 
 p 
 
 1 
 
 n 
 
 3 
 
 1? 
 
 n 
 
 3 
 
 
 p^ 
 
 <: 
 
 1 
 
 
 !^ 
 
 If 
 
 r 
 
 I 
 
 r 
 > 
 H 
 W 
 
 OCn Oi Oi 4i^ ^ 4i^ -P- OJ 
 
 !^ 
 
 
 0^ <J\ ON On ON ON^J ^J -O 
 
 -ji. asanas covp p w to 
 CO b b b\ b en 4^ -K 4^ 
 
 tocn OCyi^Ol^ ONOO 
 
 ^ ►TJ ^ ^ 13 '-d ^ "d ^ 
 
 Iron Plate, 
 Iron Rivets. 
 
 M 
 Z 
 
 s 
 
 
 "0 
 
 Ca Oi Cn Cn Oi Cn 0\ 0^^ 
 
 4i. 4^ 0\-<I COv^ to- ^ W 
 
 ?d ?fl ?o ;^ ?o ^73 ?d ?c ^ 
 
 3 
 
 1 
 
 1 
 
 0\ ON ON CT\ ON Os~<I ^J ^J 
 f>. ON p^ ON 00>O p M 10 
 
 (OCn Oai^C/i--l o^oo 
 
 i-d '-0 hj ►Td -Tj i-d TJ '-d '-d 
 
 C/5 
 
 1 
 
 
 > 1- 
 
 2 < 
 
 <i w 
 
 o r 
 
 <B-T-0^r: 
 
88 BOILERS AND FURNACES 
 
 by so doing the pitch of the rivets in the horizontal line can usually be 
 widened, especially in the case of zigzag riveting, without impairing 
 the tightness of the joint. 
 
 Referring to Table XXVI., it will be seen that with but two excep- 
 tions the percentages of strength is above 75 for ' ' iron plate and iron 
 rivets. ' ' The riveting is in every case stronger than the net area of the 
 plates ; this also holds good for "steel plate with steel rivets." In the 
 column of ' ' steel plate with iron rivets, ' ' the riveting is in every in- 
 stance weaker than the plate. Triple-riveted joints are not much used 
 except for plates less than y'^ inch thick, the preference being given to 
 butt-joints with double welts for plates ^ inch thick and over. 
 
 The strength of a triple-riveted joint of say ^-inch steel plates of 
 55,000 pounds tensile strength, rivet-holes -^ inch diameter, spaced on 4 
 inches pitch, rivets of iron having a shearing resistance of 34, 200 pounds 
 per square inch, may be calculated thus : 
 
 Thickness of plate, Yz inch 5 inch. 
 
 Diameter of rivet-hole, \% inch 9375 inch. 
 
 Area of rivet-hole . . 6903 square inch. 
 
 Pitch of rivets, 4 inches 4.000 inches. 
 
 The strength of the whole plate would be 4 X -5 X 55, 000 = 
 110,000 pounds. 
 
 The strength of the net section of plate would be (4 — -9375) X .5 
 X 55,000 ^84,219 pounds. 
 
 The strength of 3 rivets in single shear would be .6903 X 3 X 
 34,200 = 70,824 pounds, showing that the net section of the plate is 
 stronger than the riveting. 
 
 The percentage of strength of joint is 
 
 70,824 X 100 
 
 iio.ooo =64-39 per cent. 
 
 If the rivets had been of steel the shearing resistance would have 
 been increased to 44,625 pounds per square inch, or .6903 X 3 X 44,625 
 = 92,414 pounds. The shearing resistance of the rivets being greater 
 than the strength of net section of plate, we then have as the percentage 
 of joint 
 
 84,219 X 100 
 
 TToo^^ = 76.5 per cent. 
 
 Lap-joint with Reinforced Welt. — A modified form of lap-joint 
 combining some of the features of both lap- and butt-joints has a rein- 
 forced welt or strap on one side, as shown in Fig. 86. A central row of 
 rivets secures the three thicknesses of metal ; the strap and plates are 
 further secured by a row of rivets in each plate. 
 
RIVETED JOINTS 
 
90 
 
 BOILERS AND FURNACES 
 
 The central row of rivets must be spaced for tightness under pressure 
 as well as for strength of joint, so that a closer spacing is necessary than 
 in the case of zigzag riveting, whether in double- or triple-riveted joints. 
 
 The outer rivets are twice the 
 Fig. 86. pitch of those in the central row. 
 
 This is a good form of joint, 
 though expensive to make, and 
 for the reason that butt-joints 
 with double welts are simpler and 
 no more expensive than this joint 
 they are commonly preferred. 
 The efficiency of this joint is 
 about the same as a triple-riveted 
 joint for the same arrangement 
 and size of rivets. 
 
 A duplicate of this joint frac- 
 tured the upper plate through the 
 outside line of rivet-holes. Fracture began at end rivet-hole. Efficiency 
 of Joint 90. 1 per cent. 
 
 ZoiUGT' Plate.. 
 
 TABLE XXVIL 
 
 SINGLE-RIVETED LAP-JOINT WITH REINFORCED WELT, ^-INCH IRON AND 
 STEEL PLATES, WATERTOWN ARSENAL. 
 
 Material of plate 
 
 Thickness of plate, nominal, inch 
 
 Thickness of welt-plate, nominal, inch 
 
 Width of test-specimen, inches 
 
 Diameter of rivets, inch 
 
 Material of rivets 
 
 Diameter of holes, punched or drilled, inch 
 
 Number of rivets in each plate 
 
 Pitch of rivets, Fig. 86, outside rows, inches 
 
 Pitch of rivets. Fig. 86, middle row, inches 
 
 Pitch of rivets. Fig. 86, vertically, inches ... .... 
 
 Bearing surface of 9 rivets, square inches 
 
 Shearing area of 9 rivets, square inches 
 
 Tensile strength of plate, pounds per square inch 
 
 Gross sectional area of under plate, square inches .... 
 Net sectional area of under plate through line of rivet- 
 holes, B, square inches ... 
 
 Net sectional area of under plate through line of rivet- 
 holes, C, square inches 
 
 Maximum stress on joint per square inch : 
 
 Tension on gross section of under plate, pounds • . • 
 
 Tension on net section of under plate through line of 
 rivets, B, pounds ... 
 
 Tension on net section of under plate through line of 
 rivets, C, pounds ... 
 
 Compression on bearing surface of 9 rivets, pounds . 
 
 Shearing strain on 9 rivets, pounds 
 
 Efficiency of joint, per cent 
 
 Fracture, similar to Figure . 
 
 Iron. 
 
 Steel. 
 
 12 
 
 Iron. 
 
 .82 D. 
 
 9 
 
 4 
 2 
 
 steel. 
 .94 D. 
 
 9 
 
 4 
 
 2 
 
 4-75 
 
 47,180 
 
 4.69 
 
 3.^6 
 
 6.25 
 53,330 
 
 4-34 
 
 2.77 
 
 2.30 
 
 3-74 
 
 3-32 
 
 34,900 
 
 47,465 
 
 59.100 
 
 89,565 
 
 43,770 
 56,640 
 34,660 
 
 74 
 
 86 
 
 62,050 
 32,960 
 67,320 
 
RIVETED JOINTS 9 1 
 
 Another joint similar to the preceding, except that the welt-plate was 
 ^-inch steel and the ^-inch rivets of iron in 0.93-inch drilled holes, 
 failed by shearing 8 rivets in both rows and tearing out one hole in the 
 corner of the ^-inch welt-plate. Efficiency of joint 87.8 per cent. 
 
 Butt-joints. — This form of joint is commonly used for plates ^ inch 
 thick and over. It is the lap-joint repeated. The plates to be joined are 
 placed edge to edge with outer and inner welt-strips, the whole secured 
 together by through-going rivets, as shown in Fig. 87. Butt-joints may 
 be single-, double-, triple-, and quadruple-riveted, according to the thick- 
 ness of plates and steam pressure to be carried. This kind of a joint 
 has an advantage over ordinary lap-joints in the fact that the pull of the 
 joint is in the direction of the centres of the plates, and that the rivets 
 are in double instead of single shear. 
 
 In Table XXVIII. no iron plates are given, as butt-joints are seldom 
 used for plates thinner than inch. Plates of this thickness and thicker 
 are almost invariably made of mild steel. There is a choice of rivet 
 material without loss of efficiency for the diameter and pitch of rivets, as 
 given in the table. 
 
 The strength of a double-welt butt-joint, triple-riveted, as in Fig. 87, 
 plates and rivets of mild steel, may be calculated thus : 
 
 Thickness of plates, ^ inch 625 inch. 
 
 Diameter rivet-hole, IyV inches 1.0625 inches. 
 
 Area of rivet-hole 8866 square inch. 
 
 Pitch of rivets, 7^ inches 7-375 inches. 
 
 The spacing of the joint shows i rivet in single and 4 rivets in double 
 shear. 
 
 The strength of the whole plate would be 7.375 X -625 X 55, 000 
 = 253,516 pounds. 
 
 The strength of net section of plate at outer row of rivets would be 
 (7-375 — 1.0625) X .625 X 55.000 = 216,992 pounds. 
 
 The strength of i rivet in single shear. Table X 39,299 
 
 Four rivets in double shear 298,672 
 
 Total shearing resistance 337, 97 1 
 
 It will be seen that the shearing resistance of the rivets is greater 
 than the strength of net section of plate. We have, then, 
 
 216,992 X 100 
 
 253.516 = ^5.59 per cent. 
 
 ' ' Where butt-straps are used in the construction of marine boilers, 
 the straps for single butt-strapping shall in no case be less than the 
 thickness of the shell-plates ; and where double butt-straps are used, the 
 thickness of each shall in no case be less than five-eighths (^) the 
 thickness of the shell-plates," — U. S. Rule. 
 
 Method of Riveting. — Originally all riveted joints were hand- 
 made, and hand-riveting is still largely practised, because small boiler- 
 
92 
 
 BOILERS AND FURNACES 
 
 shops are not usually equipped with riveting-machines. In certain por- 
 tions of machine-riveted boilers difficult of access, repairs, etc. , are of 
 necessity performed by hand. 
 
 Fig. 87. 
 
 Riveting-machines are now in very general use, and include crank 
 and cam machines, which are not much used ; pneumatic machines, 
 which are very convenient, especially in field work ; steam-riveters, 
 which are extensively used, and hydraulic riveting-machines, which are 
 now especially in favor. Steam and hydraulic machines permit a slow 
 and gradually controlled movement of the ram, and when the rivet-head 
 is formed a pressure can be maintained upon it until it is fully set and 
 
 the rivet sufficiently cooled to permit the withdrawal of the ram without 
 risk of stretching the shank of a hot rivet by the springing apart of the 
 plates. The flow of the metal to fill the hole is usually complete if the 
 pressure is sufficient and not too rapidly applied. The complete filling 
 of the hole is a matter of the utmost importance. Sections from actual 
 plates are shown in Fig. 88, in which the left-hand illustration represents. 
 
RIVETED JOINTS 
 
 93 
 
 ■5-° = 
 
 5^ I a a 
 
 a 
 
 
 3 
 
 ^ 
 
 S 
 
 
 
 
 
 
 
 
 
 =f 
 
 rD 
 
 r/) 
 
 
 
 55 
 
 £L 
 
 '^ 
 
 3 
 
 rr 
 
 
 
 3 
 
 5= 
 
 a 
 
 Cfp 
 
 CfC! 
 
 
 p 
 
 ^ 
 
 a. 
 
 X) 
 
 3 
 
 3 
 
 
 n^ 
 
 
 
 
 
 
 
 
 CfQ 
 
 01 
 
 3 
 
 0) 
 
 
 
 "o' 
 
 
 3" 
 
 lij 
 
 3 
 
 ST- 
 
 n 
 
 Lj 
 
 - 
 
 
 
 2. 
 
 
 c 
 
 rD 
 
 0) 
 
 v; 
 
 
 ^ 
 
 
 
 ^ 
 
 ffC! 
 
 W 
 
 
 
 
 3 
 
 
 
 
 W 
 
 
 ^ 
 
 
 
 
 
 
 3- 
 
 
 
 n 
 
 fD 
 
 
 ^' 
 
 ^ 
 
 
 
 
 3 
 
 
 
 ff<5 
 
 'd 
 
 tf 
 
 
 -v 
 
 X 
 
 ? 
 
 
 
 
 
 
 1 
 S 
 
 09 
 
 3 
 
 
 
 00 
 
 ^ 
 
 
 rr 
 
 ^ 
 
 
 
 
 
 
 
 
 Ui 
 
 
 
 p 
 
 
 
 
 
 3* 
 
 
 
 ci 
 
 01 
 
 
 
 
 
 
 
 
 p 
 
 n) 
 
 
 
 ^ 
 
 [1! 
 
 
 (7 to 
 
 ^ s^^ ^ s:h ^ :H ^ SI" :^ 
 
 Thickness of Plate. 
 
 ^ J- " si:; ^ s'« ^ X ^ 
 
 > 
 
 
 > 
 
 M 
 
 M 
 $0 
 
 J ^ J- " SIS ^ Sis i^ si^ 
 
 > 
 
 
 
 
 W 
 
 Pi 
 
 K 3: ^- ^ ^ SH sH ^ 
 
 CO^^ O^O^O^OlOlCA 
 ^ ^ ^ ^ ^ K ^ 
 
 p 
 
 
 n 
 S 
 
 N 
 
 f 
 
 s 
 
 
 
 tOMMWMMMMW 
 M|N! COIM ColN Si« CslM M| , Co|t- Vm 
 
 Mlm M[<r MlK MS to]-i ml K[i-^ "Ps 
 
 B 
 
 1 
 
 ^ "^ s£ Sr ^ 1^ SI" "^ ^ 
 
 !^ 
 
 00^-<t O^C^O^C/^Cnal 
 ^ ^ ^ ^ ^ ^ ^ 
 
 P 
 
 § 
 
 n 
 
 3 
 
 n 
 
 KStOfOtOtOtOWMl-l 
 
 ^ SI- SI" ^ S'- SIS ^ s^ 
 
 w 
 
 < 
 
 Si "^ s"^ s£ ^ ll -SI" "^ ^ 
 
 y 
 
 M 0^ COOO^^ON 
 
 ^ SI- ^ ^ ^ K ';^ 
 
 r 
 
 ^ 
 
 W 
 
 1. 
 
 3' 
 op 
 
 
 
 r 
 
 H 
 ■Tl 
 
 H 
 
 M 
 
 •^ CT\ai^<ijOJ to M 
 
 K ^ ^ ^ i-\ X >!; 
 
 Vh 
 
 9 
 
 H M ^ \D 0000^ 
 
 " 
 
 3 
 
 n 
 g* 
 5' 
 
 5 
 3' 
 
 OK? 
 
 
 V- 
 
 
 00^ 0NCn-;i-4^0J K) i-H 
 
 ^ X ^ ^ is ^ 
 
 oooooooooooocoooco 
 oi oi oi oi oi c^ a\ p\ a\ 
 
 M Oj Oi io en CO b Ck> -t' 
 
 ^J MVO^J-I^ 10(^/1 M^J 
 
 y y y y y y y y y 
 
 
 C/5 
 
 1 
 ■-0 
 
 n 
 M 
 Z 
 
 > 
 
 PI 
 
 11 
 
 cococococococococo 
 
 Cn Oi Cn Oi 01 Oi p-\ p\ as 
 mCmOi (0(Ln boodo-Pi. 
 
 ^3 ;Td ^d [-d Td 7d _*D 70 ^d 
 
 1 
 
 ^ 
 
94 BOILERS AND FURNACES 
 
 a hand-driven rivet in a punched hole, the middle one a machine-driven 
 rivet in a punched hole, the right-hand one a machine-driven rivet in a 
 drilled hole. Plates should be bolted together through alternate holes, 
 metal to metal in flanged work, to get the spring out of them ; other- 
 wise a thin film of metal is likely to be forced into the space between the 
 plates, reducing frictional contact and lowering the efficiency of the 
 joint. 
 
 Heating Steel Rivets. — It is important that steel rivets be uni- 
 formly heated throughout, and not the points merely, as is the ordinary 
 method of heating iron rivets ; neither should they be heated as highly 
 as iron rivets, and should never exceed a bright cherry-red. Particular 
 attention must be given to the thickness of the fire. Steel, of whatever 
 kind, should never be heated in a thin fire, especially in one having a 
 forced blast, such as an ordinary blacksmith or riveting fire. The reason 
 for this is, that more air passes through the fire than that needed for 
 combustion, and in consequence there is a considerable quantity of free 
 oxygen in the fire, which will oxidize the steel, or, in other words, burn 
 it. If excluded from this free oxygen, steel cannot be burned ; if the 
 temperature is high enough, it can be melted, and will run down through 
 the fire ; but burning is impossible in a thick fire with moderate draft. 
 This is an important matter in using steel rivets, and should not be 
 overlooked. The same principle applies to the heating of steel plates 
 for flanging. 
 
CHAPTER IV. 
 
 WELDING AND FLANGING. 
 
 It has long been the desire of both makers and users of steam 
 boilers that a seamless shell might take the place of the aggregation of 
 plates held together by riveted joints which now constitutes the ordinary- 
 method of constructing a boiler. It is true that cylindrical shells of 
 steam boilers have been constructed by welding and without horizontal 
 and circumferential riveted seams, excepting those fastening the heads 
 and shells together, and even these are not necessary, as many digesters 
 and pressure-tanks are made with the heads welded in place. A recent 
 construction in this country is an internally fired boiler of large dimen- 
 sions, 8^ feet diameter by 27 feet long, having welded horizontal and 
 vertical seams, but with heads riveted in. Such examples are rare, but 
 they serve to show what is possible by this method of construction it 
 proper facilities are at command, coupled with knowledge of how to suc- 
 cessfully use the facilities towards the desired end. 
 
 An advantage to be gained by making the cylindrical shells of boilers 
 seamless is 'that they may be rerolled after welding, producing a per- 
 fectly cylindrical shell. This is, of course, impossible in a riveted joint. 
 So, also, if a shell could be thus welded the objectionable two thick- 
 nesses of plate in the fire would be removed, together with the trouble 
 incident to the accumulation of deposit which is likely to form around 
 the joints and rivet-heads ; further, if there is no jointed seam, the cor- 
 rosion caused by the leakage of the lap-joints or around loose or imper- 
 fectly fitted rivets could not occur. 
 
 Reduction in thickness of plates has been advocated if welded joints 
 be used instead of riveted joints, and such reduction in thickness has 
 been placed as much as one-half; but this is not possible unless a welded 
 joint is known to equal the strength of the original plate, which cannot 
 under any conditions of workmanship be assumed to be true. 
 
 The ordinary claims made for perfectly welded joints are that welding 
 approximates more nearly the original strength of the plates than the 
 best form of riveted joints, relieving the plates from loss of strength due 
 to punching and the additional loss occasioned by drifting and cold 
 hammering. Calking could be dispensed with, and thus relieve the 
 shell of incipient fractures occasioned by bad workmanship. 
 
 W^rought Iron and Mild Steel possess the property of welding 
 when brought into perfect contact while the surfaces are in a state of 
 partial fusion. To further insure a complete contact the surfaces thus 
 joined are pressed, rolled, or hammered until they are united in a single 
 
 95 
 
96 BOILERS AND FURNACES 
 
 piece at the weld. Small quantities of the impurities usually found in 
 wrought iron, such as sulphur, phosphorus, etc., exert a marked influ- 
 ence upon the properties of wrought iron. These foreign substances do 
 not wholly prevent the welding of wrought iron when parts are brought 
 together in a state of fusion, but they do have the effect of lowering the 
 efficiency of the welded joint, especially when the heating and welding 
 is undertaken in an atmosphere containing free oxygen. In a non- 
 oxidizing atmosphere these influences are less marked. What has just 
 been said in regard to wrought iron is also true of mild steel, which does 
 not readily weld except in a non-oxidizing atmosphere. After many 
 trials and many failures in attempting to weld mild-steel boiler-plates 
 Adamson found it necessary to ascertain in all cases the composition of the 
 metal before putting any labor on it, and from a large experience he con- 
 sidered it desirable that the carbon should not exceed ^ of i per cent. , 
 while the sulphur and phosphorus should, if possible, be kept as low as 
 0.04 per cent., silicon being admissible to the extent of o.i of i per cent. 
 Temperature. — A high temperature is essential in welding, ap- 
 proximately 1600° Fahr. for wrought iron ; but different irons require a 
 temperature adapted to each varying composition, because with such 
 variation in composition there is also a variation in point of fusion. 
 Temperature alone is not sufficient for securing the best results. It is 
 true that a high temperature promotes welding in a non-oxidizing atmos- 
 phere, but it is also true that in an atmosphere in which there is free 
 oxygen, the latter, being the cause of burning the metal, not only pre- 
 vents welding, but destroys the strength of the metal wherever it may 
 occur. The temperature, then, must be regulated if welding is to be 
 done in an oxidizing atmosphere, so as to insure the fusion of the metal 
 surfaces to be joined and avoid covering such surfaces with a coating of 
 iron oxide, which will either imperfectly weld or wholly prevent metallic 
 contact. 
 
 Oxide of Iron — The presence of oxide of iron in a joint is one of 
 the principal causes of non-welding. It is difficult to prepare iron plates 
 for welding without the presence of this objectionable material, and it 
 
 is for this reason that upset scarfed 
 ^^^- ^9- edges should have a swell in the 
 
 1 middle of the angular face, as shown 
 
 I in Fig. 89, the object being to bring 
 
 the metal into immediate contact at 
 
 I the middle of the weld, so that any 
 
 subsequent hammering will force out 
 
 the vitreous oxide on either side, securing a better weld than if it were 
 
 allowed to lodge on either surface of the plates, for any lodgement would 
 
 mean defective welding at that point. 
 
 Flux. — As oxidation always occurs in a greater or less degree, the 
 heated surfaces must be protected by means of a flux. The one gen- 
 
WELDING AND FLANGING 
 
 97 
 
 erally used is sand ; this is composed of silicon and oxygen. The action 
 of the flux is twofold, — in forming a vitreous coating over the iron and 
 in reducing the temperature of the parts to which it is applied. This 
 arises from the circumstance that iron is usually "scarfed" at the place 
 where it is to be welded, as in Fig. 89. We thus have a thick and a 
 thinner portion of the same plate exposed to the action of heat. Ordi- 
 narily the thinner portion of the plate is nearest the centre of the fire, 
 consequently it attains welding heat before the thicker portion does. 
 If the action of the heat was not modified in some manner this thinner 
 edge would be burned away long before the thicker portion was brought 
 to the welding point. The sand or other flux coming in contact with 
 the highly heated iron is melted and absorbs so much heat from the iron 
 that it gives the latter a vitreous coating, combining with the iron and 
 covering that portion which is of sufficiently high temperature to melt 
 the sand. Silicon, being very refractory in its nature, will last some time 
 in the fire before it burns off" the iron ; it thus serves to protect the 
 thinner parts of the iron while the thicker portion is absorbing heat and 
 arriving at a welding condition. In using sand as a flux care must be 
 exercised that it be cleaned off the faces of the joint where two scarfed 
 edges are to be welded, because its presence in the weld would prevent 
 perfect contact and weaken the joint. For small work, borax is the flux 
 generally employed in the forge for welding ; it prevents oxidation in 
 the same manner as already described for sand. 
 
 Edges of Plates. — Scarf-welding, shown in Fig. 89, is to be pre- 
 ferred to lap-welding. Fig. 90, because the strain on the scarf-joint is 
 
 direct, while on the lap-joint it is 
 ^^^- 90- indirect and tends to distort the 
 
 joint when under pressure, as in 
 Fig. 91. A scarf-weld is best made 
 
 Fig. 91. 
 
 by upsetting the edges to about double the thickness of the plate and 
 bevelling the edges to about 45°, as shown in Fig. 89. In scarfing 
 and thinning down the plate the sharp edge may be about one-sixteenth 
 of an inch thick, perhaps less. An exact thickness of the upset portion 
 is not a material part of making a good joint, neither is the thinning to 
 a sharp edge of special importance. All that is necessary is the upset- 
 ting of the edge to a thickness considerably more than that of the plate 
 itself, the object being that when the weld is made the plate may then 
 be hammered and finished down to the regular thickness. The edges 
 should be heated simultaneously to a white heat, and when joined the 
 
98 BOILERS AND FURNACES 
 
 joint should be hammered or rolled to secure perfect contact through its 
 whole length ; the swell of the joint can be afterwards worked down to 
 the thickness of the plate. This is a much better method than that 
 shown in Fig. 90, in Bertram's experiments. 
 
 Lap-welded joints as shown in Fig. 90 are not recommended, except 
 for thin plates, say ^ inch and less, because they are weaker than scarf- 
 welded, although there is no reduction of plate section through the 
 joint. This has already been pointed out and illustrated in Fig. 89. 
 The thicker the plates joined together the greater will be the distortion 
 in the joint. This fact was clearly brought out in the test, page 99, 
 which shows the ^-inch joint to be relatively weaker than the ^-inch 
 joint in the same test. 
 
 ^Velding Bars. — If a weld is to be made in a brace or stay, the 
 ends should be upset to about double the original diameter, afterwards 
 bevelling to an angle of about 45°, as in Fig. 89. This form of joint is 
 favorable to the escape of scale, flux, etc., out of the joint. After the 
 parts are joined, the swelled portion of the joint can then be hammered 
 down to the common diameter or size of the bar. Wherever possible, 
 stays and braces should be made without welds. 
 
 Welding Plates. — The heating of two plates in a well-made, open 
 fire is attended with greater risks than in the case of two bars of iron. 
 The reasons are quite obvious : the ends of the bars are easily placed in 
 the centre of the fire and entirely shut off from the injurious effects of 
 free oxygen, if the fire is properly made. When a thick fire is built 
 upon a tweer, the air passing up through it gives up its oxygen to the 
 incandescent carbon, and carbonic acid gas is the product of this union. 
 This gas in passing up through the bed of burning coal takes up another 
 equivalent of carbon, and carbonic oxide gas is formed. Nitrogen is 
 also present in the fire. But none of these gases have an injurious effect 
 on iron, so far as welding is concerned. Therefore, the two bars of iron 
 referred to above are in a highly heated chamber formed by the incan- 
 descent sides and cover of the fire. The included atmosphere being 
 non-oxidizing, the bars may be readily brought to a welding-heat with- 
 out fear of oxidation, for there is no excess of oxygen in the fire to come 
 in contact with the iron. In the case of plates it is somewhat different, 
 for the plates being hottest in the centre and of lower temperature 
 towards the edges of the fire, it is not possible to confine the heated por- 
 tion of the plates to a chamber of heated gases from which oxygen is 
 excluded, — for no such chamber exists in an ordinary fire, and cannot 
 from the nature of the case. Further, every movement of the plate 
 brings the more or less highly heated portions in contact with the air : 
 oxidation instantly occurs, forming an oxide of iron or hard cinder 
 which prevents welding. There is at the same time a partial loss of 
 iron ; but this is not a serious matter in comparison with the bad effects 
 resulting from the presence of oxide of iron in the weld. 
 
WELDING AND FLANGING 
 
 99 
 
 In the manufacture of welded boilers as a business it would be neces- 
 sary to construct a special heating apparatus, which would probably 
 consist of an external and internal gas-furnace, operating on the prin- 
 ciple of a blow-pipe, in which the flames of the burning gas would be 
 directed against such portions of the joint as needed the greater heat. 
 Such an apparatus could be made in which no free oxygen could reach 
 the heated plates, and thus welds could be made without the use of a 
 flux of any kind. The plates could be heated their whole length at one 
 time, and when brought to the point of fusion could be welded by pres- 
 sure instead of by hammering. 
 
 Localizing Heat in Welding. — Welding occurs only at the edges 
 or ends of parts to be thus joined. Heating, therefore, should be confined 
 to such portions only, and not allowed to extend over wide areas upon 
 which no work is to be done. In the absence of special appliances no 
 greater length of edge should be heated than can be conveniently and 
 properly welded, because excessive temperatures occurring where no 
 work is to be done, especially if in contact with the air, is fatal both to 
 iron and steel plates. Without special heating appliances it is probable 
 that not more than a few inches — say less than a foot — could be heated 
 at any one time ; and this heating might preferably begin at the centre 
 of the length of the joint and work from there to either end, rather than 
 begin welding at one end and work towards the other. Whichever 
 method be adopted, there are likely to be strains set up in the plates 
 thus joined, both in extension and compression, which can only be 
 eliminated by proper annealing. 
 
 Bertram's Method. — Boiler-plates were welded at the Woolwich 
 Dockyard in 1857. The edges were scarfed and placed together be- 
 tween two flames directed against either side, as shown in Fig. 92. 
 These flames were obtained by the combus- 
 tion of coal or coke, and were non-oxidizing 
 in their character. When the two plates 
 were raised to the welding temperature they 
 were united by pressure or hammering in a 
 special machine. 
 
 Tests made of these welded joints showed 
 that the lap-welded test-pieces, as in Fig. 
 90, were inferior in strength to those scarf- 
 welded, as in Fig. 89. 
 
 The specimens tested were 4 inches wide 
 t>y ^> ttj and ^ inch in thickness. The 
 lap of the joint was i^ inches, with results 
 as follows : 
 
 The strength of a scarf- welded joint. Fig. 
 89, for the yz-'mch plate was faulty ; but for the ■^- and |/^-inch plates 
 the welds were equal to that of the original plate. 
 
 Fig. 92. 
 
lOO BOILERS AND FURNACES 
 
 The strength oi the lap-welded joint, Fig. 90, was for the ^-inch 
 plate 50 per cent, of the original plate, increasing to 69 per cent, in the 
 YViiich plate, and 66 per cent, in the ^-inch plate. 
 
 From the above data it appears that the strength of joints united by 
 lap-welding is scarcely greater than that secured by single riveting, the 
 joint being about 40 per cent, weaker than the plates which compose it. 
 Scarf-welding, on the contrary, equalled the strength of the plate. No 
 doubt the shape of the joint under severe stress had much to do with the 
 lowering of its strength in consequence of the indirect pull, as shown in 
 Fig. 91. 
 
 Annealing Welded Joints. — It would be advantageous if, in the 
 case of plate work, the whole structure after welding could be placed in 
 an annealing furnace, properly heated, and allowed to cool gradually ; 
 but this is not always practicable. In lieu of this, a strip wide enough 
 to cover any excessive heating on either side of the welded joint should 
 be heated to a cherry-red and then allowed to cool gradually. 
 
 Practical Results. — Experiments on ^^-inch iron plates with 
 welded joints, specimens taken from boiler-shells so as to test the effi- 
 ciency of the welding, showed that of 23 tests 1 1 broke in the weld and 
 12 broke in the solid. The breaking strength of the solid plate was 
 46,368 pounds per square inch for the least strength, 57,792 pounds for 
 the greatest, — an average of 52,865 pounds tensile strength for the 
 whole series, showing that the iron was of good quality. The strength 
 of the welded plates was 36,960 pounds per square inch for least strength, 
 and 53,312 pounds for the greatest, or an average of 46,144 pounds. 
 The efficiency of the welded joints on the total averages was 87.3 per 
 cent. The efficiency of the weakest welded joint as compared with the 
 average strength of the original plate was 68 per cent. , showing through- 
 out the series that the joints varied in efficiency from 68 to 87.3 per 
 cent., which approximates that of good riveting. 
 
 The specifications for the United States protected cruisers ' ' Colum- 
 bia' ' and ' ' Minneapolis' ' called for welded steel pipes, a contract exe- 
 cuted by the Continental Iron- Works. These pipes varied from 10 to 
 20 inches inside diameter, and from ^^ to ^ of an inch thick ; the 
 maximum length was 16 feet, with flanges from ^ to i inch thick after 
 being faced. These pipes were to be made of plate steel, and were to 
 comply with all the requirements called for in the specifications of the 
 material used in the boilers, and be subjected to a water-test of 400 
 pounds per square inch. 
 
 The method of forming these flanged pipes was to scarf the edges, 
 roll to the required diameter, and then weld, the length of each cylinder 
 thus formed being about 8 inches shorter than the finished length of the 
 pipe over the flanges. Steel disks of the diameter and thickness re- 
 quired to form the flanges were punched and flanged like a flue-head, 
 with a cylindrical projection corresponding to that of the pipe and of the 
 
WELDING AND FLANGING 
 
 lOI 
 
 desired length, which, after scarfing in the lathe, was welded circum- 
 ferentially to the pipe. 
 
 The experimental pipe was made of ^-inch plate, as difficulty was 
 expected in making the rivets in the head remain tight under such 
 pressure as would be required to burst a vessel of the same diameter and 
 ^ of an inch thick, and this proved to be true during the course of the 
 experiment. The experimental pipe was 20 inches inside diameter and 
 42 inches long over the flanges. (Fig. 93.) The heads were yi inch 
 
 Fig. 93. 
 
 Fig. 94. 
 
 ? 
 
 yVv'e lct& gLjou7b 
 
 ■•Ci^eCTeiA-'JoIiTt. 
 
 thick, domed about 6 inches, and secured to the pipe by flanges ^ of an 
 inch thick. As the radius of the interior curve on the flange was i 
 inch, the total area to strain the rivet was that of the diameter, or 22 
 inches. The heads were held in place by 34 drilled holes, fitted with 
 rivets, machine-driven, ij^g inches diameter, countersunk on each side. 
 This experimental pipe had a water pressure applied up to 1700 pounds 
 to the square inch, when it failed, as shown in Fig. 94, measurements of 
 which are given in Fig. 95. It enlarged like a barrel, becoming over 
 24 inches in diameter at the middle of its length. The fracture occurred 
 about 6 inches from the line of the longitudinal weld. This pressure is 
 equivalent to a strain of about 68,000 pounds per square inch of section. 
 
 Three test-pieces were cut from the experimental pipe, as shown in 
 Fig. 96, two of them being cut across the welded seam, leaving the 
 welded part in the middle of the length, and the other from the un- 
 welded part of the steel. 
 
 The test of these specimens shows a maximum tensile strength for the 
 two having the weld of 58,230 pounds and 62,500 pounds respectively. 
 
I02 
 
 BOILERS AND FURNACES 
 
 and for the unwelded piece 61,470 pounds, the fracture in all the pieces 
 occurring near the ends several inches from the weld. 
 
 Fig. 95. 
 
 Fig. 96. 
 
 The specimens tested as follows : 
 
 Number of test-mark i 2 3 
 
 Condition Welded Welded No weld 
 
 Length of test-specimen, inches .88 8 
 
 Width, inches 1.105 1.008 i.oio 
 
 Thickness, inches .239 .222 .215 
 
 Area, square inches .243 .224 .217 
 
 Final length, inches 9.50 9.28 9.28 
 
 Elongation, per cent 18.75 16.00 16.00 
 
 Tensional stress on specimen, 
 
 maximum pounds 14,150 14,000 13,340 
 
 Tensional stress per square inch, 
 
 maximum pounds 58,230 62,500 61,470 
 
 At first sight there appears to be a discrepancy between the tensile 
 strength of the metal, as indicated by the testing-machine, and that 
 given from the gauge pressure. A careful examination of the outline of 
 the ruptured vessel, Fig. 95, will, however, show that the stress on the 
 metal at the time of rupture was much greater than would be obtained 
 from a consideration of the original diameter and thickness of the metal. 
 
 The outside circumference from edge to edge of the rupture, of 
 about the middle of its length, is 75y-g- inches, corresponding to an out- 
 side diameter of 24.095 inches. The thickness of the metal at this 
 
WELDING AND FLANGING 103 
 
 point was 0.206 inches, making the internal diameter at the time of 
 
 rupture 24.095 — 2 X .206 = 23.683 inches. Using these figures for 
 
 thickness and diameter, the stress on the metal when the vessel burst 
 
 becomes 
 
 1700 X 23.683 
 f^ 2 X .206 = 97,720 pounds. 
 
 Test-specimen No. i measured approximately 0.765 by 0.17 1 inch at 
 
 point of fracture ; or its area was 0.1308 square inch. The total stress 
 
 on it at the time it broke was 14,150, and, consequently, the stress per 
 
 square inch was 
 
 14,150 
 
 r, = 108,200 pounds. 
 
 Similarly, specimens Nos. 2 and 3 measured 0.775 by 0.165 ^^d 
 0.783 by 0.169 inch respectively, giving stresses at the time of fracture 
 of 109,500 and 100,800 pounds respectively, or an average for the three 
 specimens of 106,167 pounds, — a result near enough to that obtained 
 from the gauge pressure to leave little doubt of the correctness of the 
 latter. 
 
 Furnace Flues. — In an internally fired boiler it is important that 
 the main flue should be truly cylindrical, as the resistance to collapse 
 depends largely upon this. Lap-joints prevent the plates forming a true 
 circle ; it has been the practice, therefore, among the best makers to 
 employ in its construction a butt-riveted joint with the seam below the 
 grates. The objections to this arrangement are, that it is impossible to 
 perfectly calk such a seam when once in place ; if the seam of rivets be 
 along the bottom of the flue, the ready removal of ashes is prevented, 
 and more or less of them will accumulate along the whole length of the 
 furnace. Should there be a leaky joint, and this is not improbable, we 
 may almost certainly count on an accumulation of hard-baked ashes and 
 cinders, which lend themselves readily to surface corrosion. The best 
 practice at this time is to make all such furnace flues with welded joints 
 and flanged ends, placing the weld below the grate bars, so as to be 
 away from the fire ; the pressure, being wholly on the outside, tends to 
 collapse, and thus to tighten the weld. An imperfect weld might, in 
 such a flue, escape detection for a long time, but would soon make 
 itself apparent in any case where internal pressures were employed. 
 
 Strength of Welded Joints. — Theoretically, the parts joined by 
 welding should be equal in strength to the original bars or plates, and 
 many tests have'proven this to be the case. Were this true of a majority 
 of welded joints, the claims of superiority for such joints over riveted 
 joints would be realized ; unfortunately, this is not the case in common 
 practice. 
 
 The weakness of a welded joint is in part due to the unequal heating 
 of the edges to be joined and the absence of immediate contact when 
 
I04 BOILERS AND FURNACES 
 
 the parts are brought together for welding, whether by hammers, rollers, 
 or simple pressure. In the case of bars, metallic contact is usually had 
 in welding, except in large pieces necessarily handled by a crane. The 
 strength of welded bars of iron varies from 35 to 90 per cent. , with an 
 occasional bar showing full strength. In tests of iron chain cables in 
 which a weld occurs in each link the failure almost invariably occurs in 
 the weld, the best percentages of strength ranging anywhere from 70 to 
 90 per cent, of the original bar. A chain-cable link does not present the 
 most favorable conditions for welding, but the workshop appliances and 
 unusual skill on the part of the workman who is really a specialist make 
 the average of chain-riveting equal to, if not superior to, ordinary forge- 
 work where welding is not a continuous practice. 
 
 The uncertainty in regard to welded seams is equally shared by 
 makers and users of steam boilers. So far as experimental tests have 
 gone, welded seams do not average higher than single- or double-riv- 
 eted joints, say from 60 to 70 per cent, of the strength of the original 
 plate, which latter is, no doubt, injured by overheating at the time of 
 making the weld. Test-sections cut from portions thus overheated show 
 extreme brittleness and a reduction in tensile strength of the original 
 plate of from 50 to 75 per cent., — a very serious loss, some of which 
 may be restored by judicious annealing ; but furnaces large enough to 
 admit the completed shell of a boiler for the purpose of annealing must 
 be rare indeed, if they exist at all, in this country. 
 
 Welded-joint fractures do not always occur immediately in the joint, 
 though such failure is known to be due to loss of strength occasioned by 
 destructive treatment which the bar or plate receives in the operation of 
 heating preparatory to welding. This loss of tensile strength, although 
 occurring outside of the joint, is directly chargeable to the process of 
 welding, even though the break did not occur in the weld itself. In a 
 number of tests of welded bars, the break occurred both in and along- 
 side of the weld, showing in each case from 15 to 40 per cent, reduction 
 of strength. It is for reasons similar to these that thoroughly prepared 
 boiler specifications require that stays and braces, whether of iron or ol 
 steel, intended for steam boilers shall not be welded if they are to be 
 subjected to tensile stress. This exclusion of welded stays and braces 
 accords with sound judgment, based upon a varied and sometimes disas- 
 trous experience. 
 
 Efficiency. — Welded seams in boiler-plates have not ordinarily 
 yielded an efficiency higher than could have been supplied by double- or 
 triple-riveted joints, or such a joint as a designer would have selected as 
 suitable for the diameter of boiler, thickness of plate, and steam pressure 
 to be carried. 
 
 Cost. — The relative cost of welding, as compared with punching and 
 riveting, is probably in favor of welding when proper facilities are pro- 
 vided for heating, scarfing, hammering, annealing, etc. 
 
WELDING AND FLANGING 
 
 105 
 
 Flanging. — The process of bending the edge of a plate so that a 
 second plate may be fastened to it at an angle, such as a boiler-head, 
 Fig. 97, or if a second plate is at a distance from the first plate, as in 
 
 Fig. 97. 
 
 Fig 
 
 Fig. 99. 
 
 Fig. 100. 
 
 the bottom of a fire-box, Fig. 98, is called flanging. The advantages 
 of a flanged joint over that in which the parts are dimensioned and se- 
 cured by angle-iron corners, riveted as in Fig. 99, are that it is stronger, 
 simpler, because consisting of fewer parts, and less liable to leakage, 
 because there is one less riveted seam. In localities where flue boilers 
 are popular, some very intricate flanging is re- 
 quired, many of these flues being as small as 6 
 inches, see Fig. 100, and few of them larger than 
 18 inches in diameter. 
 
 Heating of Plates.— Flanging should always 
 be done at a bright-red heat, and work upon a 
 plate should never be continued after it has cooled 
 down below a dark cherry-red. Hand-flanging 
 should always be performed over a cast-iron 
 former with rounded corners. In heating a plate 
 for hand-flanging, the heat should be confined to 
 such portions of the edge of the plate as are to be 
 flanged, and should not extend very far inward 
 towards the centre. As much length of edge of 
 the plate should be heated as can be conveniently 
 flanged at one operation. This will require a 
 special construction of fire, needing only a few 
 bricks to give the necessary boundary to the 
 enclosed fire and a proper arrangement of the 
 fire itself, so that the entire edge can be inspected 
 at any time during the progress of the heating 
 of the plate. It seems almost needless to remark 
 that the fire must be uniformly hot, absolutely clean, and free from 
 ashes or clinkers, and must on no account have holes in it by which 
 the blast can escape, or a burnt plate will surely result. 
 
Io6 BOILERS AND FURNACES 
 
 Hand- Flanging. — The flanging of an iron or steel plate should be 
 done with wooden mauls, bending the plate over a cast-iron former. 
 The blows should be quick, light, and distributed over as large a surface 
 as possible in the shortest time, avoiding anything like short bends in 
 turning the flange. The heating, when done in an ordinary open flange 
 fire, must of necessity be local ; there will be required, therefore, the 
 greatest care in working. As the flanging approaches completion by 
 successive stages of heating and hammering, care must be exercised 
 that the plate, if of steel, is not ruined by splitting or cracking, which 
 may be induced by internal strains. 
 
 The operation of hand-flanging is one requiring skill and judgment 
 on the part of the workmen, — first, in the matter of heating, during 
 which, if done in an open fire, as is usually the case, there is liability of 
 overheating portions of the plate while other portions are not hot enough 
 to insure the best working ; second, in working down a hot sheet to the 
 edge of the cast-iron former, the flange must be left true and accurate, 
 free from lumps and wavy edges. Should the latter occur, the flange 
 must be reheated and worked down with sledges and flatters until the 
 entire edge is true, even, and accurately dimensioned. Attention must 
 also be given to the flat portion of a flanged plate, to see that it is in 
 presentable as well as in workable condition. Buckling is likely to 
 occur, because the operation of flanging the edges subjects them to 
 alternate strains of compression and elongation, much of which is trans- 
 mitted to the centre portion of the plate. 
 
 Machine-Flanging. — There is a certain advantage in machine- 
 flanging over hand-flanging in the one fact that the whole plate is 
 heated in a special furnace, insuring a moderate and even temperature 
 throughout. In hydraulic flanging the machine acts quickly by press- 
 ure, performing the entire operation upon a single plate in two or 
 three minutes' time and without striking any blows, the metal flowing 
 easily, naturally, and without abnormal strains into whatever shape the 
 dies may give it. The centre of the plate is flat, and, on the whole, a 
 much more satisfactory product than can be secured by hand. 
 
 Flanging-machines in which the edge of a circular plate is turned 
 by means of rollers to a right angle turn out excellent work ; but as 
 such machines are confined to circular plates only, they are not adapted 
 to general work, such as flanging irregular sheets, making flanged 
 fire-door openings, flanged manholes, etc. 
 
 Hydraulic flanging-machines are more in favor at this time than 
 any other, and, together with hydraulic riveting-machines, are now 
 considered a necessary requisite in a first-class boiler-shop. 
 
 Hand- vs. Machine-Flanging. — Fig. loi represents the thinning 
 of the curve occasioned by the stretching of the plate over the cast-iron 
 former in hand-flanging, the dotted line representing the normal curve 
 and the middle line the actual thickness of metal. Fig. 102 is a repre- 
 
WELDING AND FLANGING 
 
 107 
 
 sentation of the thickening of the curve, taken from a roller-machine- 
 flanged head. The normal curve, it will be noticed, falls considerably 
 within the actual line of the metal. The advantages gained by the 
 strengthening of the head at that particular 
 point are quite obvious and are not hkely to 
 be underestimated. 
 
 Fig. 103. 
 
 Fig. ioi. 
 
 Fig. 102. 
 
 Flanging Iron Plates. — Iron plates are more severely tested by 
 the act of flanging than by any other work done upon them. Iron, by 
 reason of its fibrous structure, requires careful manipulation to prevent 
 breaking in the bend, especially if the corner be too sharp, as is often 
 the case. Mr. Allen called attention to this defect in flanging several 
 years ago, supplemented by a sketch which is reproduced in Fig. 103, 
 assigning the sharp curve and consequent distortion of metal in the bend 
 as a direct cause of grooving. 
 
 Radius of Flange. — A defect more frequently met with in old than 
 in later flanging is in the sharpness of the bend of the flange. Ordinary 
 tubular boiler-heads, machine-flanged for the trade, have an inside radius 
 of flange for heads y\ to f inch thick, — i inch for the former to i^ inches 
 for the latter, — nearly every mill that furnishes flanged heads having 
 a radius of its own. Although there is no uniform standard for inside 
 curves, they will generally be found to be ample for the thickness of the 
 head and with no sharp corners. As the outer diameter of any tube is 
 not likely to come nearer the outside of a flanged head than 3 inches, 
 there is no reason for adhering to sharp curves ; in fact, the large curve 
 adds much to the stiffness of the head and reduces the area of flat sur- 
 face to be fitted with stays. 
 
 When it is known that only the neutral axis of the plate does not 
 change length in bending, that the outside surface must suffer extension 
 and the inside surface compression, the advantage of a large radius in 
 bending is made immediately apparent. Referring to Fig. 103, sketched 
 from an actual specimen of flanged iron plate, it will be seen that the 
 
I08 BOILERS AND FURNACES 
 
 outer layers of metal separated and slid upon each other, as indicated 
 by the transverse lines ; the outer surface was filled with small cracks 
 not unlike the season checks as seen in timber. The inside of the 
 flange, as shown in the engraving, being in a state of undue compres- 
 sion, presents the appearance of a crushed and buckled-up mass ot 
 fibres, and it is particularly this disturbance of the fibres and the laminae 
 of the iron which renders it susceptible to the corrosive action of the 
 acids present in the feed-water, which, together with the strains pro- 
 duced by expansion and contraction incident to the combined action of 
 heat and pressure, result in corrosion, grooving or channelling of the 
 flange, and possibly rupture. An examination of the engraving will 
 show that the inner radius of the flange is but little more than the thick- 
 ness of the plate. 
 
 Flanging Steel Plates. — Mild steel requires uniform heating, 
 moderate curves, and gentle working to get the best results. This is 
 best secured in machine-flanging, whether by rollers or by gently 
 forcing the plate through a die by hydraulic pressure. The process ot 
 the flanging of mild steel need not be in any respect different from 
 what has already been described relative to the subject in general. 
 
 Annealing Steel Plates. — After flanging a steel plate, whether by 
 hand or machine, it should be immediately heated to a cherry-red to 
 relieve it of all internal strains incident to working, allowing it to cool 
 slowly, not disturbing it until entirely cold. 
 
 Thickness of Flanged Edges. — When a flat disk of metal is 
 flanged, as in the case of a boiler-head, the edges of the flange, espe- 
 cially in hand-flanging, will be slightly thicker than the original plate. 
 In machine-flanging, whether by rollers or by passing the plate through 
 a die, the edges are not usually any thicker than other portions of the 
 plate, except the excess due to the space allowance in the dies above 
 the nominal thickness of the plate. This extra thickness is due to the 
 longer circumference of the flat disk being compressed to the shorter 
 circumference corresponding to the diameter of the finished head. The 
 accumulated metal is worked down to the original thickness of plate, 
 which causes an increase of depth of flange. Sexton gives the follow- 
 ing practical instructions regarding flanging : "A plate during the pro- 
 cess of flanging will gain twice its thickness in length of each flange. 
 Thus, suppose you want to flange a circular plate to have a 3-inch 
 flange all around, and to be 3 feet in diameter after being flanged and 
 ^ inch thick ; you must not add twice the width of the flange to the 
 diameter, making 3 feet 6 inches, but twice the width of the flange, less 
 four times the thickness, — making 3 feet 4.}4. inches. In marking the 
 plate line out the exact diameter you want it to be after being flanged, 
 then allow the width of the flange less twice the thickness, and when 
 flanged the centre marks should be on the flange just where the curve 
 joins the flat." 
 
 i 
 
WELDING AND FLANGING IO9 
 
 Flanging for flue-holes, manholes, etc., has just the opposite effect 
 to that given in the preceding paragraph. Preparatory to flanging an 
 opening in a plate, the whole depth of flange must be allowed when 
 laying off the work. An inner hole is cut to the inside line of flange 
 thus laid off, the plate heated, and the flange formed by forcing the 
 metal outward from the plate in the desired direction, — as, for example, 
 Fig. 100, which shows a flue-hole flanged to the inside of head, or as in 
 Fig. 216, which shows a flange turned to the outside of head. As the 
 diameter of the finished flange for flue-hole or manhole is worked from 
 the diameter and thickness of the plate, it follows that the edge of the 
 flange is thinner than that of the plate, the decrease in thickness de- 
 pending upon the depth of flange to be formed. 
 
CHAPTER V. 
 
 DETAILS AND STRENGTH OF CONSTRUCTION. 
 
 The cylindrical shell of a steam boiler is commonly an aggregation 
 of plates fastened together by riveted joints. These plates may be 
 either wrought iron or mild steel, the thickness depending upon the 
 diameter of the shell and the pressure to be carried. When boiler- 
 shells were made of iron the plates were much narrower arid shorter 
 than can now be had in mild steel, because steel plates have no grain or 
 fibre, but are homogeneous throughout, and can, therefore, be rolled 
 lengthwise or crosswise as best suits the mill-man, which was not the case 
 with wrought iron, as its fibrous nature had to be taken into account. 
 
 No difficulty is now experienced in getting mild steel plates of any 
 size that may be economically handled or worked in a boiler-shop, thus 
 effecting a reduction in the number of plates and number of joints in 
 boiler-shells of at least one-half Boiler-shells of 36 inches diameter by 
 8 feet in length can now be made of a single plate with one row of rivets 
 if desired ; boilers 60 inches in diameter and less are commonly made of 
 2 sheets with 2 horizontal rows of riveted joints, forming the shell up 
 to 16 or 18 feet in length ; boilers 72 inches in diameter are commonly 
 made with a smgle sheet in the bottom and 2 or 3 sheets forming the 
 upper half of the boiler for lengths up to 20 feet. In each of these 
 arrangements no cross-seams are exposed to the fire, but in all cases 
 where the upper and lower sheets of the boiler meet midway of the 
 circumference there will be exposed to the action of the heat, but not 
 to the direct action of the fire, 2 rows of riveted joints, unless the bottom 
 sheet is made wide enough to extend around to the ^-heating-surface 
 line, which is not a very good arrangement of spacing plates. Other 
 shell designs are made up of 3 rings, each in a single piece, the length 
 of the sheet reaching around the boiler and including the riveted joint ; 
 the 3 widths thus riveted together make the required length of the 
 boiler, usually not more than 20 feet. 
 
 Strength of Riveted Shell. — Wrought-iron boiler-plates should 
 average 45,000 pounds and mild steel 55,000 pounds 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 insertion of rivets, 
 so that for single-riveted joints the net strength for plates -^ inch thick 
 is about 60 per cent. In double-riveted joints the net strength of the 
 same thickness of plate is about 71 per cent. In triple-riveted joints for 
 the same thickness of plate the net strength is about 79 per cent. A 
 butt-joint triple-riveted, with outside and inside welt-strips, would have 
 about 86 per cent, of the strength of the original plate. 
 
DETAILS AND STRENGTH OF CONSTRUCTION III 
 
 Safe Working Pressure : Rule. — Multiply together the tensile 
 strength of the plate, the thickness of the plate in parts of an inch, and 
 the efficiency of the joint ; divide the product thus obtained by one-half 
 the diameter of the boiler multiplied by the factor of safety. 
 
 Example •' What is the safe working pressure for the cylindrical 
 
 shell of a boiler 72 inches in diameter, made of 55,000 pounds tensile 
 
 strength steel plates, y'g- inch thick, butt-joints with outside and inside 
 
 welt- strips, the efficiency of which is assumed to be 86 per cent, that of 
 
 the original plate, factor of safety 5 ? 
 
 55,000 X .4375 X .86 
 
 36 X 5 = "4-97 pounds. 
 
 Example 2 : Shell 48 inches in diameter, y^-inch plate, steel 55,000 
 pounds tensile strength, iron rivets, double-riveted lap-joint, assumed to 
 have 67 per cent, of the efficiency of the original plate, factor of safety 5 ? 
 
 55,000 X .3125 X .67 
 
 2^ X 5 ^ 9^-9 pounds. 
 
 In the first example, by referring to Table XXVIII. it will be seen 
 that for steel plates, whether rivets are of iron or of steel, there is an 
 excess of strength in the rivets : the joint would probably fail by tearing 
 the plate through the hne of rivet-holes, as in Fig. 68. 
 
 In the second example, by referring to Table XXV. it will be seen 
 that for steel plates and iron rivets the percentage of joint is less in the 
 proportion of 67 to 71 for the same spacing and diameter of rivet, — i.e., 
 ■^-inch rivets on 2^-inch centres. By referring to Table VII., page 59, 
 it will be seen that the shearing strength of a ^^-inch iron rivet com- 
 pletely driven to fill a ^-inch hole is 15,110 pounds. Turning now to 
 Table X. , page 60, the shearing strength of a steel rivet similarly di- 
 mensioned is found to be 19,715 pounds. If the joints were tested to 
 rupture, the steel plate and iron rivets would probably fail by shearing 
 the rivets, as in Fig. 58. If steel rivets were used, the failure would 
 occur by tearing the plate across the line of rivet-holes, as in Fig. 59. 
 
 Tables of Working Pressures. — In the following tables. No. 
 XXIX. gives the calculated working pressure for double- riveted lap- 
 joints for iron shells and iron rivets, steel shells and iron rivets, steel 
 shells with steel rivets, for the ordinary diameters of shell from 36 to 72 
 inches, and for the thickness of plates usually employed in such diameters, 
 — i.e., % \.o Yi inch. A factor of safety of 5 is employed in this table. 
 
 Table XXX. contains calculated working pressures when similar 
 cylindrical shells are triple-riveted. A factor of safety of 5 is employed 
 in this table. 
 
 Table XXXI. contains working pressures calculated for diameters 
 from 36 to 120 inches and for plates from ^ inch to ^ inch in thick- 
 ness. These include iron and steel shells with iron and steel rivets ; 
 the longitudinal seams have butt-joints triple-riveted. The working 
 pressure is based upon a factor of safety of 5. 
 
112 
 
 BOILERS AND FURNACES 
 
 TABLE XXIX. 
 
 WORKING PRESSURES FOR CYLINDRICAL SHELLS OF STEAM BOILERS, 
 JOINTS, DOUBLE-RIVETED. 
 
 Factor of Safety, 5. 
 
 Diameter. 
 
 Thickness. 
 
 Iron Shell, Iron 
 Rivets. 
 
 Steel Shell, Iron 
 Rivets. 
 
 Steel Shell, Steel 
 Rivets. 
 
 36 
 
 a 
 
 91 
 112 
 
 Ill 
 128 
 
 Ill 
 137 
 
 38 
 
 {'i 
 
 86 
 106 
 
 105 
 121 
 
 105 
 129 
 
 40 
 
 
 82 
 
 lOI 
 
 100 
 115 
 
 100 
 123 
 
 42 
 
 
 78 
 96 
 
 95 
 no 
 
 95 
 1x7 
 
 44 
 
 {X 
 
 74 
 91 
 
 91 
 105 
 
 91 
 112 
 
 46 
 
 {X 
 
 71 
 
 87 
 
 87 
 
 100 
 
 87 
 107 
 
 48 
 
 {ft 
 
 84 
 99 
 
 96 
 107 
 
 102 
 121 
 
 50 
 
 {% 
 
 81 
 95 
 
 92 
 103 
 
 98 
 116 
 
 52 
 
 {ft 
 
 77 
 92 
 
 89 
 
 99 
 
 95 
 112 
 
 54 
 
 /A 
 
 li 
 
 85 
 96 
 
 91 
 108 
 
 56 
 
 
 72 
 85 
 
 82 
 92 
 
 88 
 104 
 
 58 
 
 {ft 
 
 59 
 
 82 
 
 79 
 89 
 
 85 
 100 
 
 60 
 
 {ft 
 
 67 
 79 
 
 V, 
 
 82 
 97 
 
 62 
 
 {.^ 
 
 77 
 
 88 
 
 83 
 92 
 
 94 
 108 
 
 64 
 
 {« 
 
 it 
 
 81 
 89 
 
 91 
 105 
 
 66 
 
 {« 
 
 72 
 83 
 
 78 
 87 
 
 88 
 102 
 
 68 
 
 {-^ 
 
 70 
 81 
 
 t 
 
 86 
 99 
 
 70 
 
 {r; 
 
 68 
 
 78 
 
 74 
 82 
 
 83 
 96 
 
 72 
 
 iT6 
 
 66 
 
 76 
 85 
 
 72 
 
 81 
 
 93 
 104 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 113 
 
 TABLE XXX. 
 
 WORKING PRESSURES FOR CYLINDRICAL SHELLS OF STEAM BOILERS, LAP- 
 JOINTS, TRIPLE-RIVETED. 
 
 Factor of Safety, 5. 
 
 Diameter. 
 
 Thickness. 
 
 Iron Shell, Iron 
 Rivets. 
 
 Steel Shell, Iron 
 Rivets. 
 
 Steel Shell, Steel 
 Rivets. 
 
 36 
 
 
 100 
 124 
 
 121 
 139 
 
 123 
 151 
 
 38 
 
 a 
 
 95 
 
 117 
 
 "5 
 132 
 
 116 
 144 
 
 40 
 
 {'i 
 
 90 
 112 
 
 109 
 125 
 
 IIO 
 
 136 
 
 42 
 
 {i 
 
 86 
 106 
 
 104 
 119 
 
 105 
 130 
 
 44 
 
 {'i 
 
 83 
 
 lOI 
 
 99 
 114 
 
 100 
 
 124 
 
 46 
 
 a 
 
 79 
 97 
 
 95 
 109 
 
 96 
 119 
 
 48 
 
 {ft 
 
 93 
 no 
 
 104 
 118 
 
 114 
 135 
 
 50 
 
 {% 
 
 89 
 
 106 
 
 100 
 113 
 
 109 
 129 
 
 52 
 
 {ft 
 
 86 
 102 
 
 96 
 109 
 
 105 
 124 
 
 54 
 
 {ft 
 
 % 
 
 93 
 105 
 
 lOI 
 
 120 
 
 56 
 
 {ft 
 
 80 
 95 
 
 89 
 
 lOI 
 
 97 
 116 
 
 58 
 
 {ft 
 
 77 
 91 
 
 86 
 98 
 
 94 
 112 
 
 60 
 
 {ft 
 
 74 
 88 
 
 83 
 95 
 
 91 
 
 108 
 
 62 
 
 {t 
 
 85 
 98 
 
 92 
 103 
 
 104 
 120 
 
 64 
 
 it 
 
 83 
 95 
 
 89 
 
 TOO 
 
 lOI 
 
 117 
 
 66 
 
 {t 
 
 80 
 93 
 
 86 
 97 
 
 98 
 113 
 
 68 
 
 it 
 
 78 
 90 
 
 84 
 94 
 
 95 
 no 
 
 70 
 
 {t 
 
 76 
 
 87 
 
 81 
 91 
 
 92 
 107 
 
 72 
 
 « 
 
 97 
 
 79 
 
 90 
 104 
 117 
 
114 
 
 BOILERS AND FURNACES 
 
 TABLE XXXL 
 
 WORKING PRESSURES FOR CYLINDRICAL SHELLS OF STEAM BOILERS, BUTT- 
 JOINTS, TRIPLE-RIVETED. 
 
 Factor of Safety, 5, 
 
 Diameter. 
 
 Thickness. 
 
 Iron Shell, Iron 
 Rivets. 
 
 Steel Shell, Iron 
 Rivets. 
 
 Steel Shell, Steel 
 Rivets. 
 
 36 
 
 « 
 
 108 
 161 
 
 134 
 197 
 
 Ill 
 
 197 
 
 38 
 
 a 
 
 102 
 128 
 152 
 
 127 
 156 
 187 
 
 127 
 187 
 
 40 
 
 a 
 
 97 
 121 
 
 145 
 
 120 
 148 
 178 
 
 120 
 148 
 178 
 
 42 
 
 a 
 
 138 
 
 115 
 141 
 169 
 
 115 
 141 
 169 
 
 44 
 
 a 
 
 89 
 
 no 
 
 132 
 
 109 
 135 
 161 
 
 109 
 
 46 
 
 a 
 
 85 
 
 106 
 126 
 
 105 
 129 
 154 
 
 105 
 129 
 
 154 
 
 48 
 
 
 lOI 
 121 
 
 141 
 
 124 
 148 
 172 
 
 lit 
 
 172 
 
 50 
 
 1a 
 
 116 
 
 135 
 
 119 
 
 142 
 165 
 
 119 
 142 
 165 
 
 52 
 
 11 
 
 93 
 III 
 130 
 
 114 
 137 
 159 
 
 114 
 137 
 159 
 
 54 
 
 11 
 
 90 
 107 
 125 
 
 no 
 132 
 153 
 
 no 
 132 
 ^53 
 
 56 
 
 It 
 
 87 
 103 
 121 
 
 106 
 127 
 148 
 
 106 
 127 
 148 
 
 58 
 
 (ft 
 
 84 
 100 
 117 
 
 102 
 123 
 142 
 
 102 
 123 
 
 142 
 
 60 
 
 11 
 
 81 
 97 
 III 
 128 
 
 138 
 157 
 
 i?i 
 138 
 157 
 
 62 
 
 f/8 
 
 Is 
 
 93 
 109 
 124 
 
 115 
 
 133 
 152 
 
 115 
 133 
 152 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 TABLE ^'KX.h— Continued. 
 
 115 
 
 Diameter. 
 
 Thickness. 
 
 Iron Shell, Iron 
 
 Steel Shell, Iron 
 
 Steel Shell, Steel 
 
 
 
 Rivets. 
 
 Rivets. 
 
 Rivets. 
 
 
 \H 
 
 90 
 
 Ill 
 
 Ill 
 
 64 
 
 h 
 
 106 
 
 129 
 
 129 
 
 120 
 
 147 
 
 147 
 
 
 I A 
 
 135 
 
 165 
 
 165 ■ 
 
 
 \% 
 
 88 
 
 108 
 
 108 
 
 66 
 
 ^J 
 
 102 
 
 125 
 
 125 
 
 
 1^ 
 
 117 
 
 143 
 
 143 
 
 
 Ia 
 
 131 
 
 160 
 
 160 
 
 
 
 '¥ 
 
 85 
 
 105 
 
 105 
 
 68 
 
 
 
 99 
 113 
 
 121 
 138 
 
 121 
 138 
 
 
 
 .A 
 
 127 
 
 155 
 
 155 
 
 
 {Vi 
 
 83 
 
 102 
 
 102 
 
 70 
 
 k 
 
 97 
 no 
 
 118 
 134 
 
 118 
 134 
 
 
 I A 
 
 123 
 
 151 
 
 151 
 
 
 r^ 
 
 80 
 
 99 
 
 99 
 
 
 Tf 
 
 94 
 
 115 
 
 115 
 
 72 
 
 i>^ 
 
 107 
 
 131 
 
 131 
 
 
 ba 
 
 120 
 
 147 
 
 147 
 
 
 134 
 
 163 
 
 163 
 
 
 
 'tV 
 
 90 
 
 no 
 
 no 
 
 75 
 
 
 ^ 
 
 102 
 
 125 
 
 125 
 
 
 
 115 
 
 141 
 
 141 
 
 
 
 .>^ 
 
 128 
 
 157 
 
 157 
 
 
 Id 
 
 87 
 
 106 
 
 106 
 
 78 
 
 99 
 
 121 
 
 121 
 
 
 TF 
 
 III 
 
 135 
 
 135 
 
 
 l>i 
 
 123 
 
 151 
 
 151 
 
 
 r-rV 
 
 83 
 
 102 
 
 102 
 
 81 
 
 - ¥ 
 
 95 
 
 116 
 
 116 
 
 
 
 107 
 
 130 
 
 130 
 
 
 -^ 
 
 119 
 
 145 
 
 145 
 
 
 
 '}4 
 
 92 
 
 112 
 
 112 
 
 84 
 
 
 
 103 
 
 126 
 
 126 
 
 
 s^ 
 
 115 
 
 140 
 
 140 
 
 
 
 H 
 
 126 
 
 158 
 
 158 
 
 
 
 .U 
 
 137 
 
 167 
 
 167 
 
 
 r>^ 
 
 89 
 
 108 
 
 108 
 
 87 
 
 • ft 
 
 99 
 III 
 
 121 
 135 
 
 121 
 135 
 
 
 ii 
 
 121 
 
 148 
 
 148 
 
 
 Vh 
 
 132 
 
 162 
 
 162 
 
 
 -% 
 
 86 
 
 105 
 
 105 
 
 
 
 96 
 
 117 
 
 117 
 
 90 
 
 - H 
 
 107 
 
 131 
 
 131 
 
 
 Iff 
 
 117 
 
 143 
 
 143 
 
 
 128 
 
 156 
 
 156 
 
ii6 
 
 BOILERS AND FURNACES 
 TABLE XXXL - Continued. 
 
 Diameter. 
 
 Thickness. 
 
 Iron Shell, Iron 
 Rivets. 
 
 Steel Shell, Iron 
 Rivets. 
 
 Steel Shell, Steel 
 Rivets. 
 
 
 
 '% 
 
 83 
 
 lOI 
 
 lOI 
 
 
 
 
 93 
 
 114 
 
 114 
 
 " 93 
 
 i 
 
 H 
 
 103 
 
 126 
 
 126 
 
 
 
 In 
 
 114 
 
 139 
 
 139 
 
 
 
 124 
 
 151 
 
 151 
 
 
 r^ 
 
 80 
 
 98 
 
 98 
 
 
 A 
 
 90 
 
 no 
 
 no 
 
 96 
 
 \n 
 
 100 
 
 123 
 
 123 
 
 
 w 
 
 no 
 
 134 
 
 134 
 
 
 VH 
 
 120 
 
 146 
 
 146 
 
 
 {% 
 
 78 
 
 95 
 
 95 
 
 
 -h 
 
 87 
 
 107 
 
 107 
 
 99 
 
 \yi 
 
 97 
 
 119 
 
 119 
 
 
 
 107 
 
 130 
 
 130 
 
 
 ¥ 
 
 116 
 
 142 
 
 142 
 
 
 fi 
 
 75 
 
 92 
 
 92 
 
 
 85 
 
 104 
 
 104 
 
 102 
 
 
 94 
 
 115 
 
 115 
 
 
 1 xi 
 
 104 
 
 127 
 
 127 
 
 
 \.H 
 
 113 
 
 138 
 
 138 
 
 
 [% 
 
 73 
 
 90 
 
 90 
 
 
 ft 
 
 82 
 
 lOI 
 
 lOI 
 
 105 
 
 92 
 
 112 
 
 112 
 
 
 \\ 
 
 lOI 
 
 123 
 
 123 
 
 
 VYa 
 
 no 
 
 134 
 
 134 
 
 
 V4 
 
 71 
 
 87 
 
 87 
 
 
 
 80 
 
 98 
 
 98 
 
 108 
 
 \ H 
 
 89 
 
 109 
 
 109 
 
 
 H 
 
 98 
 
 120 
 
 120 
 
 
 IH 
 
 107 
 
 130 
 
 130 
 
 
 \'4 
 
 69 
 
 85 
 
 85 
 
 
 
 78 
 
 95 
 
 95 
 
 III 
 
 87 
 
 106 
 
 106 
 
 
 if 
 
 95 
 
 116 
 
 n6 
 
 
 104 
 
 127 
 
 127 
 
 
 {% 
 
 68 
 
 83 
 
 83 
 
 
 A 
 
 76 
 
 93 
 
 93 
 
 114 
 
 \% 
 
 84 
 
 103 
 
 103 
 
 
 \\ 
 
 93 
 
 113 
 
 113 
 
 
 U 
 
 lOI 
 
 123 
 
 123 
 
 
 \'4 
 
 66 
 
 80 
 
 80 
 
 
 
 74 
 
 90 
 
 90 
 
 117 
 
 ■ H 
 
 82 
 
 100 
 
 100 
 
 
 H 
 
 90 
 
 no 
 
 no 
 
 
 IK 
 
 99 
 
 120 
 
 120 
 
 
 r^ 
 
 64 
 
 78 
 
 78 
 
 
 A 
 
 71 
 
 88 
 
 88 
 
 120 
 
 \h 
 
 80 
 
 98 
 
 98 
 
 
 \k 
 
 88 
 
 108 
 
 108 
 
 
 Vh 
 
 96 
 
 117 
 
 "7 
 
DETAILS AND STRENGTH OF CONSTRUCTION II7 
 
 Philadelphia City Rules. — In estimating the strength of the 
 longitudinal seams for rating maximum working pressure on cylindrical 
 boiler-shells two rules should be applied : 
 
 Rule A. — From the pitch of the rivets in inches subtract the diam- 
 eter of holes punched to receive the rivets ; divide the remainder by the 
 pitch of the rivets. The quotient represents the percentage of strength 
 of the solid part of the sheet. 
 
 Rule B. — Multiply the area of the hole filled by the rivet by the 
 number of rows of rivets in the seam ; divide the product by the pitch 
 of the rivets multiplied by the thickness of the sheet. This product, 
 multiplied by the shearing strength of the rivet, divided by the tensile 
 strength of the sheet, will give the percentage of the strength of the 
 rivets in the seam as compared with the strength of the solid part of the 
 sheet. 
 
 The shearing strength of a rivet in a composite joint made of iron 
 rivets and steel plates shall not be considered in excess of 40,000 
 pounds. Take the lowest of the percentages as found by Rules A and 
 B and apply that percentage as the value of the seam in the following 
 rule (C), which determines the strength of the longitudinal seams. 
 
 Ride C. — Multiply the thickness of the boiler-plate in parts of an 
 inch by the value of the seam as obtained by Rules A or B and by the 
 ultimate tensile strength of the metal used in the plates ; divide this 
 product by the internal radius of the boiler in inches multiplied by the 
 factor of safety. The quotient will be the pressure per square inch at 
 which the safety-valve may be set. 
 
 Boiler-Heads. — If the radius of the curvature of the convex head 
 of the boiler be equal to the diameter of the shell of the boiler to which 
 it is attached, then the metal in the head-sheet must be of the same 
 thickness as the plates used in the shell or cylindrical part, and no 
 bracing is necessary. 
 
 United States Rule for Boiler Pressure. — "Multiply one-sixth 
 (^) of the lowest tensile strength found stamped on any plate in the 
 cylindrical shell by the thickness — expressed in inches or parts of an 
 inch — of the thinnest plate in the same cylindrical shell and divide by 
 the radius, or half-diameter, — also expressed in inches, — and the sum 
 will be the pressure allowable per square inch of surface for single rivet- 
 ing, to which add 20 per cent, for double riveting, when all the rivet- 
 holes in the shell of such boiler have been ' fairly drilled' and no part 
 of such hole has been punched. 
 
 "The hydrostatic pressure applied must be in proportion of 150 
 pounds to the square inch to 100 pounds to the square inch of the steam 
 pressure allowed." 
 
 Factor of Safety. — This is a numerical term employed to indicate 
 that proportion which the working pressure bears to the ultimate 
 strength of a boiler. This proportion has long been fixed at ^ of the 
 
118 BOILERS AND FURNACES 
 
 bursting pressure. For example, a riveted shell which is estimated to 
 fail at the joints at a pressure of 570 pounds per square inch would be 
 allowed a working pressure under a factor of safety of 6 as follows : 
 570 -i- 6 = 95 pounds. 
 
 Since mild steel has practically displaced wrought iron for boiler- 
 shells a belief has gradually possessed the minds of designers of steam 
 boilers that a factor of safety of 6 was too large ; as a result, a factor of 
 safety of 5 has been very generally adopted instead. The effect of this 
 change is to increase the working pressure 20 per cent, over what was 
 formerly allowed, taking the above example : 570 -f- 5 ^ 114 pounds, 
 instead of 95 pounds. 
 
 The elastic limit of mild steel approximates ^ its tensile strength. 
 
 The factor of safety has reference only to the ultimate strength. 
 
 Working stress must always come within the elastic limit ; if not, 
 
 elongation and permanent set will occur. In other words, elastic limit 
 
 is the yielding-point, and no stresses can be safely carried beyond that 
 
 limit. Let us take Example 2, page iii, as a case in point. The whole 
 
 strength of a 48-inch shell, -j^-inch steel, 55,000 pounds tensile strength, 
 
 would be 
 
 55,000 X .3125 . , 
 
 ^^^ — - =716 pounds. 
 
 A riveted joint 67 per cent, as strong as the plate reduces the working- 
 pressure limit as follows : 
 
 55.000 X ^3125 X .6 7 _ ^g^ p^^^^^ 
 
 If this be divided by the factor of safety, we have, 480 -^ 5 = 96 pounds 
 working pressure. The ratio which this bears to the elastic limit may 
 be determined by substituting 27,500 pounds, or^ the ultimate strength 
 of the plate, thus : 
 
 -^ — — - = 358 pounds per square inch, 
 
 as the highest permissible pressure before the plates begin to stretch. 
 The strength of the joint has to do with ultimate failure only. It also 
 is affected by the elastic limit, for when stresses occur beyond this 
 limit, stretching of plate occurs, as in Fig. 3 : therefore, 
 
 27,500 X .3125 X .67 _ ^^^ p^^^^^ 
 
 The working pressure, as found above, was 96 pounds : then, 240 -^ 96 
 = 2.5 factor of safety with reference to elastic limit. Hydraulic tests 
 seldom or never exceed ij^ times the working pressure. We have, 
 then, 96 X 1.5 = 144 pounds test pressure; which is 96 pounds per 
 square inch less than that necessary to reach the elastic limit. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1 19 
 
 Thickness of Boiler-Heads. — Externally fired cylindrical flue 
 boilers under the United States Regulations require a thickness of ma- 
 terial as follows : " For boilers having a diameter exceeding 32 inches 
 and not exceeding 36 inches, not less than ^ an inch ; for boilers exceed- 
 ing 36 inches in diameter and not exceeding 40 inches in diameter, not 
 less than y^g- of an inch ; for boilers exceeding 40 inches in diameter, not 
 less than -^ of an inch additional thickness for every 8 inches additional 
 diameter required for boilers 40 inches in diameter. 
 
 ' ' And the heads of steam- and mud-drums of such boilers shall have 
 a thickness of material of not less than ^ an inch." 
 
 Strains on a Boiler- Head. — The strains due to steam-pressure on 
 a boiler-head are to be estimated as acting at right angles to the plate. 
 Flat plates begin to bulge out at very low pressures, offering, as they do, 
 very little resistance to bending. The common practice in boiler design 
 is to transfer, by means of stays, the greater portion of the stress upon 
 the flat surfaces of the heads to some portion of the cyUndrical shell, 
 and as this transference brings additional work upon the shell, the fas- 
 tenings of the braces or stays must not be too much localized ; other- 
 wise distortion of the shell and consequent weakness may result. 
 
 In estimating the stress upon a boiler-head due to pressure alone, it 
 is customary to make no account of the strength of the flat plate to be 
 supported, but to transfer the whole of the estimated stress upon the 
 stays. As a matter of fact, the whole stress is not thus transferred, but 
 any difference between the two counts for that much additional to the 
 factor of safety of the stay. 
 
 Unstayed Flat Heads. — The working pressure to be allowed 
 when made of stamped materials on steam-drums or shells of boilers, 
 when flanged and made of wrought iron or steel or of cast steel, shall, 
 under the United States Regulations, be determined by the following 
 rule : 
 
 ' ' The thickness of plate in inches multiplied by ^ of its tensile 
 strength in pounds, which product divided by the area of the head 
 in square inches multiplied by 0.09 will give pressure per square inch 
 allowed. The material used in the construction of flat heads when 
 tensile strength has not been ofiicially determined shall be deemed to 
 have a tensile strength of 45,000 pounds." 
 
 Bumped Heads. — The pressure allowed on bumped heads under 
 the United States Regulations is determined by the following rule : 
 
 ' ' Multiply the thickness of the plate by ^ * of the tensile strength, 
 and divide by ^ of the radius to which head is bumped, which will give 
 the pressure per square inch of steam allowed. ' ' 
 
 * This means the factor of safety. If 5 be the factor employed in any boiler 
 calculation relating to working pressure, then A of the tensile strength is em- 
 ployed, instead of ^ under the United States rule. 
 
120 
 
 BOILERS AND FURNACES 
 
 '■'Example : On a bumped head 54 inches in diameter, tensile 
 strength 45,000 pounds, ^ inch thick, bumped to a radius of 54 inches, 
 pressure would be allowed as follows : 
 
 45,000 X -75 
 6X 27 
 
 208.33. 
 
 ' ' Where the circumferential seam in such head is double-riveted to 
 shell, which is entitled -to an additional pressure, there will be allowed 
 20 per cent, additional pressure on said head." 
 
 Concave Heads. — The pressure allowed for concave heads ot 
 boilers under the United States Regulations is as follows : 
 
 " Multiply the pressure per square inch allowable for bumped heads 
 attached to boilers or drums convexly by the constant 0.6, and the 
 product will give the pressure per square inch allowable in concaved 
 heads. ' ' 
 
 Boiler-Head Stays. — The end surfaces of steam boilers are stayed 
 by means of braces extending from the heads of the shell or by longi- 
 tudinal stay-bolts extending through from the front to the back heads. 
 In all flue boilers in which the flues are riveted to the heads, the flues 
 themselves act as stays, and commonly have strength enough to dis- 
 pense with other stays below the water-line, except in very large boilers 
 or for very high pressures. The holding power of wrought-iron tubes 
 expanded in the heads is sufficient to withstand any working pressure 
 occurring in that portion of stationary-engine boilers in which the tubes 
 are located ; but boilers of large diameter, such as marine boilers, are 
 commonly fitted with stay-tubes in addition. 
 
 Fig. 104. 
 
 Flanging the edges of a boiler-head increases its stiffness along the 
 outer edge. The radius of such a flange will average not far from i^ 
 inches for large sizes ; 2 inches of this outer flanged surface of the head 
 may be left to take care of itself In addition to this, the influence of 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 the flange extends inward, and no braces need be located within 4 
 inches of the flange radius for pressures less than 100 pounds per square 
 inch, as shown in Fig. 104 ; and unless in the case of curve intersec- 
 tions and odd spaces which require a brace, the regular line of braces 
 for pressures not exceeding the above need not be closer than 3 inches, 
 also shown in the illustration. 
 
 The holding power of the tubes imparts sufficient stiffness to the 
 boiler-head as not to require braces nearer than 4 inches, so that in all 
 
 ordinary calculations the 
 
 •^ Fig. 105. 
 
 di supported, area \ 
 
 ■OOOOOOOOOOOi 
 
 area to be supported would 
 be represented by a seg- 
 ment of a circle, as shown 
 in Fig. 105, of 6 inches 
 less radius than the boiler- 
 head and its base-line 4 
 inches above the tubes. 
 
 The location of stay- 
 centres is not easily 
 worked out except on the drawing-board. This may, perhaps, be best 
 shown by the illustration. Fig. 106, which represents a boiler-head 60 
 inches in diameter. The tube-line, f from the bottom, is 36 inches, and 
 to this is added the 4 inches supported by the tubes, making 40 inches 
 in all ; this leaves 20 inches to the top of the head, from which 6 inches 
 
 are deducted, this being 
 the area supported by the 
 flange. We have, there- 
 fore, a segment of a circle 
 with a chord of 43 inches 
 and a height of 14 inches, 
 as shown. Let us assume 
 that each brace will have 
 I square inch of area — this 
 corresponds to i^ inches 
 diameter — and that the steam-pressure to be carried is 100 pounds per 
 square inch. We have then as a practical limit 60 inches of surface for 
 each stay to support ; it will be a near enough approximation to call it 
 8 inches square. Whatever the system of spacing, this distance of 8 
 inches ought not to be greatly exceeded, even though a corresponding 
 reduction from centre to centre be made in the other direction. The 
 chord of 43 inches, when laid off" in 8-inch sections, as shown, will re- 
 quire 5 stays. The height of the segment, being 14 inches, will require 
 2 stays. By simply intersecting the lines passing through these centres, 
 the number and location of stays is had. These centres may be con- 
 siderably changed without in the least affecting the strength of the 
 boiler-head, so long as the work thrown upon each stay is not more than 
 
122 BOILERS AND FURNACES 
 
 6000 pounds per square inch of section. The area to be covered by any 
 
 one brace may be determined, when the steam pressure is known, by the 
 
 use of Table XXXII. 
 
 Another method is to calculate the area of the segment, which in 
 
 this case is 418 inches ; 
 ^*^' ^°''' divide it by the number 
 
 of braces to be used (8 in 
 this case) ; this will give 
 the area to be supported 
 by each : thus 418 -h 8 = 
 52. 25 square inches, corre- 
 sponding to a diameter of 
 8}i inches. These circles 
 transferred to the drawing 
 
 as in Fig. 107 approximate very closely those previously described. 
 
 The working strength assumes an ultimate strength of 6000 pounds 
 
 per square inch of section. 
 
 TABLE XXXII. 
 
 DIRECT BRACES FOR STEAM BOILERS. 
 For diagonal braces, see page 129. 
 
 Brace. 
 
 Wrought-Iron Stays. 
 
 Inches Square each Brace will Support 
 FOR Pressure per Square Inch. 
 
 Diameter, 
 Inches. 
 
 Area, 
 Square Inch. 
 
 Working 
 Strength, 
 Pounds. 
 
 Pounds. 
 
 100 
 
 Pounds. 
 
 125 
 
 Pounds. 
 
 150 
 
 Pounds. 
 
 I 
 
 r)4 
 
 .60 
 .78 
 
 •99 
 1.23 
 1.48 
 1.77 
 
 3,600 
 4,712 
 5,964 
 7,362 
 8,880 
 10,620 
 
 7- 
 7-9 
 8.9 
 9-9 
 10.7 
 11.9 
 
 6. 
 6.9 
 
 7-7 
 8.6 
 9-5 
 10.4 
 
 6.9 
 
 7-7 
 8.5 
 9.2 
 
 4-9 
 
 5-6 
 
 6.4 
 7.0 
 
 7-7 
 8.5 
 
 The United States Regulations are : ' ' No braces or stays hereafter 
 employed in the construction of boilers shall be allowed a greater 
 strain than 6000 pounds per square inch of section. Braces must be 
 put in sufficiently thick that the area in inches which each has to sup- 
 port, multiplied by the pressures per square inch, will not exceed 6000 
 when divided by the cross-sectional area of the brace or stay. 
 
 " Steel stay-bolts exceeding a diameter of ij^ inches and not exceed- 
 ing a diameter of 2^ inches at the bottom of the thread may be allowed 
 a strain not exceeding 8000 pounds per square inch of cross-section ; 
 steel stay-bolts exceeding a diameter of 2^ inches at bottom of thread 
 may be allowed a strain not exceeding 9000 pounds per square inch of 
 cross-section ; but no forged or welded steel stays will be allowed. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1 23 
 
 ' ' The ends of such stays 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." 
 Lloyd's Rule. — The working pressure allowed on flat surfaces sup- 
 ported by stays is according to Lloyd's rule as follows : 
 
 C X T" 
 
 — — — = working pressure in pounds per square inch. 
 
 Where T — thickness of plate in j\ of an inch. 
 p = greatest pitch in inches. 
 C •— 90 for yV plate and thinner fitted with screw stays with riveted 
 
 heads. 
 C = 100 for plates thicker than j\ fitted with screw stays with riveted 
 
 heads. 
 C = 1 10 for jV plate and thinner fitted with screw stays and nuts. 
 C = 120 for plates thicker than j\ fitted with screw stays and nuts. 
 C = 140 for plates fitted with stays with double nuts. 
 C ^ 160 for plates fitted with stays with double nuts and washers, at 
 
 least h thickness of plates and a diameter of f of the pitch, 
 
 riveted to the plates. 
 
 The United States Regulations differ slightly from Lloyd's rule in the 
 value of constants employed. The working pressure allowed on flat 
 surfaces fitted with screw stay-bolts and nuts, or plain bolt with single 
 nut and socket, or riveted head and socket, will be determined by the 
 following rule : 
 
 " When plates y\ inch thick and under are used in the construction 
 of marine boilers, using 112 as a constant, multiply this by the square 
 of the thickness of plate in i6ths of an inch. Divide this product by 
 the square of the pitch or distance from centre to centre of stay-bolt." 
 
 Example : A plate y^g- inch thick, with stays placed on 6-inch cen- 
 tres, using the constant 112 as above, would yield 
 
 7 X 7 X 112 , 
 
 '- — -^^^ = 152.44 pounds 
 
 as the working pressure allowed, provided the strain on stay or bolt 
 does not exceed 6000 pounds per square inch of section. 
 
 ' ' Plates above yV i^ch thick, the pressure will be determined by the 
 same rule, excepting the constant will be 120. Then, a plate ^ inch 
 thick, stays spaced 7 inches from centre, would be as follows : 120, the 
 constant, multiplied by 64, the square of thickness in i6ths of an inch, 
 equals 7680 ; which, divided by the square of 7 inches (distance from 
 centre to centre of stays), which is 49, would give 156 pounds working 
 pressure." 
 
 On other flat surfaces there may be used stay-bolts with the ends 
 threaded, having nuts on same, both on the outside and inside of plates. 
 The working pressure allowed would be as follows : 
 
124 BOILERS AND FURNACES 
 
 "A constant 140, multiplied by the square of the thickness of plate 
 in i6ths of an inch. This product divided by the pitch or distance of 
 bolts from centre to centre, squared, gives the working pressure." 
 
 Example: A plate ^ inch thick, supported by bolts 14-inch centres, 
 
 would be 
 
 140 X 144 J 1 • 
 
 7 — = 102 pounds working pressure. 
 
 Same thickness of plate, with bolts 12-inch centres, would be 
 
 — — = 140 pounds working pressure. 
 
 144 
 
 "Flat part of boiler-head plates, when braced with bolts having 
 double nuts and a washer at least ^ the thickness of head, where 
 washers are riveted to the outside of the head and of a size equal to y% 
 of the pitch of stay-bolts, or where heads have a stiffening plate cover- 
 ing the area braced, will equal the thickness of the head and washers, 
 the head and stiffening plate being riveted together, with rivets spaced 
 and of sufficient sectional area of rivets, shall be allowed a constant of 
 200, rivets to be spaced by thickness of washer on the stiffening plate. 
 Boiler-heads so reinforced will be allowed a thickness to compute pres- 
 sure allowed of 80 per cent, of the combined thickness of head and 
 washer or head and stiffening plate. ' ' 
 
 Example: A boiler-head plate ^ inch thick, with washers ^ inch 
 thick and 11% inches square, supported by bolts 14-inch centres, would 
 be allowed a working steam pressure as follows : 
 
 Thickness of plate and washers equals f -|- f = f inch ; 80 per cent, 
 of which combined thickness equals y^^ X f inch = .9 inch = 14.4 
 i6ths of an inch. Then, by rule, 
 
 200 X 14.4= 200 X 207.36 , , . 
 
 3 ~ fi — =211 pounds working pressure. 
 
 ' ' Plates fitted with double angle-iron and riveted to plate with leaf 
 at least ^ thickness of plate and depth at least ^ of the pitch would be 
 allowed the same pressure as determined by formula for plate with 
 washer riveted on." 
 
 ' ' Exam,ple : A boiler-head plate ^ inch thick, supported by angle- 
 iron }4, inch thick and 3^ inches depth of leaf, and with bolts 14-inch 
 centres, would be allowed a working steam pressure as follows : 
 
 ' ' Thickness of head and leaf of angle-iron equals | -|- i- = |^ ; 80 
 per cent, of which combined thickness equals -^-^ X f inch = i inch ^ 
 16 i6ths of an inch. Then, by rule, 
 
 200 X 16^ 200 X 256 , , , . 
 
 ^ — = = 261 pounds working pressure. 
 
 ' ' But no fiat surface shall be unsupported at a greater distance in 
 any case than 16 inches, and such flat surfaces shall not be of less 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 25 
 
 strength than the shell of the boiler, and able to resist the same strain 
 and pressure to the square inch. In allowing the strain on a screw stay- 
 bolt, the diameter of the same shall be determined by the diameter at 
 the bottom of the thread. ' ' 
 
 Details of Stays and Braces. — The stays and braces in steam 
 boilers working under high pressures require careful working out, that 
 they do not of themselves become an element of weakness. A less 
 number of strong stays is to be preferred to a larger number of weaker 
 ones. Too many braces, especially below the water-line, interfere with 
 the circulation ; the interior of the boiler above the water-line would 
 also be more or less inaccessible, and thus prevent proper inspection, as 
 well as interfere in the facility of making repairs. 
 
 The spacing of braces and the kind best adapted for any given area 
 will depend upon the extent of that area, the thickness of the plate, and 
 the steam pressure. No particular form of stay is recommended to the 
 exclusion of others, as 
 
 several kinds may be ^^^' ^°^- 
 
 included in the same 
 boiler to good advan- 
 tage. 
 
 A crowfoot is a 
 link-joint secured to 
 wrought-iron lugs, X" 
 shaped, riveted to 
 crown-sheets, heads, or 
 other flat surfaces, and 
 adapted for the fasten- 
 ing of a brace or stay 
 by means of a bolt or 
 split pin, as shown in 
 Fig. 108. 
 
 Bolts and nuts are to be preferred to split pins, because they give 
 less chance for the spreading of the eye of the brace. One or other of 
 the ends of the split pins are liable to break off when spreading, or 
 
 closing them for removal. Dififi- 
 FiG. 109. culty is sometimes experienced in 
 
 □ I \ n getting a bolt back into place after 
 ! , ~\ />\ removal by reason of the spring- 
 '\ 1 _jti__ ^^S of the boiler-plate. In such 
 
 I I \^v^ ^ ^^^^ ^ mortised drift-pin, as in 
 
 Fig. 109, may be used and left 
 in place, after the insertion of the spring cotter to prevent its working 
 back and out of place. 
 
 In case two points of support are required for a single rod, the end 
 of the stay should be made preferably of two links, as shown in Fig. 
 
126 
 
 BOILERS AND FURNACES 
 
 I lo. Wherever practicable such forgings should be made without weld- 
 ing. No objections exist to cutting them out of odd pieces of plate of 
 
 proper thickness. 
 
 Fig. iio. 
 
 When fitting the joints or other fastenings of stay-bolts the holes 
 should be drilled. The pins or bolts should be sufficiently near the 
 size of the hole to make a shearing instead of a bending stress. A not 
 uncommon practice of bending a piece of square iron over the horn ot 
 an anvil and welding it, making an end similar to A, Fig. iii, is not 
 first-class work. It is much better to make the end as at B and drill 
 the hole not more than -^-^ larger than the bolt. 
 
 Fig. III. 
 
 A gusset-stay is an iron plate forming a diagonal brace, one end of 
 which is fitted with a flanged foot for riveting to the head and the other 
 end flanged for riveting to the shell, as shown in Fig. 112. 
 
 Gusset-stays are usually cut out of plate metal, with two pieces of 
 angle-iron riveted to each end, as shown in Fig. 112, B. Less fre- 
 quently the end of the plate itself is flanged and a piece of angle-iron 
 riveted to it, as shown at A. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 127 
 
 A larger factor of safety must be given gusset-stays than for ordinary 
 oblique stays, because the tension on a stay of this kind is not uniform 
 across its area, but is greater near one edge. Its sectional area should 
 be at least four times that calculated for a round or square stay for the 
 same pressure. 
 
 Fig. 112. 
 
 Longitudinal stays are rods extending through from one end of a 
 boiler to the other. These rods are commonly fitted with nuts and 
 washers, as shown in Fig. 326, and further in detail for the same boiler 
 in Fig. 113. When such stays are of considerable length, say 18 or 20 
 feet, they should have a central support to prevent drooping in the 
 
 Fig. 113. 
 
 centre. Whenever possible longitudinal stays should be made without 
 welding ; but if welding be a necessity, the material should then be iron, 
 and not steel. Upsetting the ends to get a larger diameter of screw- 
 threads should be confined to iron only. Stays are sometimes neces- 
 sary in places where for reasons they cannot pass through the plates. 
 
128 
 
 BOILERS AND FURNACES 
 
 as in Fig. 113 ; in such cases the ends have been fitted as in Fig. 114. 
 After the stay is in place the distance piece at A is fitted to the exact 
 length between the flat plate and the stay, after which the tap-bolt is 
 inserted and screwed up tight. The latter arrangement is not as satis- 
 factory as the former one, as well as being more expensive to make. Its 
 use is confined to those places in which no other form of stay can be 
 
 used. 
 
 Fig. 114. 
 
 Diagonal Stays, — It is always desirable that stays and tie-rods 
 lead at right angles to and from the surfaces to be stayed, but this is a 
 condition not always practicable, and diagonal stays become a necessity. 
 
 Knowing the area of a direct stay required for any working pressure, 
 the area of a diagonal stay may be found for the same pressure thus : 
 Multiply the length of the proposed diagonal stay in inches by the 
 proper area in square inches required for a direct stay ; divide the 
 product thus obtained by the length in inches of a right-angle line 
 drawn from the face of the surface to be stayed to the centre of the first 
 
 Fig. 115. 
 
 rivet in the proposed diagonal stay : the quotient will give the area of 
 the smallest part of the diagonal stay. 
 
 Let the diagram (Fig. 115) represent a portion of the boiler-head to 
 be stayed, and from a common centre imagine a direct stay leading to 
 
DETAILS AND STRENGTH OF CONSTRUCTION 129 
 
 the opposite head of the boiler, also a diagonal stay, H, attaching to the 
 shell of the boiler at a distance, L. The direct stay is the best, but if 
 for any reason it cannot be used, the enlarged area of a diagonal stay 
 to carry the same load may be found in the use of the above rule, as 
 follows : 
 
 Example: A direct stay of i^ inches diameter or 1.227 square 
 inches area is sufficient for a given area of surface and steam pressure. 
 Allowing 6000 pounds per square inch of area of stay, what must be the 
 area of a diagonal stay having a distance, H^ of 48 inches and a length, 
 L, of 42 inches? Area of direct stay = 1.227 square inches X 48 = 
 58.896 -^ 42 = 1.402 square inches area, corresponding to a diameter 
 of nearly i^ inches. 
 
 Each diagonal stay must be separately calculated. The following 
 formulae may be useful in this connection. Let 
 
 A = surface to be supported in square inches. 
 
 B = working pressure in pounds. 
 
 H = length of diagonal stay in inches. 
 
 L = length of line drawn at a right angle from the surface to be supported 
 
 to the end of the diagonal stay in inches. 
 S ^ working stress per square inch on stay in pounds. 
 a ■= area required for direct stay in square inches. 
 Cj = area of diagonal stay in square inches. 
 d = diameter of diagonal stay in inches. 
 
 Then a -^^"^ 
 
 L ■ 
 
 
 a, X L 
 
 
 a 
 
 
 ^.7854 
 
 J aXH 
 ^.7854 X L 
 
 .7854 X d' XSXL 
 
 Ax H 
 
 
 or d 
 
 
 X BX H 
 
 7854 S X L* 
 
 The stress upon a diagonal stay is equal to the stress which a per- 
 pendicular stay supporting a like surface would sustain, divided by the 
 cosine of the angle which it forms with the perpendicular to that surface 
 requiring to be supported. 
 
 Let us suppose an area of 64 square inches to be supported under 
 100 pounds pressure and the angle of the stay to be 30°, as in Fig. 116, 
 what would be the stress upon such a diagonal stay ? 
 
 The cosine of 30° is 0.866. Then 
 
 100 X 64 
 
 — ^^^ = 7390 pounds, 
 
 as against 100 X 64 = 6400 pounds if a direct stay-rod could be used. 
 
I30 
 
 BOILERS AND FURNACES 
 
 Two forms of diagonal stays are shown in Fig. 117. The one recom- 
 mended is that having a T-head with a rivet on either side of the rod 
 
 Fig. 116. 
 
 connecting with the shell. The flat stay with bent end is not as rigid as 
 the former, because there is more or less yielding at the joint inside the 
 head, unless the stay is quite thick, which is usually not the case. 
 
 Fig. 117. 
 
 
 The diagonal brace (Fig. 118), by the Lukens Iron and Steel Com- 
 pany, is made of plate steel bent to the form shown in the engraving. 
 It will be seen that the bend which occurs at the end fitted for staying 
 
 Fig. 118. 
 
 the head is braced down to the lower edge of the plate by a peculiar 
 curve for securing the necessary rigidity. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 131 
 
 Longitudinal or other stays are sometimes made of two lengths, to 
 be connected in the cen- 
 tre, in which case a turn- ^*^' ^^^' 
 buckle, as in Fig. iig, is 
 used ; but this detail had 
 better be omitted if it is 
 possible to make a single- 
 link connection with pin-joints at either or both ends, as screw-threads 
 are apt to waste away inside of a steam boiler. 
 
 Fig. 120. 
 
 □■T ^"^ ^ 
 
 Crown-Bars. — Flat crown-sheets for boilers of the locomotive type 
 are commonly supported by crown-bars extending across the furnace ; 
 less frequently they extend lengthwise of the furnace. Crown-bars are 
 generally made up of two wrought-iron bars welded at the ends, as 
 
 Fig. 121. 
 
 shown in Fig. 1 20. The depth will vary from 4 to 6 inches ; the thick- 
 ness of the bars will average not far from ^ of an inch, though ^-inch 
 metal is sometimes used ; the distance apart is i inch or i}4 inches, de- 
 
32 
 
 BOILERS AND FURNACES 
 
 Fig. 122. 
 
 pending on whether % - or i -inch stay-bolts are used. A clear water-way 
 between the crown-sheet and the crown-bars of i to i^ inches is allowed ; 
 the latter distance is preferable. The ends of the crown-bars should be 
 carefully fitted by chipping and filing, so that a good bearing is had on 
 the end of the side sheets not only, but upon the flanged curve of the 
 crown-sheet as well. The crown stay-bolts seldom vary from the two 
 diameters already referred to, — viz., ^ inch or i inch, — and these are 
 placed on about 4^ -inch centres across the crown-sheet. The details 
 of the head of such stay-bolts vary somewhat ; for example, the stays 
 experimented upon by Mr. Cole (Figs. 133 to 147) are all 
 in use. The illustration. Fig. 121, represents a practice 
 of the Pennsylvania Railroad, while Fig. 122 represents 
 a stay-bolt head and neck recommended by J. G. A. 
 Meyer in his "Modern Locomotive Construction." 
 This bolt has a slight taper under the head ; the crown- 
 sheet is reamed out to fit this taper. The advantage 
 claimed for this form is, that should leaks occur the 
 bolts may be readily taken out, the tapered part extended 
 by turning a small portion off the head and refitting the 
 bolt in the crown-sheet. Cast-iron washers or distance- 
 pieces are placed between the crown-sheet and the crown- 
 bar for tightening the stay-bolts without springing the 
 crown-sheet. These washers contribute also to the stiff- 
 ness of the crown-bars, the effect being that of deepening 
 the crown-bar by including the crown-sheet, which now 
 acts as a bottom flange to the girder spanning the fire-box. 
 
 The strength of a crown-bar is usually determined by the formula 
 used when a beam is supported at both ends and uniformly loaded, in 
 which 
 
 W = load in pounds, 
 b = breadth of crown-bar in inches, 
 d = depth of crown-bar in inches. 
 1 =^ length of crown-bar in inches. 
 16,000 ■= a constant. 
 
 When a crown-bar is made of two pieces, as shown in the illustrations, 
 the breadth b includes both, — that is, if the crown-bar is ^-inch thick, 
 b = .625 X 2 = 1.25 inches. 
 
 Example: A crown-bar is 38 inches long, 4^ inches deep by 
 ^ inch in thickness, welded in pairs as in Fig. 120. The sustaining 
 power if uniformly loaded would be 
 
 16,000 X d^' X b 16,000 X 4.5" X 1.5 
 \ °^ 3^ 
 
 12,789 pounds. 
 
 If these crown-bars are to be placed on 5-inch centres, the area sup- 
 ported would be 38 X 5 = 190 square inches. Then 12,789 -i- 190 = 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 133 
 
 67.31 pounds per square inch, a pressure much too low for the modern 
 boiler of the locomotive type. Let us assume a steam pressure of 140 
 pounds per square inch to be necessary for the work. We have then 
 190 X 140 = 26,600 pounds to be supported, or 26,600 — 12,789 = 
 13,811 pounds for each girder in excess of what the girder itself can 
 safely carry. The best way to provide for this is to connect the crown- 
 bars by means of braces with the roof-sheet overhead, as shown in Fig. 
 123. If two braces be used, the stress to be borne by each will be 
 13,811 -T- 2 = 6906 pounds. If we allow 6000 pounds per square 
 inch for the brace, we have 6906 -^ 6000 = 1.151 area of cross-section 
 at the smallest part, which approximates a diameter of i ^ inches. 
 
 Fig. 123. 
 
 The crown-bars referred to in the above example are placed cross- 
 wise of the furnace, but in some boilers they are arranged to extend 
 lengthwise of the furnace. The latter arrangement affords in loco- 
 motives a better circulation over the crown-sheet ; on the other hand, 
 it adds to the length of the crown -bar, which must be made deeper and 
 will require more roofstays to prevent deflection at its centre. The 
 present tendency in locomotive designs is against crown-bars altogether 
 and towards that of radial stays, or the employment of a Belpaire fire- 
 box. 
 
 Radial Stays are employed in curved crown-sheets of internally 
 fired boilers. These extend from the crown-sheet to the outside of the 
 boiler, as shown in Fig. 124. This arrangement of stays has much to 
 commend it. The pressure of steam tends to burst the shell and to col- 
 lapse the fire-box. But by a correctly laid out scheme of radial stays 
 these pressures are made to assist in counteracting each other by bring- 
 ing the stress of both the outer and inner sheets upon the intervening 
 stays. These stays are commonly If inch in diameter, 12 threads per 
 inch, with one end enlarged to i^ inches, so that the smaller screw 
 shall pass easily through the larger hole. The sketch. Fig. 125, illus- 
 trates average practice in locomotive work. The plates are represented 
 as being further strengthened by nuts on the outside, a practice by no 
 means universal, the commonest method being to cut the stay-bolts to 
 length in place and then rivet the ends without nuts. 
 
34 
 
 BOILERS AND FURNACES 
 
 Fig. 124. 
 
 Fig. 125. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 135 
 
 1 — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Stay-bolts for flat surfaces under high pressures, such as locomotive 
 fire-boxes, are commonly placed on 4-inch centres, as shown in Fig. 
 126. Each stay-bolt 
 
 is required, therefore. Fig. 126. 
 
 to take the pressure of 
 a surface of 4 inches 
 square, as indicated by 
 the dotted lines, less 
 its own area, which is 
 commonly omitted, to 
 be on the safe side. 
 Suppose the boiler 
 pressure to be 175 
 pounds per square 
 inch. We then have 
 4 X 4 X 175 = 2800 
 
 pounds tension upon each stay-bolt. If the stay-bolts have i inch 
 threads and are ^-inch diameter between the threads, it is the area, 
 0.601 inch, due to this latter diameter which must take the load. As- 
 suming 50,000 pounds as the tensile strength of the wrought iron, we 
 have 50,000 X -601 = 30,500 pounds as the breaking strength of the bolt. 
 Therefore, 
 
 ^-^ — = 10.9 the factor of safety, 
 2800 ^ 
 
 which for bolts under bending as well as tensile strain is the least admis- 
 sible limit. A I -inch bolt is recommended instead, which will give 
 under similar conditions a factor of safety of 14. No stay-bolt less than 
 ^ inch diameter should be used in flat fire-box sheets, no matter how 
 low the pressure. 
 
 Table XXXIII. shows the proper spacing from centre to centre of 
 stay-bolts for flat surfaces for pressures of 50 to 150 pounds per square 
 inch, thicknesses of plate from j^ to ^ inch for stay-bolts ^ to i^ 
 inches in diameter. 
 
 The Strains upon a Screwed Stay- Bolt are not the same as 
 upon a rivet, or calculations regarding them would be very simple. 
 Unfortunately, such calculations are at best of little value, because stay- 
 bolts seldom or never fail in mere tensile strength. It is their inability 
 to withstand the bending stresses which centre immediately inside of the 
 outside sheet ; and it is here that almost all stay-bolt failures occur. 
 Railway master mechanics, who have had a larger experience in such 
 matters than any one else, have generally reached the conclusion, based 
 entirely upon experience, that no screwed stay-bolts less than |- inch in 
 diameter shall be used at all in locomotive fire-boxes, and then only for 
 pressures up to 150 pounds per square inch. For pressures greater than 
 that, I inch screw stay-bolts of the best quality refined iron, placed on 4- 
 
136 
 
 BOILERS AND FURNACES 
 
 inch centres, are recommended. Such staying may be counted on for at 
 least five years' service for pressures up to 175 pounds per square inch. 
 The screw threads are finer than those of the United States standard. 
 A common pitch for screw stay-bolts is 12 threads per inch. 
 
 TABLE XXXIIL 
 
 PROPORTIONS FOR STAY-BOLTS FOR FLAT SURFACES. 
 
 
 Centre to Centre of Stay-Bolts in Inches. 
 
 Pressure per 
 Square Inch. 
 
 i<-inch Plate. 
 
 xVinch Plate. 
 
 ?/8-inch Plate. 
 
 iVinch Plate. 
 
 ^-inch Plate. 
 
 
 J^-inch Stay. 
 
 J:^-inch Stay. 
 
 %-inch Stay. 
 
 i-inch Stay. 
 
 1-%-mdh Stay. 
 
 70 
 80 
 90 
 100 
 no 
 120 
 130 
 140 
 150 
 
 6 
 
 5/8 
 
 5 
 
 aH 
 
 A% 
 
 4 
 
 3^ 
 
 2>H 
 2,% 
 
 Ia 
 
 b% 
 
 aYa 
 
 A'A 
 A% 
 
 aH 
 
 A% 
 A 
 
 8 
 
 1% 
 
 6^ 
 
 6^ 
 
 h% 
 
 5A 
 
 5% 
 
 5 
 
 aA 
 
 A% 
 
 aA 
 
 9 
 
 1% 
 
 iA 
 
 6% 
 
 sA 
 5 
 
 10 
 9 
 
 &A 
 
 'A 
 
 6>^ 
 6 
 
 Stay-bolts with a drilled hole, as in Fig. 127, are quite common, 
 especially in boilers of the locomotive type. The hole may be say y^g 
 
 Fig. 127. 
 
 Tire-box. 
 
 Fig. 128. 
 
 inch in diameter and extend inward perhaps an inch or a little more. 
 
 Inasmuch as breakages almost always occur at the outside sheets, the 
 
 drilled hole should always 
 be at the outside end of the 
 stay-bolt. In the event of 
 breakage occurring imme- 
 diately inside of the outer 
 sheet, warning is given by 
 the escape of water through 
 the central hole. 
 
 It sometimes happens 
 
 that leakage occurs around a stay-bolt, not serious enough to shut down 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 137 
 
 the boiler for repairs. A soft patch may be put over it with tap-bolts, as 
 shown in Fig. 128, at any time the pressure is off 
 
 Flexible Stay-Bolts. — The outer shell and the fire-box of an inter- 
 nally fired boiler, such as a locomotive boiler, never expand alike ; and 
 it is the greater expansion 
 
 of one plate over another ^^^- ^^9- 
 
 which causes the breakage 
 of stay-bolts. To obviate 
 this, flexible stay - bolts 
 have been designed, one 
 of which is shown in Fig. 
 129. In this illustration 
 a T-iron is riveted to the 
 outer shell. Depending 
 from this at fixed intervals 
 are duplex hangers, on 
 either side of which are 
 stay-bolts passing through 
 the crown-sheet and se- 
 curely fastened by thread 
 and nut. The proper 
 tension for the stays can 
 be had by nuts at the 
 hangers, which are after- 
 wards held in place by 
 the split pins above. 
 
 One objection to the design in the preceding paragraph is the space 
 occupied, which prevents its use in places other than above crown-sheets. 
 The staying of flat surfaces along the water-leg of a boiler requires differ- 
 ent designs, one of which is shown in Fig. 130. This stay is screwed 
 into the fire-box sheet, making a joint against the button-head, as shown 
 in the engraving. At the outer end of this stay-bolt is fitted a spherical 
 washer and nut. This washer fits into a seating screwed into the outer 
 sheet. As such an arrangement is not likely to remain tight under 
 varying pressures and conditions of expansion, a cap is screwed over 
 the nut and washer, making a water-tight joint. The arrangement 
 shown in Fig. 131 is the same, except that, instead of a conical washer, 
 the under side of the bolt-head is turned spherical, adapting it to a seat- 
 ing as in the previous example. This bolt is screwed into the fire-box 
 sheet from the outside and then riveted over. In the event that either 
 of these two caps should occupy more space on the outside of the boiler 
 than can be given to them, another form of seating is shown in Fig. 132, 
 which diminishes the outer distance about one-half. It will be seen in 
 these three examples that considerable movement may occur before the 
 sides of the stay-bolt touch the seating through which they pass. 
 
38 
 
 BOILERS AND FURNACES 
 
 Cole's Experiments. — Investigations into the holding power at 
 different temperatures of various styles of locomotive fire-box crown- 
 stays was made by Mr. Francis J. Cole, the results of which were pre- 
 sented in a communication to the American Society of Mechanical 
 Engineers, 1897. The tests were made as nearly as possible under the 
 same conditions as in actual service. 
 
 The material used to represent the stays was i-inch round mild steel, 
 58,390 pounds tensile strength, an elastic limit of 38,900 pounds, and 
 an elongation of 30. 25 per cent, in 8 inches. The sheets were mild steel 
 ^ inch thick, mostly cut from the same plate ; the tensile strength was 
 59, 150 pounds, elastic limit 28,800 pounds, elongation 31.75 per cent, 
 in 4 inches when tested lengthwise of the grain ; 58,400 pounds tensile 
 strength, 28,040 pounds elastic limit, elongation of 28 per cent, in 4 
 inches when tested crosswise of the grain. 
 
 In all these tests it is assumed that the bolts are spaced 4x4 inches 
 from centre to centre, supporting an area of 16 square inches. The 
 bagging down characteristic of an overheated crown-sheet caused by 
 low water was imitated by heating the specimens to a bright red and 
 the use of a bearing-plate of ^-inch steel, 8 inches square, with a hole 
 4)^ inches in diameter bored through its centre. The area of this hole 
 is 15.9 square inches. 
 
 Specimen, Fig. 133, represents a i-inch stay, with head }i inch 
 above the sheet riveted over. This specimen developed under cold 
 test an elastic limit of 12,400 pounds, the tensile strength being 16,700 
 pounds, yielding as shown in Fig. 134. The effect of the hot test is 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 139 
 
 shown in Fig. 135. The sheet was at a bright red heat after parting, 
 the stay-bolt puUing through the sheet. The tensile strength was 3570 
 pounds. 
 
 Fig. 133. Fig. 134. Fig. 135. 
 
 -^-^ — , r^" 
 
 iZTbrecuLs 
 
 Specimen, Fig. 136, represents a i-inch stay-bolt with a j4 standard 
 nut tapped out to i inch, 12 threads, and riveted over the top of the 
 nut, the end projecting about )4 inch for that purpose. This specimen 
 under cold test had an elastic limit of 28,000 pounds and a tensile 
 strength of 43,100 pounds. The effect of the test was to break the 
 stay, as shown in Fig. 137. The hot test consisted in heating the speci- 
 men to a dull red, which then showed an elastic limit of 14,500 pounds 
 and a tensile strength of 2 1 , 500 pounds at the time of yielding. The 
 final failure is represented in Fig. 138. The plate was almost black 
 
 Fig. 136. 
 
 Fig. 137. 
 
 Fig. 138. 
 
 ^ 
 
 
 /? 
 
 TL^ 
 
 
 1^ 
 
 after parting. The holding power of a stay provided with a nut is con- 
 siderably increased, when red hot, by countersinking the nut and rivet- 
 ing the bolt end into it. 
 
 Specimen, Fig. 139, represents a i-inch stay with button head, no 
 groove under the head, the plate slightly countersunk to fit a corre- 
 sponding projection under the head. This specimen under cold test 
 broke the bolt midway, as shown in Fig. 140, the elastic limit being 
 27,000 pounds, the tensile strength 39,800 pounds. When heated to a 
 bright red the bolt parted under the head, as shown in Fig. 141, the 
 tensile strength being 8000 pounds. 
 
 The specimen. Fig. 142, was an unthreaded stay-bolt with a button- 
 head fitted into a i^ reamed hole. Under cold test this specimen 
 pulled through the sheet, as shown in Fig. 143. The elastic limit was 
 32,500 pounds, tensile strength 43,100 pounds. It will be observed 
 that the effect of the cold test was to rupture the sheet as well as to 
 
140 
 
 BOILERS AND FURNACES 
 
 distort the head. When heated to a bright red the stay-bolt pulled 
 through the sheet, as shown in Fig. 144, exhibiting a tensile strength 
 of 9700 pounds. The bolt and plate parted while bright red. 
 
 Fig. 139, 
 
 Fig. 140. 
 
 Fig. 141. 
 
 w 
 
 
 1 
 
 izTl?rmcLs 
 
 The above tests seem to indicate that the best riveted head which 
 can be formed cold, made in the usual conical shape, has a holding 
 power, hot and cold, much less than the solid head ; but the objection 
 to the use of solid heads is the liability of injury when screwed into a 
 fire-box where the holes are not tapped at right angles to the sheet, and 
 where the surface of the sheet is curved, but this objection can easily 
 be removed by properly seating the head. 
 
 Fig. 142. 
 
 Fig. 143. 
 
 Fig 144. 
 
 Specimen, Fig. 145, was a button-headed stay-bolt with i^-inch 
 tapered reamed hole, 3-inch thimble and nut. Under cold test the 
 effect of the pull was to break the bolt 7 inches from the plate, as shown 
 in Fig. 146. The elastic limit was 22,300 pounds, the tensile strength 
 40,300 pounds. Under hot test, at a bright red, the stay-bolt pulled 
 through the sheet, as shown in Fig. 147, showing a tensile strength of 
 9660 pounds, the parts being still at a bright red after their separation. 
 
 The results of the whole series of tests, of which only a few are here 
 given, seem to indicate that the average holding power of the usual 
 form of stay-bolt, at a dull red or almost black heat, would be decreased 
 from its strength when cold by about 50 per cent. ; at a bright red, to 
 about ^ of its original strength, except in specimens, Figs. 139, 142, 
 and 145, which are decreased to about }( of their original strength. 
 In the case of specimens 142 and 145 their holding power would be 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 [41 
 
 very much increased by the use of a thicker crown-sheet, as they mostly 
 fail, both hot and cold, by the head pulling through the sheet. 
 
 Mr. Cole's conclusions are, that the centre rows of the crown-stays in 
 a locomotive boiler should be provided with solid button-heads like Fig. 
 
 Fig. 145. 
 
 Fig. 146. 
 
 Fig. 147. 
 
 tzJprM^ 
 
 139, or with nuts having a countersunk cavity into which the stay-bolt is 
 riveted, to prevent pulling through in case the crown-sheet is overheated. 
 
 Material for Stay- Bolts. —Thus far the best quality of refined 
 iron has been most satisfactory. Good stay-bolt iron will equal mild 
 steel in strength, and from its fibrous character will better withstand the 
 strains of alternate heating, cooling, and bending than steel appears to 
 do. Material for stay-bolts or braces should have an ultimate tensile 
 strength of not more than 65,000 pounds per square inch, nor less than 
 50,000 pounds ; it should show 20 per cent, elongation, and not more 
 than 35 per cent, of reduction in area at point of fracture. The objec- 
 tion brought against the employment of steel for stay-bolts by master 
 mechanics of railroads is, that they break off in a few months' use, and 
 this has been especially the case in the throat-sheet and the front parts 
 of the side sheets. The reason assigned for this breakage is the crys- 
 talline structure of the steel and the repeated bendings to which the 
 stays are subjected, owing to the difference in expansion between the 
 inner and outer fire-box sheets that they brace together. 
 
 Boiler-Tubes. — These are made of wrought iron or mild steel ; 
 charcoal-iron lap-welded tubes are commonly called for in boiler specifi- 
 cations. The standard dimensions of such tubes to 21 inches diameter 
 are given in Table XXXIV. 
 
 Tubes are always measured by outside diameter, and are commonly 
 true to gauge, so that heads, tube-plates, etc. , can be bored or other- 
 wise fitted without taking measurements directly from the tubes them- 
 selves. The bursting and collapsing pressures of solid-drawn iron tubes 
 are calculated to be, according to Clark, as in Table XXXV. 
 
142 
 
 BOILERS AND FURNACES 
 
 TABLE XXXIV. 
 
 STANDARD BOILER-TUBES, LAP-WELDED, WROUGHT IRON. 
 
 Outside. 
 
 
 
 Heating Surface, 
 I Foot in Length. 
 
 Area of 
 
 Opening. 
 
 
 ~, . 
 
 ckness, 
 ches. 
 
 Weight 
 per Foot, 
 Pounds. 
 
 
 
 
 
 Diameter, 
 Inches. 
 
 Thi 
 Circum- ^" 
 ference, 
 Inches. 
 
 Outside, 
 Square 
 Feet. 
 
 Inside, 
 Square 
 Feet. 
 
 Square 
 Feet. 
 
 Square 
 Inches. 
 
 I^ 
 
 4.71 
 
 08 
 
 1.25 
 
 .393 
 
 .349 
 
 .0097 
 
 1.40 
 
 lU 
 
 5-50 
 
 10 
 
 1.67 
 
 .458 
 
 .408 
 
 .0133 
 
 I.9I 
 
 2 
 
 6.28 
 
 10 
 
 1.98 
 
 .524 
 
 .472 
 
 .0177 
 
 2.56 
 
 2% 
 
 7.07 
 
 10 
 
 2.34 
 
 .589 
 
 .540 
 
 .0230 
 
 3.31 
 
 2% 
 
 7.85 
 
 II 
 
 2.76 
 
 .655 
 
 .598 
 
 .0284 
 
 4.09 
 
 2H 
 
 8.64 
 
 II 
 
 3.05 
 
 .720 
 
 .663 
 
 •0350 
 
 5-04 
 
 3 ^ 
 
 9-43 
 
 II 
 
 3-33 
 
 .785 
 
 .729 
 
 .0422 
 
 6.08 
 
 3H 
 
 10.21 
 
 12 
 
 3.96 
 
 .851 
 
 .789 
 
 .0495 
 
 7.12 
 
 ?>% 
 
 11.00 
 
 12 
 
 4.27 
 
 .916 
 
 .854 
 
 .0580 
 
 8.36 
 
 ?>Ya 
 
 11.78 
 
 12 
 
 4.59 
 
 .982 
 
 .919 
 
 .0673 
 
 9-69 
 
 4 
 
 12.57 
 
 13 
 
 5.32 
 
 1.047 
 
 .979 
 
 .0763 
 
 10.99 
 
 ^% 
 
 14.14 
 
 13 
 
 6.01 
 
 1. 178 
 
 I.IIO 
 
 .0981 
 
 14.13 
 
 5 
 
 15.71 
 
 14 
 
 7.23 
 
 1.309 
 
 1.234 
 
 -I215 
 
 17-50 
 
 6 
 
 18.85 
 
 15 
 
 9-35 
 
 1. 571 
 
 1.492 
 
 .1771 
 
 25-51 
 
 7 
 
 21.99 
 
 17 
 
 12.44 
 
 1.833 
 
 1.743 
 
 .2417 
 
 34-81 
 
 8 
 
 25.13 
 
 18 
 
 15.11 
 
 2.094 
 
 1.998 
 
 .3180 
 
 45.80 
 
 9 
 
 28.27 
 
 19 
 
 18.00 
 
 2.356 
 
 2.254 
 
 .4048 
 
 58.29 
 
 lO 
 
 31.42 
 
 21 
 
 22.19 
 
 2.618 
 
 2.506 
 
 .4998 
 
 71.98 
 
 II 
 
 34-56 
 
 22 
 
 25.49 
 
 2.880 
 
 2.764 
 
 -6075 
 
 87.48 
 
 12 
 
 37.70 
 
 23 
 
 28.52 
 
 3.142 
 
 3.022 
 
 -7205 
 
 103-75 
 
 13 
 
 40.84 
 
 24 
 
 32.21 
 
 3.403 
 
 3-279 
 
 -8554 
 
 123.19 
 
 14 
 
 43.98 
 
 25 
 
 36.27 
 
 3.665 
 
 3-534 
 
 ■9943 
 
 143-19 
 
 15 
 
 47.12 
 
 26 
 
 40.61 
 
 3.927 
 
 3-791 
 
 I. 1438 
 
 164.72 
 
 i6 
 
 50.27 
 
 27 
 
 45.20 
 
 4.189 
 
 4.047 
 
 1.3032 
 
 187.67 
 
 17 
 
 53.41 
 
 28 
 
 49.90 
 
 4.451 
 
 4.305 
 
 1.4738 
 
 212.23 
 
 i8 
 
 56.55 
 
 29 
 
 54.82 
 
 4.712 
 
 4-560 
 
 1.6543 
 
 238.22 
 
 19 
 
 59.69 
 
 30 
 
 59.48 
 
 4.974 
 
 4-817 
 
 1.8465 
 
 265.90 
 
 20 
 
 62.83 
 
 32 
 
 66.77 
 
 5.219 
 
 5.068 
 
 2.0443 
 
 294-37 
 
 21 
 
 65.97 
 
 34 
 
 73.40 
 
 5.498 
 
 5.320 
 
 2.2522 
 
 324-31 
 
 TABLE XXXV. 
 
 BURSTING AND COLLAPSING PRESSURES (CALCULATED) FOR SOLID-DRAWN 
 WROUGHT-IRON TUBES. 
 
 Diameter. Th 
 
 ckness. 
 
 Bursting Pressures. 
 
 Collapsing Pressures. 
 
 Inch. 
 
 nch. 
 
 Pounds. 
 
 Pounds. 
 
 I>^ 
 
 083 
 
 6200 
 
 5200 
 
 X% 
 
 083 
 
 5300 
 
 4300 
 
 2 
 
 083 
 
 4500 
 
 3700 
 
 2% 
 
 095 
 
 4600 
 
 3600 
 
 2% 
 
 109 
 
 4800 
 
 3600 
 
 2y^ 
 
 109 
 
 4300 
 
 3100 
 
 3 
 
 120 
 
 4400 
 
 3000 
 
 z% 
 
 120 
 
 4000 
 
 2700 
 
 z% 
 
 134 
 
 4200 
 
 2700 
 
 2>U 
 
 134 
 
 3900 
 
 2400 
 
 4 
 
 134 
 
 3600 
 
 2100 
 
 A% 
 
 134 
 
 3200 
 
 1700 
 
 5 
 
 134 
 
 2800 
 
 1400 
 
 6 
 
 148 
 
 2600 
 
 1000 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 143 
 
 Fig. 148. 
 
 Expanders. — A thin tube can be expanded in a bored hole so as to 
 make a steam- and water-tight joint ; this practice, however, is confined 
 to tubes having a less diameter than 5 inches. Tubes 6 inches in diam- 
 eter and larger are commonly riveted to flanged 
 heads, as shown in Fig. 148. Two forms of ex- 
 panders are in use : the Prosser expander, shown 
 in Fig. 149, which consists of a number of steel 
 segments with radial joints held together by an 
 external steel spring band, the whole being so 
 arranged that the expander when collapsed is of 
 less diameter, and may thus be inserted in the 
 end of the tube to be fitted. A tapered steel pin 
 passes through the centre of these steel pieces, 
 and by driving on the end of this pin these seg- 
 ments are forced out radially against the tube. 
 By successive operations of driving and slacking, 
 turning the expander slightly after each such ex- 
 pansion, the end of the tube is stretched until it 
 accurately fills the hole. The expander is made 
 partially concave near the end, the length of this 
 groove approximating the thickness of the head 
 and about three times the thickness of the tube. 
 Tubes put in by this method, being expanded on 
 both sides of the tube-plate in one operation, serve 
 as braces, and tend greatly to stiffen the head. 
 Before the tubes are put in place they should be carefully cut to length 
 before expanding, as the chipping off the end of the tube in place is not 
 only unworkmanlike, but there is danger of splitting the tube. 
 
 The Dudgeon expander, shown in Fig. 150, is designed to expand a 
 tube by means of a continuous rotary pressure. It consists of a hollow 
 cylinder provided with openings to receive three or more steel rollers ; 
 these rollers engage the inner diameter of the tube on the outside, and 
 
 Fig. 149. 
 
 rest upon a conical mandrel on the inside. A guide-sleeve is provided 
 which bears against the tube-sheet ; this is secured to the hollow cylinder 
 by a set screw. By shifting the position of this guide-sleeve the differ- 
 
144 
 
 BOILERS AND FURNACES 
 
 ent thicknesses of tube-plate are provided for, and thus the expander will 
 answer for several thicknesses of tube-sheet. By revolving the expander 
 and at the same time gently forcing in the conical mandrel, the rollers 
 are forced gently outward against the inner circumference of the tube, 
 enlarging its diameter until it completely fills the bored hole. 
 
 Fig. 150. 
 
 Shock's observations on the Dudgeon expander indicate that it 
 might become a dangerous instrument in the hands of an inexperienced 
 or careless person, since the operation of rolling out the metal may be 
 continued until the tube-ends are entirely cut off without giving warn- 
 ing. To prevent this, the taper mandrel often carries a loose collar, which 
 may be secured in any position by means of a set screw, and thus limit 
 the distance which the mandrel may enter the tool and force out the 
 rollers. 
 
 Prosser and Dudgeon Expanders Compared. — The difference 
 in effect by the use of the two expanders is shown in Fig. 151, in which 
 A represents a tube expanded by the Prosser, and B represents work 
 done by the Dudgeon expander. Some twenty years ago a series of 
 experiments were made at the Washington Navy Yard to determine 
 
 Fig. 151. 
 
 the holding power of boiler-tubes secured by various methods. So far 
 as relates to tube-expanders, the following general conclusions were 
 reached, viz. : (i) The tubes fixed by the Dudgeon expander and 
 beaded over have a considerably stronger hold of the tube-plates than 
 those fixed by the Prosser expander, particularly with thin tube-plates ,- 
 (2) that if the tubes were not beaded over, the hold afforded by the 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1 45 
 
 Dudgeon is less than that afforded by the Prosser system of fixing ; (3) 
 that with both expanders the introduction of ferrules, Fig. 152, adds 
 very materially to the holding power 
 
 of the tubes ; (4) that on the whole the Fig. 152. 
 
 effect of ferrules is with the Dudgeon 
 expander proportionately greater in 
 thick than in thin tube-plates, while in 
 the case of the Prosser expander the 
 proportionate increase of resistance af- 
 forded by the introduction of ferrules 
 is not materially affected by the thick- 
 ness of the tube-plates. 
 
 The distortion of flue-sheets by the action of expanders was investi- 
 gated by Mr. Brown in the Dubuque shops of the Chicago, Milwaukee 
 and St. Paul Railway, in which two engines were undergoing repairs, 
 one receiving a new flue-sheet, the other a new fire-box. After the flues 
 in one engine had been set in the usual manner by the use of a Prosser 
 sectional expander, two flues were removed from the region of the upper 
 corners and the hole in the sheet calipered. It was found that the hole 
 was -3^2 ii^ch out of round. The same thing was done with the other 
 boiler, and the holes in it were found to be }i inch out of round. This 
 was before the boiler had been fired up. It was then determined to 
 make accurate measurements of a fire-box entirely new. Accurate 
 measurements taken before and after the flues were set and expanded by 
 a Prosser expander showed that the sheet was expanded upward -^ 
 inch and sideway -j^ inch. Experiments of the same kind were then 
 made with another boiler, only a Dudgeon roller was employed to 
 expand the flues. There was no distortion of sheets caused by the 
 roller expander. To amplify this test, they took a discarded flue-sheet, 
 reamed the holes true, and rolled pieces of flues in the holes. Some of 
 the pieces were rolled until they were as thin as a piece of paper, and 
 in no case was it found that the hole was distorted by the action of the 
 roller. 
 
 The manufacturers of the Dudgeon expander supplemented these 
 experiments by taking a wrought-iron tube 4 inches in diameter and 
 ■j^ inch thick, and fitted a ring )4 inch thick on one end, and then ap- 
 plied all the power they could upon a flue-roller. The pressure was so 
 great that it made the metal flow outward, but it left the holes perfectly 
 true. 
 
 It appears that the obvious lesson of these experiments is, abandon 
 the sectional flue-expander. Fire-boxes are known to have had the 
 middle of the flue-sheet forced up almost an inch above its original 
 level ; this commonly has been attributed to expansion due to the action 
 of heating and cooling. There is good reason for believing that the 
 expander was the real cause of the distortion. 
 
146 BOILERS AND FURNACES 
 
 Holding Power of Tubes. — The Hartford Steam-Boiler Inspec- 
 tion and Insurance Company had tests made for them to determine the 
 holding power of 3-inch standard wrought-iron tubes. The tubes were 
 rolled into ^-inch plate in the ordinary way, without any expanding 
 other than that produced by a Dudgeon expander : ^-inch plate is 
 thinner than is usually used for heads of boilers of ordinary dimensions. 
 A thicker head or tube-sheet would give more frictional surface and, 
 consequently, more holding power. The test consisted in determining 
 the stress necessary to draw the tubes out of the plates. Fig. 151, B, 
 represents the end of the three tubes experimented upon, which we 
 designate as «, b, c. 
 
 The greatest observed stress sustained without the tube yielding in 
 the plate was, — 
 
 Specimen a 6000 pounds. 
 
 Specimen b 4500 pounds. 
 
 Specimen c . . 7000 pounds. 
 
 The observed stress which occasioned yielding was, — 
 
 Specimen a . 6500 pounds. 
 
 Specimen b ' . 5000 pounds. 
 
 Specimen c 7500 pounds. 
 
 To ascertain the holding power of tubes in an ordinary tubular 
 boiler, multiply the holding power of one tube by the number of tubes. 
 
 Riveting over the ends of tubes is quite generally practised, and 
 when well done makes a very strong joint ; but those who are familiar 
 with this kind of work know that in many cases the ends of the tubes 
 are frayed out and split, and until the thumb-tool is brought to bear on 
 the job it has a very unpromising look. Such work yields readily to 
 the action of the heated gases, and after a time the riveting or beading 
 fractures and crumbles off and very little strength remains, a result due 
 to want of proper annealing of the ends of the tubes, but quite as often 
 to bad workmanship. 
 
 Another method of fastening tubes into the tube-sheet is to adjust 
 the tubes so that they shall project slightly beyond the tube-sheet, roll 
 them in with a Dudgeon expander, and then flare or expand the ends 
 with a suitable tool or set. These experiments extend to 3-inch tubes 
 in ^-inch plates thus expanded with results as below. 
 
 The tube was fastened in the plate by being expanded, and the end 
 of the tube, which projected ^ inch beyond the plate, was flared so that 
 the external diameter of the extreme end was 3.2 inches, while the 
 diameter of the tube where it entered the plate was expanded to 3. i 
 inches diameter. 
 
 The test was made by observing the stress required to draw the tube 
 out of the plate, but the tube was not wholly removed from the plate 
 in the specimen e. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1 47 
 
 The stress which was sustained without the tube yielding in the plate 
 was, — 
 
 For specimen d . . 20,000 pounds. 
 
 For specimen e 18,500 pounds. 
 
 The observed stress which first produced yielding was, — 
 
 For specimen d 20,500 pounds. 
 
 For specimen <? 19,000 pounds. 
 
 And the observed stress which occasioned failure was, — 
 
 For specimen d 21,000 pounds. 
 
 For specimen ^ 19,500 pounds. 
 
 From the foregoing it will be seen that the observed stress which 
 first produced yielding was 20,500 pounds and 19,000 pounds. To 
 ascertain the holding power of the tubes in an ordinary tubular boiler 
 we multiply the holding power of one tube by the number of tubes. 
 The above company, in their publication the "Locomotive," from 
 which these figures are taken, state that they have had boilers with 
 tubes set in this way under their care for some years, and have seen 
 nothing to lead them to apprehend any trouble. The ends have given 
 little or no trouble by being subjected to the heated gases, but a projec- 
 tion of the tube beyond the sheet of more than yi inch before expanding 
 is not recommended. 
 
 Stay-Tubes. — These are seldom used in stationary boiler practice, 
 and in marine practice not as much as formerly ; their use originated at 
 a time when the holding power of expanded tubes had not been experi- 
 mentally determined. It is now known that such holding power is more 
 than equal to any pressure occurring in the spaces between tubes in any 
 ordinary tube-head. 
 
 A tube of extra thickness, threaded and fitted with a nut outside the 
 tube-head, as shown in Fig. 153, is perhaps the simplest method of 
 fitting a stay-tube. Inasmuch as the outside thread over which a nut 
 must screw easily cannot have its di- 
 ameter changed, such a tube cannot be Fig. 
 expanded in place, being simply a hol- 
 low stay in which all the stress comes 
 upon the outside nut. 
 
 Upsetting each end of an extra thick 
 tube and threading both the tube and 
 the tube-sheet, screwing the former into 
 the latter, as in Fig. 154, and after- 
 wards expanding the tube so as to make 
 
 a tight joint, is a method that has been used for staying tube-heads, but 
 is not now in common use. The tube-end shown in Fig. 155, in which a 
 nut is added, giving the tube additional thread surface, which by reason 
 
148 
 
 BOILERS AND FURNACES 
 
 of the fineness of the threads is not needed, provides a lock-nut against 
 the tube-head, which assists in making a tight joint. 
 
 Fig 
 
 Fig. 155. 
 
 Furnace-Flues. — Flues subjected to external pressure must always 
 be kept perfectly cylindrical. When such flues are riveted the joints 
 should be butt-riveted, and not lap-riveted. Welding should be prac- 
 tised whenever practicable. In the designing of furnace-flues, large 
 diameters and short grates are to be preferred to small furnaces and 
 long grates, because with the larger diameter of furnace a better com- 
 bustion of fuel is had, and this contributes to higher evaporative effi- 
 ciency. Fairbairn's experiments showed that the strength of a plain 
 tube under collapsing pressure varied inversely with the length ; this 
 led to the making of tubes in short sections with flanged ends, as in 
 Fig. 156, known as Adamson's flanged seam. This method of con- 
 struction makes an excellent flue. It is very elastic and permits of free 
 expansion. The Adamson flue has sufficient strength if riveted flange 
 to flange without the intermediate ring shown in the engraving, but this 
 ring is used in order to give a calking edge on each side of the lap. 
 This design was a great advance over the plain furnace : it is sufficiently 
 flexible longitudinally to prevent destructive strains within the boiler, 
 and the flanged rings make it much stronger to resist collapse. 
 
 Fig. 156. 
 
 Fig. 157. 
 
 mm^A 
 
 The BowHng hoop is shown in Fig. 157. This is a weldless hoop, 
 made in either wrought iron or steel. It has been largely used for 
 strengthening the flues of internally fired boilers. 
 
 The T-iron ring, shown in Fig. 158, is also a weldless hoop ; it was 
 the first form of strengthening ring employed for furnace-flues. The 
 objection to it is that it holds the flue too rigidly and does not permit 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 149 
 
 of free expansion and contraction. The Bowling hoop is superior in 
 permitting such movement, but both kinds are faulty in exposing a 
 double thickness of plates and two rows of rivets to the flames from the 
 furnace. 
 
 Corrugated Flues. — These flues are now in common use and are 
 deservedly popular. The corrugations afford a resistance to collapse 
 sufficient for all pressures now used in steam engineering. The corru- 
 gations render such flues longitudinally elastic and thereby reduce the 
 local strains within a boiler to a minimum. The material of such flue 
 should be of equal thickness, that the expansion be equal throughout 
 its length. The plates should be of such thickness and the corrugations 
 of such size and form as to prevent sagging in the middle of its length. 
 Corrugated furnaces range from y^ to ^ inch in thickness. A 3 feet 
 6 inch flue, ^ inch thick, with the Morison corrugations, will carry 198 
 pounds working pressure, in which example the desirable limit of thick- 
 ness and the practical limit of steam pressure seem to meet. 
 
 The Fox corrugated furnace flue, shown in Fig. 159, was introduced 
 about twenty years ago. Its merits were quickly appreciated, and its 
 
 Fig. 159. 
 
 almost universal adoption in marine-boiler construction followed because 
 the corrugated flue was stronger than the other flues which preceded it 
 for the same weight of material. It had the further merit of readily 
 accommodating itself longitudinally to the varying stresses incident to 
 wide changes of temperature, but it has been pointed out that the ex- 
 treme longitudinal elasticity of this form of corrugation is unnecessa- 
 rily great, and detracts from the strength to resist deflection or other 
 deformation. 
 
 The Morison suspension furnace shown in Fig. 160 possesses the 
 same general characteristics as the Fox corrugated furnace, but in a more 
 pronounced and, it is claimed, more perfect development. Referring 
 to the engraving, it will be seen that this furnace consists of a series of 
 long curves projecting inward towards the fire. The strengthening pro- 
 jections are outward or towards the water space, and are thus protected 
 from the fire, each curve being approximately a catenary. The distance 
 
ISO 
 
 BOILERS AND FURNACES 
 
 between centres of ridge arch supporters and the general proportions as 
 adopted have been experimentally determined and offer great resistance 
 to distortion or collapse. This form of corrugation presents a crown 
 surface in which the facilities for lodgment of scale is reduced to a mini- 
 mum, coupled with a maximum convenience for readily removing the 
 same when formed. 
 
 Fig. i6o. 
 
 The peculiar form of the long suspension curve is emphasized as that 
 feature of the furnace which has gained for it a practical success over the 
 Fox corrugation, inasmuch as the tension on the material is more uni- 
 formly distributed in the Morison curves than in the Fox section, with 
 its series of semicircles. Recent experiments have shown the Morison 
 furnace-flue has a slightly less longitudinal elasticity than the Fox, as 
 might have been expected from the shape of the curves ; but experience 
 has proved, so it is claimed, that the Morison has a greater tendency to 
 preserve its original circular form under work, due, no doubt, to the 
 uniform stiffening effect of the ridges, which gives a little less -elasticity, 
 and corrects to an appreciable extent the disposition to sag under severe 
 conditions of work. 
 
 Strength of Flues and Tubes to Resist Collapse. — The 
 strength of a cylinder resisting internal pressure is not affected by its 
 length. The reverse is true of cylinders subjected to external pressure, 
 for such cylinders are liable to collapse through want of conformity to a 
 true circle, and this liability to collapse increases with the length of the 
 cylinder. It is for this reason that, for the same thickness of metal, a 
 lap-welded tube will withstand a higher pressure for the same length 
 than would a lap-riveted tube, because the latter must of necessity be 
 out of round the thickness of metal at the lap-joint. 
 
 Horizontal flue boilers externally fired, when made of lap-welded 
 flues, commonly range in diameter of flue from 6 to i6 inches. Boilers 
 of this kind are rarely more than 20 feet long. The standard lap-welded 
 flues in such boilers withstand collapse under ordinary working pressures 
 which probably do not average more than 75 to 90 pounds. In the case 
 of longer boiler-shells the flues must be lengthened also, the ordinary 
 practice being for flues 12 inches in diameter and larger to make riveted 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 151 
 
 flues in short sections, say 26 to 48 inches ; the circular seams, present- 
 ing two thicknesses of metal, are favorable to resist collapse. Such flues 
 are not commonly larger than 18 inches in diameter. 
 
 Under the United States Regulations, lap-welded flues for horizontal 
 flue-boilers, externally fired for Western river steamboats, over 6 inches 
 in diameter, and not exceeding 16 inches in diameter, and not longer 
 than 18 feet, the steam-pressure not exceeding 60 pounds per square 
 inch, are not required to be made in sections. If the steam pressure is 
 more than 60 pounds and does not exceed 120 pounds per square inch, 
 such flues may be allowed if made in sections not exceeding 5 feet in 
 length, and properly fitted one into the other, and substantially riveted. 
 Furnace Flues. — The external pressure allowable on flues not more 
 than 42 inches in diameter, such flues being used as furnaces in boilers, 
 under the United States Regulations may be determined by the follow- 
 ing formula : 
 
 89,600 X T' _ 
 L X D ~ ^' 
 in which D = diameter of flue in inches. 
 89,600 = a constant. 
 
 T = thickness of flue in decimals of an inch. 
 L = length of flue in feet, not to exceed 8 feet. 
 P = pressure of steam allowable in pounds. 
 
 Example : Given a flue 40 inches in diameter, 7 feet in length, and 
 y-z inch in thickness, required working pressure to be allowed. 
 
 Substituting values in the formula, and performing the operation 
 indicated, we have 
 
 89,600 X T» 
 
 89,600 X .25 _ 
 
 22,400 
 
 L X D, ~ 
 
 7 X 40 
 
 280 
 
 80 pounds pressure. 
 
 Provided, that when such flues are made in sections of less than 8 
 feet in length and flanged to a depth of not less than 2^ inches, and 
 
 Fig. 16 
 
 are substantially riveted together with wrought-iron rings between such 
 flanges, such rings having a thickness of not less than ^ inch and a 
 width of not less than 2^ inches, see Fig. 161 ; or, in lieu thereof. 
 
152 
 
 BOILERS AND FURNACES 
 
 angle-iron rings are employed, such rings having a thickness of ma- 
 terial of not less than double the thickness of the material in the flue 
 and a depth of not less than 2)^ inches, and substantially riveted in 
 position with wrought-iron thimbles between the inner surface of such 
 ring and the outer surface of the flue, at a distance from the flue not to 
 exceed 2 inches, with rivets having a diameter of not less than i^ 
 times the thickness of the material in the flue, and placed apart at a 
 
 Fig. 162. 
 
 rr 
 
 distance not to exceed 6 inches from centre to centre at the outer surface 
 of the flue, see Fig. 162, the distance between the flanges, or the distance 
 between such angle-iron rings, shall be taken as the length of the flue in 
 determining the pressure allowable. 
 
 Example : Given a flue 40 inches in diameter, 8 feet long, and y^ 
 inch in thickness, having one ring at the middle of its length, required 
 the pressure allowable by the inspectors : 
 
 Substituting values in the formula, and performing the operation, we 
 have 
 
 p 89,600 X T^ _ 89,600 X .25 22,400 
 
 160 
 
 L X D 
 
 4 X 40 
 
 140 pounds. 
 
 To determine the thickness of material required, multiply the diam- 
 eter of the flue in inches by the length of the flue in feet, and multiply 
 the product by the pressure per square inch in pounds, and divide the 
 last product by the constant, 89,600 ; then extract the square root of 
 the quotient. The answer will give the thickness of material required. 
 
 Example: Let 
 
 40 = diameter of flue or furnace in inches. 
 
 4 = length of flue or length of furnace in feet. 
 140 = pressure per square inch in pounds. 
 89,600 = a constant. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 Then we have 
 
 153 
 
 / 40 X 4 X 140 
 ^ 8q,6oo 
 
 .5, thickness of material in decimals of an inch. 
 
 Corrugated Furnace-Flues. — The strength of corrugated flues, 
 see Figs. 159 and 160, when used for furnaces (corrugation not less than 
 i}i inches deep), and provided that the plain parts at the ends do not 
 exceed 6 inches in length, and the plates are not less than -^ inch thick 
 when new, corrugated and practically true circles, may be calculated 
 from the following formula : 
 
 14,000 
 D 
 
 X T = working pressure in pounds per square inch. 
 
 14,000 = a constant. 
 
 T = thickness in inches. 
 
 D = mean diameter in inches. 
 
 Example : Given a corrugated flue 40 inches mean diameter, 
 inch thick, required the pressure allowed : 
 
 X -5 = 175 pounds working pressure. 
 
 40 
 
 To find the thickness of metal for a corrugated furnace-flue, the 
 
 formula 
 
 P X D 
 
 thickness 
 
 14,000 
 
 may be used. 
 
 Example : Given a furnace 40 inches mean diameter, to carry 175 
 pounds working pressure, required the thickness of metal : 
 
 .5 inch thickness of metal. 
 
 175 X 40 
 14,000 
 
 Manholes. — Any steam boiler having either steam-space or water- 
 space large enough to admit a man should be provided with a manhole. 
 
 Fig. 163. 
 
 Men are not all the same size, but the standard dimensions of manholes, 
 so far as such a detail can be standardized, is 11 x 15 inches, which will 
 
 II 
 
154. 
 
 BOILERS AND FURNACES 
 
 admit a good-sized man. Smaller openings, such as 9 x 14^ inches 
 and 8 X 12 inches, and so on down to small handholes, serve a useful 
 purpose in case of examination, cleaning, and repairs. 
 
 The location of a manhole should be such that a man can enter the 
 boiler easily. In horizontal tubular and flue boilers the best place for 
 a manhole is in one of the heads, as shown in Fig. 163 ; but if for any 
 _ reason the manhole should not 
 
 be placed there, it may then be 
 located in the shell, as shown in 
 Fig. 164. This arrangement is 
 not a good one, except for large 
 boilers, because of the difficulty 
 a man has in straightening him- 
 self out in the small height oc- 
 curring between the top of the 
 tubes and the inside of the shell. 
 
 Wrought - iron strengthening 
 rings of not less than 2x1 
 inches should be riveted around 
 manholes when cut through a 
 flat plate, as shown in Fig. 163, 
 to restore as much as possible the 
 loss of strength occasioned by 
 cutting an opening in the plate, 
 a recommendation which also ap- 
 plies to a steam dome-head when 
 made of plate metal. An additional strengthening plate is shown in 
 Fig. 165. It will be observed that the inside plate is double-riveted and 
 the outer ring is single-riveted. The inside of a strengthening ring 
 serves also as a bearing surface 
 for the joint between the boiler- 
 head and manhole -plate, which 
 necessitates countersinking the 
 rivet-heads . 
 
 A flanged opening, shown in 
 Fig. 166, has the merit of suffi- 
 ciently strengthening the plate 
 without the necessity of a separate ring and the insertion of a row of 
 rivets, common to other methods of reinforcement. This is an excel- 
 lent method of making a manhole, and nearly all manufacturers of 
 boiler-plate which furnish flanged heads have the necessary machinery 
 for forming flanged manholes. The additional price charged for this 
 extra work is reasonable. 
 
 A manhole in the shell of a boiler needs to have special precautions 
 taken to insure its being perfectly safe, for it must be remembered that 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 155 
 
 the strain on the circumferential shell of a boiler is much greater than 
 upon the heads. Two forms of cast-iron frames for manhole openings 
 
 Fig. 166. 
 
 are in use, the one shown in Fig. 164, known as an outside frame, and 
 Fig. 167, an inside frame. Of these two the latter is to be preferred, 
 because the strains upon the frame are those of compression and not 
 those of extension, as in the case of the former illustration. It is im- 
 portant — and this detail should 
 never be overlooked — that the 
 long diameter of the manhole 
 should be placed across the boiler 
 and never lengthwise of it, be- 
 cause the strength of the circum- 
 ferential shell is affected more by 
 longitudinal openings than by- 
 transverse ones. If, now, a stand- 
 ard II X 15 manhole opening is cut in the shell, its long diameter par- 
 allel to that of the centre line of the boiler, 15 inches of the solid plate 
 would be affected, but if the manhole be placed across the boiler, only 
 1 1 inches would be affected. 
 
 The strength of the frame surrounding the opening must make good 
 the loss of plate removed for the manhole. It is important that the 
 
156 
 
 BOILERS AND FURNACES 
 
 frame, especially if an outside one like Fig. 164, be made of some 
 more tenacious material than cast iron. It is recommended that a steel 
 casting be used instead, or, if possible, a wrought-steel frame, shown 
 in Fig. 168. This latter is one of several details published in this vol- 
 ume relating to the 
 Fig. 168. Galloway boiler by 
 
 the Edgemoor Iron 
 Company. These 
 manhole frames are 
 for the present im- 
 ported from Eng- 
 land. The inside 
 frame. Fig. 167, 
 should also be made 
 in steel casting in 
 preference to cast 
 iron. 
 The Eclipse manhole, a wrought-steel frame, shown in Fig. 169, is 
 recommended as a satisfactory device for the purpose. The fact that 
 it is pressed out of plate metal having the same physical properties as 
 that of the shell to which it is attached is much in its favor. 
 
 Whatever form of frame be chosen for a manhole in the shell of a 
 boiler, its flange must be of sufficient area of cross-section to restore any 
 loss of strength caused 
 
 openmg 
 
 Fig 
 
 by cutting the 
 in the boiler. 
 
 The placing of a man- 
 hole in the shell of a 
 boiler having been de- 
 cided upon, its location 
 should be, if possible, in 
 the centre of a wide sheet 
 and not close to a cir- 
 cumferential riveted seam. No other opening, such as steam-nozzle, 
 etc. , should be included in the same plate as that which includes the 
 manhole. 
 
 Manhole plates or covers are commonly made of cast iron and in 
 general detail as shown in Fig. 167. The pressure of the steam forces 
 the plate against the seat ; there is required nothing but some form of 
 gasket not affected by heat and moisture to complete the joint. The 
 wrought-iron loop handles are usually included in the casting, and are 
 convenient when placing manhole plates in position. The bolts may 
 be placed in cored pockets, as shown in Fig. 165. This is a preferable 
 arrangement to drilling through the plate and riveting, as shown in 
 Fig. 167. For covering openings in frames, as in Fig. 168, cast and 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 157 
 
 wroug-ht plates are both in use ; it will be observed that the pressure is 
 all taken up by the bolts, as in the case of a cylinder-head. 
 
 Plate-iron or steel manhole covers are now furnished with wrought 
 metal frames, as shown in Fig. 169, which represents a design by Roe. 
 This cover is pressed in shape by suitable dies, the object being to 
 secure lightness as well as strength and furnish a reliable cover at a low 
 price. In the illustration the yoke, which is also of plate metal pressed 
 into shape, is shown lengthwise of the manhead instead of at right 
 angles to it, a position which it occupies in use. The object in showing 
 the yoke as above was to save the additional engraving necessary to 
 show it in another position. 
 
 Swinging Manhead. — This device, furnished with the Cahill 
 boiler and shown in Fig. 170, is very simple, and is intended to save 
 the labor of taking manhole plates out of boilers and lifting them to a 
 
 Fig. 170. 
 
 place of security, and then going through the annoyance of putting them 
 back in their places again after the work in the boiler is finished. By 
 loosening the nuts on this manhead, a slight push swings the head in 
 as though it were a door, and it fits back in place against the drum 
 
:58 
 
 BOILERS AND FURNACES 
 
 without occupying any appreciable amount of room, and when the time 
 arrives to again close it, it is pulled back to its place. Being hinged, the 
 seats come together in the same place. The joint having been once 
 
 properly fitted is tight for all 
 ■^^^- -^^i- time, needing only a gasket to 
 
 complete it. 
 
 Handholes are to be treated 
 in the same manner as man- 
 holes if the opening be greater 
 than 4x6 inches. Handhole 
 plates should by preference be 
 inside of the boiler, as in Fig. 
 166. An example of a hand- 
 hole plate outside of the boiler 
 is shown in Fig. 171, a detail 
 belonging to a water- tube boiler. 
 In this case the stress is upon 
 the bolt, which must be large enough to carry the pressure due to the 
 entire opening. 
 
 Supporting Boilers in the Furnace. — Horizontal tubular boilers 
 are commonly furnished with cast-iron wings riveted on both sides of a 
 boiler and near each end, as in Fig. 172. These wings are from i inch 
 to i^ inches thick, depending on the size of the boiler, i^ inches 
 being a fair average. The location of wings as regards height is com- 
 monly such as to give three-fifths of the shell to the furnace for heating 
 
 Fig. 172. 
 
 
 surface. It is not always practicable to carry the wings up to that 
 height, in which case the furnace-walls are carried up higher between 
 the supporting piers than the lower face of the \Vings. 
 
 Detachable wings are sometimes furnished boilers for narrow open- 
 ings which prevent a boiler passing through with wings as above, in 
 which case a base is securely riveted to the boiler, as in Fig. 173, the 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 159 
 
 face of which does not project beyond the diameter of the boiler. A 
 wing is then bolted, as shown in both elevation and plan. A lug being 
 
 Fig. 173. 
 
 Fig. 174. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 included in the wing and a corresponding recess in the base brings all 
 
 the load upon the cast portions and entirely relieves the bolts from 
 
 shearing strains. Another form of suspension is shown in Fig. 174, in 
 
 which a special head is forged on the end of a hanging bolt which fits 
 
 into a corresponding recess in the casting riveted to the boiler-shell. A 
 
 plate of iron with two 
 
 tap-bolts prevents the Fig. 175. 
 
 hanging bolt working 
 
 out of place. This form 
 
 of hanger is also within 
 
 the diameter of the 
 
 boiler. 
 
 A stamped metal 
 bracket or supporting 
 wing, shown in Fig. 175, 
 made of a single plate 
 of boiler steel by the 
 Lukens Iron and Steel 
 Company, is now being 
 
 placed upon the market. It is well formed, light, and strong. Being 
 made of mild steel, it is not likely to fail by any of the methods which 
 characterize the failure of cast iron. 
 
 A hook-strap riveted to the shell, as in Fig. 176, with an eye-bolt 
 suspended from an overhead beam, is occasionally employed in hori- 
 zontal-boiler settings, but its use is not at all common. 
 
 A loop-strap riveted to the shell, as in Fig. 177, has long been in 
 use for suspending boilers of small diameter. The rivets are not in 
 
i6o 
 
 BOILERS AND FURNACES 
 
 shearing stress, but in tension, all the strain coming upon the heads of 
 the rivets. Ordinarily these are forged out of bar iron, but Fig. 178 
 shows a loop-strap forged out of plate steel. These are not in very- 
 common use, having been but recently introduced. 
 
 Fig. 176. 
 
 Fig. 177. 
 
 Boilers suspended by links or bolts from overhead are to be provided 
 with a girder extending across the boiler. Fig. 179 represents one made 
 of cast iron suspending a two-flue boiler. Fig. 180 is a representation of 
 a tubular boiler suspended from two channel-beams placed back to back, 
 
 Fig. 178. 
 
 as shown in the central section. This latter method of using channel- 
 beams is recommended. 
 
 A bracket with suspension-rod fitted with a pin in double shear, as 
 shown in Fig. 181, is a device which enables the boiler to swing free in 
 the furnace. The nuts at the four corners can be tightened up so as to 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 l6l 
 
 entirely relieve the boiler of twisting strains occasioned by walls being 
 out of level, which sometimes occurs when using plain wings resting 
 upon brick walls or piers. 
 
 The hook bracket and link shown in Fig. 182 allows a little more 
 freedom of movement in expansion 
 
 and contraction than is the case with Fig. 179. 
 
 any of the preceding methods of 
 hanging. 
 
 The expansion of a horizontal 
 boiler will be in the direction of 
 least resistance ; and as it is quite 
 undesirable that the expansion shall 
 occur from the rear of a boiler to 
 the front, it is customary to have 
 the front wings or brackets of a 
 boiler rest upon the brickwork and 
 provide the rear wings or brackets 
 with a plate and rollers, as shown in 
 
 Fig. 183. This will secure a fixed distance for the front end of the 
 boiler, and all variations in length due to expansion or contraction occur 
 at the rear end of the boiler. 
 
 The expansion of a boiler, in case the front end of it rests upon a 
 cast-iron fire-front, is often provided for by building up a brick pier near 
 the rear end and adapting to it a cast-iron top, cylindrical rollers, and a 
 cradle for the boiler to rest upon, as shown in Fig. 184. Another 
 
 Fig. 
 
 method of accomplishing the same thing is shown in Fig. 185, which 
 consists of a single roller, curved to fit the boiler and adapted to roll 
 upon a cast-iron block of the same radius as that of the boiler. 
 
 Expansion from the rear of the boiler to the front, if for any reason 
 
l62 
 
 BOILERS AND FURNACES 
 Fig. i8i. Fig. 182, 
 
 
 ' 
 
 n 
 
 fk 
 
 i 
 
 4 
 
 7 
 
 1 
 
 
 
 this is desirable, can be had by putting wings resting upon the brick- 
 work at the rear end only and suspending the front by links or bars, as 
 shown in Fig. 181 or Fig. 182, at the discretion of the designer. 
 
 Fig. 183. 
 
 Boilers are sometimes set with one end of the boiler resting upon the 
 fire-front, the rear end resting in a cast-iron stand, as shown in Fig. 
 
 Fig. I 
 
 186, without any special provision for expansion ; but this is not good 
 practice, and if employed at all should be confined to small boilers. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 163 
 
 An expansion-rocker for a large boiler is shown in section in Fig. 
 313 and in elevation in Fig. 349. 
 
 Fire-Door Openings. — Boilers of the locomotive type and vertical 
 tubular boilers, — in fact, all boilers in which there is an outer and inner 
 sheet with a water-space between, — will require either a ring around the 
 fire-door opening or a flanging of the plates to effect a closure between 
 the two plates. 
 
 Fig. 185. 
 
 These openings are usually made oval for stationary boilers, ranging 
 from 8x12 inches for portable engine boilers to 16 x 20 for large sta- 
 tionary boilers. In locomotive practice they are quite as often made 
 round, and commonly 16 inches, but occasionally as large as 18 inches 
 in diameter. 
 
 For portable boilers a ring such as shown in Fig. 187 is used more 
 than anything else. The distance between the outside and the inside 
 sheets is commonly from 2^ to 3 inches. These rings are frequently 
 
 Fig, 
 
 made of cast iron, a practice to be condemned, because of the brittle 
 nature of the material, the frequency of rivet-holes, and the unequal ex- 
 pansion to which it is subjected. It is a much better plan to use steel 
 castings, or to make them of wrought iron with a welded joint, after- 
 wards shaping the ring to fit the surfaces to be joined and drilling the 
 holes to suit the pitch of the rivets. 
 
164 
 
 BOILERS AND FURNACES 
 
 A fire-door opening made of a Z-bar, such as shown in Fig. 188, is 
 seldom used. It is not an easy ring to make, because the outer and 
 inner flanges subjected to compression and extension make it difficult 
 to bend in circular form. It offers a convenience in riveting which 
 would be advantageous in case the ring was formed in a die in a forging- 
 machine, which is really the only practical way of producing such rings 
 at a low cost. 
 
 Fig. 187 
 
 Fig. 190. 
 
 The double-flanged ring shown in Fig. 189 is not unlike a channel- 
 bar in its cross-section. This ring should preferably be made with a 
 welded joint, the flanges being turned in a flanging-machine to such 
 dimensions as will admit the rivets, not, as shown in the engraving, 
 directly opposite each other, but, to lessen the distance between the 
 two sheets, the rivets should be staggered, placing them at half-pitch on 
 opposite sides. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 165 
 
 The flanging of a fire-box sheet and the head-sheet outward with a 
 welded ring, joining the two as shown in Fig. 190, is much used in 
 locomotive practice and makes a very substantial joint. To save wear 
 at the bottom of the fire-door opening, it is recommended that a piece 
 of mild steel (say 2 inches in width by ^ inch thick) be riveted for at 
 least a third of the distance around the lower part of the opening. 
 This will form an excellent leverage upon which the fire-tools can be 
 used and save wear on the plates. 
 
 Fig. 191. 
 
 Fig. 192. 
 
 Fig. 193. 
 
 Fig. 194. 
 
 ^^ 
 
 A fire-door opening in which the outer and inner sheets are flanged 
 one inside of the other and held by a single row of rivets is shown in 
 Fig. 191. This method of flanging is open to the objection regarding 
 the wear on the lower half of the opening by fire-tools already referred 
 to, so that a bar of the same dimensions as indicated above should be 
 riveted on the lower third or half of this opening. The thickness of the 
 
i66 
 
 BOILERS AND FURNACES 
 
 flange on the outside of the inner plate will be considerably reduced by 
 reason of the stretch to which it is subjected to bring it in line with the 
 shorter flange of the outside sheet. The designer will, of course, take 
 this into account and adapt the diameter and spacing of the rivets to 
 the thinner flange. 
 
 Flanging the fire-box sheet and the outside sheet each towards the 
 other, as shown in Fig. 192, is an excellent way of joining a fire-door 
 opening. This design is largely used in locomotive practice. 
 
 For vertical tubular boilers the outside sheet is usually left cylindri- 
 cal and the fire-box sheet flanged outward to meet it, as shown in Fig. 
 193. This method of flanging is not often practised and has no advan- 
 tages over the wrought-iron ring shown in Fig. 187. 
 
 The flanging of the outer and the fire-box sheet as shown in Fig. 
 194 is occasionally met with in locomotive practice. This flanging is 
 accomplished by the use of dies in a flanging-machine, and makes a 
 very good opening. An error which crept into the drawing unobserved 
 will be readily noticed as such, — the lower rivet is shown just the reverse 
 of what it should have been. 
 
 Steam-Dome. — This is a small reservoir attached by a base-flange 
 to the shell of a boiler, as shown in Fig. 195, for the purpose of in- 
 creasing the steam-room, both in quantity and, especially, in height. 
 
 Some makers place the 
 Fig. 195. upper manhole in the 
 
 dome instead of the 
 boiler-head ; in such 
 cases the dome-head is 
 not infrequently made 
 of cast iron, as in Fig. 
 196. This arrangement 
 for entering the boiler 
 above the tubes re- 
 quires that the shell 
 have an opening large 
 enough for a man to 
 enter the main steam- 
 space. The weakening 
 of the main shell inci- 
 dent to cutting out so 
 large a portion of the 
 plate is made good by 
 flanging the shell into 
 the dome, and sometimes further strengthening the shell by riveting 
 around it a heavy wrought-iron ring, the size of which and the propor- 
 tion of the rivet area must equal the strength of the other riveted joints. 
 Steam-domes of large diameter when made similar to Fig. 195 affect 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 167 
 
 the working strength of the main shell, because the area covered by the 
 dome has a pressure on both sides of the plate. This counter pressure 
 neutrahzes the action 
 
 of the steam pressure Fig. 196. 
 
 over the area thus cov- 
 ered. The pressure ex- 
 erted by the steam act- 
 ing upon the main shell 
 on either side of the 
 dome tends to flatten 
 the main shell within 
 the dome. It is for this 
 reason that large domes 
 with flat heads usually 
 have stays reaching 
 from the dome-head to 
 the shell, as in Fig. 
 197, in order to over- 
 come any tendency to 
 flattening at that point. 
 
 A drainage-hole must be provided at the lowest point on each side 
 of the dome to conduct any water of condensation or that due to 
 priming back into the boiler, as shown in Fig. 195. The proportions 
 for steam-domes vary according to the fancy of the builder, but approxi- 
 
 FiG. 197. 
 
 mate one-half the diameter of the boiler-shell. A close approximation 
 to average practice may be had by consulting Table XXXVI. 
 
1 68 
 
 BOILERS AND FURNACES 
 
 TABLE XXXVL 
 
 PROPORTIONS FOR STEAM-DOMES. FOR lOO POUNDS PRESSURE, DOUBLE- 
 RIVETED, STEEL SHELL AND HEAD, IRON RIVETS. 
 
 
 Size op 
 
 Dome. 
 
 Thickness 
 
 OF Metal. 
 
 Diameter of 
 
 
 
 
 
 Boiler. 
 
 
 
 
 
 
 Diameter. 
 
 Height. 
 
 Shell. 
 
 Head. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Inch. 
 
 Inch. 
 
 36 
 
 20 
 
 22 
 
 X 
 
 T6 
 
 38 
 
 20 
 
 22 
 
 X 
 
 A 
 
 40 
 
 22 
 
 24 
 
 y\ 
 
 A 
 
 42 
 
 22 
 
 24 
 
 % 
 
 T6 
 
 44 
 
 24 
 
 26 
 
 % 
 
 T6 
 
 46 
 
 24 
 
 26 
 
 X 
 
 r'e 
 
 48 
 
 26 
 
 28 
 
 % 
 
 
 50 
 
 26 
 
 28 
 
 t\ 
 
 ^ 
 
 52 
 
 28 
 
 30 
 
 A 
 
 /s 
 
 54 
 
 28 
 
 30 
 
 
 Vs 
 
 56 
 
 30 
 
 32 
 
 t\ 
 
 H 
 
 58 
 
 30 
 
 32 
 
 T6 
 
 V, 
 
 60 
 
 32 
 
 34 
 
 T6 
 
 y% 
 
 62 
 
 32 
 
 34 
 
 A 
 
 
 64 
 
 34 
 
 36 
 
 
 Y% 
 
 66 
 
 34 
 
 36 
 
 tV 
 
 68 
 
 36 
 
 38 
 
 ^ 
 
 tV 
 
 70 
 
 36 
 
 38 
 
 ^ 
 
 1^6 
 
 72 
 
 36 
 
 40 
 
 ^8 
 
 A 
 
 The thicknesses given in the table for the shell of the dome are 
 those for the top, where the shell joins the head by a riveted joint. 
 Plates can be rolled on special order -^ inch thicker at the bottom 
 for] turning a flange, but the thickness given in the table is ample 
 
 •for 100 pounds press- 
 FiG. 198. ure without additional 
 
 thickness for flanging. 
 The flanging at the bot- 
 tom of a steam-dome 
 should be wide enough 
 for a double row of 
 rivets, because the shell 
 of a boiler ought to 
 have additional stiffness 
 around the base of the 
 dome to assist in coun- 
 teracting the strains 
 incident to cutting 
 through the shell, as 
 well as the flattening 
 tendency that occurs 
 when steam acts on both sides of a curved sheet. An outside and inside 
 flange, as shown in Fig. 197, is not largely employed in boiler-making ; 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 it is more expensive to make than simply extending the width of the 
 flange for two rows of rivets, and offers no additional advantages for 
 extra cost. 
 
 Stiffening rings riveted around an opening into a steam-dome are 
 shown in Figs. 195 and 197. These add to the strength of the shell 
 by restoring a portion of that lost by cutting through the plate. 
 
 A steam-dome with a flanged joint and removable top for a loco- 
 motive boiler is shown in Fig. 198. The bottom is joined to the boiler- 
 shell by a double-riveted joint. The boiler-shell has a large opening 
 reinforced by a ring riveted to it. This dome is made in halves, that 
 ready access may be 
 
 had to the throttle- Fig. 199. 
 
 valve located within it, 
 an arrangement more 
 costly and no more efli- 
 cient than the current 
 American practice of 
 making dome-heads of 
 cast iron with an open- 
 ing large enough to 
 insert, examine, or re- 
 move the throttle- 
 valve, the cover to the 
 opening containing the 
 safety-valves. Fig. 196. 
 A flanged opening in 
 the shell of a boiler to 
 fit the interior of a 
 steam -dome, as shown 
 in Fig. 199, is seldom 
 practised in stationary 
 
 boiler-work ; it is, however, largely employed in locomotive practice, 
 but the bumped head shown in the engraving is now largely employed, 
 as it needs no bracing. 
 
 Steam-Drum. — This appendage to a steam boiler has for its object 
 the increasing of the steam-room of a boiler in a separate vessel, con- 
 nected with the boiler by a small nozzle. When a steam-drum extends 
 across a single boiler, as in Fig. 200, a common proportion is to make 
 the drum one-half the diameter of the boiler, the length of the drum to 
 be twice its own diameter. A manhole should be provided for the pur- 
 pose of internal examination. Corrosion is quite as likely to occur in a 
 steam-drum as in the boiler. 
 
 Nozzles for steam-drums may be of cast iron, riveted to the shells of 
 both drum and boiler, the flanges faced for making a joint with a suita- 
 ble gasket. The interior diameter of nozzles may be for boilers of 36 to 
 
 12 
 
170 
 
 BOILERS AND FURNACES 
 
 Fig. 200. 44 inches diameter, 4 inches ; 
 
 of 46 to 54 inches diameter, 5 
 inches ; of 56 to 72 inches diam- 
 eter, 6 inches. The metal in 
 nozzles should be about an inch 
 thick, — not for strength to resist 
 the pressure of steam, but the 
 better to withstand the eifects of 
 rough handling, etc. The thick- 
 ness of shells for steam-drums 
 may be ^ inch for all sizes up to 
 36 inches diameter. The lon- 
 gitudinal seams for diameters 
 24 inches and larger should be 
 double-riveted. It is recom- 
 mended that bumped heads be 
 used rather than flat ones. 
 
 A cross steam-drum to con- 
 nect two or more boilers is in 
 very common use in connection with horizontal tubular and flue boilers, 
 as shown in Fig. 201. The diameter may be one-half that of the boiler 
 
 Fig. 201. 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 171 
 
 with which it is connected ; the length may be that which the boilers 
 measure from outside to outside of shells, as indicated in the engraving. 
 Where two or more boilers are included within the same furnace, no 
 other fittings in the steam connections are necessary than are shown,, 
 but if the boilers are separately set, as in Fig. 202, so that any one of 
 
 Fig. 202. 
 
 them can be withdrawn from service, a stop-valve should be placed 
 between each of the boiler and steam-drum nozzles ; if so, the safety- 
 valve must not be placed on the steam-drum, but on a separate nozzle 
 on the shell of the boiler. 
 
 Longitudinal steam-drums are often used on horizontal tubular 
 boilers when the latter are 5 feet and larger in diameter. It is a common 
 practice to fit such drums with cast-iron connecting-nozzles, as shown in 
 Fig. 203. This arrangement is not without objection ; the principal one 
 
 Fig. 203. 
 
 
 urged against it is, that the expansion of the boiler and drum are not 
 likely to be coincident, and thus unequal strains occur at the riveted 
 intersections, the boiler-joint frequently becoming leaky, and a leak thus 
 caused is difficult, if not impossible, to stop. For this reason some de- 
 
172 
 
 BOILERS AND FURNACES 
 
 signers use one steam-nozzle only, the other being simply a stand to 
 carry the weight of that end of the drum. When one nozzle only 
 is used it is not infrequently made of plate metal, and is commonly 
 larger in diameter than tabulated above, and riveted to both boiler and 
 drum. 
 
 Nozzles. — The three sizes of nozzles referred to above are shown 
 in Fig. 204. The writer is not partial to either steam-domes or steam- 
 
 drums when they exist for no other reason than to furnish additional 
 steam-room ; if, however, either seem desirable, it is better to use a 
 drum, because the opening in the boiler is less, there are fewer rivet- 
 holes, and these are confined to a smaller diameter. The sizes of nozzles 
 as given above are ample for any steam service that will ever be required 
 of them. 
 
 Mud-Drums. — This appendage to a steam boiler is located under- 
 neath and commonly at the rear end, as shown in Fig. 205, which repre- 
 sents the drum attached to a horizontal tubular boiler, the end of the 
 drum passing through the rear end of the furnace-walls. The boiler 
 must be suspended in such a manner that no portion of its weight comes 
 upon the mud-drum. In building a wall around a mud-drum, no por- 
 tion of the brickwork should be in contact with it. A good method is 
 to make the opening in the rear wall about ^ inch larger all around than 
 the mud-drum, and after completion drive in a piece of asbestos packing- 
 rope or plaited gasket to fill the hole : this will be at once tight, flexible, 
 and enduring. 
 
 A cross mud-drum, as shown in Fig. 206, must be free from the 
 weight of the boiler, as already referred to. Instead of providing for 
 expansion at the rear end of the boiler, the better plan is to put expan- 
 sion-rollers, shown in Fig. 183, under the front wings of the boiler, so 
 as to throw all the expansion towards the front, the rear end having no 
 free movement. Where one mud-drum serves for two boilers attached, 
 as shown in Fig. 201, this same remark regarding expansion applies, if 
 anything, with added emphasis. 
 
 The size of mud-drums is not fixed by any definite rules. The fol- 
 lowing proportions agree closely with average practice : The diameter 
 of mud-drum for a 36-inch boiler may be 12 inches ; for 38- to 52-inch 
 boilers, the mud-drum may be 14 inches ; for 54- to 66-inch boilers, 16- 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1 73 
 
 Fig. 205. 
 
174 
 
 BOILERS AND FURNACES 
 
 inch mud-drums ; and for 72-inch boilers and larger, 18-inch mud- 
 drums. The length of the mud-drum will depend upon whether it leads 
 through the rear wall, Fig. 205, or crosswise of the furnace-setting and 
 into the two side-walls, Fig. 206, and whether or not two or niore boilers 
 are attached to the same mud-drum. These lengths will be fixed by the 
 draughtsman preparing the designs. 
 
 Material for Mud-Drums. — In ordinary horizontal tubular- and 
 flue-boiler installations, the mud-drum is commonly made of the same 
 material as the boiler to which it is attached. In water-tube boilers the 
 mud-drum is almost invariably made of cast iron, this material being less 
 injuriously acted upon by corrosive gases, water, etc., than is the case 
 with wrought plates, whether of iron or mild steel. 
 
 Functions of a Mud-Drum. — It was formerly, and is to some 
 extent at present, the practice to have the feed-water enter the boiler 
 through the mud-drum, in which case the latter performs the functions 
 of a heater. The feed-water entering the boiler through the mud-drum 
 does so at a much less velocity than if the feed-pipe led directly into the 
 boiler. This slow movement of the water through the mud-drum, in 
 contact with the heated gases, brings its temperature up to that within 
 
 the boiler or nearly so ; it 
 ^^^- ^°7- thus relieves the latter in 
 
 part from the bad effects of 
 cold water localized within 
 the boiler, such as leaky 
 tubes and riveted joints 
 when occasioned by the 
 contraction of metal in the 
 vicinity of the water- inlet. 
 This relief, if any, is simply 
 due to a higher tempera- 
 ture of feed-water. The 
 proper function of a mud- 
 drum is not to heat the 
 feed-water, but if the water 
 contains mud or any other 
 foreign substance which 
 may be precipitated by 
 heating the water, the ex- 
 pectation is that much of 
 this foreign matter will be 
 lodged by a downward cir- 
 culation in the mud-drum 
 and blown out from there. 
 To accomplish this to the best advantage, the mud- drum should be pro- 
 tected from the atmosphere of heated gases which surrounds it. One 
 
DETAILS AND STRENGTH OF CONSTRUCTION 
 
 175 
 
 method of doing this in connection with a horizontal tubular boiler is 
 shown in Fig. 207, where a brick arch is thrown over the mud-drum, 
 thus wholly removing it from the direct action of the hot gases. In the 
 illustration of the Stirling boiler, Fig. 378, the construction of furnace 
 to protect the mud-drum from the direct action of the heat is especially- 
 noticeable. 
 
 Water-Surface. — In designing steam boilers of whatever type 
 there should be provided a large water-surface for the disengagement 
 of steam. Such a surface secures a steadiness of water-level and pre- 
 vents foaming. Boilers having a large water-surface for the liberation 
 of steam furnish dryer steam than those boilers in which there is a de- 
 ficiency of liberating surface. This is notably the case in small vertical 
 tubular boilers, in which the liberating surface is small as compared with 
 the total heating surface. 
 
 Steam-Room. — In horizontal tubular and flue boilers the steam- 
 room is nearly constant for a given diameter. The upper line of the top 
 row of tubes is commonly two-thirds of the diameter of the boiler from the 
 bottom ; over this is from 2 to 4 inches of water, depending on the size 
 of the boiler ; all above this is steam-room. In water-tube boilers the 
 water is carried up to about one-half the diameter of the combined water- 
 and steam-drum. In ascertaining the steam-room in any cylindrical 
 shell, it is the area of the segment above the water multiplied into the 
 length of the steam-space. Knowing the diameter of the boiler and the 
 height of the segment, the area of the latter may be calculated by means 
 of Table XXXVII. 
 
 TABLE XXXVII. 
 
 
 AREA OF 
 
 CIRCULAR segments; 
 
 DIAMETER OF CIRCLE, I. 
 
 
 Height. 
 
 Area. 
 
 Height. 
 
 Area. H 
 
 eight. 
 
 Area. H 
 
 eight. 
 
 Area. 
 
 .100 
 
 .0409 
 
 .205 
 
 .1158 
 
 310 
 
 .2074 
 
 415 
 
 .3081 
 
 .105 
 
 •0439 
 
 .210 
 
 .1199 
 
 315 
 
 .2120 
 
 420 
 
 •3130 
 
 .110 
 
 .0470 
 
 .215 
 
 .1240 
 
 320 
 
 .2167 
 
 425 
 
 .3180 
 
 •115 
 
 .0502 
 
 .220 
 
 .1281 
 
 325 
 
 .2213 
 
 430 
 
 .3229 
 
 .120 
 
 •0534 
 
 .225 
 
 •1323 
 
 330 
 
 .2260 
 
 435 
 
 •3279 
 
 .125 
 
 .0567 
 
 .230 
 
 •1365 
 
 335 
 
 .2307 
 
 440 
 
 •3328 
 
 .130 
 
 .0600 
 
 •235 
 
 .1407 
 
 340 
 
 •2355 
 
 445 
 
 •3378 
 
 •135 
 
 .0634 
 
 .240 
 
 .1449 
 
 345 
 
 .2403 
 
 450 
 
 .3428 
 
 .140 
 
 .0668 
 
 •245 
 
 .1492 
 
 350 
 
 .2450 
 
 455 
 
 •3478 
 
 •145 
 
 .0703 
 
 .250 
 
 •1536 
 
 355 
 
 .2498 
 
 460 
 
 •3527 
 
 .150 
 
 •0739 
 
 •255 
 
 ■1579 
 
 360 
 
 .2546 
 
 465 
 
 •3577 
 
 •155 
 
 •0775 
 
 .260 
 
 .1623 
 
 365 
 
 •2594 
 
 470 
 
 .3627 
 
 .160 
 
 .0811 
 
 .265 
 
 .1667 
 
 370 
 
 .2642 
 
 475 
 
 •3677 
 
 .165 
 
 .0848 
 
 .270 
 
 .1711 
 
 375 
 
 .2690 
 
 480 
 
 •3727 
 
 .170 
 
 .0885 
 
 •275 
 
 •1755 
 
 380 
 
 •2739 
 
 485 
 
 •3777 
 
 •175 
 
 .0923 
 
 .280 
 
 .1800 
 
 385 
 
 .2787 
 
 490 
 
 • 3827 
 
 .180 
 
 .0961 
 
 .285 
 
 .1845 
 
 390 
 
 .2836 
 
 495 
 
 •3877 
 
 .185 
 
 .1000 
 
 .290 
 
 .1891 
 
 395 
 
 .2885 
 
 500 
 
 •3927 
 
 .190 
 
 .1039 
 
 295 
 
 •1936 
 
 400 
 
 •2934 
 
 
 
 •195 
 
 .1078 
 
 .300 
 
 .1982 
 
 405 
 
 .2983 
 
 
 
 .200 
 
 .1118 
 
 •305 
 
 .2028 
 
 410 
 
 .3032 
 
 
 
1/6 
 
 BOILERS AND FURNACES 
 
 To use the table, divide the height of the segment by the diameter 
 of the boiler ; find the quotient in the table opposite to which is the cor- 
 responding area for a circle whose 
 ^^'^- ^°^- diameter is i ; square the diam- 
 
 eter of the boiler and multiply it 
 by the area found as above : the 
 product will be the area of the 
 segment. 
 
 Exainple : Let Fig. 208 rep- 
 resent a boiler in which the diam- 
 eter D = 48 inches and the height 
 of the segment B = 14 inches. 
 What will be the contents of the 
 steam-room in cubic feet if the boiler be 14 feet long ? 
 
 = .292. 
 
 The nearest number in the tabular heights is 0.290, and its corre- 
 sponding area is 0.1891. The total area of the segment is 48 X 48 X 
 .1891 = 435.69 square inches. Length, 14 feet = 14 X 12 = 168 
 inches. Then 
 
 168 X 435.69 . u- r . 
 
 p; = 42. ^6 cubic leet. 
 
 1728 ^ ^ 
 
 Heating Surface. — All that portion of a boiler subject to the direct 
 action of the heat of the furnace, and that traversed by the heated 
 products of combustion after leaving the furnace, is reckoned as water- 
 heating surface when the opposite side of such a surface is in contact 
 with water, and superheating surface where such surfaces pass through 
 the steam-room of a boiler. Unless otherwise stated, heating surface 
 means water-heating surface. 
 
 The most effective heating surface is that of the fire-box in internally 
 fired boilers and the exposed shell surface in externally fired boilers. 
 Flue- and tube-surfaces are less effective than either of the above types, 
 since the limiting figure for one boiler might not be suitable for another. 
 
 Experiments made with a locomotive boiler on the Northern Railway 
 of France, the boiler being divided into several sections, showed a rapid 
 decrease in the rate of evaporation as the distance from the furnace in- 
 creased, so that the tubes at their extreme ends scarcely evaporated any 
 water at moderate rates of combustion. The furnace-sheets and a small 
 portion of the contiguous tube-surface forming the first section evapo- 
 rated at the rate of 44.6 pounds of water hourly per square foot of 
 heating surface. 
 
 Horse-Power of Boilers. — There is no standard for measuring 
 boiler power by extent of heating surface, nor can there be, because the 
 
DETAILS AND STRENGTH OF CONSTRUCTION 1/7 
 
 evaporating power for similar areas throughout the boiler is very un- 
 equal. By experiment the efficiency of certain types of boilers has been 
 ascertained with considerable accuracy, so that it is a well-established 
 commercial practice to sell steam boilers by extent of heating surface as 
 a basis of horse-power. 
 
 The common rating of horizontal tubular boilers is 15 square feet of 
 heating surface per horse-power. The heating surface includes two- 
 thirds of the shell and all of the tube-surface. 
 
 Flue boilers are similarly rated at 12 square feet. 
 
 Vertical tubular boilers are rated at 12 square feet. 
 
 Cylinder boilers are rated at 9 square feet per horse-power. 
 
 For horizontal tubular boilers the usual allowance for each horse- 
 power is, — 
 
 Steam for heating, etc 15 square feet heating surface. 
 
 For plain throttle engine 15 square feet heating surface. 
 
 For single Corliss engine 12 square feet heating surface. 
 
 For compound Corliss condensing . 10 square feet heating surface. 
 
 Hence a boiler for furnishing steam for 
 
 Plain slide-engine, with 1500 square feet surface . 100 horse-power. 
 
 Simple Corliss engine, same boiler 125 horse-power. 
 
 Compound engine, same boiler 150 horse-power. 
 
 The best method is to compare boilers by their evaporative efficiency 
 and not by heating surface. 
 
 The following is an approximate consumption of steam per indicated 
 horse-power per hour for engine : 
 
 Plain slide engine 60 to 70 pounds, 
 
 High-speed automatic engine 30 to 50 pounds, 
 
 Simple Corliss engine 25 to 35 pounds. 
 
 Compound Corliss engine 15 to 20 pounds, 
 
 Triple-expansion engines 13 to 17 pounds, 
 
 depending upon the horse-power, steam pressure, condition of engine, 
 load, etc. 
 
 The Centennial Standard. — The Centennial Exhibition occurred 
 in 1876. The Committee of Judges to whom was committed the trials 
 of competing boilers adopted as a unit of horse-power ' ' 30 pounds of 
 water evaporated into dry steam per hour from feed-water at 100° Fahr, , 
 and under a pressure of 70 pounds per square inch above the atmos- 
 phere." These conditions were considered by the committee to fairly 
 represent average practice. 
 
 The American Society of Mechanical Engineers' standard is the 
 Centennial standard in another form, — viz., "34^ pounds of water 
 evaporated from a feed-water temperature of 212° Fahr. into steam of the 
 same temperature." 
 
CHAPTER VI. 
 
 EXTERNALLY FIRED BOILERS. 
 
 Externally fired boilers are usually cylindrical shells with or with- 
 out flues or tubes and set in brickwork, the furnace being located under- 
 neath one end of the shell of the boiler. 
 
 A Cylinder Boiler consists of a plain cylindrical shell with closed 
 ends, no flues or tubes being used. Cylinder boilers are seldom used 
 singly, but commonly in batteries of from two to four, and occasionally 
 six, with one furnace common to all. The diameters range from 26 to 
 40 inches, having lengths of from 10 to 15 diameters, according to loca- 
 tion, service, etc. They are not as economical of fuel as either flue or 
 tubular boilers ; their use is restricted, therefore, to the lumber and coal 
 regions, or such other localities where the feed-water is bad and the fuel 
 abundant and cheap. 
 
 A cross-section and foundation plan of a battery of three cylinder 
 boilers, 36 inches in diameter by 36 feet long, by the Jeanesville Iron 
 Works, for an anthracite mine in Pennsylvania, is shown in Fig. 209. 
 These boilers have two points of support placed at 8 feet 3 inches from 
 each end, a cast-iron suspension girder extending across the furnace 
 walls, shown in Fig. 210, carrying one end of the three boilers. Cylinder 
 boilers are commonly fitted with cast-iron heads, and this is the case in 
 the present example. This head is detailed in Fig. 211. The lower 
 nozzle is for the feed-water, the upper one for the steam connection ; 
 the piping may be at either the front or rear of the boiler, as may be 
 most convenient. The feed-pipe may extend some distance back from 
 its entrance and upward, towards the centre of the boiler, to prevent 
 local contraction, which would occur if the cold water came directly 
 upon the plates at either end. The chimney is situated at the rear of 
 the boilers, suitable flues being constructed in brickwork, as shown in 
 plan in Fig. 209 and in sectional elevation in Fig. 212. 
 
 One objection to this type of boiler is the excessive ground-room 
 required as compared with other types when large heating surfaces are 
 necessary to furnish a given quantity of steam. In the boiler illustrated 
 above, as well as cylinder boilers generally, only one-half the circum- 
 ference is available for the transmission of heat to the water. We have, 
 then : ^ circumference of 3 feet ^4.71 feet ; length, 36 feet X 4.71 
 feet = 169.56 square feet for one boiler, or 169.56 X 3 = 508.68 square 
 feet total heating surface. The ground-room occupied, apart from the 
 178 
 
EXTERNALLY FIRED BOILERS 
 
 179 
 
 chimney and flues leading thereto, is 472.29 square feet. Allowing 9 
 square feet of heating surface per horse-power, we have 
 
 508.68 
 
 = 56.52 horse-power. 
 
 This same power could have been furnished by a horizontal tubular 
 boiler 60 inches dianieter by 14 feet long on the basis of 15 square feet 
 
i8o 
 
 BOILERS AND FURNACES 
 
 per horse-power, which would have required a ground space of 9 x 19 
 feet =171 square feet, the former being 2.76 times as much. 
 
 Fig. 210. 
 
 Another objection to this type of boiler is the incomplete utilization 
 of heat as compared with other types, the direct escape of gases from 
 the furnace to the chimney resulting in a large waste of heat, which, 
 
 Fig. 211. 
 
EXTERNALLY FIRED BOILERS 
 
 l8l 
 
 going on continuously, could not be borne under any circumstances 
 which involved purchase of fuel. The temperature of escaping gases 
 when the fires are forced is not infrequently over 800° Fahr., showing 
 a wasteful expenditure of fuel. 
 
 Fig. 212. 
 
 Double-Deck Cylinder Boilers. — The lack of sufficient heating 
 surface in a cylinder boiler to bring the temperature of the escaping 
 gases down to that of a flue or tubular boiler, when the boiler is worked 
 up to its capacity, has led to the placing of an additional cylinder under 
 the main boiler, back of the bridge-wall, as shown in Fig. 213. The 
 lower shell is in reality an exaggerated mud-drum and heater, its office 
 being to absorb heat from the outgoing gases. Whatever waste heat 
 can thus be reclaimed is gain. Such boilers are not in common use : 
 the difference in temperature of water in the upper and lower cylinders 
 produces unequal expansion, and this brings additional and variable 
 stresses upon the joints at the necks connecting the two shells, resulting 
 in troublesome leaks and corrosion induced by unusual strains. 
 
 The French Boiler, or Elephant Boiler, is one large cylinder with 
 two or three smaller cylinders underneath connected by suitable nozzles, 
 as shown in Fig. 214. The lower cylinders are filled with water and 
 almost wholly surrounded by heated gases. The furnace being under 
 these small cylinders, the products of combustion act first upon them 
 throughout their length, then return to the front end of the boiler along 
 one side of the main cylinder, and finally pass to the chimney along the 
 other side of the main cylinder. The particular boiler here illustrated 
 is one in which the famous Mulhouse experiments were made. The 
 principal dimensions are : 
 
 Main cylinder, 3 feet 9 inches diameter by 20 feet 6^ inches long ; 
 three lower cylinders, 19.7 inches diameter by 32 feet 9^ inches long ; 
 grate surface, 20 square feet ; total heating surface, 607.6 square feet. 
 
l82 
 
 BOILERS AND FURNACES 
 
 *6P 
 
 
EXTERNALLY FIRED BOILERS 
 
 183 
 
 RESULTS OBTAINED. 
 
 Equivalent evaporation from and at 212° Fahr. per pound of coal . 8.97 pounds. 
 Equivalent evaporation from and at 212° Fahr. per pound of net 
 
 combustible 10.37 pounds. 
 
 Equivalent evaporation from and at 212° Fahr. per hour per 
 
 square foot of heating surface 3.28 pounds. 
 
 Weight of air supplied per pound of coal consumed 14.89 pounds. 
 
 Mean temperature of gases entering chimney . . 425° Fahr. 
 
 This shows a good rate of evaporation, but it does not surpass that 
 of a good tubular boiler, which can be bought for much less money. 
 
 Two-Flue Boilers. — These boilers are extensively used in the 
 Western States, especially in the river towns and coal-mining districts. 
 Many of the light-draught steamboats which navigate the Western 
 rivers use boilers of this design. 
 
 In diameter such boilers range from 36 to 48 inches, and from 16 to 
 24 feet in length, with an occasional increase for the larger diameters to 
 30 feet. The flue diameter is commonly 12 inches for 36- and 38-inch 
 shells, 13 inches for 40-inch shells, 14 inches for 42-inch shells, 15 
 inches for 44-inch shells, 16 inches for 46- and 48-inch shells. For ordi- 
 nary land service the pressures are not much above 75 pounds per 
 square inch, and flues are made ^ inch thick for all diameters. Lap- 
 welded tubes can be had in either steel or iron up to 20 feet in length ; 
 but the advantages accruing from a double thickness of plate at riveted 
 circumferential seams 
 
 to resist collapse keeps Fig. 216. 
 
 the riveted flue still in 
 favor for the larger di- 
 ameters. Lap-welded 
 flues can be cut in 
 lengths and made up 
 with riveted circum- 
 ferential seams to ac- 
 commodate any length 
 of shell. The United 
 States Regulations fix 
 the greatest length of 
 such sections for diam- 
 eters of 12 to 18 inches 
 
 at 3 feet. The thickness of flues from 12 to 15 inches diameter rnay be ^ 
 inch ; 16 to 18 inches diameter, -^^ inch. Such 12- and 14-inch flues will 
 pass United States inspection for 150 pounds working pressure. For 
 16- to 18-inch flues, — the 16 x -^-inch flues will be passed for 150 
 pounds working pressure, the 17-inch flues for 141 pounds, and the 18- 
 inch flues for 134 pounds. 
 
 The flanging of two-flue boiler-heads is commonly as shown in Fig. 216, 
 
1 84 
 
 BOILERS AND FURNACES 
 
 the illustration representing a lap-welded flue. The rear head is flanged 
 in, and that end of the flue is first riveted in place. The front head is 
 flanged out and flue riveted on the outside. 
 
 kill,,! 
 
 A longitudinal elevation of a boiler of 40 inches diameter with 2 flues 
 of 13 inches diameter, each boiler 20 feet long, is shown in Fig. 217. A 
 cross-sectional elevation showing two such boilers set in the same furnace 
 
EXTERNALLY FIRED BOILERS 
 
 185 
 
 is given in Fig". 218. The grate-bars, adapted for burning bituminous 
 coal, are placed 36 inches below the bottom of the boilers, thus afford- 
 ing a roomy combustion-chamber above the fuel. 
 
 Fig. 218. 
 
 Five-Flue Boilers. — The power of resistance to prevent collapse 
 of a boiler-flue decreases with its diameter. For this reason several 
 smaller flues are preferred to two large ones. When more than two 
 flues are wanted, the next practical number is five. Three flues are 
 placed in the upper row and two in the space underneath. 
 
 In some localities the flue-diameters vary for a given boiler ; for 
 example, a 
 
 44-inch boiler may have 3 8-inch and 2 lo-inch flues. 
 46-inch boiler may have 2 8-inch, 2 9-inch, and i 13-inch flues. 
 48-inch boiler may have 2 8-inch, 2 lo-inch, and i 12-inch flues. 
 50-inch boiler may have 2 8-inch, 2 lo-inch, and i 14-inch flues. 
 
 This arrangement of flue-diameter is not known to possess any advan- 
 tage over the simpler one in having all the flues of the same diameter. 
 There is a disadvantage in the fact that escaping gases flow in the direc- 
 tion of least resistance, the larger flues robbing the smaller ones of their 
 proportion of heated gases. The same remarks apply here regarding 
 details of construction and collapsing resistance of flues as were given in 
 the section on two-flue boilers. If the flues are all of the same diameter, 
 7-inch flues will answer for 36- and 38-inch shells, 8-inch flues for 40- 
 inch shells, 9-inch flues for 42-inch shells, and lo-inch flues for 44- to 
 
 13 
 
[86 
 
 BOILERS AND FURNACES 
 Fig. 219. 
 
 Fig. 220. 
 
 48-inch shells. Boilers of this type are commercially rated at 12 square 
 
 feet of heating surface per horse-power. 
 
 Six-Inch-Flue Boilers. — This design of 
 boiler has long been a favorite one in the Western 
 States ; and when a choice is made away from a 
 two-flue boiler it usually passes over to this, be- 
 cause a larger heating surface is had than in two- 
 flue boilers, and there are larger spaces between 
 the flues than is the case in tubular boilers, making 
 it a compromise boiler well adapted to water 
 heavily charged with scale-making impurities. 
 The tubes are lap welded and riveted into the 
 heads, as shown in Fig. 219. The rear head is 
 flanged in and the front head flanged out. A 
 detail of a riveted flue and section of head is shown 
 in Fig. 220. A few special tools are needed for 
 holding the rivets in place, and perhaps a few 
 special riveting-hammers. The rivets should be 
 y^, inch diameter for 6-inch flues. The holes in 
 the head should be countersunk, as shown in Fig. 
 220, so that the rivet-point shall project into the 
 flue as little as possible. When putting the flues 
 in the boiler, begin with the bottom ones, insert 
 the rivets from the inside of the boiler, and rivet 
 
EXTERNALLY FIRED BOILERS 
 
 187 
 
 them on the inside of the flue. This can be done with Uttle difficulty 
 from the outside of the boiler. 
 
 The number of 6-inch flues that can be gotten into a given diameter 
 of head depends entirely upon the number of flanged openings which 
 can be made in it. In a properly constructed machine there is no diffi- 
 culty in flanging them within 3 inches of each other and keeping the 
 face of the boiler-head flat. The number of 6-inch lap-welded flues for 
 each diameter of boiler may be as follows : 
 
 Diameter of boiler . . . 
 
 . 40 
 
 44 
 
 48 
 
 54 
 
 60 
 
 66 
 
 72 
 
 Number of 6-inch flues . 
 
 . 6 
 
 8 
 
 10 
 
 12 
 
 18 
 
 22 
 
 26 
 
 The commercial rating of a 6-inch-flue boiler is 12 square feet of 
 heating surface per horse-power. 
 
 TABLE XXXVIII. 
 
 PRINCIPAL DIMENSIONS OF SIX-INCH-FLUE BOILERS. 
 
 
 Shell. 
 
 uTC 
 
 Thickness. 
 
 
 Dome. 
 
 
 
 ^.2 
 
 
 
 
 
 
 
 
 
 
 
 
 "5 
 
 - 
 
 
 
 E| _ 
 
 
 
 
 
 Thickness. 
 
 s 
 
 • 
 
 u 
 
 
 
 
 
 u 
 
 
 
 
 
 •"-£* 
 
 
 
 
 
 
 
 
 
 
 
 
 II 
 
 B 
 .2 
 P 
 
 
 1^' 
 
 1 
 
 1 
 
 S 
 Q 
 
 1 
 
 1 
 
 i 
 
 a 
 
 i 
 
 H.-P. 
 
 Inch. 
 
 Feet. 
 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Sq. Feet. 
 
 Pounds. 
 
 40 
 
 48 
 
 16 
 
 ID 
 
 A 
 
 tV" 
 
 26 
 
 28 
 
 A 
 
 H 
 
 380 
 
 12,065 
 
 45 
 
 48 
 
 18 
 
 10 
 
 t\ 
 
 t'^ 
 
 26 
 
 28 
 
 A 
 
 H 
 
 426 
 
 12,465 
 
 50 
 
 54 
 
 16 
 
 12 
 
 T6 
 
 
 30 
 
 34 
 
 A 
 
 Yi 
 
 454 
 
 14,025 
 
 55 
 
 54 
 
 18 
 
 12 
 
 A 
 
 7 
 T6 
 
 .SO 
 
 34 
 
 A 
 
 Y 
 
 508 
 
 14,725 
 
 60 
 
 54 
 
 20 
 
 12 
 
 t\ 
 
 
 ?,o 
 
 34 
 
 A 
 
 H 
 
 566 
 
 15,825 
 
 65 
 
 60 
 
 16 
 
 18 
 
 ii 
 
 T6 
 
 ,S2 
 
 36 
 
 A 
 
 Y 
 
 624 
 
 17,120 
 
 75 
 
 60 
 
 18 
 
 18 
 
 H 
 
 tV 
 
 .^2 
 
 36 
 
 A 
 
 Vs 
 
 702 
 
 18,370 
 
 80 
 
 60 
 
 20 
 
 18 
 
 II 
 
 tV 
 
 .S2 
 
 36 
 
 ft 
 
 Vs 
 
 780 
 
 19,700 
 
 85 
 
 66 
 
 18 
 
 22 
 
 ^'? 
 
 36 
 
 40 
 
 
 827 
 
 21,800 
 
 95 
 
 66 
 
 20 
 
 22 
 
 n 
 
 .^6 
 
 40 
 
 Y% 
 
 A 
 
 919 
 
 23,370 
 
 90 
 
 72 
 
 16 
 
 26 
 
 T6 
 
 Yz 
 
 .^6 
 
 40 
 
 Y% 
 
 t'« 
 
 854 
 
 22,800 
 
 100 
 
 72 
 
 18 
 
 26 
 
 T6 
 
 y^ 
 
 .^6 
 
 40 
 
 H 
 
 A 
 
 961 
 
 24,050 
 
 no 
 
 72 
 
 20 
 
 26 
 
 tV 
 
 /2 
 
 36 
 
 40 
 
 n 
 
 A 
 
 1067 
 
 25,750 
 
 Tubular Boilers. — The diameter of tubes in an ordinary horizon- 
 tal tubular boiler ranges from 3 to 4 inches, seldom above or below 
 these diameters ; the ordinary diameters for boiler shells lie between 36 
 and 72 inches. For 36-inch boilers up to 48 inches diameter, 3-inch 
 tubes are commonly employed ; from 48 to 60 inches diameter of shell, 
 the tubes are commonly 2,/4 inches ; 4-inch tubular boilers usually range 
 
 * Fixtures comprise : Full front with anchors, coking-plate and stack-plate, 
 grate- and bearing-bars, flue-plate or arch-bars, wall-binder bars and rods, expan- 
 sion-plates and rollers to go under boiler-brackets, rear ash-door and frame, and 
 smoke-stack with bands and guys. 
 
BOILERS AND FURNACES 
 
 from 54 to 72 inches diameter. The dividing Hne for changing tube- 
 diameter is a variable one and rests largely upon the judgment of the 
 
 designer, but the above 
 ■ ^^^' covers the usual practice. 
 
 The number of tubes 
 for a given head requires 
 correct judgment and lib- 
 eral spacing. A common 
 defect in the construction 
 of horizontal tubular boil- 
 ers is the insertion of too 
 many tubes, and for large 
 boilers the tubes are some- 
 times too small in diam- 
 eter. An example is here 
 given in Fig. 221, which 
 represents a 60-inch boiler 
 fitted with 91 tubes 3 inches 
 in diameter, arranged zig- 
 zag. This boiler - head 
 ought not to have contained more than 76 tubes 3 inches in diameter, 
 and preferably the tubes should have been placed in vertical rows. A 
 much better arrangement would have been to select tubes 2,% inches in 
 diameter, which would have j)ermitted the insertion of 56 when placed 
 in vertical rows. Over- 
 crowding of tubes pre- ^^^- ^'^'^■ 
 vents a proper circulation 
 of water and increases the 
 resistance to the free es- 
 cape of steam by imped- 
 ing the upward movement 
 of the heated water from 
 the shell to the steam- 
 space above. The spacing 
 of tubes to insure a proper 
 circulation of water in the 
 boiler ought to bear some 
 relation to the diameter 
 of the tube, and in care- 
 fully prepared specifica- 
 tions this is usually the 
 case. 
 
 A central water-space 
 is sometimes provided in the arrangement of tubes by separating them 
 into two groups, as shown in Fig. 222. 
 
EXTERNALLY FIRED BOILERS 
 
 This method of tube distribution was formerly believed to have pe- 
 culiar merit, because it permitted a vertical movement of the water in 
 the centre of the boiler, rising in the centre and descending at the sides ; 
 but it is quite probable that the reverse of this is true. In any event, 
 what is wanted is not a circulation of water around a group of tubes, but 
 a circulation between them, and as water, Uke every other fluid, will flow 
 in the direction of least resistance, these circular currents will be set up 
 within the boiler, one on either side, to the practical exclusion of water- 
 currents through the mass of tubes. The objection to such an arrange- 
 ment is, that scale is likely to form around the tubes, and this will, by 
 reason of its granular surface, retain such floating matter as may be in 
 mechanical suspension in the water, thus accumulating a coating much 
 thicker in a given time than would be the case if there were a freer circu- 
 lation of water between the tubes themselves. The recommendation is, 
 to dispense with this central space and set the tubes wider apart in their 
 horizontal centres. 
 
 The distance from the side of a tube to the inside of the boiler-shell 
 should not in the case of a 36-inch boiler be less than 2 inches, for a 
 48-inch boiler it should be not 
 
 less than 2^ inches, and for ^^^- ^^3- 
 
 50 inch boilers and larger the 
 distance should be not less 
 than 3 inches, to secure good 
 circulation of water. 
 
 The distance from the side 
 of a tube to the shell of a 
 boiler is limited by the radius 
 of the flange of the head, as 
 shown in Fig. 223, which re- 
 presents a 36-inch machine- 
 flanged head ^ inch thick, 
 with an outside radius of i^ 
 inches and a 3-inch tube. As 
 there is more or less distortion 
 of the head where the flange- 
 curve merges into the flat 
 plate, the nearest distance at 
 which it is advisable to locate 
 a tube is ^ inch from the cen- 
 tre of the flange-curve ; this 
 makes the distance from the 
 
 inside of the shell to the tube 2 inches, as shown. The flange of a 
 72-inch head, >^ inch thick, with a 4-inch tube, is shown in Fig. 224, in 
 which the outside radius is 2 inches, and the distance from the inside of 
 the shell to the side of the tube is 2^ inches, which is the least practi- 
 
190 
 
 BOILERS AND FURNACES 
 
 cal limit in locating the tube. These limiting facts govern the location 
 of tubes with reference to the shell ; but the matter of circulation must 
 not be overlooked, and a further allowance of clear waterway will greatly 
 
 facilitate the circulation at 
 Fig. 224. the point where it is par- 
 
 ticularly needed to ac- 
 commodate the ascending 
 currents caused by the 
 contact of water with the 
 hot shell, and such in- 
 creased allowance will add 
 sufficiently to the effi- 
 ciency of the boiler to 
 fully compensate for any 
 loss of heating surface oc- 
 casioned by the omission 
 of a few tubes which, if 
 retained, might prove a 
 positive detriment to the 
 boiler. 
 
 The horizontal dis- 
 tance between tubes 
 should not in any case 
 be less than one-third the 
 diameter of the tube for 
 the smallest diameter of 
 shell in which tubes are to 
 be used ; and this distance 
 should increase in pro- 
 portion to the depth of 
 water below the top row 
 of tubes, — in other words, 
 the distance increases with the diameter of the shell. If tubes are too 
 close together horizontally, the circulation is interfered with and the 
 efficiency of the boiler lowered. The writer has in mind a boiler-shell 
 which bagged no less than three times in as many weeks under mod- 
 erately heavy firing, the reason for which was afterwards traced directly 
 to a want of proper circulation caused by the tubes being too close to- 
 gether to begin with, and this meagre allowance of space was further 
 lessened by a slight accumulation of scale, which so interfered with the 
 water circulation that the boilers utterly failed at a much less power- 
 rating than could have been secured had there been a proper allotment 
 of space between the tubes themselves, and between the outside of the 
 tube and the inside of the shell. 
 
 The horizontal spacing of tubes in the accompanying tables is based 
 
EXTERNALLY FIRED BOILERS 
 
 191 
 
 upon many years of observation, experiment, and trial. It is recom- 
 mended that for 3-inch tubes the horizontal centres be 4 inches for shells 36 
 to 40 inches diameter ; 4^-inch centres for shells 42 to 52 inches diameter ; 
 4^ -inch centres for shells 54 to 60 inches diameter. See Table XXXIX. 
 
 For 3^ -inch tubes the horizontal centres should be 5 inches for 48- 
 to 56-inch shells, 5^-inch centres for shells 58 to 62 inches diameter, 
 5^-inch centres for shells 64 and 66 inches diameter. See Table XLI. 
 
 For 4-inch tubes the horizontal centres may be 6 inches for all di- 
 ameters from 54 to 72 inches inclusive. See Table XLIII. 
 
 The vertical spacing of tubes is of less importance than the hori- 
 zontal spacing. In the tables above referred to this distance is uniformly 
 I inch between the outside diameters of the tubes, which is as little 
 metal as should intervene between them. 
 
 The upper limit of tubes varies somewhat, and may be said to lie 
 between three-fifths and two-thirds of the height from the bottom in 
 the best practice. The drawings from which the accompanying tables 
 were prepared had a uniform limit of two-thirds, the only variation being 
 in one or two cases where the tubes were carried about one-half inch 
 above this line in order to get 
 
 a better arrangement of tubes Fig. 225. 
 
 at the bottom of the boiler. 
 One of these was the 70-inch 
 boiler with 4-inch tubes, a size 
 of boiler quite unusual in trade 
 and not likely to be called for 
 in any engineering specifica- 
 tions. The two-thirds hmit 
 ought not to be encroached 
 upon, as the steam-room is 
 none too large after allowing 
 say three inches for depth of 
 water above the tubes. 
 
 The location of a tube with 
 reference to a flanged hand- 
 hole or manhole will depend 
 upon the radius of the curve 
 if flanged as in Fig. 225. The 
 illustration represents a 4 x 6 
 handhole in a 36-inch head, 
 the latter being ^ inch thick. 
 The radius for such a flanged 
 opening would probably be 
 j}{ inches, as shown, to which 
 
 an allowance of }4 inch is added, making the nearest distance from the 
 inside of the opening to the side of the tube i^ inches, to which an 
 
192 
 
 BOILERS AND FURNACES 
 
 TABLE XXXIX. 
 
 3-inch tubular boiler heads. 
 
 Fig. 226. 
 
 ,3 
 
 Diameter. 
 
 Manhole. 
 
 Tubes Centre 
 TO Centre. 
 
 u 
 
 Centre of 
 Boiler to 
 Centre of 
 Manholes. 
 
 1 
 H 
 
 '0 
 % 
 
 a 
 1 
 
 1 
 
 c 
 Si 
 
 Head. 
 
 Tube. 
 
 Upper. 
 
 Lower. 
 
 Hori- 
 zontal. 
 
 Verti- 
 caL 
 
 Upper. 
 
 Lower. 
 
 i 
 
 
 A. 
 
 B. 
 
 C. 
 
 D. 
 
 E. 
 
 F. 
 
 G. 
 
 H. 
 
 I. 
 
 s 
 
 ►J 
 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 
 Feet. 
 
 I 
 
 36 
 
 3 
 
 6 X 10 
 
 4x 6 
 
 4 
 
 4 
 
 A% 
 
 12 
 
 13 
 
 28 
 
 8 to 12 
 
 2 
 
 38 
 
 3 
 
 6 X 10 
 
 6 X 10 
 
 4 
 
 4 
 
 5 
 
 13 
 
 13 
 
 32 
 
 8 to 12 
 
 3 
 
 40 
 
 3 
 
 9X I4;5^ 
 
 6 X 10 
 
 4 
 
 4 
 
 5^ 
 
 i2y. 
 
 14 
 
 34 
 
 8 to 12 
 
 4 
 
 42 
 
 3 
 
 9x I4>^ 
 
 6 X 10 
 
 4X 
 
 4 
 
 5}^ 
 
 I3>^ 
 
 15 
 
 36 
 
 8 to 12 
 
 5 
 
 44 
 
 3 
 
 9 X 14/^ 
 
 9X I4>^ 
 
 4X 
 
 4 
 
 5^ 
 
 14)^ 
 
 i4>^ 
 
 38 
 
 10 to 12 
 
 6 
 
 46 
 
 3 
 
 9 X I4>^ 
 
 9xi4>^ 
 
 4X 
 
 4 
 
 6 
 
 ^5% 
 
 15^ 
 
 42 
 
 10 to 12 
 
 7 
 
 48 
 
 3 
 
 II X 15 
 
 9X 14^ 
 
 4X 
 
 4 
 
 6>^ 
 
 ^5% 
 
 i6>^ 
 
 46 
 
 10 to 12 
 
 8 
 
 50 
 
 3 
 
 II x 15 
 
 9X I4>^ 
 
 4)4 
 
 4 
 
 6^ 
 
 16% 
 
 i7>^ 
 
 52 
 
 10 to 12 
 
 9 
 
 52 
 
 3 
 
 II X 15 
 
 9x i4>^ 
 
 A% 
 
 4 
 
 7X 
 
 17/2 
 
 i8>^ 
 
 54 
 
 10 to 12 
 
 10 
 
 54 
 
 3 
 
 II X 15 
 
 9xi4>^ 
 
 A% 
 
 4 
 
 7>^ 
 
 i8>^ 
 
 i9>^ 
 
 60 
 
 10 to 12 
 
 II 
 
 56 
 
 3 
 
 II X 15 
 
 II X 15 
 
 A% 
 
 4 
 
 8 
 
 19/3 
 
 I9>^ 
 
 64 
 
 10 to 12 
 
 12 
 
 58 
 
 3 
 
 II X 15 
 
 II X 15 
 
 A'A 
 
 4 
 
 83^ 
 
 20 >^ 
 
 20>^ 
 
 70 
 
 10 to 12 
 
 13 
 
 60 
 
 3 
 
 II X 15 
 
 II X 15 
 
 AVz 
 
 4 
 
 8>^ 
 
 2I>^ 
 
 21/2 
 
 76 
 
 10 to 12 
 

 
 
 
 
 ZE OF Grate. 
 
 Commercial Horse-Power. 
 
 Length. 
 
 
 
 
 
 
 15 Feet. 
 
 12 Feet. 
 
 10 Feet. 
 
 Calculate 
 Feet. 
 
 Proposed 
 
 i Feet and 
 
 Inches. 
 
 
 2.75' @ 7 
 
 3'o" 
 
 14.74 
 
 18.43 
 
 22.11 
 
 3-15' @ 8' 
 
 3' 3" 
 
 18.43 
 
 23-03 
 
 27.64 
 
 3-54' @ 9 
 
 3' 6" 
 
 22.11 
 
 27.64 
 
 33-17 
 
 2.98' @ 7 
 
 3'o" 
 
 16.59 
 
 20.74 
 
 24.88 
 
 3-41' @ 8 
 
 3' 6" 
 
 20.74 
 
 25.92 
 
 31.10 
 
 3.83'® 9 
 
 4'o" 
 
 24.88 
 
 31-10 
 
 37-32 
 
 3-01' @ 7 
 
 3'o" 
 
 17-59 
 
 21.99 
 
 26.39 
 
 3.44' @ 8 
 
 3' 6" 
 
 21.99 
 
 27.49 
 
 32.98 
 
 3-86' @ 9 
 
 4'o" 
 
 26.39 
 
 32.98 
 
 39.58 
 
 3.04' @ 7 
 
 3'o" 
 
 23-25 
 
 29.06 
 
 34-87 
 
 3-47' @ 8 
 
 3' 6" 
 
 27.90 
 
 34-87 
 
 41-85 
 
 3.91' @ 9 
 
 4' 0" 
 
 32.55 
 
 40.69 
 
 48.82 
 
 3.05' @ 7 
 
 3'o" 
 
 24.50 
 
 30.63 
 
 36.76 
 
 3.49'® 8 
 
 3' 6" 
 
 29.40 
 
 36.76 
 
 44.11 
 
 3-92' @ 9 
 
 4' 0" 
 
 34-30 
 
 42.88 
 
 51.46 
 
 3.23'® 7 
 
 3' 3" 
 
 26.81 
 
 36.51 
 
 40.22 
 
 3.69'® 8 
 
 3' 9" 
 
 32.17 
 
 40.22 
 
 48.26 
 
 4.16' @ 9 
 
 4' 3" 
 
 37-54 
 
 46.92 
 
 56-30 
 
 340' ® 7 
 
 3' 6" 
 
 29.11 
 
 39-39 
 
 43-67 
 
 3.88' ® 8 
 
 4'o" 
 
 34-93 
 
 43-67 
 
 52.40 
 
 4-37' ® 9 
 
 4' 6" 
 
 40.76 
 
 50-95 
 
 61.14 
 
 3-68' @ 7 
 
 3' 9" 
 
 32.46 
 
 40.58 
 
 48.69 
 
 4.20'® 8 
 
 4' 3" 
 
 38.95 
 
 4S.69 
 
 58.43 
 
 4-73' @ 9 
 
 4' 9" 
 
 45-44 
 
 56.81 
 
 68.17 
 
 3.69' ® 7 
 
 3' 9" 
 
 33-72 
 
 42.15 
 
 50.5S 
 
 4.21' @ 8 
 
 4' 3" 
 
 40.47 
 
 50-58 
 
 60.70 
 
 4.74' @ 9 
 
 4' 9" 
 
 47.21 
 
 59-01 
 
 70.81 
 
 3.94' ® 7 
 
 4'o" 
 
 37-07 
 
 46.34 
 
 55-60 
 
 4.50'® 8 
 
 4' 6" 
 
 44-48 
 
 55-60 
 
 66.73 
 
 5-06'® 9 
 
 5'o" 
 
 51-90 
 
 64.87 
 
 77-85 
 
 4-05'® 7 
 
 4' 0" 
 
 39-38 
 
 49.22 
 
 59-07 
 
 4.63'® 8 
 
 4' 9" 
 
 47-25 
 
 59-07 
 
 70.88 
 
 5.20' @ 9 
 
 5' 3" 
 
 55-13 
 
 6S.91 
 
 82.69 
 
 4.28' @ 7 
 
 4' 3" 
 
 42-73 
 
 53-41 
 
 64.09 
 
 4.89' ® 8 
 
 5'o" 
 
 51-27 
 
 64.09 
 
 76.91 
 
 5-50' ® 9 
 
 5' 6" 
 
 59.82 
 
 74-77 
 
 89.72 
 
 4.49' @ 7 
 
 4' 3" 
 
 46.07 
 
 57-59 
 
 69.11 
 
 S-H' @ 8 
 
 5' 2" 
 
 55-29 
 
 66.11 
 
 82.93 
 
 5.78'® 9 
 
 5' 9" 
 
 64-50 
 
 80.63 
 
 96-76 
 
 193 
 
 ; manhole, 
 in a larger 
 
 Dituminous 
 diameters ; 
 od modern 
 orced draft 
 onger tube 
 in is ordi- 
 dth natural 
 
 I arrange- 
 ater-tubes, 
 •■ over each 
 g. 227, or 
 as in Fig. 
 :tically set- 
 :he former, 
 e the latter 
 Dyed. The 
 probably 
 belief that 
 1 current 
 le water to 
 leat on its 
 2. as would 
 
TABLE XL. 
 
 HORIZONTAL TUBULAR BOILERS, FITTED WITH 3-INCH TUBES. 
 
 
 Boiler 
 
 -Shell. 
 
 Number 
 of 3-inch 
 Tubes. 
 
 
 
 Heating Surface. 
 
 Grate Area. 
 
 Size of Grate. 
 
 Commercial Horse-Power. 
 
 1 
 
 i 
 
 Q 
 
 § 
 ^ 
 
 Internal 
 Area of 
 Whole 
 Number of 
 Tubes. 
 
 Tubes. 
 
 1 Shell. 
 
 Total. 
 
 Proport 
 
 ons to Tu 
 
 be Area. 
 
 Ratio of Grate Surfac 
 Heating Surface w 
 Area is 7, 8, and 9 
 of the Tube Area. 
 
 e to Total 
 hen Grate 
 imes that 
 
 Width in 
 Feet. 
 
 Length. 
 
 15 Feet. 
 
 12 Feet. 
 
 10 Feet. 
 
 1 
 
 7 Times. 
 
 "'-■ 
 
 9 Times. 
 
 7 Times. 
 
 8 Times. 
 
 9 Times. 
 
 Calculated 
 Feet. 
 
 Proposed 
 Feet and 
 Inches. 
 
 
 
 Inches. 
 
 Feet. 
 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 
 
 
 
 
 
 
 
 
 
 (-36 
 
 8 
 
 28 
 
 1. 18 
 
 175-93 
 
 45-20 
 
 221.13 
 
 8.26 
 
 9-44 
 
 10.62 
 
 26.77 : I 
 
 23-42 : I 
 
 20.82 : 1 
 
 3'o" 
 
 2.75' @ 7' 
 
 3'o" 
 
 14.74 
 
 18.43 
 
 22.11 
 
 I 
 
 J36 
 
 10 
 
 28 
 
 
 18 
 
 219.91 
 
 56-50 
 
 276.41 
 
 8.26 
 
 9-44 
 
 10.62 
 
 33-47 
 
 29.28 
 
 26.03 
 
 (3-0') 
 
 3-15' @ 8' 
 
 3' 3" 
 
 iS.43 
 
 23-03 
 
 27.64 
 
 
 •.36 
 
 12 
 
 28 
 
 
 18 
 
 263.89 
 
 67-80 
 
 331-69 
 
 8.26 
 
 9-44 
 
 10.62 
 
 40.16 
 
 35-14 
 
 31-23 
 
 
 3.54' ® 9' 
 
 3' 6" 
 
 22.11 
 
 27-64 
 
 33.17 
 
 
 f38 
 
 8 
 
 32 
 
 
 35 
 
 201.06 
 
 47-76 
 
 248.82 
 
 9-45 
 
 10.80 
 
 12.15 
 
 26.34 
 
 23-04 
 
 20.48 
 
 3' 2" 
 
 2-98' @ 7' 
 
 3'o" 
 
 16-59 
 
 20.74 
 
 24.88 
 
 2 
 
 ■ 38 
 
 10 
 
 32 
 
 
 35 
 
 251-33 
 
 59-70 
 
 311-03 
 
 9-45 
 
 10.80 
 
 12.15 
 
 32.91 
 
 28.80 
 
 25.60 
 
 (3.17') 
 
 3-41' @ 8' 
 
 3' 6" 
 
 20.74 
 
 25-92 
 
 31.10 
 
 
 ••38 
 
 12 
 
 32 
 
 
 35 
 
 301-59 
 
 71-64 
 
 373-23 
 
 9-45 
 
 10.80 
 
 12-15 
 
 39-50 
 
 34-56 
 
 30-72 
 
 
 3-83'® 9' 
 
 4'o" 
 
 24.88 
 
 31-10 
 
 37.32 
 
 
 j'40 
 
 8 
 
 34 
 
 
 43 
 
 213-63 
 
 50-24 
 
 263-87 
 
 10.01 
 
 11.44 
 
 12.87 
 
 26.36 
 
 23.07 
 
 20.50 
 
 3' 4" 
 
 3-0i' @ 7' 
 
 3'o" 
 
 17.59 
 
 21.99 
 
 26.39 
 
 3 
 
 ■ 40 
 
 10 
 
 34 
 
 
 43 
 
 267.04 
 
 62.80 
 
 329.84 
 
 10.01 
 
 11.44 
 
 12.87 
 
 32-95 
 
 28.83 
 
 25-63 
 
 (3-33') 
 
 3-44' @ 8' 
 
 3' 6" 
 
 21.99 
 
 27.49 
 
 32.98 
 
 
 ^o 
 
 12 
 
 34 
 
 
 43 
 
 320.44 
 
 75-36 
 
 395-80 
 
 10.01 
 
 11.44 
 
 12.87 
 
 39-54 
 
 34-60 
 
 30-75 
 
 
 3-85' @ 9' 
 
 4'o" 
 
 26.39 
 
 32.98 
 
 39.58 
 
 
 ■42 
 
 10 
 
 36 
 
 
 52 
 
 282.74 
 
 66.00 
 
 348-74 
 
 10.64 
 
 12.16 
 
 13-68 
 
 32-78 
 
 28.68 
 
 25-50 
 
 3' 6" 
 
 3-04' @ 7' 
 
 3'o" 
 
 23.25 
 
 29.06 
 
 34.87 
 
 4 
 
 ■ 42 
 
 12 
 
 36 
 
 
 52 
 
 339-29 
 
 79.20 
 
 418.49 
 
 10.64 
 
 12.16 
 
 13-68 
 
 39-33 
 
 34-42 
 
 30-59 
 
 (3-50') 
 
 3-47' @ 8' 
 
 3' 6" 
 
 27.90 
 
 34-87 
 
 41.85 
 
 
 '-42 
 
 14 
 
 36 
 
 
 52 
 
 395-84 
 
 92.40 
 
 488.24 
 
 10.64 
 
 12.16 
 
 13.68 
 
 45-89 
 
 40.15 
 
 35-69 
 
 
 3-91' @ 9' 
 
 4'o" 
 
 32.55 
 
 40.69 
 
 48.82 
 
 
 44 
 
 10 
 
 38 
 
 
 60 
 
 298-45 
 
 59.10 
 
 367-55 
 
 11.20 
 
 12.80 
 
 14.40 
 
 32.82 
 
 28.71 
 
 25-54 
 
 3' 8" 
 
 3-05' @ 7' 
 
 3'o" 
 
 24.50 
 
 30-63 
 
 36-76 
 
 5 
 
 " 44 
 
 12 
 
 38 
 
 
 60 
 
 358-14 
 
 82.92 
 
 441-06 
 
 11.20 
 
 12.80 
 
 14-40 
 
 39-38 
 
 34-46 
 
 30-63 
 
 (3-67') 
 
 3-49'® 8' 
 
 3' 6" 
 
 29.40 
 
 36.76 
 
 44.11 
 
 
 ^44 
 
 14 
 
 38 
 
 
 60 
 
 417-83 
 
 96-74 
 
 514-57 
 
 11.20 
 
 12.80 
 
 14.40 
 
 45-94 
 
 40.20 
 
 35-73 
 
 
 3-92'® 9' 
 
 4'o" 
 
 34.30 
 
 42.88 
 
 51-45 
 
 
 (•46 
 
 10 
 
 42 
 
 
 77 
 
 329.87 
 
 72-30 
 
 402.17 
 
 12.39 
 
 14.16 
 
 15-93 
 
 32-46 
 
 28.40 
 
 25-25 
 
 3' 10" 
 
 3-23'®/ 
 
 3' 3" 
 
 26.81 
 
 36.51 
 
 40.22 
 
 6 
 
 46 
 
 12 
 
 42 
 
 
 77 
 
 395-84 
 
 86.76 
 
 482.60 
 
 12.39 
 
 14.16 
 
 15-93 
 
 38-95 
 
 34-08 
 
 30-30 
 
 (3-83') 
 
 3-69'® 8' 
 
 3' 9" 
 
 32.17 
 
 40.22 
 
 48.26 
 
 
 Us 
 
 14 
 
 42 
 
 
 77 
 
 461.82 
 
 101.22 
 
 563-04 
 
 12.39 
 
 14.16 
 
 15-93 
 
 45-44 
 
 39-76 
 
 35-34 
 
 
 4.16'® 9' 
 
 4' 3" 
 
 37.54 
 
 46.92 
 
 56-30 
 
 
 (-48 
 
 10 
 
 46 
 
 
 94 
 
 361-29 
 
 75-40 
 
 436.69 
 
 13-58 
 
 15-52 
 
 17-46 
 
 32-16 
 
 28.14 
 
 25.01 
 
 4'o" 
 
 3-40'®/ 
 
 3' 6" 
 
 29.11 
 
 39.39 
 
 43-67 
 
 7 
 
 ■ 48 
 
 12 
 
 46 
 
 
 94 
 
 433-54 
 
 90.48 
 
 524-02 
 
 13-58 
 
 15-52 
 
 17.46 
 
 38-59 
 
 33-83 
 
 30-01 
 
 (4-00') 
 
 3-88' @ 8' 
 
 4'o" 
 
 34.93 
 
 43.67 
 
 52.40 
 
 
 I- 48 
 
 14 
 
 46 
 
 
 94 
 
 505-80 
 
 105-56 
 
 611.36 
 
 13-58 
 
 15-52 
 
 17-46 
 
 45-02 
 
 39-39 
 
 35-01 
 
 
 4-37' @ 9' 
 
 4' 6" 
 
 40.76 
 
 50.95 
 
 5I.I4 
 
 
 ("5° 
 
 10 
 
 52 
 
 
 19 
 
 408.41 
 
 78.50 
 
 486.91 
 
 15-33 
 
 17-52 
 
 19.71 
 
 31-76 
 
 27.79 
 
 24.70 
 
 4' 2" 
 
 3-68' @ 7' 
 
 3' 9" 
 
 32.46 
 
 40.58 
 
 48.69 
 
 8 
 
 ■ 50 
 
 12 
 
 52 
 
 
 19 
 
 490.09 
 
 94.20 
 
 584.29 
 
 15-33 
 
 17-52 
 
 19.71 
 
 38-n 
 
 33-35 
 
 29.64 
 
 (4-17') 
 
 4.20' @ 8' 
 
 4' 3" 
 
 38.95 
 
 48.69 
 
 58-43 
 
 
 '■SO 
 
 14 
 
 52 
 
 
 19 
 
 571-77 
 
 109.90 
 
 681.67 
 
 15-33 
 
 17-52 
 
 19.71 
 
 44-47 
 
 38.91 
 
 34-58 
 
 
 4.73' ® 9' 
 
 4' 9" 
 
 45.44 
 
 55.81 
 
 68.17 
 
 
 (52 
 
 10 
 
 54 
 
 
 28 
 
 424.12 
 
 81.70 
 
 505-82 
 
 15-96 
 
 18.24 
 
 20.52 
 
 31-69 
 
 27-73 
 
 24-65 
 
 4' 4" 
 
 3-69'® 7' 
 
 3' 9" 
 
 33.72 
 
 42.15 
 
 50.58 
 
 9 
 
 52 
 
 12 
 
 54 
 
 
 28 
 
 508.94 
 
 98.04 
 
 606.98 
 
 15-96 
 
 18.24 
 
 20.52 
 
 38-03 
 
 33-28 
 
 29.58 
 
 (4-33') ■ 
 
 4.2.' @ 8' 
 
 4' 3" 
 
 40.47 
 
 50.58 
 
 60.70 
 
 
 ^2 
 
 14 
 
 54 
 
 
 28 
 
 593-76 
 
 114-38 
 
 708.14 
 
 15-96 
 
 18.24 
 
 20.52 
 
 44-37 
 
 38-27 
 
 34-51 
 
 
 4-74'® 9' 
 
 4' 9" 
 
 47.21 
 
 59.01 
 
 70.81 
 
 
 (54 
 
 10 
 
 60 
 
 
 53 
 
 471.24 
 
 84-80 
 
 556-04 
 
 17.71 
 
 20.24 
 
 22-77 
 
 31.40 
 
 27-47 
 
 24.42 
 
 4' 6" 
 
 3-94' ® 7' 
 
 4' 0" 
 
 37.07 
 
 46.34 
 
 55-60 
 
 10 
 
 54 
 
 12 
 
 60 
 
 
 53 
 
 565-49 
 
 101.76 
 
 667-25 
 
 17.71 
 
 20.24 
 
 22.77 
 
 37-68 
 
 32-97 
 
 29-30 
 
 (4-50') 
 
 4-50'® 8' 
 
 4' 6" 
 
 44.48 
 
 55.60 
 
 66.73 
 
 
 ^4 
 
 14 
 
 60 
 
 
 53 
 
 659-74 
 
 118.72 
 
 778-46 
 
 17-71 
 
 20.24 
 
 22.77 
 
 43-96 
 
 38-46 
 
 34-19 
 
 
 5-06'® 9' 
 
 5'o" 
 
 51-90 
 
 64.87 
 
 77-85 
 
 
 (56 
 
 10 
 
 64 
 
 
 70 
 
 502.66 
 
 88.00 
 
 590-66 
 
 18.90 
 
 21.60 
 
 24.30 
 
 31-25 
 
 27-35 
 
 24-31 
 
 4' 8" 
 
 4-05'® 7' 
 
 4' 0" 
 
 39-38 
 
 49.22 
 
 59-07 
 
 II 
 
 56 
 
 12 
 
 64 
 
 
 70 
 
 603.19 
 
 105.60 
 
 708.79 
 
 18.90 
 
 21.60 
 
 24-30 
 
 37-50 
 
 32-81 
 
 29.17 
 
 (4-67') 
 
 4-63'® 8' 
 
 4' 9" 
 
 47-25 
 
 59.07 
 
 70.88 
 
 
 ^6 
 
 14 
 
 64 
 
 
 70 
 
 703-72 
 
 123.20 
 
 826.92 
 
 18.90 
 
 21.60 
 
 24.30 
 
 43-75 
 
 3S-28 
 
 34-03 
 
 .... 
 
 5.20' @ 9' 
 
 5' 3" 
 
 55-13 
 
 68.91 
 
 82.59 
 
 
 
 10 
 
 70 
 
 
 95 
 
 549-78 
 
 91.10 
 
 640.88 
 
 20.65 
 
 23-60 
 
 26.55 
 
 3104 
 
 27.11 
 
 24,14 
 
 4' 10" 
 
 4.28' @ 7' 
 
 4' 3" 
 
 42-73 
 
 53.41 
 
 64.09 
 
 12 
 
 {58 
 
 
 70 
 
 
 95 
 
 659-74 
 
 109.32 
 
 769.06 
 
 20.65 
 
 23.60 
 
 26-55 
 
 37-24 
 
 32.59 
 
 28.96 
 
 (4-83') 
 
 4.89'® 8' 
 
 5' 0" 
 
 51-27 
 
 64.09 
 
 76.91 
 
 
 Us 
 
 ,4 
 
 7" 
 
 
 95 
 
 769-69 
 
 127-54 
 
 897.23 
 
 20.65 
 
 23-60 
 
 26-55 
 
 43-45 
 
 38.02 
 
 33-79 
 
 
 5-50'® 9' 
 
 5' 6" 
 
 59-82 
 
 74-77 
 
 89-72 
 
 
 reo 
 
 10 
 
 76 
 
 
 
 596.90 
 
 94.20 
 
 691.10 
 
 22.47 
 
 25.68 
 
 28.89 
 
 30.76 
 
 26.91 
 
 23.92 
 
 5'o" 
 
 4-49'® 7' 
 
 4' 3" 
 
 46.07 
 
 57.59 
 
 69.11 
 
 13 
 
 ■ 60 
 
 12 
 
 76 
 
 3 
 
 21 
 
 716.28 
 
 113-04 
 
 829.32 
 
 22.47 
 
 25.68 
 
 28.89 
 
 36-91 
 
 32-29 
 
 28.71 
 
 (5-00') 
 
 5.14'® 8' 
 
 5' 2" 
 
 55-29 
 
 66.11 
 
 82,93 
 
 
 '-60 
 
 M 
 
 76 
 
 3 
 
 21 
 
 835-67 
 
 131-88 
 
 967-55 
 
 22.47 
 
 25-68 
 
 28.89 
 
 43-06 
 
 37-68 
 
 33-49 
 
 
 5.78'® 9' 
 
 5' 9" 
 
 64-50 
 
 S0.63 
 
 96.76 
 
192 
 
 
 Diameter. 
 
 !Z 
 
 
 
 n 
 
 Head. 
 
 Tube. 
 
 <B 
 
 
 
 (2 
 
 A. 
 
 B. 
 
 
 Ins. 
 
 Ins. 
 
 I 
 
 36 
 
 3 
 
 2 
 
 38 
 
 3 
 
 3 
 
 40 
 
 3 
 
 4 
 
 42 
 
 3 
 
 5 
 
 44 
 
 3 
 
 6 
 
 46 
 
 3 
 
 7 
 
 48 
 
 3 
 
 8 
 
 50 
 
 3 
 
 9 
 
 52 
 
 3 
 
 10 
 
 54 
 
 3 
 
 II 
 
 56 
 
 3 
 
 12 
 
 58 
 
 3 
 
 13 
 
 60 
 
 3 
 
EXTERNALLY FIRED BOILERS 
 
 t93 
 
 extra ^ or ^ inch might be added if practicable. If a large manhole, 
 such as the standard ii x 15 inches, be employed, it will be in a larger 
 boiler with a thicker head, 
 
 say Y-z inch or more in ^^^- ^^7- 
 
 thickness, the outer ra- 
 dius of the curve will be 
 greater and may vary 
 from i^ to 2 inches, de- 
 pending upon the dies 
 used, so that it is alto- 
 gether probable that no 
 less distance could be 
 safely made than 2^ 
 inches instead of i^, as 
 shown in Fig. 225. 
 
 The length of tube is 
 governed somewhat by 
 its diameter, and for the 
 three sizes in common 
 use, viz., 3, 3^, and 4 
 inches, this length ap- 
 proximates 50 diameters, if to be used in connection with bituminous 
 coal. For anthracite coal the length may be increased to 60 diameters ; 
 and these two proportions, with slight variations, represent good modern 
 
 practice. If a forced draft 
 Fig. 228. be employed a longer tube 
 
 can be used than is ordi- 
 narily the case with natural 
 draft. 
 
 The vertical arrange- 
 ment of the water-tubes, 
 whether directly over each 
 other, as in Fig. 227, or 
 placed zigzag, as in Fig. 
 228, is now practically set- 
 tled in favor of the former, 
 though for a time the latter 
 was much employed. The 
 zigzag spacing probably 
 originated in a belief that 
 an intercepted current 
 would enable the water to 
 take up more heat on its 
 way upward than if it proceeded directly to the water-surface, as would 
 be the case when the tubes are placed in vertical rows. 
 
194 
 
 BOILERS AND FURNACES 
 
 TABLE XLL 
 
 3>^-inch tubular boilers. 
 
 Fig. 229. 
 
 E 
 
 Diameter. 
 
 Manhole. 
 
 Tubes Centre 
 TO Centre. 
 
 Z S t, 
 WMfe 
 U 
 
 Centre of 
 Boiler to 
 Centre of 
 Manholes. 
 
 1 
 1 
 
 1 
 
 
 Head. 
 
 Tube. 
 
 Upper. 
 
 Lower. 
 
 Hori- 
 zontal. 
 
 Verti- 
 cal. 
 
 Upper. 
 
 Lower. 
 
 "0 
 
 ■5, 
 
 i 
 
 A. 
 
 B. 
 
 C. 
 
 D. 
 
 E. 
 
 F. 
 
 G. 
 
 H. 
 
 I. 
 
 s 
 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 
 Feet. 
 
 14 
 
 48 
 
 3/2 
 
 II X 15 
 
 9xi4>^ 
 
 5 
 
 A% 
 
 6^ 
 
 15^ 
 
 163^ 
 
 34 
 
 12 to 14 
 
 15 
 
 50 
 
 Z% 
 
 II X 15 
 
 9xi4>^ 
 
 5 
 
 A% 
 
 6>^ 
 
 16/^ 
 
 -^1% 
 
 38 
 
 12 to 14 
 
 16 
 
 52 
 
 3% 
 
 II X 15 
 
 9X I4;5^ 
 
 5 
 
 A% 
 
 7X 
 
 Ij/z 
 
 18% 
 
 46 
 
 12 to 14 
 
 17 
 
 54 
 
 Z% 
 
 II X 15 
 
 9X 14;^ 
 
 5 
 
 4K 
 
 7X 
 
 18/2 
 
 ^9% 
 
 47 
 
 12 to 14 
 
 18 
 
 56 
 
 3% 
 
 II X 15 
 
 II X 15 
 
 5 
 
 A% 
 
 1% 
 
 ^9% 
 
 19^ 
 
 50 
 
 12 to 14 
 
 19 
 
 58 
 
 z% 
 
 II X 15 
 
 II X 15 
 
 5X 
 
 4^ 
 
 IYa 
 
 20% 
 
 20 j^ 
 
 52 
 
 12 to 14 
 
 20 
 
 60 
 
 2,% 
 
 II X 15 
 
 II X 15 
 
 5X 
 
 aY^ 
 
 8X 
 
 21K 
 
 2I>^ 
 
 56 
 
 14 to 16 
 
 21 
 
 62 
 
 Z% 
 
 II X 15 
 
 II X 15 
 
 5X 
 
 A'A 
 
 8Y 
 
 22% 
 
 22>^ 
 
 60 
 
 14 to 16 
 
 22 
 
 64 
 
 2>% 
 
 II X 15 
 
 II X 15 
 
 5K 
 
 A% 
 
 9 
 
 23>^ 
 
 23>^ 
 
 64 
 
 14 to 16 
 
 23 
 
 66 
 
 Z% 
 
 II X rs 
 
 II X 15 
 
 <>%. 
 
 A% 
 
 9X 
 
 24^ 
 
 24^ 
 
 70 
 
 14 to 16 
 

 Boiler-Shell, ize of Grate. 
 
 Commercial Horse-Power. 
 
 s 
 
 3 
 
 
 
 Length. 
 
 
 
 
 1 
 
 B 
 G 
 
 g 
 
 
 
 15 Feet. 
 
 12 Feet. 
 
 10 Feet. 
 
 Pi 
 
 Calculate 
 Feet. 
 
 Proposed 
 
 1 Feet and 
 
 Inches. 
 
 
 
 Inches. 
 
 Feet. 
 
 
 
 
 
 
 
 (-48 
 
 12 
 
 3-45' ® 7 
 
 3' 6" 
 
 30.95 
 
 38.68 
 
 46.42 
 
 1-1 
 
 4S 
 
 14 
 
 3-94' @ 8 
 
 4' 0" 
 
 36.11 
 
 45.13 
 
 54.16 
 
 
 L» 
 
 i6 
 
 443'® 9 
 
 4' 6" 
 
 41.26 
 
 51. 58 
 
 61.89 
 
 
 (50 
 
 12 
 
 3-69' @ 7 
 
 3' 9" 
 
 34-13 
 
 42.66 
 
 51.19 
 
 15 
 
 5. 
 
 14 
 
 4.22' @ 8 
 
 4' 3" 
 
 39-81 
 
 49.77 
 
 59.72 
 
 
 I 50 
 
 i6 
 
 4-75' @ 9 
 
 4' 9" 
 
 45-50 
 
 56.88 
 
 68.25 
 
 
 V-52 
 
 12 
 
 4-32'®' 7 
 
 4' 6" 
 
 40.24 
 
 50.31 
 
 60.37 
 
 i6 
 
 14 
 
 4-93' @ 8 
 
 5'o" 
 
 46.95 
 
 58.69 
 
 70.43 
 
 
 i6 
 
 5-55' ® 9 
 
 5' 6" 
 
 53-66 
 
 67.08 
 
 80.49 
 
 
 (-54 
 
 12 
 
 4-25' @ 7 
 
 4' 3" 
 
 41.23 
 
 51.53 
 
 61.84 
 
 17 
 
 54 
 
 14 
 
 4.85'® 8 
 
 5'o" 
 
 48.10 
 
 60.12 
 
 72.15 
 
 
 u 
 
 x6 
 
 5-46' @ 9 
 
 5' 6" 
 
 54-97 
 
 68.71 
 
 82.45 
 
 
 (-56 
 
 12 
 
 4-35' @ 7 
 
 4' 6" 
 
 43-68 
 
 54.60 
 
 65.52 
 
 i8 
 
 56 
 
 14 
 
 4-97' ® 8 
 
 5'o" 
 
 50.96 
 
 63.70 
 
 76.44 
 
 
 (-56 
 
 i6 
 
 5-59' @ 9 
 
 5' 6" 
 
 58.24 
 
 72.S0 
 
 87.36 
 
 
 (-58 
 
 12 
 
 4-38'® 7 
 
 4' 6" 
 
 45-39 
 
 56.74 
 
 68.09 
 
 19 
 
 5S 
 
 14 
 
 5.00' @ 8 
 
 5'o" 
 
 52.96 
 
 66.20 
 
 79-44 
 
 
 (.58 
 
 i6 
 
 5-63' @ 9 
 
 5' 9" 
 
 60.52 
 
 75.66 
 
 90.79 
 
 
 CO 
 
 Leo 
 
 12 
 
 4-55' ® 7 
 
 4' 6" 
 
 48.57 
 
 60.72 
 
 72.86 
 
 20 
 
 14 
 
 5.20' @ 8 
 
 5' 3" 
 
 56-67 
 
 70.84 
 
 85.00 
 
 
 i6 
 
 5.85' ® 9 
 
 6' 0" 
 
 64.78 
 
 80.96 
 
 97-15 
 
 
 (-62 
 
 12 
 
 ! 4-71'® 7 
 
 4' 9" 
 
 51-76 
 
 64.70 
 
 77-64 
 
 21 
 
 62 
 
 14 
 
 5.38'® 8 
 
 5' 6" 
 
 60.39 
 
 75.48 
 
 90.58 
 
 
 (.62 
 
 i6 
 
 : 6.06' @ 9 
 
 6'o" 
 
 69.01 
 
 86.27 
 
 103.52 
 
 
 (-64 
 
 12 
 
 I 4-86' @ 7 
 
 5'o" 
 
 54-94 
 
 68.67 
 
 82.41 
 
 22 
 
 6. 
 
 14 
 
 5-57'® 8 
 
 5' 6" 
 
 64.10 
 
 80.12 
 
 96.14 
 
 
 .[64 
 
 i6 
 
 6.20' @ 9 
 
 6' 3" 
 
 73.26 
 
 91.57 
 
 109.88 
 
 
 (-66 
 
 12 
 
 5-17' @ 7 
 
 5' 3" 
 
 59-59 
 
 74.49 
 
 89-39 
 
 23 
 
 66 
 
 14 
 
 5.91' @ 8 
 
 6' 0" 
 
 69.52 
 
 86.91 
 
 104.29 
 
 
 (66 
 
 i6 
 
 6.64' @ 9 
 
 6' 9" 
 
 79.46 
 
 99.32 
 
 119.1S 
 
TABLE XLII. 
 
 HORIZONTAL TUBULAR BOILERS, FITTED WITH 3>^-INCH TUBES. 
 
 ^ 
 
 Boiler-Shell. 
 
 Number 
 
 of 
 3^^-inch 
 Tubes, 
 
 Internal 
 Area of 
 Whole 
 Number of 
 Tubes. 
 
 Heating Surface. 
 
 Grate Area. 
 
 SI 
 
 ZE OF Grate. 
 
 Commercial Horse-Power. 
 
 i 
 
 i 
 
 1 
 
 
 Tubes. 
 
 1 Shell. 
 
 Total. 
 
 Proport 
 
 ons to Tube Area. 
 
 Ratio of Grate Surface to Total 
 Heating Surface when Grate 
 Area is 7, 8, and 9 times that 
 of the Tube Area. 
 
 Width in 
 Feet, 
 
 Length. 
 
 15 Feet. 
 
 12 Feet, 
 
 10 Feet. 
 
 pi 
 
 7 Times. 
 
 8 Times. 
 
 9 Times. 
 
 7 Times. 
 
 8 Times. 
 
 9 Times. 
 
 Calculated 
 Feet. 
 
 Proposed 
 Feet and 
 Inches, 
 
 
 
 Inches. 
 
 Feet. 
 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 
 
 
 
 
 
 
 
 
 
 r48 
 
 12 
 
 34 
 
 1.97 
 
 373-73 
 
 90.48 
 
 464.21 
 
 13-79 
 
 15-76 
 
 17-73 
 
 33-66 : I 
 
 29.45 : I 
 
 26.18 : I 
 
 4'o" 
 
 3-45' ® 7' 
 
 3' 6" 
 
 30.95 
 
 38,68 
 
 46-42 
 
 lA 
 
 .« 
 
 14 
 
 34 
 
 1.97 
 
 436.02 
 
 105.56 
 
 541-58 
 
 13-79 
 
 15-76 
 
 17-73 
 
 3927 
 
 34-36 
 
 30-,55 
 
 (4-0') 
 
 3-94' @ 8' 
 
 4'o" 
 
 36,11 
 
 45-13 
 
 54-16 
 
 
 (^8 
 
 16 
 
 34 
 
 1.97 
 
 498.30 
 
 120.64 
 
 618.94 
 
 13.79 
 
 15-76 
 
 17-73 
 
 44.88 
 
 39-27 
 
 34-90 
 
 
 4-43' @ 9' 
 
 4' 6" 
 
 41,26 
 
 51 -58 
 
 61.89 
 
 
 rso 
 
 12 
 
 38 
 
 2.20 
 
 417.70 
 
 94.20 
 
 511.90 
 
 15.40 
 
 17.60 
 
 19.80 
 
 33-24 
 
 29.09 
 
 25-85 
 
 4' 2" 
 
 3-69'® 7' 
 
 3' 9" 
 
 34-13 
 
 42.66 
 
 51-19 
 
 15 
 
 50 
 
 14 
 
 33 
 
 2.20 
 
 487-31 
 
 109.90 
 
 597-21 
 
 15.40 
 
 17.60 
 
 19.S0 
 
 3S.78 
 
 33-93 
 
 30.16 
 
 (4-17') 
 
 4.22' @ 8' 
 
 4' 3" 
 
 39-81 
 
 49-77 
 
 59-72 
 
 
 (50 
 
 16 
 
 38 
 
 2.20 
 
 556-93 
 
 125.60 
 
 6S2.53 
 
 15-40 
 
 17.60 
 
 19.80 
 
 44-32 
 
 38.78 
 
 34-47 
 
 
 4-75' @ 9' 
 
 4' 9" 
 
 45-50 
 
 56.88 
 
 68.25 
 
 
 rs^ 
 
 12 
 
 46 
 
 2.67 
 
 505-63 
 
 98.04 
 
 603.67 
 
 18.69 
 
 21.36 
 
 24-03 
 
 32-30 
 
 28.27 
 
 25.12 
 
 4' 4" 
 
 4-32' @ 7' 
 
 4' 6" 
 
 40,24 
 
 50-31 
 
 60,37 
 
 i6 
 
 r^ 
 
 14 
 
 46 
 
 2.67 
 
 5S9.90 
 
 ,14-38 
 
 704.28 
 
 18.69 
 
 21.36 
 
 24.03 
 
 37-68 
 
 32-97 
 
 29,31 
 
 (4-33') 
 
 4-93' @ 8' 
 
 S'o" 
 
 46,95 
 
 58.69 
 
 70,43 
 
 
 (5. 
 
 16 
 
 46 
 
 2.67 
 
 674.18 
 
 130.72 
 
 804.90 
 
 18.69 
 
 21.36 
 
 24.03 
 
 43-07 
 
 37-68 
 
 33.50 
 
 
 5-55' @ 9' 
 
 5' 6" 
 
 53-66 
 
 67.08 
 
 80.49 
 
 
 p^ 
 
 12 
 
 47 
 
 2-73 
 
 516.62 
 
 101.76 
 
 618.38 
 
 19.11 
 
 21.84 
 
 24-57 
 
 32-36 
 
 28.31 
 
 25-17 
 
 4' 6" 
 
 4-25' ® 7' 
 
 4' 3" 
 
 41-23 
 
 51-53 
 
 61.84 
 
 17 
 
 54 
 
 14 
 
 47 
 
 2-73 
 
 602.73 
 
 118.72 
 
 721.45 
 
 19.H 
 
 21.84 
 
 24-57 
 
 37-75 
 
 33-03 
 
 29-36 
 
 (4-50') 
 
 4.85'® 8' 
 
 5'o" 
 
 48,10 
 
 60.12 
 
 72.15 
 
 
 (54 
 
 i5 
 
 47 
 
 2.73 
 
 688.83 
 
 135-68 
 
 824.51 
 
 19.11 
 
 21.84 
 
 24-57 
 
 43-15 
 
 37-75 
 
 33-56 
 
 
 5-46'® 9' 
 
 5' 6" 
 
 54-97 
 
 68.71 
 
 82.45 
 
 
 (56 
 
 12 
 
 50 
 
 2.90 
 
 549-60 
 
 105.60 
 
 655-20 
 
 20.30 
 
 23.20 
 
 26.10 
 
 32.28 
 
 28,24 
 
 25.10 
 
 4' 8" 
 
 4-35' ® 7' 
 
 4' 6" 
 
 43-68 
 
 54-60 
 
 65-52 
 
 18 
 
 5a 
 
 14 
 
 50 
 
 2.90 
 
 641.20 
 
 123.20 
 
 764.40 
 
 20.30 
 
 23.20 
 
 26.10 
 
 37-66 
 
 32-95 
 
 29,29 
 
 (4-670 
 
 4,97' @ 8' 
 
 s'o" 
 
 50-96 
 
 63.70 
 
 76.44 
 
 
 (-56 
 
 16 
 
 50 
 
 2.90 
 
 732.80 
 
 140.80 
 
 873-60 
 
 20.30 
 
 23.20 
 
 26.10 
 
 43-04 
 
 37-66 
 
 33-47 
 
 
 5-59' ® 9' 
 
 5' 6" 
 
 58.24 
 
 72.S0 
 
 87-36 
 
 
 rss 
 
 12 
 
 52 
 
 3-02 
 
 571-5S 
 
 109.32 
 
 680.90 
 
 21.14 
 
 24.16 
 
 27.18 
 
 32.21 
 
 28.18 
 
 25-05 
 
 4' 10" 
 
 4-38'® 7' 
 
 4' 6" 
 
 45-39 
 
 56.74 
 
 68.09 
 
 J9 
 
 5S 
 
 14 
 
 52 
 
 3.02 
 
 666.85 
 
 127-54 
 
 794-39 
 
 21.14 
 
 24.16 
 
 27-18 
 
 37.58 
 
 32.88 
 
 29,23 
 
 (4-87') 
 
 5-00' @S' 
 
 s'o" 
 
 52,96 
 
 66,20 
 
 79-44 
 
 
 (5B 
 
 16 
 
 52 
 
 3.02 
 
 762.11 
 
 '45-76 
 
 907.87 
 
 21.14 
 
 24.16 
 
 27.18 
 
 42-95 
 
 37-58 
 
 33-40 
 
 
 5-63' @ 9' 
 
 S'9" 
 
 60.52 
 
 75-66 
 
 90-79 
 
 
 /-60 
 
 12 
 
 5fi 
 
 325 
 
 615-55 
 
 113-04 
 
 728.59 
 
 22.75 
 
 26.00 
 
 29.25 
 
 32.02 
 
 28.02 
 
 24,91 
 
 S'o" 
 
 4-55' @ 7' 
 
 4' 6" 
 
 48,.57 
 
 60,72 
 
 72,86 
 
 20 
 
 .0 
 
 14 
 
 56 
 
 3-25 
 
 718.14 
 
 131.8S 
 
 850.02 
 
 22.75 
 
 26.00 
 
 29.25 
 
 37-36 
 
 32.69 
 
 29,06 
 
 (5-0') 
 
 5,20'® 8' 
 
 5' 3" 
 
 56.67 
 
 70,84 
 
 85,00 
 
 
 (fio 
 
 16 
 
 56 
 
 3-25 
 
 820.74 
 
 150.72 
 
 971.46 
 
 22.75 
 
 26.00 
 
 29.25 
 
 42.70 
 
 37-36 
 
 32.81 
 
 
 5-85' @ 9' 
 
 6' 0" 
 
 64,78 
 
 80,96 
 
 97-15 
 
 
 re. 
 
 12 
 
 60 
 
 348 
 
 659-52 
 
 116.88 
 
 776.40 
 
 24-36 
 
 27.84 
 
 31-32 
 
 31-87 
 
 27-89 
 
 24.79 
 
 5' 2" 
 
 4-71' @ 7' 
 
 4' 9" 
 
 51 -76 
 
 64-70 
 
 77-64 
 
 21 
 
 .2 
 
 14 
 
 60 
 
 3.48 
 
 769.44 
 
 136.36 
 
 905.80 
 
 24-36 
 
 27.84 
 
 31-32 
 
 37-18 
 
 32.54 
 
 28.92 
 
 (5-17') 
 
 5-38' @ 8' 
 
 5' 6" 
 
 60.39 
 
 75-48 
 
 90.58 
 
 
 (52 
 
 16 
 
 60 
 
 3.48 
 
 879.36 
 
 155-84 
 
 1035.20 
 
 24-36 
 
 27.84 
 
 31-32 
 
 42-50 
 
 37-18 
 
 33-05 
 
 
 6,06' @ 9' 
 
 6'o" 
 
 69.01 
 
 86.27 
 
 103-52 
 
 
 (64 
 
 12 
 
 64 
 
 3-71 
 
 703-49 
 
 120.60 
 
 824-09 
 
 25-97 
 
 29.68 
 
 33-39 
 
 31-73 
 
 27.77 
 
 24.68 
 
 5' 4" 
 
 4,86' @ 7' 
 
 S'o" 
 
 54-94 
 
 68.67 
 
 82.41 
 
 22 
 
 a. 
 
 14 
 
 64 
 
 3-71 
 
 S20.74 
 
 140.70 
 
 961.44 
 
 25-97 
 
 29.68 
 
 33-39 
 
 37-02 
 
 32-39 
 
 28.79 
 
 (5-33') 
 
 5-57'® 8' 
 
 5' 6" 
 
 64.10 
 
 80.12 
 
 96.14 
 
 
 .(64 
 
 16 
 
 64 
 
 3-71 
 
 937-98 
 
 160.80 
 
 1098.78 
 
 25-97 
 
 29-68 
 
 33-39 
 
 42.31 
 
 37.02 
 
 32,91 
 
 
 6,20' @ 9' 
 
 6' 3" 
 
 73-26 
 
 91-57 
 
 log.88 
 
 
 ,-66 
 
 12 
 
 70 
 
 4.06 
 
 769.44 
 
 124.44 
 
 893-88 
 
 28.42 
 
 32.48 
 
 36.54 
 
 31-45 
 
 27-52 
 
 24,46 
 
 5' 6" 
 
 5,17'® 7' 
 
 5' 3" 
 
 59-59 
 
 74-49 
 
 89-39 
 
 23 
 
 66 
 
 14 
 
 70 
 
 4.06 
 
 S97.68 
 
 145.18 
 
 1042.86 
 
 28.42 
 
 32.48 
 
 36.54 
 
 36.69 
 
 32.11 
 
 28,54 
 
 (5-50') 
 
 5,91'® 8' 
 
 6'o" 
 
 69,52 
 
 86.91 
 
 104.29 
 
 
 (66 
 
 16 
 
 70 
 
 4.06 
 
 1025.92 
 
 165-92 
 
 1191.84 
 
 28.42 
 
 32.48 
 
 36-54 
 
 41-94 
 
 36.69 
 
 32,62 
 
 
 6,64' ® 9' 
 
 6' 9" 
 
 79.46 
 
 99-32 
 
 119,18 
 
EXTERNALLY FIRED BOILERS 1 95 
 
 Ratio of Tube Area to Grate Area. — This area is often referred 
 to in essays on steam boilers as if the tube area in a boiler was to be 
 fixed after the grate surface had been determined upon ; as a matter of 
 fact, this is not the method by which this ratio is obtained. For a given 
 diameter of boiler a certain diameter of tube is decided upon. The 
 number of tubes to be placed in the head will be the greatest that the 
 upper limit of water-space and distance from boiler-shell will allow, 
 provision being made for the proper manhole or handhole under the 
 tubes. 
 
 It was found at an early date that boilers built with a certain ratio 
 of tube area to grate surface did better work than others of different 
 proportions, and that with a certain fixed relation of tube area to grate 
 surface the evaporative efficiency of a boiler could be approximately 
 forecast. 
 
 If a tubular boiler has been designed upon extent of heating surface 
 alone, the tube area is, of course, known in advance. The proportionate 
 grate area for such tube area will depend somewhat upon the kind of 
 fuel to be used and upon its rate of combustion. For anthracite coal 
 the grate area may be 8 times the sectional area of tubes for ordinary 
 .draft. For bituminous coal the grate area may be 7 times, or under 
 exceptional conditions as little as 6 times, the sectional area of tubes. 
 If the fuel be of indifferent quality or the draft be sluggish, the propor- 
 tion may be 9 of grate to i of tube area. A very good average is 8 of 
 grate to i of tube area. 
 
 Heating Surface and Grate Area. — The total heating surface of 
 a boiler, as commonly determined, is two-thirds the superficial area of the 
 shell, to which is added the superficial area of all the tubes. If less than 
 two-thirds the circumference of the boiler is exposed to the action of the 
 heated gases, then the shell heating surface will be in whatever propor- 
 tion of circumference of shell is thus exposed multiplied by the length 
 of the boiler. All heating surfaces are expressed in square feet. 
 
 The proportion of heating surface to grate area will vary according 
 to the kind of boiler. A two-flue boiler 48 inches by 26 feet, for ex- 
 ample, would have a ratio probably of 14 feet of heating surface to i of 
 grate; a 60-inch by 24-feet boiler with 18 6-inch tubes would have a 
 ratio of heating surface to grate area of about 32 to i ; a 72-inch by 20- 
 feet boiler with 68 4-inch tubes would have a ratio of heating surface to 
 grate area of about 54 to i. All the above are taken from actual 
 examples. Neglecting the first, we have a proportion varying from 32 
 to I to 54 to I, and between these two limits most of the tubular boilers 
 in use are likely to be found. 
 
 The coal burnt per hour per square foot of grate in each of the above 
 boilers was : For the two-flue boilers, 14.5 pounds ; for the 6-inch tubular 
 boiler, 41.38 pounds ; for the 4-inch tubular boiler, 35 pounds. Bitu- 
 minous coal was used in each of the above. 
 
196 BOILERS AND FURNACES 
 
 The coal burnt per hour per square foot of heating surface was : For 
 the two-flue boiler, 0.934 pounds ; for the 6-inch tubular boilers, 1.277 
 pounds ; for the 4-inch tubular boiler, 0.650 pounds. 
 
 The temperature of the escaping gases was : For the two-flue boiler, 
 735° Fahr. ; for the 6-inch tubular boiler, 542° Fahr. ; for the 4-inch 
 tubular boiler, 585° Fahr. Of these, the temperature of the first is too 
 high ; the second approaches closely that temperature at which the 
 chimney gives its best working results ; the third is rather too high for 
 boilers of this type for best economy. 
 
 The heating surface in square feet required to develop i horse-power 
 (34/^ pounds of water evaporated per hour from and at 212° Fahr.) 
 was : For the two-flue boiler, 3.34 ; for the 6-inch tubular boiler, 4.74 ; 
 for the 4-inch tubular boiler, 6.61. 
 
 For small tubular boilers, the ratio of heating surface to grate area 
 will be found to vary from 25 to i up to 40 to i. All things considered, 
 a ratio of from 30 to 35 square feet of heating surface to i square foot 
 of grate surface will be found to give good results. Much will depend 
 upon the intensity of the draught, because this determines how many 
 pounds of coal shall be burned per square foot of grate in a given time. 
 The higher the rate of combustion the greater proportionally may be 
 the extent or ratio of heating surface. Excess of heating surface may 
 have the eflect of reducing the temperature of gases to a point so low 
 that the chimney will not give sufficient draft to get the most economical 
 results of the furnace. 
 
 The ratio of tube area and heating surface to grate surface in the case 
 of a 48-inch boiler fitted with 46 3-inch tubes : if the proportion of the 
 tube area to that of the grate be 7 to i, the grate area would contain 
 13.58 square feet ; if 8 to i, 15.52 square feet ; and if 9 to i, 17.46 
 square feet. As 3-inch tubular boilers vary in length, we will select 
 three, — viz., 10, 12, and 14 feet. There must be a ratio of grate to 
 total heating surface most economical for a given draft. For example, 
 the ratio of heating surface to grate surface when the latter is 7 times 
 that of the tube area, would be 32.14 to i of grate for a lo-foot boiler ; 
 for a 12-foot boiler this ratio is increased to 38.65 to i ; and for a 14-foot 
 boiler to 45.03 to i. 
 
 If the grate area be 8 times that of the tube area, then the ratio of 
 heating surface to grate surface would be : For a boiler 10 feet long, 
 28.12 to I of grate ; if 12 feet long the ratio would be 33.82 to i ; and 
 if 14 feet long the ratio would be 39.40 to i. 
 
 If the ratio of grate area to that of tube area be 9 to i, the ratio of 
 heating surface to grate area would be : For the lo-foot boiler, 25 to i ; 
 for the 12-foot boiler, 30.06 to i ; and for the 14-foot boiler, 35.03 to i. 
 
 An examination of these three lengths and three proportions of tube 
 area to grate area show that good average results will be secured if the 
 boiler is 1 2 feet long and the ratio of grate to tube area 8 to i . 
 
EXTERNALLY FIRED BOILERS 1 97 
 
 A 3^ -inch tubular boiler, 60 inches diameter, for example, will show 
 nearly the same results, — that is : If the ratio of grate to tube area is 7 
 to I , the best length approximates 1 2 feet ; if 8 to i , the best length 
 approximates 14 feet ; and if 9 to i, the best length approximates 16 
 feet. Each of these is a little less than 35 feet of heating surface to i of 
 grate. In case the ratio of grate be 8 to i of tube area, the length 
 might be 16 feet, provided the draught is good. 
 
 Referring now to the 4-inch tubular boilers : If a diameter of 66 
 inches be selected, this contains 56 4-inch tubes. If the proportion of 
 the grate area to tube area be 7 to i, the greatest length approximates 
 16 feet, and if 8 to i, the length might be increased to 18 feet, a ratio 
 of 36.08 to I, and if the ratio be 9 to i, a length of 20 feet would be 
 required to secure a ratio of 35.91 square feet of heating surface to i 
 of grate surface. By an examination of the accompanying table, in 
 which the ratio of total heating surface to grate surface is approxi- 
 mately 35 to I, it will be seen that the length of tube approximates 
 very closely 18 feet, which has been found in practice to be a very satis- 
 factory length for boilers from 5 to 6 feet in diameter fitted with 4-inch 
 tubes. 
 
 Tables. — The leading properties of horizontal tubular boilers from 
 36 to 72 inches diameter are conveniently arranged for reference in the 
 annexed tables. The strength of shells for corresponding and larger 
 diameters, together with the several kinds of riveting best suited to the 
 strength of plates used, has already been given in Tables XXIX., 
 XXX., and XXXI. The number and arrangement of tubes as given 
 in Tables XXXIX. to XLIV. accords with what has been said in this 
 chapter regarding horizontal spacing and the distance of top of upper row 
 of tubes from the bottom of the boilers. Regarding the latter, it may 
 be said that two-thirds is the highest permissible limit ; some designers 
 place it at three-fifths instead, in order to get a lower water level. Be- 
 tween these two limits lies the best modern practice. The tube area, 
 total heating surface, and grate area are all expressed in square feet. 
 The ratio of tube area to grate surface is given as 7, 8, and 9, these be- 
 ing the ordinary ratios in practice. The width of grate is in all cases 
 that of the diameter of the boiler ; the length is calculated for the fixed 
 ratios of 7, 8, and 9 feet. As these present fractional dimensions, a 
 proposed length of grate-bars for common use is appended. 
 
 Retarders for Fire-Tubes. — A retarder is a thin strip of sheet 
 iron fitting loosely in and running the whole length of a tube ; this iron 
 is twisted one or more spiral convolutions in its whole length. See 
 Fig. 231. 
 
 The insertion of retarders in fire-tubes or flues is but rarely practised 
 in this country. The selection of the word "retarder" is not a happy 
 one, though, of course, it does obstruct the flow of gas through a tube, 
 as any obstruction in a tube will retard the flow of the gases through 
 
BOILERS AND FURNACES 
 
 TABLE XLin. 
 
 4-inch tubular boilers. 
 
 Fig. 230. 
 
 g 
 
 Diameter. 
 
 Manhole. 
 
 Tubes Centre 
 TO Centre. 
 
 u 
 
 Centre of 
 Boiler to 
 Centre of 
 Manholes. 
 
 i 
 
 s 
 1 
 
 
 s 
 
 Head. 
 
 Tube. 
 
 Upper. 
 
 Lower. 
 
 Hori- 
 zontal. 
 
 Verti- 
 cal. 
 
 Upper. 
 
 Lower. 
 
 ■0 
 
 .£3 
 
 1 
 
 A. 
 
 B. 
 
 C. 
 
 D. 
 
 E. 
 
 F. 
 
 G. 
 
 H. 
 
 I. 
 
 J 
 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 
 Feet. 
 
 24 
 
 54 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 7 
 
 I8>^ 
 
 ^.%% 
 
 36 
 
 16 to 20 
 
 25 
 
 56 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 7X 
 
 I9K 
 
 ^9% 
 
 38 
 
 16 to 20 
 
 26 
 
 58 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 IYa 
 
 2oy2 
 
 20;5^ 
 
 40 
 
 16 to 20 
 
 27 
 
 60 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 8 
 
 2^.% 
 
 21^ 
 
 47 
 
 16 to 20 
 
 28 
 
 62 
 
 4 
 
 11 X 15 
 
 II X 15 
 
 6 
 
 5 
 
 8X 
 
 22% 
 
 22;5^ 
 
 49 
 
 16 to 20 
 
 29 
 
 64 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 8^ 
 
 23>^ 
 
 23>^ 
 
 51 
 
 16 to 20 
 
 30 
 
 66 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 9 
 
 24;5^ 
 
 24>^ 
 
 56 
 
 16 to 20 
 
 31 
 
 68 
 
 4 
 
 II X 15 
 
 11 X 15 
 
 6 
 
 5 
 
 9% 
 
 2h% 
 
 25^ 
 
 62 
 
 16 to 20 
 
 32 
 
 70 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 9K 
 
 26% 
 
 26% 
 
 64 
 
 16 to 20 
 
 33 
 
 72 
 
 4 
 
 II X 15 
 
 II X 15 
 
 6 
 
 5 
 
 10^ 
 
 21% 
 
 21% 
 
 74 
 
 16 to 20 
 
u 
 
 Boiler-Shell. 
 
 |zE OF Grate. 
 
 Commercial Horse-Power. 
 
 s 
 
 3 
 
 
 
 ^ Length. 
 
 
 
 
 o 
 
 5 
 
 ! 
 
 
 
 15 Feet. 
 
 12 Feet. 
 
 10 Feet. 
 
 1 
 
 Calculated 
 Feet. 
 
 Proposed 
 Feet and 
 Inches. 
 
 
 
 Inches. 
 
 Feet. 
 
 
 
 
 
 
 
 r54 
 
 16 
 
 4-26'® 7' 
 
 4' 3" 
 
 49.28 
 
 61.56 
 
 73.88 
 
 24 
 
 5. 
 
 18 
 
 4-89' @S' 
 
 5'o" 
 
 5541 
 
 69.26 
 
 83.11 
 
 
 u 
 
 20 
 
 5.48'® 9' 
 
 5' 6" 
 
 61.56 
 
 76.95 
 
 92.34 
 
 
 (-56 
 
 16 
 
 4-35'® 7' 
 
 4' 6" 
 
 51-83 
 
 64.78 
 
 77-74 
 
 25 
 
 56 
 
 18 
 
 4-97' @ 8' 
 
 5'o" 
 
 58.30 
 
 72.89 
 
 87.46 
 
 
 (56 
 
 20 
 
 5-59'® 9' 
 
 5' 6" 
 
 64.78 
 
 80.98 
 
 97.17 
 
 
 rss 
 
 16 
 
 4-42'®/ 
 
 4' 6" 
 
 54-39 
 
 67.99 
 
 81.58 
 
 26 
 
 5S 
 
 18 
 
 5-05' @ 8' 
 
 5'o" 
 
 61.19 
 
 76.49 
 
 91.78 
 
 
 (58 
 
 20 
 
 5-68'® 9' 
 
 5' 9" 
 
 67.99 
 
 84.98 
 
 101.98 
 
 
 reo 
 
 16 
 
 5-01' @ 7' 
 
 5'o" 
 
 62.54 
 
 78.17 
 
 93.81 
 
 27 
 
 ]eo 
 
 18 
 
 5-73' ©8' 
 
 5' 9" 
 
 70.35 
 
 87.94 
 
 105.53 
 
 
 Leo 
 
 20 
 
 6.44' @ 9' 
 
 6' 6" 
 
 78.17 
 
 97.72 
 
 117.26 
 
 
 r62 
 
 16 
 
 5.05'© 7' 
 
 5'o" 
 
 65.11 
 
 81.38 
 
 97.67 
 
 2S 
 
 .2 
 
 18 
 
 5-79' ©8' 
 
 5' 9" 
 
 73.25 
 
 91.56 
 
 109.88 
 
 
 (.62 
 
 20 
 
 6.51' ©9' 
 
 6' 6" 
 
 81.39 
 
 101.74 
 
 122.09 
 
 
 r64 
 
 16 
 
 5.11'©/ 
 
 5' 3" 
 
 67.68 
 
 84.60 
 
 101.52 
 
 29 
 
 64 
 
 .s 
 
 5.84' ©8' 
 
 6'o" 
 
 76.15 
 
 95.17 
 
 114. 21 
 
 
 (64 
 
 20 
 
 6.5/ ©9' 
 
 6' 6" 
 
 84.60 
 
 105.75 
 
 126.89 
 
 
 Ue 
 
 (.65 
 
 16 
 
 5-43' © 7' 
 
 5' 6" 
 
 73.60 
 
 92.00 
 
 110.40 
 
 30 
 
 18 
 
 6.21'® 8' 
 
 6' 3" 
 
 82.80 
 
 103.50 
 
 124.20 
 
 
 20 
 
 6.99'® 9' 
 
 7' 0" 
 
 92.00 
 
 115.00 
 
 138.00 
 
 
 .-68 
 
 16 
 
 ' 5-84'©/ 
 
 6' 0" 
 
 80.63 
 
 100.79 
 
 120.95 
 
 31 
 
 6S 
 
 18 
 
 6.67' @ 8' 
 
 6' 9" 
 
 90.71 
 
 113.39 
 
 136.07 
 
 
 (68 
 
 20 
 
 7-51' ©9' 
 
 7' 6" 
 
 100.79 
 
 125-99 
 
 151-19 
 
 
 (70 
 
 16 
 
 5-86'©/ 
 
 6'o" 
 
 83.21 
 
 104.01 
 
 124.81 
 
 32 
 
 70 
 
 18 
 
 6.70' @ 8' 
 
 6' 9" 
 
 93-61 
 
 117.01 
 
 140.41 
 
 
 (70 
 
 20 
 
 7-53'® 9' 
 
 7' 6" 
 
 104.01 
 
 130.01 
 
 156.02 
 
 
 (-72 
 
 16 
 
 6.58'©/ 
 
 6' 6" 
 
 94-71 
 
 119.22 
 
 142.06 
 
 33 
 
 r^ 
 
 18 
 
 7.52' ©8' 
 
 7' 6" 
 
 106.55 
 
 133-1S 
 
 159.82 
 
 
 (72 
 
 20 
 
 8.46'® 9' 
 
 8' 6" 
 
 118.38 
 
 147.98 
 
 177-58 
 
TABLE XLIV. 
 
 HORIZONTAL TUBULAR BOILERS, FITTED WITH 4-INCH TUBES. 
 
 Number 
 of 4-inch 
 Tubes. 
 
 Internal 
 Area of 
 Whole 
 Number of 
 Tubes. 
 
 Proportions to Tube Area. 
 
 imes. 9 Times. 
 
 Ratio of Grate Surface to Total 
 Heating Surface when Grate 
 Area is 7, 8, and 9 times that 
 of the Tube Area. 
 
 Size of Grate. 
 
 Commercial Horse-Power. 
 
 Proposed 
 Feet and 
 Inches. 
 
 r 60 
 } 60 
 
 Ceo 
 
 4.27 
 4.27 
 4.27 
 
 4-73 
 4-73 
 4.73 
 
 Sq. Feet. 
 603.07 
 678.46 
 753-84 
 636.58 
 716.15 
 795-72 
 670.08 
 753-84 
 837.60 
 787-34 
 885-76 
 
 923-45 
 1026.06 
 
 854-35 
 
 961.15 
 1067.94 
 
 938.11 
 1055-3'S 
 1172.64 
 1038.62 
 1168.45 
 1298.28 
 1072.13 
 
 1340.16 
 1239-65 
 1394.60 
 
 135-68 
 152.64 
 169.60 
 140.S0 
 158.40 
 176.00 
 145-76 
 163.98 
 182.20 
 
 155-84 
 175-32 
 
 160.80 
 1S0.90 
 
 207,40 
 170.88 
 192.24 
 213.60 
 176.00 
 
 Sq. Feet. 
 738.75 
 831.10 
 923-44 
 777-38 
 874-55 
 971.72 
 815.84 
 917.82 
 
 1019.80 
 938-06 
 
 1055-32 
 
 1172.58 
 976.69 
 
 1098.77 
 
 1360.69 
 1511.88 
 
 Sq. Feet. 
 19-25 
 19-25 
 19-25 
 
 25-13 
 25-13 
 25-13 
 
 26.18 
 27.23 
 27.23 
 27.23 
 29.89 
 
 34-16 
 39-55 
 
 24.40 
 24.40 
 24.40 
 28.72 
 28.72 
 28.72 
 29.92 
 29.92 
 29,92 
 
 34.16 
 34.16 
 34-16 
 37-84 
 37-84 
 37-84 
 39-04 
 
 45-20 
 45-20 
 
 Sq. Feet. 
 24-75 
 24-75 
 24-75 
 
 27-45 
 27-45 
 27-45 
 32-31 
 32-31 
 32-31 
 33-66 
 
 42-57 
 42-57 
 42.57 
 43-92 
 43-92 
 43-92 
 50-85 
 50.85 
 50-85 
 
 42.99 
 47-77 
 37-43 
 42.11 
 46.79 
 37-31 
 41-95 
 46-63 
 37-28 
 41-94 
 46,60 
 36-94 
 41-55 
 46-17 
 36-53 
 41.10 
 45-66 
 36-54 
 41.10 
 45-67 
 35-98 
 
 45-23 
 
 3369 : 
 37-90 
 42.11 
 33-51 
 37-70 
 41.88 
 33-44 
 37.62 
 41.80 
 32-75 
 36-85 
 40.94 
 32-64 
 36-72 
 40.80 
 
 36.36 
 40,40 
 31.96 
 35-96 
 39-95 
 31-97 
 35-97 
 39-67 
 31-49 
 35-36 
 39-56 
 
 29.96 : 
 33-70 
 37-45 
 29-78 
 33-51 
 37-23 
 29.72 
 33-44 
 37-15 
 29.11 
 32-75 
 36-39 
 29.02 
 32-64 
 36-27 
 29.00 
 32.62 
 36.25 
 28.73 
 32.32 
 35-91 
 28.41 
 31-96 
 35-52 
 28.42 
 31-97 
 35-52 
 27.99 
 31-49 
 
 4' 10" 
 (4-83') 
 
 50" 
 (5-0') 
 
 5' 2" 
 (5-17') 
 
 5' 4" 
 (5-33') 
 
 5' 6" 
 (5-50') 
 
 5' 8" 
 (5.67') 
 
 5' 10" 
 (5-83') 
 
 6'o" 
 (6.0') 
 
 4.26' ( 
 4.89' ( 
 
 4-35' a 
 
 5-01' 
 5-73' 
 6.44' @ 9' 
 5-05' @ 7' 
 
 5-79' @ 
 6,51'® 9' 
 5-11'® 7' 
 5-84' @ 8' 
 6.57' @ 9' 
 5-43' @ 7' 
 6.21' @ 
 6.99' @ 9' 
 
 7-53' 
 
 6.5S' @ 7' 
 7.52' @ 8' 
 8.46' @ 9' 
 
 5' 6" 
 4' 6" 
 
 5' 6" 
 4' 6" 
 
 5' 9" 
 6' 6" 
 5' 0" 
 5' 9" 
 6' 6" 
 5' 3" 
 
 7' o" 
 6' o" 
 6' 9" 
 7' 6" 
 
 7'6" 
 6' 6" 
 7' 6" 
 
 61.56 
 51-83 
 58-30 
 64.78 
 54-39 
 61.19 
 67.99 
 62.54 
 70-35 
 78.17 
 65.11 
 73.25 
 81.39 
 67.68 
 76.15 
 84.60 
 73-60 
 
 90.71 
 100.79 
 
 83.21 
 
 93.61 
 104.01 
 
 94-71 
 
 69.26 
 76-95 
 64.78 
 
 67.99 
 76.49 
 84.98 
 78-17 
 
 '-55 
 
 95-17 
 10575 
 
 113-39 
 125-99 
 
 119.22 
 133-18 
 
EXTERNALLY FIRED BOILERS 
 
 199 
 
 it ; but the object of a retarder is a wholly different one, and 
 has nothing to do with the flow of the gases. 
 
 Retarders* are intended to increase the amount of 
 heat transmitted to the tube surface from the hot gases, 
 and it does it in two ways : first, by a mixing action upon 
 the gas in the tubes. The friction upon the surface of the 
 retarder aids in stirring up the gas in its passage through 
 the tube and in mixing the hot gas at the centre with the 
 cold film next the surface of the tube. Also in every hori- 
 zontal tube there is a tendency for the gases to be cooler 
 in the upper part of the tube and hotter in the lower, for 
 the upper part of the tube extracts heat far more readily 
 from the gases than the lower half. The twist of the re- 
 tarder has the effect of repeatedly turning over the gas in 
 the tube as it flows along. 
 
 In the second place, the retarder acts by direct radia- 
 tion of heat to the tube surface. While this action may 
 not be apparent at first sight, it is of such importance that 
 it should be clearly understood. To this end the fact 
 should first be realized that the temperature of the tube 
 surface exposed to the fire in any steam boiler is practi- 
 cally the same as that of water in contact with it, no 
 matter what the temperature of the gas on the other side, 
 supposing, of course, the tube surface to be clean. The 
 reason is that water absorbs heat many times as rapidly as 
 gas. Now suppose we place in the tube any solid body, 
 of any shape whatever. Manifestly, as it is surrounded 
 and bathed on all sides by gases at a temperature of say 
 1000° Fahr. , it will, if it loses no heat, soon become of the 
 same temperature as the hot gases. Suppose the surface 
 of the tube is of a temperature of 300° Fahr., the hot 
 body at the centre of the tube will energetically radiate 
 heat to the walls of the tube and will materially increase 
 the amount of heat transmitted to the water. 
 
 Whitham's Experiments. — A trial of retarders on a 
 100 horse-power horizontal tubular boiler was conducted by 
 J. M. Whitham, to ascertain under what conditions, if any, 
 they would add to the efficiency of the boiler. The boiler 
 was 60 inches diameter by 20 feet long, fitted with 44 tubes 
 4 inches in diameter. The water-heating surface was 11 37 
 square feet. Ratio of heating surface to grate surface was 
 42.6 to I. 
 
 * Jay M. Whitham and C. W. Baker, "Transactions Ameri- 
 can Society of Mechanical Engineers," vol. xvii. 
 
200 
 
 BOILERS AND FURNACES 
 
 Fig. 232. 
 
 The conclusions reached were : 
 
 Retarders in fire-tubes of a boiler interpose a resistance varying with 
 the rate of combustion. 
 
 Retarders result in reducing the temperature of the waste gases and 
 in increasing the effectiveness of the heating surface of the tubes. 
 
 Retarders show an economic advantage when the boiler is pushed, 
 varying in the tests from 3 to 18 per cent. 
 
 Retarders should not be used when boilers are used very gently and 
 when the stack draught is small. 
 
 It is probable that retarders can be used with advantage in plants 
 using a fan- or steam-blast under the fire, or a strong natural or induced 
 chimney draught, when burning either an- 
 thracite or bituminous coals. 
 
 Retarders may often prove to be, for an 
 existing plant, as economical as an econo- 
 mizer, and will not, in general, interpose as 
 much resistance to the draught. 
 
 The economic results obtained on the 
 boiler tested are ideal, showing that it was 
 clean, the coal good in quality, and the firing 
 skilful. 
 
 With retarders, the tubes are more effec- 
 tively cleaned than without their use. 
 
 Baker's Experiments. — These were 
 made to test the efficiency of radiators with 
 the apparatus shown in the accompanying 
 sketch. Fig. 232. It represents a section of 
 a single tube of a vertical boiler. The water 
 space surrounding it is well protected by 
 non-conductors. Through the tube a cur- 
 rent of hot gas is caused to flow from a lamp, 
 gas-jet, or other suitable source, and the 
 amount of heat transmitted to the water in a 
 given time is measured. The test is then 
 repeated under identical conditions, except 
 that a radiator of the form shown at the 
 upper part and at the side of Fig. 232 is 
 placed in the tube. The increased amount 
 of heat transmitted to the water is taken as 
 the amount due to the radiation from the 
 internal piece. 
 
 Experiments with the apparatus showed 
 the following general results : 
 That the percentage of increase in heat transmitted due to radiation 
 increases with increase in temperature of gases passing through the tube. 
 
EXTERNALLY FIRED BOILERS 
 
 Fig 
 
 That the percentage of increase in heat transmitted due to radiation 
 is larger in vertical tubes than in horizontal, on account of the fact that 
 a given area of heating surface in a horizontal tube absorbs heat faster 
 by direct contact with the gases than the same area in a vertical tube. 
 
 Experiments on actual boilers indicate that either device is most 
 useful on boilers with short tubes of not too 
 small diameter and with an abundance of draught. 
 
 With either device the tube surface must be 
 kept clean, otherwise the increased efficiency 
 will soon disappear, as is the case with the Serve 
 ribbed tube, an illustration of which is given in 
 Fig. 233, when care is not taken in this respect. 
 
 The economic gain by the use of either 
 radiators or retarders depends entirely upon the 
 temperature at which the boiler is discharging 
 its hot gases. It may be assumed that every 
 
 100° Fahr. reduction in the temperature of the waste gases represents 
 from 5 to 10 per cent, saving in fuel. 
 
 In general it will not usually be found worth while to introduce either 
 retarders or radiators in the tubes of any boiler, unless the thermometer 
 shows its hot gases to be discharging at a temperature of over 550° Fahr. 
 
 Double-Deck Boilers. — These boilers consist of an upper and 
 lower horizontal cylindrical shell, as shown in Fig. 234. These two 
 
 Fig. 234. 
 
202 
 
 BOILERS AND FURNACES 
 
 Fig. 235. 
 
 shells are connected by two or more necks ; the present illustration 
 shows three necks, which is the common practice. The lower shell is 
 fitted with as many tubes as can be conveniently arranged in it, having 
 due regard for the manholes necessary for internal examination and 
 cleaning. The upper shell is fitted with no tubes, and is intended to be 
 about half filled with water ; these details are shown in the half-sectional 
 elevation, Fig. 235. The connecting necks are usually made of flange 
 
 steel, and are as large in diameter as 
 can be conveniently adapted to the 
 two shells : the larger the diameter 
 the better the circulation ; but how- 
 ever large these necks may be, the 
 circulation in a boiler of this type is 
 not as good as in the ordinary horizon- 
 tal tubular type, in which the steam- 
 space is included in the upper part of 
 the boiler. The only apparent reason 
 for constructing a boiler in this manner 
 is that of increasing the total heating 
 surface by carrying the fire-line nearer 
 the top of the outer shell, as well as 
 increasing the number of tubes in a 
 boiler of a given diameter. The cir- 
 cuit of the products of combustion 
 from the furnace is underneath and 
 around the sides of the lower shell to 
 the rear end, returning through the 
 fire-tubes to the front, thence under 
 the upper drum to the rear of the boiler ; the gases then pass off to the 
 chimney. A test of a boiler of this type, in which the lower shell was 
 54 inches in diameter by 12 feet in length, having 118 tubes 3 inches in 
 diameter, the upper drum being 32 inches in diameter and of the same 
 length, measured 1281 square feet of total heating surface. This boiler 
 evaporated 10 pounds of water from and at 212° Fahr. per pound of coal, 
 showing that boilers thus constructed have not sufficient advantages 
 over the ordinary horizontal tubular boilers to pay for their extra cost. 
 
 Triplex Boiler. — This boiler differs from the one previously de- 
 scribed in having two lower shells mounted side by side, and with liberal 
 connections from these two lower shells upward into a combined water- 
 and steam-drum above, as shown in Figs. 236 and 237. The circula- 
 tion of the water is upward through the front connections, returning 
 downward to the shells by similar rear connections, these being ar- 
 ranged, as shown in the drawings, to induce and maintain a longitudinal 
 circulation in a circuit through the upper drum and lower shells. To 
 avoid the difficulties ordinarily met in supporting boilers of this type, 
 
EXTERNALLY FIRED BOILERS 
 Fig. 236. 
 
 203 
 
 Fig. 237. 
 
204 BOILERS AND FURNACES 
 
 and to allow for expansion and contraction, the front ends of the tubular 
 shells are held in a fixed position, the rear ends being free to move back- 
 ward and forward. The weight of the boiler and contents is sustained 
 at the back partly by rollers underneath, and partly by slings which 
 suspend the weight from points overhead. The slings are equipped 
 with springs which, while maintaining a nearly uniform strain on the 
 slings and girders, allow by their elasticity for imperfections in work- 
 manship and adjustment, as well as for settling or other changes which 
 may occur ; they provide also for vertical expansion and contraction 
 of the parts by changes of temperature. The slings are so adjusted that 
 they will support about three-fourths of the weight, leaving the other 
 quarter to run on the rollers. 
 
 The principal dimensions of the boiler illustrated, from designs by 
 J. T. Fanning, are here given : 
 
 Diameter of each lower shell 58 inches. 
 
 Length between heads and length of tubes ... 16 feet 9 inches. 
 Number of tubes in each shell, 4 inches outside 
 
 diameter 62. 
 
 Diameter of steam-drum 48 inches. 
 
 Length of drum 16 feet. 
 
 Diameter of necks connecting shell and drums . 15 inches. 
 
 Area of heating surface, one boiler 2705 square feet. 
 
 Area of grate surface, upper 41.8 square feet. 
 
 Area of grate surface, lower 49.5 square feet. 
 
 Total area of grate surface 91.3 square feet. 
 
 Ratio of heating surface to total grate surface . . 29.6 to i. 
 
 The boilers were set with the Hawley down-draft furnace, described 
 in the next chapter. These boilers were erected under a guarantee by 
 the contractor that each boiler should develop 250 standard horse-power, 
 and evaporate 10^ pounds of water per pound of Youghiogheny coal. 
 The guarantees of the Hawley Down-Draft Furnace Company were that 
 each boiler should develop 300 horse-power when the chimney-draught 
 was 0.6 of an inch, evaporate 10.5 pounds of water per pound of com- 
 bustible from and at 212° Fahr. when burning Pennsylvania (bitumi- 
 nous) coal, and consume 95 per cent, of the smoke. 
 
 The results of tests made by G. H. Barrus, based on dry coal, are 
 given in the table of comparative performance of modern boilers. From 
 this it appears that the evaporation per pound of coal from and at 212° 
 Fahr. was 10.846 pounds, and the evaporation per pound of combustible 
 from and at 212° Fahr. was 11. 551 pounds, meeting the guarantees of both 
 contractors in respect to economy. The boiler developed 315.2 horse- 
 power with a chimney-draught of 0.25 of an inch water pressure for 1% 
 hours, meeting the conditions of contract relating to capacity, — no smoke 
 whatever escaping from the chimney except at times of firing, or when the 
 bed of coal was disturbed by the use of the poker ; even then the quan- 
 tity was exceedingly small and within the guarantee, and its color barely 
 
EXTERNALLY FIRED BOILERS 
 
 205 
 
 perceptible. Assuming that the boiler develops 250 horse-power and 
 operates 24 hours per day, the quantity of lump coal consumed on the 
 basis of this test would be 10 
 
 tons per day, and the quan- Fig. 238. 
 
 tity of slack 11. 11 tons per 
 day. At $3.65 per ton for 
 lump coal and $2.60 per ton 
 for slack coal, the cost of fuel 
 for a day's run of 24 hours on 
 each boiler when developing 
 250 horse-power would be 
 $36.50 for the lump coal and 
 $28.89 for the slack, — the dif- 
 ference between the two being 
 $7.51 in favor of the slack coal. 
 A vertical tubular boiler 
 externally fired, as shown in 
 Fig. 238, is not often met 
 with. It presents some good 
 features, however : one of 
 which is, that an equal amount 
 of heating surface is provided 
 within the combustion-cham- 
 ber as would have been pro- 
 vided and without the compli- 
 cations involved in the design 
 of a fire-box. This is simply 
 a cylindrical shell fitted with 
 tubes of a diameter and in 
 number suited to the diameter 
 of the boiler. Around the 
 outer shell of the boiler is riv- 
 eted a band of iron of sufficient 
 thickness and depth to sup- 
 port the boiler in place, and 
 without bringing undue strain 
 upon the rivets by which this 
 band is fastened to the boiler. 
 The furnace consists of a cir- 
 cular grate, as shown in the engraving, the height of the furnace being 
 proportioned to the length of the shell, which in this case is a little less 
 than one-half its height. A cast-iron cap covers the top of the furnace 
 walls, through which, and upon an annular ring included in the cap, the 
 boiler is suspended. The performance of this boiler is the same as that 
 of horizontal tubular boilers having the same heating surface. 
 
CHAPTER VI I. 
 
 BOILER FURNACES AND SETTINGS, 
 
 Before entering upon the design of a boiler furnace it is necessary 
 to know among other things what kind of a boiler is to be used, its size, 
 the kind of fuel to be used, and the rate of combustion. It is needless 
 to say that many boiler plants throughout the country are developing 
 less power than they ought because of defective designs for the furnace 
 and poor construction. 
 
 Batteries of Boilers. — If we select the horizontal tubular boiler 
 as an example, we shall have probably the most popular boiler in use at 
 this time. Such boilers are practically limited as to diameter from 36 
 to 72 inches, in length from 10 to 20 feet, there being, of course, occa- 
 sional variations outside these dimensions. For large powers the ques- 
 tion is raised at the outset whether two or more boilers shall be set in 
 a single furnace as a " battery' ' or whether each boiler shall be set singly. 
 Both methods are practised ; both have reasons for and against. Let 
 us take the case of a steam plant requiring six boilers to do the work, 
 with spare boilers for cleaning and repairs. If the boilers are set singly, 
 one spare boiler will be enough, or seven boilers in all ; if the boilers are 
 set in pairs, eight boilers will be required ; if three boilers are set in a 
 single furnace, nine boilers will be required. The settings for the single 
 boilers will be most expensive, the boilers in pairs less so, and for the 
 three boilers in battery the cheapest. Whatever is saved in furnace 
 walls can be applied towards the cost of the additional boilers. Two or 
 three cylinder and two-flue boilers, which are always of small diameter 
 relatively to tubular boilers, may preferably be set in a single furnace. 
 So also tubular boilers up to 48 inches diameter may be set in pairs ; 
 for larger diameters the writer prefers that they be set singly. 
 
 The Size of the Grate. — The ratio of heating surface to grate area 
 will vary with the kind of boiler : a cylinder boiler, for example, will 
 require less grate area than one presenting a larger heating surface, if 
 economic evaporation only be considered ; but such boilers are not run 
 along economic lines so much as that of capacity, and this necessitates 
 a larger grate area. In the battery of cylinder boilers. Fig. 212, which 
 represents actual practice, there are three boilers, each 3 x 36 feet. The 
 heating surface at f of the shell is 610 square feet for the three ; the 
 grate surface is 9 feet 8 inches wide by 6 feet long, — say 58 square feet, 
 — a ratio of heating surface to grate area of 10.87 to i. One reason for 
 the large grate area is that non-merchantable refuse is commonly burnt 
 206 
 
BOILER FURNACES AND SETTINGS 20/ 
 
 at coal mines, saw-mills, etc. , where cylinder boilers are mostly in use. 
 If refuse coal is burnt, the rate of combustion is much less than is the 
 case with a better grade of fuel, and this requires a larger grate area. 
 One important fact must not be overlooked in this connection, shell 
 heating surface is much more effective than tube surface. A tubular 
 boiler having a ratio of heating surface to grate area of 32.6 to i would 
 not be three times as effective as the cylinder boilers now under con- 
 sideration. 
 
 If a change be made from cylinder to two-flue boilers to do the same 
 work, there would be required three boilers 42 inches in diameter by 20 
 feet long, each boiler fitted with two 14-inch fiues. The combined heat- 
 ing surface is 882 square feet, or 272 square feet more than the cylinder 
 boilers. The grate surface for the two-flue boilers in a battery of three 
 would be about 12 feet wide by 4 feet long, — say 48 square feet ; then 
 882 -s- 48 = 18.37 to I is the ratio of heating surface to grate area, 
 about 70 per cent, greater than in the case of cylinder boilers. The 
 estimated steam-producing capacity of the two batteries of boilers is 
 practically alike. 
 
 To do the same work with a horizontal tubular boiler would require 
 one boiler say 66 inches in diameter by 16 feet long, with 56 4-inch tubes, 
 or a tube area of 4.27 square feet, the total heating surface of which 
 would be 1 104 square feet. Taking into account the inferior quality of 
 fuel, as was the case with the cylinder boiler, the ratio of grate area to 
 tube area may be 9 to i ; that is, the tube area being 4.27 square feet, 
 the grate area would be 38.43 square feet. If the length of the grate 
 be fixed at 6 feet as a maximum, we have 38.43 -^ 6 = 6.4 feet as its 
 width, — say 6 feet 5 inches. The ratio of total heating surface to grate 
 area would be 1104 -=- 38.43 = 28.74 to i. With a better quality of 
 fuel and good draft the ratio of grate area to tube area might be made 
 7 to I, in which case the grate area would be 4.27 X 7 ^ 29.9 square 
 feet. If the length of grate be 5^ feet, we have 29.9 -f- 5.5 = 5.43, — 
 say 5 feet 6 inches square. The ratio of heating surface to grate area 
 would be 1 104 -^- 29.9 = 38.6 to i. 
 
 To recapitulate : We have for the cylinder boiler i square foot of 
 grate for each 10.87 ^^^t of heating surface ; two-flue boilers, i square 
 foot of grate for each 18.37 ^et of heating surface ; tubular boiler, i 
 square foot of grate for each 28.74 ^^t of heating surface, on the basis 
 of 9 square feet of grate to i of tube area, or i square foot of grate for 
 each 38.60 square feet of heating surface, on the basis of 7 square feet 
 of grate to i of tube area. Taking the first three results, we have for 
 a given evaporation, or, as commonly stated, a given horse-power, with 
 three different sizes of grates, viz. : 10.87, 18.37, and 28.74 square feet. 
 The explanation is that the last figure will yield both high economy and 
 capacity, the middle figure yields less capacity and less economy, and 
 the first figure slightly less capacity and very much less economy ; the 
 
208 BOILERS AND FURNACES 
 
 temperature of the products of combustion being highest with the cyHn- 
 der boiler, perhaps 800° Fahr. , lower with the two-flue boiler, perhaps 
 600'' Fahr. , and least with the tubular boiler, perhaps 500° Fahr. 
 
 The rate of combustion will affect the proportions of grate relatively 
 to the total heating surface. Bituminous coal with natural draught will 
 burn from 10 to 45 pounds per hour per square foot of grate, a differ- 
 ence wide enough to show that for ordinary steam-boiler furnaces it is 
 not worth while to enter upon this problem without knowing all the 
 conditions which affect combustion, and these cannot usually be known 
 in advance. 
 
 The practical considerations which affect the size of the grate are 
 the diameter of the boiler, which determines the width of the furnace, 
 and the impossibility of keeping the fire clean at the bridge-wall if the 
 grates exceed 7 feet, which establishes the maximum limit of length in 
 hand-firing, and this ought to be shortened to 6 feet to get the best 
 results. For horizontal tubular boilers it is customary to set the side 
 walls 4 inches away from the boiler, as shown in Fig. 272. If this dis- 
 tance be brought vertically downward the greatest ordinary width of 
 furnace will be had ; the least width is seldom less than the diameter of 
 the boiler, see Fig. 271. One-thirty-fifth of the total heating surface 
 will make a good average area for the grate surface, or eight times the 
 combined area of the tubes will approximate good working conditions. 
 
 Grate-Bars. — These are usually made of cast iron, and consist of 
 alternate supports for the fuel to be burned and spaces for supplying the 
 air needed for combustion. Every portion of the entire area of the grate 
 surface should be made up of these alternate ' ' lands' ' and spaces. The 
 top surface of the grate should be perfectly level, otherwise it will be 
 difficult to properly clean the fire with ordinary hand-tools. The grates 
 must rest upon proper supports, and these must be of such form, dimen- 
 sions, and so located under the grates as to prevent their warping or 
 getting out of shape by the action of the radiant heat from the fire. 
 The grates should have at least one end free for expansion ; otherwise 
 they will get out of shape. 
 
 Fig. 239. 
 
 The clear opening through the grates is usually as large and as much 
 subdivided as possible, that each portion of the fire shall have its proper 
 supply of air. They commonly have from 3/8- to ^-inch openings, with 
 only as much lateral obstruction at the ends and centre of the bars as is 
 necessary to enable them to keep their shape. Such a grate is shown in 
 Fig. 239, the metal and spaces being about equally divided. Mr. Barrus's 
 
BOILER FURNACES AND SETTINGS 209 
 
 observations are, that " Grates with 50 per cent, air-space gave 3.2 per 
 cent, better results based on coal and 1.7 per cent, better results based 
 on other combustibles than grates with 60 per cent, air-space. 
 
 ' ' The smaller the rate of combustion, the smaller should be the 
 opening for draught through the grates. The gain probably comes 
 about by preventing the introduction of too great an excess of air over 
 that required for combustion. ' ' 
 
 The practical spacing of grate-bars for the admission of air will de- 
 pend on the kind of fuel to be burned. For bituminous coal it may 
 be ^ or f^ inch, but for anthracite coal — and especially when burning 
 the finer grades, such as rice, buckwheat, and pea coal — the distance 
 must be less, and may approximate inch in width. The larger the 
 area of opening, the less will be the resistance of the air passing into the 
 fire ; but if the draught is good, this is a matter of less moment than the 
 resistance of the air passing through the burning fuel. In the case of 
 fine anthracite this resistance becomes so great that it is necessary to 
 employ a forced draught in order to burn the fuel at a sufficiently rapid 
 rate to develop the full power of a boiler. 
 
 The distance for the level of the grates above the fire-room floor will 
 vary from 24 to 30 inches. The first is probably the best distance, all 
 things considered ; and nearly all fire-fronts approximate this height. 
 
 The slope of grate-bars is usually from the furnace-front towards the 
 bridge-wall in the proportion of about one inch in twenty of the length 
 of the bar. No advantage accrues in this ; and for stationary boilers, 
 externally fired, as good results may be had if the grates are set level. 
 Inclining grates to the rear probably originated in marine practice, in 
 order to protect the fireman from live coals rolling out of the furnace 
 when stoking in rough weather. 
 
 Fig. 240. 
 
 CT \[ 
 
 Grate-bars are commonly cast in pairs, as this adds much to their 
 stiffness over single bars, and are thus better able to resist bending and 
 twisting. Cast-iron bars should have a groove on the top along their 
 whole length, as in Fig. 239. This groove, filling with ashes, prevents 
 in a measure the clinkers adhering to the grate. Short bars suffer less 
 in distortion by overheating than long ones. Thin and deep grate-bars, 
 such as shown in Fig. 240, give excellent results. The bottom edge 
 
2IO 
 
 BOILERS AND FURNACES 
 
 should be no thicker than will insure a sound casting. Such bars are 
 
 No difference need be made in 
 the construction of grates be- 
 tween anthracite and bitumi- 
 
 usually cast singly and not in pairs. 
 Fig. 241. 
 
 nous coals. 
 
 A grate for a circular fire- 
 box is shown in Fig. 241 
 which does not differ essen- 
 tially from the one referred 
 to above, except that instead 
 of being confined to two or 
 three bars a larger number 
 are included in a single cast- 
 ing. The engraving shows 
 the grate to be made up of 
 three sections. It is obvious, 
 however, that the number of 
 sections will depend upon the 
 diameter of the fire-box. The 
 width of the sections ought not 
 to greatly exceed 12 inches. 
 A stationary grate with 
 tilting section at the end next the bridge-wall is shown in Fig. 242. 
 This is a very convenient arrangement for dropping ashes, clinkers, etc. , 
 into the ash-pit underneath. Grates as long as 6 feet should be provided 
 with some device similar to this to prevent ashes from accumulating 
 at the bridge-wall, which accumulation has the effect of shortening the 
 furnace and diminishing the grate area. 
 
 A herring-bone grate, shown in Fig. 243, is extensively used in this 
 country, its popularity not being confined to any one locality. The 
 angular spacing readily permits expansion and contraction to take place 
 
 Fig. 242. 
 
 J^'^^uuyuilljllllljyiiii^^^"' 
 
 Gi 
 
 m 
 
 without breakage. It withstands the effects of repeated heating and 
 cooling quite successfully. An adaptation of this grate for a circular 
 fire-box is shown in Fig. 244. 
 
 A revolving grate is shown in Fig. 245. This is a very convenient 
 arrangement for shaking the ashes out of the fire, for breaking up 
 masses of clinker, and for dumping the contents of the furnace into the 
 
BOILER FURNACES AND SETTINGS 211 
 
 ash-pit underneath. The two middle grates in the upper row show how 
 this is accompHshed ; the side bars show how the grates are kept level 
 by locking upon one side wall and then upon each other. 
 
 Fig. 243. 
 
 Shaking- or moving-grates are designed to facilitate cleaning the 
 fires and, without opening the fire-doors, the removal of loose ashes 
 and clinkers which would ordinarily fall into the ash-pit by the use of a 
 slice-bar, also to break up and loosen the bed of burning fuel if the latter 
 has a caking tendency. 
 
 Fig. 244. 
 
 Some very absurd claims have been put forth regarding the great 
 saving in fuel by the use of such grates, but the following extract from 
 a letter puts the question in a way too often overlooked, and that is in 
 reference to the protection afforded the fireman himself, who asks : "It 
 
 Fig. 245. 
 
 15^ # # ®®? 
 
 the doors have to be opened to clean the fires, is there not a loss in 
 the admission of cold air to the boiler, to say nothing regarding the 
 injury to the boiler itself? Again, can any fireman with a slice-bar, no 
 
BOILERS AND FURNACES 
 
 matter how dexterous he may be, thoroughly stir out the ashes from 
 every part of the grate, and if not, is not the supply of air to that 
 extent cut off? If a fireman can, with comparatively little effort and 
 without exposure to intense heat, thoroughly clean every inch of his 
 grate by using the shaking-grate, why, in the interest of humanity and 
 economy, should he not be furnished one ?' ' 
 
 Shaking-grates will break up a bed of non-caking coal or anthracite 
 coal very satisfactorily, but they do not always succeed in breaking up 
 large masses of coal which may have fused together because of its caking 
 quality. The fireman who relies upon a shaking-grate to break up such 
 masses of caked coal will waste a great deal of fine coke in process of 
 burning, which will drop into the ash-pit through the openings occa- 
 sioned by the tilting of the grate-bars. 
 
 The Butman grate, shown in Fig. 246, has long been used in the 
 Western States in furnaces burning bituminous coal. The grate is ex- 
 
 FiG. 246. 
 
 Q H o zzz^ CI zzzz!! o zzi~zz o zmzn o jo 
 \J \J KJ \J \J KU O 
 
 ceedingly simple and very effective ; it consists of two side-bars notched 
 to receive the grates, which are provided on their upper surface with 
 oblique cutting edges and interlocking fingers, which also form parallel 
 cutting edges. The cross-bars to which these interlocking fingers are 
 attached are corrugated in the direction of their length, and taper from 
 the top to the bottom. The spacing of the interlocking fingers is such 
 that a series of irregular openings occur over the whole surface, and these 
 openings are so dimensioned that a full supply of air is had for the fuel. 
 The fingers, having semicircular ends, always preserve the same distance 
 whether stationary or being rocked, this detail of construction preventing 
 the loss of small fuel by its falling through into the ash-pit when cleaning 
 the fire. Each rocking-bar has an arm projecting downward, attached 
 to a connecting bar operated from the front of the furnace. When the 
 fire is being cleaned by this rocking motion the whole surface is broken 
 
BOILER FURNACES AND SETTINGS 
 
 213 
 
 up at the same time, any accumulation of clinker being prevented by 
 the cutting action of the locking fingers. 
 
 A shaking-grate known as the ^tna is shown in Fig. 247. Upon 
 ordinary bearing-bars, common to all furnaces, are laid two or more 
 stationary bars, A (de- 
 pending upon the width Fig. 247. 
 of the furnace). Near 
 either end is an open- 
 ing, shown in one of 
 the detail drawings, in 
 which are placed two 
 rockers. These spring 
 freely like a scale bear- 
 ing upon pivoted edges. 
 Upon these rockers are 
 placed the moving bars, 
 B and C. When at 
 rest the grate-bars pre- 
 sent the same level sur- 
 face as an ordinary 
 
 grate. Attached to the front rocker is a socket, shown in the two 
 upper drawings, which reaches almost to the ash-pit door. A lever 
 put in this socket and worked imparts to both sets of bars at the 
 same time a vertical and horizontal movement over the entire surface. 
 By this action the entire surface of the fire is cleaned and opened, a 
 condition which can be maintained with little effort by one or two 
 movements of the lever at periodic intervals. 
 
 The Rose grate is shown in sectional elevation in Fig. 248. It differs 
 from the preceding designs in that a vertical movement is given each 
 alternate bar by rocking-levers located underneath the grate. 
 
214 
 
 BOILERS AND FURNACES 
 
 Each bar has V-shaped corrugations along its entire length. The 
 bars are of unusual depth and taper from top to bottom, and are well 
 calculated, therefore, to resist the action of fire. A plan of a set of 
 grates is shown in Fig. 249. A centre-bar extends the length of the 
 
 furnace, dividing it into 
 Fig. 249. two sections. This cen- 
 
 tre-bar locks into the two 
 end-pieces by a dovetail 
 joint, and thus serves to 
 hold the frame together. 
 Whatever number of bars 
 are assembled to make a 
 grate, usually from two 
 to three feet in width, are 
 set in a frame which is 
 entirely independent of 
 the boiler or brick-work. 
 Two rocking-levers are 
 suspended in each section 
 of the frame and coupled together, as shown in Fig. 248 ; additional to 
 this coupling-bar, there is still another, extending out through the fire- 
 front and there connected to a bell-crank provided with a socket for the 
 insertion of a hand-lever for operating or shaking the grates. A locking- 
 dog hinged to the front, the free end dropping into a socket in the bell- 
 crank, holds the locking levers in neutral position, but when the grates 
 are to be shaken it is lifted out of its socket and thrown back against 
 the front ; a to-and-fro movement of the hand-lever will raise or lower 
 and at the same time give a slight end movement to each alternate 
 grate-bar for each movement of the hand. The grate-bars in their up- 
 and-down movement pass each other at a sufficient distance to sift out all 
 ashes and dirt accumulated under the fire, but not far enough to allow 
 coal to fall through into the ash-pit. By the same movement of the 
 grates the fire-surface is lifted, allowing air to pass more freely through 
 the bed of fire than is common to the ordinary form of stationary grates. 
 Deterioration in Grates. — The principal cause which contributes 
 to the rapid burning out of grate-bars is the action of the furnace heat, 
 which will in time destroy any set of grates, but the want of a proper 
 flow of air through the grates will cause overheating, whether it occurs 
 through too little air-space in the grates themselves, or by these spaces 
 becoming obstructed through any cause, thus preventing the cooling 
 effect of the air on its passage to the fire ; but a more common reason 
 is found in the impurities of the coal, and especially in the chemical 
 combinations of sulphur and iron, which abound in more or less quan- 
 tity in all red-ash coals. Any coal which forms an easily fused clinker 
 will soon injuriously affect the grates. 
 
BOILER FURNACES AND SETTINGS 215 
 
 Effect of Water-Pan in Ash-Pit. — There is no advantage in 
 the use of a water-pan in stationary boilers, and its use is not recom- 
 mended. The common notion that the vapor rising from the surface 
 of the water and passing upward through the fire is decomposed and 
 assists combustion is a mistake. Action and reaction are equal, and 
 just as much energy is expended in disassociating the gases as is 
 afterwards obtained by their recombination. The ash-pans underneath 
 locomotive fire-boxes and marine boilers of the locomotive pattern have 
 water-pans as security against fire, which may be caused by live coals 
 dropping down upon wood-work. 
 
 Stationary boiler ash-pits should be kept dry and reasonably clean. 
 As all the heat of the cinders and ashes is taken up by the air on its way 
 to the grates, it will be seen that a hot, dry ash-pit is the best. 
 
 Mechanical Stokers. — The earliest mechanical stoker for steam- 
 boilers was probably that designed by James Watt (1785), which was 
 simply a device to push the coal after it was coked at the front end 
 of the grate back towards the bridge. The suggestion of Watt was 
 followed by numerous patents in Europe for mechanical stokers, but 
 none of these have been adopted to any considerable extent in this 
 country because of their supposed lack of adaptability to American 
 boiler practice. Their non-adoption here was partly because the mech- 
 anism was of too complicated a nature to be entrusted to the average 
 fireman, or the rate of speed at which the machine would be required to 
 run was such that it would be wanting in durability. Another objec- 
 tion was, that in many cases the grates were so placed as to be almost 
 inaccessible, either for examining or altering the condition of the fire by 
 hand or for the renewing of worn-out parts. 
 
 The mechanical stokers in use in this country are almost entirely of 
 American design, and are well adapted for the service to which they are 
 applied, whether for burning graded small coals or mine refuse, such as 
 screenings or other unmerchantable coal. A mechanical stoker is par- 
 ticularly suited to the burning of such refuse material, and instances are 
 plentiful in which more steam was supplied when using bituminous slack 
 or refuse coal in a furnace provided with an automatic stoker than the 
 ordinary grate-surface supplied when using lump coal and when hand- 
 fired, the boilers and all other accessories being similar, this increase in 
 performance being due to the constant supply of coal to the fire without 
 any disturbing influences. The labor required for boiler attendance 
 with a mechanical stoker is much less than that required for the ordinary 
 flat grate of like capacity and burning similar fuel. The time consumed 
 in cleaning a fire under a boiler provided with such a machine is claimed 
 to be less than one-tenth that required for similar work with ordinary 
 furnaces. 
 
 Fuel economy is one of the claims persistently urged by makers of 
 mechanical stokers, — not that a furnace fed by machine will accomplish 
 
2l6 
 
 BOILERS AND FURNACES 
 
 results different than if fed by hand under the best conditions, but hand- 
 firing occurs at irregular intervals, and usually more coal is put on the 
 fire than is necessary. A mechanical stoker with continuous feed will 
 supply the coal as needed without disturbing the progress of the fire, 
 not cooling it by an excess of coal, nor by the admission of cold air in 
 the furnace by the opening of the fire-door. 
 
 The prevention of smoke by the use of a mechanical stoker is largely 
 brought about by a uniform delivery of fuel to the grates, maintaining a 
 high temperature at all times in the furnace and combustion-chamber. 
 This constant fuel supply, especially in the case of fine coal, permits the 
 gradual distillation of the gases from the coal as it enters the furnace, 
 and the automatic movement of the grates keeps the air-spaces open, 
 insuring the combination of the oxygen of the air with the hydro- 
 carbon gases in the furnace, and the complete utilization of the latter 
 in heat-making. 
 
 The Murphy Furnace. — This furnace was specially designed for 
 the use of small-sized bituminous coal and slack. A cross-sectional ele- 
 vation is shown of it in Fig. 250, where the magazines into which the 
 coal is put are at the sides of the combustion-chamber. At the bottom 
 
 Fig. 250. 
 
 of each magazine is a casting, part of which is used as a coking-plate ; 
 the inclined grates rest against it at their upper ends. On the central 
 part of this plate is an inverted open box (called the stoker-box) with 
 a rack under each end. A shaft worked from the exterior has a pinion 
 geared to each rack, whereby the stoker-box is moved back and forth 
 on the coking-plate as rapidly or as slowly as may be required by the 
 rate of combustion. The stoker-boxes, through the motion thus com- 
 municated, push the coal from the edge of the coking-plates to the 
 grates below. 
 
BOILER FURNACES AND SETTINGS 
 
 217 
 
 Immediately over the coking-plate is the arch-plate. At the line 
 where the fire-brick comes in contact with the arch-plate are ribs an 
 inch apart ; these form the skewback on which the arch rests ; the spaces 
 thus formed between the ribs are air-ducts. The admission of air is 
 regulated by a register at the front. The air passes through the flues, 
 up and over the arch, taking up heat from the front, arch, and arch- 
 plate ; it then passes down through the small openings in the arch-plate 
 and supplies the coking fuel with the proper quantity of air and at a 
 high temperature. This secures the immediate combustion of the un- 
 combined gases. A longitudinal sectional elevation of this furnace is 
 shown in Fig. 251. 
 
 This furnace has sufficient coking capacity to feed 50 pounds of coal 
 per square foot of grate per hour, and slowly enough to give ample 
 time for the complete expulsion of the volatile gases. The air, which 
 is delivered in numerous jets, is so distributed as to thoroughly mix 
 with the liberated gases, resulting in the complete prevention of smoke 
 escaping from the furnace. By the time the coal is fairly on the grates 
 the gaseous part of the fuel is consumed ; what remains is coke, and 
 this is not smoke-producing. The coke gets its needed supply of air 
 through the grates and is burned in the usual way. The grates are 
 kept constantly in motion. The incoming air to the furnace passes be- 
 tween the bars and thus supplies the lower bed of fuel. The incandes- 
 cent coke as it burns is moved down towards the centre of the furnace, 
 at the bottom of which is a clinker-breaker for grinding any clinker and 
 refuse to be deposited in the ash-pit below. 
 
 Fig. 251 
 
 A single or double engine with automatic governor and proper gear- 
 ing is placed on one side of a battery of boilers for operating a recipro- 
 cating bar across the outside of the entire front, and to which all the 
 working parts are attached by links which may be detached or attached 
 
 15 
 
2l8 
 
 BOILERS AND FURNACES 
 
 at the will of the operator. Any one or all of the movable parts may 
 be detached from the automatic bar, and the furnaces may be success- 
 fully operated by hand, the fuel being fed in the same manner as when 
 controlled by the engine. 
 
 The Roney Mechanical Stoker is shown in sectional elevation in 
 Fig. 252. A hopper for the coal is attached to the boiler front; in the 
 lower part of the hopper is a pusher, to which is attached, by a flexible 
 connection, the feed-plate forming the bottom of the hopper. The 
 pusher, by a vibratory motion, carrying with it the feed-plate, gradually 
 forces the fuel over the dead-plate to the grates below. These grates 
 
 Fig. 252. 
 
 are horizontal flat-surfaced bars running from side to side of the furnace ; 
 they are carried on inclined side bearers extending from the throat of 
 the hopper to the rear and bottom of the ash-pit. The grates, there- 
 fore, in their normal condition form a series of steps, upon the top step 
 of which coal is fed from the dead-plate. These steps at the inclination 
 given, however, prevent the free descent of the coal, but each bar rests 
 in a concave seat in the bearer and is capable of a rocking motion 
 through an adjustable angle. All the grate-bars are coupled together 
 by a rocker-bar, the notches of which engage with a lug on the lower 
 rib of each grate-bar, pin connections being made with two of the grate- 
 bars only for the purpose of holding the rocker-bar in position. A 
 variable back-and-forth motion being given to the rocker-bar through 
 a connecting rod, the grate-bars necessarily rock in unison, now forming 
 
BOILER FURNACES AND SETTINGS 219 
 
 a series of steps and now approximating to an inclined plane, with the 
 grates partly overlapping. The depending webs of the grate-bars are 
 perforated with longitudinal slots, so placed that the condition of the 
 fire can be seen at all times and free access be had to all parts of the 
 grate to assist, when necessary, the removal of clinker. The slots also 
 serve an important purpose in furnishing an abundant supply of air 
 for combustion. 
 
 Assuming the grates to be covered by a bed of coal and fresh fuel 
 being fed in at the top, it is obvious that when the grates rock forward 
 the fire will tend to work down in a body. But before the coal can 
 move too far the bars rock back to the stepped position, checking the 
 downward motion, breaking up the bed of fuel over the whole surface, 
 and admitting a free volume of air through the fire. The rocking 
 motion is slow, being from seven to ten strokes per minute, according 
 to the kind of coal. This alternate starting and checking motion, being 
 continuous, keeps the fire constantly stirred and broken up from under- 
 neath and finally lands the cinder and ash on the dumping-grate below. 
 By releasing the dumping-rod the dumping-grate tilts forward, throwing 
 the cinder into the ash-pit, after which it is again closed, ready for further 
 operation. The dumping-grate is made in two parts, so that each half 
 can be dumped separately. 
 
 The actuating mechanism is simple. All motion is taken from one 
 driving-shaft. In a single stoker this shaft may either be driven through 
 a worm gear from a small engine attached to the boiler front, or it may 
 be driven by a link-belt from any convenient point of the nearest shaft. 
 In large batteries of boilers the driving-shaft is extended across all the 
 boiler fronts, delivering power to each stoker ; this, with the coal ele- 
 vators and conveyers, is driven by a small independent engine. The 
 largest stoker can easily be turned over by hand, indicating the small 
 amount of power necessary to operate it. The worm-gear shaft carries 
 a disk and wrist-pin from which a link couples to the agitator, shown 
 attached to the boiler front, underneath the dead-plate. Through the 
 eye of the agitator passes a stud screwed into the pusher, on which 
 stud is a feed-wheel by which the stroke of the pusher, and conse- 
 quently the amount of feed, is regulated. The agitator having a fixed 
 stroke, it is apparent that if the feed-wheel is run down against it in the 
 position shown in the engraving, the pusher will be given its full traverse 
 and the greatest feed. If run back to clear the travel of the agitator, 
 the pusher will, of course, have no motion, and the feed will stop. 
 Between these extremes any desired rate of feed can be given. 
 
 In a like manner the rock of the grate-bars can be adjusted between 
 any limiting angles, and over a range of motion from no movement to 
 full throw, by means of the sheath-nut and jam-nuts on the connecting 
 rod. By these two adjustments the whole action of the stoker is con- 
 trolled, the fires forced, checked, or banked at will. There are poker- 
 
220 
 
 BOILERS AND FURNACES 
 
 doors in the front, on each side of the hopper, through which the whole 
 grate can be seen and the condition of the cinder on the dumping-grate 
 determined. If the location is convenient, a cleaning-door may be 
 introduced into the side of the furnace to facilitate examination of the 
 boiler. 
 
 The ^A^ilkinson Mechanical Stoker is shown in sectional ele- 
 vation in Fig. 253 and in front elevation in Fig. 254. The grate-bars 
 are a series of hollow castings approximating a rectangular cross-section, 
 placed side by side, and inclined towards the bottom of the furnace at an 
 
 Fig. 253. 
 
 angle of about twenty-five degrees with the horizontal. The upper end 
 is open to admit the blast-pipe ; it projects through, and is supported 
 by the stoker front, the lower ends sliding on and supported by a hollow 
 cast-iron box, as shown. This lower box, or bearing-bar, has finger- 
 grates about 15 inches long secured to its rear face The bars are 4-inch 
 centres, so that, practically, the air-openings are restricted to the risers. 
 Throughout the inclined length and on the face of the bar is cast a suc- 
 cession of steps. Through the rise of each step a vent of about }{ by 
 3 inches is provided to admit air through the fire to the combustion- 
 chamber. 
 
 The feeding is accomplished by the motion of the grates. The 
 pusher shown in the bottom of the hopper and resting on the grate-bars 
 
BOILER FURNACES AND SETTINGS 221 
 
 is secured to each alternate bar by a dowel-pin and moves with them, 
 feeding the fuel in measured quantities from the hopper to the upper 
 end of the grate. The continuous back-and-forth motion of the grate- 
 bar is for the purpose of maintaining a uniform thickness of fire by a 
 gradual descent of the fuel from the top to the bottom of the grate, 
 depositing the clinker and ash on the stationary grate shown projecting 
 from the bearer-bar at the ash-pit. The accumulated ash is pushed off 
 this stationary grate into the ash-pit by the reciprocating motion of the 
 bars, to be removed in the usual manner or by special appliance. The 
 mechanism for effecting the entire operation of the stoker consists of a 
 pulley, compound gearing, toggle-shaft, and quadrant, all of which are 
 shown in the two engravings above referred to. 
 
 Fig. 254. 
 
 S 
 
 rirsntrinrinnrtn r\ ni r\ ni r\A 
 
 V3 C-T3 C7 — O O < ^ < ^ 1^ C7 O — C7 'U '^ [ 
 
 m 
 
 )00000000000000[ 
 
 s, 
 
 y 
 
 The blast is saturated steam, through a nozzle of y^g-inch opening, 
 giving an induced current of air controlled by a regulating valve . This 
 method of injecting the air into a hollow bar, from which there is no 
 escape except into and through the fire, is set forth as one of the merits 
 of this machine. The steam is decomposed by the incandescent fuel 
 and fills the combustion-chamber with burning gases, resulting, accord- 
 ing to Whitham, in a more uniform distribution of the effective heating 
 surface of the boiler, reduces local injury, gives a more uniform expan- 
 sion to the parts of the boiler, but is apt to cause a loss of heat in the 
 stack if not properly controlled. 
 
 The cost of the steam-blast is from 5 to 1 1 per cent, of the steam 
 
222 BOILERS AND FURNACES 
 
 generated by the boilers, estimated at $1000 per year for a 1000 horse- 
 power plant, on a lo-hour basis, when fuel is $2.00 per ton. 
 
 An automatic damper-regulator should not be used with this stoker ; 
 simply observe the best position of damper suited to the work and keep 
 it there. When a strong natural draft exists, it may be well to partly 
 close the damper at the chimney and force the steam-jets harder. 
 
 When running too slow, the fire will burn out before reaching the 
 bottom of the grate. When running too fast, live coal will be pushed 
 out with the ash. Between these two extremes a speed may be found 
 best suited to the requirements of any boiler plant. 
 
 The object of blowing steam into the grate is in part to assist the 
 combustion and prolong the life of grate-bars. Most anthracite coals 
 of inferior sizes clinker badly if the fire is forced ; such clinkers ' ' freeze' ' 
 to the sides of the furnace and to the grates ; the efifect of a steam-blast 
 is to chill the grates and non-combustible material against them, so that 
 a clinker cannot form. It is at all times necessary with this machine to 
 have some steam passing through the jets when fired up, therefore the 
 steam-jets should all be open and blowing alike. When the furnace is 
 in operation, the damper and steam-jets should never be entirely closed. 
 
 The Playford Mechanical Stoker is shown in longitudinal sec- 
 tional elevation in Fig. 255 and in cross-sectional elevation in Fig. 256. 
 As will be seen, it consists of an endless revolving chain of grate-bars, 
 operated by power, preferably by a small engine taking steam from the 
 boiler. A fire-brick arch is sprung across the boiler front under which 
 this machine is located when in use ; a similar arch is sprung across the 
 rear end of the furnace adjoining and attached to the bridge- wall. The 
 grate-bars are attached to each other by links, shown in perspective in 
 Fig. 257. The sprocket-wheels engage between the links, giving the 
 whole grate surface a movement inward, the rapidity of which is con- 
 trolled by the attendant to suit the needs of the furnace. The grates, 
 bearing-shaft, etc. , are mounted on a frame provided with wheels resting 
 upon rails, shown in both engravings. This permits withdrawal of the 
 whole grate surface and bed of fire from the furnace at any time. 
 
 A coal-hopper is provided at the boiler front ; an adjustable gauge 
 regulates the amount of coal which may pass into the furnace from the 
 hopper. As the movement of the grates is very slow, the volatile matter, 
 in the case of bituminous coal, is driven ofi" by the intense heat of the 
 front arch, and these gases are not only here brought to the point ot 
 ignition, but are consumed in the combustion-chamber immediately 
 adjoining. The incandescent carbon remaining on the grate is burnt 
 during the interval required for the passage of the grates up to the rear 
 arch, under which the grates begin their return to the front, the ashes 
 falling off into the ash-pit underneath, and being brought forward to the 
 front of the boiler or wherever else desired by the spiral ash-conveyer, 
 shown in Fig. 255. 
 
BOILER FURNACES AND SETTINGS 
 
 223 
 
224 
 
 BOILERS AND FURNACES 
 
 The Coxe Mechanical Stoker is shown in sectional elevation 
 through a furnace in Fig. 258. It consists of an endless travelling grate 
 which receives the fuel at one end, burns it as it moves slowly along, 
 
 Fig. 258. 
 
 and deposits the ashes at the other end. It was originally designed to 
 burn the finest sizes of anthracite coal and is shown arranged for this 
 grade of fuel, but it is capable of handling bituminous coal as well. 
 
 A coal-hopper, into which the coal is placed by the fireman, is 
 
BOILER FURNACES AND SETTINGS 
 
 225 
 
 located immediately in front of the boiler. The depth of coal upon the 
 grates is regulated by a sliding gate, which can be raised or lowered as 
 a thicker or thinner bed of fire is required. 
 
 ■ After the coal passes under the regulating gate, it flows over fire- 
 brick "ignition blocks" set on an incline immediately inside the fire- 
 front and above the grates, which blocks are used only with anthracite 
 coal ; these blocks, becoming incandescent, or nearly so, retain sufficient 
 heat to insure ignition of the incoming supply of fresh coal. In burning 
 bituminous coal these ignition blocks are omitted and a hopper is put 
 on that admits of the fuel being deposited directly on the grates, as no 
 extra care is necessary to insure perfect ignition of bituminous coal. 
 
 The entire grate is set travelling at the rate of four to six feet an 
 hour, or at any other like speed, according to the demands upon the 
 boiler, etc., and thus the fuel which has been fed in from the hopper 
 at the forward end of the furnace is burned completely during its slow 
 passage to the dumping end, where the ashes are left, while the grate 
 reverses underneath and comes back, as shown, to the front end of the 
 furnace, to pass in again like an endless belt. 
 
 The smaller sizes of anthracite coal pack down closely and cannot 
 be burned unless a forced draught is used, and as this draught should 
 vary in intensity according to the state of the fire, all the air that enters 
 the fire is conducted into the largest chamber at the centre of the frame, 
 see Fig. 258 ; and the only air that can enter the adjacent chamber must 
 pass through dampers in the partitions separating these chambers, which 
 are closed at the bottom, so that the supply can be regulated to suit 
 
 Fig. 259. 
 
 existing conditions. In practice, the maximum amount of air is sup- 
 plied at the centre, and this is diminished to a minimum at each end, 
 the dampers being regulated so that the pressure is diminished in each 
 succeeding chamber. The handles which regulate these dampers are 
 shown in Fig. 259 projecting from the sides of the frame. Preference 
 
226 
 
 BOILERS AND FURNACES 
 
 is given to a dry-air- or fan-blast, because with the fan-blast the rate of 
 combustion per square foot per hour is greater than with a steam-jet. 
 The percentage of carbon left in the ash is less, and as it is not necessary 
 to prevent the formation of clinker, which is the chief reason for using 
 steam, the loss and waste due to the steam-jet are avoided. It is claimed 
 for this furnace that the finer coals are burned with a less excess of air 
 than is common, and the results with all fuels are above the average. 
 
 The American Stoker is shown in sectional elevation in Fig. 260. 
 The coal-hopper is placed at the front of the boiler, in common with other 
 
 Fig. 260. 
 
 iU-'vcLOalz 
 
 stokers. Underneath this coal-hopper is a spiral conveyer, located in a 
 trough with a half-round bottom, shown in Fig. 261, which extends 
 nearly the entire length of the magazine. Immediately beneath is lo- 
 cated the wind-box, to which is connected the piping for air-blast ; the 
 other end of the wind-box opens into the air-space between the maga- 
 zine and outer casing. The upper end of the magazine is surrounded 
 by tuyeres, or air-blocks, these being provided with openings for the 
 discharge of air-blast. 
 
 The air is delivered in the approximate proportions of 1 50 cubic feet 
 of air to each pound of coal fed, and at a pressure ranging from ^ ounce 
 to I ounce at the tuyeres. This pressure is only such as to admit of the 
 thorough mixing of the air with the coal, and must not be confused with 
 the ordinary forced draught. A wind-gate controlled by a lever, shown 
 underneath the conveyer, or hopper. Fig. 260, enables the operator to 
 regulate the supply of air to suit the amount of coal fed. Being tlius 
 independent of natural draught in the air supply, the supply of coal 
 
BOILER FURNACES AND SETTINGS 
 Fig. 261. 
 
 !27 
 
 Fig. 262. 
 
 being also under complete control, the fire can be forced at a moment's 
 notice, or it can be as quickly reduced. 
 
 The operating mechanism, apart from the fan for furnishing the blast, 
 is located beneath and in front of the 
 hopper. A steam motor for opera- 
 ting the stoker is shown in end eleva- 
 tion in Fig. 262. The only moving 
 parts of this motor are the reciproca- 
 ting piston and rod, the valve move- 
 ment being similar in its construc- 
 tion and operation to that of a single 
 steam-pump. The piston-rod carries 
 a cross-head, which by means of suit- 
 able connecting links operates a pawl 
 mechanism, which in turn actuates the 
 ratchet-wheel mounted on the con- 
 veyer-shaft. 
 
 The space on each side of the 
 stoker, between the tuyere blocks and 
 the side-walls of the furnace, is oc- 
 cupied by dead plates or air- tight 
 grates. 
 
 The tuyere blocks shown in Fig. 
 260 are substantial castings. The air 
 
 passes through the blocks, protecting them to a considerable degree ; 
 but in case a block is damaged by the action of the heat it can be easily 
 
228 
 
 BOILERS AND FURNACES 
 
 and inexpensively replaced. These blocks are practically the only part of 
 the stoker subject to renewal. 
 
 The operation of this stoker is as follows : The coal is fed into the 
 hopper by hand or by any convenient mechanism ; from there it is 
 carried by the conveyer into the magazine, and then forced upward by 
 the action of the spiral conveyer until it overflows on both sides of the 
 tuyeres, spreading upon and over the dead-grates the entire width of the 
 furnace. The entire mass of the coal above the tuyeres and all of that 
 above the dead-grates is in active combustion, consisting usually of a 
 bed of burning coke from 14 to 18 inches in depth. The feeding of the 
 coal is continuous, thus keeping the incandescent body of fuel slightly 
 agitated and preventing the formation of large clinkers ; this agitation 
 also allows the air to thoroughly mix through the burning coal. The 
 amount of coal fed is regulated by the speed at which the motor is 
 run. 
 
 In cleaning the fires, the non-combustible is usually found in the 
 shape of a vitrified clinker deposited upon the two sides of the fur- 
 nace, and is removed through the ordinary feed-doors. By the use 
 of a slice-bar this clinker can be raised from the grate and afterwards 
 pulled out with a hook. The central portion of the fire is not dis- 
 turbed. 
 
 The Jones Under-Feed Mechanical Stoker is shown in Fig. 
 263. It consists of a hopper to receive the coal ; underneath is a cylin- 
 
 FiG. 263. 
 
 der, in which is fitted a plunger or ram. A steam cylinder immediately 
 adjoining is furnished with a piston, the rod of which, passing through 
 a stufiing-box, attaches and gives motion to the plunger or ram under 
 
BOILER FURNACES AND SETTINGS 
 
 229 
 
 the coal-hopper. A slide-valve, operated by hand, admits steam on 
 either side of the piston for the forcing of coal into the furnace or the 
 withdrawal of the plunger for a fresh charge. The central trough, or 
 retort, shown in Fig. 
 
 264, into which the Fig. 264. 
 
 coal is forced, is in- 
 clined upward at the 
 angle which gives a 
 nearly uniform thick- 
 ness or spread of coal 
 throughout the length 
 of the furnace. On 
 either side of this retort 
 are tuyere pipes or 
 openings, shown in the 
 cross-sectional engrav- 
 ing, Fig. 264. Ordi- 
 nary grate - bars on 
 either side of the re- 
 tort fill out the width 
 of the furnace, the 
 length and width of 
 
 the furnace being suited to the size of the boiler and rate of combus- 
 tion. 
 
 Additional to the stoker, the equipment comprises a blower and 
 engine, steam and air piping, and usually a new grate-bearing bar. A 
 blower must be used that will hold a 4-ounce pressure at the tuyeres at 
 all times, delivering 150 cubic feet of air for each pound of coal burned, 
 estimating upon the greatest amount of coal to be burned per hour. 
 The engine for furnishing power to maintain the requisite amount and 
 pressure of air should preferably be independent. The blower should 
 not be driven from the main shaft, because air cannot be supplied to the 
 stoker in starting, nor at any time, except the main engine be running. 
 Suppose the furnace is to fired and with no steam in the boiler. The 
 retort is first filled with coal level with the top of the tuyere pipes. 
 Fire is then started on the side grates, as usual, until steam is raised. 
 The ash-pit doors that admit air to the side grates are then closed. 
 Then the coal is placed in the hopper outside of the boiler-front. The 
 steam ram is then withdrawn by shifting the lever. The desired quan- 
 tity of coal then falls from the hopper in front of the ram, and upon its 
 return stroke is forced into the retort. Air under pressure is then 
 admitted into the tuyere pipes. The air issues through the slots shown 
 in Fig. 264, over the top of the fuel in the retort, but under and through 
 the burning fuel. The result is that the heat from the burning fuel over 
 the retort slowly liberates the gas from the green fuel in the retort. 
 
230 
 
 BOILERS AND FURNACES 
 
 The gas, being thoroughly mixed with the incoming air before it passes 
 the burning fuel above, results in a bright, clear fire, free from smoke, 
 and the complete consumption of all the heat-producing elements in 
 the fuel. The retort being practically air tight from below, and the fuel 
 being in a compact mass in the retort, the air will find its way in the 
 direction of least resistance, which is upward ; consequently combustion 
 takes place only above the air-slots. Hence the castings of the retort 
 are always cool and not subject to the action of the fire. The incoming 
 fresh fuel from the retort forces the resulting ash and clinker over the 
 top of the tuyere pipes on the side grates, from whence they may be 
 removed at any time without interfering with the fire in the centre of 
 the furnace. 
 
 The furnace should not be charged with more than loo pounds of 
 coal at a time. It is better to put in small charges at regular intervals, 
 giving the coal ample time to coke before forcing it over the air-slots 
 and into the fire. 
 
 Hawley Do\A/'n-Draft Furnace. — This furnace consists essentially 
 of two water-drums, one located immediately inside the boiler -front and 
 
 Fig. 265. 
 
 another back the proper depth of the furnace. These two water-drums 
 are connected with each other by wrought-iron pipes, forming a water- 
 grate, as shown in Fig. 265. The forward drum is in water communi- 
 cation with the bottom of the boiler through a wrought-iron pipe con- 
 
BOILER FURNACES AND SETTINGS 
 
 231 
 
 nection. The rear drum is similarly connected, but the pipe connection 
 
 to the boiler has its opening higher up, see Fig. 266. This arrangement 
 
 of pipes, drums, and 
 
 water-grates is conducive Fig. 266. 
 
 to water circulation and 
 
 adds to the efficiency of 
 
 the boiler. 
 
 The wrought-iron 
 pipes forming the water- 
 grates are sometimes ar- 
 ranged in a zigzag row 
 in the drums and some- 
 times in a single row. It 
 matters little which ar- 
 rangement is employed, as 
 the results are practically 
 the same for the same 
 amount of tube heating 
 surface thus employed. 
 
 The front water-drum 
 is usually 8 or 10 inches 
 in diameter, and must be 
 provided with a handhole 
 for examination and clean- 
 ing. The rear drum is 
 
 made from 10 inches diameter, the Chicago practice, to 20 inches diam- 
 eter, corresponding to St. Louis practice. For a drum as large as 20 
 inches a manhole can be inserted in the head, and this has been found 
 useful in examining the condition of the water-grate pipes connecting 
 the two drums, as by attaching a candle to a stick and passing it across 
 the intersection of the openings in the small drum, a man in the large 
 drum can ascertain the exact condition of each tube. A constructional 
 
 Fig. 267. 
 
 advantage which the large drum has over the small one is that when a 
 zigzag row of tubes is desired no flattening of the drum is necessary 
 to get the proper amount of thread for screwing in the water-grate 
 pipes, which is necessary in the case of small drums. See Fig. 267. 
 
232 BOILERS AND FURNACES 
 
 The pipes forming the water-grates are commonly 2 inches in diam- 
 eter. In the earlier forms they were set level, but it was found that by 
 .placing them on an incHne of 2^ to 3 inches per foot an improved 
 circulation was had, and the probability of burning off the tubes was 
 greatly reduced. 
 
 The lower-grate surface is formed of common grate-bars, placed 18 
 inches below the lowest line of the upper grates ; it is set level, or 
 nearly so, one end resting on the boiler-front, the rear end on the 
 bridge-wall. 
 
 The ash-pit is located under the lower-grate bars, and quite fre- 
 quently at a level slightly below the fire-room floor. This arrangement 
 is made necessary by the use of a fire-front three doors high, as the 
 firing is at the upper door only. This necessarily raised the fire-door 
 some eighteen inches higher than that to which firemen are accustomed, 
 and to save the fireman the extra labor thus occasioned, the ash-pit 
 floor was placed a few inches below the fire room level, as shown in the 
 engraving. This depression in front of the boiler is about three feet 
 wide, which does not inconvenience the fireman as much as lifting the 
 coal to an unaccustomed height. The ash-pit slopes downward towards 
 the bridge-wall, which facilitates the removal of ashes. 
 
 The fire-front is made three doors high. The upper one receives all 
 the fresh fuel, the middle one opens into the lower grate, the lower one 
 into the ash-pit. The middle door should always be kept closed, 
 except when it is necessary to clean the lower fire, — perhaps three or 
 four times a day. 
 
 The furnace combustion proceeds in two stages : first, the raw fuel 
 is fed into the upper furnace, the air supply coming in through the 
 upper-furnace door directly over the fresh charge of fuel. The direc- 
 tion of the current of gases is directly opposite that of an ordinary 
 furnace, being downward instead of upward. The gaseous products 
 of the upper furnace pass down into the lower furnace. These gases 
 are combustible because the combustion in the upper furnace is incom- 
 plete. It is, in fact, a chamber for the distillation of gas from coal, 
 rather than a furnace in the true sense of the word. The water-grates 
 being set wide apart, the coked fuel of the upper furnace falls through 
 the upper grates into the lower furnace, and is the source from which it 
 receives its fuel supply. This brings us to the second stage, in which 
 the heat from the incandescent body of coke burning on the lower 
 grate raises the incoming flow of cooler gases from the upper furnace 
 to that temperature necessary for their ignition, there being enough 
 oxygen supplied by the excess of air mixed with the gases to complete 
 the combustion begun in the upper furnace. 
 
 No raw fuel should be used in the lower furnace ; if the upper fur- 
 nace is properly managed it will supply all the coke necessary for the 
 lower one. As about 90 per cent, of the entire work is done by the 
 
BOILER FURNACES AND SETTINGS 233 
 
 upper grate, the lower grates do not need a large fuel supply ; by slicing 
 the upper fire enough half-burnt fuel will drop down in the lower furnace 
 to supply its needs. Additional air supply for the lower furnace should 
 be had through the ash-pit. An even fire 6 to 8 inches deep should be 
 kept on the upper grates, the thickness depending on the size of the 
 coal (which should be about the size of one's fist) and the intensity of 
 the draught, a fine coal with moderate or poor draught requiring the 
 thinnest fire. 
 
 The chimney draught must be ample to maintain a quick fire, other- 
 wise the action of the furnace will be sluggish, producing smoke, and 
 of less efficiency generally. 
 
 The standard form of construction is similar to that shown in the 
 engraving, in which the furnace is under the front end of the boiler. 
 Some 50 to 60 square feet of shell heating surface is thus made non- 
 effective for evaporation, which has suggested the construction of an 
 external furnace with its attendant advantages in capacity and efficiency. 
 
 Evaporative tests made by W. H. Bryan indicate that this furnace 
 adds to the efficiency of improved water-tube forms of boilers, although 
 the percentage of increase is not so great as with ordinary horizontal 
 tubular boilers. The claims for this furnace are that it prevents at least 
 95 per cent, of smoke, saves 10 to 40 per cent, of fuel, and increases 
 boiler capacity 25 to 50 per cent. 
 
 Reynolds Furnace. — A horizontal return tubular boiler with a 
 furnace externally fired is shown in Fig. 268. It is well known that a 
 much higher furnace temperature can be had if its walls are made of 
 refractory and non-absorbing material than if the top or sides be made 
 of an absorbing body like the shell or fire-box of a steam boiler. Com- 
 bustion is more easily perfected in a furnace at a high temperature than 
 at a low one. Smoke abounds most in furnaces of low temperature. 
 This furnace has for its object better combustion and a reduction of 
 smoke where the fuel is bituminous coal ; this end is sought to be at- 
 tained by the highest temperature which can be had by the combustion 
 of bituminous coal in a separate chamber. 
 
 This furnace as applied to a steam boiler, shown in Figs. 268 and 
 269, consists of three parts, — the furnace proper where the coal is burnt 
 upon the grates, a combustion- chamber at the rear of the furnace, and a 
 diffusion-chamber underneath the boiler. The products of combustion 
 on leaving the external brick furnace pass through a contracted opening 
 or throat, where they mingle with a supply of air which is admitted 
 through the brick walls. The entrance of this air not only enhances 
 the combustion, but carries back into the furnace some of the heat 
 which would otherwise escape through radiation from the brickwork, 
 and it thus serves a twofold purpose. The products of combustion then 
 come in contact with an overhanging brick arch, by means of which 
 their direction of motion is changed and they emerge into a combustion- 
 
 16 
 
234 
 
 BOILERS AND FURNACES 
 
 chamber, from which they escape through a checker-work arch into the 
 space beneath the boiler-shell. 
 
 Experiments by W. H, Bryan with this furnace and setting applied 
 to a 66-inch x 1 8-foot horizontal tubular boiler with 56 4-inch tubes, a 
 total heating surface of 1276 square feet, the grate surface 25 square feet, 
 a ratio of heating to grate surface of 51 to i, resulted as follows : The 
 coal used was Cherokee slack, having a calorific value of 11,335 thermal 
 units per pound, with 15 per cent, of ash, burning 30.6 pounds per 
 square foot of grate per hour. The equivalent evaporation per pound 
 
 Fig. 268. 
 
 V/// ////// /// ///////// M. 
 
 of combustible from and at 212° Fahr. was 7.83 pounds ; the tempera- 
 ture of the escaping gases was 671° Fahr. ; the efficiency or percentage 
 of total calorific power utilized was 56.6 per cent. This is not a bad 
 showing considering the kind of fuel used and the high temperature of 
 the escaping gases. 
 
 As proof of the excellent working of the furnace, the smoke escaping 
 from the chimney in a daily run varied, according to the St. Louis stand- 
 ard, from 0.6 of I per cent, to 2.8 per cent., according to the amount 
 of work done. 
 
BOILER FURNACES AND SETTINGS 235 
 
 Admission of Air over the Fire. — It has long been thought 
 essential to the proper and economical combustion of bituminous coal 
 that air should be admitted over the fire, and within certain limitations 
 this is true and is recommended ; but an excess of air is wasteful, be- 
 cause it lowers the temperature of the furnace at a time when the air is 
 not needed to assist in the work of combustion. An opinion based upon 
 practical results has been gaining, that if an excess of air is to be per- 
 mitted in the furnace at all, it is better for it to pass up through the fire 
 than to be admitted over it ; and the reason is quite obvious, for whatever 
 oxygen is needed for the combustion of the incandescent fuel on the 
 grate is taken up by this fuel during the passage of air through it, and 
 any excess of air simply passes through the fuel, its temperature being- 
 raised to that of the other products of combustion. Should this heated 
 air come in contact with combustible gases not yet combined, the con- 
 ditions for ignition and combustion are more favorable than if the air 
 were admitted over the fire instead of through it. It is for this reason 
 principally that the admission of air through the sides of a furnace, 
 through the bridge-wall, or in the combustion-chamber back of the 
 bridge-wall has so often failed of its purpose, such currents of air acting 
 as a cooling medium rather than uniting with the combustible gases 
 under the conditions necessary for making the combustion complete. 
 
 Distance between Under Side of Boiler and Top of Grate. — 
 Neglecting all other fuels than anthracite and bituminous coals, it may 
 be stated generally that anthracite coal requires the least distance, and 
 for horizontal tubular or flue boilers this approximates 24 inches when 
 the larger sizes are used for steam purposes, such as Nos. i or 2 chest- 
 nut ; but coal of this size is now seldom used in steam-boiler furnaces on 
 account of its cost ; pea, buckwheat, and rice sizes, being much cheaper, 
 are used instead ; such coals may have the grates placed within 20 inches 
 of the boiler-shell. For bituminous coal the grates may be say 30 
 inches for non-caking coals like Indiana block and kindred varieties, to 
 36 or 48 inches for fatty or gaseous coals ; 36 inches is a good distance 
 for average bituminous coals ; this gives ordinarily all the cubic space 
 needed for a thorough mixing of the gases, and such furnaces, if not 
 forced too hard, are commonly smokeless. This latter quality ought 
 not to be ignored, even though no direct saving in fuel occurs when the 
 distance exceeds 36 inches. Anthracite and bituminous coals cannot be 
 economically burnt in the same furnace, therefore furnaces should be 
 adapted for either one or the other variety of coal. 
 
 The height of the level of the grate surface above the fire-room floor 
 is commonly 24 inches, occasionally 26 inches ; this latter distance is 
 rarely exceeded. 
 
 Ash-Pit. — The vertical depth of the ash-pit is fixed by the height ol 
 the grate surface above the fire-room floor ; this distance will, therefore, 
 approximate 20 inches clear space if the ash-pit floor be level, which is 
 
236 BOILERS AND FURNACES 
 
 ordinarily the case. Sometimes the ash-pit floor indines to the rear, 
 making it deeper at the bridge-wall than at the furnace front. This 
 makes an easier angle for the removal of ashes, especially if the grate- 
 bars are 6 feet or so in length. The bottom of the ash-pit should have 
 a tight brick or cement floor. 
 
 Bridge- Wall. — This wall must be strong enough to take the thrust 
 of the implements used in cleaning the fire, and must be thick enough 
 that the joints do not loosen by this action. The top of the bridge- 
 wall should be not less than 13 inches, or i^ bricks, in thickness, and 
 never less than 18 inches thick where it receives the weight and thrust 
 of the grate-bars. The bridge-wall should be faced next the fire and 
 capped with fire-bricks laid in fire-clay. The top of the bridge-wall is 
 sometimes curved to follow that of the boiler, so as to present an equal 
 area for the passage of the gases. This is altogether a matter of fancy, 
 and no additional valuable results are had over carrying the bridge- 
 wall straight across the furnace. This latter has long been, and is now, 
 the common practice in building bridge-walls. It is customary to make 
 a sloping surface from the top of the grates to the top of the bridge- 
 wall, as shown in Fig. 293. So far as affecting combustion, it makes 
 no difference whether this wall is straight, or inclined. Hollow spaces 
 are frequently provided in bridge-walls by which air is introduced into 
 the furnace for the better combustion of bituminous coal. Whilst this 
 device has in many cases been productive of good results, it is not 
 in itself sufficient to make good the loss occasioned by having the 
 furnace itself too small. 
 
 Fire-Brick Lining. — Steam-boiler furnaces should be Hned with 
 fire-brick at least as far back as the rear end of the bridge-wall. This 
 lining should be capable of repair or complete removal without disturb- 
 ing the side walls. In making provision for relining a furnace with fire- 
 brick, one or two courses of fire-brick "headers" should be used in 
 finishing the top courses of the lining, as shown in Fig. 270. This will 
 permit the removal of the lower without disturbing the upper courses 
 in the wall. One thickness, 4 or 4^ inches, depending on the width 
 of the brick, will suffice ; these bricks should be set in fire-clay mortar. 
 The fire-brick lining should begin at the bottom of the grate-bars, and 
 should include the stepped courses where the top of the furnace is closed 
 in upon the side of the boiler, as in Fig. 270, which represents a wall 
 extending vertically downward from that distance allowed at the top of 
 the furnace for the hot gases to come in contact with the shell of the 
 boiler near the water-line. When the furnace is narrowed to the width 
 of the boiler, or sometimes less, the fire-brick should not be trimmed to 
 fit the angle of the furnace sides, but should be used their full width 
 and backed up by the red brick walls, as in Fig. 271. 
 
 The fire-brick lining between two boilers when set singly is carried 
 out in the same manner as that just described for the side walls. 
 
BOILER FURNACES AND SETTINGS 
 Fig. 270. Fig. 271. 
 
 237 
 
 The distance from the side of the boiler at its centre Une to the side 
 wall of furnace, as at A, Fig. 272, varies in practice from 3 inches for 
 36-inch boilers to 6 inches for 72-inch boilers. This latter distance is 
 
 Fig. 272. 
 
 greater than is necessary ; it need not, for all practical purposes, be more 
 than 4 inches. The under side of the top line of fire-bricks where they 
 join the boiler, as at B, should not be much higher than the top of the 
 
238 BOILERS AND FURNACES 
 
 upper line of tubes ; if this distance were extended up to the water-level 
 no harm could occur, but if above the water-level there would be danger 
 of burning the boiler along that line when forcing the fire before steam 
 was raised in the boiler. 
 
 Thickness of Walls. — For single walls the thickness should not 
 be less than 2 bricks, varying from 17 to 18 inches, depending upon 
 the locality, and this must be exclusive of any fire-brick lining subject 
 to renewal. The thickness of walls between two boilers ought not in 
 any case to be less than i^ bricks, say 13 inches, to which is to be 
 added the two fire-brick finings of 4 or 4)^ inches each, adding 8 or 9 
 inches more to the thickness, as shown in Fig. 272. 
 
 Double Walls. — A considerable economy can be effected by build- 
 ing the furnace-walls double, with an air-space between the outer and 
 inner walls, as shown in Fig. 270. Walls thus constructed prevent 
 radiation of heat and are, therefore, recommended. When constructing 
 hollow walls headers should extend occasionally, say every 2 feet or so, 
 to give support to the walls and prevent the air-space closing up. These 
 headers should merely touch, and must in no case enter or be fastened 
 to the opposite wall. 
 
 The width of the air-space may be 3 inches for boilers up to 46 
 inches, and 4 inches for boilers 48 inches and larger in diameter. 
 
 When two or more boilers are set singly in a battery the thickness 
 of the division walls should be such that an air-space is had of the same 
 width as that of the outer walls. 
 
 The walls when made hollow are commonly thinner than when they 
 are single. Any cracking of the interior wall is not likely to cause 
 much leakage into the furnace, because the outer wall may still be 
 intact ; if not, the cracks are in plain sight and can be easily remedied. 
 If the walls carry straight down, as in Fig. 270, each wall and the air- 
 space will be parallel from top to bottom, but if the grate is diminished 
 to the diameter of the boiler, as in Fig. 273, the air-space becomes less 
 in width at the top ; this applies also in the case of the dividing walls, 
 as shown in the same engraving. The rear wall of the boiler setting 
 should have an air-space as well as the side walls. See Fig. 207. 
 
 Air-spaces in furnace-walls have been objected to by some engineers, 
 who never include them in furnace designs, the objection being that two 
 thin walls are weaker than a single wall equalling the combined thick- 
 ness of the two, the thin wall is more liable to crack than a thicker one, 
 and that the saving in radiant heat is not worth the extra cost of the 
 brickwork ; but these objections are not generally held by engineers : 
 when single walls are used, it is usually because the construction costs 
 less money. 
 
 Brickwork. — Nearly all furnaces for steam boilers are constructed 
 of red bricks. These should be quite hard, so as to produce a ringing 
 sound when struck ; they must be flat, square to each other when laid 
 
BOILER FURNACES AND SETTINGS 
 
 239 
 
 together, and of uniform size and color. The sizes vary for different 
 locaHties, but will average not far from 8^ x 4JE^ x 2j{ inches each. 
 The harder the brick the less water it will absorb and the greater will 
 be its ability for carrying a heavy load. The safe working load on a 
 fair quality of hard-burnt red brickwork well bedded in domestic cement 
 mortar is approximately 10 tons per square foot ; but furnace-walls are 
 not usually thus laid, the mortar being ordinarily of lime and sand, with 
 very little cement, if any be used at all. For light-colored red bricks 
 in common mortar the safe working load would not be more than hall 
 the above. Buff or salmon bricks when resulting from deficient burning 
 should not be used. 
 
 Fig. 273. 
 
 Mortar Joints. — The thickness of mortar joints will approximate 
 }( inch ; that is, the average height of four courses of common brick 
 work is 10 inches. 
 
 The mortar used for boiler settings has commonly been of lime only, 
 but a stronger wall is had if a small proportion of cement be added to 
 it. A very good mortar is made by mixing three parts of good lime 
 mortar with one part of hydraulic cement mortar. The cement may be 
 of domestic manufacture. The sand should be clean and sharp. 
 
 Fire-bricks should be laid only in fire-clay. Arch bricks should be 
 carefully fitted dry, and when the arch is completed the bricks should 
 
240 
 
 BOILERS AND FURNACES 
 
 Fig. 274. 
 
 then be set with a thin fire-clay paste, fi-om the same clay as that of the 
 bricks if practicable. The sizes of fire-bricks vary according to locality, 
 but they are commonly larger in every way than red bricks. 
 
 Buck- Staves. — The cracking of any brick wall, hot on one side 
 and cold on the other, cannot be entirely prevented. Thick walls offer 
 a greater resistance to cracking, but it 
 is only a question of time when the 
 continued action of the furnace-heat will 
 cause any furnace-wall to crack, regard- 
 less of its thickness. When furnace- 
 walls begin to open, the joints should 
 be filled with a thin fire-clay grout, 
 poured into the opening, which will pre- 
 vent the admission of cold air through 
 the walls into the furnace. A buck- 
 stave is shown in Fig. 274. Stay-rods 
 extend through from side to side of the 
 furnace at top and bottom, as shown. 
 Buck-staves are commonly made of cast 
 iron. Their use is to prevent the spread- 
 ing of the furnace-walls ; to best secure 
 this end, the web should be of con- 
 siderable depth, from 4 to 6 inches, 
 depending on the diameter of the boiler. 
 These are usually placed 4 to 5 feet 
 apart on the side walls. 
 Combustion-Chamber.^-The space back of the bridge-wall is 
 popularly known as a combustion-chamber, but it is doubtful if much, 
 or if any, combustion takes place after the gases pass the bridge-wall, or 
 beyond the immediate vicinity of the bridge-wall, in case hot air is ad- 
 mitted at that point. This space is thought by some to be useful as 
 a reservoir for the heated gases, breaking up the rapid current which 
 would otherwise flow in a line parallel to the bottom of the boiler and 
 thence through the tubes to the chimney, not giving out as much heat as 
 would be the case if this volume of hot gases could be interrupted in its 
 passage and proceed with a slower movement, but this argument is far- 
 fetched and probably not true. Boiler- tests made with a combustion- 
 chamber, as shown in Fig. 275, have given practically the same results as 
 when filled in, as shown in Fig. 217, showing that it is immaterial whether 
 a combustion- chamber is provided or not. The flow of furnace gases 
 over the top of the bridge-wall will not work downward to fill this large 
 space with heated gases at the same temperature as the furnace, but will 
 follow along the boiler-shell to the rear end, where they turn upward and 
 flow through the tubes to the chimney. As no combustion takes place 
 in this chamber, and the temperature is much higher than that of the 
 
BOILER FURNACES AND SETTINGS 
 
 241 
 
 atmosphere, the side 
 walls must of necessity- 
 radiate heat. For this 
 reason it has become 
 a common practice to 
 fill in the combustion- 
 chamber with earth, as 
 shown in Fig. 295, 
 leaving room enough 
 at the rear end for a 
 cleaning door for taking 
 out the ashes which ac- 
 cumulate there. 
 
 Back Connec- 
 tions. — These are con- 
 structed usually in two 
 ways, the practice being 
 quite evenly divided 
 between the plate, as 
 shown in Fig. 277, and 
 the arched connections, 
 as shown in Fig. 278. 
 The arch consists of a 
 cast-iron skeleton filled 
 in with fire-brick and 
 afterwards with earth to 
 make it air-tight. The 
 back plate affords easy 
 access to the end of the 
 boiler and plenty of 
 light at the same time. 
 Either of the two are 
 good, but the cover- 
 ing of the joints with 
 earth should not be neg- 
 lected, or cold draughts 
 of air at that point will 
 lower the temperature 
 of the gases passing into 
 the tubes. 
 
 The arched connec- 
 tion at the rear of the 
 boiler is made up of cast- 
 iron segments, as shown 
 in Fig, 279. These rest 
 
242 
 
 BOILERS AND FURNACES 
 Ftg. 277. Fig. 278. 
 
 upon the rear wall and against the boiler, an angle-iron about 2x2 inches 
 being attached to the boiler-head to form a suitable support. Another 
 form of arched connection is shown in Fig. 280, in which an arch springs 
 from the two side-walls ; this is also shown in Fig. 299. 
 
 Fig. 279. 
 
 
 Boiler Covering. — Any non-conducting covering will answer, pro- 
 vided it contains nothing that will act injuriously on the shell of the 
 boiler in case of a steam leak. Boilers ought not to be covered until 
 
 Fig. 280. 
 
 they have steamed sufficiently to be sure that no leaks occur in the 
 joints. A good covering consists in laying narrow strips, say 2 or 
 3 inches wide, of pine boards i inch thick over the top of the boiler 
 
BOILER FURNACES AND SETTINGS 
 
 243 
 
 from the brick ledges on either side, and then covering all with a 4-inch 
 brick arch resting on these narrow strips, as shown in Fig. 281. The 
 only objection to this covering is, that in case of a leak almost the whole 
 top must come off to get at a joint. Another method is, to cover the 
 whole surface of the top of the boiler with asbestos board about ^-inch 
 thick, and on top of this narrow strips of pine, say 2 inches wide and i 
 inch thick, and covering again with clay to a depth of 4 to 6 inches. 
 
 This clay can be easily shovelled off when not wanted, and the pine- 
 strips removed, the asbestos board preventing the loose clay coming in 
 contact with the boiler. Asbestos board with additional hair-felt clothing 
 makes a good covering. 
 
 Carrying the products of combustion over the top of a boiler, as 
 shown in Fig. 275, is quite generally practised in some localities. This, 
 of course, prevents any radiation from the boiler, because the tempera- 
 ture of the escaping gases is always higher than that of the steam. 
 
 A jacket of hot gas over the top of the boiler, as shown in Fig. 282, 
 is frequently employed, and affords complete protection against heat 
 radiation from the boiler. 
 
 Both of these styles of boiler settings have been seriously opposed 
 because of the supposed liability of the hot gases passing over the top 
 of a boiler-shell to cause overheating, especially in the interval between 
 starting a fire under a cold boiler and filling the steam-space with steam. 
 Referring to Fig. 275, the products of combustion from the grate pass 
 along the under side of the boiler to the rear, then return through the 
 tubes to the front, and thence back again, along and over the top sheets 
 of the boiler-shell to the rear, where they are finally conducted to the 
 chimney. The overheating of the top sheets, so far as the writer is 
 aware, has never occurred, nor ought it to be expected in any properly 
 designed boiler setting not using a powerful fan-blast. No injury could 
 accrue to the shell at a temperature below red heat (900° Fahr.), and it 
 is scarcely possible that any such temperature, or more than the half ot 
 it, ever reaches the top portion of the shell of a boiler not under steam. 
 In any boiler setting in which there is a ratio of 30 square feet of heat- 
 
244 
 
 BOILERS AND FURNACES 
 
 ing to I square foot of grate surface, employing a natural draft, there is 
 probably no danger whatever that the top of the sheets will ever become 
 overheated. 
 
 Fig. 282. 
 
 Smoke-Connections. — Flue boilers are almost invariably and 
 tubular boilers quite frequently made with flush ends, without smoke- 
 box extension. The smoke-box 
 Fig. 283. jg therefore made to bolt directly 
 
 to the end of the boiler, and for 
 a single boiler is similar to that 
 shown in Fig. 283, in which case 
 an iron chimney may start directly 
 from the smoke-box, or a pipe 
 may convey the products of com- 
 bustion into a chimney alongside. 
 A design for a smoke-con- 
 nection for two or more boilers is 
 shown in Fig. 284, Each boiler 
 should be provided with its own 
 damper when they are set sepa- 
 rately. These smoke-connections 
 empty into a common breeching, from which a central chimney of 
 wrought iron is intended to be fitted, but if a brick or other chimney 
 be located at the side of the boiler, a combined smoke-box and breech- 
 ing, as shown in Fig. 285, is commonly fitted. 
 
 Half- Arch Front. — For tubular boilers the very general practice is 
 to extend the front sheet sufficiently to make a smoke-box by simply 
 
 'I I I 
 
 I 
 I 
 
BOILER FURNACES AND SETTINGS 
 Fig. 2S 
 
 245 
 
 Fig. 285. 
 
246 
 
 BOILERS AND FURNACES 
 
 closing the end with a cast-iron front with a hinged door, as shown in 
 Fig. 286. In this case, whatever distance the smoke-box projects be- 
 yond, the head extends into the fire-room, and is of more or less annoy- 
 ance to the fireman. The extension of the shell makes a stronger base 
 
 Fig. 286. 
 
 for the wrought-iron chimney to rest upon than when this smoke-box is 
 bolted to the front head ; when a brick chimney is employed a wrought- 
 iron pipe connects it with the smoke-box. 
 
 Full Square Front. — Another method is to bring the whole front 
 out far enough to include the depth of the smoke-box between the 
 outer line of the fire-front and the inner face of the fire-brick lining. 
 Such a construction is shown in Fig. 287. One objection to this method 
 of boiler-setting is that the fire-brick lining of the front must equal the 
 depth of the smoke-box, and this necessitates fire-door liners from 13 
 to 18 inches deep, making it somewhat more difficult to attend and 
 clean the fires than is the case with the ordinary depth of linings. A 
 cast-iron plate. Fig. 288, serves as a base for a wrought-iron chimney 
 or to receive a breeching-pipe connecting with a brick chimney at one 
 side. This plate is anchored in the brickwork immediately back of the 
 fire-front. This style of front presents a neat appearance and usually 
 receives more or less of decoration, according to the fancy of the designer. 
 
 Nay lor' s fire-front, shown in Fig. 289, overcomes the objection to 
 the deep fire-brick lining, referred to above, by recessing the cast-iron 
 front, bringing the fire-door to within the narrow thickness of a half- 
 arch front, and filling in the space between the arch over the fire-door 
 
BOILER FURNACES AND SETTINGS 
 
 247 
 
 and the bottom of the boiler with fire-brick, — a detail clearly shown in 
 the two sectional drawings which accompany the front elevation. 
 
 Fire-Doors. — For boilers of the locomotive type and internally 
 fired boilers generally these are commonly fitted to cast-iron frames 
 bolted to the shell of the boiler. Such a door is shown in Fig. 290. 
 
 Fig. 2S7. 
 
 The door and frame should have planed surfaces to make a tight joint ; 
 the hinges should be of unusual strength to withstand the rough usage 
 of a fire-room. The perforated lining shown may be of cast iron ; the 
 perforations are simply round holes about ^ inch in diameter, and may 
 be drilled in the pattern, so that the rough casting will have all the holes 
 included in it. A bolt passing through the lining, distance-piece, and 
 the door make a simple and substantial fastening. A butterfly register 
 opening is usually included in the door-casting. 
 
 Fig. 288. 
 
 Butman's fire-door is shown in Fig. 291. The frame of the door is 
 bolted against the fire-front, or against the boiler, as the case may be. 
 Unlike the ordinary fire-door, this door is hinged to open upward, the 
 counter -weight being slightly more than sufficient to overbalance the 
 door and raise it when the latch underneath the door is carried down 
 a sufficient distance to release it. The weight has a segment of a gear 
 
248 
 
 BOILERS AND FURNACES 
 Fig. 289. 
 
 o "oroj i 
 
 On 0X0 ^ 
 
 P lo"oi l 
 
 [ M o^ob 
 
 ; 
 
 Fig. 290. 
 
BOILER FURNACES AND SETTINGS 
 
 249 
 
 cast inside, as shown in the engravings, into which a similar toothed 
 segment, fitted to the door, is geared, and thus the movements of the 
 door and weight are controlled by each other. On the same central 
 shaft, around which the weight oscillates, is also secured a deflecting- 
 plate. When the door is closed, as in Fig. 291, the deflecting-plate is 
 wholly within the housing. A butterfly register is attached to the door, 
 through which a greater or less quantity of air may be admitted. When 
 the deflecting-plate is down, as shown in Fig. 291, the air passes under- 
 neath its lower edges, and is thus brought into close surface contact 
 with the burning fuel. When the furnace door is opened to supply 
 fresh fuel the deflecting-plate is thrown out horizontally, as shown in 
 
 Fig. 291. 
 
 Fig. 292 
 
 Fig. 292. The object of this deflecting-plate is to prevent the cold air 
 from impinging directly against the bottom of the boiler, but to so direct 
 its course that it shall mingle with the heated gases immediately over 
 the fire. 
 
 Examples of Furnace Construction. — In Chapter VI., Fig. 
 209, is given an illustration of the ordinary method of setting cylinder 
 boilers ; Fig. 217, that of flue boilers, except that the earth filUng is not 
 always included ; indeed, for flue boilers this filling is the exception 
 rather than the rule. 
 
 Horizontal tubular boilers are set with a considerable variety of 
 detail. Taking the country at large, the ordinary setting shown at 
 Fig. 293 is, perhaps, more largely employed than any other ; it repre- 
 sents a boiler suspended on side wings and with a half front, as shown 
 in Fig. 286. The distance from the underside of the boiler to the top 
 of the grates may be 24 inches for anthracite coal, 26 to 30 inches for 
 semi-bituminous coals, and 36 to 48 inches for rich bituminous coals. 
 
 17 
 
250 
 
 BOILERS AND FURNACES 
 Fig. 293. 
 
 Fig. 294. 
 
BOILER FURNACES AND SETTINGS 
 
 251 
 
 The distance from boiler to top of bridge-wall may be 16 inches for all 
 furnaces, regardless of diameter of boiler or kind of fuel used. The 
 distance from bottom of boiler to top of rear division wall may be 12 
 inches for all sizes of boilers. It has long been the practice to curve 
 the top of this wall, as shown in the curved line in Fig. 273, but no 
 special advantage is had by so doing. 
 
 It is becoming the usual practice to fill in the space between the 
 bridge-wall and the rear division wall with earth, as shown in Fig. 294, 
 covering the top of it with a single layer of red brick ; in this case the 
 top of the wall is not curved, but extends straight across the furnace, 
 the same as the bridge-wall. 
 
 A design of furnace for horizontal tubular boilers by the Hartford 
 Steam-Boiler Inspection and Insurance Company is shown in sectional 
 elevation in Fig. 295. In 
 
 this design the rear divi- ^^^- ^96- 
 
 sion wall is omitted ; the 
 space back of the bridge- 
 wall is filled in at the in- 
 chnation shown in the 
 drawing and afterwards 
 covered with brick. A 
 cleaning-door is shown at 
 the rear end of the boiler- 
 wall. A cross - section 
 through the furnace show- 
 ing the width of grate and 
 the details of the fire-brick 
 lining is given in Fig. 
 296. A plan of the boiler 
 showing the air-space be- 
 tween the walls is given 
 in Fig. 297 ; so also the 
 points of suspension of the 
 
 boiler, showing the details of brickwork upon which the side wings rest. 
 A large number of boilers are set in this manner and yield excellent 
 results. 
 
 A furnace design for carrying the products of combustion over the 
 top of the boiler is shown in Fig. 275. In this drawing there is no rear 
 division wall and no filling in back of the bridge-wall. This furnace was 
 designed for burning bituminous coal, and it was thought that this rear 
 combustion-chamber might be advantageous in affording a better ad- 
 mixture of gases and better combustion, — results which were probably 
 not realized. 
 
 A fiirnace designed for two boilers, set singly, by Charles Edgerton, 
 is shown in longitudinal elevation in Fig. 298. In this case the com- 
 
252 
 
 BOILERS AND FURNACES 
 
 bustion-chamber is filled up to the level of the grates. It will be noticed 
 that the bridge-wall is of unusual depth. The products of combustion 
 pass to the rear of the boiler, thence through the tubes to the front, and 
 over the top of the boiler to the chimney. A damper is placed near the 
 exit of the gases for controlling the draft. 
 
 Fig. 297. 
 
 Fig. 299 gives two cross-sectional elevations : one through the fur- 
 nace showing details of the fire-brick lining, as well as a flue over the top 
 of the boiler for conducting the furnace gases to the chimney ; the other 
 half of this engraving illustrates a detail of fire-brick lining to the cast- 
 iron fronts, forming also the smoke-chamber around the sides and at the 
 front end of the boiler. A plan of the flues on top of the boilers is 
 shown in the sectional drawing, Fig. 300. This drawing also shows the 
 location of the chimney, which in this case was made of metal and 
 attached to a cast-iron base-plate resting upon one corner of the boiler 
 setting, as shown. 
 
 The horizontal tubular boiler setting shown in sectional elevation, 
 Fig. 301, is by W. Barnet LeVan. The boiler has a superheating drum 
 attached by a wrought-metal connection, also shown in the illustration. 
 This detail was prepared in recognition of the fact that steam cannot 
 ordinarily be superheated when in contact with the water from which it 
 was generated. The drum is, therefore, isolated and attached to the 
 main shell by a single connection, located at the front end of the boiler. 
 The delivery of steam is from the rear end of the superheating drum. 
 The latter is of sufficient size to allow ample time for the highly heated 
 gases which outwardly surround and also pass through the tubes in the 
 drum to evaporate any entrained water in the steam -drum before it 
 passes ofl" to the engine. 
 
 It will be noted that in the furnace details a bridge-wall of somewhat 
 unusual construction is provided. The combustion-chamber beyond is. 
 
BOILER FURNACES AND SETTINGS 
 
 253 
 
 Fig. 298. 
 
 Fig. 299. 
 
254 
 
 BOILERS AND FURNACES 
 
 Fig. 300. 
 
 Fig. 301. 
 
BOILER FURNACES AND SETTINGS 
 
 555 
 
 In this combustion-chamber is a 
 openings, as shown in the cross- 
 
 FiG. 302. 
 
 on the same level with the grates, 
 rear division wall, with checkered 
 sectional elevation, Fig. 302. 
 This wall becoming highly heated 
 is intended to act as a regenerator 
 and to promote combustion of the 
 uncombined gases on their way to 
 the rear of the boiler. The gases 
 are baffled in their flow, and in 
 passing through these openings 
 are thoroughly mingled and re- 
 ceive an additional supply of 
 heated air conducted into this 
 chamber through register open- 
 ings not shown In the engraving. 
 After the gases pass through this 
 division wall, they are then in the 
 rear chamber ; from thence they 
 pass through the tubes of the 
 boiler to the front end. They 
 are then conducted upward into 
 another chamber, in which is lo- 
 cated the superheating drum. The 
 
 gases not only surround the drum, but pass through the tubes with which 
 this superheating drum is fitted. They then pass through a damper ol 
 somewhat unusual construction, and from thence to the chimney. 
 
CHAPTER VIII. 
 
 INTERNALLY FIRED BOILERS. 
 
 Internally fired boilers are those in which the furnace is included 
 within the structure of the boiler itself The commonest varieties of 
 internally fired boilers are : 
 
 The ordinary vertical tubular boiler, a cylindrical shell in which is 
 enclosed a cylindrical fire-box, with numerous tubes leading directly 
 from the furnace to the upper head of the shell. Sometimes vertical 
 boilers are made with a single flue, especially in localities where the 
 water is bad. 
 
 The Cornish boiler, a cylindrical horizontal shell fitted with a single 
 horizontal flue, one end of which is made the furnace. This boiler is 
 rarely met with in this country. 
 
 The Lancashire boiler, a cylindrical horizontal shell with two hori- 
 zontal flues and furnaces. 
 
 There are a variety of designs for marine and land boilers, consisting 
 in the main of a shell of large diameter in which are one or more large 
 flues fitted with grate-bars ; the furnace being constructed within the 
 flue, the products of combustion are returned from the combustion- 
 chamber at the rear end of the boiler forward to the fire-room, and from 
 thence to the chimney. 
 
 The rectangular enclosed fire-box combined with an outer shell, to 
 which is attached a cyHndrical shell containing numerous horizontal 
 tubes leading directly from the furnace to the smoke-box. It is a type 
 of boiler well known to engineers as the locomotive boiler, though its 
 use is not restricted to such service. Such boilers are constructed in 
 great variety, both as to size and detail. 
 
 Vertical Tubular Boilers. — The ordinary construction of a ver- 
 tical tubular boiler consists of an outer shell containing a cylindrical fire- 
 box and vertical tubes, as shown in Fig. 303. Boilers of this type are 
 made in considerable numbers for small powers in combination with 
 steam engines, pumps, etc. These boilers are ordinarily fitted with 
 tubes 2 inches in diameter, and the largest diameter of tubes even for 
 very large boilers seldom exceeds 2^ inches. The upper end of the 
 tubes are located in the steam-space, and result in giving the steam a 
 slight superheating. 
 
 For ordinary vertical boilers as many tubes are placed in the top of 
 the furnace or crown-sheet as can be conveniently arranged ; the diam- 
 256 
 
INTERNALLY FIRED BOILERS 
 
 257 
 
 Fig, 
 
 eter of furnace is made as large as the outer shell will permit. The 
 aggregate tube area will approximate one-fourth that of the grate area, 
 a higher ratio than obtains in horizontal boiler practice. 
 
 The fire-box affords an excellent evaporating surface, and for that 
 reason it should be as high as possible and maintain a good circulation 
 between it and the outer shell. The circulation will be improved if the 
 fire-box be made slightly conical. The furnace height will be governed 
 somewhat by the fuel to be used : if bitumi- 
 nous coal, the height may be from 30 to 48 
 inches, according to the size of the boiler ; if 
 anthracite coal, the height may be from 24 to 
 36 inches. In general, furnaces for soft coal 
 should be high enough to form an ample com- 
 bustion-chamber and prevent the formation of 
 smoke, the usual accompaniment of low fur- 
 naces and small combustion -chambers. 
 
 The length of tubes in vertical boilers will 
 vary according to the purpose for which they 
 are designed. These boilers are largely used 
 by contractors, who move them around from 
 place to place, according to the necessities of 
 their business. As economy of fuel is not so 
 generally practised in contracting as in a fixed 
 business, an increase in diameter with its larger 
 fire-box is usually preferred to that of height, 
 which gives merely additional length of tube. 
 If, however, it is to be used as a stationary 
 boiler, the height should be as great as is per- 
 missible for the location. Short tubes under 
 heavy firing are apt to burn out at the upper 
 tube-sheet, and this is one argument for in- 
 creasing their length, that more absorbing 
 surface may intervene, with the consequent re- 
 duction in temperature of the gases escaping at the chimney. Notwith- 
 . standing the increased length of the tube, there is danger at all times in 
 overheating the tops of the tubes when raising steam from cold water, 
 and for this reason the fire should not be urged until after steaming has 
 begun. To obviate this fault, boilers are sometimes made with sub- 
 merged tubes, as in Figs. 304 and 305, which show two methods of 
 constructing the upper chamber. The conical one is to be preferred, 
 as giving more steam-room than would be the case with a chamber 
 having vertical sides and a flanged head. The cost of manufacture 
 would also be less, but as no protection is afforded the plate surface, 
 which is thus made to take the place of the tube surface, the change is 
 at best one of doubtful utility. 
 
258 
 
 BOILERS AND FURNACES 
 
 TABLE XLV. 
 
 PRINCIPAL DIMENSIONS OF VERTICAL BOILERS WITH FULL-LENGTH TUBES AS 
 FURNISHED BY THE TRADE, FIG. 303. 
 
 to 
 
 
 Sheli-. 
 
 
 Fttrnatk. 
 
 Flanged 
 
 Tubes 
 
 2 Inches 
 
 
 ^ 
 
 rt 
 
 
 
 
 
 
 
 Heads. 
 
 IN Diameter. 
 
 g 
 
 S 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 i 
 
 v: 
 
 
 i 
 
 0) 
 
 
 
 (J) 
 
 
 
 u 
 
 
 
 
 
 
 
 c 
 
 
 A 
 
 
 
 
 S 
 
 n 
 
 •s 
 
 -i^ 
 
 E 
 
 .S? 
 
 
 ^ 
 
 be 
 g 
 
 •b 
 
 i 
 
 .3 
 
 U 
 
 a 
 
 ffi 
 
 H 
 
 P 
 
 ffi 
 
 H 
 
 E- 
 
 kJ 
 
 ^ 
 
 ffi 
 
 Q 
 
 H.-P. 
 
 Ins. 
 
 Ft. 
 
 In. 
 
 Ins. 
 
 Ins. 
 
 In. 
 
 In. 
 
 Ins. 
 
 
 Sq. Ft. 
 
 Ins. 
 
 4 
 
 24 
 
 4 
 
 3^ 
 
 20 
 
 24 
 
 '4 
 
 
 24 
 
 31 
 
 44 
 
 12 
 
 5 
 
 24 
 
 5 
 
 54 
 
 20 
 
 24 
 
 K 
 
 / 
 
 36 
 
 31 
 
 60 
 
 12 
 
 6 
 
 24 
 
 6 
 
 Ji 
 
 20 
 
 24 
 
 V4 
 
 ^8 
 
 48 
 
 31 
 
 75 
 
 12 
 
 8 
 
 30 
 
 5 
 
 X 
 
 25 
 
 27 
 
 X 
 
 H 
 
 33 
 
 55 
 
 92 
 
 14 
 
 10 
 
 30 
 
 6 
 
 X 
 
 25 
 
 27 
 
 X 
 
 H 
 
 45 
 
 55 
 
 121 
 
 14 
 
 12 
 
 30 
 
 7 
 
 Ji 
 
 25 
 
 27 
 
 ^ 
 
 H 
 
 57 
 
 55 
 
 150 
 
 14 
 
 15 
 
 36 
 
 6K2 
 
 X 
 
 31 
 
 27 
 
 J4 
 
 H 
 
 51 
 
 77 
 
 189 
 
 15 
 
 18 
 
 36 
 
 7 
 
 ^ 
 
 31 
 
 27 
 
 ^ 
 
 H 
 
 57 
 
 77 
 
 210 
 
 15 
 
 20 
 
 36 
 
 8 
 
 X 
 
 31 
 
 27 
 
 X 
 
 
 69 
 
 77 
 
 250 
 
 15 
 
 25 
 
 42 
 
 7% 
 
 /^ 
 
 37 
 
 27 
 
 X 
 
 H 
 
 60 
 
 109 
 
 307 
 
 18 
 
 30 
 
 42 
 
 8X 
 
 A 
 
 37 
 
 27 
 
 X 
 
 Vs 
 
 72 
 
 109 
 
 364 
 
 18 
 
 35 
 
 42 
 
 q^ 
 
 ^^ 
 
 37 
 
 27 
 
 X 
 
 ^8 
 
 84 
 
 109 
 
 422 
 
 18 
 
 40 
 
 48 
 
 8/. 
 
 
 43 
 
 30 
 
 X 
 
 n 
 
 72 
 
 149 
 
 496 
 
 20 
 
 45 
 
 48 
 
 9 
 
 A 
 
 43 
 
 30 
 
 X 
 
 y. 
 
 7« 
 
 149 
 
 535 
 
 20 
 
 50 
 
 48 
 
 10 
 
 A 
 
 43 
 
 30 
 
 X 
 
 y 
 
 90 
 
 149 
 
 613 
 
 20 
 
 60 
 
 54 
 
 9 
 
 T^U 
 
 48 
 
 30 
 
 X 
 
 Y% 
 
 7« 
 
 201 
 
 716 
 
 24 
 
 TABLE XLVL 
 
 PRINCIPAL DIMENSIONS OF VERTICAL BOILERS WITH SUBMERGED TUBES, AS 
 FURNISHED BY THE TRADE, FIG. 304. 
 
 ho 
 
 
 Shell. 
 
 
 Furnace. 
 
 Flanged 
 Heads. 
 
 Tubes 2 Inches 
 IN Diameter. 
 
 1 
 
 a 
 
 
 i 
 
 & 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "3 
 
 
 
 
 
 
 
 
 
 
 U 
 
 
 'r, 
 
 "2 
 
 S 
 
 
 ^ 
 
 !r; 
 
 
 S 
 
 8 
 
 
 u 
 
 'o 
 
 
 fc 
 
 s 
 
 s 
 
 J3 
 
 
 t; 
 
 I 
 
 c 
 
 
 5 
 '^ 
 
 
 •a 
 
 .s 
 
 
 
 "S 
 
 
 
 
 2 
 
 2 
 
 s 
 
 3 
 
 ■ s 
 
 s 
 
 .3 
 
 u 
 
 
 
 ffi 
 
 H 
 
 5 
 
 ffi 
 
 H 
 
 H 
 
 J 
 
 ^ 
 
 ffi 
 
 K 
 
 U 
 
 H.-P. 
 
 Ins. 
 
 Ft. 
 
 In. 
 
 Ins. 
 
 Ins. 
 
 In. 
 
 In. 
 
 Ins. 
 
 
 
 Sq. Ft. 
 
 In. 
 
 4 
 
 24 
 
 5J^ 
 
 X 
 
 20 
 
 24 
 
 X 
 
 Vz 
 
 24 
 
 31 
 
 18 
 
 44 
 
 12 
 
 5 
 
 24 
 
 6 
 
 J4 
 
 20 
 
 24 
 
 '4 
 
 y% 
 
 30 
 
 31 
 
 18 
 
 52 
 
 12 
 
 6 
 
 24 
 
 6J^ 
 
 X 
 
 20 
 
 24 
 
 % 
 
 n 
 
 36 
 
 31 
 
 18 
 
 60 
 
 12 
 
 8 
 
 30 
 
 6 
 
 X 
 
 25 
 
 27 
 
 X 
 
 y 
 
 27 
 
 55 
 
 18 
 
 83 
 
 14 
 
 10 
 
 30 
 
 6J/3 
 
 X 
 
 25 
 
 27 
 
 X 
 
 y% 
 
 33 
 
 55 
 
 18 
 
 98 
 
 14 
 
 12 
 
 36 
 
 6^2 
 
 X 
 
 31 
 
 27 
 
 X 
 
 /8 
 
 33 
 
 77 
 
 18 
 
 133 
 
 15 
 
 15 
 
 36 
 
 7 
 
 X 
 
 31 
 
 27 
 
 K 
 
 ^ 
 
 39 
 
 77 
 
 18 
 
 
 15 
 
 18 
 
 36 
 
 8 
 
 X 
 
 31 
 
 27 
 
 X 
 
 /8 
 
 51 
 
 77 
 
 18 
 
 196 
 
 15 
 
 20 
 
 42 
 
 7J^. 
 
 «H. 
 
 37 
 
 27 
 
 X 
 
 
 39 
 
 109 
 
 24 
 
 215 
 
 18 
 
 25 
 
 42 
 
 8 
 
 /^ 
 
 37 
 
 27 
 
 X 
 
 >^ 
 
 45 
 
 109 
 
 24 
 
 244 
 
 18 
 
 30 
 
 42 
 
 9 
 
 A 
 
 37 
 
 27 
 
 X 
 
 ^8 
 
 57 
 
 109 
 
 24 
 
 301 
 
 18 
 
 35 
 
 48 
 
 9 , 
 
 VIT 
 
 43 
 
 30 
 
 X 
 
 y% 
 
 51 
 
 149 
 
 27 
 
 370 
 
 20 
 
 40 
 
 48 
 
 9J^2 
 
 fk 
 
 43 
 
 30 
 
 X 
 
 y% 
 
 57 
 
 149 
 
 27 
 
 409 
 
 20 
 
 45 
 
 48 
 
 10 
 
 
 43 
 
 30 
 
 X 
 
 y& 
 
 63 
 
 149 
 
 27 
 
 448 
 
 20 
 
 
 54 
 
 9^2 
 
 T6 
 
 48 
 
 30 
 
 X 
 
 y?, 
 
 54 
 
 201 
 
 30 
 
 518 
 
 24 
 
 60 
 
 54 
 
 IO>^ 
 
 A 
 
 48 
 
 30 
 
 X 
 
 y 
 
 66 
 
 201 
 
 30 
 
 623 
 
 24 
 
INTERNALLY FIRED BOILERS 
 
 259 
 
 A vertical flue boiler, as shown in Fig. 306, was formerly much used 
 in the South and West for small powers in localities where the water 
 was bad. The absence of tubes made it a comparatively easy matter to 
 rid the boiler of any accumulated scale on the crown-sheet. The fire- 
 box in boilers of this kind was commonly higher than was the case with 
 tubular boilers ; the average height was approximately one-half the ver- 
 tical height of the boiler. In some cases an internal sheet-metal lining 
 was inserted in the central flue as far down as the water-line, the object 
 of which was to prevent overheating the flue when getting up steam 
 from cold water. This kind of boiler yields fairly good results, but is 
 not as much in use as formerly. 
 
 Fig. 304. 
 
 Fig. 305. 
 
 Fig. 306. 
 
 The expansion of the interior portion of a vertical tubular boiler over 
 that of the outer shell, when arranged as in Fig. 303, is considerable, 
 caused by the upward movement of the fire-box as well as the lengthen- 
 ing of the tubes, these being of higher temperature than the outer shell, 
 to which they are both attached. This expansion generally manifests 
 itself by the continual leakage of tubes in the crown-sheet. 
 
 Economy of floor space combined with moderate first cost has 
 contributed much towards making vertical boilers popular when the 
 conditions are favorable for their installation, the water actually evap- 
 orated per pound of coal equalling that of any other type. The 
 
26o 
 
 BOILERS AND FURNACES 
 
 principal objection to vertical boilers of large power is the height, a 
 loo-horse-power boiler approximating 24 feet. On the other hand, 
 vertical boilers secure an economy of floor space not equalled by any 
 other type. The boiler above referred to would require not more than 
 7 feet square for its foundation ; a horizontal tubular for the same power 
 would require a space of about 10 x 20 feet. 
 
 Cleaning a vertical boiler and removing any deposits on the crown- 
 sheet is at all times a difficult and uncertain operation, the causes for 
 which are easily to be seen by an inspection of the arrangement of 
 tubes shown in Fig. 307, which fairly represents the common method 
 of tube spacing and construction. A better arrangement of tubes is 
 
 Fig. 307. 
 
 Fig. 308. 
 
 that designed by Reynolds for vertical tubular boilers, and shown in 
 Fig. 308. Handhole-plates must be inserted at the crown-sheet level, 
 as well as at the bottom of the water-leg, for cleaning purposes, not less 
 than three at each line. 
 
 Priming is a common fault in vertical tubular boilers, due mostly 
 to the insertion of an unnecessarily large number of fire-tubes in the 
 fire-box head. 
 
 The commercial rating of vertical tubular boilers is 12 square feet of 
 heating surface per horse-power, but this is a misleading factor, because 
 one-quarter to one-half of the tube surface is above the water-level, and 
 in no wise assists evaporation. The tube surface not being as effective 
 as the fire-box surface for evaporation, the area through the tubes need 
 not, therefore, be more than that necessary for the proper escape of the 
 gases. The fire-box heating surface is highly effective and does the 
 greater part of the work, but, making allowance for this, no less than 
 12 square feet of heating surface should be allowed per horse-power. 
 Retarders or radiators may be used to advantage in vertical-boiler 
 tubes, descriptions of which are given on pages 197 and 199. 
 
INTERNALLY FIRED BOILERS 26 1 
 
 The Reynolds Boiler. — The tubes in this design are set in rows 
 radiating from a large manhole located over the fire-door and bottom 
 tube-sheet, as shown in Fig. 308. The tubes and crown-sheet over 
 the furnace can be inspected and cleaned when the manhole cover is 
 removed. Handholes are located opposite the manhole for admitting 
 light for inspecting and inserting a hose-nozzle for washing the tubes 
 and crown-sheet. Handholes are placed at intervals around the base, 
 whereby any sediment collected in the water-legs may be removed. 
 The feed-water is pumped into the internal reservoir through the feed- 
 pipe, shown in the sectional elevation, Fig. 309. This reservoir being 
 closed at the bottom, the discharge into the boiler is over the top, and, 
 it being so much larger than the feed-pipe, the current upward is very 
 slow ; consequently the feed-water gains the same temperature as the 
 water in the boiler before it is discharged into the boiler. This action 
 is effective in precipitating nearly all of the heavy impurities carried in 
 with the feed-water, which can be blown out of the reservoir by a 
 blow-off arranged for the purpose. By carrying the water in the boiler 
 slightly above the top of the reservoir, it can be utilized as a surface 
 blow-off to free the boiler of scum or light impurities collected on the 
 surface of the water. 
 
 The smoke-hood on top of the boiler is furnished with a revolving 
 top having a movable cover. For the purpose of cleaning the flues 
 this cover is removed, and only a small portion of the total number of 
 flues is exposed at one time. This arrangement enables the fireman to 
 clean the flues while the boilers are in operation. 
 
 Vertical boilers usually furnish dry steam by reason of the tube sur- 
 face above the water-level. Frequent tests of this boiler show from 10° 
 to 40° Fahr. superheating. 
 
 The Manning Boiler. — This boiler is represented in sectional ele- 
 vation in Fig. 310, accompanied by three cross-sectional drawings rep- 
 resenting sections through the fire-box, the tubes, and the smoke-box. 
 It will be observed that the fire-box is much larger than is common in 
 vertical tubular boiler construction. In all cases it is larger than the 
 diameter of the waist, or upper cyHndrical shell. This is made possible 
 by the double-flanged connection between the waist and the outer fire- 
 box shell. This connection serves another purpose, in providing for 
 the expansion and contraction between the tubes and the outer shell. 
 The ability to increase the diameter of the fire-box to the exact point 
 where the proportion of grate area to heating surface is such as to give 
 the best possible results with the most economical firing is a valuable 
 one. 
 
 The tubes have their length so proportioned to their diameter that 
 the temperature of the escaping gases is not higher than that necessary 
 to produce a good draught, about 500° Fahr., the proportion in the 
 case of a 100 horse-power boiler being 15 feet in length for a 2^-inch 
 
262 
 
 BOILERS AND FURNACES 
 
 ^ X 
 
INTERNALLY FIRED BOILERS 
 
 263 
 
 Fig. 310. 
 
264 
 
 BOILERS AND FURNACES 
 
 tube, or 72 to I. This proportion insures ample draught and prevents 
 
 injury to the tubes in the top head 
 that are unprotected by water. 
 Too much draught in a vertical 
 boiler is worse than too little, as 
 the fuel is wasted and, what is 
 even worse, the tubes are soon 
 burnt out and destroyed. With 
 an insufficient draught the boiler 
 will, of course, fail in efficiency. 
 
 The outer fire-box shell is 
 carried well above the head, and 
 handholes are placed exactly on 
 a line with the crown-sheet. The 
 tubes are placed in straight rows, 
 and at right angles to one another 
 extend two cleaning-channels of 
 ample size, shown in sections A 
 and B. A bent tube connected 
 with a hose can be inserted 
 through the handholes and be- 
 tween the rows of tubes, whereby 
 the crown-sheet can be thor- 
 oughly washed and cleaned. 
 
 The outer-shell plates, which 
 bear the greatest strain and which 
 it is impossible to brace or 
 strengthen by stay-bolting, re- 
 ceive no heat from the fire and 
 can therefore be made of any 
 required thickness.. In the fire- 
 box, on the other hand, where 
 ^^ the sheets come in direct contact 
 /H,\ ijiljiiil j ^^^j!i ^ I with the fire, they can be made 
 
 \J ',' ijijjiliji \^^^i!i w I thin enough to prevent their 
 
 ■ '''■ ' burning, and the requisite 
 
 strength is gained by stay-bolts. 
 The tubes, being of standard di- 
 mensions, have a thickness far in 
 excess of that required to enable 
 them to bear any collapsing 
 strain to which they may be sub- 
 jected. 
 
 In the water-leg are placed a 
 
 number of handholes and a cleaning-chain, by means of which any 
 
INTERNALLY FIRED BOILERS 265 
 
 sediment that may have accumulated there can be stirred up and 
 removed. 
 
 The Cornish Boiler, Figs. 311 and 312, derives its name from the 
 circumstance that boilers of this type were first used in the Cornish 
 mines ; they are still used in England for small and medium powers. It 
 consists of an outer shell, within which is a flue of sufficient size to 
 permit its being fitted with a furnace for the combustion of the fuel. 
 The products of combustion pass to the rear end of the boiler, where 
 they divide and return along the sides of the boiler to the front, where 
 they are again united and pass into a flue underneath the shell to the 
 rear end of the boiler and from thence to the chimney. The external 
 flues alongside the boiler are built of ordinary red bricks lined with fire- 
 bricks. 
 
 By reason of the large diameter of the flue and its liability to collapse 
 under a high pressure, the latter was formerly restricted to 45 pounds 
 per square inch, but with improved construction these boilers are now 
 made for any ordinary pressure, though commonly not more than 100 
 pounds. 
 
 The principal dimensions of the ordinary sizes used in England are : 
 
 Diameter of shell, 3 feet 6 inches, 4 feet 3 inches, 5 feet, 5 feet 6 inches, 6 feet. 
 Length of shell, 8 feet, 12 feet, 15 feet, 18 feet, 22 feet. 
 
 Diameter of flue, 2 feet 2 inches, 2 feet 4 inches, 2 feet 9 inches, 3 feet 3 inches, 
 3 feet 6 inches. 
 
 The efficiency of the Cornish boiler was tested incidentally in a series 
 of fuel experiments made in England, in which it appeared that in a 
 mean of thirty-seven experiments, using Welsh coal, 9 pounds of water 
 at 212° Fahr. were converted into steam at any working pressure per 
 pound of fuel ; a mean of eighteen experiments, using Newcastle coals, 
 gave 8.37 pounds of water evaporated as above ; and a mean of twenty- 
 eight experiments, using Lancashire coals, gave 7.94 pounds evaporation 
 under the same conditions. 
 
 A Cornish boiler 6 feet in diameter by 28 feet long, having a single 
 flue 3 feet 6 inches in diameter, fitted with grates 4 feet 6 inches long, 
 yielding 15 square feet of net grate surface, burned 7.24 pounds of coal 
 per square foot of grate, or 0.19 pound per square foot of heating sur- 
 face per hour, evaporating 9.9 pounds of water from 51° Fahr. into 
 steam at 60 pounds pressure by gauge, an evaporation of 11.86 pounds 
 of water from and at 212° Fahr. per pound of bituminous coal. Taking 
 into account the calorimetric value of the coal, the efficiency of the boiler 
 was 0.77. 
 
 Lancashire Boilers. — When the outer shell of a boiler on the 
 Cornish plan exceeds 5^ or 6 feet in diameter, a flue of excessive 
 diameter would be required to get a proper width of grate. It is well 
 known that flues are less able to resist a collapse as they increase in 
 
 18 
 
266 BOILERS AND FURNACES 
 
 diameter. Inasmuch as a proper width of grate can be secured by 
 the use of two smaller flues without the risks attending the use of one 
 large flue, it is a better construction if boilers are thus fitted, and this 
 particular construction of the Cornish boiler is called a Lancashire 
 boiler. 
 
 The principal dimensions of the three leading sizes used in England 
 are here given : 
 
 Diameter of shell, 6 feet, 6 feet 6 inches, 7 feet. 
 
 Length of shell, 20 feet to 28 feet, 20 feet to 30 feet, 24 feet to 30 feet. 
 
 Diameter of each flue, 2 feet 3 inches, 2 feet 6 inches, 2 feet 9 inches. 
 
 In practical working it is customary to fire the furnaces alternately, 
 so that while the one is giving off smoke and unburnt hydrocarbon gases, 
 the other is burning briskly and with the greatest heating effect. By 
 this arrangement, when the gases from the two furnaces mix in the 
 external flues, the unburnt gases given off by the freshly charged fire 
 are burned by the excess of air which has passed through the other 
 furnace, being raised to the point of ignition by the great heat of the 
 gases from the bright fire. 
 
 The Galloway Boiler. — This boiler, as manufactured by the 
 Edgemoor Iron Company, is a modification of the Lancashire boiler. 
 A sectional elevation of a boiler 7 feet in diameter by 28 feet in length 
 is given in Fig. 313. It has two furnaces 2 feet 10 inches in diameter 
 by 7 feet 9 inches in length ; the metal is ^ inch thick, welded seams, 
 flanged at each end to form three lengths riveted together with a central 
 stiffening ring similar to the detail shown in Fig. 161. These two 
 furnaces merge into a combustion-chamber of segmental cross-section, 
 shown in Figs. 314 and 315. This chamber is fitted with tapered water- 
 tubes for the purpose of increasing the effective heating surface of the 
 boiler and to promote a better circulation of water ; they also act as 
 stays, largely increasing the strength of the flue to which .they are fitted. 
 The diameters of the flanges and necks of these tubes are such that an 
 opening which will allow the smaller flange to pass through will be of 
 proper diameter for the inside of the larger flange. They are riveted 
 in place as shown. 
 
 A sectional-plan view of the boiler is given in Fig. 316, in which the 
 arrangement of the two furnaces and the conical water-tubes is clearly 
 shown. This view also shows six corrugations along each side of the 
 combustion-chamber, adding to the stiffness of the latter and assisting in 
 the deflecting and diffusion of gases back of the furnace. 
 
 The shell of the boiler is ^ inch thick, heads y^g inch thick, the fire- 
 box and conical tubes are ^ inch thick ; all the other plates are y^ inch 
 thick except the gusset-plates, which are -^ inch. The gusset-plates are 
 shown in Figs. 317 and 318, which represent front- and rear-end views 
 of this boiler. 
 
INTERNALLY FIRED BOILERS 
 
 267 
 
268 
 
 BOILERS AND FURNACES 
 
INTERNALLY FIRED BOILERS 
 
 269 
 
 Brick Setting. — The boiler is supported at the front by means of 
 a brick pier, which is built tight to the boiler for about one-third of its 
 
 circumference. See Figs. 313 
 
 and 3 1 9. The remainder of the 
 brickwork at the front end of 
 the boiler is built free from the 
 shell, allowing a space of about 
 one-half inch to be filled with 
 asbestos- fibre or other non con- 
 ducting and elastic substance. 
 After the inner side and end 
 walls and the first 9 - inch 
 arch-covering is laid, they are 
 covered with a layer of mortar, 
 and a wrought-iron case (No. 
 22 black iron) is then put on 
 and bedded into the layer of 
 mortar. The iron sheets form- 
 ing the case are united with 
 roofers' standing joints, which 
 are laid down flat with the 
 rest of the casing. This casing 
 covers the ends, sides, and top 
 arch, and after it is put on the 
 external courses of brick are 
 laid. See Fig. 320. This iron 
 casing prevents air being drawn 
 into the boiler chamber and 
 insures the temperature of the 
 gases surrounding the boiler to 
 be higher than that of the con- 
 tained steam. 
 
 The rear end of the boiler 
 is supported on a cast-iron ex- 
 pansion-rocker, shown in Fig. 
 313, behind which is built a 
 small brick pier, which does not 
 touch the boiler by about three 
 quarters of an inch. This pier is 
 built as a precautionary measure 
 against accident to the rocker- 
 casting. Without this pier, a 
 failure of the rocker support 
 would allow the boiler to fall 
 with disaster to itself and its setting. The boiler set as above described 
 
2/0 
 
 BOILERS AND FURNACES 
 
INTERNALLY FIRED BOILERS 
 
 271 
 
 is free to expand and contract without touching the brickwork of the 
 setting, except at the front end, where it is built fast and is stationary. 
 The brickwork, being exposed to but low temperatures, will not be 
 affected by the expansion and contraction of the boiler. It can neither 
 be cracked by the heat nor racked by the motion of the boiler, as must 
 occur with the ordinary brick setting. 
 
 Fig. 319. 
 
 Fig. 
 
 320. 
 
 The nozzles and manhole project through the covering arch in the 
 brick setting. A space is left in the brickwork surrounding the nozzle 
 to be filled with asbestos-fibre or other non-conducting and elastic 
 substance impervious to the passage of air. 
 
 The damper-flue leading to chimney is shown relatively to the boiler 
 in Figs, 313 and 319. So, also, the location of the bottom-blow under 
 the grating at the boiler front ; the location of the surface-blow is shown 
 in Fig. 318, The feed-pipe enters either from the top, as shown in Fig, 
 313, or from the front head, as in Fig, 318. A fusible plug, which will 
 melt out when the water-line is about three inches deep over the fire- 
 boxes, is screwed into each fire-box. The Philadelphia Rules and 
 Regulations require two safety-valves to be set for service on each 
 steam boiler. Fig. 313 shows one of these to be of the lever and weight 
 variety, the outer a spring safety-valve similar to Fig, 418, 
 
 The steam- and water-spaces are very large, — larger, perhaps, than 
 
272 BOILERS AND FURNACES 
 
 any other type of boiler of similar capacity. Large steam-liberating 
 surface, if the circulation underneath is good, is usually productive of 
 dry steam, tending to coal economy and higher efficiency in the engine. 
 
 Scotch Boiler. — An internally fired boiler by the Continental Iron- 
 Works is shown in sectional elevation, together with two cross-sections, 
 front and rear views, in Figs. 321-325. It will be observed that the 
 boiler consists of a horizontal cylindrical shell having an internal fur- 
 nace-flue, in one end of which the fire-grate is located, the other end ter- 
 minating in a fire-brick-lined back connection or combustion-chamber, 
 which is contained in the casing forming an extension of the shell 
 proper. The products of combustion are conducted therefrom through 
 horizontal tubes and delivered into the sheet-iron breeching attached to 
 the front head of the boiler. This breeching may be connected directly 
 with the smoke-stack, or, where a group of boilers is employed, it may 
 be attached to an uptake, leading to a common chimney. 
 
 Several modifications of this design may be made. The boiler, il 
 large enough in diameter, may be arranged to contain two or more fur- 
 naces, or it may be provided with a comparatively short furnace, ter- 
 minating in a combustion-chamber, from which the tubes lead direct to 
 a smoke-connection at the rear of the boiler without returning forward 
 over the furnace, forming what is known as a gun-boat boiler, numerous 
 examples of which are in use in the United States navy and by several 
 water-woirks corporations. Fig. 326 being an illustration of recent in- 
 stallations by the Philadelphia Water Department. 
 
 The furnace is the important part of an internally fired boiler. Mor- 
 ison's suspension furnace is the one illustrated here. It is necessary 
 that the furnace be able to resist the collapsing pressure to which it is 
 subjected. It must be longitudinally elastic, and of such external form 
 as to admit of its being readily freed from accumulations of scale, grease, 
 or other deposit. The furnace shown in Fig. 321 is corrugated, the 
 details of which are better illustrated in Fig. 160. It meets successfully 
 all the above somewhat exacting conditions, and, in addition, the undu- 
 lations in the furnace surfaces act as baffle-plates to the passage of the 
 furnace gases, causing them to mechanically mix, thus producing a 
 better combustion than is practicable with smooth furnaces. The addi- 
 tional heating surface in the most active portion of the boiler, due to the 
 corrugations, adds to its evaporative power. 
 
 Gun- Boat Boiler. — The city of Philadelphia recently installed 24 
 boilers, from designs by J. E. Codman, in the Queen-Lane Water- Works 
 Pumping Station, of which a longitudinal sectional elevation is shown in 
 Fig. 326, and a cross- sectional elevation in Fig. 327, the latter being on 
 a larger scale than the former. This design will be recognized as one 
 much used for marine purposes. The boiler illustrated is 8 feet 6 inches 
 in diameter by 20 feet long. The shell-plates are ^ inch thick, with 
 double-welt butt-joints on the longitudinal seams and double-riveted in 
 
INTERNALLY FIRED BOILERS 
 Fig. 321. 
 
 273 
 
274 
 
 BOILERS AND FURNACES 
 
INTERNALLY FIRED BOILERS 
 
 275 
 
2/6 
 
 BOILERS AND FURNACES 
 
 the circumferential seams ; a detail of this joint is given in Fig. 328. 
 
 The heads are ^ inch thick. There are two Fox corrugated furnaces 
 
 3 feet 6 inches in diameter by 8 feet long, the longitudinal seams of 
 
 which are welded. The metal is ^| inch thick ; the corrugations are 
 
 on 6-inch centres. 
 
 Fig. 328. 
 
 'O 
 
 O 
 
 ff-e, 
 
 o 
 
 -& 
 
 rP 
 
 j'/^i 
 
 ^■^M- 
 
 o -it-o- 
 
 o 1 cb I o 
 
 0000-- 
 000 \io -i 
 
 o 9 o o o o o- 
 o I o o o o o 
 
 o o o >o -O-v^ o ! o 
 000 o---C^- o CD 
 
 0000 
 00000- 
 
 cb I 9 
 
 I o --^"o- 
 
 I --O ^"vl; — - 
 
 
 o\ 
 
 -O/ 
 
 
 ^si^-^^ 
 
 The combustion-chamber is situated midway in the boiler ; its form 
 is clearly shown in the two elevations. The front head is flanged to 
 receive the rear ends of the two furnaces ; the back head is fitted with 
 90 tubes 4 inches in diameter. The front and rear heads, as well as the 
 plates joining them, are ^ inch thick. The combustion-chamber is 
 stiffened longitudinally by crown-bars 5^ x i^ inches placed on 6-inch 
 
 centres, shown in Fig. 
 Fig. 329. 329. Each crown-bar 
 
 carries 3 stay-bolts i inch 
 in diameter on 7-inch 
 centres, also shown in 
 place in Fig. 326, but 
 enlarged in separate de- 
 tail in the upper right- 
 hand corner. Fig. 327. 
 Double angle-irons 4 x 
 3>^ X y^ inch riveted to 
 the bottom shell of this 
 chamber on 8-inch centres perform the same office. Connecting stays 
 provide for the transfer of collapsing strains, to which these plates are 
 subjected, to the outer shell, the latter being fitted with T-sections 4 x 
 4. X y inch to make a substantial link connection. 
 
 The 90 tubes above mentioned are 9 feet 4^ inches to outside of 
 heads, spaced on 6-inch horizontal and 5-inch vertical centres. 
 
INTERNALLY FIRED BOILERS 
 
 277 
 
 The bracing of the heads of this boiler is clearly shown in both sec- 
 tional engravings. Four 3^ x 4 x }4 inch angles extend across the 
 front and back heads in pairs, as shown in both sectional drawings ; 
 these are detailed on a larger scale in Fig. 330. The angle-irons are 
 riveted to the heads by ^-inch rivets on 4-inch centres. Between each 
 pair of angles are 5 through-going stay-rods 2j^ inches in diameter with 
 upset ends to 2^ inches in diameter. Each end is fitted with nuts, a 
 collar being furnished at the inside to clear the angles. Two similar 
 bolts are placed under the furnace on either side of the under manhole. 
 
 Fig. 330. 
 
 Three 11 x 15 inch manholes are provided, one at the top and one 
 near the bottom of each head. For the bottom openings a j-\-inch plate 
 24 X 27 inches, with the corners clipped, as shown in Fig. 327, is secured 
 to the head by 32 rivets, to restore in part the weakness occasioned by 
 the opening. Over the rear manhole is a 3)^ x 6 x ^ inch angle riveted 
 to the head by ^-inch rivets on 5 -inch centres. This angle extends 
 across the head to the beginning of the flange-curve. The manhole on 
 the top of the boiler has a ring 45^ inches wide by ^ inch thick, double- 
 riveted to the shell. 
 
 The steam-dome is 30 inches in diameter by 3 feet 6 inches high 
 over all ; the shell is -^-^ inch thick, double-riveted to the boiler ; the 
 dome-head is also y^g- inch thick, bumped to a radius of 3 feet. An 8- 
 inch diameter opening is made in both the boiler and dome-head ; the 
 latter is fitted with a cast ring riveted to the dome-head, and to which 
 the steam fittings are bolted. The opening into the boiler is reinforced by 
 a ring 4 inches wide by ^ inch thick, double-riveted to the boiler-shell. 
 
 The details of the riveting are so clearly shown on the dimensioned 
 drawing, Fig. 328, as to require no further explanation. 
 
 Locomotive Boiler. — ^This name is applied to a certain design in 
 boilers much used in stationary-engine practice because of its universal 
 adoption as a type in locomotive construction. Fig. 331 represents in 
 sectional elevation a boiler designed for stationary purposes by the 
 Pennsylvania Railroad. This boiler is used for testing and for other 
 purposes. The fire-box is well adapted for burning bituminous coal. 
 
278 
 
 BOILERS AND FURNACES 
 
INTERNALLY FIRED BOILERS 
 
 279 
 
 The dimensions are : 49^ inches in width at the bottom ; the top 
 measures 42 inches in width at the front and 38 inches at the rear ; the 
 taper at the sides is well illustrated in Fig. 332 ; the inside length of the 
 furnace is 72 inches ; the height from the top of the grates to the under- 
 side of crown-sheet is 50 inches ; the crown-sheet is ^ inch ; the other 
 inside sheets are -^ inch thick ; the water-space around the bottom of 
 the fire-box is 4 inches ; the fire-door opening is 16 inches in diameter 
 and is flanged similar to Fig. 192 ; the bottom ring is 2 inches deep. 
 
 Fig. 332. 
 
 Fig. 333- 
 
 The barrel of the boiler is 50 inches in diameter, straight on top, 
 and provides a distance of 18 inches from crown-sheet to shell. The 
 laps of the sheets are such that the drainage will be into the water-space 
 around the fire-box without pocketing. The barrel-sheets are ^ inch 
 thick, so also the sheets forming the outside of the fire-box. There are 
 74 tubes 3 inches in diameter by 11 feet in length, making the total 
 length of the boiler 20 feet, including 26 inches depth of smoke-box. 
 The fire-box tube-sheet is }4 inch thick and constructed somewhat dif- 
 ferently from the ordinary, in being a separate head flanged around the 
 edges and riveted to a corresponding flanged opening in the fire-box 
 sheet, clearly shown in Fig. 331. 
 
 The steam-dome is 30 inches in diameter by 30 inches in height, 
 plates inch thick. The circumferential sheet to which it is riveted is 
 flanged to receive the dome, as shown in Fig. 196. Underneath the 
 opening for the dome are five cross-stays 2 x ^ inches of cross-section 
 extending across the barrel and riveted at both ends, as shown in Fig. 
 
 333- 
 
 The stays for both front and rear heads are secured at one end by 
 bolts passing through angle-irons riveted to the head, at the other end 
 by riveting to the circumferential shell, both of which details are shown 
 
28o 
 
 BOILERS AND FURNACES 
 
 in Fig. 334. These stay-rods are i}i inches diameter. The fire-box 
 stay-bolts are j4 inch diameter, placed on approximately 4-inch centres. 
 
 Fig. 334. 
 
 The crown-bars are double, each 4 inches deep by 2^ inches in 
 width over all, carrying 7 stay-bolts of ^ inch diameter. The arrange- 
 ment of head, thimble, washer, and nut for one of the stay-bolts is 
 shown in Fig. 335, which also shows the proportionate dimensions of 
 crown-bars. There are 14 horizontal through-going stays, one between 
 each crown-bar. These are j4 inch diameter, with one end enlarged 
 to I inch for passing the ^-inch stay through the larger tapped hole. 
 These pass directly over the crown-sheet, as shown in Fig. 331. There 
 are 15 crown-bars, every other one being fitted with two links, one on 
 either side of the centre, connecting with the outer shell, as shown in 
 Fig, 332 and in enlarged view in Fig. 335. 
 
 Fig. 335. 
 
 The sectional area of the tubes is 3.123 square feet. The grate 
 area is 22.2 square feet, the ratio of tube to grate area being 22.2 -=- 
 3.123 = 7.1 to I. - The total heating surface is 734 square feet, of which 
 96 is in the fire-box and 638 square feet in tube heating surface, the ratio 
 of total heating surface to grate surface being 734 -^ 22.2 = 33 to i. 
 
 A shaking-grate is provided, the details of which are shown in Fig. 
 331. Three bars are cast together with one inch air-space between, the 
 whole width of one grate being 5^ inches. A lever projects downward 
 from each grate for making connection with a link-bar, to which all the 
 grates are attached. This link -bar is operated by a lever pivoted to the 
 
INTERNALLY FIRED BOILERS 
 
 281 
 
 outside of the fire-box as shown. On each end of each grate-bar is a 
 trunnion i^ inches diameter resting in a groove included in the side- 
 bar castings, fastened to the sides of the fire-box. 
 
 The smoke-box extends 26 inches beyond the tube ends ; it is 
 fitted with smoke-doors and a saddle for making connection with the 
 chimney. In this case the smoke-flue is downward, an underground 
 flue leading to the chimney, 
 
 TABLE XLVII. 
 
 PRINCIPAL DIMENSIONS OF PORTABLE BOILERS WITH OPEN BOTTOM AND 
 WATER-FRONT. 
 
 iii 
 a 
 
 1 
 
 Furnace. 
 
 TuBKS 3 Inches 
 IN Diameter. 
 
 Thickness. 
 
 Dome. 
 
 1 
 
 
 
 Pi 
 
 PQ 
 
 
 
 
 
 
 
 
 
 
 
 (fi 
 
 
 3 
 
 ^ 
 
 
 
 
 
 
 
 
 1 
 
 
 
 'o 
 
 o^- 
 
 
 
 
 
 
 
 1-' 
 
 
 
 
 
 
 
 rt 
 
 
 
 .a 
 
 
 
 
 (U 
 
 
 
 (fi 
 
 
 
 
 J3 ™ 
 
 g 
 
 
 
 .c 
 
 .a 
 
 
 J3 
 
 
 
 
 
 J3 
 
 
 
 E 
 
 s 
 
 bi 
 
 ^ 
 
 .5? 
 
 
 S 
 
 d 
 
 g 
 
 i 
 
 S 
 
 5 
 
 i 
 
 ^> 
 
 
 
 
 
 
 
 V 
 
 
 
 
 3 
 
 
 
 
 S 
 
 U 
 
 S 
 
 J 
 
 ^ 
 
 X 
 
 J 
 
 ^ 
 
 w 
 
 ti, 
 
 H 
 
 ■q 
 
 m 
 
 5 
 
 J 
 
 H.-P. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Feet. 
 
 
 In. 
 
 In. 
 
 In. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Feet. 
 
 25 
 
 40 
 
 48 
 
 34 
 
 33 
 
 8 
 
 34 
 
 ¥ 
 
 ? 
 
 H 
 
 22 
 
 24 
 
 18 
 
 i3i4r 
 
 30 
 
 42 
 
 50 
 
 36 
 
 34 
 
 8 
 
 40 
 
 A 
 
 A 
 
 y. 
 
 22 
 
 24 
 
 20 
 
 14 
 
 35 
 
 44 
 
 50 
 
 38 
 
 36 
 
 8/, 
 
 44 
 
 A 
 
 A 
 
 H 
 
 26 
 
 28 
 
 20 
 
 I4>^ 
 
 40 
 
 44 
 
 50 
 
 38 
 
 36 
 
 10 
 
 44 
 
 A 
 
 A 
 
 % 
 
 26 
 
 28 
 
 20 
 
 16 
 
 50 
 
 48 
 
 54 
 
 42 
 
 40 
 
 10/2 
 
 54 
 
 j% 
 
 A 
 
 y% 
 
 26 
 
 28 
 
 22 
 
 i6|< 
 
 60 
 
 54 
 
 60 
 
 48 
 
 44 
 
 II 
 
 60 
 
 -5^ 
 
 A 
 
 Yz 
 
 30 
 
 34 
 
 24 
 
 18 
 
 70 
 
 56 
 
 60 
 
 50 
 
 44 
 
 12 
 
 66 
 
 -1 
 
 
 H 
 
 30 
 
 34 
 
 26 
 
 19 
 
 80 
 
 58 
 
 60 
 
 52 
 
 48 
 
 12 
 
 76 
 
 -i 
 
 T^^ 
 
 
 32 
 
 36 
 
 26 
 
 19 
 
 90 
 
 
 60 
 
 52 
 
 48 
 
 14 
 
 76 
 
 T6 
 
 y% 
 
 32 
 
 36 
 
 26 
 
 21 
 
 100 
 
 62 
 
 60 
 
 56 
 
 50 
 
 14 
 
 90 
 
 J 
 
 A 
 
 y^ 
 
 32 
 
 40 
 
 30 
 
 21^. 
 
 no 
 
 64 
 
 60 
 
 
 52 
 
 14 
 
 100 
 
 ^8 
 
 T6 
 
 y. 
 
 36 
 
 40 
 
 30 
 
 21% 
 
 125 
 
 66 
 
 60 
 
 60 
 
 58 
 
 15 
 
 108 
 
 /8 
 
 A 
 
 A 
 
 36 
 
 40 
 
 32 
 
 22% 
 
 The Belpaire Fire-Box is unlike the ordinary fire-box of the loco- 
 motive type in many of its details of construction, but especially in the 
 method of bracing, the former having no crown-bars and no radial stay- 
 bolts made to accommodate curved surfaces seldom or never parallel to 
 each other. The crown-sheet in the Belpaire fire-box, as applied to 
 locomotive boilers, is flat, and the roof- sheet lies parallel to it, as shown 
 in Fig. 336. Stay-bolts extend in straight lines and at right angles from 
 the crown-sheet to the roof-sheet. In locomotive practice, for 160 
 pounds steam pressure, these stay-bolts are ]/q inch diameter in the body, 
 with screw-ends enlarged to i inch, 12 threads, and placed on 4^- to 
 4|^-inch centres. Horizontal tie-rods extend across the fire box to stay 
 the flat sides ; in locomotive boilers these are usually three in number, i 
 inch diameter in the body with screw-ends i}i inches. They are spaced 
 longitudinally to the same centres as the stay-bolts between which they 
 are located ; vertically the distance varies according to the height over 
 the crown-sheet, but the approximate centre to centre is about 5^^ 
 
 19 
 
282 
 
 BOILERS AND FURNACES 
 
 € I 
 
 Fig. 336. 
 
 g B Q g 
 
 @— # 
 
 =E I 
 
 :1^ 
 
 %, 
 
 0000000000000 
 0000000000000 
 0000000000000 
 0000000000000 
 0000000000000 
 0000000000000 
 . 00000000000 , 
 \ 000 000 / 
 
 fl: 
 
 r 
 
 inches, with the centre of the lower tie-rod about two inches above the 
 crown-sheet. This method of staying is advantageous, in having all 
 
 the stays and tie-rods in 
 tension in straight lines, 
 as compared with radial 
 stays fitted to curved 
 sheets, in which it fre- 
 quently happens that not 
 more than one or two full 
 threads are had in the 
 outer sheet, whereas by 
 placing the sheets par- 
 allel full threads are had 
 in both. 
 
 A Belpaire Fire- 
 Box Boiler, differing in 
 many respects from the 
 preceding, and shown in 
 the longitudinal sketch. 
 Fig. 337, is from designs 
 by E. D. Leavitt, Jr., 
 who has long used this 
 type of boiler with that 
 success which character- 
 izes his engineering de- 
 signs. The boiler here 
 illustrated is 7 feet in diameter of shell by nearly 35 feet in length from 
 out to out, and is intended to work under a continuous steam pressure 
 of 135 to 140 pounds per square inch. 
 
 Two fire-boxes are included in this design, with a central water-leg 
 between, as shown in Fig. 338. Each furnace is connected with a large 
 central combustion-chamber, in which a thorough admixture of the gases 
 and final combustion occurs before these gaseous products enter the 
 tubes. The crown-sheets are flat, but the roof-sheet of the fire-box end 
 is curved to the same radius as the barrel of the boiler ; this necessitates 
 a somewhat unusual scheme of bracing, that the vertical stay-bolts shall 
 pass through no curved surfaces in the roof-sheet. This is accomplished 
 by the introduction of forged frames placed on about 5-inch centres. One 
 of these frames is shown in place in Fig. 338 ; the two ends of it rest 
 upon and are riveted to the two crown-sheets ; nearly vertically above 
 these riveted joints the frame is riveted to the roof-sheet. Horizontal 
 offsets are then made for a stay-bolt on each side of the boiler, and a ver- 
 tical offset for a riveted stay passing through a distance-piece on each 
 side of the boiler, after which the frame extends horizontally across the 
 fire-boxes. Six vertical stay-bolts, with distance-pieces between, fix the 
 
 &i 
 
 O 
 
INTERNALLY FIRED BOILERS 
 
 283 
 
 horizontal portion of this forged 
 frame and transmit any strains 
 received to the curved roof-sheet 
 above. This arrangement per- 
 mits the insertion of 8 vertical 
 stay-bolts, 4 on each crown-sheet, 
 with a central space through 
 which a man may crawl for in- 
 spection of the interior ; it also 
 permits the removal and re- 
 placing of any of the vertical 
 stays. The front, central, and 
 outside water legs have ^-inch 
 stay-bolts on 4^ -inch horizontal 
 by 4^-inch vertical centres. The 
 outer and inner sheets of the front 
 head are flanged for two fire-door 
 openings, each 21 inches wide 
 by 24 inches high, the bottom 
 of each opening being about 12 
 inches above the bottom of the 
 boiler. The front head above the 
 crown-sheets is stayed by means 
 of two pairs of angles and one 
 T-iron, all riveted to the head, 
 as shown in Fig. 339. Eight 
 tie rods extend from end to end 
 of the boiler ; these are ij4 inches 
 ■diameter with 2^ -inch ends. The 
 arrangement of nuts and distance 
 pieces between the angle-irons is 
 also shown in the above illustra- 
 tion. The relative positions of 
 these rods from centre to cen- 
 tre, as arranged for carrying the 
 strains, is shown in Fig. 340 and 
 Fig- 338. The sides of the fire- 
 box swell from 7 feet diameter of 
 barrel to 9 feet spread at the bot- 
 tom of the fire-box. This intro- 
 duces some difficulties not usu- 
 ally experienced in boiler design. 
 The ring of plates correspond- 
 ing to the gusset-sheet has to be 
 greatly strengthened. The three 
 
284 
 
 BOILERS AND FURNACES 
 
INTERNALLY FIRED BOILERS 
 
 285 
 
 thicknesses and throat flange are shown in longitudinal section in Fig. 
 339, as are also the brackets for stiffening the fire-box sheet around the 
 throat. 
 
 Fig. mo. 
 
 The combustion-chamber is shown in cross-sectional elevation in 
 Fig. 341. Its crown-sheet is on the same level as that of the furnace. 
 Inasmuch as there is no central water-space at this point, additional stay- 
 bolts are inserted at the centre of this crown-sheet. The opening pro- 
 vided over and between the fire-box crown-sheets is here continued, 
 whereby a man may crawl through the centre of the bracing over the com- 
 bustion-chamber by employing a frame similar to that already described 
 for the fire-box crown-sheets and suspending another frame below it ; this 
 provides for five short vertical stays, shown in the drawing. An end view 
 is given of the tube-sheet which closes the rear end of the combustion- 
 chamber ; it provides for 118 tubes 2>/4 inches in diameter. These tubes 
 are 18 feet long inside the heads. Underneath the combustion-chamber 
 is a space large enough to permit a man to examine or clean that portion 
 of the boiler. Forged stays, fitting the outer and inner curves of the plates 
 to be joined, and which do not admit of radial stays, are also shown. The 
 details of the rear head are shown in Fig. 342, and further supplemented 
 
286 
 
 BOILERS AND FURNACES 
 
 foobooOoooOooo^^- 
 
 •OOOOOOOOOOOOOOO' 
 ' OOOOOOOOOOOOOOO 
 'OOOOOOOOOOOOOOO. 
 ^OOOOOOOOOOOOOOO^ 
 
 ooooooooooooo 
 
 jpOOOOOOOOOOOOi 
 
 by the longitudinal section, Fig. 339. Angle- and T-irons are riveted to 
 the head above the tubes, with provision for the eight longitudinal tie-rods. 
 
 The feed-pipe enters 
 Fig. 341. through a nozzle at 
 
 the top of the boiler, 
 branches to either side, 
 and continues forward 
 between the shell of the 
 boiler and the tubes 
 to within a couple ol 
 feet of the combustion- 
 chamber, shown by 
 dotted line in Fig. 337. 
 These pipes are of brass 
 with perforations along 
 the submerged horizon- 
 tal branches for dif- 
 fusing the feed-water 
 throughout the entire 
 length. The fire-tubes 
 are sufficiently above 
 the bottom of the barrel 
 of the boiler to admit a man through its whole length. A series oi 
 brackets on about 6-inch centres are placed radially around that portion 
 of the head not suffi- p^^. ^42. 
 
 ciently stayed by the 
 fire-tubes. These brack- 
 ets are riveted to both 
 head and shell, and are 
 shown in Figs. 339 and 
 342. Brackets for simi- 
 larly supporting the 
 Iront head are shown 
 in the left-hand portion 
 of Fig. 339- A rear 
 elevation is shown in 
 Fig. 343. A cast-iron 
 plate closes the entire 
 end, in which is a suita- 
 ble opening for attach- 
 ing the plate-metal con- 
 nection leading to the 
 chimney, underneath 
 which are two hinged doors giving access to the tubes. 
 
 The fittings and attachments include a manhole-opening under the 
 
 ^OOOOOOOOOOOOOOO, 
 ^OOOOOOOOOOOOOOO, 
 
 OOOOOOOOOOOOOOO' 
 ^OOOOOOOOOOOOOOO 
 
 OOOOOOOOOOOOOOO 
 
 ooooooooooooo 
 .ooooooooooooo 
 00000000000 
 
 .000 
 
 00. 
 
INTERNALLY FIRED BOILERS 
 
 287 
 
 tubes ; the usual handhole-openings in the water-legs and at the surface 
 of the crown-sheet ; a blow-off connection under the barrel of the boiler 
 near the rear end ; an 18-inch ring manhole-opening on the top of the 
 barrel of the boiler ; a safety-valve nozzle ; a steam-pipe nozzle. Two 
 dry pipes, perforated along their upper sides, extend fore and aft from 
 the latter nozzle, shown in Fig. 337. A cast-iron front, shown in Fig. 
 339) covers that end of the boiler, making an attractive finish and pro- 
 tecting the non-conducting covering at a point always liable to injury. 
 
 Fig. 343. 
 
 The covering consists of a coating of plaster-of-Paris and sawdust 2^ 
 inches thick, over which is placed a layer of hair-felt i inch thick and a 
 painted canvas cover. The fire-box doors are cast iron and are divided 
 horizontally, as shown in Fig. 340. They have perforated baffle-plates 
 and register openings in each half. 
 
 The plates of the boiler are ^-^ inch for the main shell and -^ inch 
 for the fire-boxes, all of mild steel of 60,000 pounds tensile strength, 
 with an elastic limit of 40,000 pounds. The joints of the shell are 
 double butt-joints, the horizontal seams being triple-riveted, the vertical 
 ones double-riveted. 
 
288 
 
 BOILERS AND FURNACES 
 
 The expansion of this boiler is towards the rear, the front end resting 
 upon a cast-iron ash-pit provided with cleaning-doors, shown in Figs. 
 337 and 340, no part of which is below the floor level. Three additional 
 supports are provided under the barrel of the boiler, the latter resting 
 upon cast-iron cradles, shown in Fig. 343 . These cradles have a broad 
 base at each end, supported upon balls which permit a free movement 
 in any direction, and are provided with adjusting screws, shown in Fig. 
 339. by means of which the weight can be equally distributed over the 
 entire surface of the balls. 
 
 Fig. 344- 
 
 ft ft IL F-'-'^"-^ ^^ "ff^ 
 
 A design for a Belpaire fire-box, by F. W. Dean, is shown in cross- 
 section through the fire-box in Fig. 344, the details being those in 
 sight when looking towards the rear end of the boiler. This design 
 applies to those fire-boxes which are approximately rectangular in form 
 over the crown-sheet, as indicated in the engraving. The usual longi- 
 tudinal tie-rods extending from front to rear heads are shown above the 
 crown-sheet in section. Additional to these are transverse tie-rods 
 extending across the steam-space. Vertical stay-bolts connect the roof- 
 sheet of the boiler with the crown-sheet below, as shown. It is almost 
 imperative in large boilers that the fire-box bracing be accessible for 
 examination or repairs. This demands a central space through which 
 a man may crawl into and examine the interior of the boiler. In order 
 to accomplish this it is necessary to do away with the central transverse 
 
INTERNALLY FIRED BOILERS 
 
 289 
 
 tie-rod commonly used and also two or more of the central vertical 
 stay-bolts. The removal of these tie-rods and stay-bolts is a source of 
 weakness ; therefore, to make good the loss of strength occasioned by 
 the omission of these parts, trussed connecting-bars are here employed, 
 and these are provided with a central hub through which a bolt passes 
 and by which the outer shell of the fire-box is securely bolted to the 
 trussed connecting-bars on the inside. In this manner the fire-box 
 shell is as well strengthened as if a through-going tie-rod was used. 
 The horizontal tie-rods have outside and inside nuts, by means of which 
 the trussed connecting-bars are firmly secured to the outer shell. The 
 loss of strength occasioned by the removal of any two of the central 
 vertical tie-rods is made good by increasing the diameters of the two 
 inner tie-rods and in providing a trussed connecting-bar similar to the 
 one above referred to, except that the inner nuts for clamping it to the 
 outer shell are omitted. A series of these trussed connecting-bars are 
 located centrally at the top of the boiler. Connecting-bars of some- 
 what different shape, but answering the same purpose, span the central 
 water-space ; vertical bolts riveted to the inner side of the central water- 
 leg pass up through these connecting-bars, each bolt being supplied 
 with nuts, as shown. 
 
 Fig. 345. 
 
 Portable Engine Boilers. — These are usually of the locomotive 
 pattern, and are very generally provided with a water-bottom, as shown 
 in Fig. 345. The object of the water-bottom is to prevent live coals 
 falling upon the ground, thereby lessening the fire risk in agricultural 
 districts, where portable engines of small and medium powers are' prin- 
 cipally used. 
 
 This type of boiler permits of furnace length and height suited to 
 any kind of fuel. The width of furnace is governed by the diameter 
 of the barrel and the water-space at the sides and bottom of the furnace. 
 
290 BOILERS AND FURNACES 
 
 The sides of the fire-box are somewhat deficient in circulation, and as 
 a result are liable to fill with deposits if hard water is used ; for this 
 reason handhole plates should be provided at the corners that any 
 sediment may be easily removed. 
 
 A water-front is shown in the sectional illustration, Fig. 345, which 
 also shows a water-bottom. The outer fire-box head is reversed so as 
 to afford the best facility for riveting the head and shell together. The 
 head next to the smoke-box has its flange turned outward, as at the 
 fire-box end. A handhole should always be provided for cleaning the 
 barrel of the boiler. 
 
 The fire-door opening is commonly made with a wrought-iron ring 
 riveted through, as shown in Fig. 187. Some of the cheaper boilers in 
 the market have cast-iron rings, but such rings are not always safe, 
 because liable to crack without warning. 
 
 Another method of constructing a furnace is shown in Fig. 346, in 
 which the furnace-sheet extends through and is riveted to the outside 
 sheet at the fire-box end, the latter being flanged to receive it. This 
 opening is covered by a cast-iron front, in which is included the fire- 
 door opening, a bearing-bar for the grates, and an ash-pit opening. 
 As this cast plate is subjected to an intense heat, its interior must be 
 covered with fire-brick tiles. The fire-door is fitted with a perforated 
 plate, as in stationary-boiler practice. 
 
 Portable engine boilers when mounted on wheels are commonly from 
 8 to 15 horse-power, seldom more than 20. The diameter of the waist 
 or barrel for such boilers ranges from 26 to 32 inches, consequently the 
 width of the fire-box varies from 21 to 28 inches. This easily permits 
 the use of an arched crown- sheet supported by radial stays instead of 
 crown-bars necessary to the proper supporting of flat crown-sheets. 
 Radial stays are much to be preferred to crown-bars for portable boilers 
 because of the little room occupied by them in the boiler as well as the 
 lesser weight. Such stays provide a satisfactory transfer of the collapsing 
 strains upon the crown-sheet to the outer shell. Radial stays obstruct 
 the circulation much less than crown-bars ; they also permit a greater 
 thoroughness of inspection and cleaning. The stay-bolts joining the 
 outer and inner plates around the fire-box are commonly % inch in 
 diameter for ^-inch plates and ^ inch in diameter for y%- to |4-inch 
 plates. The radial stays and those used in the side and bottom of the 
 fire-box are commonly of the same diameter. For long stay-bolts the 
 outer plate should be tapped enough larger than the inner one that the 
 main body of stay shall pass through without being obliged to screw 
 the stay its whole length through the outside hole. This necessitates 
 the use of two taps, the upper one of which is a hob secured to a shank, 
 forming the smaller tap. Both threads must be of the same pitch and 
 so placed that the threads shall start at the same point in a revolution ; 
 this necessitates the same provision being made in threading the stay- 
 
INTERNALLY FIRED BOILERS 
 
 291 
 
 bolt. The front and rear heads are braced similarly to those for hori- 
 zontal tubular boilers ; that is to say, the stays are riveted to the head, 
 extend back, and are then riveted to the shell of the boiler. 
 
 The tubes are commonly 3 inches in diameter for all portable and 
 semi-portable boilers, whatever may be their size. 
 
 Fig. 346. 
 
 c 
 
 Hi II life 
 
 A steam-dome is commonly furnished portable boilers, the opening 
 into which is not usually any larger than the diameter of the pipe leading 
 to the engine. Small holes should always be provided at the bottom 
 where the dome and barrel intersect for the drainage of any water ol 
 condensation back into the boiler. See Fig. 195. 
 
 The smoke-box is commonly an extension of the barrel of the boiler 
 and fitted with a cast-iron frame and door, giving easy access to the tubes. 
 
 A saddle for the smoke-stack attaches to the opening provided at the 
 top of the smoke-box. In the case of portable boilers it is further pro- 
 vided with a hinged ring, permitting the stack to be laid down upon the 
 boiler when transporting it from place to place. 
 
 Handhole plates must be inserted in the water-leg for the removal 
 of scale or mud which may accumulate there. 
 
 In some tests made for the writer, a very inferior bituminous coal 
 being used for fuel, the feed-water entering the boiler at 65° Fahr., an 
 eight horse-power boiler evaporated 12 cubic feet of water per hour, a 
 ten horse-power boiler evaporated 15 cubic feet in the same time. The 
 evaporation was under a pressure of 80 pounds per square inch. The 
 boilers were in the condition in which they are usually delivered to the 
 trade, and the firing in the test was as near as possible the same as it 
 would have been in the hands of the purchaser, except that it was con- 
 ducted with a view to ascertain the actual evaporative capacity of the 
 boiler instead of being an economy trial. 
 
2(^'. 
 
 BOILERS AND FURNACES 
 
 TABLE XLVIIL 
 
 PRINCIPAL DIMENSIONS OF PORTABLE BOILERS WITH WATER-BOTTOM AND 
 CAST-IRON FRONT, FIG. 346. 
 
 ^ 
 
 •5 
 
 Furnace. 
 
 Tubes 3 
 
 Inches 
 
 Tb 
 
 ICKNESS. 
 
 Dome. 
 
 .iii 
 
 s. 
 
 3 
 
 rt 
 
 
 
 IN Diameter. 
 
 
 
 
 
 rt 
 
 •3 
 
 
 
 
 
 
 
 
 
 
 • 
 
 
 
 
 
 .2 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 . 
 
 y 
 
 i! 
 
 
 
 
 
 • 
 
 
 .1 
 
 
 u 
 
 
 tu 
 
 =3 
 
 J 
 
 1 
 
 5 
 
 
 ■g 
 
 .CI 
 
 bo 
 
 J 
 
 .0 
 
 S 
 3 
 
 1 
 
 
 
 .2 
 Q 
 
 •a 
 
 5 
 
 
 H.-P. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Feet. 
 
 
 In. 
 
 In. 
 
 In. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Feet. 
 
 6 
 
 26 
 
 34 
 
 21 
 
 29 
 
 5 
 
 15 
 
 y^ 
 
 A 
 
 3/8 
 
 12 
 
 15 
 
 12 
 
 93/ 
 
 8 
 
 28 
 
 36 
 
 22 
 
 33 
 
 6 
 
 18 
 
 % 
 
 A 
 
 3/8 
 
 14 
 
 16 
 
 14 
 
 II 
 
 10 
 
 30 
 
 3« 
 
 24 
 
 3,S 
 
 7 
 
 20 
 
 % 
 
 A 
 
 3/8 
 
 16 
 
 18 
 
 14 
 
 12 
 
 12 
 
 32 
 
 38 
 
 26 
 
 3.S 
 
 6>^ 
 
 24 
 
 % 
 
 ^\ 
 
 3/8 
 
 18 
 
 20 
 
 16 
 
 II3^ 
 
 15 
 
 32 
 
 44 
 
 26 
 
 3.S 
 
 7 
 
 24 
 
 % 
 
 A 
 
 3/8 
 
 18 
 
 20 
 
 16 
 
 13 
 
 20 
 
 34 
 
 ,S2 
 
 28 
 
 37 
 
 8 
 
 28 
 
 Va 
 
 t\ 
 
 3/8 
 
 20 
 
 22 
 
 16 
 
 14% 
 
 25 
 
 36 
 
 52 
 
 30 
 
 40 
 
 8i^ 
 
 32 
 
 % 
 
 A 
 
 3/8 
 
 20 
 
 22 
 
 18 
 
 15 
 
 30 
 
 40 
 
 60 
 
 34 
 
 43 
 
 8 
 
 38 
 
 A 
 
 
 3/8 
 
 22 
 
 24 
 
 20 
 
 15 
 
 35 
 
 40 
 
 60 
 
 34 
 
 43 
 
 9 
 
 40 
 
 
 A 
 
 3/8 
 
 22 
 
 24 
 
 20 
 
 16 
 
 40 
 
 40 
 
 60 
 
 34 
 
 43 
 
 loJ^ 
 
 40 
 
 385- 
 
 A 
 
 3/8 
 
 22 
 
 24 
 
 20 
 
 17/2 
 
 50 
 
 44 
 
 64 
 
 3« 
 
 .SO 
 
 ii>^ 
 
 46 
 
 fV 
 
 t'V 
 
 3/8 
 
 22 
 
 24 
 
 22 
 
 19M 
 
 60 
 
 48 
 
 64 
 
 42 
 
 52 
 
 12 
 
 52 
 
 A 
 
 T^^ 
 
 y% 
 
 26 
 
 28 
 
 22 
 
 20 
 
 The "Electric" return tubular portable boiler," shown in Fig. 347, 
 is not unlike the locomotive fire-box pattern of portable boiler, except 
 that the crown-sheet is placed lower in order that return tubes may be 
 
 Fig. 347. 
 
 located above it. The fire-box does not differ fi-om the ordinary ; it is 
 provided with a water-bottom and a cast-iron fi-ont lined with fire-brick 
 
INTERNALLY FIRED BOILERS 
 
 293 
 
 Fig. 348. 
 
 tiles. The tubes are of two diameters, — the lower ones being 4 inches, 
 the upper ones 3 inches, — the combined areas for each diameter approxi- 
 mately equalling those of the other. The 
 products of combustion pass through the 
 shorter and larger tubes to the back con- 
 nection and from thence return through 
 the smaller and longer tubes to the front 
 connection leading into the smoke-stack. 
 The crown-sheet, which is usually the high- 
 est part of the locomotive type of boiler, 
 has the least water to protect it and is the 
 first part exposed to low water. The pro- 
 tection afforded the crown-sheet in this 
 design is all that can be desired ; in general, 
 it is a compact boiler for the power de- 
 veloped and has good steaming qualities. 
 
 The ' ' Economic' ' return tubular boiler, 
 shown in Fig. 349, is of somewhat unusual 
 construction. The front end is cylindrical 
 in form, the rear end is oval, the length of 
 each portion being about one-half that of 
 
 the entire length of the boiler. A furnace is attached to and underneath 
 the front end. The lower portion of the rear end extends down far enough 
 
 Fig. 349. 
 
 to hold the short tubes leading from the furnace to the back connection. 
 The lower half of the front end of the boiler, being over the fire, forms 
 
294 
 
 BOILERS AND FURNACES 
 
 Fig. 350. 
 
 the crown-sheet : it will be noticed that it occupies a position relatively 
 
 to the fire quite the reverse of ordinary portable boiler practice, the 
 latter commonly having the crown-sheet 
 arched over the fire and not recessed 
 towards it ; this design insures that the 
 crown-sheet is always well supplied with 
 water. 
 
 The furnace consists of iron plates at- 
 tached to the cylindrical shell of the boiler, 
 which plates extend to any depth suitable 
 for the location of the grates adapted to 
 fuel to be burned. The side walls and 
 front of the furnace are lined with fire- 
 brick, and these are held in place by iron 
 rods protected from the fire. Both rods 
 and fire-brick can be removed and replaced 
 whenever necessary. Fig. 350. 
 
 The advantages claimed for this boiler 
 over the ordinary internally fired portable 
 boiler of the locomotive type is that it 
 occupies no more space, is an equally rapid 
 steamer, there are no water-legs in which 
 to deposit sediment, and that it combines 
 
 with the safety of the stationary return tubular boiler the convenience 
 
 and portability of those of the locomotive type. 
 
 An internally fired boiler of about 40 horse-power, by Tonkin, shown 
 
 in Fig. 351, is of somewhat unusual construction for a semi-portable 
 
 Fig. 351. 
 
 boiler. This boiler is 54 inches in diameter and about 15)^ feet long. 
 The fiirnace is 36 inches in diameter and 54 inches long, approximating 
 
INTERNALLY FIRED BOILERS 295 
 
 13.5 feet of surface. It is closed by a cast-iron front, which includes 
 the fire-door opening, grate bearing-bar, and an ash-pit opening. The 
 bridge-wall, or what corresponds to a bridge-wall in ordinary boilers, 
 is made of cast iron and perforated for the admission of air through 
 slots provided in the casting. Those included in the inclined portion of 
 the casting are located under the fire, and practically answer the purpose 
 of an extension of the grate-bars so long as they are kept covered with 
 fuel. The openings at the rear admit air into the combustion-chamber 
 between the end of the bridge-wall and the tube-sheet. 
 
 There are sixty tubes 3 inches in diameter by 9 feet in length. The 
 ratio of tube area* to grate surface is 13.5 -e- 2.5 = 5.4 to i. 
 
 The total heating surface is 474 square feet, the ratio to grate surface 
 being 474 -h 13.5 = 35.1 to i. 
 
 This boiler is set at an inclination to the horizontal for the purpose 
 of increasing the depth ol water at the fire-box end. 
 
CHAPTER IX. 
 
 SECTIONAL AND WATER-TUBE BOILERS. 
 
 Sectional boilers are those made up of a number of units similar 
 in size and shape and so constructed that when suitably assembled they 
 will admit of performing all the functions of a steam boiler. To the late 
 Joseph Harrison, Jr., is due the credit for the invention of what may be 
 considered an ideal sectional boiler, consisting, as it does, of a series of 
 simple units, spherical in form, convenient in size, and capable of aggre- 
 gation to any extent required in practical operation. 
 
 Mr. Harrison's invention dates back some thirty years. At that time 
 boiler pressures did not often exceed 75 pounds per square inch. Mild 
 steel had not yet been introduced, but the quality of cast iron was in all 
 respects satisfactory for his purpose, because such castings were not only 
 of ample strength, but were found to better resist the corrosive action 
 of acids in the feed-water than wrought iron, the only other material 
 then in use for land boilers. Castings were also less affected by oxida- 
 tion in the furnace and were free from blisters. 
 
 This boiler possessed at that time the unique distinction of being 
 constructed wholly of cast iron, a material hitherto considered unsafe 
 in boiler construction. Its use was adopted, and since continued, be- 
 cause it contributes better than any other material to the manufacture of 
 spherical units with their connecting necks for grouping into "slabs," 
 as these aggregations of units are called. Harrison's selection of this 
 material was the result of certain guiding principles in design, which led 
 him to the adoption of the spherical form of unit because its strength 
 is in no respect dependent upon any system of stays or braces ; further, 
 these spherical units need not be of any great size or weight, allowing 
 the easy displacement and replacement or interchange of one or more 
 of these units without disturbing the remaining portion of the structure. 
 
 The spherical form was adopted because it combined greatest strength 
 with greatest heating surface, the hollow sphere being twice as strong as 
 a tube of equal diameter. These spheres as now made are 8 inches in 
 external diameter, with curved necks 4 inches in external diameter, the 
 thickness varying from ^ inch in the necks to ^ inch in the central 
 line of the sphere. These units are subjected to a hydrostatic pressure 
 of 350 pounds per square inch before they are passed for the erecting- 
 shop. Two, three, or four spheres are included in a single casting, 
 joined together by suitable necks, an illustration of which is given in 
 296 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 297 
 
 Fig"- 352- These units must of necessity be uniform and duplicate. 
 No present means are known whereby either wrought iron or mild steel 
 can be economically wrought into such shape. The units have milled 
 faces made by special machinery, and are combined in vertical sections 
 by means of bolts 
 
 passing through their ^^^- 352- 
 
 centres which are en- 
 tirely surrounded by 
 water. The joints so 
 made are perfectly 
 steam- and water- 
 tight, iron to iron, 
 without the aid of 
 any packing what- 
 ever. In case of re- 
 pair, it is accom- 
 plished by the inser- 
 tion of a new unit 
 of exactly the same 
 size and shape as the 
 defective one removed, which does not detract from the original con- 
 struction. Cast iron easily permits all parts to be tooled to a uniform 
 standard gauge, and such parts will go together without forcing. 
 
 A slab, as already explained, is an aggregation of units held together 
 by wrought-iron bolts, the heads of which are protected from the fire by 
 round caps, shown in Fig. 352, while their upper ends extend through 
 the common caps and terminate with a screw-thread and close-ended 
 nut. The slabs to the required number, and of varying sizes according 
 to the size of the boiler, are suspended side by side from a suitable iron 
 framework, and are also connected at top and bottom to form a proper 
 steam and water coupling. The slabs are hung one inch apart, and are 
 slightly inclined upward from front to back, so as to secure a uniform 
 area of steam -liberating surface, while the height of the water-line may 
 vary. 
 
 Safety was the first consideration in Harrison's mind when designing 
 this boiler ; his one object was to make it absolutely secure from de- 
 structive explosions, even when carelessly used. From the method of 
 joining, a rupture cannot extend beyond the unit in which it originates. 
 The effect of the rupture is, therefore, localized, so that the adjacent 
 parts of the boiler are not affected except in so far as they may be in- 
 jured or displaced by the rupture of a defective unit. A boiler of this 
 construction was tested by a committee of the Franklin Institute, during 
 which the pressure was increased to 875 pounds per square inch, when 
 a sudden discharge of steam took place, after which the pressure fell to 
 450 pounds, at which it stood when the fire was drawn. 
 
298 BOILERS AND FURNACES 
 
 The Wharton-Harrison Boiler, a modified form of the original 
 Harrison boiler, is shown in Fig. 353, which, it is claimed, possesses 
 certain advantages over the earlier forms. For example, the changed 
 relative positions of the units to the radiant heat and impinging hot 
 gases is an improvement, and with this there is an increased depth of 
 water over the fire which admits of a gauge-glass one-half as long again 
 
 Fig. 353. 
 
 in this form of slab as was possible in the earlier forms, relieving the 
 attendant from anxiety about the position of the water-level. Injury to 
 the boiler from low water, whether due to neglect or accident, and which 
 would have caused leaky joints in the old style, is practically obviated in 
 the present design. 
 
 A longer travel for the steam after its liberation from the water im- 
 pinging against the highly heated surfaces in its flow outward from the 
 boiler results in greater dryness than was the case formerly. 
 
 The cost of repairs is reduced to a small percentage of what it once 
 was, because they are more easily made. 
 
 Longer travel for the products of combustion between the slabs and 
 around the units before escaping to the chimney-flue results in the better 
 absorption of heat by the boiler, and the consequent reduction in tem- 
 perature of the waste gases ; there is thus secured an increase in the 
 amount of power for a given space. 
 
 The present design affords increased facility for thorough interior 
 
SECTIONAL AND WATER-TUBE BOILERS 299 
 
 cleaning, as, when the bolts are removed, any soft sediment or loose 
 scale can be washed down through the vertical openings. 
 
 The brickwork for this type of boiler, when compared with other 
 boilers, is, by reason of its nearly cubical form, quite economical, the 
 cost being much less. As the entire weight of the boiler proper is 
 borne by the metal columns and beams, the brickwork is reduced to 
 mere enclosing walls carrying no load : they can be set, repaired, or 
 removed without disturbing the boiler. The covering over the boiler 
 is formed by special tiles. 
 
 The construction of these boilers is favorable for introduction into 
 cellars and other inconvenient places, or for mule-back transportation 
 over mountainous localities, because the boiler can be shipped in sec- 
 tions, no one of which exceeds 200 pounds in weight. 
 
 Water-Tube Boilers. — This type of boiler has a history extending 
 back for more than a century ; but the water-tube boiler of to-day can 
 hardly be said to extend further back than the invention of Wilcox in 
 1856, which invention, changed in form but without loss of identity, 
 appeared again in 1867 in the Babcock & Wilcox design, one widely 
 known and probably more generally copied than any other. 
 
 A water-tube boiler consists usually of an assemblage of tubes filled 
 with water, located in a furnace, and connected with two receivers, — the 
 lower one of small diameter called the mud-drum, and an upper and 
 usually larger one called the steam-drum. The mud-drum is placed in 
 the coolest part of the furnace, sometimes protected from the heat of 
 the furnace by a non-conducting covering, and not infrequently it is 
 placed outside of the furnace altogether. The upper receiver, or steam- 
 drum, is placed in the hottest part of the furnace, and performs the 
 double function of a storage for water needed for circulation, as well as 
 providing a large disengaging surface for the liberation of steam. The 
 water-surface being usually at the centre of the drum, the upper half is 
 steam-room, from which the supply of steam is taken. 
 
 The tubes are inclined to the horizontal, though in some cases the 
 inclination is from the vertical, and so arranged that they receive water 
 through the lowest headers, sometimes from the mud-drum, discharging 
 the water through the upper headers into the steam-drum, the tubes 
 being so located in the furnace that the flame and hot gases pass over 
 and between them, usually at right angles to their axis. The tubes are 
 both straight and curved, depending on the design of the boiler. The 
 Babcock & Wilcox, Heine, Cahall, and variations of these designs 
 have straight tubes ; the Hogan, Stirling, Morrin, and boilers of that 
 class have bent tubes. 
 
 The manufacturers of boilers with bent tubes claim that the bending 
 is advantageous, because such tubes adapt themselves readily to the 
 expansion of the boiler and prevent leaky joints. The manufacturers 
 of boilers with straight water-tubes claim that such tubes can be more 
 
300 BOILERS AND FURNACES 
 
 easily cleaned, examined, and renewed than is the case with bent 
 tubes. 
 
 The present tendency in boiler design for steam pressures higher 
 than ICO pounds per square inch is in the direction of water-tube boilers, 
 of which there is now a great variety of designs in active competition. 
 This type of boiler, when properly designed and built, is well calculated 
 to carry high steam pressures. The tubes are seldom more than four 
 inches in diameter ; they are not only quite thin, but possess a large 
 factor of safety. The evaporative efficiency is fully equal to any type 
 of boiler yet introduced. It permits also a large power-rating in mod- 
 erate space. 
 
 Circulation.- — Water-tube boilers require a free and abundant sup- 
 ply or circulation of water through them to prevent overheating. This 
 circulation adds directly to the efficiency of the boiler ; it also con- 
 tributes to both durability and safety. The cause of circulation in water- 
 tube boilers has been attributed to a difference in density caused by a 
 difference in temperature of the water within the boiler by the action of 
 the fire, such that one portion of the boiler becomes more highly heated 
 than another. The circulating movement is very gentle at first, because 
 all the water in the boiler is at approximately the same temperature ; 
 but as the temperature rises and the tubes become filled with foam or 
 steam-bubbles, it becomes more intense, partly because of the entraining 
 action of the bubbles, until a powerful circulation is had throughout the 
 whole interior, which, when once secured, may be indefinitely prolonged 
 by the continued appHcation of heat. 
 
 Wet steam is a common fault in water-tube boilers. To obviate this 
 the steam generated in the tubes should have a free and direct passage 
 through suitably designed and properly located headers leading directly 
 into the steam-drum. Attempts have been made to counteract this fault 
 by the employment of superheating surfaces which were not altogether 
 satisfactory, though some designs permit superheating with less risk 
 than others. 
 
 This lack of sufficient steam- and water-room is a serious defect ot 
 many water-tube boilers. A small water- space may be attended by a 
 capacity for quickly getting up steam to a high pressure when all 
 escapes for steam are shut off, but unless there is a sufficiently large 
 volume of heated water in reserve ready to convert itself into steam 
 as rapidly as the draught of steam is made from the steam-drum, the 
 pressure will fall quicker than it was raised in spite of a fairly good fire, 
 unless the boiler is of more liberal capacity than usually furnished for a 
 given power. Consequently a closer attention will be required on the 
 part of the fireman than with ordinary flue or tubular boilers. 
 
 Draught Area. — This receives considerable attention in horizontal 
 tubular- boiler practice, but apparently less in the case of water-tube 
 boilers. The cross-sectional area in fire-tube boilers varies from y to |r 
 
SECTIONAL AND WATER-TUBE BOILERS 301 
 
 that of the grate surface, an average of }i being considered good prac- 
 tice. Too large a draught opening in any boiler has a tendency to re- 
 duce its efficiency. An editorial in vol. xxxi. of the Engineering Record 
 calls attention to this defect in furnace proportion, and makes reference 
 to water-tube boilers known to have been erected with a draught opening 
 as large as ^ of the grate surface, resulting in a wasteful temperature 
 of chimney gases. The draught area for water-tube boilers need not be 
 greater than has been found necessary for return tubular and internally 
 fired boilers. In the water-tube boiler most largely in use the direc- 
 tion of the current of gases is subjected to less abrupt change than in 
 return tubular boilers, for in the latter the current is forced to change 
 to an exactly opposite direction as the gases enter the tubes, while in 
 the former the motion is continuously forward, though not in a continu 
 ous straight line. Consequently a less force of draught and a smaller 
 total area should suffice in the water-tube boiler than in the return 
 tubular. 
 
 The heating surfaces in water-tube boilers are of the most effective 
 kind, because the heat acts upon the large aggregate tube surface and 
 a comparatively small quantity of water for any given cross-section 
 through the tubes ; if the circulation is not interfered with, such heating 
 surfaces permit steam to be quickly and continuously made. 
 
 The facilities for inspection, cleaning, and repairing are of the utmost 
 importance, because these have a direct bearing upon the efficiency and 
 durability of the boiler. Water-tube boilers are complicated structures 
 at best, and if to this be added inaccessibility for cleaning, the efficiency 
 will be quickly lowered and result in permanent injury. One of the 
 advantages possessed by nearly all the later designs of water-tube 
 boilers is the facility afforded for the renewal of parts by a system of 
 duplication which obtains in all modern and properly equipped manu- 
 factories. 
 
 Freedom from disastrous effects of explosion is one of the merits 
 particularly emphasized by makers of water-tube boilers. It is true 
 that no overwhelming disaster and wreckage of property, such as often 
 accompanied the explosion of Cornish, Lancashire, horizontal tubular 
 and cylinder boilers has been chargeable to water-tube boilers, nor is it 
 ever likely to be, but the casualty list resulting through the breakage of 
 headers and other constructive details is by no means small. 
 
 The transportation and setting up in place of water-tube boilers is 
 in most cases both easy and economical ; there are, however, excellent 
 water-tube boilers which cannot be taken apart for easy transportation, 
 nor more easily erected in places difficult of access than is the case with 
 horizontal tubular boilers. For ordinary supply this is a matter of little 
 consequence, but if a boiler is to be erected in a city basement or in an 
 inconvenient mountain district, it must be selected with special reference 
 to the facts governing the case. 
 
302 
 
 BOILERS AND FURNACES 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 303 
 
 The Babcock & Wilcox \A^ater-Tube Boiler. — The general 
 design of this boiler is shown in Fig. 354. It is composed of an assem- 
 blage of lap-welded wrought-iron _ 
 tubes, an overhead horizontal 
 drum extending the whole length 
 of the tubes, and a mud-drum 
 below extending across the rear of 
 the boiler at the bottom of the 
 combustion-chamber, all of which 
 are connected together in one 
 complete circulatory system. The 
 particular example here illustrated 
 is 9 tubes high and 7 tubes wide by 
 20 feet in length, the tubes being 
 placed at an inclination of 15° to 
 the horizontal, as shown. 
 
 These tubes are 4 inches in 
 diameter, arranged in series by 
 means of a vertical connection 
 called a header. A number of 
 these headers can be placed side 
 by side to make any requisite 
 width, as shown in Fig. 355. In 
 this boiler the headers are of such 
 form that the tubes are staggered, 
 or placed zigzag, so that the flow of 
 gases upward among the tubes is 
 interrupted and made to impinge 
 against the sides and bottom of 
 each tube instead of passing through a straight opening. The holes 
 are bored and reamed tapering in the headers, into which the tubes are 
 expanded so as to make a tight joint under pressure : a detail of this 
 joint is shown in Fig. 356. Opposite each of these openings is a hand- 
 hole, detailed in Fig. 357. The handhole plates and headers are faced 
 by milling the surfaces to accurate metal contact, making a tight joint 
 without the use of packing of any kind. The front header is attached 
 to the water- and steam-drum in two ways, one underneath the drum by 
 a cross-box, as in Figs. 355 and 356, the other in the head of the drum, 
 as in Fig. 358. 
 
 A drum cross-box is riveted to the underside and to each end of the 
 upper drum. After having been bored and reamed, these receive the 
 nipples which connect the headers, as shown in Fig. 355, In case 
 connection is made directly into the head of the drum, as shown in 
 Fig. 358, a cast-iron head is employed. 
 
 The headers are made of cast iron and of forged steel. Both kinds 
 
304 
 
 BOILERS AND FURNACES 
 
SECTIONAL AND WATER-TUBE BOILERS 305 
 
 are now in use, — the former for pressures approximating 100 pounds, the 
 latter for higher pressures. The connection of each of the rear headers 
 to the drum is in all respects similar to that at the front just described, 
 having a pipe instead of a nipple connecting the rear cross-drum to the 
 headers arranged below. From the bottom of each of the rear headers is 
 a nipple, forming a connection with the mud-drum below, which is located 
 at the lowest part of the combustion-chamber back of the bridge-wall. 
 
 The drum is of a diameter and length suited to the size of the boiler. 
 It is usually made up of three sheets, as shown in Fig. 354, the longi- 
 tudinal seams being butt-strapped inside and out, with two rows of rivets 
 passing through both straps and the shell, and two rows through the 
 shell and inner strap only, as shown in Fig. 359. Manhole-openings 
 are provided in the heads, as 
 
 shown in Fig. 356, for such as Fig. 359. 
 
 are made of plate metal, and 
 in Fig. 358 for those made of 
 cast iron. A deflector is placed 
 in the front end of the drum 
 when connection is made from below, as shown in Fig. 356, the object 
 of which is to prevent a violent agitation of the water at the front end, 
 caused by the high velocity of the upward current from the headers in 
 the body of the water in the drum, the deflector changing the direction 
 of the upward current and directing it towards the rear. The steam- and 
 water-gauge fittings are attached to the front of the drum in the ordinary 
 manner. The feed-pipe enters the front head of the drum, as shown in 
 Fig. 354, extending through the deflector, so that the end of the feed- 
 pipe is under it and to the rear of the openings in the cross-box above 
 the line of header connections. 
 
 The completed boiler is suspended in the furnace by a strap at the 
 front and rear end of the drum, as shown in Fig. 354. The ends of 
 these straps are fitted with screw-threads passing through cast-iron 
 washers resting upon channel-beams, which extend across from side to 
 side of the furnace and rest upon iron columns, which carry the weight 
 of the boiler on suitable foundations below the ground level. By this 
 means the entire boiler may be, and is, suspended in place before the 
 brickwork is begun. The furnace-walls may, therefore, be constructed 
 independently of the boiler ; so, also, any portion of the furnace-walls may 
 be repaired, taken down, or rebuilt without affecting the boiler itself 
 
 The furnace front has large doors, by which complete access can be 
 had to all of the handhole plates in the front headers. The ordinary 
 fire- and ash-pit doors are underneath, as is common to stationary boilers. 
 
 The mud-drum is made of cast iron, commonly 12 inches in diameter, 
 bored and reamed for expanding the wrought-iron nipples, as is done 
 with the headers. It is tapped for blow-off connections and furnished 
 with handhole for cleaning. 
 
3o6 
 
 BOILERS AND FURNACES 
 
 Operation : The fire is made under the front and higher end of the 
 tubes. The products of combustion pass up between and around the 
 tubes into a combustion-chamber above and underneath the steam- and 
 water-drum. From thence the gases pass down between the tubes, 
 thence once more up through the spaces between the tubes, and off to 
 the chimney. The water inside the tubes, absorbing heat from the gases, 
 is heated and tends to rise towards the higher end ; if it receives heat 
 enough, the water is converted into steam. The mingled column of steam 
 and water, being of less specific gravity than the solid water at the back 
 of the boiler, rises through the headers into the drum above, where the 
 steam completely separates from the water, the latter flowing back to the 
 rear of the drum and down through the connecting tubes into the lower 
 headers and inclined tubes in a continuous circulation. If the passages 
 are all large and free, this circulation will be very rapid, sweeping away 
 the steam as fast as formed and supplying its place with water ; absorb- 
 ing the heat of the fire to good advantage ; causing a thorough com- 
 mingling of the water throughout the boiler and a consequent equal 
 temperature ; preventing also to a degree the formation of deposits or 
 incrustations upon the heating surfaces, these being deposited in the 
 general course of circulation in the mud-drum, M^hence they are blown 
 out. 
 
 Fig. 360. 
 
 Zell Water-Tube Boiler. — This boiler is shown in sectional ele- 
 vation in Fig. 360. It consists of an assemblage of tubes and a water- 
 drum placed at an angle of about 15° to the horizontal. These tubes 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 307 
 
 are equally spaced both vertically and horizontally, and arranged in 
 zigzag vertical rows which are connected at the rear by suitable header 
 
 Fig. 361. 
 
 connections, which also attach to a mud-drum located at the bottom of 
 the combustion-chamber. At the front end of the boiler these tubes 
 connect with similar headers, which are joined together by suitable con- 
 nections with the water-drum. 
 
 The tubes are standard 4- F^^- 362. 
 
 inch lap-welded wrought-iron 
 boiler tubes arranged in sec- 
 tions or headers of four tubes 
 each. An elevation and sec- 
 tion of one of these headers is 
 shown in Fig. 361. These 
 headers are made of cast iron 
 bored and reamed to receive 
 the tubes which are expanded 
 in the holes thus prepared ; 
 each header is fitted with two 
 internal handhole plates, each 
 of which covers an opening 
 across two tubes. The hand- 
 hole plates are fitted to the 
 inside of each header, as 
 shown, the plates being held 
 in position by T-bolts, the 
 heads of which fit into slots 
 on the outer faces of the plates. 
 The pressure of steam holds 
 the plates up to the joints 
 without the aid of the bolts, a 
 much safer method than that 
 of putting these plates or caps 
 on the outside and securing 
 them by bolts from the inside, for if the threads slip or if the bolts break 
 under steam pressure, the plate will fly off and steam and boiling water 
 
3o8 
 
 BOILERS AND FURNACES 
 
 would rush out, to the danger of hfe and property. The headers are 
 connected in series vertically by means of nipples expanded into bored 
 holes, as shown in Fig. 362. The number of these headers to be placed 
 in vertical and horizontal rows will depend upon the size of the boiler or 
 the power required. 
 
 A horizontal mud-drum 12 inches in diameter is placed immediately 
 below the bottom of the rear headers and below the line of circulation ; 
 it is connected with the several headers extending across the furnace by 
 means of short expanded wrought-iron nipples. Since cast iron has 
 been proved to be the best material to resist the action of corrosion, 
 these mud-drums are made of that material. They are tested up to 500 
 pounds hydraulic pressure, and have the necessary handholes and plates 
 for cleaning. The plates closing these handholes have the same internal 
 surface-bearing and general arrangement as those closing the handholes 
 in the headers. Suitable blow-off and feed openings for flushing, with 
 proper fittings, are placed at opposite ends of the drum. 
 
 The inclined water-drum at the top of the tubes is attached to the 
 headers by wrought-iron nipples expanded into bored and reamed holes, 
 
 as shown in Fig. 362. 
 F^^- 363- A cast-iron saddle is 
 
 EL__0__^ riveted to the front 
 
 water-drum, and each 
 saddle has sufficient 
 area of opening for the 
 tubes below. These 
 are connected with the 
 top headers by means 
 of expanded wrought- 
 iron nipples in the 
 same manner that the 
 headers are connected 
 together. This drum is made of mild steel plate with convex heads 
 and manhole. A saddle is riveted to the top of the highest part of the 
 water-drum, to which is attached by an expanded nipple connection 
 a manifold having a number of openings for pipes which convey the 
 steam to the steam-drum at the rear of the boiler, shown in Fig. 362 
 and on enlarged scale in Fig. 363. 
 
 The steam-drum is located to the rear and at the top of the boiler 
 and furnace. It is made of mild steel. The heads are convex in shape 
 and will withstand high pressure without bracing. There is a manhole 
 at one end of the drum, the cover-plate for which has its bearing on the 
 inside of the drum in a similar manner to the headers. This drum is 
 located in a horizontal position at the rear end of the water-drums, 
 and its length is somewhat greater than the combined width of the 
 
SECTIONAL AND WATER-TUBE BOILERS 309 
 
 headers. It acts as a superheater and reservoir for the storage of 
 steam and contains no water, thereby differing from the usual water- 
 and steam-drums which accompany other forms of water-tube boilers. 
 A connection is made from it to the water-drum below by means of 
 wrought-iron tubes expanded at each end into the drums. These con- 
 necting tubes are incased in cast-iron columns, not shown in the en- 
 graving ; they are flanged at the ends, giving a uniform bearing on each 
 drum. These columns are the support of the steam-drum and relieve 
 the wrought-iron tubes from all strains of that nature. Being made in 
 half sections and bolted together, they can readily be removed to inspect 
 the enclosed tubes. The proper openings and fittings are provided for 
 safety-valves and steam piping. 
 
 Superheating pipes connect the manifold, which extends across the 
 whole width of the front of the boiler, to the steam-drum at the rear ; 
 these are lap-welded wrought iron, placed in a single row and expanded 
 in both the manifold and the steam-drum. 
 
 The boiler is supported at the rear end by resting upon two or more 
 cast-iron saddles placed under the mud-drum ; these castings are set 
 upon a substantial brick wall. The front end of the boiler rests upon 
 a roller, which is placed upon the top of the arch-box, shown in Fig. 
 362, thus allowing freely for expansion and contraction. This arch-box, 
 through which there is a constant circulation of air, is covered on the 
 side nearest the furnace with large and especially moulded fire-brick, 
 which protects it from injury by the heat. It rests upon cast-iron stands, 
 one on each side of the furnace, which are embedded in brickwork and 
 anchored to the foundations. The entire weight of the boiler is carried 
 by the foundation supports mentioned above ; that is, it is not suspended, 
 but supported directly from underneath. The manufacturers of this 
 boiler claim that all water-tube boilers that are suspended from above 
 must have the combined weight of the boiler and its contained water 
 carried by the wrought-iron nipples in the expanded joints which tie the 
 structure together, subjecting such joints to strains for which they were 
 never designed or intended, which strains are mainly eliminated when 
 supporting the boiler from below. 
 
 The flame-plates, shown in Fig. 360, consist of sections made of cast 
 iron and are faced with fire-brick ; they are usually two in number, and 
 extend from side to side of boiler and from top to bottom of tubes. 
 
 The feed-water pipe connects at the rear end of the water-drum. 
 The water descends into the rear headers and is then distributed through 
 the tubes. By this means of feeding it is expected that more or less of 
 the impurities in the feed-water will be precipitated on contact with the 
 hot water in the boiler and, gravitating downward, will collect in the 
 mud-drum below. 
 
 The operation is such that from the fire in the furnace the products 
 of combustion pass up between the staggered tubes and water-drums, 
 
3IO 
 
 BOILERS AND FURNACES 
 
 then down and around the same between the flame-plates to the bottom 
 of the combustion-chamber, thence up again, making three runs across 
 the water-tubes, and out through the space between the steam- and 
 water-drums to the flue, and thence to the chimney. The water in the 
 incHned tubes next to the furnace is first heated and raised to a rapid 
 state of ebuUition. This mingled body of steam and water, being lighter 
 than the solid water in the rear of the drum, is forced rapidly up through 
 the front headers into the water-drum, where the steam and water sepa- 
 rate ; the steam passes up through the manifold into the superheating 
 tubes, which, being surrounded with hot gases, absorb additional heat, 
 evaporating any entrained water in its passage to the steam-drum ; this 
 latter being jacketed by the waste gases in their passage to the chimney 
 and having a temperature of a few degrees higher than the steam within 
 the drum, more heat is absorbed by the steam and it becomes further 
 superheated. 
 
 Fig. 364. 
 
 The Gill Boiler in its general features is not unlike the Babcock & 
 Wilcox boiler, and is shown in Fig. 364. The water-tubes are 4 inches 
 in diameter and spaced about 3 inches apart ; they incline at an angle of 
 about 15° to the horizontal. The tubes are grouped in sets of four or 
 more, which are expanded into bored holes in a cast-iron box or header 
 at each end in such a way that the tubes are staggered instead of being 
 placed one above the other, shown in Fig. 365. By making these 
 boxes short, and by connecting them by slightly flexible tubes, the 
 danger of breakage is practically eliminated. In the engraving it will be 
 seen that the column of headers at the left hand shows a 5- and 6-hole 
 header with the caps removed, together with the curved tube-connection 
 with the steam-drum ; the column at the right shows similar headers, 
 with their inside handhole, caps, bolts, and dogs in place. The mid- 
 dle column shows a 4- and 6-hole header in section at the middle 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 311 
 
 of the headers, showing the manner of connecting the headers with 
 each other and with the steam-drum. The small illustration at the 
 bottom shows a section through the three headers and four tubes at 
 the line A. Handholes are provided in 
 the headers at the front and rear of all 
 the tubes through which a scraper can be 
 used to remove any sediment or scale 
 that may have lodged in the tubes or in 
 the headers. The tubes may also be re- 
 moved through these handholes when 
 necessary and new ones inserted. These 
 openings are closed with caps placed in- 
 side of the headers and packed with thin 
 rubber gaskets. The assemblage of tubes 
 is divided into two or three sections by 
 one or two flame-walls made of fire-brick, 
 secured by rods running through them 
 and into the side-walls. The flame and 
 heated gases are thereby compelled to 
 traverse the tubes two or three times 
 before finding their exit at the rear end 
 of the furnace. 
 
 The steam-drum is made of mild 
 steel, in diameters varying from 30 to 
 50 inches, or so proportioned as to 
 allow a cubic-foot capacity for steam- 
 space and a cubic foot for water storage 
 in the drums for each horse-power to 
 be developed. Thirty-inch drums have 
 been found to be the smallest diameter 
 that will allow of the production of dry 
 steam under conditions of rapid driving. 
 The water-tubes located over the fire 
 receive the most intense part of the 
 heat, and absorb a considerable portion 
 of it before it reaches the drum ; the 
 
 latter is not, therefore, subjected to the same destructive energies that 
 would otherwise occur, and thus large drums may be employed with 
 comparative safety in water-tube boilers. 
 
 A mud-drum, usually 18 inches in diameter, is placed under the rear 
 row of headers, the connection being made of short nipples. This 
 drum serves to collect the sediment floating in the water and hold it 
 until it is blown off. 
 
 A structural framework composed of wrought-iron I-beam columns 
 and channel-bar cross-beams is erected, within which the drums, tubes. 
 
312 
 
 BOILERS AND FURNACES 
 
 and headers are suspended, thus allowing the working parts to expand 
 and contract independently of the structural parts. 
 
 The Caldwell "Water-Tube Boiler, Fig. 366, differs principally 
 from those already described in the construction of the header and in 
 
 Fig. 366. 
 
 the distribution of the current of gases in the furnace, the combination 
 of the horizontal water- and steam-drum, the inclined assemblage of 
 tubes, and the mud-drum underneath being substantially that of the 
 Babcock & Wilcox design. 
 
 Fig. 367. 
 
 The tubes are of wrought iron, 4 inches in diameter, built up in 
 sections of 4 tubes, each with a cast header at each end of the tubes, 
 into which they are secured by expanding. These sections are set over 
 each other three, four, and five high, and nippled together by means of 
 short pieces of 4-inch extra heavy tubing, expanded into bored holes of 
 uniform diameter, making as a whole a series of flexible members yield- 
 ing to the inequality of expansion among the tubes. 
 
 The header castings are quadrangular in shape, and fully detailed in 
 Fig. 367. The openings and handholes in front, through which access 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 313 
 
 is had to the tubes, are covered with plates which close on an inner lip, 
 whereby a tight joint is secured, the pressure assisting. 
 
 The baffle-bricks used in giving direction to the current of gases in 
 the furnace are shown in Fig. 368. The uses of these bricks are to 
 absorb the heat of the furnace while 
 the furnace doors are shut, to give 
 it out again when they are opened 
 and the cold air is rushing in during 
 the firing, thus saving the tubes 
 from some of the evil effects of 
 sudden changes in temperature to 
 which such boilers are commonly 
 subjected. It is claimed they also 
 offer a more positive impingement 
 of the gases on the tubes than is 
 gained by merely staggering the 
 tubes. It is further claimed that 
 the gases passing upward between 
 the bricks and the tubes tend to 
 keep the tops of the tubes clean, 
 
 whereas in water-tube boilers not thus equipped there is always a ridge 
 of soot which gathers on the upper side of each tube. The baffle-bricks 
 are only used in the first and last passes of the tubes, these being upward 
 passes, and not in the middle, which is a downward pass. Baffle-bricks 
 are not used on the three or four rows of tubes immediately over the 
 fire, these being within the influence of the radiant heat from the fire, 
 and require no baffling to hold the gases back. The baffling is provided 
 for on the top rows only. These tubes being away from the furnace and 
 radiant heat, the circulation of gases between the upper and lower tubes 
 is thus equalized and their efficiency increased. The baffle-bricks lie 
 on and partly embrace the top portion of the tubes, as shown in Fig. 368 ; 
 there is, therefore, a constant absorption of heat by the bricks, to be 
 afterwards given out by them to the upper portion of the tubes. Inas- 
 much as flame and heat in ascending form the cone-shaped flame of a 
 candle, so also the last pass of the gases at the upward and outward 
 corners of the assemblage of tubes do not get as much heat as the 
 middle tubes, unless the gases are directed towards these places by means 
 similar to that just described. It was found when operating at rated 
 capacity that the temperature of flue gases was lowered from 480° to 
 410° Fahr. by a proper distribution of the baffle-bricks in the last pass 
 of the gases. They also serve to reduce the draught between the 
 tubes, which would ordinarily be too great to attain the best results in 
 water-tube boilers ; they also furnish a ready means for cutting down 
 the draught area in the last compartment in the furnace to that required 
 for a proper distribution of the heated gases over and around the tube 
 
314 
 
 BOILERS AND FURNACES 
 
 surfaces, a necessity for realizing the highest economy in water-tube 
 boilers. 
 
 The Root Boiler. — The general design of this boiler is shown in 
 Fig"- 369- It may be briefly described as consisting of an assemblage 
 of inclined water-tubes connected at their front and rear ends with a 
 series of horizontal overhead water- and steam-drums ; the fire being 
 applied under the front and raised ends of the tubes, its action is to 
 cause an upward and forward flow in the tubes, the water thus forced 
 from the tubes into the front end of the horizontal drums being con- 
 stantly replaced by a downward flow from the rear end of the same 
 drums. This continuous circulation is common to other water-tube 
 boilers, and is an essential requisite to successful operation. 
 
 Fig. 369. 
 
 The tubes are expanded in pairs at both ends into bored holes in cast- 
 iron headers, the pair of tubes with their headers forming what the manu- 
 facturers call a package. See Fig. 371. The two tubes in one pack- 
 age are on the same horizontal level when set up, and are in identical 
 conditions as regards exposure to flame and temperature of contents ; 
 therefore, the expansion and contraction of both must be identical. The 
 headers are stacked alongside of and over each other in such a manner 
 as to form straight horizontal rows of tubes, while the rows of tubes up 
 and down are staggered as shown. There is no connection between 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 315 
 
 any one header and another lying either to the right or left of it, but 
 
 each header is connected with the one above it by means of a return 
 
 bend, shown in Fig. 371. The uppermost header of each vertical row, 
 
 front and back, is connected respectively with the front and back of an 
 
 overhead horizontal drum, of which latter there are as many in a boiler 
 
 as there are vertical rows of headers. 
 
 A horizontal drum at the rear with its underlying row of headers 
 
 and the tubes therein form a separate section of the boiler. The rows 
 
 of headers at the front and back of each section are all connected with 
 
 the ends of the overhead drums to insure a complete circulation. 
 
 The joints between headers ^ 
 
 ,11 1 • 1 r 1 Fig. 370. 
 
 and bends are made m the lol- 
 
 lowing manner : in line with 
 the tubes and on the face of 
 the header are milled recesses, 
 in which fit rings of gun -metal 
 which have a bearing at the 
 bottom of the recesses. These 
 rings are bored on a taper, and 
 into this taper is engaged the 
 end of the return bend, turned 
 to the same taper. Lugs are 
 cast on the bend, whereby 
 two T-headed bolts have their 
 heads pocketed in recesses cast 
 on the outer face of the head- 
 ers, and these serve to draw 
 the bend home into the gun- 
 metal ring. The other end of 
 the bend is likewise secured to 
 the next header (above or be- 
 low), so that each bend is held in place by four bolts against two separate 
 headers. No other packing than this gun-metal ring is used in making 
 the joint. Much importance is attached to the elasticity of this joint, 
 for if one package of tubes expand more than the one above it, the 
 gun-metal rings in the two joints on the connecting-band yield enough 
 to adjust themselves to the difference in expansion without leakage. 
 The return bends, with similar metallic joints, are also used to connect 
 the uppermost front header of each section with the front end of the 
 overhead drum. 
 
 The expansion and contraction of the tubes and drums take place at 
 the front end of this boiler. All the rear headers are locked together 
 by means of plates or wedges, engaging under suitable lugs cast on the 
 outer face of the headers. The front headers, being held by the bends 
 with flexible joints, can yield slightly without leakage, permitting every 
 
3i6 
 
 BOILERS AND FURNACES 
 
 pair of tubes in the boiler and each one of the overhead drums to 
 expand and contract independently. 
 
 A cross steam-pipe lies over the top and at the rear of the overhead 
 drums, being connected with each by means of a nipple expanded into 
 bored holes. From this common steam-pipe two outlets connect with 
 an overhead cross steam-drum located over the centre of the boiler, to 
 which are attached the main stop- and safety-valves. This drum is 
 provided with a drip-pipe leading below the water-line, so that any 
 condensation may drain back into the boiler. 
 
 Fig. 371. 
 
 The steam -drum is enclosed in brickwork and surrounded by an air- 
 space connecting with the combustion-chamber, but far enough removed 
 from the direct path of the gases as to prevent injury to the metal in 
 the drum by overheating. 
 
 From the rear end of each one of the overhead drums is a connec- 
 tion with the rear headers of its corresponding section. In the smaller 
 boilers, having tubes 10, 12, or 15 feet long, the connection is made 
 direct from the drum to the uppermost header, and the undermost rear 
 headers of all the sections are connected with a common mud-drum 
 located at the lowest point of the boiler. In the larger boilers, having 
 tubes 18 feet long, the downtakes from the several overhead drums all 
 lead into one horizontal drum, to which the feed-pipe is attached, and 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 317 
 
 which is connected with the mud-drum set at the lowest part of the 
 boiler by means of two large standpipes ; the mud-drum, as in the case 
 of smaller boilers, is connected with the undermost rear headers of all 
 the sections. The path of the downward circulation in the large boiler 
 is, therefore, from each overhead drum through its connected downtake 
 into the common feed-drum, from this through the standpipes to the 
 mud-drum, and from the latter the water is distributed to the lower 
 headers of all the sections. 
 
 The connections of the mud-drum are all made at the top, so that 
 the contents are not stirred up by the circulation. The object of this 
 arrangement of feed-drum and standpipes is to heat the entering feed- 
 water by mingling with the descending currents from the overhead 
 drums, both in the feed-drum and in the standpipes, the cross-section 
 of which latter in proportion to downtakes is made sufficiently large to 
 retard the velocity of flow and give the feed-water time enough to 
 become heated to the point where it will precipitate most of its lime in 
 the mud-drum before entering the boiler. 
 
 Fig. 372. 
 
 The Heine Boiler belongs to the water-tube variety, as indicated 
 by the illustration, Fig. 372, consisting, as it does, of an assemblage 
 of inclined water-tubes over the furnace, these tubes connecting at front 
 and rear with an overhead water- and steam-drum, the connections cor- 
 responding to headers in ordinary water-tube boilers being water com- 
 
318 
 
 BOILERS AND FURNACES 
 
 partments, commonly known as water-legs, extending across the furnace 
 and into which the tubes are expanded, thus making a continuous water 
 circulation similar to that of the water-tube boilers previously described. 
 The water-legs are of approximately rectangular shape, drawn in at 
 top to fit the curvature of the shells, as shown in Fig. 373. The area 
 of throat or opening at the top of each water-leg is approximately equal 
 
 to the combined area of 
 ^^^- 373- tubes for that leg. Each is 
 
 composed of a head-plate 
 and a tube-sheet, flanged all 
 around and joined at bot- 
 tom and sides by a butt- 
 strap of the same material 
 strongly riveted to both, as 
 shown in Fig. 374. The 
 water-legs are further stayed 
 by hollow stay-bolts of hy- 
 draulic tubing, of large di- 
 ameter, so placed that two 
 stay-bolts support each tube 
 and handhole, also shown in 
 Fig- 374- The water-legs 
 are joined to the shell by 
 flanged and riveted joints 
 and the drum is cut away at 
 these two points to make 
 connection with the inside 
 of water-leg, the opening 
 thus made being strength- 
 ened by bridges and special stays, so as to preserve the original strength, 
 the details of which are shown in Fig. 375. The tubes extend through 
 the tube-sheets, into which they are expanded with roller expanders. 
 Opposite the end of each, in the head-plates, is placed a handhole of 
 slightly larger diameter than the tube and through which it can be with- 
 drawn. These handholes are closed by small cast-iron plates shown in 
 Fig. 374- 
 
 The water- and steam-drum is a cylinder with heads dished to form 
 part of a true sphere and requiring no stays. Both the cylinder and its 
 spherical heads are free to follow their natural lines of expansion when 
 put under pressure. To the bottom of the front head a flange is riveted, 
 into which the feed-pipe is screwed. This pipe is shown in Fig. 376 with 
 an angle-valve and check-valve attached. 
 
 On top of the drum, near the front end, is riveted a steam-nozzle, 
 to which is bolted a T-fitting ; this fitting carries the steam-valve on its 
 horizontal branch, the safety-valve being placed on top. Just under 
 
 0°0°0°0°0<''^' 
 
 ogogogog4| ' 
 
 odddo 
 
 ©==^ 
 
SECTIONAL AND WATER-TUBE BOILERS 
 Fig. 374. 
 
 319 
 
 the steam-nozzle is placed a dry-pipe, and underneath that an inclined 
 deflecting-plate which extends from the front head of the drum to some 
 distance beyond the mouth or throat of the front water-leg. The rear 
 
 Fig. 375. 
 
 B 
 
 
 
 fi 
 
 
 
 
 B 
 
 
 
 head carries a blow-off-flange of about the same size as the feed-flange 
 and a manhead curved to fit the head, the manhole supported by a 
 strengthening ring on the outside. On each side of the drum a tile-bar 
 
320 
 
 BOILERS AND FURNACES 
 
 ^- 
 
SECTIONAL AND WATER-TUBE BOILERS 32 1 
 
 rests loosely in flat hooks riveted to the drum. This bar supports the 
 side tiles, whose other ends rest on the side walls, thus closing in the 
 furnace on top. The top of the tile-bar is 2 inches below low-water line. 
 The bars rise from front to rear at the rate of i inch in 12. When the 
 boiler is set the tile-bars must be exactly level, the whole boiler being 
 then on an incline, — i.e.^ with a fall of i inch in 12 from front to rear ; 
 this makes the height of the steam-space in front about two-thirds the 
 diameter of the drum, while at the rear the water occupies two-thirds 
 of the drum, the whole contents of the drum being equally divided 
 between steam and water. 
 
 The mud-drum is located inside of the water- and steam-drum ; it is 
 placed well below the water-line, usually parallel to and 3 inches above 
 the bottom of the shell. See Fig. 376. It is thus completely immersed 
 in the hottest water in the boiler. It is of oval section, as shown in the 
 engraving at the top of Fig. 376, and slightly smaller than the manhole ; 
 it is made of strong sheet iron' with cast-iron heads ; it is entirely closed 
 except about 18 inches of its upper portion at the forward end, which is 
 cut away nearly parallel to the water-line. The feed-pipe enters the 
 mud-drum through a loose joint in front ; the blow-off-pipe is screwed 
 tightly into its rear head and passes by a steam-tight joint through the 
 rear end of the main drum. 
 
 In setting this boiler the front leg is placed firmly on a set of cast-iron 
 columns bolted and braced together by the door-frames, dead-plates, 
 etc. , and forming the fire-front. This is the fixed end. The rear water- 
 leg rests on rollers which are free to move on cast-iron plates firmly set 
 in the masonry of the low and solid rear wall. Wherever the brickwork 
 closes in to the boiler broad joints are left which are filled in with tow or 
 waste saturated with fire-clay or other refractory but pliable material. 
 Thus the boiler and its walls are each free to move separately during 
 expansion or contraction without loosening the joints in the masonry. 
 
 Fig. 377. 
 
 Light fire-brick tiles, shown in Fig. 377, are placed between the 
 upper tubes. The lower tier extends from the front water-leg to within 
 a few feet of the rear one, leaving there an upward passage across the 
 rear ends of the tubes for the flame, etc. The upper tier closes in to the 
 rear water-leg and extends forward to within a few feet of the front one, 
 
322 
 
 BOILERS AND FURNACES 
 
 thus leaving the opening for the gases in front, as shown in Fig. 372. 
 The side tiles extend from side of walls to tile-bars on the drum and 
 close up to the front water-leg and front wall, and leave open the final 
 uptake for the waste gases over the back part of the shell, also shown 
 in Fig. 372. The rear wall of the setting and one parallel to it are 
 arched over the shell a few feet forward to form the uptakes ; on these 
 and the rear portion of the side walls is placed a light sheet-iron hood, 
 from which the breeching leads to the chimney. When an iron stack 
 is used this hood is stiffened by L- and T-irons, so that it becomes a 
 truss, carrying the weight of such stack and distributing it to the side 
 walls, as shown in Fig. 372. 
 
 Fig. 378. 
 
 The Stirling Boiler as now made consists of three upper or steam- 
 drums and one lower or mud-drum, all connected together by means of 
 tubes, which are bent slightly so as to allow them to enter the drums 
 radially, as shown in Fig. 378, and in sectional and half-front elevation 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 323 
 
 in Fig. 379. All of the upper or steam-drums are connected by steam- 
 circulating tubes, but the front and middle drums only are connected by 
 water-circulating tubes. The tubes used are 3JE^-inch lap- welded mild 
 steel ; the drums are also made of mild steel. These drums and tubes 
 form the boiler proper, no cast metal entering into its construction. 
 
 Fig. 379. 
 
 The front and middle bank of tubes, through which rapid circula- 
 tion takes place, receive their supply of water from the mud-drum. 
 Fig. 380 is an ideal representation of the ascending and descending cur- 
 rents induced by the action of the furnace. The tubes, D, receiving heat 
 sufficient to vaporize the contained water, cause an upward movement ot 
 the water into the drum, B, having water communication with the adjoin- 
 ing drum, A, and it by descending pipes with the mud-drum, C, at the 
 rear of the bridge-wall. A circulation once established is maintained so 
 long as fire is continued in the furnace. 
 
324 
 
 BOILERS AND FURNACES 
 
 The feed-water enters at the top of the rear upper drum. This being 
 the coolest part of the boiler, the temperature of the feed-water is gradu- 
 ally brought to the steaming-point in its descent through the rear bank 
 of tubes to the mud-drum below, these being surrounded by the hot 
 gases escaping to the chimney. 
 
 The mud-drum, protected from the fierce heat of the furnace by an 
 ample bridge-wall, acts as a settling-chamber, the circulation in that por- 
 
 FiG. 380. 
 
 tion of the boiler being comparatively 
 slight. Reaching the mud-drum, the 
 impurities in the feed-water descend to 
 the bottom in the form of sludge or mud, 
 which can be blown out as often as may 
 be found necessary. Whatever solid 
 impurities or precipitate adhere to the 
 interior surfaces of the rear bank of tubes 
 must be washed off with a hose or re- 
 moved with a scraper. 
 
 The interior surfaces of this boiler are 
 rendered accessible by the removal of 
 four manhole plates, which exposes to 
 view the two ends of every tube in the 
 boiler and, for all practical purposes, the entire area of the interior heat- 
 ing surface. The boiler attendant can then enter the drums and remove 
 the mud and other deposits, scraping the tubes from end to end if need 
 be with a chain or jointed scraper adapted for that purpose. 
 
 This boiler is erected entirely independent of the brickwork, so that 
 the latter may be removed or replaced without disturbing the boiler or 
 its connections. The three upper or steam-drums are supported by 
 wrought-iron beams resting on wrought-iron columns, with cast-iron 
 bases properly secured, whilst the mud-drum is suspended and left free 
 to allow for contraction and expansion. 
 
 The Hogan Boiler, shown in Fig. 381, is made up of a water-and- 
 steam-drum at the top, a distributing-drum at the bottom, and a series 
 of bent tubes connecting these two drums, the shape of the tubes being 
 such that the heat will not injuriously strain the expanded joints or dis- 
 tort the form of the tubes. There are no headers or stayed tube-plates 
 to prevent the free expansion of all tubes. 
 
 The circulation begins in the upward movement of the water over 
 the fire and into the water-and-steam-drum at the top of the furnace, 
 the water returning through the rear tubes, marked ' ' water-heating 
 tubes" and "circulating-tubes," both series of which are connected 
 with the upper and lower drums as shown. 
 
 It will be observed that the steaming-tubes, which are 2 inches in 
 diameter, deliver the steam formed in them above the water-level of the 
 upper drum. A mechanical extractor relieves the steam from any en- 
 
SECTIONAL AND WATER-TUBE BOILERS 
 Fig. 381. 
 
 325 
 
 ^s^^^^^^^^^^^^^i^^&^i^i^:^^;^^ 
 
326 BOILERS AND FURNACES 
 
 trained water, which falls back into the drum and passes down the 3-inch 
 circulating-tubes to the distributing-drum below. The extent of water 
 surface from which steam has to escape as it is produced in ordinary- 
 boilers is the same under all conditions of firing. The effect of this 
 limitation of water surface is great agitation of the water, which takes 
 place under moderate or forced conditions of firing, as indicated by the 
 motion of water in the gauge-glass and by the wet condition of the steam. 
 
 The feed-water enters the inductors in the upper drum and passes 
 thence to the circulating-tubes, where precipitation takes place below 
 the water-level and at a temperature within a few degrees of that of 
 the steam. The circulating-tubes are not exposed to the heat of the 
 gases. If the foreign substances and sediments held in suspension in 
 water are not allowed to precipitate or come in contact with the sur- 
 faces in steam boilers which are exposed to the fire and the gases no 
 scale will form on these surfaces. It is claimed for this boiler that the 
 precipitation occurring in the distributing-drum, which is not heated, 
 passes to the mud-drum, which is an external vessel and not exposed to 
 any heat, where it may be blown out at intervals. 
 
 The heating- tubes in this boiler are not claimed to be self-cleaning ; 
 on the contrary, the makers state that no water-tube in a steam boiler, 
 be it straight or bent, can be self-cleaning ; but what is claimed for it 
 is, that the precipitation of the scale-forming substances is so nearly 
 accomplished by the methods here shown that deposits of scale do not 
 occur. A soft deposit has been found in the distributing- and mud- 
 drums, but no hard scale has been brought into existence. 
 
 The Hazelton Boiler, shown in Fig. 382, has a central vertical 
 cylinder or standpipe which varies in diameter and height according to 
 the power required ; it rests upon a circular cast-iron foundation-plate, 
 placed upon a supplementary foundation of brick, raised one course 
 above level of foundation, so as to prevent any water in the ash-pit 
 from coming in contact with the boiler ; nor is it fastened to the founda- 
 tion, but left free to expand and contract according to variations of 
 temperature. That portion of the stand-pipe below the grate-bars forms 
 the mud-drum, into which a manhole is placed for entering the boiler 
 and which permits access to every portion of its interior surface. 
 
 Radial tubes extend outwardly from the standpipe at regular inter- 
 vals, as shown in Fig. 383 ; the diameter, length, and number depend 
 upon the size of the boiler. The outside end of each tube is closed upon 
 itself by welding, forming a hemispherical end ; in the process of closing 
 this end is slightly thickened. The open end of the tube is expanded 
 into the standpipe. The tube, extending outward horizontally and 
 being secured at one end only, can expand and contract without strain. 
 
 The steam-pipe and steam-drying tubes are located at the top of the 
 boiler, as shown in Fig. 382. A wrought-iron flange is riveted to the 
 under side of the top head of standpipe, and abundant threadhold secured 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 327 
 
 by cutting a thread 
 through both head and 
 flange. A heavy nip- 
 ple is screwed in from 
 the outside, the lower 
 end with a long thread 
 extending below the 
 head two or three 
 inches. At the upper 
 end of this nipple is a 
 spherical tee, from one 
 outlet of which is the 
 steam-pipe extending 
 horizontally to the 
 outer line of brickwork 
 to receive steam-valve ; 
 similarly, another pipe 
 extends in the opposite 
 direction, which con- 
 nects with the pop 
 safety-valve. At the 
 lower end of this cen- 
 tral nipple and inside 
 of the standpipe, con- 
 nected with it by a pair 
 of flanges bolted to- 
 gether, is a vertical 
 pipe open at the upper 
 and closed at the lower 
 end ; into this vertical 
 pipe are screwed a 
 series of wrought-iron 
 pipes of small diameter 
 open at both ends ; 
 these extend radially al- 
 most to the outer end of 
 the uppermost steam- 
 drying tubes in the 
 standpipe, as shown in 
 Fig. 382. The steam 
 as it becomes disen- 
 gaged from the surface 
 of the water must enter 
 the steam-drying tubes ■ 
 and pass to their outer 
 
 Fig. 
 
328 
 
 BOILERS AND FURNACES 
 
 ends before entering these small pipes, which convey the dry steam to 
 
 the steam-pipe. 
 
 The feed-water pipe is led beneath the grate-bars and screws into the 
 
 standpipe, passing through 
 which it is then carried up- 
 ward to a short distance 
 below the water-line, thence 
 downward to about the grate 
 level, where it discharges 
 into the lower part of the 
 standpipe, that portion 
 having little or no circula- 
 tion and where precipita- 
 tion is most likely to occur, 
 hence called the mud-drum. 
 The blow-off-pipe is sim- 
 ilarly connected to the 
 standpipe, but the exten- 
 sion in this case is down- 
 ward to within a short dis- 
 tance of the bottom head. 
 A plan of the furnace and grate-bars is shown in Fig. 384, the grate 
 
 area showing the herring-bone pattern, this being the manufacturer's 
 
 preferred style of grate. 
 
 The shaded portion back Fig. 384. 
 
 of the standpipe represents 
 
 a bridge-wall, and is not 
 
 counted in the grate area, 
 
 but any portion of this 
 
 could be made available for 
 
 grates if necessary. That 
 
 portion of the brickwork 
 
 lining of the square furnace 
 
 casing, from the floor level 
 
 of the ash-pit to a point a 
 
 little above the top of the 
 
 grates, is built of uniform 
 
 thickness on the four sides, 
 
 and by means of corbelling, 
 
 shown in Fig. 382, the 
 
 brickwork is brought in so 
 
 as to form a circular deflect- 
 ing furnace. The furnace casing is reinforced by angle-irons and braces, 
 
 and at its top supports a circular metal shelf, upon which rests the upper 
 
 brick lining inside of the circular upper jacket of the boiler. A space of 
 
SECTIONAL AND WATER-TUBE BOILERS 
 
 329 
 
 several inches is left between the top of the deflecting furnace and this 
 metal shelf for expansion of furnace and to remove the weight of the 
 superstructure from the fur- 
 
 nace. 
 
 The smoke-hood has a 
 door placed in it for ready- 
 access to exterior of upper 
 partofstandpipe. The steam 
 and pop safety-valve pipes 
 extend outward through 
 openings in this hood, and 
 are covered with movable 
 sUdes, permitting the stand- 
 pipe to expand and con- 
 tract without straining the 
 joints. 
 
 The Adams Boiler, 
 shown in Fig. 385, has some 
 points of similarity to the 
 boiler previously described, 
 yet is widely divergent in 
 its details. Water - tube 
 boilers having radially pro- 
 jecting tubes closed at their 
 outer ends have many 
 stanch advocates, who as- 
 sert that such boilers will 
 evaporate more water per 
 pound of coal and raise 
 steam more rapidly than 
 boilers of any other type. 
 The two principal objec- 
 tions raised against boilers 
 of this type have been : 
 first, that on account of 
 the small settling-chamber, 
 where water containing im- 
 purities in large degree is 
 used, the tubes are liable to 
 fill up, requiring constant 
 cleaning ; second, when 
 crowded beyond their rated 
 capacity they are very 
 liable to give wet steam. 
 This boiler has been de- 
 
 FiG. 385. 
 
330 
 
 BOILERS AND FURNACES 
 
 signed to overcome these difficulties not only, but to embody such other 
 necessary requisites as shall make it a durable and economical steam 
 generator. 
 
 A feed-water reservoir is formed by extending the centre tube up 
 into the steam-dome, as shown. This is practically a live-steam heater 
 and purifier. As the feed-water falls into the reservoir it is raised to the 
 same temperature as the steam, and the impurities settle in the form of 
 sediment to the bottom of the reservoir, whence they can easily be re- 
 moved. From the reservoir the hot 
 ^ ' water passes slowly through a large 
 
 down-pipe or outside connection 
 not shown in the engraving from 
 the extension of the upper dome 
 to the extension from the bottom 
 of the centre tube. Should any of 
 the impurities held in suspension 
 flow over from the reservoir, they 
 will be retained and settle in the 
 bottom extension formed at the 
 bottom of the centre tube, thus 
 leaving the inside of the steam-tubes 
 clean and free from scale. To pro- 
 vide against the possibility of im- 
 perfect separation of the steam from the water, a large dome is placed on 
 top of the centre tube above the water-line ; this secures ample room 
 for separation and a large storage for steam, insuring a steam-supply 
 free from saturation and to some extent superheated. Manholes at 
 the outer ends of the two extensions allow access to both top and 
 bottom of the boiler. Fig. 386 shows a plan of the furnace. 
 
 The Morin Boiler, shown in Fig. 387, has a vertical central stand- 
 pipe with bumped heads. The lower end of the bottom sheet extends 
 so far below the head that the latter carries no weight and is above con- 
 tact with the foundation. The weight of the boiler and its contents is 
 borne by a heavy cast-iron ring riveted to the projecting sheet below 
 the bottom head. This ring has a broad base resting upon the brick 
 foundation underneath. 
 
 The lower part of the vertical shell, having little or no circulation, 
 acts as a mud-drum, in which foreign substances, held in solution and 
 precipitated by the action of the heat, accumulate and may be blown 
 out as required. 
 
 The construction of the cylindrical shell into which the tubes are 
 inserted is the same as that of any boiler-shell, with the exception that 
 it is welded longitudinally instead of riveted, which allows a uniform 
 spacing of the holes for the tubes. 
 
 The tubes are from i^ to 3 inches in diameter, depending upon the 
 
SECTIONAL AND WATER-TUBE BOILERS 
 Fig. 387. 
 
 331 
 
332 
 
 BOILERS AND FURNACES 
 
 size of the boiler. They are bent to the shape indicated in Fig. 388, 
 in addition to which they also have a vertical twist, so that the bent 
 tube re-enters the cylinder at about 18 inches above the lower entrance. 
 The general appearance of a series of tubes when put together in the 
 boiler is that of a number of spirals forming annular rings in series one 
 above the other to such a height as may be required for a given boiler. 
 The tubes must be identical in their shape to fit into place in the series 
 to which each belongs. To insure exact duplication they are bent over 
 formers, giving each tube the precise shape needed for the boiler for 
 which it is intended. 
 
 Fig. 388. 
 
 The tubes, being expanded in the main shell, allow for unequal ex- 
 pansion and contraction, while the bends provide for unequal expansion 
 in the individual tubes. The expanded joints are well protected from 
 the action of the fire, being for the most part under water ; consequently 
 they are not only well protected, but any tube can be easily replaced, 
 and at small expense. The lower tubes and the central shell are filled 
 with water, as shown in Fig. 387. The heat circulates among the tubes, 
 around the central shell, and thence to the chimney-flue. The upper 
 tubes and the central shell above the water-line act as superheating 
 surfaces. 
 
SECTIONAL AND WATER-TUBE BOILERS 333 
 
 A manhole is placed in the upper head, by which access is had to 
 the interior of the boiler for examination or repairs. A handhole-plate 
 is located in the mud-drum for convenience in cleaning. 
 
 A deflector-plate is inserted in the central shell a short distance 
 above the water-level, which tends to throw back any water that may 
 be carried up by the steam ; and a series of diaphragms divide the upper 
 portion of this cyhnder, forming a series of superheating-chambers, 
 through which the steam is successively compelled to circulate by the 
 connecting bent tubes, drying the steam and slightly superheating it. 
 The steam then passes into a reservoir provided for it above the top 
 row of tubes. 
 
 A water-heating coil or economizer of i^- to 3-inch pipe is located 
 above the central shell, resting on the top row of tubes ; the length may 
 be from 100 to 300 feet, according to the size of the boiler. This coil 
 is welded and without screw-joints. The feed-water enters and flows 
 through this coil of pipe before it enters the boiler, absorbing consider- 
 able heat from the waste gases, and then passes to the central shell, 
 where it is taken up by the curved tubes, in which a rapid circulation 
 is maintained by reason of their upward course and the intense heat by 
 which they are surrounded. 
 
 The fire-box surrounds the central vertical cylinder, and is, therefore, 
 annular in form. It is enclosed in a casing of iron, bolted together in 
 sections, and lined with fire-brick. Ordinarily three or four fire-doors 
 are provided, according to the size of the boiler. 
 
 The outside casing of the boiler is sectional and bolted together ; it 
 can be easily removed, wholly or in part, should it be necessary to 
 remove a defective or substitute a new tube. Each section of the out- 
 side casing of the boiler is provided with a cleaning-door, through which 
 the ashes and soot may be blown off" the tubes by means of a steam- 
 hose. 
 
 In erecting this boiler the central shell is stood upon the cast-iron 
 base-plate secured to the foundation, after which the work of placing 
 the tubes in the shell is commenced. After all the tubes have been ex- 
 panded, the whole is subjected to a hydrostatic test of 300 pounds per 
 square inch, replacing any defective tubes ; this completed, the castings, 
 which have been lined with fire-brick, are then placed in position and 
 securely bolted. The smoke-hood is then placed and the usual boiler 
 attachments made, until every part is in proper position. 
 
 The Cahall Boiler, shown in sectional elevation in Fig. 389, con- 
 sists of two steel drums arranged one above the other, and connected 
 with 4-inch lap-welded tubes. These tubes are vertical, are perfectly 
 straight throughout their entire length, and are expanded into the drums 
 at each end. The upper or steam-drum has an opening through its 
 centre for the exit of waste gases. The water-line in the upper drum is 
 about a foot above the bottom of the drum, the drum itself being about 
 
334 
 
 BOILERS AND FURNACES 
 
 Fig. 389. 
 
 Blwqffc 
 
 §W^msm^^mmm!n^ms!mm^^mmMMm 
 
SECTIONAL AND WATER-TUBE BOILERS 335 
 
 6 feet high, leaving a space of 5 feet between the surface of water and 
 the point at which the steam is drawn off from the boilers. A circu- 
 lating-pipe is shown at one side of the engraving ; this is attached to the 
 upper or steam-drum, just below the water-level, and is carried down- 
 ward, outside the brickwork, to a point just below the tube-sheet of the 
 lower drum, where it enters that drum. Through some oversight these 
 connections were not shown in the engraving. There is no steam in 
 this external circulating-pipe and no possibility of making any, but the 
 tubes connecting the two drums are steam making tubes, consequently 
 the water in the external pipe has a greater specific gravity than the mix- 
 ture of steam and water in the steaming-tubes, and a rapid and positive 
 circulation is thus set up. The water in the tubes connecting the drums, 
 ascending to the steam-drum, delivers the mixture of water and steam 
 there ; the steam separating at once from the water, the latter enters the 
 circulating-pipe and is carried down to the mud-drum, and again rises 
 with its mixture of steam. As this mixture of steam and water coming 
 from the upper end of the tube is about half steam and half water in 
 bulk, and as steam at 100 pounds pressure will occupy about 218 times 
 the space occupied by the water itself, the water in the boiler will circu- 
 late through the boiler 218 times before finally becoming steam. This 
 insures not only a rapid and steady circulation, but also insures a uni- 
 form temperature of water in all the tubes, as fresh feed-water, being 
 thus circulated 218 times before evaporation, must necessarily mingle in 
 such minute parts with the water already present in the boiler that the 
 water in one ascending tube cannot be different in temperature to that 
 in the others. The boiler is thus relieved from any possibility of de- 
 structive strains from unequal expansion. 
 
 A central opening is provided in the upper drum through which the 
 gases escape to the chimney ; the upper tube-sheet has, therefore, a 
 circular opening in its centre, leaving a central open space between the 
 tubes, which gradually narrows to the bottom tube-sheet. Advantage 
 is taken of this space, which is in the form of an inverted cone, to intro- 
 duce deflecting-plates, which cause the gases to be alternately thrown 
 out and in throughout the whole heating surface, giving them a sweep 
 at nearly right angles to the tubes, thereby extracting from these gases 
 their heat until they come to very nearly the temperature of the water 
 contained in the boiler. 
 
 Manholes are provided in both the upper and lower drums, to which 
 are fitted the swinging manheads shown in Fig. 170. By simply taking 
 off the nuts from the manheads and swinging them open, the attendant 
 can place a light in the lower drum of the boiler and get into the upper 
 drum and examine the condition of the tubes and that of the interior 
 of the boiler. In case scale is discovered in any of the tubes, he can 
 run a scraper through them, for which purpose the one in use is made 
 in sections a trifle less than 6 feet long. Four of these sections are used, 
 
336 
 
 BOILERS AND FURNACES 
 
 and the man who is cleaning the boiler takes them into the upper drum 
 and pushes the first section down as far as it will go, then simply hooks 
 the second section to that, and continues doing this until the scraper 
 has gone entirely through the tube, forcing any scale that may have 
 been deposited on the sides of the tube straight through to the bottom 
 drum, where it can be removed through the bottom manhole. 
 
 The boiler rests upon four iron brackets riveted to the lower or mud- 
 drum, supported upon four piers of the foundation, the entire structure 
 standing without contact with the brickwork, thus allowing the boiler 
 every freedom for expansion without in any way straining the brick 
 setting. In all places where pipe connections are made to the boilers 
 through the walls they are encased in expansion-boxes. 
 
 Defective tubes may be removed from the boiler in the following 
 manner : In the upper or top head of the steam-drum there are placed 
 six handholes, not shown in the engraving, which are closed by means 
 of plate, yoke, and bolt in the usual manner ; there are in addition two 
 other holes, — one for the steam-pipe connection, the other for the pop 
 safety-valve connections. By means of these eight openings any of the 
 tubes needing removal, after having been cut loose from the tube-sheets, 
 
 can be pushed up through the 
 Fig. 390. tube-hole from which it has just 
 
 been cut and through the most 
 convenient of these openings in 
 the top head and removed from 
 the boiler. The new tube to re- 
 place the defective one is passed 
 into the boiler through the same 
 openings. 
 
 The furnace is external, as 
 shown in Fig. 389. The com- 
 bustion-chamber is roofed with 
 a heavy fire-brick arch, which 
 becomes incandescent shortly 
 after the boiler is fired and radi- 
 ates its contained heat directly on 
 top of the green coal ; further- 
 more, owing to the direct upward 
 passage of all gases and iiill free 
 openings, a comparatively short 
 stack will furnish a draught press- 
 ure not usually had with most 
 other boilers ; for instance, it is 
 claimed that in tests with a stack only 50 feet high a draught pressure 
 in the furnace of over ^ inch was attained, but the temperature of the 
 escaping gases is not furnished. This is a result which could hardly be 
 
SECTIONAL AND WATER-TUBE BOILERS 337 
 
 expected from any other water-tube boiler with a stack loo feet high. 
 The heavy draught causes a rapid combustion of fuel per square foot of 
 grate, with the consequent high initial temperature of gases, both of 
 
 Fig. 391. 
 
 which are requisite to either efficiency or economy in boiler practice. 
 A plan of the furnace is shown in Fig. 390 ; a half-front and half-sectional 
 elevation through the arch of the furnace is shown in Fig. 391, 
 
 BOILER PERFORMANCE. 
 
 ADDITIONAL TO CHAPTERS VI, VIII, IX. 
 
 It was no part of the original scheme to enter upon questions 
 relating to boiler performance, whether for economy, capacity, or dura- 
 bility, but rather to confine the subject-matter of this book to the details 
 of construction and arrangement of parts. As the work progressed, 
 however, it became more and more apparent that a table of comparative 
 evaporative tests would prove interesting even though it had no special 
 value. Table XLIX. is not, strictly speaking, comparative, but is 
 simply an aggregation of boiler performances arranged in the order 
 in which the several types of boilers are considered in the preceding 
 chapters, all of which are in use in this country. The tests from which 
 this table was compiled varied from eight to twenty-four hours in 
 
__../^s 
 
 338 BOILERS AND FURNACES 
 
 duration ; but to form a better basis of comparison, all the tests were 
 reduced to a common basis of ten hours. 
 
 As nearly all of the tabular numbers have been changed from the 
 final sheets prepared by the experts who conducted the tests, it did not 
 seem fair to use their names in connection with the tests without pre- 
 senting the figures as originally prepared. We will say, however, that 
 the experts thus quoted are all well known and recognized as being 
 thoroughly competent for conducting such expert trials. 
 
 From the mass of material at our disposal, only such boilers were 
 selected as seemed to meet the ordinary conditions of service. Ex- 
 traordinary results, above or below what is recognized as good per- 
 formance, have not been used, except in the case of one or two boilers 
 which were overfired, as indicated by the high temperature of the 
 escaping gases. One object in the preparation of this table was to 
 present as nearly as possible the conditions under which the several 
 steam boilers are operated in actual service. 
 
 A, Cylinder Boilers. — Number of shells, three ; each shell 36 inches 
 in diameter by 30 feet in length ; area of water-heating surface, 518 
 square feet ; area of grate surface, 47.5 square feet ; ratio of water- 
 heating surface to grate surface, 10,9 to i. Arrangement of furnace 
 and setting similar to Figs. 209 to 212. Fuel, anthracite pea coal; 
 moisture, 3.2 per cent. 
 
 B, Two-Flue Boilers. — Number of shells, three ; each shell 48 
 inches in diameter by 36 feet in length ; two 16-inch flues in each shell ; 
 area of water -heating surface, 1225 square feet ; area of grate surface, 
 90 square feet ; ratio of water-heating surface to grate surface, 13.6 to 
 I. Arrangement of furnace and setting similar to Fig. 217. Fuel, 
 bituminous coal, Pittsburgh run of mine. 
 
 C. Three- Inch Tubular Boiler. — Number of boilers, one ; shell, 
 48 inches in diameter by 16 feet in length, with 50 tubes 3 inches in 
 diameter ; area of water-heating surface, 763 square feet ; area of grate 
 surface, 25 square feet ; ratio of water-heating surface to grate surface, 
 30.52 to I. Arrangement of furnace and setting similar to Fig. 293. 
 Fuel, anthracite pea coal ; moisture, 3.5 per cent. 
 
 D. Three-and- One- Half - Inch Tubular Boilers. — Number of boilers, 
 two ; each shell 60 inches in diameter by 16 feet in length, with 62 
 tubes 3^ inches in diameter in each boiler, or 124 in all ; area of water- 
 heating surface, 2100 square feet ; area of grate surface, 52 square feet ; 
 ratio of water-heating surface to grate surface, 40.38 to i. Arrangement 
 of furnace and setting similar to Fig. 293. Fuel, bituminous coal, 
 Pittsburgh run of mine. 
 
SECTIONAL AND WATER-TUBE BOILERS 339 
 
 E. Four- Inch Tiibular Boilers. — Number of boilers, two; each shell 
 72 inches in diameter by 20 feet in length, with 68 tubes 4 inches in 
 diameter in each boiler, or 136 in all ; area of water-heating surface, 
 3385 square feet; area of grate surface, 62.24 square feet; ratio of 
 water-heating surface to grate surface, 54.39 to i. Arrangement of fur- 
 nace and setting similar to Fig. 293. Fuel, bituminous coal, Mount 
 Olive lump. 
 
 F. Six- Inch Tubular Boilers. — Number of boilers, two; each shell 
 60 inches in diameter by 24 feet in length, with 18 lap-welded tubes 
 6 inches in diameter in each boiler, or 36 in all ; area of water-heating 
 surface, 1879 square feet ; area of grate surface, 58 square feet ; ratio ol 
 water-heating to grate surface, 32.4 to i. Arrangement of furnace and 
 setting similar to Fig. 294. Fuel, bituminous coal, small lump. 
 
 G. Double-Deck Horizontal Tubular Boiler. — Number of boilers, 
 one, of a series of four ; diameter of boiler-shells, 60 inches and 50 
 inches ; length of boiler-shell, 15 feet ; number of 4-inch tubes in each 
 boiler, 60 ; area of water-heating surface, 1286 square feet each ; area ol 
 steam-heating surface, 141 square feet each ; area of grate surface, each 
 boiler, 24 square feet ; ratio of grate to water-heating surface, 53.6 to i. 
 Arrangement of furnace and setting similar to Fig. 234. Fuel, anthracite 
 coal, rice, 3 per cent, moisture, Wilkinson stoker, force-blast. 
 
 H. Triplex Boiler (see Fig. 236). — The two lower shells are 58 
 inches in diameter by 16 feet 9 inches long, with 62 4-inch tubes in each 
 shell. The upper drum is 48 inches in diameter by 16 feet in length. 
 The connecting necks are 15 inches in diameter. In the test here given 
 the Hawley down-draft grate was used. The original design was for a 
 standard grate which would give data for the test as follows : area of the 
 standard grate, 11x5 feet, 55 square feet ; total area of heating surface, 
 2705 square feet ; ratio of heating surface to grate area, 49.2 to i ; coal 
 consumed per hour, 893.8 pounds ; coal per hour per square foot ot 
 standard grate, 16.25 pounds ; capacity of boiler, as stated in the con- 
 tract, 250 horse-power; capacity of boiler averaged during test 277.1 
 horse-power (excess 11 per cent.) ; capacity of boiler in preliminary test, 
 315.2 horse-power (excess 26 per cent.). 
 
 The lump coal when weighed contained 1.6 per cent, of moisture. 
 
 The total heat of combustion by calorimeter per pound of dry coal 
 in B.T.U. was 12,765. 
 
 Efficiency, 80.6. 
 
 I. Fire-Box Tubular Boiler. — Trial 48 hours ; test records results 
 of one boiler out of a battery of five boilers working together. Grate 
 surface, 54^ square feet each ; water-heating surface, 1562 square feet 
 each ; ratio of water-heating to grate surface, 28.7 to i. Arrangement 
 
340 BOILERS AND FURNACES 
 
 of boiler details similar to Fig. 331. Fuel, anthracite coal, bird's eye ; 
 7.4 per cent, moisture. 
 
 K. Belpaire Boiler, Double Furnace. — Two boilers in use ; record 
 of one boiler. Diameter of shell, smallest inside, 82 inches ; inside 
 length of fire-box, 8 feet i)^ inches ; inside width of each fire-box (2), 
 7 feet 6 inches ; length of combustion-chamber, 7 feet ; length of tubes, 
 16 feet; diameter of tubes, outside, 3 inches; number of tubes, 159; 
 length of grate, 7 feet i inch ; width of grate during trial bricked up to 
 5 feet 3 inches ; heating surface, 2240 square feet ; grate surface during 
 trial, each boiler, 37.185 square feet; ratio of heating to grate surface 
 during trial, 60.24 to i. Pocahontas coal ; average moisture, 2.60 per 
 cent. ; calorific value of i pound of coal by analysis, 14,924 B.T.U,; 
 design of boiler approximates Fig. 337. 
 
 L. Gun-Boat Boiler (Fig. 326). — Results of one boiler in a battery of 
 two boilers working together. Diameter of boiler, 8 feet 6 inches ; length 
 of boiler, 20 feet ; furnaces in each boiler, 2 ; type of furnace. Fox cor- 
 rugated ; diameter of furnaces, 3 feet 6 inches ; length of furnaces, 7 
 feet 6 inches ; number of 4-inch tubes in each boiler, 90 ; length of tubes, 
 10 feet; heating surface, i boiler, 11 19 square feet; heating surfaces, 
 2 furnaces, 136 square feet ; heating surface, combustion-chamber, 47 
 square feet ; heating surface of tubes, 936 square feet ; area of cross- 
 section of tubes, I boiler, 8 square feet ; ratio of tube cross-section to 
 grate area, 5.25 to i ; grate surface, i boiler, 42 square feet; ratio of 
 water-heating to grate area, 26.67 to i. Semi-anthracite pea coal; 9 
 per cent, moisture. Heavy rain during test ; coal was brought in from 
 yard perfectly saturated. 
 
 M. Vertical Tubular Boiler. — One boiler ; shell 48 inches in diameter 
 by 9 feet in height ; 148 tubes, 2 inches in diameter, 72 inches long ; fire- 
 box, 42 inches in diameter by 36 inches in height. Heating surface : fire- 
 box, 35 ; water-tubes, 349 ; superheating, 116 ; head, 7 ; making a total 
 of 505 square feet. Upper end of tubes pass through steam-space for 
 18 inches. Area of grate surface, 9.62 square feet ; ratio water -heating 
 to grate surface, 40.44 to i. Arrangement of boiler details similar to 
 Fig. 303. Fuel, anthracite stove coal. 
 
 N. Manni7ig Vertical Boiler. — Results of one boiler in a battery ot 
 six boilers. Diameter of shell, 5 feet ; diameter of fire-box, 6 feet ; 
 height of fire-box, 3 feet 6 inches ; number of tubes, 180 ; diameter, 
 2^ inches; length of tubes, 15 feet; grate area, 28.7 square feet; 
 water-heating surface, 1383 square feet ; superheating surface, 471 square 
 feet ; ratio of water-heating surface to grate, 48.2 to i ; ratio of super- 
 heating surface to grate, 16.4 to i. Arrangement of boiler details similar 
 to Fig. 310. Fuel, bituminous coal, Pocahontas ; 4.75 per cent, moisture. 
 
SECTIONAL AND WATER-TUBE BOILERS 34 1 
 
 O. Galloway Boiler. — Diameter, 85 inches, by 28 feet in length : 
 number of conical tubes, 33 ; diameter of conical tubes, 5^^ and 10 
 inches ; length of conical tubes, 2 feet 1 1 inches ; diameter of furnace, 
 2 feet 10 inches ; grate, 6 feet 7 inches long by 2 feet 9 inches wide 
 total heating surface, 1058 square feet. Arrangement of boiler and set- 
 ting similar to Fig. 313. Fuel, bituminous coal, Cumberland. 
 
 P, Wharton- Harris 071 Boiler. — The details of this boiler are in all 
 respects as described on page 298 and following. Figs. 352 and 353 
 illustrate the details common to all boilers of this manufacture. 
 
 O. Babcock & Wilcox Boiler. — Boiler drum, 36 inches in diameter 
 by 16 feet 6 inches in length ; grate, 60 inches wide by 72 inches long, 
 30 square feet ; 64 tubes 4 inches in diameter by 16 feet in length ; total 
 water-heating surface, 1253 square feet; ratio of water-heating surface 
 to grate surface, 41.8 to i. Arrangement of furnace and setting similar 
 to Fig. 354. Fuel, anthracite coal. 
 
 R. Caldwell Water- Tube Boiler. — Fig. 366 shows the general ar- 
 rangement of this boiler. It is distinguished from the Babcock & 
 Wilcox boiler in the design and grouping of the headers, a detail of one 
 of which is shown in Fig. 367 ; also a novel design of baffle bricks used, 
 shown in Fig. 368. On a capacity test by G. H. Barrus, additional to 
 the one given in the table, a 206 horse-power boiler developed 65.7 per 
 cent, more than its rated power, and this was done with a draught of 
 0.47 inch in the flue, which corresponds to that of an ordinary good 
 chimney. The actual draught at command was more than double this 
 force, and in this particular instance the excess of capacity could, if 
 desired, have been increased much beyond the amount realized. 
 
 S. Heine Safety Boiler (Fig. 372). — Number of boilers, one ; i shell 
 48 inches in diameter by 19 feet 9 inches in length ; 92 tubes 33^ inches 
 in diameter by 16 feet in length ; grate, 81 inches wide by 48 inches 
 long, 27 square feet ; water-heating surface, 1406 square feet ; ratio of 
 water-heating surface to grate surface, 52 to i. Fuel, bituminous coal. 
 
 T. Stirling Boiler. — Number of boilers, one ; number of water- 
 drums, I ; number of steam- and water-drums, 3 ; diameter of water- 
 tubes, 3j^ inches ; grate surface, 52 square feet ; water-heating surface, 
 2300 square feet ; ratio of water-heating surface to grate surface, 44. i 
 to I. Arrangement of furnace and setting similar to Fig. 378. Fuel, 
 Youghiogheny bituminous coal. 
 
 U. Hogan Boiler (Fig. 381). — Number of boilers in test, one ; diam- 
 eter of steam-drum, 42 inches ; diameter of distributing-drum, 24 inches ; 
 diameter of mud-drum, 10 inches ; number of steaming- and heating- 
 
342 BOILERS AND FURNACES 
 
 tubes, 384 ; number of circulating-tubes, 36 ; grate surface, 80 square 
 feet ; water-heating surface, 4100 square feet ; ratio of water- heating 
 surface to grate surface, 51.25 to i. 
 
 V. Hazelto7i Boiler. — Shell, 42 inches diameter by 34 feet in height ; 
 1070 tubes 4 inches in diameter by 3 feet long ; grate, 129 inches outside 
 by 44^ inches inside diameter, 80 square feet ; water-heating surface, 
 2371 square feet ; superheating surface, 1049 square feet. Arrangement 
 of furnace and setting similar to Fig. 382. Fuel, bituminous coal. 
 
 W. Morin Climax Boiler. — Total heating surface, 10,000 square 
 feet; grate surface, 113.6 square feet; ratio of total heating to grate 
 surface, 88 to i. Arrangement of furnace and setting similar to Fig. 
 387. Fuel, Youghiogheny bituminous coal. 
 
 X. Cahall Water- Tube Boiler. — Number of boilers in test, one ; 
 diameter of upper drum, 6 feet 8 inches ; diameter of internal flue, 34 
 inches ; 108 tubes 4 inches in diameter by 18 feet long ; area of grate 
 surface, 40 square feet ; area of water-heating surface, 2064 square feet ; 
 area superheating surface, 50 square feet ; ratio of water-heating to grate 
 surface, 51.6 to i ; draught area through internal flue, 6.3 square feet, 
 or a trifle less than one-sixth of the grate. Arrangement of furnace and 
 setting similar to Fig. 389. Fuel, Pocahontas bituminous coal ; 3.3 per 
 cent, moisture. 
 
Dimensions and Proportions. 
 
 Grate surface, area in square feet .... 
 
 Water-heating surface, square feet . . . 
 
 Superheating surface, square feet . . . 
 
 Ratio of water-heating surface to grate . 
 Average Pressures. 
 
 Steam pressure by gauge 
 
 Chimney draught in inches of water . . 
 Average Temperatures. 
 
 External air, degrees Fahr 
 
 Escaping gases from boiler, degrees F. | 
 
 Feed-water entering boiler, degrees F. . 
 Fuel. 
 
 Kind and size •! 
 
 Coal fed to furnace, deducting moisture, 
 pounds 
 
 Percentage of ash 
 
 Total combustible, pounds 
 
 Coal burned per square foot of grate 
 surface per hour 
 
 Coal burned per square foot of water- 
 heating surface per hour 
 
 Quality of Steam. 
 
 Moisture in steam, per cent 
 
 Superheated, degrees Fahr 
 
 Water. 
 
 Weight of water fed to boiler, pounds . 
 
 Equivalent water evaporated into dry 
 
 steam from and at 212° Fahr., pounds 
 
 Evaporation. 
 
 Water evaporated per pound of coal 
 under actual conditions 
 
 Equivalent evaporation from and at 
 2X2° Fahr. per pound of coal 
 
 Equivalent evaporation from and at 
 212° Fahr. per pound of combustible . 
 
 Equivalent evaporation from and at 
 212° Fahr. per square foot of water- 
 heating surface per hour, pounds . . . 
 
 Commercial Horse-Power. 
 
 On a basis of 34 J^ pounds of water evap- 
 orated per hour from and at 212° Fahr. 
 
 Square feet of water-heating surface 
 per H.-P 
 
 51-3 
 2300 
 
 44.8 to I 
 
 453 
 39 
 
 Poca- 
 hontas, 
 
 6644 
 7-1 
 6170 
 
 12.9 
 
 .29 
 
 .11 
 
 57,528 
 69,839 
 
 8.658 
 10.511 
 11.314 
 
 3-03 
 
 27 
 1407 
 
 123.3 
 ■65 
 
 90.5 
 644 
 84 
 
 Bitu- 
 minous 
 
 7866 
 31-7 
 
 77,995 
 91,878 
 
 9.12 
 10.74 
 11.60 
 
 6.53 
 
 T. 
 
 52 
 2300 
 
 44.1 to 
 129 
 
 Youghi 
 ogheny 
 
 6929 
 5-5 
 6549 
 
 13-32 
 
 71,784 
 78,574 
 
 10.36 
 "•34 
 
 229 
 10.04 
 
 4100 
 
 None. 
 
 51.25 to I 
 
 18,115.41 
 
 32.14 
 
 12,293.12 
 
 22.52 
 
 •439 
 
 •58 
 
 103,285.14 
 120,245.45 
 
 5^71 
 6.64 
 9^79 
 
 348.53 
 11.76 
 
 2371 -j 
 1049 
 29.63 to 
 
 82.5 
 498 
 91 
 
 Bitu- 
 minous. 
 
 12,821 
 4.21 
 12,281 
 
 16 
 
 io9,9«o 
 127,906 
 
 8.58 
 9^99 
 10.28 
 
 3^74* 
 
 370.8 
 6.39 
 
 113.6 
 Total 
 Surfaces. 
 10,000 
 88 to I 
 
 37 
 501 
 35^3 
 
 Bitu- 
 minous 
 
 36,252 
 3-75 
 33,123 
 
 31^9 
 
 •363 
 
 6.81 
 
 31,433 
 38,034 
 
 8.34 
 10.07 
 
 1058.8 
 9M 
 
 40 
 I 2064 
 
 50 
 51.6 to I 
 
 110.7 
 .14 
 
 Bitu- 
 minous. 
 
 6507 
 6 
 
 16.3 
 •315 
 
 4.4 
 64,772 
 77,108 
 
 9-97 
 11.84 
 12.60 
 
 223.4 
 9.28 
 
TABLE XLIX. 
 
 EVAPORATIVE TESTS OF STEAM BOILERS. 
 
 
 A. 
 
 B. 
 
 c. 
 
 D. 
 
 E. 
 
 F. 
 
 G. 
 
 H. 
 
 ^• 
 
 K. 
 
 L. 
 
 M. 
 
 N. 
 
 0. 
 
 P. 
 
 Q. 
 
 R. 
 
 s. 
 
 T. 
 
 u. 
 
 V. 
 
 w. 
 
 X. 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 3 
 
 -OH c 
 
 m 
 
 k 
 
 i 
 
 £u 
 
 1 
 
 i 
 
 1 
 
 1 
 
 ? 
 
 li 
 
 
 .2 
 
 .1 
 
 s 
 
 1 
 
 s 
 
 .2 
 
 
 1 
 
 .B 
 
 4 
 % 
 
 .11 
 
 P 
 
 
 
 §a5 
 
 II 
 
 
 II 
 
 i 
 
 
 .S 
 1 
 
 1 
 
 1 
 
 If 
 
 1 
 
 ! 
 
 5 
 
 1 
 
 1 
 
 •5 
 
 1 
 
 1 
 1 
 
 
 u 
 
 H 
 
 ci. 
 
 ■^ 
 
 z 
 
 ° 
 
 H 
 
 £ 
 
 m 
 
 
 
 > 
 
 s 
 
 
 
 X 
 
 « 
 
 u 
 
 a 
 
 U) 
 
 X 
 
 K 
 
 1 
 
 (J 
 
 Dimensions and Proportions. 
 
 
 
 
 
 
 
 24 { 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Grate surface, area in square feet .... 
 
 47-S 
 
 90 
 
 25 
 
 52 
 
 62.24 
 
 58 
 
 Total. 
 
 }54-5 
 
 37-185 
 
 42 
 
 9-62 
 
 28.7 
 
 36-25 
 
 46.25 
 
 30 
 
 51-3 
 
 27 
 
 52 
 
 So 
 
 80 
 
 v^-f 
 
 40 
 
 Water-heating surface, square feet . . . 
 
 518 
 
 1225 
 
 763 
 
 2100 
 
 3385 
 
 1879 
 
 1286 
 
 
 1562 
 
 2240 
 
 III9 
 
 389 
 
 I3S3 
 
 960 
 
 1872 
 
 1253 
 
 2300 
 
 1407 
 
 2300 
 
 4100 
 
 237. { 
 
 1049 
 
 29.63 to I 
 
 Tol»l 
 Surfaces. 
 
 }2064 
 
 Superlieating surface, square feet . . . 
 
 
 
 
 
 
 
 141 
 
 53.6 to 1 
 
 
 
 
 None. 
 
 116 
 
 471 
 
 48.2 to I 
 
 199 
 
 26.48 to I 
 
 585 
 
 
 
 
 
 None. 
 51-25 to I 
 
 
 Ratio of water-heating surface to grate . 
 
 10.9 to I 
 
 13.6 to I 
 
 30.52 to I 
 
 40.38 to I 
 
 54-39 to I 
 
 32.4 to I 
 
 29.6 to I 
 
 28.7 to I 
 
 60.24 to I 
 
 26.67 to 1 
 
 40.44 to I 
 
 41.8 to I 
 
 44-8 to I 
 
 52 to I 
 
 44.1 to I 
 
 10,000 
 
 50 
 51.6 to I 
 
 Average Pressures. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Steam pressure by gauge 
 
 75 
 
 101.2 
 
 88 
 
 95 
 
 96 
 
 ",«, 
 
 Forced 
 
 86,3 
 } -=5 
 
 79.8 
 
 140 
 
 48,2 
 
 85 
 
 124-3 
 
 69.9 
 
 87.14 
 
 90 
 
 77-1 
 
 123-3 
 
 129 
 
 92 
 
 90 
 
 68.8 
 
 110.7 
 
 Cliimney draught in inclies of water . . 
 
 
 
 
 
 -64 
 
 
 -5 
 
 .42 
 
 
 
 .69 
 
 
 •75 
 
 -45 
 
 -65 
 
 
 • 143 
 
 
 1-23 
 
 .14 
 
 Average Temperatures. 
 
 
 
 
 
 .88 
 
 
 
 
 
 
 
 External air, degrees Fahr 
 
 M?lts 
 Lead. 
 
 32 
 
 72 
 
 64 
 
 37 
 
 89 
 
 83 
 
 
 31 
 
 
 49-8 
 
 76 
 
 
 46 
 
 47-7 
 
 67 
 
 
 90-5 
 
 50 
 
 90 
 
 82.5 
 
 37 
 
 
 Escaping gases from boiler, degrees F. | 
 
 816.3 
 
 520 
 
 512 
 
 439 
 
 542-5 
 
 
 388.5 
 
 358 
 
 457-5 
 
 623 • 
 
 
 478 
 
 
 375-4 
 
 605 
 
 453 
 
 644 
 
 480 
 
 530 
 
 498 
 
 501 
 
 619 
 
 Feed-water entering boiler, degrees F. . 
 
 
 149.9 
 
 120 
 
 138 
 
 71 
 
 167 
 
 148 
 
 169 
 
 76.4 
 
 144-5 
 
 94 
 
 64 
 
 159-7 
 
 114 
 
 51-93 
 
 181 
 
 39 
 
 84 
 
 168 
 
 83 
 
 91 
 
 35-3 
 
 71.1 
 
 Fuel. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Anth. 
 
 Bitu- 
 
 Anth. 
 
 Bitu- 
 
 Bitu- 
 
 Bitu- 
 
 Anth. 
 
 Youghi- 
 
 Anthra- 
 
 Poca- 
 
 Semi- 
 
 Anthra- 
 
 Bitu- 
 
 Bitu- 
 
 Anthra- 
 
 Anthra- 
 
 Poca- 
 
 Bitu- 
 
 Youghi- 
 
 
 Bitu- 
 
 Bitu- 
 
 BiUi- 
 
 
 Pea. 
 
 
 Pea. 
 
 
 
 minous. 
 
 Pea. 
 
 ogheny. 
 
 cite. 
 
 hontas. 
 
 Anth. 
 
 cite. 
 
 
 
 cite. 
 
 cite. 
 
 hontas. 
 
 minous. 
 
 ogheny. 
 
 
 
 
 minous. 
 
 Coal fed to furnace, deducting moisture, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 pounds 
 
 3900 
 
 16,875 
 
 2758 
 
 6817 
 
 11,250 
 
 24,000 
 
 3542 
 
 8938 
 
 3292 
 
 4275 
 
 4533 
 
 859 
 
 3480 
 
 8188 
 
 5809 
 
 7220 
 
 6644 
 
 8550 
 
 6929 
 
 18,115.41 
 
 12,821 
 
 36,252 
 
 6507 
 
 Percentage of ash 
 
 12.63 
 3704 
 
 12 
 
 13 
 2400 
 
 12.6 
 
 6.49 
 10,520 
 
 13-35 
 20,796 
 
 16.46 
 
 8247 
 
 17.4 
 2720 
 
 4.62 
 
 20.8 
 
 13-2 
 746 
 
 8-33 
 3073 
 
 9.18 
 7436 
 
 17.4 
 
 19.8 
 5789 
 
 7-1 
 
 3 
 
 5-5 
 
 32-14 
 
 4.21 
 
 3.75 
 
 6 
 
 Total combustible, pounds 
 
 14.879 
 
 5958 
 
 2959 
 
 4075 
 
 3589 
 
 4807 
 
 6.70 
 
 7866 
 
 6549 
 
 12,293.12 
 
 12,281 
 
 33,123 
 
 
 Coal burned per square foot of grate 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 surface per hour 
 
 
 18.75 
 
 11.03 
 
 I3-II 
 
 18 
 
 41.38 
 
 14. 
 
 9-79 
 
 
 11.50 
 
 
 8.93 
 
 
 
 12.82 
 
 
 12.9 
 
 31.7 
 
 13.32 
 
 22.52 
 
 16 
 
 31 -9 
 
 16.3 
 
 Coal burned per square foot of water- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 heating surface per hour 
 
 7-53 
 
 1.38 
 
 .361 
 
 .325 
 
 •332 
 
 1.277 
 
 -275 
 
 -330 
 
 .211 
 
 -195 
 
 -77 
 
 .221 
 
 -252 
 
 -853 
 
 .310 
 
 •576 
 
 .29 
 
 .608 
 
 -332 
 
 -439 
 
 •541 
 
 •363 
 
 -315 
 
 Quality of Steam. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1.32 
 
 
 ' 
 
 
 ' 
 
 -0055 
 
 .0136 
 
 .... 
 
 
 
 
 ' 
 
 
 -75 
 
 .91 
 
 -.58 
 
 
 6.81 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 17.I 
 
 
 
 
 
 
 
 
 1.2 
 
 
 4-4 
 
 Water. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Weight of water fed to boiler, pounds . 
 
 24.438 
 
 102,700 
 
 23,002 
 
 61,626 
 
 83.369 
 
 125,982 
 
 32,073 
 
 88,363 
 
 26,500 
 
 41.123 
 
 34,382 
 
 7654 
 
 36,225 
 
 71,464 
 
 47,782 
 
 53,092 
 
 57,528 
 
 77,995 
 
 71,784 
 
 103,285.14 
 
 109,980 
 
 31,433 
 
 64,772 
 
 Equivalent water evaporated into dry 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 steam from and at 212° Fahr., pounds 
 
 28,54s 
 
 113,440 
 
 26,063 
 
 68,784 
 
 97.536 
 
 136,816 
 
 35,402 
 
 95,609 
 
 30.687 
 
 45,742 
 
 39,584 
 
 9114 
 
 40,065 
 
 81,411 
 
 57,291.6 
 
 56,400 
 
 69,839 
 
 91,878 
 
 78,574 
 
 120,245.45 
 
 127,906 
 
 38,034 
 
 77,108 
 
 Evaporation. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Water evaporated per pound of coal 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 under actual conditions 
 
 5.42 
 
 6.086 
 
 8.34 
 
 9.04 
 
 7-31 
 
 5-25 
 
 9-05 
 
 9-845 
 
 7-93 
 
 9-57 
 
 7.58 
 
 8.91 
 
 10.42 
 
 8.73 
 
 8.22 
 
 7^30 
 
 8.658 
 
 9.T2 
 
 10.36 
 
 5-71 
 
 8.58 
 
 8.34 
 
 9-97 
 
 Equivalent evaporation from and at 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2i2°Fahr. per pound of coal 
 
 7-32 
 
 6.72 
 
 9-45 
 
 10.09 
 
 8.67 
 
 5-70 
 
 10.00 
 
 10.646 
 
 9-32 
 
 10.70 
 
 8.73 
 
 10.61 
 
 10.9s 
 
 9-71 
 
 9.86 
 
 7.81 
 
 10.51 1 
 
 10.74 
 
 11-34 
 
 6.64 
 
 999 
 
 10.07 
 
 11.84 
 
 Equivalent evaporation from and at 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ro.28 
 
 
 12.60 
 
 212° Fahr. per pound of combustible . 
 
 7.71 
 
 7.62 
 
 10.86 
 
 "•54 
 
 9.27 
 
 6.06 
 
 11.97 
 
 II-55I 
 
 11.28 
 
 11.22 
 
 11.02 
 
 12.22 
 
 12.90 
 
 10.92 
 
 11.92 
 
 9^75 
 
 I1.314 
 
 11.60 
 
 12.00 
 
 9-79 
 
 11.02 
 
 Equivalent evaporation from and at 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 212° Fahr. per square foot of water- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3^74* 
 
 3-^* 
 
 
 heating surface per hour, pounds . 
 
 5-51 
 
 9.26 
 
 3.42 
 
 3.28 
 
 2.8S 
 
 7.28 
 
 2.75 
 
 3.07 
 
 1.96 
 
 2.04 
 
 3-53 
 
 2.34 
 
 2.90 
 
 8.46 
 
 3-06 
 
 4^50 
 
 3-03 
 
 6.53 
 
 3-41 
 
 2.93 
 
 3-2 
 
 Commercial Horse-Power. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 On a basis of 34J4 pounds of water evap- 
 orated per hour from and at 212° Fahr. 
 
 82.75 
 
 328.81 
 
 75-54 
 
 199-37 
 
 282.71 
 
 396-57 
 
 I02.6 
 
 277-1 
 
 88.98 
 
 132-55 
 
 "5 
 
 26.42 
 
 116.1 
 
 235-97 
 
 i66 
 
 i63^5 
 
 202.4 
 
 280 
 
 229 
 
 348.53 
 
 370^8 
 
 1058.8 
 
 223-4 
 
 Square feet of water-heating surface 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 n.76 
 
 6^39 
 
 
 9.28 
 
 per H.-P 
 
 6.26 
 
 3-73 
 
 10.10 
 
 10.53 
 
 11.97 
 
 4-74 
 
 12.53 
 
 9-77 
 
 17-55 
 
 16.89 
 
 9-73 
 
 14.72 
 
 IT. 91 
 
 4.07 
 
 11.28 
 
 7.66 
 
 11.4 
 
 5-03 
 
 10.04 
 
 9^44 
 
 ■■ Includes superheating surface. 
 
CHAPTER X. 
 
 BOILER MOUNTINGS AND SAFETY APPARATUS. 
 
 The necessary attachments to a steam boiler include the feed- and 
 blow-pipes, safety-valve, gauge-cocks and water-gauge, pressure-gauge, 
 steam -delivery pipe, and the connections for operating the damper. 
 
 The Feed-Pipe may enter the boiler at any convenient place, but 
 the valves and other feed-controlling devices should, if possible, be 
 placed in the fire-room rather than elsewhere, because the water-level 
 requires constant attention on the part of the fireman. There is no 
 uniformity regarding the location for the admission of feed-water, but 
 almost any place is better than directly over the furnace or crown-sheet. 
 The coolest part of the boiler is generally thought to be at the bottom 
 of the rear end, probably because of the greater amount of sediment 
 which collects there, indicating that the circulation is least in that 
 portion of the boiler. This location has long been a favorite one for 
 the admission of feed-water. When a mud-drum is used the feed-water 
 sometimes enters there, but this is not now considered good practice. 
 One might easily infer from certain articles in technical journals on this 
 subject that the normal condition of the water in a steam boiler was in 
 layers of unequal temperature, which is probably not true, except as to 
 the underlying fact that it requires a difference in temperature to begin 
 and continue circulation. The important thing is to prevent local injury 
 to plates or tubes by not admitting 
 cold water. ^^^- 392- 
 
 An internal feed-pipe is recom- 
 mended. The opening into the 
 boiler may be at any convenient 
 place — the front or back head will 
 answer as well as any. After tap- 
 ping the head put in a nipple with 
 a long thread and screw a coupling 
 on the inside of the boiler, as 
 shown in Fig. 392, to which a 
 pipe may be attached leading to 
 within a few inches of the other 
 end of the boiler ; an elbow will 
 then permit the pipe to be ex- 
 tended upward between the tubes (if the latter are not placed zigzag), 
 with another elbow leading towards the opposite end of the boiler. 
 The far end of this last pipe may be bent downward so as to direct the 
 
 343 
 
344 
 
 BOILERS AND FURNACES 
 
 Fig 
 
 ■^3^ 
 
 feed-water down among the tubes, where it will be taken up by the 
 general circulation. Feed-water entering the boiler in this manner will 
 have acquired the same temperature as that of the water within the 
 boiler before it is discharged into the general circulation. 
 
 The diameter of the feed-pipe should be quite liberal. The larger 
 the pipe within the boiler the slower the movement of the water within, 
 and the more heat it will have absorbed before it is discharged into the 
 boiler. The least diameter recommended for the internal feed- or circu- 
 lating-pipe is l}{ inches for hori- 
 zontal tubular boilers up to 48 
 inches diameter, and i^ inches 
 for larger diameters. 
 
 Feeding into the steam-room 
 of a boiler is now quite common. 
 Moore's device, shown in Fig. 
 393, has been in use some twenty 
 years. The water is admitted, as 
 shown by the overhead pipe and 
 direction of the arrow, through 
 the shell of the boiler into a 
 return fitting, discharging up- 
 ward at the ball check -valve. A guard over the top of the check-valve 
 prevents any spray touching the overhead shell of the boiler. The feed- 
 water, falling upon the water underneath, is immediately taken up by the 
 circulation without coming in contact with any portion of the boiler. 
 
 A suspended pan in the steam-room, with the feed-pipe leading 
 directly into it, as shown in Fig. 394, has been used in a number of 
 boilers by the writer 
 
 during the past twenty ^^" ^^^' 
 
 years with good results. 
 The pan may be 3 or 4 
 feet long, 12 to 18 inches 
 wide, and 3 to 5 inches 
 deep, according to the 
 size of the boiler. The 
 top edge is serrated like 
 a saw, so that the over- 
 flow will occur by drops 
 and not in a stream, the 
 drops being taken up 
 by the circulation proceeding along the surface of the water. 
 
 Feeding into an inverted frustum of a cone which overflows into a 
 shallow pan and from the latter into the boiler, as developed in Ford' s 
 device, is shown in Fig. 395. This device, wherever used, has given 
 satisfaction. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 345 
 
 Ftg. 395. 
 
 A purifying-chamber in the rear end of the water- and steam-drums 
 of the water-tube boilers, made by the Standard Boiler Company, is 
 unique in providing an 
 apartment over which the 
 pans are suspended in the 
 steam-space, the overflow 
 of which is not taken up 
 directly by the circulation, 
 but falls into a chamber 
 partitioned off, so as to 
 be uninfluenced by the 
 circulation, as shown in 
 Fig. 396. This purifying- 
 chamber permits a pre- 
 cipitation of the scale- 
 bearing matter into a 
 separate compartment, 
 where it can be discharged 
 
 through a bottom-blow, also shown in the engraving. By this means 
 the purifying-chamber can be completely emptied, if so desired, without 
 disturbing the water-level in the boiler, a great advantage in any locality 
 where the feed-water contains salts of lime and magnesia. 
 
 Fig. 396. 
 
 Check- Valves permit a flow of water into the boiler and prevent 
 its escape through the feed-pipe upon withdrawal of pressure. They- 
 are made in considerable variety as to exterior form, but the principle 
 of operation is always the same. For small powers the globe check- 
 valve, shown in Fig. 397, is probably used more than any other. The 
 globe, cap, and valve are all made of hard gun-metal ; the ends of the 
 globe are tapped for standard wrought-iron pipe-threads. The valve 
 
 23 
 
346 
 
 BOILERS AND FURNACES 
 
 and seat are flat in this engraving, but they are also made with a bevel, 
 as shown in the section of an angle valve, Fig. 398. Whether a globe 
 
 Fig. 397. 
 
 Fig. 398. 
 
 or an angle check-valve shall be used will depend upon the details of 
 piping for any given boiler. 
 
 A ball check-valve is shown in Fig. 399. This valve is not in as 
 common use as the one preceding, and is not generally regarded with 
 the favor accorded valves having flat or conical faces, which may be 
 easily reground after distortion occasioned by wear. 
 
 The swinging check-valve, shown in Fig. 400, is now much used 
 and has given general satisfaction. It gives a straight way for the 
 passage of the feed-water. The valve is free to revolve on the face, 
 having a pivoted connection in the swinging arm, as shown. These 
 valves are wholly made of hard gun-metal. 
 
 Fig. 399. 
 
 Fig. 400. 
 
 An angle check-valve, much used with injectors, is shown in Fig. 
 401. It is provided with a tapered nipple for screwing directly into 
 the shell of a boiler ; the lower connection is that of a standard union 
 common to wrought-iron pipe-work. The valve is winged and fitted to 
 a bevelled seat. All the parts of this valve are made of hard gun-metal. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 347 
 
 Fig. 402. 
 
 An angle valve, shown in Fig. 402, is much used in large installations 
 where several boilers are connected with the same feed-pipe. The main 
 shell and cap are of cast iron ; the valve 
 and bushing forming the seat are of hard 
 gun-metal. The shell is bored the depth 
 of the bushing with a recess, say of % inch, 
 into which the bushing, having a corre- 
 
 FiG. 401. 
 
 sponding projection, is forced. The cap is provided with a yoke tapped 
 to receive either a brass or an iron spindle, which, passing through [a 
 stuffing-box, also included in the cap, can be screwed down upon the 
 check-valve so as to prevent 
 
 it rising except at such times Fig. 403. 
 
 as feed-water may be re- 
 quired in the boiler to which 
 it is attached. This valve is 
 flanged, and must be fitted 
 to corresponding flanges in- 
 cluded in the system of 
 piping, the whole being se- 
 cured by through -going bolts 
 and nuts. 
 
 Culver's stop- and check- 
 valve combined is shown in 
 Fig. 403. The valve body is 
 made of iron and in two 
 parts, the gun-metal seat 
 being located between the 
 two, and all bolted together through flanges outside of the seat. This 
 design is to all intents a straightway valve. An inspection of the draw- 
 
348 BOILERS AND FURNACES 
 
 ing will make clear that it is first of all a stop-valve consisting of two 
 parts, a small auxiliary valve having its seat on the back of the larger 
 valve, and this larger valve having its seat on the central gun-metal ring 
 lying between the two halves of the main body. The small valve is at- 
 tached to the spindle operated by the hand-wheel the same as any globe 
 valve. Raising the screwed spindle, however slightly, opens the small 
 valve and water enters the space between the main valve and the check- 
 valve below, thus balancing the upper disk. Upon a further raising of 
 the spindle the nut under the upper valve is brought in contact with the 
 latter and easily raised from its seat, because the pressure is the same 
 on both sides. A square lug on the bottom of the check-valve passes 
 through an opening in the lower half of the body, which allows it to 
 work up and down freely and guides it to its seat, this guide keeping 
 the check in perfect alignment with its seat. By unscrewing the bottom 
 valve-cap and rotating the valve on its seat the check-valve can be re- 
 ground in case of leakage. The seats are renewable by removing the bolts 
 and springing the valve body far enough apart to remove the brass seat- 
 ring and slip another in its stead ; this can be done in less time than 
 would be required to grind the valve if the leakage should be very great. 
 
 Gate Valves are much used for water-valves because they give a 
 straightway passage of full diameter of the connecting-pipes. A sec- 
 tional elevation of a Chapman valve is given in Fig. 404. It consists 
 of a plug or gate in one piece, guided closely in the shell by means of 
 ribs or splines, the latter taking all the strains, relieving the faces of the 
 seats until the central plug is seated, the splines insuring a true and easy 
 vertical movement of plug. The seats are hard gun -metal, and when 
 placed in their proper positions in the body of the shell are held to their 
 exact line by means of a screw-gland inserted through the threaded 
 ends. These glands can be worked forward and back by means of 
 a spanner fitting the sphnes in the inside of these screw-glands. 
 
 The plug is double-faced and equally tight on either face, and either 
 end of the valve may be used for inlet or outlet. This valve is shown 
 with a stationary spindle ; that is, the valve rises and falls on the spindle ; 
 in other forms the spindle rises and falls through the stuffing-box corre- 
 sponding to the position of the valve. 
 
 A Jenkins gate valve with travelling spindle is shown in Fig. 405. 
 The details of the spindle are not shown, but they are similar to those 
 of Fig. 450. The body has one vertical and one inclined side, both 
 of which are faced true. The valve or plug has of necessity a ver- 
 tical and inclined side to fit the seat. The back of the valve fits, metal 
 to metal, against the body ; the face is furnished with a vulcanized 
 disk, shown black in the engraving ; this has a metal guard outside of 
 it, as shown ; these disks are renewable by unscrewing the cap and 
 removing the valve from the body. The spindle is independent of the 
 disks, therefore not liable to stick in opening or closing the valve. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 349 
 
 A Bottom Blow is used for emptying the boiler ; it is placed, there- 
 fore, at the lowest part of the boiler. If the latter is supplied with a 
 mud-drum the bottom blow discharges from that. Horizontal boilers are 
 usually set with an inclination of two to three inches to the rear. The 
 
 Fig. 404. 
 
 Fig. 405. 
 
 O 
 
 oh 
 
 blow-oif should be attached by preference at the bottom of the boiler 
 rather than through the rear head, because, for practical reasons, the 
 drilling and tapping into a boiler-head must occur above the curve of 
 the flange, which leaves a couple of inches of water underneath the 
 pipe that cannot be drained. The blow-off opening into the boiler in 
 the case of pipes larger than two inches should be reinforced by a plate 
 riveted to the shell, the pipe-tap screwing through both plates, or a 
 pipe-flange should be riveted to the boiler instead. 
 
 It is sometimes necessary to insert an elbow immediately under the 
 boiler, as shown in Fig. 295, that the blow-off pipe may pass through 
 the furnace-walls at a higher level than the bottom of the combustion- 
 chamber. Some manufacturers furnish a cast-iron pipe, as shown in 
 Fig. 406. In either case the pipe and fittings are wholly enveloped by 
 hot gases. But a much better method is to carry the pipe vertically 
 downward to the bottom of the combustion-chamber, where the tem- 
 perature is much lower, and connect the elbow near or below the floor- 
 level. An example of such an arrangement is shown in Fig. 298. A 
 brick protecting wall is built around the blow-off pipe to the bottom of 
 
350 
 
 BOILERS AND FURNACES 
 
 the combustion-chamber, and a long^ curve joins this pipe to the front 
 of the boiler. 
 
 Another method is to build a pier around the pipe, as shown in Fig. 
 
 Fig. 406. 
 
 /fffr/','ff^/^ff/-/A 
 
 407. As this requires only the length of one brick it does not obstruct 
 the flow of gases. Still another method is to wrap the blow-off pipe 
 with coils of plaited asbestos packing ; the latter, being non-combustible, 
 affords considerable protection to the blow-off" pipe. 
 
 The diameter of the bottom blow-off" pipe should be i^ inches for 
 boilers up to 42 inches in diameter, 2 inches for 44- to 60-inch boilers, 
 and 2^-inch pipe for larger diameters. The blow-off" pipe must be laid 
 so as to completely drain at the discharging end, or it may in time fill 
 with scale or mud, "When using the bottom blow the velocity of flow 
 should be rapid, otherwise scale is liable to lodge in the turns or other 
 pipe-fittings and eventually clog the pipe so as to prevent full flow ; for 
 this reason, when blowing, the valve or plug-cock should be wide open 
 or nearly so. 
 
 A Surface Blow is needed for boilers in which the feed-water con- 
 tains impurities liable to separate during the process of ebullition and 
 form a thick scum on the surface of the water. This scum is greatest 
 in waters in which the carbonates of magnesia predominate. The sur- 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 351 
 
 face blow should be located at the water-level and in that part of the 
 boiler in which the surface agitation of the water is least. It should have 
 a receiving or collecting funnel with a wide surface across the boiler, the 
 wider the better. 
 
 An arrangement of surface and bottom blows as applied to a vertical 
 boiler, shown in Fig. 408, was illustrated in the technical journals a few 
 years ago, which, as a 
 scheme of piping, is here 
 reproduced. When the 
 valve at the top of the 
 boiler under the hori- 
 zontal branch - pipe and 
 the lower valve outside of 
 the bottom tee are opened 
 the water will be blown 
 from the bottom of the 
 blow-oft pipe, located at 
 about an inch above the 
 surface of the crown-sheet, 
 the latter being swept by 
 the current of escaping 
 water as it passes across 
 it and up the pipe. So, 
 also, any oil or scum that 
 may be floating upon the 
 surface of the water, and 
 which would otherwise be 
 deposited upon the crown- 
 sheet, is blown ofl" in the 
 same way by opening the 
 surface blow. After all 
 the water that can be 
 reached by the surface 
 blow is carried off in this 
 manner, the top valve 
 may be closed and the 
 bottom valve next the 
 boiler opened, when the 
 remainder of the water in 
 
 the boiler may be blown out of the bottom blow. After all the water 
 has been blown out of the boiler, the top blow-off valve being closed 
 and the uppermost valve shown in the engraving opened, a hose can be 
 coupled to the upper pipe and the crown-sheet washed off, the water 
 striking the centre of the crown-sheet and washing all sediment into the 
 water-legs, where it can be removed through the handholes. 
 
352 
 
 BOILERS AND FURNACES 
 
 The Hotchkiss surface blow, illustrated in Fig. 409, has been in use 
 in this country for the past twenty years. It consists of a funnel, an 
 
 up-flow pipe, a reservoir, a 
 Fig. 409. return pipe, and a blow-oft 
 
 pipe. The funnel is made 
 of iron, and as large as will 
 pass through the manhole. 
 The reservoir is a cast-iron 
 spherical vessel, with a capac- 
 ity of about eighteen gallons 
 and of sufticient thickness to 
 withstand the boiler press- 
 ure. Its action is as follows : 
 in a boiler with the cleaner 
 attached, the funnel is set 
 near the surface, but partly 
 submerged, and in such po- 
 sition that its opening will 
 intercept the currents of hot 
 water flowing towards it. By 
 the action of gravity the hot 
 surface water entering the 
 funnel will flow into the reservoir through the up-flow pipe, displacing 
 an equal quantity of cooler water therein, which latter returns to the 
 boiler by the pipe shown between the flues. A constant circulation ol 
 water through the cleaner is thus maintained by the unbalanced columns 
 of water so long as firing is kept up under the boiler. Any sediment 
 once deposited in the reservoir remains there until removed through 
 the blow-pff" pipe under control of the 
 engineer. 
 
 Plug- Cocks are generally used for 
 the bottom blow of steam boilers. An 
 objection to the use of ordinary plug- 
 cocks has been the tendency to leak 
 around the bottom of the plug, the loss 
 of water not being of so much account 
 as the drip upon the floor, which makes 
 in some fire-rooms a disagreeable slop ; 
 also the tendency of the plug to stick 
 to the shell if driven down slightly after 
 closing it, making it difficult to after- 
 wards move the plug when required. 
 Both of these objections are over- 
 come in the plug-cock shown in Fig. 410, which has a cast bottom 
 through which no drips can occur. The sticking is less objectionable 
 
 Fig. 410. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 353 
 
 Fig. 
 
 because the collar-nut at the top of the plug, which forces the latter 
 down into its seat, will by a slight turn cause a vertical movement of 
 the plug and thus loosen it sufficiently to move it with an ordinary 
 wrench. 
 
 An asbestos-packed plug-cock is shown in horizontal and vertical 
 sections in Fig. 411, the latter on a larger scale than the former. This 
 cock has a closed bottom, 
 the asbestos being driven 
 solidly in the dove-tail 
 grooves in the body of the 
 cock and, being elastic, 
 fits against the plug, mak- 
 ing a tight joint with but 
 little friction. Asbestos, 
 being unaffected by heat 
 and moisture, is quite 
 durable. These cocks 
 are in good repute wher- 
 ever used. 
 
 Ordinary globe valves should not be used as a bottom blow for 
 steam boilers ; a gate valve or one of the forms of angle valves in 
 which no lodgment of scale can occur is much to be preferred. A 
 Jenkins valve, similar to Fig. 412, or the Eastwood valve, shown in 
 
 Fig. 413, have both been 
 
 Fig. 412. 
 
 used with satisfactory 
 results. Myers's blow- 
 off valve as improved 
 by Mowry is shown in 
 Fig. 414. The seat of 
 this valve is removable. 
 Double wheels are so 
 arranged that the large 
 wheel opens and seats 
 the valve. The small 
 wheel rotates the valve 
 on its seat without either 
 raising or lowering the 
 disk. After the valve 
 has been opened and 
 water blown off", the 
 large wheel is used to 
 close the valve upon its 
 seat ; then, reversing this wheel about an eighth of a turn, the smaller 
 wheel is revolved to clean off" both the valve-disk and the ring on which 
 it is seated. 
 
354 
 
 BOILERS AND FURNACES 
 
 Safety- Valves. —The object of a safety-valve is to relieve the boiler 
 and prevent accumulation of steam pressure above the limit at which 
 the valve is set. The grate surface is now the commonly accepted unit 
 by which to determine the size of a safety-valve. The rate of com- 
 bustion varies with different boiler furnaces and the kind of coal used. 
 
 Fig. 414. 
 
 Fig. 413. 
 
 Records show from 12 to 40 pounds of coal per square foot of grate 
 per hour, but furnaces with ordinary draught do not often burn more 
 than 20 pounds. Good evaporation may be assumed to be 10 pounds 
 of water per pound of coal. 
 
 Each boiler should have its own safety-valve. Some municipalities, 
 Philadelphia, for example, require safety-valves to be in duplicate. 
 Safety-valves should be attached directly to the shell or to the steam- 
 dome of the boiler. There should never be a stop-valve intervening 
 between the boiler and the safety-valve if, by closing the former, the 
 latter is no longer in communication with the boiler pressure. 
 
 The United States Regulations for steam-vessels require that lever 
 safety-valves shall have an area of not less than i square inch to 2 
 square feet of grate surface in the boiler, and the seats of all such 
 safety-valves shall have an inclination of 45° to the centre line of their 
 axes. These proportions obtain in good stationary-engine practice. 
 
 Fig. 415 shows a combined safety-valve and stop- valve, useful for 
 situations in which several boilers deliver the steam into a common 
 reservoir. Spring-loaded safety-valves, constructed so as to give an 
 increased lift by the operation of steam, after the valve is raised from its 
 seat, shall be required, according to the United States Regulations, to 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 355 
 
 have an area of not less than i square inch to 3 square feet of grate 
 surface of the boiler, 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 
 one-eighth of the diameter of the valve opening. The seats of all such 
 safety-valves shall have an angle of inclination to the centre line of their 
 axes of 45 degrees. But in no case shall any spring-loaded safety-valve 
 be used in lieu of the lever-weighted safety-valve without first having 
 been approved by the Board of Supervising Inspectors. 
 
 Fig. 415. 
 
 The diameter of a safety-valve is not a test of its efficiency. A valve 
 is effective in direct proportion to its lift, other conditions being equal. 
 Professor Burg, of Vienna, found by actual measurements that a lever 
 safety-valve of 4 inches diameter rises from its seat according to the 
 laws stated below : 
 
 With a boiler pressure of . . 
 
 The rise of a common valve 
 
 is, in parts of an inch . . . 
 
 12 
 
 20 
 
 35 
 
 45 
 
 50 
 
 60 
 
 70 
 
 80 
 
 aV 
 
 is 
 
 sk 
 
 eV 
 
 A 
 
 A 
 
 Th 
 
 Tk 
 
 90 pounds 
 
 Other reliable authorities do not fix the rise of such a valve from its 
 seat at more than ^ oi i inch when loaded at any pressure between 
 12 pounds and 90 pounds. And further experiments of Mr. Burg 
 proved beyond a doubt that the higher the pressure the less will a 
 common safety-valve rise ; and in not rising it simply obeys the action 
 of the forces exerted upon it. 
 
 According to the table of Professor Burg, the actual size of the 
 venting capacity of a common lever and weight safety-valve of 3 inches 
 diameter is but 0.56 of a square inch area at 70 pounds of steam. 
 Pressure ought, therefore, to be taken into account when fixing upon 
 the size of a safety-valve, especially for large powers, or in all cases 
 
356 
 
 BOILERS AND FURNACES 
 
 where the flow of steam is intermittent and pressures are likely to 
 accumulate nearly or quite to the danger limit. 
 
 The Philadelphia regulations for fixing the size of safety-valves take 
 the pressure into account, and order that the least aggregate area of 
 the two safety-valves required by law (being the least sectional area for 
 the discharge of steam) to be placed upon all stationary boilers with 
 natural or chimney draught may be expressed by the formula : 
 
 ^ ~ P + 8.62, 
 in which A is the area of combined safety-valves in inches ; G is the 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 : 
 
 Pressure per Square Inch. 
 
 10 20 30 40 50 60 70 80 90 100 no 120 150 175 
 
 Valve Area in Square Inches, corresponding to i Square Foot of Grate. 
 
 •79 
 
 58 .46 .38 .33 
 
 29 .25 .23 
 
 .19 
 
 .14 
 
 Calculating the Load on a Lever Safety- Valve. — Two things 
 are usually known in advance, — the size of the valve and the pressure 
 at which it is to work. 
 
 Reference letters for lever safety-valves are given in Fig. 416. Let 
 
 A = area of valve in inches. 
 P = pressure at which valve is to lift. 
 S = short arm of lever in inches. 
 L = long arm of lever in inches. 
 W = weight of ball in pounds. 
 w = weight of lever in pounds. 
 g =^ fulcrum to centre of gravity in inches. 
 V = weight of valve and spindle in pounds, to which add : 
 T ^ actual work to be done, expressive of : 
 
 ^ C^ ^ ), explained further on. 
 
 zi^ X g 
 S 
 
 A X P — 
 
 / = the effect of z' -f 
 
 in pounds. 
 
 These letters of reference are as far as practicable included in the 
 accompanying sketches. 
 The ordinary formula. 
 
 W = 
 
 A X P X S 
 
 is not complete, because it does not take into account the weight of the 
 lever and valve. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 357 
 
 The correct formula would be 
 
 W = PxA-(z. + ''^) X l- 
 
 This somewhat complicated-looking formula is made necessary be- 
 cause of the weight of the valve and the additional influence of the lever. 
 The weight of the valve and spindle, v, may be had by the simple pro- 
 cess of direct weighing. The influence of the lever upon the valve will 
 be the weight of the lever, w, acting at its centre of gravity, ^; this latter 
 
 Fig. 416. 
 
 ,g^^-^_!_ 
 
 
 ^ A ^ 
 
 1- 1 1 
 
 ^^^-^-Th— ^^ ^-^^ 
 
 'iJiM 
 
 -^ 
 
 IT 
 
 ( ^ 
 
 may be found by simply balancing it on a sharp edge, as in Fig. 416. 
 The combined eflect of the weight of the valve and that of the lever 
 acting upon it as just indicated is that expressed in the formula as 
 
 S 
 
 which is to be deducted from the work required of the weight, W. 
 
 However simple all this may appear to those who have it frequently 
 to do, many practical men never feel wholly safe in such calculations, 
 and much prefer to weigh the lever and valve, also shown in Fig. 416. 
 Whatever the weight in pounds thus observed may be is simply de- 
 ducted from the total pressure (A X P). This amount is that repre- 
 sented in the letters of reference as T. 
 
 To illustrate the foregoing, let us assume a 4-inch lever safety-valve 
 of the following proportions : 
 
 Area of valve, A 12.57 sq. in. 
 
 Pressure, P 100 pounds per sq. in. 
 
 Short lever, S 4 inches. 
 
 Long lever, L 36 inches. 
 
 Weight of lever, w 10 pounds. 
 
 Fulcrum to centre of gravity, g- 16 inches. 
 
 Weight of valve and stem, v 6 pounds. 
 
 Effect of z^ + ^ ^ represented by / .... 46 pounds. 
 
 The value of T for this particular valve as calculated by the formulae 
 given in the letters of reference is 
 
 T = 12.57 X 100 — (6 + I2j<_i6\ ^ J2II pounds. 
 
 V 4 / 
 
358 BOILERS AND FURNACES 
 
 Then 
 
 w = 3^^, or ze/ = ^^"^^ 4 ^ 134.5+ pounds. 
 L 36 
 
 Other calculations from the same data may be made thus : 
 
 T = W_^^ ^^ T _ ^34.5 X 36 _ J2II pounds. 
 S 4 
 
 The long arm of the lever thus : 
 
 L =. 5^^^, or L = i^i^LX^ = 36 inches. 
 w 134-5 
 
 The short arm of the lever thus : 
 
 S _ WXL ^^ s _ 134-5X36 _ i„,hes. 
 T 1211 
 
 To find the pressure, P, we may proceed thus : 
 WX_L ^ ^ ^34-5 X 36 ^ ^6 
 
 P = , or P = ^ = 100 pounds. 
 
 A 12.57 
 
 Other sizes and other proportions may be similarly worked out. 
 
 Spring-Loaded Safety- Valves are those in which pressure is regu- 
 lated by the tension of a spring instead of a weight and lever. Safety- 
 valves of this kind do not ordinarily have plain bevelled valves, but are 
 so constructed that when the pressure reaches the point at which it is to 
 blow, the valve opens slightly at first and allows steam to escape. This 
 escaping steam enters an annular chamber around the valve, and whilst 
 the steam is not wholly prevented from escaping into the atmosphere, it 
 is sufficiently retarded to accumulate pressure in this chamber, to which 
 must also be added the work done by the escaping steam by impinging 
 against the larger surface of the valve before it can make its escape 
 downward, both of which effects, added to the valve already in balance, 
 because of the pressure underneath, combine to force the valve suddenly 
 upward to its full height of lift, producing a sound which has given this 
 valve its characteristic name, — the pop safety-valve. The valve, being 
 thus at its full height, quickly relieves the boiler of its pressure, and as 
 the latter falls the valve is returned to its seat by the spiral spring above. 
 
 The American pop safety-valve is shown in sectional elevation in 
 Fig. 417, which shows a flat valve on a flat seat, the kind usually fur- 
 nished for stationary boilers, but for marine boilers the valves and seats 
 are bevelled at an angle of 45°. The valve is originally set for any 
 desired pressure by screwing down on top of the spring the sleeve-nut 
 shown at the top of the case, afterwards securing it by the lock-nut, also 
 shown at the top of the boss through which the sleeve-nut is adjusted. 
 A combination of levers is shown by which the valve can be raised to 
 a height equal to one-eighth of its diameter, to conform to the United 
 States Regulations. To reset the valve, first loosen the lock-nut, then 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 359 
 
 screw down the sleeve-nut to get increased pressure, or unscrew it to 
 get a lower pressure. The blow-down is adjusted by a relief ring fitted 
 with adjusting screws, shown on either side of the valve, which screws 
 are adjustable in the outside flange ; by raising or lowering this ring the 
 distance between it and the outer lip of the valve can be adjusted to any 
 desired area, and thus any desired height of lift of valve when blowing off. 
 
 Fig. 418. 
 
 The consolidated safety-valve is shown in Fig. 418. The flanged 
 base is cast iron, into which is screwed a gun-metal seat ; a winged valve 
 is fitted into this seat, having a bevelled edge of 45°. The valve has a 
 projecting annular lip, underneath which and screwed upon the outside 
 of the valve-seat is an adjustable ring for regulating the lift of the valve 
 when blowing off. The central stem by which the pressure of the spring 
 is brought upon the valve has its lower end pivoted well below the valve- 
 face, so as to insure a vertical movement at all times. This stem has 
 two collars with curved faces, against which the spring is compressed, 
 the thrust at the upper end being taken by an adjustable screw-bolt 
 passing through the upper part of the cast-iron case, the screw being 
 fixed at any determined point by means of a lock-nut, all of which are 
 shown in the engraving. A lever is also provided by which the valve 
 may be lifted off its seat one-eighth of its diameter. 
 
 Safety-valves should blow directly into the boiler-room, and not 
 
360 BOILERS AND FURNACES 
 
 through a pipe into a chimney or other outlet. The objection to piping 
 a safety-valve outlet is, that unless it passes directly upward through the 
 roof there is always a possibility that water will accumulate in the valve 
 chamber and be a direct cause of external corrosion, unless provided 
 with suitable drips. Neglect to properly drain such pipes might end 
 quite disastrously in the winter by freezing, if the pipes and valves are 
 in an exposed position. 
 
 Low- Water Alarms. — These appendages to a steam boiler have 
 long been in use, and are, therefore, of varying detail. They all have 
 a common object, which is to sound an alarm in the event that the water 
 in the boiler to which it is attached falls below its proper level. The 
 alarm is commonly operated in one of three ways : by melting a fusible 
 plug, by a differential expansion movement, as in the case of a wrought- 
 iron and a copper pipe, or mechanically, by the use of floats ; to which 
 might be added, as a fourth, electrical devices. But these have never 
 come into general use and are not now in favor. 
 
 The Ashcroft low- water detector is shown in Fig. 419, attached to 
 the top of the shell of a horizontal tubular boiler. The ordinary water- 
 level is shown ; the alarm-level corresponds to that of the bottom of the 
 pipe above the tubes ; this distance is to be fixed for each particular 
 case, and will probably vary from ^ to i inch, depending on the size of 
 the boiler and the rate of evaporation. The plug-cock is always to be 
 left open ; this permits the water to flow into the upper chamber by the 
 steam pressure acting upon the surface of the water below ; there being 
 no circulation in this chamber, the water cools to a temperature much 
 below that of the boiler. A fusible plug which is shown, by double 
 hatching closes the orifice leading into the water-chamber, and is held in 
 place by a screw-plug having a central hole in it leading to the attached 
 steam whistle. In the event of the water getting below the pipe, shown 
 in the engraving, the upper chamber and its connecting-pipe is emptied, 
 steam takes its place, the temperature of which is sufliciently high to 
 melt the fusible alloy, and escaping through the whistle sounds an alarm. 
 After a plug has melted the intermediate cock must be closed to stop 
 the alarm. Water is then pumped up to its proper level. To fix a new 
 fusible plug, unscrew the hollow plug by which it is to be held in place, 
 and after carefully removing all fragments of the former fusible plug, in- 
 sert a new one and replace the hollow plug, screwing it up sufliciently to 
 make the joint water-tight ; then turn on the water slowly to give it time 
 to cool sufficiently so as not to melt the plug. When the pressure is on 
 there must be no leak or drip, otherwise a circulation would be established, 
 and the plug melted by the hot water flowing into the upper chamber. 
 
 Fusible plug-alarms have been objected to on the ground that they 
 can never be tested, and new plugs have to be replaced at intervals of a 
 few months at great inconvenience, the renewals being necessary on 
 account of certain molecular changes which occur in the metal, which 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 361 
 
 becomes non-fusible in use. They are also rendered inoperative by scale 
 or sediment. 
 
 The Hardwick low-water alarm is shown in Fig. 420. This device 
 is operated by the different expansions for the same temperature caused 
 by steam entering two pipes, one of which is of wrought iron and the 
 other of brass, copper, or other material in which the rate of expansion 
 
 Fig. 419 
 
 Fig. 420. 
 
 is more than that of iron. The shell of the boiler is tapped to receive 
 a casting having two pipe-openings, into one of which is screwed a 
 wrought-iron pipe extending upward and terminating in a whistle ; the 
 other pipe, being made of brass or copper, also extends upward, and is 
 fitted with a cap at the top. Near the top of this pipe and attached to 
 the iron pipe is an arm carrying the fulcrum of a lever. This lever is in 
 the form of a right angle, the upper end of which is intended to operate 
 a steam-whistle ; the lower end, or horizontal member of the lever, is 
 
 24 
 
362 BOILERS AND FURNACES 
 
 fitted with an adjusting screw. Underneath the part which screws into 
 the boiler is a short piece of pipe fitted to that opening which leads to 
 the expansion-pipe. The length of this short piece of pipe is governed 
 by the height to which the water is to be carried in the boiler, and which 
 governs the sounding of the alarm. 
 
 When steam is raised on the boiler the water is forced up into the 
 expansion-pipe and, as there is no circulation in this pipe, it is soon 
 cooled, and will remain cool so long as the pressure is maintained in the 
 boiler. In the event that the water should fall below the bottom of this 
 pipe, the water contained in the expansion-pipe will fall by gravity and 
 steam will take its place ; the expansion of the brass or copper pipe, 
 being greater than that of the iron pipe under the same steam pressure, 
 will push upward the horizontal arm of the lever, which in turn will 
 push the valve of the whistle inward and sound an alarm. It will be 
 observed that there is a dry-steam connection from the top of the boiler 
 directly to the whistle-valve, so that the operation of the whistle will 
 not be interfered with by a combined mixture of water and steam ; this 
 might result if the steam were drawn from the water-line of the boiler, 
 which is always in a more or less agitated state by reason of the disen- 
 gagement of steam going on at that point. When the water-level is re- 
 stored the brass pipe contracts in length and the whistle ceases blowing. 
 
 Expansion-tube devices have been objected to, notwithstanding the 
 fact that nothing is more certain than the expansion of metal under 
 heat, but the fact that the expansion of the metal in a low-water alarm 
 is dependent upon steam taking the place of water — and the sound- 
 ing of the whistle is dependent upon this condition and the proper 
 adjustment and rigidity of the parts — makes, it is claimed, the greater 
 expansion of one of the metals a matter of somewhat remote importance, 
 also that the adjustment of these devices is commonly too fine for prac- 
 tical use in a boiler-room, inasmuch as the difference in temperature 
 between hot water and steam is not sufficient to cause very great elon- 
 gation of the tube. These strictures, while rather severe, are apparently- 
 supported by facts in the case of failure of expansion devices not wholly 
 confined to low-water alarms. Particular care should be taken with such 
 devices to see that they are at all times in good working order. 
 
 The Ashley low-water alarm combined with a water-column is shown 
 in Fig. 421. The principle of its operation is based upon the difference 
 in weight of a body suspended in air and immersed in water. The alarm 
 has two connections with the boiler, — a steam- and a water-connection 
 common to all water-columns. The top cover of the column is fastened 
 in place by tap-bolts, is removable, and to this is attached all the mech- 
 anism of the device. Suspended from the under side of the cover is a 
 valve working in combination with a double-ended lever, having its 
 fulcrum between its two ends. From the ends of the lever two cylinders 
 are suspended, the upper and smaller one of solid iron and the larger 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 363 
 
 one hollow, with holes in its top, so that it is filled with water. When 
 the water stands at the desired level in the boiler and column the solid 
 cylinder is heavier than the larger immersed cylinder, and consequently 
 keeps the valve closed. As the water in the column lowers, the hollow 
 cylinder filled with water overbalances the weight of the solid cylinder 
 
 Fig. 421. 
 
 Fig. 422. 
 
 when deprived of the buoyancy of the surrounding water and opens 
 the valve, admitting steam to the alarm whistle. The solid cylinder 
 regains its counterbalance on the admission of water to the boiler and 
 column, closing the valve. Its working can be tested at any time by 
 simply opening the valve at the bottom of the column, thus lowering 
 the water in the column and sounding the whistle. By removing the 
 bolts in the cover the whole of the mechanism can be lifted out for 
 cleaning or inspection, and easily replaced without disturbing the boiler 
 connections. The body of this alarm should always be attached in a 
 vertical position with no valve between the boiler and the alarm. The 
 position of the bottom of the gauge-glass should be at the line where 
 the whistle is to blow, in which case the blow would occur when the 
 
364 
 
 BOILERS AND FURNACES 
 
 water is about one inch from the bottom of the glass. The alarm can 
 be tested by blowing the water out of the column, thus allowing the 
 mechanism in the alarm to operate and blow the whistle. 
 
 The Pittsburgh high- and low-water alarm is operated by means of 
 a copper float placed in a water-column, as shown in Fig. 422. As the 
 water falls to the level of the bottom of the gauge-glass the float falls 
 with it, the collar on the vertical rod resting upon and afterwards lower- 
 ing the central end of the lever through which it passes ; the other end, 
 being fulcrumed above, opens the steam-valve by lifting it from its seat, 
 sounding the alarm for low water. In the event that too much water is 
 fed into the boiler, the float, rising in the water-column, carries with it 
 a lower collar attached to the vertical rod near the float, which lifts the 
 
 central end of the lever up- 
 FiG. 423. ward ; the lever, centring upon 
 
 the opposite fulcrum in the 
 valve-chamber above, causes 
 the alarm-valve to move up- 
 ward with the float and blow 
 the whistle. These collars can 
 be adjusted to any desired va- 
 riation in water-level. 
 
 The Reliance high- and low- 
 water alarm is shown in Fig. 
 423. There are two floats, one 
 in the steam- and the other in 
 the water-room of a vertical 
 water-column, to which are also 
 attached the gauge-cocks and 
 glass water-gauge common to 
 all combined water-gauges. A 
 bell crank-lever and rod con- 
 nects each float with a whistle- 
 valve ; when the water is at the 
 proper height, the lower float, 
 being submerged, presses up- 
 ward, its steam-valve remaining 
 closed ; if from any cause the 
 water gets low enough to rob 
 the float of its support, it sinks 
 of its own gravity, thus open- 
 ing the valve and blowing the 
 whistle. 
 The high-water alarm is simply the low-water alarm reversed. A 
 bell crank-lever is turned over so that the weight of the float holds the 
 valve closed until the water rises and carries the float up with it, thus 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 365 
 
 opening the whistle-valve. The water cannot pass either the upper or 
 lower limit without automatically blowing the whistle. 
 
 The spherical extension of the main water-column at the bottom is 
 a sediment-chamber, into which all heavy particles fall, to be blown out 
 into the ash-pit through a blow-ofif pipe. The sediment cannot get back 
 into the column when the valve is opened, on account of the contracted 
 neck connecting the sediment-chamber with the column proper. 
 
 Fusible Plugs. — The United States Regulations require that " cyl- 
 inder boilers with flues shall have one plug inserted in one flue of each 
 boiler ; and also one plug inserted in the shell of each boiler from the 
 inside, immediately before the fire-line, and not less than four feet from 
 the forward end of the boiler. All fire-box boilers shall have one plug 
 inserted in the crown of the back connection, or in the highest fire 
 service of the boiler. All upright tubular boilers used for marine pur- 
 poses shall have a fusible plug inserted in one of the tubes at a point 
 at least two inches below the lower gauge-cock, and said plug may be 
 placed in the upper head-sheet when deemed advisable by the local 
 inspectors. All fusible plugs, unless otherwise provided, shall have an 
 external diameter not less than that of a one-inch gas- or steam-pipe 
 screw-tap, except when such plugs shall be used in the tubes of upright 
 boilers. Plugs may be used with an external diameter of not less than 
 that of a three-eighths of an inch gas- or steam-pipe screw-tap, said 
 plugs to conform in construction with plugs now authorized to be used 
 by this Board ; and it shall be the duty of the Inspectors to see that 
 these plugs are filled with Banca tin at each annual inspection." 
 
 A fusible plug as applied to land boilers is ordinarily a brass shell 
 fitted with block-tin. Babbitt-metal, or other metal having its melting- 
 point below the temperature of red-hot iron. These plugs are inserted 
 in the crown-sheet, the upper part of a flue, or such other portion of a 
 steam boiler as will be liable to dangerous overheating in case of low 
 water. In order to fully protect a boiler, the fusible plug may extend 
 upward a short distance inside of the boiler, as shown at Fig. 425. 
 No escape will occur so long as the top of the plug is covered with 
 water, but when the water gets below the upper surface the soft metal 
 melts and runs out of the shell, followed by the steam, which either 
 deadens or completely puts out the fire, or at least serves to warn 
 the fireman that the water has reached the danger line in the boiler. 
 Fusible plugs need to be carefully looked after, as a layer of scale or 
 mud over the top will prevent the escape of steam, even though the 
 metal underneath be melted and gone. 
 
 The insertion of a soft metal rivet, usually of lead, as at Fig. 424, 
 in the crown-sheet or flue of a boiler was formerly much used for this 
 purpose, but has been superseded by the better arrangement of a re- 
 movable brass shell filled with soft metal. Fig. 425. Either of these will 
 be effective if the top is not allowed to become covered with scale. 
 
366 
 
 BOILERS AND FURNACES 
 
 The Bailey plug, shown at Fig. 426, is an improvement over both 
 the others. The body of the plug is permanently fixed. A screw-cap 
 holds the fiisible disk in place. The upper part of this disk is protected 
 
 Fig. 427. 
 
 by a copper cap, shown in the drawing. This cap is intended to prevent 
 the water coming in direct contact with the soft metal, thus maintaining 
 its normal point of fusion. 
 
 Parry's safety-plug is shown in Fig. 427. The fusible metal is inter- 
 posed between two brass shells, which protect it from the water and 
 expose only a line of its surface to the furnace gases. 
 The outer shell and the fusible metal fit against the 
 sheeting in a hollow brass plug screwed into the 
 crown of the furnace, and are held firmly in place by 
 a cotter driven through slots in two lugs on the plug. 
 The entire device can be cleaned both inside and out 
 without any difficulty and can be taken to pieces 
 without a wrench. 
 Gauge- Cocks are for the purpose of ascertaining the water-level in 
 a steam boiler. They are usually three in number and, whenever prac- 
 ticable, should be placed directly in the boiler-shell, the centre one on 
 the proposed water-line, the lower one about an inch above the top of 
 the tubes, and the upper one at that point beyond which it is desirable 
 that the water should not go. 
 
 The Mississippi gauge-cock, shown in Fig. 428, is largely used, and 
 is of the simplest possible construction, being a hollow plug screwed 
 
 Fig. 428. 
 
 into the boiler, through which is a rod terminating in a valve having a 
 bevelled face fitting into a corresponding seat in the end of the hollow 
 plug, the pressure of the steam keeping the valve against its seat. 
 When the valve is pressed off its seat the blow will indicate whether 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 367 
 
 there is water or steam in the boiler at the level of that particular gauge- 
 cock. These gauge-cocks are easily ground without removing them 
 from the boiler. 
 
 The Williams rotating gauge-cock, shown in Fig. 429, is similar to 
 the above, except that it has spiral wings attached to[the valve-stem for 
 
 Fig. 429. 
 
 the purpose of making it self-grinding and self-cleaning. The water 
 escaping from the boiler impinges against these wings, causing a rotary 
 movement to the valve, which prevents it from seating twice in the 
 same place, the rotary motion keeping the valve and seat both clean 
 and tight. 
 
 The Bingham rotating gauge-cock is shown in Fig. 430. It is con- 
 structed on the same principle as the Mississippi gauge-cock, except 
 that the wings are located back of the valve and next to the water in 
 
 Fig. 430. 
 
 the boiler. It is provided with a lever for opening the valve by means 
 of a cord or jack-chain. A rotary motion is given the valve when in 
 use by the spiral wings making it self-grinding and self-cleaning. 
 
 The Register gauge-cock is shown in Fig. 431. It consists of a 
 hollow plug screwed into the water- or steam-space, at the outer end of 
 which is hung a weighted 
 
 lever, having a strip of vul- ^^^- 43i- 
 
 canized rubber immediately 
 over the central orifice of 
 the plug ; the boiler is tested 
 by simply lifting the weight, 
 thus raising the rubber 
 valve off its seat, causing 
 
 the blow. For large boilers this gauge- cock is very convenient, as it 
 can be lifted from its seat by a pole or rod from below, is an excellent 
 form of gauge-cock, and not liable to get out of order, needing only a 
 new strip of vulcanized rubber occasionally to make it good as new. 
 
368 
 
 BOILERS AND FURNACES 
 
 The Reliance gauge-cock, shown in Fig. 432, has its valve-stem 
 passing through a chamber outside of the hollow plug leading to the 
 boiler. One end of this chamber is fitted with a screw-cap. The valve- 
 stem is provided with a collar, between which and the inside end of the 
 
 Fig. 432. 
 
 chamber is a spiral spring of sufficient tension to lift the chain and open 
 the valve when no steam is in the boiler by simply lifting the weight. 
 A right-angled lever with chain and weight keep the valve against the 
 seat. The valve is not closed by the pressure behind it, but by means ol 
 the weight on the long end of the lever. By lifting the weight the com- 
 bined action of the boiler pressure and that of the spring opens the valve, 
 and the water-level is determined. Where these gauge-cocks are made 
 up of sets of three, the levers are of graded lengths, the top one being 
 
 longest, the bottom one short- 
 FiG. 433. est, so that the chains and 
 
 handles hang clear of each 
 other. 
 
 The compression gauge- 
 cock, shown in Fig. 433, is 
 fitted with a disk of soft metal 
 or vulcanized rubber. The com- 
 pression is had by turning the 
 hand-wheel, the movement of the valve being effected by means of the 
 screw attached to the hand-wheel. For small boilers where the gauge- 
 cocks are within reach of the attendant, the compression-cock is the one 
 commonly selected. 
 
 A Water- Gauge consists of an upper and lower angle valve, with 
 a glass tube connecting the two, as shown in Fig. 434. When both 
 valves are open, the water-level in the boiler is indicated by a corre- 
 sponding level in the gauge-glass. It is important that the openings be 
 kept free from obstructions of every kind, otherwise the true water-level 
 will not be indicated. The small cock under the bottom of the valve is 
 for the purpose of blowing through, to ascertain whether or not the 
 openings in the boiler are free. If not free, the true water-level will not 
 be indicated in the glass. If the bottom opening is completely ob- 
 structed, the steam will blow through, and the water will not rise to its 
 proper level upon closing the pet-cock. In the event of a glass breaking, 
 the bottom- or water-valve should be closed first, and then the steam- 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 369 
 
 valve. The ends of the broken 
 glass tubes may now be removed 
 and a new glass inserted, with new 
 packings in the stuffing-boxes. 
 
 Glass tubes for water gauges 
 must be made of clear, transparent 
 glass and of considerable tough- 
 ness. Those most in use, prob- 
 ably, are the imported Scotch 
 tubes, easily known by a certain 
 fibrous appearance in the glass 
 lengthwise of the tube. Occa- 
 sional breakage of glass tubes is 
 to be expected, owing to the brit- 
 tle nature of the material and to 
 the unequal expansion incident to 
 the action of the hot water and 
 steam inside and the cold air 
 on the outside. As glass tubes 
 never give previous warning of 
 failure, spare tubes should always 
 be kept on hand. 
 
 The lower valve and fittings of 
 a glass water-gauge are shown in 
 Fig. 435 ; the pet-cock underneath is 
 valve under control of the hand- wheel 
 
 Fig. 435. 
 
 Fig. 434. 
 
 for the purpose of blowing out. A 
 on the outside opens or closes com- 
 munication with the boiler 
 as desired. The glass water- 
 tube need not project below 
 the bottom of the stuffing- 
 box farther than is shown 
 in the drawing. This box 
 is commonly filled with a 
 soft packing, usually ot 
 plaited lamp wick. 
 
 Upon the breakage of 
 a water-glass, especially if 
 the fittings are at a con- 
 siderable height, it is not 
 only difficult, but sometimes 
 dangerous, to approach the 
 gauge for the purpose of 
 shutting ofi" the water and 
 steam blowing into the fire- 
 room. Several devices have 
 
370 
 
 BOILERS AND FURNACES 
 
 been brought out having an automatically closing valve which will remain 
 open so long as the glass is whole, but which will immediately close upon 
 the breakage of a tube. Such a fitting is shown in Fig. 436. The lower 
 
 of the two valves shown is a 
 
 Fig. 436. 
 
 r^ 
 
 metal ball resting against a 
 pin to prevent its rolling 
 into the boiler or into the 
 barrel of the water-gauge 
 fixture. So long as there is 
 no circulation in the water- 
 glass the ball-valve will re- 
 main in the position shown, 
 but as soon as a current 
 is set up, whether by the 
 water or steam, the ball- 
 valve will roll forward and 
 close the orifice leading to 
 the water-glass, remaining 
 there as long as the pressure 
 continues behind it. The 
 upper one of the two valves shown is of a different type, being a mitred 
 valve with wings arranged spirally about its body. The steam and 
 water will pass through the spiral openings and indicate the true water- 
 level in the glass. Should this glass break, the pressure within will 
 force the valve up against the seat and prevent the flow of either steam 
 or water into the fire-room. Once this valve is closed, it requires to be 
 opened from without. A ready means for opening this valve is had by 
 projecting the valve-rod through the seat and forcing the valve back 
 against a pin shown in the drawing. It will be understood that in this 
 illustration the valve controlled by the handle should be withdrawn 
 further than appears in the drawing ; if not, it would be impossible for 
 the inside valve with the spiral wings to seat itself, and would not, 
 therefore, prevent the flow of either steam or water into the fire-room. 
 
 Glass tubes may be cut to any desired length by using a 6-inch half- 
 round fine-cut Stubb's file. Place the glass tube on an even surface, 
 such as a flat board ; then place the sharp edge of the file at the point 
 at which the tube is to be cut and bear on lightly at first, rolling the 
 tube back and forth by pushing and pulling on the file, having the 
 cut or mark meet around the tube ; bear on a little harder as the cut 
 grows deeper and roll until the tube flies apart. Tubes thus cut have 
 square ends, and when properly managed need never result in failure, 
 even for short lengths. Another method, and a good one, is to take 
 a piece of steel wire ; sharpen at one end and bend that end into a 
 right-angle hook and harden same ; run this hook into the inside of the 
 glass and make a scratch around the interior at the point where the 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 371 
 
 break is to be made. The glass can then be readily broken without 
 cracking. 
 
 There is a want of uniformity in the matter of locating the glass 
 water-gauge fixtures with reference to top of tubes, fire-box, or other 
 heating surface liable to be exposed in case of low water. A glass 
 water-gauge ought to show when the water is near or, perhaps, at the 
 danger limit. In case of low water it is particularly important that the 
 exact height of water above the line of heating surface be definitely 
 known. A permanent marker should be attached to the lower glass 
 gauge fixture to accurately indicate the top of the highest heating sur- 
 face inside the boiler. The height of water above this line is the one 
 important fact the fireman must have constantly before him. 
 
 A Combined Water-Gauge consists of a barrel to which are 
 attached the gauge-cocks and the glass water-gauge. This barrel has 
 two pipes connecting with the steam and water portions of the boiler ; 
 these pipes are of considerably larger diameter than would ever be used 
 with any of the fittings taken singly. The lower pipe being removed 
 from the path of circulation, there is less disturbance of water-level in 
 the barrel of the combined water- 
 gauge during foaming than occurs F^^- 437- 
 within the boiler itself The illus- 
 tration. Fig. 437, shows the glass 
 tube and the gauge-cocks oppo- 
 site each other ; they are not 
 usually thus placed, but at right 
 angles to each other. 
 
 There are no fixed proportions 
 for water-gauge barrels. They 
 vary from 2}^ to 5 inches in di- 
 ameter, with glass gauges ranging 
 from 12 to 20 inches in length. 
 The smallest barrels are not often 
 fitted with less than ^-inch pipe 
 connections to the boiler, the 
 larger sizes being tapped for i- 
 inch and i^-inch wrought- iron 
 pipe. Stationary-engine boilers 
 more than 48 inches in diameter 
 should have water-gauge barrels 
 not less than 4 inches in diameter 
 and should have pipe connections 
 
 not less than i^ inches. The gauge-glass for such a barrel need not 
 be more than 12 inches between stufiing-boxes. 
 
 The piping of a water-column must be properly done or the apparatus 
 will fail to indicate the true water-level in the boiler. The steam-pipe 
 
372 
 
 BOILERS AND FURNACES 
 
 should lead from a point above the highest gauge-cock and the water 
 connection to a point below the lowest gauge-cock. The drainage ot 
 these pipes must be carefully attended to, and must be arranged to drain 
 completely dry, either into the boiler or into the water-column. No 
 water-pockets or trapping must be permitted. A blow-ofif and drain- 
 pipe should lead from the bottom of the water-column into the ash-pit ; 
 this pipe will be useful when blowing out any sediment which might have 
 been carried into the water-barrel. The bottom gauge-cock should be 
 placed at what is to be considered the danger line. In Fig. 434 it is 
 shown at the upper side of the top row of tubes. This is too low down ; 
 it should be at least i inch above for boilers 46 inches in diameter or 
 less, and at least i^ inches above for 48-inch and larger boilers. The 
 middle gauge-cock should be placed on the water-line at which the 
 boilers are intended to work ; this may be from 2 to 4 inches above the 
 lower gauge-cock, depending upon the diameter of the boiler. The 
 upper gauge-cock is usually placed at the same distance above the 
 central gauge-cock that it is above the bottom one. 
 
 Safety Water- Columns are those in which a signal is given auto- 
 matically to warn the attendant when the water in the boiler is either 
 more or less than the established water-line, and for which provision 
 was made in the apparatus itself The sectional elevations. Figs. 421, 
 422, 423, show the interior arrangements of such safety water-columns 
 as are in common use. The barrel includes the gauge-cocks and glass 
 water-gauge usual in all water-columns. In addition to these there are 
 devices arranged for operating a small steam whistle. 
 
 The Piping of Water-Col- 
 umns, and, in fact, all the piping 
 for a steam boiler where subject to 
 an accumulation of sediment, should 
 have crosses fitted with plugs at the 
 turns rather than elbows. By the 
 removal of a plug opposite the pipe 
 likely to accumulate sediment, an 
 iron bar or scraper can be forced 
 
 Fig. 439. 
 
 C 
 
 Fig. 438. 
 
 3 
 
 through and the pipe kept open. If globe valves are used, they should 
 be placed with the spindle horizontal instead of vertical in all steam-pipes 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 373 
 
 likely to trap the condensed water. The effect of this condensation 
 and trapping is shown in Fig. 438, and the non-accumulation of water 
 by placing the valve-spindle horizontal in Fig. 439. 
 
 A Steam-Pressure Gauge is an instrument for indicating the 
 pounds pressure per square inch in a steam boiler above that of the 
 atmosphere. 
 
 The Schaeffer diaphragm pressure-gauge is shown in Fig. 440. This 
 gauge is the pioneer of all spring gauges, being the first one to super- 
 sede the old-style mercury gauges. It consists of a corrugated steel 
 diaphragm secured between two concaved flanges. The fluid pressure is 
 admitted underneath this diaphragm, which will, by reason of the flexi- 
 bility allowed by its corrugations, permit a central rise, the extent of 
 which is determined by the amount of pressure and the stiffness of 
 the diaphragm. A central rod connects this diaphragm with a sector 
 attached to the cylindrical case above. This sector has teeth which 
 
 engage a small pinion, the shaft of which 
 Fig. 440. has fastened to it one end of a hair- 
 
 spring, the other end of the spring 
 being fastened to the case. This hair- 
 
 FiG. 441. 
 
 spring brings the pointer back to the zero mark on the dial when the 
 pressure is removed. To this pinion-shaft is also attached a hand or 
 pointer for indicating the pounds pressure per square inch on a circular 
 dial, not shown in the engraving. 
 
 The commonest form of pressure-gauge now in use is a modification 
 of the Bourdon gauge. This originally consisted of an oval tube bent 
 as in Fig. 441. One end of this tube is fastened to a fixture at the 
 bottom of the case, the other end is free to move according to the 
 pressure within it. Steam-gauges thus constructed are objectionable in 
 one respect, — the free end of the tube, nearly one-half of it, is below the 
 line of drainage, which is liable to freeze and injure the tube ; even if 
 
374 
 
 BOILERS AND FURNACES 
 
 it does not burst it, the accuracy of the gauge will be seriously im- 
 paired, and, in consequence, it will be utterly worthless for the purpose 
 intended. 
 
 Gauges are now made with two oval tubes, as in Fig. 442. As these 
 tubes never exceed a semicircle, a perfect drainage is had at any time 
 by simply opening the drainage-cock below the case. An important 
 advantage is had in another respect. These tubes, being shorter, are 
 less sensitive to vertical shocks, as in locomotive service. The lever 
 and rack being attached to the free end of the tubes, and not pivoted to 
 the case, the horizontal vibrations of the free ends of the tubes are not 
 violently transmitted to the pointer, as they were when a single tube 
 was employed and the quadrant pivoted to the case. 
 
 Each boiler should have its own steam-gauge. The connection 
 should be with the steam-room of the boiler direct, and not with the 
 steam-pipe leading to the engine. No steam-gauge should be used with- 
 out a siphon between it and the boiler to protect the bent tubes from 
 
 Fig. 442. 
 
 Fig. 443. 
 
 Fig. 444. 
 
 expansion by heat. The siphon interposes a body of water between the 
 steam and the gauge which fully protects the latter from injury. The 
 simplest form of siphon is a bent pipe, shown in Fig. 443. Such a pipe 
 cannot drain dry. It must, therefore, be protected from the frost or fitted 
 with a pet-cock to completely empty it when not in use in cold weather. 
 When a steam-gauge is piped, as shown in Fig. 444, the loop thus 
 formed will fill with water of condensation and protect the gauge without 
 a siphon attachment or the formation of other water-pocket. The inser- 
 tion of a pet-cock in the lower opening of the bottom tee will afford 
 complete drainage. There are a number of combined siphons and stop- 
 cocks in the market, of which Fig. 445 may be taken as representative 
 of the class. The section shows a loose cap over the central pipe which 
 extends into the chamber. This deflects the entering steam, which, 
 condensing in the chamber, effectually prevents the live steam reaching 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 375 
 
 Fig. 446. 
 
 the spring of the gauge. The cap over the pipe falls as the pressure is 
 removed, making a siphon which empties the water from the chamber, 
 thus preventing danger of bursting from the action of frost. 
 
 Shaw's mercury gauge is shown in sectional elevation in Fig. 446. 
 The only piece of moving mechanism in the gauge is a double-headed 
 piston inserted between two flexible diaphragms. This piston has ends 
 of unequal areas. The larger piston is 
 placed at the top, or measuring- chamber, 
 the smallest one at the bottom, or enter- 
 ing-chamber, this latter communicating 
 with the boiler or other vessel containing 
 the fluid pressure to be measured. The 
 steam pressure acts on the lower dia- 
 phragm, forcing it upward, carrying the 
 double-headed piston with it. The larger 
 end of this piston communicates to its 
 diaphragm the same upward movement. 
 
 Fig. 445- 
 
 This upward movement has the effect to diminish the volume of the 
 upper chamber, thereby forcing its contained mercury into the vertical 
 glass tube, where it records measurements in pounds per square inch by 
 means of a graduated scale, the divisions of which correspond to the 
 applied pressures. It will be seen that the principle of action is that of 
 differential areas, analogous to the short and long arms of a lever, — a 
 difference of 10 to i, for example, in area of the gauge-pistons, and their 
 corresponding diaphragms give the same result as 10 to i in a lever. 
 In either case the employment of one pound on the long end of the lever 
 or on the larger area in the gauge will balance ten pounds on the short 
 end of the lever or the small area of the gauge. 
 
376 
 
 BOILERS AND FURNACES 
 
 Dry Pipe. — Such a pipe is sometimes attached to the interior of 
 stationary steam boilers not provided with a steam-dome or steam-drum. 
 It consists of a pipe with closed ends, located inside of the boiler in the 
 steam-space and close to the upper side of the shell, as in Fig. 447, which 
 shows a dry pipe suspended from the shell, usually near the rear end of 
 
 Fig. 447- 
 
 -t^ 
 
 the boiler. Fig. 448 shows a dry pipe leading out of the boiler through 
 one of the heads. The drawing does not show it, but the flanges at the 
 head should be so bolted that the steam-pipe may be wholly detached 
 without disturbing the flanged joint of the dry pipe next the boiler- 
 head. There is no rule for either diameter or length, so that both 
 dimensions are widely variable in practice, diameters varying from three 
 to six inches and from one-eighth to seven-eighths that of the boiler 
 
 Fig. 448. 
 
 for length, but one-fourth the length of the boiler is not far from the 
 average. The upper surface of the pipe is drilled or slotted with holes 
 aggregating a little less than the area of the stop- valve ; one small hole 
 should be drilled in the bottom at each end for drainage. By having a 
 considerable number of holes distributed over a large pipe surface in a 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 377 
 
 direction different from that taken by the rising water when priming, 
 but httle water will be carried over by the steam into this pipe, and dryer 
 steam will be supplied than if this dry pipe was not attached. 
 
 Edgerton's separator for steam boilers is shown in Fig. 449. At 
 any convenient place on the top of the boiler-shell is riveted a pipe- 
 flange, in the lower 
 
 half of which is a ^i^. 449. 
 
 nipple extending into 
 the boiler. On the 
 bottom of this nipple 
 is screwed the sepa- 
 rator, which consists 
 of three pans of the 
 design shown in the 
 engraving, the lower 
 pan having its flange 
 turned upward, the 
 upper pan having its 
 flange turned down- 
 ward ; intermediate between these two is the third pan. The distance 
 apart at which these pans are to be set is fixed by wrought-iron ferrules 
 on the outside of the bolts which hold the pans together. The course 
 of the steam is around and downward over the flange of the lower 
 pan ; then, changing direction, it passes upward around and inside of 
 the flange of the upper pan, and from thence to the steam-pipe and 
 beyond. Small holes are drilled in the two lower pans, that any water 
 of condensation may drop back into the boiler. 
 
 Steam Stop- Valves. — The diameter of steam outlets for horizontal 
 tubular boilers is commonly 2 inches for a 36-inch boiler up to 6 inches 
 for a 72-inch boiler. The least size of opening in a steam boiler should 
 be such that the velocity of steam issuing through the steam-pipe shall not 
 exceed 100 feet per second when the boilers are worked up to their limit. 
 
 A flange should be riveted to the top of the boiler-shell with which the 
 steam-pipe is to be connected. This flange should be located preferably 
 at least as far back as the centre or between the centre and the rear end 
 of the boiler rather than immediately over the furnace. If a dry pipe 
 is used, the steam outlet is, of course, wherever the dry-pipe outlet may 
 be fixed by the designer of the boiler. The steam stop-valve should 
 always be next to the boiler, so that if for any reason one boiler out 
 of a battery of several is to be withdrawn from service, the valve will 
 completely shut off the steam from such boiler. 
 
 A stop-valve must never be placed underneath a safety-valve if the 
 closing of the former also closes communication with the latter ; but if 
 for any reason it is imperative that a stop-valve and safety-valve be 
 connected with the same opening, then a combined valve, such as shown 
 
 25 
 
378 
 
 BOILERS AND FURNACES 
 
 in Fig. 415, may be used, or a tee attached to the boiler with the safety- 
 valve on the top and stop-valve on the horizontal branch. 
 
 Globe valves threaded for wrought-iron pipe are in very general use 
 for steam-piping. These are usually of gun-metal for sizes less than 
 four inches in diameter, and cast-iron bodies with gun-metal mountings 
 for larger sizes. A globe valve is shown in section in Fig. 450, having 
 a metal valve loosely attached to a screwed spindle, by which it may be 
 opened or closed by turning the hand-wheel attached to the same 
 spindle. The valve is loosely fitted to the spindle, that it shall be free to 
 revolve around the stem and to secure a flexible joint favoring complete 
 contact with the seat when screwed down. Leaky globe valves when 
 fitted with hard-metal valves may be restored by regrinding. For grind- 
 ing hard brass moderately fine emery may be used for the preliminary 
 grinding, to be followed by and finished with powdered glass. 
 
 Fig. 451. 
 
 Soft metal disks, as well as those made of vulcanized rubber, are 
 now much used as valve-faces for stop-valves. These have undergone 
 the test of many years and have stood it well ; such a disk is shown 
 by double-hatching in the angle valve. Fig. 451. These disks, of hard 
 rubber or soft metal, are renewable ; a new disk can be easily applied 
 by unscrewing the upper half of the valve, taking out the worn-out 
 disk, and replacing it by a new one. The angle valve, just referred 
 to, is a good form of valve to use on top of a boiler, because there are 
 no pockets for tlie accumulation of water of condensation. 
 
 The Eastwood stop-valve, shown in Fig. 452, is not unlike an ordi- 
 nary globe valve in its general features, but differs in the details of the 
 valve-seat, which is bevelled outward. The valve-face is a copper disk 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 379 
 
 fitting into a recess in the valve to give it a substantial backing. The 
 inner face of this copper disk and the outer one of the valve-seat are 
 ground to a tight joint. In the event of repairs incident to long wear 
 or through any other cause a new valve-face is required, it can be easily- 
 furnished by unscrewing the old copper disk and inserting a new one 
 to take its place. 
 
 Flanged stop-valves are not commonly employed for steam-pipes 
 less than four inches in diameter. The body is ordinarily made of cast 
 iron, into which is screwed a hard gun-metal seat. Such a valve is 
 shown in Fig. 453. In this case the seat has radial arms, with a central 
 
 Fig. 452. 
 
 boss for guiding the spindle attached to the upper valve. The valve is 
 made of hard gun-metal, reinforced by a cast-iron back, clearly shown 
 in the illustration. The collar at the lower end of the spindle for raising 
 and lowering the valve is convex, and, being loosely fitted in the back 
 of the valve, with a gland above it, allows perfect freedom of movement 
 between the stem, valve, and the valve-seat ; the valve can, therefore, 
 be closed without any lateral strains. Large stop-valves are usually 
 fitted with a yoke extending above the bonnet for the purpose of fixing 
 a nut, through which a screwed spindle passes for operating the valve- 
 disk within. This arrangement prevents any injurious action of the 
 steam upon the screw-threads by removing them completely from the 
 interior of the valve. 
 
 A combined stop- and check-valve for steam-pipe is shown in Fig. 
 454. This valve is so designed that the valve is closed when the press- 
 ure in the boiler is below that of the steam main, although it may be 
 positively closed at all times if desired. This check-valve principle is, 
 of course, a decided advantage in case of an accident to one of the 
 
38o 
 
 BOILERS AND FURNACES 
 
 Fig. 454. 
 
 boilers, the valve closing and preventing a loss of pressure in the 
 mains and the liability of crippling the entire plant. Referring to 
 the sectional view of the valve, it may be 
 seen that the valve-disk is free to move 
 within certain limits upon the stem. The 
 upper part of the valve-disk is bored out 
 and fitted with a cover which screws on. 
 The valve-stem passes through a hole in 
 the cover and through the lower part of 
 the valve and into a spider-bearing or guide 
 at the throat of the valve. A piston con- 
 taining two rings is screwed upon the stem, 
 the periphery of the rings and piston bear- 
 ing against the inner surface of the valve- 
 disk as shown to prevent it from chattering, 
 the whole acting like a dash-pot. 
 
 Expansion-Joints. — One effect of heat 
 upon iron is to increase its volume ; expan- 
 sion, then, especially in the case of long 
 steam-pipes, must not be overlooked. The 
 expansion of iron pipe for each 100 feet in 
 length from the freezing temperature will approximate as follows : 
 ... 32° 212° 259' 
 
 287° 307° 324° 335° 
 o 20 40 60 80 100 
 1.44 1.82 2.04 2.20 2.34 2.42 
 
 Fig. 455. 
 
 Temperature, Fahr 
 
 Steam pressure above atmosphere . 
 Elongation in inches 
 
 It will be seen that the linear expansion of steam-pipes must be 
 provided for or leaky joints, perhaps broken fittings, will result. 
 
 Steam-pipes fitted with expansion-joints should be securely anchored 
 somewhere near the extreme ends of the pipe, leaving the centre free 
 to expand and contract according to the tem- 
 perature of the steam. The necessity for 
 anchoring at each end is owing to the fact that 
 the two ends of a common expansion-joint are 
 not bolted together. The internal area of the 
 pipe multiplied by the steam pressure gives 
 that pressure exerted at each end of each pipe 
 tending to force them apart. 
 
 A copper pipe bent as shown in Fig. 455, 
 with flanges joining the ends of wrought-iron 
 pipe, is now used with satisfactory results. The 
 distance from face to face of the flanges may 
 be 3 feet for 2-inch pipe up to 7 feet for 6-inch 
 pipe ; the distance from the centre of the line 
 
 of wrought-iron pipes to the centre of the bent copper pipe at its highest 
 point in the bend may be 18 inches for a 2-inch pipe up to 48 inches for 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 381 
 
 a 6-inch pipe. The radius for bending copper pipes should not be less 
 than 12 inches for a 2-inch pipe nor less than 30 inches for a 6-inch pipe, 
 and both of these should be greater if space will permit. Dimensions 
 for pipes of intermediate diameters to the above can be had by simple 
 interpolation. 
 
 The expansion-joint shown in Fig. 456 is probably in more general 
 use than any other. The engraving shows flanged ends, and this detail 
 
 Fig. 456. 
 
 v/////////////////^^^^^ 
 
 is recommended, though for sizes up to 2^ inches screwed ends are 
 commonly used when the expansion-joint is made of brass. The ordi- 
 nary traverse of the slip-joint for a 2-inch pipe is 2^ inches, and for a 
 6-inch pipe it is 5 inches. Special traverse greater than the above can 
 be had when ordering an expansion-joint, but the above will cover all 
 ordinary requirements. 
 
 Pearson's expansion-joint is shown in Fig. 457. The steam mains 
 are anchored in two places, and at a point midway is placed the expan- 
 sion-joint, which was designed to eliminate the strain on anchorages 
 
 Fig. 457. 
 
 r W/M^^^^^^M^^^MJ^m 
 
 common to the usual form of cylindrical slip-joint, which the pressure 
 on the ends of a pipe containing an ordinary expansion-joint tends to 
 pull apart. The entering pipe in this joint slips through stuffing-boxes 
 
382 
 
 BOILERS AND FURNACES 
 
 made in the ordinary manner. One end of this slip-pipe is open for 
 connecting with the steam-pipe, the other end is fitted blank. Openings 
 are made through the slip-pipe, as shown in the engraving, permitting 
 a flow of steam into the branched pipe carrying the stufling-boxes, the 
 two branches being united at the further end into a common opening 
 and flanged, to connect with the steam-pipe to any point beyond. As 
 there are no free ends in this expansion-joint open to the atmosphere, 
 it will be seen that this arrangement constitutes a balanced joint. 
 
 Smith's balanced expansion-joint, shown in Fig. 458, consists oi 
 three parts : First, a sliding pipe having two diameters, the larger di- 
 ameter being of a short length, making it similar to a piston centrally 
 located on a large hollow piston-rod. The sliding pipe corresponds to 
 
 Fig. 458. 
 
 the normal diameter of the steam-pipe to which it is attached and ol 
 which it is a continuation. The larger or piston diameter is such that 
 its area is just double that of the sliding pipe, measured by its outside 
 diameter. Second, a sleeve bored to fit both the piston and the sliding 
 pipe ; this sleeve is provided with a stuffing-box at each end. There 
 are openings in the sliding pipe by which steam has access behind the 
 piston, the pressure tending to force the piston outward through the 
 large stufling-box. Third, immediately outside of the piston stuffing- 
 box is another bored sleeve accurately fitting the sliding pipe. This 
 sleeve is also provided with a stuffing-box, and is held by a suitable 
 flange at a fixed distance from the opposite or piston stuffing-box by 
 means of distance pieces made of iron pipe, through which pass the 
 bolts for holding both sleeves rigidly together. The relative areas ot 
 these two openings into the atmosphere are as 2 to i. The rear end ol 
 this sleeve is fitted with a flange for connection with steam-pipe. 
 
 In an ordinary expansion-joint the steam exerts a pressure equiva- 
 lent to the outside area of the sliding pipe multiplied by the steam 
 pressure. This must be resisted by an abutment somewhere in the pipe 
 system, or the sliding pipe will be blown out of its stuffing-box. In the 
 expansion-joint now under consideration the steam pressure is balanced, 
 because the tendency of the piston to move in one direction is offset by 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 383 
 
 that of the sliding pipe to move in the opposite direction, both areas 
 being alike in extent and subjected to the same pressure, the two forces 
 thus counterbalancing each other. This joint is thus free to move ac- 
 cording to the expansion of the pipes to which it may be attached, but 
 is in equilibrium so far as internal steam pressure is concerned. The 
 movement, then, of the sliding pipe through the stuffing-boxes will only 
 be such as is due to the expansion of the two pipes thus joined, which 
 amount depends upon the temperature of the steam and the length of 
 the pipe. The expansion-joint herein illustrated was designed for a 
 working steam pressure of 225 pounds per square inch, the steam-pipe 
 being 6 inches in diameter, the joint permitting a maximum travel of 3 
 inches. 
 
 Damper. — This may be located at any convenient place between 
 the exit of the gases from the boiler setting and the chimney, or in some 
 cases it is placed in the chimney itself. Two forms of dampers are 
 shown. Fig. 459 represents the sliding form and Fig. 460 the butterfly 
 
 Fig. 459. 
 
 ^'.^^^'v','s','v\^'.^V^^'-^V 
 
 Fig. 460. 
 
384 
 
 BOILERS AND FURNACES 
 
 type. The latter is generally preferred because of its ease of operation, 
 there being no sliding friction, consequently little liability for it to stick 
 fast or become otherwise inoperative. 
 
 Regulating the furnace draught by a damper is better than doing it 
 with the ash-pit doors, unless the latter are absolutely air-tight. The 
 reason needs simply to be suggested. Combustion can only proceed so 
 long as the fire receives fresh accessions of oxygen ; by closing the 
 damper the non-supporting products of combustion completely envelop 
 the fuel and combustion stops ; no air can reach the fire, because it can- 
 not dislodge the gases there present. If, on the other hand, the damper 
 be wide open, and the ash-pit doors as ordinarily constructed be closed, 
 the chimney draught will carry off the products of combustion above the 
 fire and form a partial vacuum in the ash-pit. The ash-pit doors, not 
 being absolutely air-tight, permit a leakage, and combustion goes on 
 the same as usual, except as to intensity, governed by the quantity of 
 air thus admitted. 
 
 An Automatic Damper Regulator is advantageous if the auto- 
 matic device is a good one ; it will furnish a more uniform boiler pressure 
 than is likely to be had by hand regulation. Automatic damper regu- 
 lators are all constructed on one general principle, the boiler pressure 
 acting upon a flexible diaphragm or a piston, which, by suitable con- 
 nections, operates the damper, opening or closing it according to the 
 steam pressure at which the damper regulator is intended to work. 
 Among the earlier inventions for regulating the draught by the pressure 
 of steam in the boiler was that of Clark (1854). This damper regulator 
 was open to some objection and was improved upon some twenty years 
 
 Fig. 461. 
 
 later by Le Van, whose regulator is shown in Fig. 461, the improve- 
 ment consisting in a protection offered the flexible diaphragm from the 
 injury which it sustained by the direct contact of the steam. There is 
 combined with the cylinder or pressure-chamber of the diaphragm a 
 water-chamber, a steam-supply pipe connected near its top, a water- 
 pipe extending from the lower surface of the diaphragm to the interior 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 385 
 
 of the water-chamber at a point below the opening of the steam-supply 
 pipe. 
 
 In the operation of the device it is obvious that the water of con- 
 densation from the steam-supply pipe will be forced up into the pro- 
 jecting pipe and against the diaphragm by the steam pressure acting 
 upon the surface of the water. This method of operation effectually 
 prevents the direct contact of the steam with the diaphragm, because 
 there is always a body of water interposed between the diaphragm and 
 the steam. Should the pressure in the boiler rise above the desired 
 limit, the weighted lever will rise, and by means of a chain or rope 
 leading to the damper close the latter and keep it closed until the 
 pressure of steam falls below its assigned limit, after which the lever 
 will fall and, by reversing the direction of the chain or rope, will open 
 the damper and permit a more rapid rate of combustion. 
 
 Hydraulic Damper- Regulator. — The engraving. Fig. 462, repre- 
 sents the Mason regulator for controlling a damper by the variation 
 
 of boiler pressure, but the 
 '* * motive-power employed in 
 
 opening or closing the damper 
 is water pressure from the 
 street main, from an over- 
 head tank, or from the boiler 
 itself. The advantage of using 
 water pressure for damper 
 regulation is the constant and 
 non-variable movement ob- 
 tained. In this regulator the 
 steam from the boiler enters 
 through a connection shown 
 to the left of the base and 
 passes into a chamber in the 
 bottom of the base and under 
 a heavy rubber dia- 
 phragm. The steam 
 pressure forces the 
 \\~o) diaphragm up- 
 ward ; this upward 
 pressure is counter- 
 balanced by a heavy 
 weight placed on 
 the lower horizontal 
 lever ; this weight can be moved out or in to suit any boiler pressure. 
 When the boiler pressure rises above the normal, at which the weighted 
 lever is set, the lever is lifted upward, carrying with it a vertical valve- 
 rod, shown attached to the weighted lever at the bottom and the com- 
 
386 
 
 BOILERS AND FURNACES 
 
 pensating lever at the top, this movement opening the water-valve and 
 allowing the water pressure to pass through the ports of the valve- 
 chamber to the top of the main piston in the cylinder forming the cen- 
 tral part of this apparatus. This piston by its downward movement 
 winds up a chain to which the damper is attached, thereby closing the 
 damper. As the chain-wheel travels around, the compensating arm 
 
 shown at the top of the appa- 
 FiG. 463. ratus is thrown outward by a 
 
 cam attached to the chain- 
 wheel. Raising one end of 
 the lever fulcrumed on the 
 small valve-rod at the top 
 tends to close the port of the 
 water-valve. This serves as 
 a compensating arrangement, 
 and does not allow the damper 
 to be entirely closed on slight 
 changes of pressure. The re 
 verse action takes place when 
 the boiler pressure drops and 
 the water pressure is shut off. 
 A weight is placed on the 
 damper in the flue heavy 
 enough to draw the wheel 
 back again to its first position. 
 The bottom of the water-valve 
 chamber and the main cylin- 
 der are connected with a pipe 
 which carries away whatever 
 water remains after the piston 
 has worked in either direction, 
 also shown in the engravings. 
 Whistle.— Nearly all lists of boiler furnishings include a steam 
 whistle. Fig. 464 is a sectional drawing representative of steam whistles 
 as a class. A whistle as shown consists of a sounding-bell attached to 
 a spindle by the base, in which is a thin annular orifice for the escape of 
 the steam into the atmosphere, and a valve, usually included in the base, 
 for regulating or controlling the pressure of steam escaping from the base. 
 The valve is opened by the action of the lever forcing it off its seat 
 against the steam pressure. The spring on the back of the valve keeps 
 the latter closed and counterbalances any tendency of the lever to keep 
 the valve open when there is no pressure behind it. The central spindle 
 rising vertically from the base has screw-threads at its top, on which is 
 screwed the sounding-bell. The object of this screw is to secure a ver- 
 tical adjustment of the bell suited to the steam pressure and tone desired. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 387 
 
 Fig. 464. 
 
 The sound from a steam whistle is produced by vibrations set up injthe 
 bell incident to the action of the steam escaping into the atmosphere 
 from a thin annular opening in the 
 base of the whistle. A vibrating 
 body, before it can act as a sounding 
 body, must produce alternate com- 
 pression and rarefaction in the air, 
 and these must be well marked. 
 Relatively deep, grave sounds are 
 produced by slower vibrations. The 
 larger the diameter and the greater 
 the height of the bell, the deeper or 
 lower will be the note sounded. 
 Small diameters and short heights of 
 bell produce a shrill note, and be- 
 tween these two may be had any tone 
 desired. 
 
 Malleable - Iron Unions are 
 much used for coupling together such 
 wrought-iron pipes as are likely to 
 afterwards require disconnecting. 
 The sectional elevation, Fig. 465, 
 shows an ordinary union. It consists 
 of a collared fitting which screws on 
 one of the pipes to be joined. On 
 this fitting is an internally threaded 
 coupling free to revolve about the 
 under side of the collar, against which it makes close contact. The 
 upper fitting is internally and externally threaded, the inside thread for 
 the pipe it is to connect, the outside thread fitting into that of the loose 
 coupling below. A vulcanized rubber or soft metal washer makes a 
 
 1 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 D 
 
 Fig. 465. 
 
 Fig. 466. 
 
 Fig. 467. 
 
 flexible seating for the two halves of the union, which are drawn together 
 by the loose coupling, the exterior of the latter being provided with 
 hexagon or octagon sides for the use of a wrench. 
 
 The Eastwood ground-joint union is shown in Fig. 466. This union 
 has a conical joint which can be ground tight and dispense with packing. 
 
388 
 
 BOILERS AND FURNACES 
 
 An elbow union is shown in Fig. 467. It is much used in steam- 
 heating practice, and deserves a wider adoption in the smaller sizes of 
 steam- and water-piping, such as for pumps, injectors, etc. 
 
 Steam-Pipe. — Lap-welded wrought-iron pipes with cast-iron fit- 
 tings are almost exclusively employed in the piping of steam-plants. 
 Such pipes are made in all needed diameters and are uniform in all work- 
 ing dimensions. The following. Table L., gives the standard dimen- 
 sions for nominal diameters from }i inch to 12 inches : 
 
 TABLE L. 
 
 STANDARD DIMENSIONS OF WROUGHT-IRON PIPE. 
 
 
 Diameters. 
 
 
 Circumfer- 
 ence. 
 
 Area. 
 
 S . 
 
 i 
 
 Threads. 
 
 1 
 
 
 
 
 
 
 
 
 si 
 
 0: 
 
 S. ^ 
 
 H 
 
 1^- 
 
 >> 
 
 
 ■3 
 
 i 
 
 
 -5 
 
 
 -• 
 
 ad 
 
 
 
 u "^ 
 
 ; u 
 
 ^l 
 
 
 5 
 
 a 
 
 a 
 
 1 
 
 
 I 
 
 1 
 
 
 
 111 
 
 
 % 
 
 
 
 
 
 w 
 
 H 
 
 a 
 
 W 
 
 a 
 
 W 
 
 J 
 
 ^ 
 
 'Z 
 
 i 
 
 ^ 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Sq. In. 
 
 Sq. In. 
 
 Feet. 
 
 Lbs. 
 
 1 
 
 ns. 
 
 Ins. 
 
 ys 
 
 .27 
 
 .41 
 
 .068 
 
 .85 
 
 1.27 
 
 .06 
 
 •13 
 
 2500 
 
 .24 
 
 27 
 
 19 
 
 ^ 
 
 H 
 
 .36 
 
 .54 
 
 .088 
 
 1. 14 
 
 1.70 
 
 .10 
 
 -23 
 
 1385 
 
 •42 
 
 18 
 
 29 
 
 <f 
 
 H 
 
 .49 
 
 .68 
 
 .091 
 
 1-55 
 
 2.12 
 
 •19 
 
 .36 
 
 751-5 
 
 .56 
 
 18 
 
 30 
 
 n 
 
 H 
 
 .62 
 
 .84 
 
 .109 
 
 1.96 
 
 2.65 
 
 -31 
 
 -55 
 
 472.4 
 
 .85 
 
 14 
 
 39 
 
 H 
 
 'X 
 
 .82 
 
 1.05 
 
 •113 
 
 2.59 
 
 3- 30 
 
 •53 
 
 .87 
 
 270.0 
 
 I.I3 
 
 14 
 
 40 
 
 H 
 
 I 
 
 1.05 
 
 1.32 
 
 •134 
 
 3-29 
 
 4-13 
 
 .86 
 
 1.36 
 
 166.9 
 
 1.67 
 
 11/2 
 
 51 
 
 it 
 
 iX 
 
 1.38 
 
 1.66 
 
 .140 
 
 4-34 
 
 5-22 
 
 1.50 
 
 2.16 
 
 96.25 
 
 2.26 
 
 IlK 
 
 54 
 
 
 i>^ 
 
 I.61 
 
 1.90 
 
 •145 
 
 5.06 
 
 5-97 
 
 2.04 
 
 2.84 
 
 70.65 
 
 2.69 
 
 11^ 
 
 55 
 
 \% 
 
 2 
 
 2.07 
 
 2.38 
 
 •154 
 
 6.49 
 
 7.46 
 
 3-36 
 
 4-43 
 
 42.36 
 
 3.67 
 
 11^ 
 
 
 2/2 
 
 2.47 
 
 2.88 
 
 .204 
 
 7.75 
 
 9-03 
 
 4.78 
 
 6.49 
 
 30.11 
 
 5-77 
 
 8 
 
 89 
 
 ij^ 
 
 3 ^ 
 
 3.07 
 
 3-50 
 
 .217 
 
 9.64 
 
 11.00 
 
 7-39 
 
 9.62 
 
 19.49 
 
 7.55 
 
 8 
 
 95 
 
 i^ 
 
 3/2 
 
 3-55 
 
 4.00 
 
 .226 
 
 11.15 
 
 12.57 
 
 9-89 
 
 12.57 
 
 1456 
 
 9.06 
 
 8 I 
 
 00 
 
 1^8 
 
 4 
 
 4.03 
 
 4.50 
 
 •237 
 
 12.65 
 
 14.14 
 
 12.73 
 
 15-90 
 
 II.31 
 
 10.73 
 
 8 I 
 
 05 
 
 'V 
 
 A'A 
 
 4-51 
 
 5-00 
 
 .246 
 
 14-15 
 
 15-71 
 
 15-94 
 
 19.64 
 
 9-03 
 
 12.49 
 
 8 I 
 
 10 
 
 Jv 
 
 
 5.05 
 
 5.56 
 
 .259 
 
 15.85 
 
 17.48 
 
 19.99 
 
 2430 
 
 7.20 
 
 14.56 
 
 8 I 
 
 16 
 
 
 6 
 
 6.07 
 
 6.63 
 
 .280 
 
 19-05 
 
 20.81 
 
 28.89 
 
 34.47 
 
 4.98 
 
 18.77 
 
 8 I 
 
 26 
 
 7 
 
 7.02 
 
 7.63 
 
 .301 
 
 22.06 
 
 23-95 
 
 38.74 
 
 45.66 
 
 3-72 
 
 23.41 
 
 8 I 
 
 36 
 
 i|^ 
 
 8 
 
 7.98 
 
 8.63 
 
 .322 
 
 25.08 
 
 27.10 
 
 50.04 
 
 58.43 
 
 2.88 
 
 28.35 
 
 8 I 
 
 46 
 
 1% 
 
 9 
 
 8.93 
 
 9-63 
 
 .348 
 
 28.07 
 
 30.24 
 
 62.73 
 
 72.76 
 
 2.26 
 
 33.70 
 
 8 I 
 
 57 
 
 2 
 
 10 
 
 10.02 
 
 10.75 
 
 .366 
 
 31.48 
 
 33-77 
 
 78.84 
 
 90.76 
 
 1.80 
 
 40.64 
 
 8 I 
 
 68 
 
 2% 
 
 II 
 
 11.22 
 
 12.00 
 
 .388 
 
 35-26 
 
 37.70 
 
 98.94 
 
 113.10 
 
 1.46 
 
 47.73 
 
 8 I 
 
 80 
 
 2% 
 
 12 
 
 12.18 
 
 13.00 
 
 .410 
 
 38.26 
 
 40.84 
 
 116.54 
 
 132.73 
 
 1.24 
 
 54.66 
 
 8 I 
 
 88 
 
 2^8 
 
 Pipes I inch and below are butt-welded and proved to 300 pounds 
 per square inch hydraulic pressure. 
 
 Pipes i}l inches and larger diameters are lap-welded and proved to 
 500 pounds per square inch hydraulic pressure. 
 
 Threads for wrought-iron pipes are cut tapering at ^ inch per foot 
 of total taper for all sizes up to and including 9 inches ; for pipes 10 
 inches in diameter and larger the taper is y% inch per foot. 
 
 The table gives dimensions of 11 -inch pipe ; this is regarded in the 
 trade as an odd size, and fittings are rarely, if ever, made for it except to 
 order. The common practice is to ignore it and take 12 inch pipe instead. 
 
BOILER MOUNTINGS AND SAFETY APPARATUS 
 
 389 
 
 Flange Unions. — No uniformity of dimensions is followed by- 
 manufacturers of cast-iron flanges for wrought-iron pipe. The differences 
 may not be great, but they 
 
 are sufficient to prevent inter- Fig. 468. 
 
 change of parts. Two sizes 
 of flanges are commonly 
 necessary in putting up a 
 line of large steam-pipe, — 
 one set of flanges to match the 
 flanged valves and another 
 set for ordinary connections, 
 the valve-flanges being com- 
 monly of larger diameter 
 than the size usually adopted 
 for merely connecting one 
 pipe with another. The il- 
 lustration, Fig. 468, and the tabular dimensions accompanying relate to 
 pipe-connections only. Valve-flanges and cast-iron connections difler 
 so much in diameter that special flanges usually have to be made for 
 them. 
 
 TABLE LI. 
 
 FLANGE UNIONS. 
 
 Reference letters correspond to those shown in Fig. 468. M, 
 up to 6 inch, and J^-inch for larger pipes. 
 
 inch for all sizes 
 
 A. 
 
 B. 
 
 C. 
 
 D. 
 
 E. 
 
 ^^ 
 
 G. 
 
 H. 
 
 I. 
 
 J. 
 
 K. 
 
 L. 
 
 
 
 ^ 
 
 
 
 
 i*- 
 
 
 
 U-, 
 
 ^ 
 
 >- , 
 
 fe 
 
 u 
 
 
 
 
 
 .i 
 
 
 
 0) 
 
 
 
 
 ° 3 
 
 o_c 
 
 -■« <^ 
 
 
 1- <D 
 
 
 •*.-d 
 
 "& s 
 
 
 
 
 
 fe ° 
 
 
 bI^ 
 
 ^li? 
 
 |l 
 
 2| 
 
 It 
 
 c^ 
 mU^ 
 
 %^ 
 
 U 
 
 
 
 ^S 
 
 i! - 
 
 III 
 
 %^ ■ 
 
 la's 
 
 ■fOo 
 
 .sE 
 
 &E 
 
 %t 
 
 •1^ 
 
 a 
 
 ■B 
 
 ^ 
 
 Ho 
 am 
 
 .1°^ 
 
 z 
 
 
 
 
 
 
 
 Q 
 
 ^ 
 
 5 
 
 n 
 
 M 
 
 ^ 
 
 
 
 Q 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 Ins. 
 
 
 Ins. 
 
 Ins. 
 
 I 
 
 1-3 
 
 5 
 
 V& 
 
 Ya 
 
 A 
 
 2% 
 
 3^ 
 
 -1 
 
 3 
 
 2^ 
 
 ^y 
 
 iX 
 
 1-7 
 
 sH 
 
 % 
 
 U 
 
 t/ 
 
 2y2 
 
 4 
 
 -r 
 
 3 
 
 S/s 
 
 1% 
 
 l/^ 
 
 1-9 
 
 5H 
 
 I 
 
 'A 
 
 
 2% 
 
 4/8 
 
 -i 
 
 4 
 
 2>y2 
 
 2% 
 
 2 
 
 2.4 
 
 ey. 
 
 1% 
 
 
 \ 
 
 3/8 
 
 5 
 
 -■ 
 
 4 
 
 Ay 
 
 2y 
 
 2/2 
 
 2.9 
 
 7% 
 
 T-yi 
 
 
 % 
 
 4 
 
 5^ 
 
 _. . 
 
 4 
 
 5 
 
 3% 
 
 3 
 
 3-5 
 
 7H 
 
 iX 
 
 \% 
 
 
 A% 
 
 6X 
 
 
 6 
 
 5% 
 
 3H 
 
 3/2 
 
 4.0 
 
 s% 
 
 iX 
 
 r% 
 
 
 5% 
 
 7 
 
 
 6 
 
 6% 
 
 4/8 
 
 4 
 
 4-5 
 
 9V2 
 
 i^ 
 
 iX 
 
 
 5H 
 
 7% 
 
 
 6 
 
 ey 
 
 4^ 
 
 4/2 
 
 5-0 
 
 ^0% 
 
 1^8 
 
 ^Y 
 
 
 6/8 
 
 8H 
 
 
 6 
 
 7% 
 
 sy 
 
 5 
 
 5-6 
 
 IIX 
 
 I>^ 
 
 iX 
 
 \% 
 
 7 
 
 9 
 
 
 6 
 
 8 
 
 sy 
 
 6 
 
 6.6 
 
 12^ 
 
 ^yi 
 
 1^8 
 
 iVs 
 
 s% 
 
 1014: 
 
 it 
 
 6 
 
 9% 
 
 ey 
 
 7 
 
 7.6 
 
 13;^ 
 
 lU 
 
 i>^ 
 
 iX 
 
 9% 
 
 113^ 
 
 
 6 
 
 loy 
 
 7y 
 
 8 
 
 8.6 
 
 14^ 
 
 m 
 
 i^ 
 
 iVz 
 
 lo/s 
 
 I2>^ 
 
 -i 
 
 8 
 
 iiy 
 
 8y 
 
 9 
 
 9.6 
 
 ^5H 
 
 2 
 
 iH 
 
 I>^ 
 
 11)^ 
 
 13^ 
 
 -i 
 
 8 
 
 12^ 
 
 9Ya 
 
 10 
 
 10.8 
 
 I7X 
 
 2}i 
 
 2 
 
 1% 
 
 12^ 
 
 15 
 
 -^1 
 
 10 
 
 14 
 
 II 
 
 12 
 
 13.0 
 
 20 
 
 2Y2 
 
 2% 
 
 IH 
 
 i5>^ 
 
 i7>^ 
 
 ItV 
 
 12 
 
 16X 
 
 13X 
 
CHAPTER XI. 
 
 CHIMNEYS. 
 
 A CHIMNEY is employed in steam engineering as an attachment to 
 a steam-boiler furnace for the purpose of creating and maintaining a 
 draught through the body of burning fuel. Its practical value as com- 
 pared with other methods is in the certainty and simplicity of its action, 
 adapting itself automatically to the demands of a lesser or greater number 
 of furnaces, with little or no loss of efficiency for either under- or over- 
 load as compared with its normal power-rating. 
 
 The fuels used for steaming purposes in this country are anthracite, 
 semi-anthracite, and bituminous coals. The first of these is a very hard 
 coal, and requires a powerful draught for its rapid combustion. Bitu- 
 minous coals are found in every grade of quality from good to bad ; 
 they burn readily, and thus a less intensity of draught is needed. It is 
 for this reason that a less height of chimney is permissible. 
 
 The rate of combustion per square foot of grate surface per hour 
 will ordinarily vary from 9 to 13 pounds for anthracite coal, and from 12 
 to 20 pounds for bituminous coal, — the latter occasionally running much 
 higher, nearly or quite doubling these figures. We shall not go far 
 wrong if we assume an average rate of combustion of 12 pounds per 
 hour for anthracite coal and 15 pounds per hour for bituminous coal in 
 ordinary steam-boiler furnaces with good draught. The evaporative 
 economy falls off rapidly under the higher rates of combustion of bitu- 
 minous coal. 
 
 The temperature of air supplied boiler furnaces may vary in certain 
 localities from 10'^' below zero in winter to 95° Fahr. in summer. Assum- 
 ing 62° Fahr. to be an average temperature, the corresponding weight 
 of air will be 0.076 pounds per cubic foot, or 13.14 cubic feet to the 
 pound at atmospheric pressure (14.7 pounds) ; in relation to water its 
 density is about i to 820. If the combustion were perfect, with the 
 minimum supply of air (12 pounds) the weight of the products of com- 
 bustion would be 13 pounds, but for chimney calculations it is customary 
 to assume the maximum of air supply, or 25 pounds of gas. 
 
 Chimney Draught is an effect produced by the difference in specific 
 gravity of the cold air entering the furnace under the grate and the 
 heated products of combustion escaping from the chimney. The differ- 
 ence in weight is explained by the fact that gases expand by heat ; the 
 atmosphere of an active fire must, therefore, be of less density than the 
 outer air. 
 
 The draught of a chimney depends upon its height, and the differ- 
 390 
 
CHIMNEYS 391 
 
 ence between the specific gravity of the gases with which it is filled and 
 that of the outside air. Draught properly begins at the level where the 
 air passes through the fire, and not at the level of the ground at the base 
 of the chimney. Its action increases at first with the temperature, but 
 afterwards gradually diminishes with the temperature of the gases. For 
 example, from 32° Fahr. to 300° Fahr. the draught augments very 
 rapidly ; from 300° Fahr. to 750° Fahr. the draught varies little, but 
 arrives at the maximum when the temperature is about 585° Fahr. 
 Some investigators place the temperature higher than this, but for steam- 
 boiler purposes this temperature is as high as the products of combus- 
 tion should be allowed to attain, and from 550° to 600° Fahr. may be 
 taken as the maximum temperature in any ordinary calculations. Some 
 steam boilers do not part with their gases until the temperature is 
 reduced to 450° Fahr., and only occasionally is the reported tempera- 
 ture less. 
 
 Intensity of draught denotes the velocity of flow of air through the 
 furnace. This property is secured by height of chimney, or by high 
 temperatures of escaping gases, or both. The larger of the small sizes 
 of anthracite coal require an intensity of draught corresponding to not 
 less than seven-eighths to one inch of water, and even this is scarcely 
 enough at times to secure the desired rate of combustion ; whereas 
 with ordinary free-burning coals in lumps as large as a hickory-nut or 
 larger a draught of three-eighths to five-eighths inch of water will suffice 
 for ordinary boiler settings. 
 
 Chimney Draught from Absolute Zero. — The best practice when 
 dealing with problems relating to heated gases is to reckon temperatures 
 from absolute zero (which on the Fahrenheit scale is — 460°), because 
 the expansion of a gas due to a given rise of temperature is found to be 
 exactly proportional to the absolute temperature. 
 
 The best chimney draught is had (reckoning temperature from abso- 
 lute zero) when the temperature of gas in the chimney is to that of external 
 air as i is to 2. The absolute temperature of air at 62° Fahr, is 62 -\- 
 460 = 522° ; therefore we get the best draught when 522 X 2 = 1044° 
 absolute, or 1044° — 460° = 584° Fahr., a temperature above the 
 melting-point of bismuth (518° Fahr.) and below that of lead (630° 
 Fahr.). In general, when the escaping gases leave the boiler at a 
 temperature of melting lead it indicates that heat is going to waste. 
 
 Example: Suppose a chimney 120 feet high, the temperature of 
 escaping gases = 584° Fahr. and that of the atmosphere 62° Fahr., the 
 draught in inches of water may be found thus : 
 
 120 X 46° + 584 _ 125,280 _ ^^^ f^^^_ 
 
 460 + 62 522 
 
 The height of a column of escaping gas at 584° Fahr. equals the weight 
 
 of a column of air at 62° Fahr. outside a chimney 120 feet high ; then 
 
 240 — 120 = 120. We will call this remainder the "motive column." 
 
392 BOILERS AND FURNACES 
 
 The relation of weight as compared with water (820 times heavier) 
 may be expressed thus : 
 
 120 
 If we divide the motive column by this amount we have 
 
 120 r ^ 
 
 — — = .0732 foot, 
 1640 
 
 or ^ inch nearly, as the height of a column of water lifted by the action 
 of a chimney corresponding to the height and temperature here given. 
 
 Temperature. — If we assume that air entering the ash-pit is in- 
 creased in temperature from 62° Fahr. to 2500° Fahr. in its passage 
 through the bed of burning fuel, its volume will have been increased 
 5.663 times ; that is, the volume occupied by i pound of air at 62° Fahr. 
 is 13.14 cubic feet and would be expanded to 74.40 cubic feet at the 
 increased temperature ; or, the weight of i cubic foot would be less in 
 the proportion of 0.0761 to 0.0134 pounds. This sudden expansion, 
 occurring as it does within a distance of 6 inches or thereabouts, and 
 probably in less than a half second of time, means the acceleration of 
 its velocity, due to increase in volume alone, from i foot below the grates 
 to more than 5^ feet per second at the surface of the fire. 
 
 This sudden rise in temperature does not immediately affect the 
 draught, for it is confined to the furnace alone, and this is always partially, 
 and sometimes wholly, surrounded by absorbing surfaces, so that after 
 the passage of the heated gases along the absorbing surfaces the tem- 
 perature of the escaping products of combustion is lowered from the 
 original temperature to about j^ or less at the moment of entering the 
 chimney. It is this final temperature only which is to be taken into ac- 
 count in draught calculations. This continual loss of heat by coming in 
 contact with absorbing surfaces has the effect to retard the flow of gases 
 from the furnaces to the chimney, but another and more serious retarda- 
 tion of flow of air into the furnace occurs in the passage of air through 
 the fire itself, and this is greatly intensified in the case of fine anthracite 
 coals or of caking bituminous coals, the former often requiring a strong 
 and positive blast to force the air through the fuel. With the latter fuel 
 a frequent breaking up of the fire becomes almost a necessity to main- 
 tain a proper rate of combustion. For natural draught the temperature 
 of the escaping gases should not generally exceed 550° Fahr. ; all things 
 considered, this is regarded as a fair and economical working tempera- 
 ture. Variations above and below this temperature, say 50° Fahr., have 
 but little effect on the draught so far as the economical result of furnace 
 working is concerned. 
 
 An approximate method of determining high temperatures consists 
 in suspending strips of metal, the melting-points of which are known, 
 in the flues leading to the chimney, or perhaps in the chimney itself 
 
CHIMNEYS 393 
 
 Three different metals, such as tin, which melts at 455° Fahr., bismuth, 
 which melts at 518° Fahr., and lead, which melts at 630° Fahr., to 
 which may be added zinc, which melts at 793° Fahr., and antimony, 
 which melts at 810° Fahr., are generally employed in making such tests. 
 In the selection of metals one is chosen of which the melting-point is 
 nearest the supposed temperature, then another piece the melting-point 
 of which is below, and a third of which the melting-point is above it. 
 For example, if the temperature be supposed to be about 600° Fahr., 
 then lead will be chosen for the central strip, and one of bismuth for 
 the lower, and one of zinc for the higher temperature. The tempera- 
 ture can readily be approximated when it is known that it is between 
 two of the three metals exposed. If the metal having the lowest 
 melting-point is melted and the other two are uninjured by the heat, 
 the temperature must be assumed to be above the one and below the 
 other ; similarly, if two of the bars are melted, the temperature must 
 then be somewhere between the highest and the intermediate points of 
 fusion. This method is not entirely accurate, but sufficiently so for all 
 ordinary purposes. 
 
 The Area of a Chimney was made the subject of experimental 
 research by Mr. Isherwood, who found that the ordinary variable limits 
 were from one-sixth to one-ninth that of the grate area. Upon these 
 experiments is based the common recommendation that the area of a 
 chimney be one-eighth that of the grate surface. 
 
 The grate area for a steam boiler bears an approximately fixed rela- 
 tion to that of the heating surface, averaging not far from one-thirty- 
 fifth that of the latter. 
 
 The chimney area must bear some relation to the quantity of coal 
 burnt. In practice it is found that for sizes up to 1000 horse-power the 
 most satisfactory chimneys are those in which from i^ to 2 square 
 inches of chimney area are had for each pound of coal burnt per hour. 
 According to this rule a chimney suitable for 1000 pounds of coal per 
 hour would vary between 1000 X 1.5= 1500 square inches, or 43^^ 
 inches diameter, and 1000 X 2 := 2000 square inches, or 50^ inches 
 diameter. 
 
 Example: Required a grate and chimney area suitable for 1000 
 pounds of coal per hour, the rate of combustion to be 12 pounds per 
 square foot of grate per hour : 
 
 ^'^ = 10.4 square feet = 1498 square inches = 43^ inches, diameter of chimney. 
 
 which corresponds to i ^ square inches of chimney area for each pound 
 of coal burnt per hour. It also corresponds to one-eighth that of the 
 grate area. 
 
 26 
 
394 BOILERS AND FURNACES 
 
 Height of Chimneys. — In order to give the required draught to a 
 steam-boiler furnace of say 50 horse-power, the chimney should not be 
 less than 60 or 65 feet in height above the ground surface at its base, 
 and need not exceed 150 feet for boiler furnaces of 1000 horse-power, 
 unless there is higher land or buildings in the immediate neighborhood 
 which would affect its draught. Anthracite coals require a higher 
 chimney than free-burning bituminous coals. The area having been 
 fixed, either by quantity of coal burned or in proportion to grate area, 
 a common rule is to make the height of a small chimney twenty-five 
 times its diameter, with a gradual decrease in the ratio for larger 
 chimneys ; thus, a 4-foot chimney may be 100 feet in height, a 5-foot 
 chimney 120 feet in height, a 6-foot chimney 135 feet in height, an 
 8-foot chimney 160 feet, and a lo-foot chimney 175 feet. 
 
 The formulas usually given for the height of chimneys do not always 
 furnish the proportions which certain localities seem to require, and 
 as a result there is more of empiricism than calculation in this detail. 
 Many large chimneys are higher than necessary when considered merely 
 as a means of obtaining or maintaining a good draught. It is doubtful 
 whether there are any advantages sufficient to pay the extra cost for 
 making a steam-boiler chimney more than 150 feet high for internal 
 diameters less than 100 inches. 
 
 Horse-Power of a Chimney.— By this is meant a chimney of 
 suitable diameter and height for the grate surface required for the com- 
 bustion of fuel necessary to develop the stated amount of horse-power 
 in steam boilers of well-known types. The standard horse-power in 
 steam engineering is the evaporation of 30 pounds of water per hour 
 from a feed-water temperature of 100° Fahr. into steam at 70 pounds 
 gauge pressure. This is equivalent to 34}^ pounds of water evaporated 
 per hour from and at 212° Fahr. 
 
 For the chimney calculations in Table LII., the tabular dimensions 
 of horizontal tubular boilers given in Tables XL., XLIL, and XLIV. 
 are assumed to meet all the requirements of ordinary single-cylinder 
 engines, non-condensing and non-compound. For single-cylinder con- 
 densing and for compound engines the boiler-power will be somewhat 
 in excess of the requirements of the engines. 
 
 The diameter of chimney given for each power is ample for any fuel, 
 whether anthracite or bituminous coal. No length of side is given for 
 a square chimney of equal area because the same diameter should be 
 used for both. The corners of a square chimney, especially for those 
 below 60 inches diameter, count for very little, the general configura- 
 tion of flow of gases upward is in a round column. It is doubtful if 
 the corners serve any useful purpose, and may be wholly neglected in 
 computing chimney area. 
 
 The height of chimneys for the several powers given will be found 
 suitable for the stipulated grades of bituminous coal and for anthracite 
 
CHIMNEYS 
 
 395 
 
 coal not finer than buckwheat size. If finer coals are used the fire may- 
 require assistance, which can be fiirnished by either a steam jet or fan 
 blower. 
 
 TABLE LII. 
 
 TABLE OF CHIMNEY DIMENSIONS FOR A SINGLE BOILER AND FURNACE FROM 
 20 TO 100 HORSE-POWER, BASED UPON THE TABULAR DIMENSIONS GIVEN 
 IN TABLES XL., XLII., AND XLIV., IN WHICH THE GRATE AREA IS ASSUMED 
 TO BE NINE TIMES THAT OF THE TUBE AREA FOR THE SMALLEST BOILER, 
 DIMINISHING TO SEVEN TIMES THE TUBE AREA FOR THE LARGEST BOILER. 
 A COMMERCIAL HORSE-POWER RATING APPROXIMATING I5 SQUARE FEET 
 OF HEATING SURFACE PER HORSE-POWER IS ASSUMED FOR ALL BOILERS 
 IN THIS SERIES. 
 
 
 
 Boiler and Furnace Details. 
 
 Chimney. 
 
 
 Boilers. 
 
 
 Grate. 
 
 
 V V 
 
 Height 
 
 n Feet. 
 
 
 3 
 
 
 
 
 
 
 i 
 
 t; c! 
 
 
 
 ^• 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 S 
 
 
 
 
 
 
 
 ^ 
 
 & 
 
 2 
 
 1 
 
 S 
 
 % 
 
 K^ 
 
 
 o.Su 
 
 •si's 
 
 < 
 
 G 
 
 11 
 
 Ml 
 
 1 3 
 
 •£ N 
 
 2. 
 
 
 .2 
 
 V 
 
 15 
 
 u 
 
 ^x^ 
 
 2 
 
 Soi 
 
 Box. 
 
 cUm 
 
 K 
 
 oi 
 
 P 
 
 J 
 
 ^ 
 
 < 
 
 ^ 
 
 u 
 
 ^ 
 
 m 
 
 <; 
 
 
 
 Inches. 
 
 Feet. 
 
 Sq. Feet. 
 
 Sq. Feet. 
 
 
 Sq. Feet. 
 
 Inches. 
 
 
 
 20 
 
 2 
 
 38 
 
 10 
 
 311 
 
 12.15 
 
 25.60 
 
 2.02 
 
 20 
 
 50 
 
 60 
 
 25 
 
 3 
 
 40 
 
 12 
 
 396 
 
 12.87 
 
 30-75 
 
 2.15 
 
 20 
 
 50 
 
 60 
 
 30 
 
 4 
 
 42 
 
 14 
 
 488 
 
 13.68 
 
 35-69 
 
 2.28 
 
 20 
 
 55 
 
 65 
 
 35 
 
 5 
 
 44 
 
 14 
 
 515 
 
 14.40 
 
 35-73 
 
 2.40 
 
 22 
 
 55 
 
 65 
 
 40 
 
 7 
 
 48 
 
 14 
 
 611 
 
 17.46 
 
 35-OI 
 
 2.91 
 
 24 
 
 55 
 
 70 
 
 45 
 
 8 
 
 50 
 
 14 
 
 682 
 
 19.71 
 
 34-58 
 
 3-29 
 
 25 
 
 55 
 
 70 
 
 50 
 
 10 
 
 54 
 
 14 
 
 778 
 
 22.77 
 
 34-19 
 
 3.67 
 
 26 
 
 60 
 
 70 
 
 55 
 
 II 
 
 56 
 
 14 
 
 827 
 
 24.30 
 
 34-03 
 
 3-8o 
 
 27 
 
 60 
 
 75 
 
 60 
 
 19 
 
 58 
 
 16 
 
 908 
 
 24.16 
 
 37-58 
 
 3-80 
 
 27 
 
 60 
 
 75 
 
 65 
 
 20 
 
 60 
 
 16 
 
 971 
 
 26.00 
 
 37-36 
 
 3-94 
 
 27 
 
 60 
 
 80 
 
 70 
 
 27 
 
 60 
 
 18 
 
 1055 
 
 28.72 
 
 36.85 
 
 4-35 
 
 28 
 
 65 
 
 80 
 
 75 
 
 29 
 
 64 
 
 18 
 
 II42 
 
 31.12 
 
 36.70 
 
 4-58 
 
 29 
 
 65 
 
 80 
 
 80 
 
 30 
 
 66 
 
 18 
 
 1242 
 
 34.16 
 
 36-36 
 
 4.88 
 
 30 
 
 65 
 
 85 
 
 85 
 
 30 
 
 66 
 
 18 
 
 1242 
 
 34.16 
 
 36.36 
 
 4.88 
 
 30 
 
 65 
 
 85 
 
 90 
 
 31 
 
 68 
 
 18 
 
 1361 
 
 37-84 
 
 35-76 
 
 5.00 
 
 30 
 
 70 
 
 90 
 
 95 
 
 33 
 
 72 
 
 16 
 
 1421 
 
 39-55 
 
 35-98 
 
 5-00 
 
 30 
 
 70 
 
 90 
 
 100 
 
 33 
 
 72 
 
 18 
 
 1598 
 
 39-55 
 
 40.48 
 
 5-00 
 
 30 
 
 70 
 
 90 
 
 The diameter for chimneys from lOO to looo horse-power, as given 
 in Table LIII., is estimated from an entirely different stand-point from 
 that of Table LII. It is here assumed that 4 pounds of coal per horse- 
 power per hour may be required, and for large boiler plants this quan- 
 tity is correct for anthracite coal, and averages tolerably close for bitu- 
 minous coals, inasmuch as the latter varies widely in evaporative power. 
 The rate of combustion, assumed to be 12 pounds per square foot of 
 grate surface per hour, is, for steam plants for the last half of the table, 
 somewhat less than the average for bituminous coal. There are steam- 
 boiler furnaces burning double that weight, and occasionally more than 
 three times the assumed rate ; but as capacity tests and economy tests 
 are often widely at variance with each other, no opinion can be here 
 
396 
 
 BOILERS AND FURNACES 
 
 expressed upon what might be considered excessive rates of combustion 
 unless the ratio of grate to heating surface is known, and especially the 
 temperature of the escaping gases. If the temperature of the latter be 
 higher than 600° Fahr. , heat is going to waste, and the rate of combus- 
 tion ought to be lowered until the temperature is reduced to a more 
 economical figure. This table, in common with the preceding one, 
 makes no difference in the diameter of a round chimney and that of 
 a square chimney to make the areas equal, for the reason given on 
 page 394. 
 
 TABLE LIU. 
 
 TABLE OF CHIMNEY DIMENSIONS FOR TWO OR MORE BOILERS SET IN BATTERY 
 AND WORKING TOGETHER ; HORSE-POWER IS BASED ON 4 POUNDS OF COAL 
 PER HOUR PER HORSE-POWER. THE RATE OF COMBUSTION IS ASSUMED 
 TO BE 12 POUNDS PER SQUARE FOOT OF GRATE SURFACE PER HOUR ; THE 
 PROPORTION OF GRATE TO CHIMNEY AREA VARIES FROM ONE-SEVENTH 
 FOR THE lOO HORSE-POWER BOILER TO ONE-TENTH FOR THE lOOO HORSE- 
 POWER BOILER. 
 
 
 
 
 
 
 Hkight 
 
 IN Feet. 
 
 
 Coal per 
 
 Area of 
 
 Grate at 12 
 
 Pounds per 
 
 Square Foot 
 
 per Hour. 
 
 Area of 
 
 Nearest 
 Diameter, 
 
 
 
 Horse- 
 
 
 
 Power. 
 
 Hour. 
 
 Chimney. 
 
 Round 
 
 Bituminous 
 
 Anthracite 
 
 
 
 
 Chimney. 
 
 Coal, free 
 
 Coal, small 
 
 
 
 
 
 burning. 
 
 sizes. 
 
 
 
 Square Feet. 
 
 Square Feet. 
 
 Inches. 
 
 
 
 lOO 
 
 400 
 
 33-33 
 
 4-76 
 
 30 
 
 70 
 
 90 
 
 125 
 
 500 
 
 41.67 
 
 5-i8 
 
 31 
 
 75 
 
 90 
 
 150 
 
 600 
 
 50.00 
 
 6.82 
 
 36 
 
 75 
 
 95 
 
 175 
 
 700 
 
 58.33 
 
 7.78 
 
 38 
 
 80 
 
 100 
 
 200 
 
 800 
 
 66.67 
 
 8.69 
 
 40 
 
 80 
 
 100 
 
 250 
 
 1000 
 
 83-33 
 
 10.64 
 
 44 
 
 85 
 
 105 
 
 300 
 
 1200 
 
 100.00 
 
 12.50 
 
 48 
 
 85 
 
 105 
 
 350 
 
 1400 
 
 116.67 
 
 14.18 
 
 51 
 
 90 
 
 no 
 
 400 
 
 1600 
 
 133-33 
 
 16.00 
 
 55 
 
 90 
 
 115 
 
 450 
 
 1800 
 
 150.00 
 
 17-65 
 
 57 
 
 90 
 
 115 
 
 500 
 
 2000 
 
 166.67 
 
 19-25 
 
 60 
 
 95 
 
 120 
 
 550 
 
 2200 
 
 183-33 
 
 20.65 
 
 62 
 
 95 
 
 120 
 
 600 
 
 2400 
 
 200.00 
 
 22.22 
 
 64 
 
 100 
 
 125 
 
 650 
 
 2600 
 
 216.67 
 
 23-65 
 
 66 
 
 100 
 
 125 
 
 700 
 
 2800 
 
 233-33 
 
 25-01 
 
 68 
 
 105 
 
 130 
 
 750 
 
 3000 
 
 250.00 
 
 26.32 
 
 70 
 
 105 
 
 135 
 
 800 
 
 3200 
 
 266.67 
 
 27.61 
 
 72 
 
 no 
 
 135 
 
 850 
 
 3400 
 
 283.33 
 
 28.82 
 
 73 
 
 no 
 
 140 
 
 900 
 
 3600 
 
 300.00 
 
 30.00 
 
 74 
 
 115 
 
 145 
 
 950 
 
 3800 
 
 316.67 
 
 31.67 
 
 76 
 
 115 
 
 145 
 
 1000 
 
 4000 
 
 333.33 
 
 33-33 
 
 78 
 
 120 
 
 150 
 
 Percentage of Chimney Area. — The larger percentage of chim- 
 ney area for the lOo horse-power boiler over the succeeding ones is on 
 account of the friction of the gases against the sides of a small chimney. 
 The usual recommendation is to add a constant of 4 inches to any diam- 
 eter found to be necessary for a chimney ; but this increase in diameter 
 is not necessary for the larger chimneys, especially when based on one- 
 
CHIMNEYS 397 
 
 eighth the grate area, — in fact, the necessity for any increase in diameter 
 for friction is not apparent for chimneys over 48 inches in diameter, 
 when based on the above proportion, and at 60 inches in diameter a 
 gradual reduction may be made in the ratio of chimney area to grate 
 area. 
 
 Chimney Design. — A chimney-shaft may be round, octagonal, 
 square, or any other form to suit the taste of the designer or its adapta- 
 tion to the conditions which call for its erection. A round chimney is 
 to be preferred to any other interior cross-section, because that is the 
 natural form of a column of ascending gases from a fire. Square chim- 
 neys are a practical or commercial necessity when of small diameters, 
 say 48 inches and less, because of the rectangular shape of common 
 bricks and the fact that bricks for circular walls, except fire-bricks, must 
 be made to order. Some square chimneys have a round inner hning, 
 and this detail has much to commend it ; the lower part is commonly of 
 fire-brick 6 inches thick, the upper of red brick 4^ inches thick. This 
 lining is not built into, but just clears the inner side of the square, 
 leaving the corners open. It may extend to the top of the chimney 
 with advantage. Such a chimney is easy of construction and of com- 
 paratively low cost. 
 
 The inside diameter of a chimney need not be otherwise than parallel 
 from bottom to top, although chimneys are made both larger and smaller 
 at the top than at the bottom, and without any perceptible loss of effect. 
 When the lining of a chimney extends only a portion of the height, the 
 area is considerably increased where the lower diameter debouches into 
 the upper portion of the shaft ; this is not known to have any injurious 
 effect upon the draught. 
 
 The Stability of a Chimney is of the utmost importance. It 
 seems almost superfluous to say that a chimney should be made of the 
 best materials and constructed in the best manner. The wind pressure 
 is the greatest resistance to be overcome by a chimney, and this is 
 reckoned at 55 pounds per square foot as a maximum. Except in the 
 case of whirlwinds or cyclones, no such wind pressure is experienced in 
 this country. Its ability to withstand wind pressure varies with the 
 shape of the chimney, a round chimney offering least resistance to the 
 wind, an octagon a greater resistance, and a square chimney most of all. 
 Without entering upon the mathematics of the stability of chimneys, it 
 may be said that complete stability may be had in brick chimneys if the 
 outside diameter at the base is one-tenth of its height for a square chim- 
 ney, one eleventh of the height for an octagon chimney, and one-twelfth 
 of its height for a round chimney, with a uniform taper of ^ inch per 
 foot throughout the whole height, excepting, of course, the base or 
 pedestal upon which the shaft rests. The thickness of the brickwork 
 at the top of a chimney should not in any case be less than one brick, 
 say 8 to 9 inches, depending upon the commercial sizes of bricks in the 
 
398 BOILERS AND FURNACES 
 
 locality where the chimney is to be erected ; and this thickness is ample 
 for all heights up to i6o feet, above which the walls should be a brick 
 and a half thick at the top. The thickness here given may continue 
 downward from the top for say 25 feet, when a half-brick should be 
 added to the thickness for another 25 feet, and so on down to the 
 foundation or to the pedestal upon which the shaft rests. 
 
 Metal Chimneys. — The superiority of metal over brick for high 
 chimneys is now being urged by many engineers. The arguments, in 
 general, are these : the enormous quantity of brick and material neces- 
 sary for high brick chimneys increases the cost and weight of such 
 construction, requiring massive foundations and much space ; should 
 these foundations sink but a trifle, there will result cracks in the chimney, 
 thus impairing the draught and endangering the stability ; the sudden 
 changes in temperature to which all chimneys are subjected are also 
 destructive to brick chimneys ; being unprotected by metal shells, the 
 sudden contraction of the side exposed to the cold blast of wind, rain, 
 or sleet causes them to crack ; in a brick chimney the only resistance to 
 wind pressure is that due to its weight, and as most brick chimneys are 
 square or octagonal in cross-section the resistance is greater, since the 
 surface exposed to the wind is flat. 
 
 Metal chimneys are built with or without fire-brick lining, depending 
 upon the temperature of the escaping gases, the thickness of the lining 
 varying for the different diameters and height. The weight of metal 
 chimneys when lined is in most cases sufficient to withstand overturning 
 by ordinary wind pressure, but the precaution of bolting securely to a 
 good foundation should not be omitted ; this, it will be understood, is 
 in addition to the customary fastening to the foundation plate, upon 
 which the metal shell rests. 
 
 A Ladder should be provided for a chimney, which may be either 
 on the outside or the inside, so as to gain access to the top for any 
 examination or repairs that may be required in after service. 
 
 A Manhole with an iron door and frame should always be con- 
 structed in the base of a chimney for admission for examination, clean- 
 ing, and repairs. The top of this opening should be arched, so that 
 the strength of the chimney shall not be impaired. 
 
INDEX. 
 
 Adams's boiler, 329. 
 
 Adamson's flanged flue, 148. 
 
 Air, admission over fire, 235. 
 
 Air, composition of, 11. 
 
 Air, excess of in furnace, 17. 
 
 American stoker, 226. 
 
 Annealing mild steel, 47, 108. 
 
 Anthracite coal, 14. 
 
 Aqueous vapor, 17. 
 
 Area of chimney, 393. 
 
 Area of circular segments, 175. 
 
 Ashcroft low-water detector, 360. 
 
 Ashes, 17. 
 
 Ashley low-water alarm, 362. 
 
 Ash-pit, 235. 
 
 Babcock & Wilcox boiler, 303, 341. 
 Back connections, 241. 
 Back stand for boilers, 162. 
 Baffle-bricks for water-tube boilers, 
 
 313- 
 Bagasse, heating power of, 12. 
 Bailey's fusible plug, 366. 
 Baker, C. W., quoted, 200. 
 Batteries of boilers, 206. 
 Belpaire boiler, 281, 340. 
 Bending test, wrought iron, 34. 
 Bingham rotating gauge-cock, 367. " 
 Bituminous coal, 13. 
 Bituminous coal, calorific value, 19, 20. 
 Bituminous coal, proximate analysis, 
 
 Blisters, 37. 
 Block coal, 14. 
 Blow-holes in castings, 27. 
 Blow-off, arrangement of, 351. 
 Boiler furnaces and settings, 206. 
 Boiler-head, pressure for bumped, 119. 
 Boiler-head, pressure for concave, 120. 
 Boiler-head stays, 120. 
 Boiler-head, strains on, 119. 
 Boiler-head, thickness of, 119. 
 
 Boiler-head, working pressure, 119. 
 
 Boiler mountings, 343. 
 
 Boiler performance, examples of, 337. 
 
 Boiler-plate, qualities of, 35. 
 
 Bottom blow, 349. 
 
 Bourdon pressure-gauge, 373. 
 
 Bowling hoop for flues, 148. 
 
 Braces for steam boilers, 122. 
 
 Brackets, detachable, 159. 
 
 Brackets, or wings, for boilers, 158. 
 
 Brand of plate iron, 36. 
 
 Breeching for boilers, 245. 
 
 Brickwork, 238. 
 
 Bridge- wall, 236, 251. 
 
 British thermal unit, 20. 
 
 Brown coal, 13. 
 
 Buck-staves, 240. 
 
 Bulging test, mild steel, 46. 
 
 Bumped heads, 119. 
 
 Burg, Professor, on safety-valves, 355. 
 
 Butt-joints, 72, 91. 
 
 Butt-joints, proportions for, 93. 
 
 Cahall boiler, 333, 342. 
 Cahall swinging manhead, 157. 
 Caking coal, 13. 
 Caldwell boiler, 312, 341. 
 Calorific value of coal, 22. 
 Cannel coal, 14. 
 Carbon, 10. 
 
 Carbon, heat units in, 20. 
 Carbon in cast iron, 25. 
 Carbon in mild steel, 38. 
 Carbonic acid gas, 16. 
 Carbonic oxide gas, 16. 
 Cast iron, 25. 
 
 Cast-iron boiler-head, 180. 
 Cast iron for steam boilers, 30. 
 Cast iron in the fire, 30. 
 Cast iron, objections to, 30. 
 Chain riveting, 85. 
 Chapman gate valve, 348. 
 
 399 
 
400 
 
 INDEX 
 
 Charcoal hammered iron, 36. 
 
 Charcoal iron, 35. 
 
 Check- and stop-valve combined, 347. 
 
 Check- valves, 345. 
 
 Chemical changes, 9. 
 
 Chemical properties, wrought iron, 35. 
 
 Chimney, 390. 
 
 Chimney design, 397. 
 
 Chimney dimensions, tables, 395, 396. 
 
 Circulation in water-tube boilers, 300. 
 
 Clinker, 18. 
 
 Coal, various analyses, 23, 24. 
 
 Codman, J. E., 272. 
 
 Coke, 14. 
 
 Cold-bending test, mild steel, 45. 
 
 Cold-short wrought iron, 35. 
 
 Cole, F. J., quoted, 138. 
 
 Collapse of flue and tubes, 150. 
 
 Combined carbon in iron, 25. 
 
 Combined water-gauge, 371. 
 
 Combustion-chamber, 240. 
 
 Combustion, rate of, 208. 
 
 Compression gauge-cock, 368. 
 
 Concave heads, 120. 
 
 Continental Iron-Works, 100, 272. 
 
 Cooling strains in castings, 27. 
 
 Cornish boiler, 265. 
 
 Corrugated flues, 149. 
 
 Corrugated flues, pressure allowed, 153. 
 
 Countersunk rivets, 55. 
 
 Covering for boilers, 242. 
 
 Cox, E. T., quoted, 19. 
 
 Coxe mechanical stoker, 224. 
 
 Crow-foot, 125. 
 
 Crown-bars, 131, 276. 
 
 Crushing strength, cast iron, 29. 
 
 Culm, 15. 
 
 Culver's stop- and check-valve, 347. 
 
 Cylinder boiler, 178, 338. 
 
 Cylinder boiler, double-deck, 181. 
 
 Damper, 383. 
 
 Damper regulator, automatic, 384. 
 
 Dean, F. W., 288. 
 
 Details and strength of construction, 
 
 no. 
 Diagonal brace, Lukens, 130. 
 Diagonal stays, 128. 
 Domes, 166. 
 
 Double-deck boilers, 201, 339. 
 Double-riveted butt-joint tests. Table 
 
 XXI. 
 
 Double-riveted lap-joints, 85, Table 
 
 XIV. 
 Double walls, 238. 
 Draught from absolute zero, 391. 
 Draught in chimneys, 390. 
 Drifting test, mild steel, 45. 
 Drilled holes, 52. 
 Drum, mud-, 172. 
 Drum, steam-, 169. 
 Dry pipe, 376. 
 Ductility, wrought iron, 33. 
 Dudgeon's tube-expander, 144. 
 Dynamical value of combustion, 20. 
 
 Eastwood blow-off valve, 353. 
 Eastwood stop-valve, 378. 
 Eastwood union, 387. 
 Eclipse manhole, 156. 
 Economic portable boiler, 293. 
 Edgemoor Iron Company, 266. 
 Edgerton's boiler setting, 251. 
 Edgerton's separator, 377. 
 Efficiency of boiler, 21. 
 Elastic limit, cast iron, 29. 
 Elastic limit, mild steel, 43. 
 Elastic limit, wrought iron, 33. 
 Elastic ratio, 43. 
 Elbow union, 388. 
 Electric portable boiler, 292. 
 Elephant boiler, 181. 
 Elongation, mild steel, 43. 
 Expander, distortion by, 145. 
 Expander, Dudgeon's, 143. 
 Expander, Prosser's, 144. 
 Expansion-joints, 380. 
 Externally fired boilers, 178. 
 
 Factor of safety, 117. 
 
 Fairbairn's experiments on flues, 148. 
 
 Fanning, J. T., boiler design, 203. 
 
 Feeding water in steam-space, 344. 
 
 Feed-pipe, 343. 
 
 Feed-water, purifying, 345. 
 
 Ferro-manganese in cast iron, 27. 
 
 Ferrules in lap-welded tubes, 145. 
 
 Fibre in wrought iron, 33. 
 
 Final area, 44. 
 
 Fire-box boiler performance, 339. 
 
 Fire-brick lining, 236. 
 
 Fire-door, Butman's, 247. 
 
 Fire-door openings, 163. 
 
 Fire-doors, 247. 
 
INDEX 
 
 401 
 
 Fire-front, half-arch, 244. 
 
 Fire-front, full square, 246. 
 
 Fire-front, Naylor's, 248. 
 
 Five-flue boilers, 185. 
 
 Flame, 11. 
 
 Flanged edges, thickness of, 108. 
 
 Flange iron, 36. 
 
 Flange, radius of, 107. 
 
 Flange union, 389. 
 
 Flanging, 105. 
 
 Flanging and welding, 95. 
 
 Flexible stay-bolts, 137. 
 
 Flues, corrugated, 149. 
 
 Flues, flanging of heads for, 183. 
 
 Flues for furnaces, 148. 
 
 Flues, furnace, pressure allowed on, 
 
 151. 
 Flues, lengths for, 183. 
 Flues, pressures allowed on, 183. 
 Flues, strength to resist collapse, 150. 
 Flues, thickness for, 183. 
 Flux in welding, 96. 
 Fox's corrugated flue, 149. 
 Free air in furnace, 17. 
 Free-burning coal, 14. 
 French boiler, 181. 
 Friction in riveted joints, 65. 
 Fuel, defined, 12. 
 Furnace combustion, 10. 
 Furnace construction, examples of, 249. 
 Furnace-flues, 148. 
 Furnace-flues, pressure allowed, 151. 
 Furnace-wall, thickness of, 238. 
 Fusible plug-alarms, objections to, 360. 
 Fusible plugs, 365. 
 
 Galloway boiler, 266, 341. 
 
 Gas from coal, heat units in, 19. 
 
 Gate valves, 348. 
 
 Gauge-cocks, 366. 
 
 Gill's boiler, 310. 
 
 Glass tubes, cutting to length, 370. 
 
 Glass tubes for water-gauges, 369. 
 
 Globe valves, 378. 
 
 Grain in mild steel, 43. 
 
 Graphitic carbon in iron, 25. 
 
 Grate, ^Etna shaking-, 213. 
 
 Grate area and heating surface, 195. 
 
 Grate area to tube area, 195. 
 
 Grate-bars, 208. 
 
 Grate, Butman's shaking-, 212. 
 
 Grate, circular, 210. 
 
 Grate, distance from boiler, 235. 
 
 Grate, height above floor, 209. 
 
 Grate, herring-bone, 211. 
 
 Grate, plain, 209. 
 
 Grate, rate of travelling, 225. 
 
 Grate, revolving, 211. 
 
 Grate, Rose's shaking-, 213. 
 
 Grate, shaking-, 280. 
 
 Grate, size of, 206. 
 
 Grates, deterioration in, 214. 
 
 Grates, slope to rear, 209. 
 
 Gun-boat boiler, 272, 340. 
 
 Gusset-stay, 126. 
 
 Hammer test, wrought iron, 34. 
 
 Hand-flanging, 106. 
 
 Handholes, 158. 
 
 Hardwick low-water alarm, 361. 
 
 Harrison, Joseph, Jr., 296. 
 
 Hartford boiler setting, 251. 
 
 Hartford Steam Boiler Inspection and 
 
 Insurance Company, quoted, 146. 
 Hawley down-draft furnace, 230. 
 Hazelton boiler, 326, 342. 
 Head, cast-iron, 180. 
 Heat developed by combustion, 18. 
 Heat, mechanical equivalent, 20. 
 Heating of plates, 105. 
 Heating surface, 176. 
 Heating surface and grate area, 195. 
 Height of chimney, 394. 
 Heine boiler, 317, 341. 
 Hogan boiler, 324, 341. 
 Holding power of tubes, 146. 
 Horse-power, 9. 
 Horse-power of boilers, 176. 
 Horse-power of chimney, 394. 
 Horse-power standard, 177. 
 Hotchkiss's surface blow, 352. 
 Hot test, riveted joint, 75. 
 Hydrogen, 10. 
 Hydrogen and carbon, 19. 
 Hydrogen, heat units in, 20. 
 
 Ignition, 11, 
 
 Incandescence, color and intensity, 10. 
 
 Internally fired boilers, 256. 
 
 Iron boiler-plates, defects in, 37, 
 
 Iron for stay-bolts, 141. 
 
 Iron in single and double shear, 64. 
 
 Iron plates and iron rivets, 57. 
 
 Iron plates, flanging of, 107. 
 
402 
 
 INDEX 
 
 Iron plates, strength of, 50. 
 Iron rivets, properties of, 59. 
 Iron tubes, lap-welded, 141. 
 
 Jenkins's blow-off valve, 353. 
 Jenkins's gate valve, 348. 
 Jones mechanical stoker, 228. 
 
 Ladder for chimneys, 398. 
 
 Lancashire boiler, 265. 
 
 Lap-joint with reinforced welt, 88. 
 
 Leavitt, E. D., Jr., boiler design, 282. 
 
 LeVan's boiler setting, 252. 
 
 Le Van's damper regulator, 384. 
 
 Lignite, 13. 
 
 Link for stays, 126. 
 
 Lloyd's rule for boiler-stays, 123. 
 
 Locomotive boiler, 277. 
 
 Low-water alarms, 360. 
 
 Machine-flanging, 106. 
 
 Manganese in cast iron, 27. 
 
 Manganese in mild steel, 39. 
 
 Manhole for chimneys, 398. 
 
 Manhole plates or covers, 156. 
 
 Manhole rings and plates, 154. 
 
 Manholes, 153. 
 
 Manning's vertical boiler, 261, 340, 
 
 Mason's damper regulator, 385. 
 
 Materials of construction, 25. 
 
 Mechanical stokers, 215. 
 
 Metal chimneys, 398. 
 
 Mild steel, 37. 
 
 Mild steel, physical properties of, 39. 
 
 Mild-steel plates, punching, 53. 
 
 Mild steel, strength of, 51. 
 
 Mild steel, welding of, 95. 
 
 Mississippi gauge-cock, 366. 
 
 Morin boiler, 330, 342. 
 
 Morison's corrugated flue, 149. 
 
 Mortar joints, 239. 
 
 Mud-drum in Heine boiler, 321. 
 
 Mud-drum, materials for, 174. 
 
 Mud-drums, 172. 
 
 Mud-drums, functions of, 174. 
 
 Murphy furnace, 216. 
 
 Myers's blow-off valve, 353. 
 
 Natural gas, composition of, 15. 
 Natural gas, evaporative power, 16. 
 Natural gas vs. coal, 16. 
 Nitrogen, 11. 
 
 Nitrogen, specific gravity, 17. 
 Non-caking coals, 14. 
 Nozzles, dimensions for, 172. 
 
 Open-hearth steel, 38. 
 Oxide of iron in welding, 96. 
 Oxygen, 11. 
 
 Parry's safety-plug, 366. 
 
 Pearson's expansion-joint, 381. 
 
 Peat, 13. 
 
 Percentage of chimney area, 396. 
 
 Petroleum at World's Fair, 15. 
 
 Petroleum, composition of, 15. 
 
 Petroleum, evaporative power of, 15. 
 
 Phosphorus in cast iron, 26. 
 
 Phosphorus in mild steel, 39. 
 
 Physical changes, 9. 
 
 Physical properties of wrought iron, 31. 
 
 Pipe for bottom blow, 350. 
 
 Piping of water-columns, 372. 
 
 Pittsburgh high- and low-water alarm, 
 
 364. 
 Playford mechanical stoker, 222. 
 Plug-cocks, 352. 
 
 Portable boiler, dimensions of, 281, 292. 
 Portable-engine boilers, 289. 
 Pressure allowed on flues, 151. 
 Pressure-gauge, 373. 
 Pressure on convex heads, 117. 
 Pressure, Philadelphia rules for, 117. 
 Pressure, safe working, rule, iii. 
 Pressure, table of working, 112. 
 Pressure, U. S. rule, 117. 
 Products of combustion, 16. 
 Products of combustion over boiler, 243. 
 Prosser's tube-expander, 144. 
 Punch and die, 52. 
 Punched holes, 52. 
 Punching steel plates , effect of, 53. 
 
 Quadruple-riveted butt-joints, tests, 
 
 Table XXI. 
 Quenching test, mild steel, 46. 
 
 Radial stays, 133. 
 Radiators for tubes, 200. 
 Rate of combustion, 21. 
 Rear arch, skeleton for, 241. 
 Red-short wrought iron, 35. 
 Reduction of area, loss by, 51. 
 Reduction of area, mild steel, 43. 
 
INDEX 
 
 403 
 
 Register gauge-cock, 367. 
 Reliance gauge-cock, 368. 
 Reliance high- and low-water alarm, 
 
 364. 
 Retarders for fire-tubes, 197. 
 Reynolds's furnace, 233. 
 Reynolds's vertical boiler, 261. 
 Riveted joint, failure in, 68. 
 Riveted joint, friction in, 65. 
 Riveted joint, hot test, 75. 
 Riveted joint, properties of, 74. 
 Riveted joint, proportioning, 82. 
 Riveted joint, results of tests, 66. 
 Riveted joints, 50. 
 Riveted joints, efficiencies of, 73. 
 Riveted shells, strength of, no. 
 Riveting, example of, 276. 
 Riveting, method of, 91. 
 Rivet-holes, 52, 56. 
 Rivet iron, strength of, 58. 
 Rivet points, 62. 
 Rivet steel, strength of, 58. 
 Rivet steel, tests of, 60. 
 Rivets, countersunk, 55. 
 Rivets, dimensions of, 61, 62. 
 Rivets in double shear, 65. 
 Rivets in single shear, 63. 
 Rivets, iron in iron plates, 57. 
 Rivets, length of, 63. 
 Rivets, pitch of, 56. 
 Rivets, shearing strength of, 63. 
 Rivets, size of, 55. 
 Rivets, steel, and steel plates, 57. 
 Rivets, steel, properties of, 60. 
 Rivets, tests and inspection of, 60. 
 Rivets, tests of mild steel, 64. 
 Root boiler, 314. 
 
 Safety- and stop-valve combined, 354. 
 
 Safety apparatus, 343. 
 
 Safety-valve, 354. 
 
 Safety-valve, American, 358. 
 
 Safety-valve, calculating load on, 356. 
 
 Safety-valve, consolidated, 359. 
 
 Safety-valve, lift of, 355. 
 
 Safety-valve, Philadelphia regulations, 
 356. 
 
 Safety-valve, piping of, 360. 
 
 Safety-valve, spring-loaded, 358. 
 
 Safety-valves in duplicate, 354. 
 
 Safety-valves, United States Regula- 
 tions, 354. 
 
 Safety water-columns, 372. 
 
 Schaeffer diaphragm-gauge, 373. 
 
 Scotch boiler, 272. 
 
 Sectional boilers, 296. 
 
 Sectional flue-expanders, 145. 
 
 Segments, area of circular, 175. 
 
 Semi-anthracite coal, 14. 
 
 Semi-bituminous coal, 14. 
 
 Separator, Edgerton's, 377. 
 
 Serve ribbed tube, 201. 
 
 Shaw's mercury gauge, 375. 
 
 Shell iron, 36. 
 
 Shells, butt-joints, pressures for, 114. 
 
 Shells, double-riveted, pressures for, 
 
 112. 
 Shells, triple-riveted, pressures for, 113. 
 Shrinkage of cast iron, 28. 
 Silicon and cast iron, 26. 
 Silicon in mild steel, 39. 
 Single-riveted butt-joints, 77. 
 Single-riveted butt-joints, efficiencies 
 
 of, 76. 
 Single-riveted butt-joints, tests, 78. 
 Single-riveted lap-joint, reinforced welt, 
 
 9°- 
 Single-riveted lap-joint, tests. Table 
 
 XIII. 
 Single-riveted lap-joints, 82. 
 Single-riveted lap-joints, efficiencies of, 
 
 83. 
 Siphon pressure-gauge, 374. 
 Six-inch flue boilers, 186. 
 Slow cooling of castings, 27. 
 Smith's balanced expansion-joint, 382. 
 Smoke, 17. 
 
 Smoke connections, 244. 
 Spiral punch, 53. 
 Stability of chimney, 397. 
 Standard Boiler Company, 345. 
 Stay-bolt, material for, 141. 
 Stay-bolt patch, 136. 
 Stay-bolt, strains on, 136. 
 Stay-bolt with drilled hole, 136. 
 Stay-bolts, Cole's experiments, 138. 
 Stay-bolts, flexible, 137. 
 Stay-bolts for flat surfaces, 135. 
 Stay-bolts, proportions for, 136. 
 Stay-centres, locating, 121. 
 Stay, end fastenings for, 130. 
 Stay, longitudinal, 127. 
 Stay, radial, 133. 
 Stay-tubes, 147. 
 
404 
 
 INDEX 
 
 Stay, working pressure by Lloyd's rule, 
 
 123. 
 Stays and braces, details of, 125. 
 Stays, diagonal, rules for, 129. 
 Stays for boiler-head, 120. 
 Stays or braces, United States rule, 122. 
 Steam-blast, cost of, 221. 
 Steam-dome, proportions for, 168. 
 Steam-domes, 166. 
 Steam-drum, 169. 
 Steam-pipe, 387. 
 Steam-room in boilers, 175. 
 Steam stop-valve, 377. 
 Steel for stay-bolts, 141. 
 Steel plates and steel rivets, 57. 
 Steel plates, annealing, 108. 
 Steel plates, flanging of, 108. 
 Steel rivets, chemical analysis, 59. 
 Steel rivets, heating of, 94. 
 Steel rivets, physical qualities, 59. 
 Steel rivets, properties of, 60. 
 Steel stay-bolts. United States rule for, 
 
 122. 
 Steel, texture of, 37. 
 Stirling boiler, 322, 341. 
 Stop- and check-valve combined, 379. 
 Straps for suspending boilers, 159. 
 Strengthening manholes, 153. 
 Strength of castings, 28. 
 Strength of iron and steel plates, 50. 
 Sulphur, 10, II. 
 Sulphur in cast iron, 26. 
 Sulphur in mild steel, 39. 
 Sulphurous acid, 17. 
 Sulphurous oxide, 17. 
 Supporting boilers in furnace, 158. 
 Surface blow, 350. 
 Swinging manhead, 157. 
 
 Tan, heating power of, 12. 
 Temperature in chimney, 392. 
 Temperature in welding, 96. 
 Temperature of fire, 20. 
 Tensile strength of cast iron, 29. 
 Tensile strength of wrought iron, 32. 
 Tensile test of mild steel, 39, 44. 
 Test-piece, American Society of Civil 
 
 Engineers, 41. 
 Test-piece, effect of length, 41. 
 Test-piece, 8-inch, 41. 
 Test-piece, short form, 40. 
 Test-piece, standard, 42. 
 
 Test-pieces, wrought-iron, 32. 
 Thermal unit, 20. 
 Tiles for water-tube boiler, 321. 
 Tonkin portable boiler, 294. 
 Transverse strength of cast iron, 29. 
 Triple-riveted butt-joints, tests, Table 
 
 XXI. 
 Triple-riveted lap-joint, 72. 
 Triple-riveted lap-joint, proportions, 
 
 87, 89. 
 Triplex boiler, 202, 339. 
 Tube area to grate area, 195. 
 Tube, distance from shell, 189. 
 Tube-expanders, 143. 
 Tube, proper length of, 193. 
 Tube-sheet, distortion by expander, 
 
 145- 
 Tube-spacing, defective, 188. 
 Tube, vertical spacing, 191. 
 Tubes as stays, 147. 
 Tubes, boiler, 141. 
 Tubes, boiler, standard dimensions, 
 
 table, 142. 
 Tubes, central water-space between, 
 
 188. 
 Tubes, ferrules in, 145. 
 Tubes, holding power of, 146. 
 Tubes, horizontal, distance between, 
 
 190, 192, 194, 198. 
 Tubes, strength of, 142. 
 Tubes, threaded for nut, 147. 
 Tubes, vertical arrangement of, 193. 
 Tubular boiler-head, 3-inch, 192. 
 Tubular boiler-head, 3>^-inch, 194. 
 Tubular boiler-head, 4-inch, 198. 
 Tubular boiler, performance, 338. 
 Tubular boiler, vertical, externally 
 
 fired, 205. 
 Tubular boilers, 187. 
 Turnbuckle, 131. 
 Two-flue boiler, 183, 338. 
 
 Unions, malleable iron, 387. 
 Unit of work, 9. 
 
 Vertical flue boiler, 259. 
 Vertical tubular boilers, 256, 340. 
 
 Water, composition of, 10. 
 Water-gauge, 368. 
 Water-gauge barrel, 371. 
 Water-pan in ash-pit, 215. 
 
INDEX 
 
 405 
 
 Water surface in boilers, 175. 
 Water-tube boilers, 296, 299. 
 Welded joint, fractures in, 104. 
 Welded joint, strength of, 103. 
 Welded joints, annealing, 100. 
 Welding and flanging, 95. 
 Welding bars, 98. 
 Welding, Bertram's method, 99. 
 Welding, cost of, 104. 
 Welding, efficiency of, 104. 
 Welding furnace flues, 103. 
 Welding, localizing heat in, 99. 
 Welding plates, 98. 
 Welding, practical results, 100. 
 Welding, tests of, 102. 
 
 Welding wrought iron, 33. 
 Wharton-Harrison boiler, 298, 341. 
 Whistle, 386. 
 
 Whitham, J. M., quoted, 199. 
 Wilkinson mechanical stoker, 220. 
 Williams rotating gauge-cock, 367. 
 Wings, or brackets, for boilers, 158. 
 Wood, composition of, 12. 
 Wood, heating power of, 12. 
 Wrought iron, 31. 
 Wrought-iron pipe, 387. 
 Wrought iron, welding of, 95. 
 
 Zell water-tube boiler, 306. 
 Zigzag riveting, 85. 
 
 THE END. 
 
r r y r 
 
 IN PREPARATION FOR IMMEDIATE PUBLICATION 
 
 CHIMNEYS 
 
 OF BRICK . . . 
 . . . AND METAL 
 
 CONSIDERED IN THEIR RELATIONS TO 
 STEAM ENGINEERING. 
 
 VOL. 11.^ STEAM ENGINEERING SERIES. 
 
 EDITED BY 
 
 WILLIAM M. BARR, 
 
 MEMBER AMERICAN SOCIETY MECHANICAL ENGINEERS. 
 
 With upwards of 750 illustrations, including 75 full-page plates, covering designs 
 
 for chimneys from 25 to 5000 horse-power from drawings 
 
 executed expressly for this work. 
 
 One Volume. Octavo. Uniform with Boilers and Furnaces. By Subscription only. 
 Price, $3.00, which includes free delivery. 
 
 EACH VOLUME IN THIS SERIES COMPLETE IN ITSELF. 
 
 P. O. Box 803. 
 
 THE FLORENCE COMPANY, 
 
 PHILADELPHIA, PA. 
 
SPECIMEN PLATE.