. 
 
 v 
 
 mSm 
 
 v .< 
 
 .> 
 
 nr *>" 
 
 
 m 
 M 
 
 
 
LIBRARY 
 
 OF THE 
 
 UNIVERSITY OF CALIFORNIA. 
 
 Class 
 
STEAM BOILERS: 
 
 THHE 
 
 DESIGN, CONSTRUCTION, AND MANAGEMENT. 
 
 BY 
 
 WILLIAM H. SHOCK, 
 
 EXGINEER-IN-CHIEF, U. 8. N., 
 CHIEF OF BUREAU OF STEAM ENGINEERING, U. S. NAVY. 
 
 NEW YORK : 
 
 D. VAN NOSTKAND, PUBLISHER, 
 
 23 WAEEEX AXB 27 MUEKAY STREETS. 
 1880. 
 
Copyrighted by 
 
 D. VAN NOSTRAND, 
 
 1880. 
 
 H. J. HEWITT, PRINTER AND ELECTROTYPER, 27 EOSE STREET, NEW YORK. 
 
TO THE 
 
 HONORABLE R. W. THOMPSON, 
 
 SECRETARY OF O. S. NAVY, 
 
 RECOGNITION OF HIS EMINENT PUBLIC SEBVICBS AND PBIVATE WORTH, 
 
 THIS BOOK 
 
 IS RESPECTFULLY INSCRIBED 
 BY THE AUTHOR. 
 
 292054 
 
PREFACE. 
 
 THE author of this work indulges the hope that he has in some degree supplied a 
 long-felt need in this particular branch of engineering and construction. 
 His thanks are due to Chief Engineer Frederick G. McKean and Passed Assistant 
 Engineer Charles R. Roelker, United States Xavy, for the cordial assistance given him 
 in its preparation ; especially to Engineer Roelker for the great care and excellent judg- 
 ment exercised in arranging and classifying the various formulae and data used. He 
 would also tender to Chief Engineer B. F. Ishenvood, United States Navy, his thanks 
 for valuable suggestions. 
 
 In quoting authorities he has endeavored to give due credit where it belonged, and 
 any omission in this particular is to be attributed to unintentional oversight. 
 
SYNOPTICAL INDEX. 
 
 CHAPTER I. 
 
 LSTBODTJCTOBY EEMARKS. 
 
 Materials used in the construction of boilers in the early days of the steam-engine. Eeasons 
 why copper was superseded by plate-iron. Latest changes in the form and construction of 
 boilers in consequence of the introduction of the compound engine and of high steam-pres- 
 sures. 
 
 The essential parts of a steam-boiler and their functions. The form of boilers may be varied 
 almost infinitely. General conditions which determine the peculiar features of marine boilers. 
 Special conditions affecting the efficiency of boilers in war-steamers. 
 
 CHAPTER II. 
 
 COMBUSTION. 
 
 Section 1. Elementary Constituents of Fuels. 
 
 Definition of the terms combustion and combustibles. Chief combustible constituents of fuels. 
 Chemical equivalents. Table I., exhibiting the principal chemical and physical properties of 
 various elementary substances found in different fuels, and of the most important compounds 
 resulting from their combustion. 
 
 Section 2. Temperature of Ignition. 
 Section 3. Combustion of the Constituents of Fuels. 
 
 Combustion of free carbon forming carbonic acid or carbonic oxide. Combustion of hydro- 
 carbons. Formation of soot, smoke, and flame. Effect of the presence of oxygen and hydro- 
 gen in fuel. Nitrogen. Sulphur. Ash and clinker. 
 
 Section 4. Total Heat of Combustion. 
 
 British thermal unit. Table II., containing quantities of heat developed by combustion of carbon, 
 hydrogen, and their compounds, and the weight of oxygen and of atmospheric air required for 
 their combustion. Calorific power of hydrocarbons. Dulong's law of the calorific power of 
 fuels. General formulae for computing the theoretical calorific power of fuels. The actual 
 calorific power of coals cannot be determined with exactness by these formulas. Vaporific 
 power of coals in a steam-boiler. W. R. Johnson's experiments. 
 
4 SYNOPTICAL INDEX. 
 
 Section 5. Fuel as a Source of Power. 
 
 Joule's equivalent;. Various determinations of the mechanical equivalent of heat. Energy de- 
 veloped by the combustion of a pound of coal. 
 
 Section 6. Air required for Combustion. 
 
 Conditions essential to the perfect combustion of a fuel. Formula for calculating the quantity of 
 oxygen required for the complete combustion of fuels. Weight of atmospheric air required 
 for the combustion of a pound of coal. Admission of air in excess of the amount theoreti- 
 cally required for the complete oxidation of the fuel. 
 
 Section 7. Temperature of Combustion. 
 
 Theoretical calorific intensity of fuels. Table III., containing the theoretical temperatures pro- 
 duced by the perfect combustion of various substances. Bunsen's experiments on dissociation. 
 Probable furnace-temperatures in marine boilers. 
 
 Section 8. Volume of Products of Combustion. 
 
 Volume of gases of combustion nearly equal to that of the air supplied to the furnace. Volume of 
 air. Formula for calculating the volume of gises at different temperatures. Table IV., ex- 
 hibiting the volumes of furnace-gases per pound of fuel at different temperatures. 
 
 Section 9. Rate of Combustion. 
 
 Eates of combustion of different fuels in marine boilers with natural and with artificial draught. 
 Bituminous coals compared with hard anthracites in respect of rate of combustion. 
 
 Section 10. Draught of Furnaces. 
 
 What produces the draught of a boiler. Peclet's formula for the head which measures the draught 
 of a boiler. Resistances to the flow of the gases of combustion in the furnace and flues of a 
 
 boiler. Additional loss of head in marine boilers. 
 
 
 
 Section 11. Cliimney-draught. 
 
 Head produced by the draught of a chimney. Formulae for calculating the weight, in pounds, of a 
 cubic foot of air and of a cubic foot of chimney-gas. Formula for calculating the unbalanced 
 pressure due to a chimney of a given height. Formula for calculating the velocity with which 
 the air flows to the grate of a furnace. Conclusions drawn from the latter equation. Practi- 
 cal limits to the height and temperature of the chimney of marine boilers. 
 
 Section 12. Artificial Draught. 
 Head produced by a blast-pipe. Work to be done by a fan or other blowing-machine, 
 
 Section 13. Efficiency of Furnace. 
 
 Conditions affecting the rate of combustion. Regulating the draught by a damper. Thickness of 
 the bed of fuel. Absorption of heat by incombustible matter and by moisture contained in 
 the fuel. Waste of unburnt combustible matter in the solid state. Losses from an admission 
 
SYXOPTICAL IXDEX. 5 
 
 of excessive quantities of air to the furnace. Experiment made by a board of United States 
 naval engineers to determine the loss due to this cause. Determination of the loss of effi- 
 ciency resulting from opening the furnace-doors in cleaning the fires. Waste of uuburnt fuel 
 in the gaseous and smoky states. 
 
 Table V., showing the character and efficiency of American coals, as determined by W. R. Johnson. 
 
 Table VI., showing the character and efficiency of English coals, as determined by De la Beche and 
 Playfair. 
 
 Table VI. a, showing results of experiments on various coals of the carboniferous and cretaceous 
 periods, made by Chief-Engineer B. F. Isherwood, TJ.S.N. 
 
 CHAPTER III. 
 
 TRANSMISSION OF HEAT AND EVAPORATION. 
 
 Section 1. Laics of Transmission of Heat. 
 
 Conditions affecting the quantity of heat transmitted and the rate of its transmission. Radiation. 
 Conduction. Convection. Formula for the rate of internal conduction through a solid 
 body. Coefficients of thermal resistance. The thermal conductivity of wrought-iron deter- 
 mined by Forbes. Formula for the rate of conduction through several layers of different 
 substances. Formula for the rate of external conduction between a solid body and a fluid. 
 
 Section 2. Experiments on the Transmission of Heat, by B. F. Isherwood, U.S.N. 
 
 Description of the manner of making these experiments. General results. Thermal conductivity 
 of copper, brass, wronght-iron, and cast-iron. 
 
 Section 3. Experiments on the Transmission of Heat, by Peclet. 
 
 How the results obtained by Peclet differ from those obtained by Isherwood. What caused this 
 difference. Peclet's experiments on the cooling of vessels exposed to the air. 
 
 Section 4. Transmission of Heat in a Steam-boiler. 
 
 Evaporative power of the heating-surfaces in a steam-boiler. Rankine's approximate formula for 
 the quantity of heat transmitted by the heating-surfaces of a boiler. Circulation of the water 
 in boilers. Heat-absorbing capacity of water. 
 
 Section 5. Efficiency of Heating-surfaces in a Steam-boiler. 
 
 Measure of the efficiency of hea'ing-snrfaces. Conditions affecting the efficiency of heating-sur- 
 faces. Armstrong's experiment on the efficiency of heating-surfaces. Differences of tempe- 
 rature in a steam-boiler. Transmission of heat through the plates forming the furnace and 
 the combustion-chamber. Radiation of heat from incandescent carbon. Evaporation from 
 the heating-surfaces of the furnaces and back-connections of return-tube boilers, determined 
 by Isherwood. Evaporation from the heating-surfaces of tubes. Differences of temperature 
 of gases discharged from the upper and lower rows of horizontal fire-tubes. 
 
6 SYNOPTICAL INDEX. 
 
 Section G. Loss of Efficiency of Boilers by External Radiation and Conduction. 
 
 The loss of heat by radiation from the furnace and from the chimney is trifling. Experiments on 
 the loss of heat by radiation and conduction from steam-boilers, pipes, etc., protected by dif- 
 ferent thicknesses of felting. Experiments to determine the effect of covering the boiler with 
 felt on the economic evaporation, made at the Navy- Yard, New York, in 1863 
 
 Section 7. Efficiency of Boilers. 
 
 Measure of the efficiency of boilers. Heat expended in the production of chimney-draught. 
 Rankine's formula for the efficiency of boilers. 
 
 Section 8. Influence of the Rate of Combustion on the Evaporative Efficiency of Boilers. (From 
 "Experimental Researches," Vol. II., by Chief-Engineer B. P. Isherwood, U.S.N.) 
 
 Table VII , showing the economical and potential evaporation of the horizontal fire-tube boiler 
 with anthracite consumed with different rates of combustion. 
 
 Section 9. Superheated Steam. 
 
 Drying and superheating steam. Tute and Fairbairn's experiments on the density of superheated 
 steam. Increase of the dynamic efficiency of steam in the engine by drying and superheating. 
 Results of Isherwood's experiments with superheated steam. TT.S.S. Mackinaw. U.S.S. 
 Eutaw. Steamer Georgeanna. Practical limits to the degree of superheating. 
 
 Section 10. Efficiency of Superheating Surfaces. 
 
 Superheating surfaces in a boiler. Different methods of superheating steam. Relative economic 
 value of water-heating and superheating surfaces in marine boilers. Value of superheaters 
 practically considered. 
 
 Table VIIL, showing properties of water and steam. 
 
 CHAPTER IV. 
 
 MATERIALS. 
 
 Section 1. Relative Value of Materials for Bt>iler Construction. 
 Section 2. Copper. 
 
 Advantages and disadvantages connected with the use of copper as a material for boilers. Present 
 use of copper in boiler-making. 
 
 Section 3. Composition. 
 General character. Brass. Brass boiler-tubes. Bronze. Phosphor-bronze. 
 
 Section 4. Tenacity of Metals at Various Temperatures. 
 
 Experiments made at Portsmouth Dockyard, England. Table IX., showing tenacity of various 
 metals at different temperatures up to 500 Fahr. 
 
SYNOPTICAL INDEX. 7 
 
 Section 5. Cast-iron. 
 Section 6. Wrought-iron. 
 Its use in boiler-making. Qualities which, it must possess for this purpose. 
 
 Section 1. Brands of Plate- iron used in Boiler-making. 
 
 American charcoal-irons. Shell-iron. Flange-iron. Firebox-iron. Special brands of boiler-iron. 
 Sizes of boiler-plates. English boiler-iron. 
 
 Section 8. Steel. 
 
 Processes of manufacture. Qualities of mild steel used in boiler-making. Steel compared with 
 wrought-iron. Tenacity of steel boiler-plates. Effects of severe strains and of corrosion. 
 Extracts from a report by the surveyors to Lloyd's Registry on the use of steel for boilers. 
 
 Table X., exhibiting certain physical and mechanical properties of various metals. 
 
 Table XI.. weight of wrought-iron plates and bars (square and round). 
 
 Table XII., weight of flat bar-iron per foot. 
 
 Table XIII., weight of sheet and plate iron. 
 
 Table XIV., weight of wrought angle-iron. 
 
 Table XV., weight of wrought T-iron. 
 
 Table XVI., wrought-iron bolts with square heads and nuts. 
 
 Table XVII., standard sizes of washers. 
 
 Table XVIII., showing number of Burden's rivets in 100 pounds. 
 
 CHAPTER V. 
 
 TESTING THE MATERIALS. 
 
 Section 1. General Character of Tests. 
 
 Section 2. United States Laics and Regulations regarding the Tests of Boiler-plates. 
 United States Revised Statutes, sec?. 4430 and 4431. Rules 3 and 4 of General Rules and Regu- 
 lations prescribed by the Board of Supervising Inspectors of Steam Vessels, 1879. 
 
 Section 3. The Rodman Testing-machine. 
 
 Description of the testing-machine at the Ordnance Department of the Navy- Yard, Washington, 
 D. C. Manner of making tests with this machine. 
 
 Section 4. Form and Dimensions of Tent-specimens. 
 
 Sectional area. Length. Form. Experiments on the influence of length and form on the appa- 
 rent strength of specimens. Care to be taken in preparing specimens. 
 
 Section 5. Effects produced by Stress. 
 
 Rate of elongation of wrought-iron under a tensile stress. The interior and exterior portions of a 
 wrought-iron bar not in equilibrium. Stretching of ductile and fibrous materials. Flow of 
 solids. 
 
8 SYNOPTICAL INDEX. 
 
 Section 6. Experiments on the Effects of Hammering and Rolling on the Strength of Bars. 
 
 Results deduced from experiments by the U. S. Test- Board. Table XIX., showing the effect of 
 variation and of uniformity in the rate of reduction from pile to bar on the tensile strength 
 and the elastic limit of wrought-irou bars. Experiments made by Chief-Engineer William H. 
 Shock, U.S.N., on the influence of hammering and rolling on the tenacity and ductility of a 
 wrought-iron bar. 
 
 Section 7. Appearance of Fractures. 
 
 Quality of iron and steel bars indicated by the appearance of their grain or fibre when fractured. 
 Appearance of the same bar may be changed by altering the manner of fracturing it. 
 
 Section 8. Hot and Cold Forge-tests. 
 Forge-tests applied to boiler-plates. Applying the bending-test to plates. Testing rivets. 
 
 Section 9. Directions for Testing Bar-iron. 
 Section 10. Testing Steel Boiler-plates (French Government tests). 
 
 Section 11. Tests for Plate, Beam, Angle, Bulb, and Bar Steel used in Building Ships for Her 
 
 Majesty's Navy (Admiralty, 9th January, 1879). 
 
 Section 12. Examining Boiler-plates. 
 CHAPTER VI. 
 
 PEINCIPLES OF THE STRENGTH OF BOILERS. 
 
 Section 1. Resistance of Spherical Shells to an Internal Fluid Pressure. 
 
 Stress produced by an internal fluid pressure. Strength of thin spherical shells. Stress in thick 
 spherical shells. 
 
 Section 2. Resistance of Cylindrical Shells to an Internal Fluid Pressure. 
 
 Stress in a longitudinal direction. Stress in a tangential direction. End-attachment of cylindri- 
 cal shells. Navier's experiment on strength of wrought-iron shells. 
 
 Section 3. Resistance of Cylindrical Shells to an External Fluid Pressure. 
 
 Fairbairn's formula. Table XX., containing the 2.19th power of several numbers. Belpaire's in- 
 vestigation of the collapsing strength of flues. Table XXI., containing numerical values for 
 factor S of Belpaire's formula. Grashof's formula. 
 
 Section 4. Experiments made on the Resistance of Cylindrical Flues to an External Fluid Pressure. 
 
 Description of apparatus used in experiments made at the Navy- Yard, Washington, D. C., in 1874. 
 Results of experiments. Application of Fairbairn's and Belpaire's formulae to these expe- 
 riments. 
 
SYNOPTICAL INDEX. 9 
 
 Section 5. Strength of Flat Plates. 
 
 Investigation of the strength and stiffness of flat plates. Formulae for calculating the strength of 
 unstayed flat, circular, rectangular, and square plates. Stiffness of flanged flat plates. 
 Stiffness of buckled plates. 
 
 Section 6. Strains on Braces and their Attachments. 
 Perpendicular braces. Oblique braces. Strains in a system of oblique braces. 
 
 Section 7. Strains on Circular Arcs. 
 CHAPTER VII. 
 
 DESIGN", DBAWINGS, AND SPECIFICATIONS. 
 
 Section 1. General Considerations governing the Design of Marine Sailers. 
 
 Section 2. Boiler Power. 
 
 Relative evaporative efficiency. Actual power of a boiler. Consumption of steam in marine 
 engines. Number of indicated horse-powers per square foot of grate. 
 
 Section 3. Various Types of Marine Boilers. 
 
 Various forms of boiler-shelK Water-tube and fire-tube boilers. Rectangular boilers of United 
 States naval vessels. Boiler of U.S.S. Lacka wanna. Rectangular boilers with two tiers of 
 furnaces. Arrangement of tubes in oval and circular shells. Size and number of furnaces in 
 cylindrical boilers. Cylindrical boiler with the tubes arranged at the sides of a single furnace. 
 Double-end cylindrical boiler. Boiler of U.S.S. Daylight. Boiler of U.S.S. Mahaska. Lo- 
 comotive types of boiler designed for marine and railroad purposes. Vertical fire-tube boiler. 
 Dickerson's marine boiler. 
 
 Table XXII., showing the dimensions, proportions, and weights of boilers of various types. 
 
 Table XXIII., showing the economic evaporation of boilers of various types under different con- 
 ditions. 
 
 Section 4. Space and Weight required for Boilers of a Given Power. 
 
 Relative proportions of weight and space required for boiler, fire-room, and fuel. Solution of the 
 problem for a given horizontal fire-tube boiler. Determination of the best rate of combustion 
 for the given boiler. Table XXIV., exhibiting the space and weight required with the hori- 
 zontal fire-tube boiler, having a rectangular shell and the tubes arranged above the furnaces, 
 and with anthracite with one-sixth refuse, to furnish a given supply of steam per hour for 200 
 hours, with different rates of combustion. 
 
 Section 5. Proportioning the Parts of a Boiler. 
 
 Length and width of the grate. Ashpit. Furnace. Calorimeter over the bridge-wall. Back- 
 connection. Calorimeter through or between the tubes. Calorimeter of the chimney. Heat- 
 ing-surface. Water-spaces. Water and steam room. 
 
10 SYNOPTICAL INDEX. 
 
 Section 6. Relative Value of Various Forms for Boiler Construction. 
 Spherical shell. Cylindrical form. Flat surfaces. 
 
 Section 7. Factor of Safety. 
 
 Factor of safety variable. Allowance for corrosion. Stiffness of stayed surfaces. Limit of high- 
 est working pressure. Strains in consequence of variations of temperature. Factor of safety 
 for steel boiler-plates. Section 4433 of United States Revised Statutes. Factor of safety of 
 flues subjected to compression. Stays and braces. 
 
 Section 8. Drawings and Specifications. 
 
 Section 9. Specifications for Sailers of U.S.S. " Lackawanna." 
 
 Section 10. Extract from Specifications for Engines of U.S.S. " Miantonomoh." 
 
 Section 11. Specifications of Boilers (Iron Shells) for Vessels of the English Navy. 
 
 Section 12. Material for Six Boilers of U.S.S. " Nipsic." 
 Section 13. List of Steel and Iron Plates, Rivets, Tubes, etc., for Boiler of Steamer " Lookout." 
 
 CHAPTER VIII. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDIN.G, ETC. 
 
 Section 1. Laying-off. 
 
 Levelling plates. Marking lines and rivet-holes. Laying-off the front-head and tube-sheet of the 
 boilers for U.S.S. Nipsic. Finding the length of plates for a cylindrical shell with butt-joints. 
 Laying-off a cylindrical shell or flue formed of alternate inner and outer rings with lap-joints- 
 Laying-off plates for conical tubes. Laying-off a cylindrical shell or flue formed of rings 
 lapping telescopically. Laying-off plates for a cylindrical shell with a wedge-shaped portion 
 cut off. Laying-off plates for the cylindrical shell of a steam-dome. 
 
 Section 2. Shearing and Planing. 
 
 Cutting openings for manholes, furnaces, etc., in boiler-plates. Shearing-machine. Relative ad- 
 vantages of shearing and planing. 
 
 Section 3. Bending. 
 Arrangement of bending-rolls. Bending plates to a circular shape. 
 
 Section 4. Flanging. 
 
 Bending plates cold. Flanging iron and steel plates. Practical instructions regarding flanging. 
 Circular flanges. 
 
 Section 5. Punching. 
 
 Process of punching plates. Forms of punches. Kennedy's helical punch. Sizes of punch and 
 die. Power required to punch steel plates. Plates to be punched from the faying sur- 
 faces. Form of punched holes. Loss of tenacity due to punching. 
 
SYNOPTICAL INDEX. 11 
 
 Section 6. Drilling. 
 
 Precautions in drilling boiler-plates. Drilling compared with punching. Harvey's boiler-drilling 
 machine. Method of drilling plates of cylindrical boilers of U.S.S. Galena. 
 
 Section 7. Riveting. 
 
 Half-blind rivet-holes. Drifting. Operation of hand-riveting. Effect of percussion on the 
 strength of iron and steel. Steam and hydraulic riveting machines. Hand-riveting and ma- 
 chine-riveting compared. Steam riveting-machine of the Providence Steam-Engine Company, 
 Providence, R. I. Tweddell's hydraulic machine tools. Stationary and portable hydraulic 
 riveting-machines. Number of rivets driven per day's work. Cold riveting. Steel rivets. 
 
 Section 8. Forms of Rivets. 
 
 Dimensions of a f-inch rivet. Forms of rivet-points. Table XXV., giving lengths of shank re- 
 quired to form different rivet-points. Snap points. Conical points. Countersunk rivets. 
 
 Section 9. Stylts of Joints. 
 Section 10. Friction in Riveted Joints. 
 
 Reed's experiments to determine the amount of friction in riveted joints. Friction in its relation 
 to the strength of riveted joints. Causes tending to diminish the friction of riveted joints. 
 
 Section 11. Straining Action on Riveted Joints. 
 
 Conditions producing the strongest joints. Apparent breaking stress defined. Unequal distribu- 
 tion of stress due to eccentricity of load. Local action of contiguous material. Influence of 
 holes on distribution of stress and on the apparent strength of plates. Strained zones around 
 punched holes. Excessive crushing action between rivets and plates. Stress in multiple- 
 riveted joints of materials of different elasticities. 
 
 Section 12. Strength of Materials in Riveted Joints. 
 
 Loss of tensile strength in riveted iron and steel plates due to punching. Shearing strength of iron 
 and steel rivets according to W. R. Browne, Fairbairn, David Greig, and Max Eyth. Formula 
 expressing relation between crushing pressure and shearing resistance of rivets. Limit of 
 crushing pressure in riveted joints. Maximum crushing pressure of the rivets on the plate. 
 Influence of crushing pressure on apparent tenacity of riveted plates. Probable influence of 
 the relative hardness of the rivets and plates. 
 
 Section 13. Proportioning Riveted Joints. 
 
 Different modes of fracture of riveted joints. Elements affecting the efficiency of riveted joints. 
 Formula for finding the diameter of rivets for a given thickness of plate, with the rivets in 
 single-shear and in double-shear. Usual practice in proportioning the diameter of rivets to 
 the thickness of plates for iron plates, and for steel plates. Width of lap. 
 
 Section 14. Lap-joint*. 
 Single-riveted lap-joint. Its characteristic features. Formula for finding the pitch of rivets. 
 
12 SYNOPTICAL INDEX. 
 
 Double-riveted lap-joint. Formula for finding the pitch of rivets. Zigzag riveting. Chain-rivet- 
 ing. 
 Treble and quadruple riveted lap-joints. 
 
 Section 15. Experiments on the Strength of Lap-joints. 
 
 Fairbairn's experiments. Eeasons why Fairbairn's conclusions regarding the strength of riveted 
 joints cannot be accepted as trustworthy. Experiments by W. Bertram published and dis- 
 cussed by D. K. Clark. Decrease of the apparent tenacity of riveted lap-joints as the thickness 
 of plates increases. Mean results of experiments with single-riveted joints. Loss of apparent 
 tenacity of joints with punched and drilled holes. Mean shearing resistance of rivets. Mean 
 efficiency of single-riveted joints. 
 
 Mean apparent tenacity and efficiency of double-riveted joints. R. V. J. Knight's experiments with 
 double-riveted joints made of ^-inch and 1-inch punched iron plates. Average results of ex- 
 periments with double-riveted steel plates. 
 
 Experiments by Kirkaldy on treble and quadruple riveted steel plates. Table XXVI., results of 
 experiments with treble and quadruple riveted lap-joints, steel plates. Tested by Kirkaldy. 
 
 Table XXVII., proportions of single-riveted joints. 
 
 Table XXVII a, French practice in single-riveted joints. 
 
 Table XXVIII., proportions of double-riveted lap-joints. 
 
 Section 16. Various Forms of Lap-joints. 
 
 Single-riveted diagonal lap-joint. Single-riveted joint with oval rivets. Single-riveted lap-joint 
 with covering-plate. Plates with thickened edges. 
 
 Section 17. Butt-joints. 
 
 Single-welt butt-joint. Double-welt butt-joints. Formulae for finding the pitch of rivets in single- 
 riveted and double-riveted double-welt butt-joints. Proportions of single-riveted double-welt 
 butt-joints according to the practice of the Crewe Works (England). Modified proportions 
 when double-welt butt-joints and lap-joints occur on the same plate. Table XXIX., Wilson's 
 table of proportions of double-riveted butt-joints with two covering-plates. 
 
 Table XXX., Kirkaldy's experiments on the comparative strength of chain and zigzag riveting in 
 double-welt butt-joints. 
 
 Table XXXI., strength of rivet- work in single and double riveted lap and butt-joints, with drilled 
 and punched plates. Experiments with chain and zigzag riveted double-welt butt-joints, steel 
 plates and steel rivets . 
 
 Table XXXII., results of experiments with single and double riveted double-welt butt-joints. 
 
 Section 18. Calking. 
 
 Method of calking the lap-edges of joints. Ordinary calking- tool. Connery's calking-tool. Calk- 
 ing the butts of plates. Precautions to be observed in calking. 
 
 Section 19. Welding. 
 
 Definition of welding. Defective welding due to the presence of oxide of iron. Effect of chemical 
 composition on the welding of iron. A. L. Holley on welding iron. 
 
SYNOPTICAL INDEX. 13 
 
 Section 20. Welding Boiler-plates. 
 
 Bertram's method of welding boiler-plates. Scarf-weld. Lap-weld. Welded seam with covering- 
 plate. Welding furnace-flues. Welding the longitudinal seams of a cylindrical boiler-shell. 
 Welding the front plates of a boiler. Welding angle-iron rings. Practical suggestions con- 
 cerning welding. Difficulties connected with the welding of plates. 
 
 Section 21. Strength of Welded Plates. 
 
 Results of experiments on the tensile strength of welded joints recorded by Kirtley. Experiments 
 on the strength of welded joints by Gillott. 
 
 CHAPTER IX. 
 
 SHELL, FURNACES, AND BACK- CONNECTIONS. 
 
 Section 1. Various Forms of Shells. 
 
 C. E. Emery's connected-arc marine boiler. Thickness of cylindrical plates. Thickness of flat 
 plates. Stiffening plates around manholes. Quality of iron for boiler-shells. 
 
 Section 2. Rectangular Shells. 
 
 Rectangular boilers of IT. S. naval vessels. Thickness of plates. Manner of connecting the plates. 
 
 Section 3. Cylindrical Shells. 
 
 Manner of building up cylindrical shells. Back and front heads. Lloyd's rules for determining 
 the strength of circular shells. Rules of the surveyors of the Board of Trade (England) for 
 determining the strength of cylindrical shells. 
 
 Section 4. Furnaces. 
 
 General arrangement of seams, laps, and braces on furnaces. Quality of iron for furnaces. Use 
 of steel and copper for furnaces. Corrugated furnace-flues. 
 
 Section 5. Furnaces of Rectangular Boilers. 
 
 Circulation of water with flat and arched furnace-crowns. Relative advantages of various forms of 
 furnace- oro WQB. Manner of securing furnaces to the front of boilers and to the back-connec- 
 tions. Furnace of a marine tubular boiler built by Laird & Son, Birkenhead, England. 
 Furnace for a boiler of the U. S. Tugboat Glance. Manner of securing furnaces to the shell 
 when the water-spaces are narrow. Usual method of securing furnaces in rectangular boilers 
 of U. S. naval vessels. Water-bottoms. Water-legs of dry-bottom boilers. 
 
 Section 6. Cylindrical Furnaces. 
 
 Construction of cylindrical furnaces. The Adamson joint. T-iron rings for furnace-flues. The 
 Bowling hoop. Directions regarding the application of strengthening-hoops to furnace-flues. 
 Manner of securing furnace-flues to the boiler-shell. Description of furnaces of boiler illus- 
 trated on Plate XV. Rules of Lloyd's and of the Board of Trade (England) for the strength 
 of circular flues. 
 
14: SYNOPTICAL INDEX. 
 
 Section 7. Combustion-chambers and Back-connections. 
 
 Combustion-chamber in various type8 of boilers. Bridge- walls. Arrangement of back-connections. 
 Construction of back-connection in horizontal tubular and vertical return-tube boilers. 
 
 Section 8. Systems of Boiler-building. 
 
 General rules observed in boiler-building. Method pursued in building the cylindrical boiler repre- 
 sented on Plate XII. Building rectangular boilers. 
 
 CHAPTER X. 
 
 STAYS AND BRACES. 
 
 Section 1. Systems of Bracing. 
 
 Surfaces to which bracing has to be applied. Methods of staying and bracing. Spacing of braces. 
 Various devices in bracing to keep the interior of the boiler accessible. Care required to 
 prevent distortion of plates to which braces are attached. System of bracing in cylindrical 
 boilers. System of bracing in rectangular boilers of U. S. naval vessels. Bracing in English 
 horizontal return-tube boilers. Bracing in horizontal return-tube boilers of U.S.S. Tippeca- 
 noe. System of bracing in rectangular boilers of French naval vessels. 
 
 Section 2. Rules for Proportioning Braces. 
 
 Area of plate supported by a brace. Factor of safety. Corrosion of braces. Lloyd's rules for 
 determining the strength of stays and flat plates. Board of Trade's (English) rules for deter- 
 mining the strength of stays and flat plates. Weisbach's formula for the thickness of flat 
 stayed plates. Eatio of diameter of stays to thickness of plate. Gussets. Girder-stays. For- 
 mulae for the strength and depth of girder-stays. Board of Trade's (English) rule for finding 
 the strength of girder-stays. Rankirie's rules for the strength of easy-fitting and tight-fitting 
 fastenings of braces. Bearing-surface of connecting-pins or bolts. Experiments by Charles 
 Fox. 
 
 Section 3. Screw-stays and Socket-bolts. 
 
 General arrangement and relative advantages of various forms of stay-bolts and rivets. Sockets. 
 Screw-stays. 
 
 Section 4. Various Forms of Stays and Modes of Fastening Them. 
 
 Stay used in narrow water-spaces of French boilers. Staying the vertical sides of tube-boxes or 
 back-connections by lugs and pins. Form of stay used in boilers of U.S.S. Lancaster. Short 
 stays connecting surfaces that are not parallel. Forked stays. Triangular stays attached to the 
 lower rounded corners of furnaces. Stay-rods secured by nuts and washers. Staying the flat 
 ends of cylindrical boilers. Rod-braces of boilers of U.S.S. Terror. Do. of U.S.S. Amphitrite. 
 Do. of U.S.S. Miantonomoli. Do. of a boiler built by R. Napier & Co., Glasgow. Braces 
 with T-ends. Flexible braces. Braces of boilers of U.S.S. Monadnock. Eye-bars. Braces 
 with oblique branches. Attachment of braces to adjacent furnace-crowns by triangular links 
 and frames. Cheap forms of branch -braces. Bolts and split pins. Adjustable braces. 
 
SYNOPTICAL INDEX. 15 
 
 Section 5. Fitting and Adjusting Rod-braces. 
 
 Forging braces. Forming the ends of braces. Flat braces. Sagging of long braces. Setting-up 
 braces. 
 
 Section 6. Girder-stays, Gusset-stays, Stay-plates, Stay-domes, etc. 
 
 Various forms of girder-stays. Stay for front plate of back-connection of boilers of U.S.S. Mian- 
 tonomoh and class. Staying flat surfaces with angle and T-irons. Gusset-plates. Stay-domes. 
 
 Section 7. Experiments on the Shear ing -strength of Wrought-iron Bolts. 
 
 Section 8. Experiments made to Determine the proper Dimensions of Pins, Eyes, and Shanks of 
 
 Boiler-braces. 
 
 Section 9. Experiments on Screw Stay-bolts. 
 CHAPTER XI. 
 
 FLUES AXD TUBES. 
 
 Section 1. Flue-boilers. 
 
 Early flue-boilers. Return-flue boilers. Drop-flue boilers. Galloway tubes. Lamb and Sumner 
 boilers. 
 
 Section 2. Relative Advantages of Flues and Tubes for Marine Boilers. 
 Section 3. Various Types of Tubular Boilers. 
 
 Arrangement of tubes. Considerations governing the arrangement and location of tubes in marine 
 boilers. 
 
 Section 4. Dimensions and Spacing of Tubes. 
 
 Conditions governing the dimensions and spacing of tubes. Vertical water-tubes. Tubes of boiler 
 of U.S.S. Lackau-anna. Arrangement of vertical water-tubes in zigzag rows. Diameter and 
 length of fire-tubes. Spacing of horizontal fire-tubes. Stimer's differential tubular boiler. 
 Ferruling and swaging tubes. 
 
 Section 5. Iron, Steel, Brass, and Copper Tubes. 
 
 Relative advantages of boiler-tubes made of different materials. Lap-welded iron tubes and seam- 
 less drawn brass tubes. Experiments with boiler of U.S.S. Wyoming. Steel tubes. Draw- 
 ing tubes. Tapering tubes. 
 
 Talle XXXIII., sizes and weights of lap-welded iron boiler-tubes of standard gauge manufactured 
 by the National Tube-Works Company. 
 
 Table XX XIV., standard dimensions of lap-welded American charcoal-iron boiler-tubes manu- 
 factured by Morris, Tasker & Co. 
 
 Table XXXV., regular sizes and weights of seamless drawn brass and copper tubes, manufactured 
 by the American Tube- Works, Boston, Mass. 
 
 Table XXXVI., Stub's wire-gauge. 
 
16 SYNOPTICAL INDEX. 
 
 Section 6. Methods of Expanding Tubes. 
 
 Ordinary method and process of securing tubes in the tube-plates by expanding their ends. Ray- 
 mond's patent recessed tube-sheet. Prosser's expanding-tool. Dudgeon's tube-expander. 
 TweddelPs hydraulic tube-expander. Kiveting over the tube-ends. Selkirk's tube-beader. 
 Precautions to be taken in expanding tubes. 
 
 Section 7. Stay -tubes. 
 
 Stay- rods for tube-plates unnecessary. Stay-tubes of boilers for U.S.S. Nipsic. English practice. 
 Stay-tubes for boilers of steamer Atrato, built by James Watt & Co. Stay-tubes for boilers 
 of steamer Lord of the Isles. 
 
 Section 8. Devices for Rendering Boiler-tubes Removable. 
 
 Removing tubes secured by expanding their ends. French iron boiler-tubes. Removable tubes in 
 boilers built at the West Point Foundry, Cold Spring, N. Y. Pauksch's boiler-tubes. Re- 
 movable tubes used in French naval boilers ; viz., systeme Internet et Gouttes, systeme Toscer, 
 and systeme Langlois. Removable tubes used in some United States naval boilers. 
 
 Section 9. Experiments on the Holding-power of Boiler -tubes secured by various Methods. 
 Section 10. Sectional or Water-tube Boilers, Hanging-tubes, Double-tubes, etc. 
 
 General character of stationary water-tube boilers. Principal advantages claimed for these boilers. 
 Failure of attempts to use these boilers on board of vessels. Perkins tubulous boiler. 
 Howard marine water-tube boiler. Belleville water-tube boiler for despatch vessel L' Active. 
 Herreshoff coil-boiler of the steam-yacht Estelle. Performance of the Herreshoff coil-boiler. 
 Davey-Paxman boiler. Hanging tubes. 
 
 CHAPTER XII. 
 
 UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWEES, ETC. 
 
 Section 1. Smoke-connections and Uptake. 
 
 General form and arrangement of smoke-connections. Front smoke-connections and uptakes of 
 rectangular boilers. Front smoke-connections and uptakes of cylindrical boilers. Front-con- 
 nections and uptake of boilers of U.S.S. Nipsic. Front-connections and uptake of boilers of 
 U.S.S. Trenton. Importance of making uptakes and their doors air-tight. 
 
 Section 2. Forms and Dimensions of Chimneys. 
 
 Chimneys of circular and oval cross-section. Subdividing the cross-area of chimneys by partitions. 
 Dimensions of chimneys. Telescopic chimneys for war-vessels. Friction of gases in and 
 radiation of heat from chimneys. Damper. 
 
 Section 3. Fixed Chimneys. 
 
 Construction of fixed chimneys. Manner of securing them to uptake. Chimney-stays. Casing. 
 Water-tank in hatch. Chimney for U.S.S. Algoma and class. 
 
SYNOPTICAL INDEX. 17 
 
 Section 4. Hoisting Chimneys. 
 
 Hoisting chimney illustrated by Ledieu. Chimney of U.S.S. Plymouth. Hoisting-gear for chim- 
 neys. Chimney of U.S.S. Nipsic. Double-hoist and single-hoist chimneys compared. Chim- 
 ney of U.S.S. Quinnebaug and class. 
 
 Section 5. Artificial Draught : Blast-pipe, Steam-jets, Fan-blowers. 
 
 Artificial compared with natural draught. Efficiency of mechanism for producing artificial draught 
 Various methods of producing artificial draught. Action of fluid-on-fluid-impulse machines. 
 Blast-pipe of locomotives. Steam-jets. Jet-arrangement for chimney of U.S.S. Algoma 
 and class. Fan-blowers. Air-tight fire-room. Koerting's jet-apparatus. 
 
 Section 6. Experiments with Artificial Draught in Marine Boilers. 
 Experiments made at the U. S. Nary- Yard, New York, 1865-66. 
 
 Table XXXYIL, showing results of experiments made at the Navy- Yard, New York, in the years 
 1865-66, with various devices for producing artificial draught in marine boilers. 
 
 CHAPTER XIII. 
 
 STEAM-BOOM AND SUPERHEATERS. 
 
 Section 1. Capacity of Steam-room. 
 
 Influence of capacity of steam-room on the efficiency of the engine. Bourne on the capacity of 
 marine boilers. Capacity of steam-room in French naval boilers. Height of steam-room. 
 
 Section 2. Steam-drums. 
 
 Form and arrangement of steam-drams. Weakening effect of steam-domes built on cylindrical 
 shells, and methods of strengthening them. 
 
 Section 3. Superheaters. 
 
 Superheating surface of steam-drams and steam-pipes. Tubular superheaters in uptakes. Effi- 
 ciency and disadvantages of tubular superheaters in uptakes. Special superheating boilers of 
 United States naval vessels. Superheating arrangement of U.S.S. Eutaw. Its efficiency and 
 disadvantages. 
 
 CHAPTER XIV. 
 
 SETTING AXD ERECTION OF BOILERS. 
 
 Section 1. Setting of Boilers. 
 
 Boiler-kelsons. Setting boilers on wooden platforms. Cements used in setting boilers. Care to be 
 taken in setting boilers. Saddles for dry-bottom boilers of United States naval vessels. Sup- 
 ports for dry-bottom boilers' of U.S.S. Yantic. Saddles for cylindrical boilers. 
 
18 SYNOPTICAL INDEX. 
 
 Section 2. Securing Boilers. 
 
 Section 3. Erection of Sailers in the Vessel. 
 
 Preparing a bed for the boilers and drawing the lines to determine their position. Putting the 
 boilers on board. Verifying their correct position. Securing them in place. Making the 
 final connections. Putting the chimney in position. 
 
 CHAPTEK XV. 
 
 BOILER-MOUNTINGS AND ATTACHMENTS. 
 Section 1. Grate. 
 
 Usual form of grate. Dead-plate. Arrangement of grate-bars. Wrought-iron grate-bars. Shape 
 and dimensions of grate-bars. Spaces between grate-bars. Bearing-bars. 
 
 Section 2. Moving Orates. 
 Various forms of moving grates. The Murphy shaking-grate. The Martin or Ashcroft grate. 
 
 Section 3. Bridge-walls. 
 Forms of bridge-walls. Air-admission through bridge-wall. 
 
 Section 4. Furnace-doors and Door-frames. 
 
 Furnace door-openings. Furnace door-frames. Wrought-iron furnace-doors. Cast-iron furnace- 
 doors. The Martin or Ashcroft furnace-door. Pridaux's furnace-door. Air-admission 
 through furnace door. 
 
 Section 5. Connection-doors, Ashpit-doors, and Ashpans. 
 
 Section 6. Manhole and Handhole Plates. 
 
 Position, form, and dimensions of man and handholes. Strengthening-rings around manholes. 
 How to determine the size of strengthening-rings. Angle-iron strengthening-rings. Cast- 
 iron manhole-plates.- Manhole-plates for cylindrical shells. Wrought-iron manhole-cover by 
 Maudslay Sons and Field. Composition manhole-covers for United States naval boilers. 
 
 Section 7. Steam Stop-valves, Dry-pipes, and Steam-pipes. 
 
 General arrangement of steam stop-valves on boilers. Area of steam stop-valves and steam-pipes. 
 Steam stop-valves for boilers of U.S.S. Nipsic. Dry-pipes. Steam-pipes. Arrangement 
 of steam stop-valves and steam-pipes on boilers of U. S. S. Nipsic. 
 
 Section 8. Check-valves and Feed-pipes. 
 
 Arrangement and form of feed and check valves. Feed and check valves of the boilers of U.S.S. 
 Nipsic. Introduction of feed- water into the boiler. Feed-pipes. Arrangement of feed-pipes 
 for boilers of U.S.S. Nipsic. 
 
SYNOPTICAL INDEX. 19 
 
 Section 9. Slow-valves and Pipes. 
 
 Brine-pumps. Blow-valves of U. S. naval boilers. Blow-off cocks. Regulation of the Board of 
 Trade (English) regarding blow-off cocks and sea-connections. Bottom blow-valves. Surface 
 blow-valve. Arrangement of blow-valves and pipes of boilers of TJ.S.S. Nipsic. Material for 
 blow-pipes. 
 
 Section 10. Instruments and Attachments for Measuring and Indicating the Height and Density 
 of the Water, and the Pressure and Temperature of the Steam. 
 
 Water-gauges. Rules and Regulations of the Supervising Inspectors of Steam-vessels regarding 
 water-gauges. Location and forms of gauge-cocks. Water-gauge glasses. Water-gauges for 
 U. S. naval boilers. Percussion water-gauge. Floats. Fusible plugs. 
 
 Salinometer-pots as provided for TJ. S. naval boilers. Long's salinometer-pot Fithian's salino- 
 meter-pot. 
 
 Steam-gauges. Thermometers. 
 
 Section 11. Safety-valves. 
 
 Arrangement of safety-valves on boilers. Manner of applying load to safety-valves. Conditions to 
 be fulfilled by a safety-valve. Weight of steam discharged per second from an orifice. Effec- 
 tive opening of safety-valves. Rules for determining the area of safety-valves. Various forms 
 of safety-valves. Directions for the construction of safety-valves given by the Board of 
 Supervising Inspectors of Steam-vessels. Safety-valve for boilers of TJ.S.S. Nipsic. Ashcroft 
 spring safety-valve. Directions for the construction of spring safety-valves given by the Board 
 of Trade (England). Formulae for calculating the steam pressure at which a lever safety-valve 
 will open, the weight of the load required for a given length of lever, and the length of lever 
 for a given load. Practical method of loading a lever safety-valve. 
 
 Section 12. Miscellaneous Attachments of Boilers. 
 
 Escape-pipes. Justice's quieting-chambers. Shaw's spiral nozzles. Bleeding valves and pipes. 
 Reverse or vacuum valves. Auxiliary stop- valves and steam-pipes. Drain-cocks. Drain-pipes 
 in U.S.S. Nipsic. 
 
 Section 13. Covering for Boilers. 
 
 General considerations concerning the covering of boilers. Air-tight iron casings. Cow-hair felt. 
 Mastic compositions. Mineral wool. 
 
 Section 14. Feed-water Heaters and Filters. 
 
 Advantages of, and saving effected by heaters. Feed- water heater of U.S.S. Waiash. The Berry- 
 man heater. Selden's filter. 
 
 Section 15. Feed-pumps and Injectors. 
 
 Size of feed-pumps. Simplest form of the feed- water injector. Its action. Fixed-nozzle injector 
 and adjustable injector. Lifting power of the injector. Quantities of steam and water ad- 
 mitted. Formulae. Duty of an injector. Experiments made on Irwin's injector. Relative 
 efficiency of injectors and pumps. Use of injectors for feeding marine boilers. Seller's 
 self-adjusting injector. Koerting's universal lifting injector. 
 
20 SYNOPTICAL INDEX. 
 
 CHAPTER XVI. 
 TESTS, INSPECTIONS, AND TKIALS OF STEAM BOILERS. 
 
 Section 1. Testing Boilers. 
 
 Hydraulic test-pressure. Section 4418 of the U. S. Revised Statutes. Testing U. S. naval boilers. 
 Testing French boilers. Regulations of Board of Trade (English) regarding tests of boilers. 
 Anderson on testing boilers. Manner of applying the hydraulic test. Examination of boil- 
 ers during the test. Duration of tests. Cold and hot water tests. Test by expansion of 
 water. Testing new boilers by steam. 
 
 Section 2. Inspection of Boilers. 
 
 Necessity of careful inspection. Section 4418 of TJ. S. Revised Statutes. Regulations of the Board 
 of Trade (English) for the survey of marine boilers. Particular points to which the attention 
 of inspecting officers should be directed. Hammer test. Drilling plates. 
 
 Section 3. Trials of Boilers. 
 
 Objects of trials. Boiler experiments made under the direction of the Bureau of Steam Engineer- 
 ing, U. S. Navy Department. Rules to be observed in making boiler-experiments. Duration 
 of experiments. Reduction of results to a uniform standard Extract from the report on 
 the Murphy grate, describing manner of making experiment. Extract from report on the 
 Ashcroft furnace-doors and grate-bars, describing an approximate method for determining 
 the temperature of the gases in the uptake. Methods for eliminating errors due to presence of 
 unvaporized water in the steam. 
 
 CHAPTER XVII. 
 
 MANAGEMENT OF BOILERS. 
 
 Section 1. Getting up Steam. 
 
 Closing the boiler. Making the joint of manhole-plates. Examination of valves. Filling the 
 boiler. Preparing the boiler for starting fires. Charging the furnaces. Starting fires. Time 
 required for getting up steam. 
 
 Section 2. Firing. 
 
 Thickness of the bed of fuel. Frequency of firing. Back-draught. Cleaning fires. Burning slack 
 and coal-dust. Burning bituminous coal. 
 
 Section 3. Management of Boilers under Steam. 
 
 How to obtain a uniform supply of steam. Increasing the evaporation. Using a fraction of the 
 boiler-power. Banking fires. Diminishing the steam supply temporarily. Stopping the en- 
 gines. Opening the safety-valves. Care of water-gauges Foaming. Low water. Leaks. 
 Plugging leaky tubes. Putting a boiler out of use. Blowing-down. 
 
SYNOPTICAL INDEX. 21 
 
 Section 4. Foaming: Its Cause*, Effects, and Prevention. 
 
 Indications of foaming. Foaming caused by dirty water. Foaming prevented by the use of molten 
 tallow. Foaming due to insufficient and low steam-room. Foaming due to defective circula- 
 tion. Methods of improving the circulation. Increasing the steam-room by cutting out 
 tubes. Foaming caused by sudden reduction of pressure. The evil of foaming cured by the 
 use of dry-pipes, steam-drums, and superheaters. 
 
 Section 5. Constituents of Saline Matter in Sea-water. 
 
 Density of sea water, and proportion of salt in waters of different seas. Analysis of the water of 
 the English Channel at Brighton by Dr. Schweitzer. Analysis of the water of the Mediterra- 
 nean by Dr. Laurens. Carbonate of lime. Sulphate of lime. Chloride of calcium. Chloride 
 of sodium. Chloride of magnesium. Sulphate of magnesia. 
 
 , Section 6. Composition of Boiler-scale. 
 
 Section 7. Const? s Tfieory of the Formation of Deposits in Steam Boilers. 
 Section 8. Prevention of the Formation of Scale in Boilers. 
 
 Section 9. The Hydrometer. 
 
 Principle of its action. Description of the hydrometer used in the United States Navy. Method 
 of graduating the instrument. The scale varies with the temperature. The divisions of the 
 scale are not uniform. Increase of boiling-point. 
 
 Section 10. Influence of Temperature and Pressure on the Limit of the Saturation of Water in a 
 
 Boiler. 
 
 Formation of deposits of sulphate of lime at high temperatures. Effect of blowing-off under cer- 
 tain conditions. Table XXXVIII., containing the conditions accompanying the saturation 
 of sea-water with regard to sulphate of lime at various pressures and temperatures. 
 
 Section 11. Calculation of the Quantities of Water and Heat Lost by Blowing-off. 
 
 Section 12. Cleaning and Scaling Boilers. 
 
 Sweeping tubes. Tube-brushes. Tube-scrapers. Removing salt from, tubes. Cleaning uptakes, 
 back-connections, furnaces, and chimneys. Cleaning outside of boilers. Painting boilers. 
 Scaling boilers. Washing out the boiler. Drying the boiler. 
 
 Section 13. Repairing Boilers. 
 
 Locating a leak. Leaky seams and rivets. Leaky stay-bolts. Patching a boiler. Blisters, 
 Cracks. Repairing a furnace-crown which has come down. Leaky tubes. Removing a tube. 
 Bulged tube-plate. Iron cement for leaky seams of old boilers. Use of oatmeal for stop- 
 ping leaks. Filling the water-bottoms with cement. 
 
22 SYNOPTICAL INDEX. 
 
 Section 14. Preservation of Boilers. 
 
 Forming a protective coating of scale. Portland cement a substitute for scale. Coating the inte- 
 rior of boilers with paint or oil. Systems used in the English navy for the preservation of 
 boilers not in use. 
 
 Section 15. Extract from "Instructions for the Care and Preservation of the Steam Machinery 
 
 of United States Naval Vessels " (1879). 
 
 CHAPTER XVIII. 
 
 CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 
 
 Section 1. General Causes of the Deterioration of Boilers. 
 
 Principal causes of deterioration. Greater durability of stationary boilers compared with that of 
 marine boilers. Speedy deterioration of boilers of naval vessels. Manner in which deteriora- 
 tion takes place. Oxidation of iron by superheated steam. Corrosion of steel and wrought- 
 iron. Corrosion a! tacks surfaces in an irregular manner. Pitting. Grooving. 
 
 Section 2. Deterioration caused by Overheating and by the Corrosive Action of the Gases of Com- 
 bustion. 
 
 Oxidation in furnaces and combustion-chambers. Sulphurous fuel. Mechanical action of cinders 
 and fine coal in locomotives.- Superheating surfaces. Overheating of plates when bared of 
 water. Blisters. Spheroidal condition of the water. 
 
 Section 3. Strains produced by Sudden Variations and Great Differences of Temperature. 
 
 Intensity of stress produced by differences of temperature. Difficulty experienced with steel boiler- 
 plates. Cracks in laps. Long furnace-flues. Difference of temperature in upper and lower 
 half of furnace-flues and of cylindrical boiler-shells. Regulation of Board of Supervising In- 
 spectors of Steam-vessels regarding temperature of feed-water. Specimens of rivets from the 
 shell of a cylindrical flue-boiler. Strains in flat stayed surfaces of shell and fire-box of loco- 
 motive-boilers. 
 
 Section 4. Formation of certain Saponaceous Deposits in Land Boilers. 
 Extract from an article by Maurice Jourdain. Comments of L. Delaunay on the foregoing account. 
 
 Section 5. Corrosion of Steam Boilers by Sulphuric Acid present in the Soot. 
 
 Extract from an article in the Annales des Mines et des Fonts et Chaussees (1876). Extract from 
 an essay on " The Acid Products of the Combustion of Coal," by M. Vincotte. 
 
 Section 6. Corrosion Due to the Presence of Oxygen and Carbonic Acid in Water. 
 
 Most common cause of corrosion, or rusting of iron. Pure distilled sea-water. Experiments made 
 by Scheurer-Kestner and Meunier-Dollfus. Experiments on the oxidation of iron by Pro- 
 fessor P. Crau-Cal vert. Extracts from Third Report of the Admiralty Committee on Boilers. 
 
SYNOPriCAL INDEX. 23 
 
 Section 7. Corrosive Action of the Chloride of Magnesium. 
 
 Decomposition of chloride of magnesium by heat. Experiments on the decomposition of chloride 
 of magnesium by the Admiralty Committee on Boilers. Reactions between chloride of magne- 
 sium and iron. Sweating of boilers. 
 
 Section 8. Corrosive Action of Fatty Acids. 
 
 Decomposition of a fatty body in the presence of certain calcium and magnesium salts. Water- 
 sapouification. Professor A. W. Hoffmann on the corrosive influence of water and fatty acids 
 upon iron. Professor V. Wartha on the action of fatty acids on iron. Professor A. W. Hoff- 
 mann on the action of fatty substances on copper. Corrosive action of fatty acids in steam 
 boilers. Corrosion of steam-drums of U.S.S. Swatara. Means of preventing corrosion of 
 steam boilers by fatty acids. Difference in the action of tallow or vegetable oils and of mine- 
 ral oils upon copper. 
 
 Section 9. Corrosion of Boilers by Galvanic Action. 
 
 Galvanic action by contact of electro-heterogeneous metals. Galvanic action of copper in boilers. 
 Galvanic action of lead in boilers. Means of preventing corrosion of boilers by galvanic action. 
 Black magnetic oxide of iron. Professor Barff's method. Electro-negative character of the 
 black oxide of iron. 
 
 Section 10. The Use of Zinc for the Prevention of Corrosion and Incrustation of Boilers. 
 
 Extract from Third Report of the Admiralty Committee on Boilers relating to the use of zinc in 
 boilers. Extract from an article by Brossard de Corbiguy in the Annales des Mines (1877) re- 
 lating to the use of zinc for preventing the formation of adhesive scale. Manner of securing 
 the zinc in boilers. 
 
 Section 11. Action of Various Substances upon the Incrustative and Corrosive Ingredients of Feed- 
 waters. 
 
 Various processes of purification of feed-waters. Difficulties connected with the chemical treat- 
 ment of feed-water for marine boilers. Oil-cakes, potatoes, and other starchy matter. Glue, 
 hoofs, horns, tobacco-juice, Irish moss, peat, tow, hemp. Clay. Varnishes or lacquers. Pe- 
 troleum or paraffine oil. Glycerine. Milk of lime and caustic lime. Hetet's process. Carbon- 
 ate of soda, caustic soda, and potash. Hyposulphite and oxalate of soda. Proto-chloride of 
 tin, silicate, phosphate and arseniate of soda. Chloride of ammonium. Tannic acid. Tannate 
 of soda. Acetic acid. Rogers's process. Organic matter, sewage, bilge- water. 
 
 CHAPTER XIX. 
 
 BOILER-EXPLOSIOXS. 
 
 Section 1. Causes of Boiler-explosions. 
 
 Rupture due to weakness or to excessive pressure. Causes of weakness in a boiler. Causes produc- 
 ing an excess of pressure. Circumstances making safety-valves inoperative. Overheating of 
 
24 SYNOPTICAL INDEX. 
 
 plates. Violent shocks. Detonation of inflammable gases in flues. Explosions in consequence 
 of injuries to the shell of boilers and their attachments. Local damages. Rupture producing 
 an explosion. 
 
 Section 2. Various TJieories concerning Boiler-explosions. 
 
 Explosions ascribed to obscure causes. Electrical theory. Detonation of hydrogen and oxygen 
 gases. Decomposition of water by heat alone. Spheroidal state of water. Superheating 
 water deprived of air. Superheated steam. Overheated plates. Forces generated in the ex- 
 plosion of a boiler. R. H. Thurston's calculation of the work done in exploding a boiler. 
 
 Section 3. Phenomena of Boiler-explosions. 
 
 Conditions affecting the character of an explosion. Influence of position of fracture. Influence of 
 strength of material in vicinity of original fracture. Rupture of stays. Collapse of flues and 
 tubes. Collapse of furnace crowns. Ruptures in cylindrical and rectangular boilers. Explo- 
 sion on the Thunderer. Explosions of locomotive-boilers. Rupture at longitudinal seams of 
 cylindrical shells. Explosion on the steamer Westfield. 
 
 Section 4. Investigation of Boiler-explosions ( Wilson). 
 Section 5. Experimental Steam-boiler Explosions. 
 
 Experiments at Sandy Hook, N. J., in 1871. Explosion of a return-flue boiler. Explosion of an 
 experimental flat box. Report of Board of U. S. Naval Engineers on explosion of an old return- 
 tube boiler. 
 
LIST OF PLATES. 
 
 PLATE CHAPTER 
 
 I. Rodman's Testing-machine at Washington Navy- Yard V. 
 
 II. Experiments on Tensile Strength of Wrought-iron, conducted at the 
 
 Washington Navy- Yard by Chief Engineer Win. H. Shock, TT-S.N.. V. 
 IIL Fig. 1. Boiler for U. S. S. Daylight. fig. 2. Two Boilers for U. S. S. 
 
 Kansas. Fig. 3. Two Boilers for F. S. S. Mahaska VII. 
 
 IV. Boiler for Passenger Locomotive VII. 
 
 V. Boiler for Consolidation Freight Locomotive VII. 
 
 VI. Boiler for U. S. S. Lackawanna VII. 
 
 VII. Boiler for U. S. S. Lackawanna VII. 
 
 VEIL Six Boilers for TJ. S. S. Miantonomoh. and Class VII. 
 
 IX. Bracing of Boilers for U. S. S. Miantonomoh and Class VII. 
 
 X. Plates for Boiler of Steamer Lookout. Plates for Boilers of U. S. S. 
 
 Nipsic VII. 
 
 XL Boiler for Steamer Lookout VII. 
 
 XII. Six Boilers for U. S. S. Nipsic VII. 
 
 XIII. Details of Riveting of Boilers for U. S. S. Nipsic VII. 
 
 XIV. Fig. 1 and 2. Method of drilling Cylindrical Boilers, Navy-Yard, Nor- 
 
 folk, Va., 18T9 VIII. 
 
 XV. Two Boilers for S. S. Lord of the Isles IX. 
 
 XVI. Horizontal Boiler for 8' by 8' Engine, U. S. S. Cutters IX. 
 
 XVII. Four Boilers for TJ. S. S. Plymouth IX. 
 
 XVIII. Bracing of Boilers for TJ. S. S. Plymouth IX. 
 
 XIX. Details of Boilers for U. S. S. Plymouth. Fig. I. Connection Doors. 
 
 Fig. 2. Furnace and Ashpit Doors. Fig. 3 to 9. Details of Grate. IX. 
 
 25 
 
26 
 
 LIST OP PLATES. 
 
 PLATE CHAPTER 
 
 XX. Experiments on shearing Wronght-iron Bolts, conducted at the Wash- 
 ington Navy- Yard, 1868, by Chief Engineer Wm. H. Shock, U.S.N. X. 
 XXI. Fig. 1. Marine Flue-boiler. Fig. 2. Boiler for U. S. S. Shockokon. 
 
 Fig. 3. Boiler for U. S. S. Morse XL 
 
 XXII. Fig. 1 to 5. Methods of securing Boiler-tubes. Fig. 6. Tweddell's Hy- 
 draulic Tube-expander. Fig. 7. Selkirk's Tube-beader XI. 
 
 XXIII. Experiment with Brass Tubes, conducted at the Navy- Yard, Washington, 
 
 D. C., by Wm. H. Shock, Chief Engineer, U.S.N. , 1877 XL 
 
 XXIV. Experiment with Iron Tubes, conducted at the Navy- Yard, Washing- 
 
 ton, D. C., by Wm. H. Shock, Chief Engineer, U.S.N., 1877 XL 
 
 XXV. Attachment for Experiments with Tubes XL 
 
 XXVI. Fig. 1. The Perkins Tubular Boiler. Fig. 2. Marine Boiler by T. and 
 
 F. Howard XL 
 
 XXVII. Fig. 1. The Herreshoff Coil Boiler. Fig. 2. The Belleville Boiler XL 
 
 XXVIII. Fig. 1. Boiler for U. S. S. Cutter. Fig. 2. The Davey-Paxman Boiler. . XL 
 
 XXIX. Fig. 1 and 2. Uptake and Furnace-doors of Boilers for U. S. S. Nipsic. XIV. 
 
 XXX. Arrangement of Boilers in U. S. S. Nipsic XIV. 
 
 XXXI. Fig. 1 and 5. Manhole-plates. Fig. 2, 3, 4. Boiler-saddles XIV. 
 
 XXXII. Fig. 1, 2, 3. Steam Stop- valves and Feed valve of Boilers for U. S. S. 
 
 Nipsic XV. 
 
 XXXIII. Water-gauge of Boilers for U. S. S. Nipsic XV. 
 
 XXXIV. Fig. 1. Safety-valve of Boilers for U. S. S. Nipsic. Fig. 2. Safety-valve 
 
 approved by the Board of Supervising Inspectors of Steam-vessels. 
 
 Fig. 3. Ashcroft's Safety-valve XV. 
 
 XXXV. Fig. 1. Koerting's Jet Apparatus. Fig. 2. Sellers' Self-adjusting Injec- 
 tor. Fig. 3. Koerting's Universal Lifting Injector XV. 
 
 XXXVI. Specimens of Rivets and Rivet-heads, from Boilers of Copper-rolling 
 
 Mill, Navy- Yard, Washington, D. C., 1879 XVIII. 
 
LIST OF ILLUSTRATIONS INSERTED IN TEXT. 
 
 FIGURES. SUBJECT. PAGE 
 
 1 and 2. Forms of test-specimens 96 
 
 3 and 4. Forms of test-specimens 97 
 
 5. Experimental flue for tests made at Washington Navy- Yard 113 
 
 6. Experimental flue for tests made at Washington Navy- Yard 114 
 
 7 and ' 8. Diagrams of strains on oblique braces 118 
 
 9. Diagram of strains on circular arcs 119 
 
 10. Diagram of strains on circular arcs 120 
 
 11 to 14. Diagrams of strains on circular arcs 121 
 
 10a to 16a. Forge-tests of angle and T-irons for boilers of English naval vessels 150 
 
 15. Front-head of boiler for U. S. S. Nipsic 158 
 
 16. Cylindrical flue with inner and outer courses 159 
 
 17 to 19. Diagrams for laying-off conical tubes 160 
 
 20 to 22. Diagrams for laying-off shell of cylindrical boiler 161 
 
 23 and 24. Diagrams for laying-off cylindrical shell of steam-dome 162 
 
 25 and 26. Arrangement of bending-rolls 164 
 
 27 to 30. Forms of punches 166 
 
 31. Shape of rivet in punched hole 167 
 
 32. Drifted rivet-hole 171 
 
 33. Steam riveting-machine of Providence Steam-engine Company, Providence, 
 
 E, 1 172 
 
 34. Form and dimensions of f-inch rivet 176 
 
 35 to 37. Forms of rivet-points 176 
 
 38. Countersunk rivet-hole 177 
 
 39 to 44. Forms of riveted lap and butt joints 178 
 
 45. Arrangement for determining friction in riveted joints 179 
 
 46. Distortion of lap-joint 179 
 
 Z! 
 
28 LIST OF ILLUSTRATIONS INSERTED IN TEXT. 
 
 FIGURES. SUBJECT. PAGE 
 
 47 to 50. Distribution of stress in riveted joints 181 
 
 51. Strained zones around punched holes 182 
 
 52 and 53. Strains in multiple-riveted joints 183 
 
 54 to 56. Modes of fracture in riveted joints 186 
 
 57. Excessive calking of lap-joints 188 
 
 58 and 59. Fractures of zigzag-riveted joints 189 
 
 60. Distorting strain in countersunk-riveted joint 191 
 
 61 to 63. Diagrams of fracture in multiple-riveted joints 193 
 
 64. Diagonal lap-joint 198 
 
 65. Oval rivets 199 
 
 66. Covering-plate for lap-joint 199 
 
 67. Forked covering-plate for lap-joint 199 
 
 68 and 69. Plates with thickened edges 199 
 
 70 to 72. Diagrams of experimental butt-joints 204 
 
 73. Ordinary calking-tools -.... 205 
 
 74. Connery's calking-tool 205 
 
 75. Calking-tool for butt-joints 206 
 
 76. Calked butt-joint 206 
 
 77. Calked lap-joint and rivet 206 
 
 78. Bertram's method of welding boiler-plates 208 
 
 79 to 82. Forms of welded joints 208 
 
 83 and 84. Methods of welding flues 208 
 
 85. Method of welding cylindrical boiler-shell 209 
 
 86 to 88. Forms of welded joints 210 
 
 89. Method of welding front plates of boilers 210 
 
 90 and 91. Welded angle-iron rings 210 
 
 92. C. E. Emery's connected-arc marine boiler 214 
 
 93 and 94. Circulation of water with flat and arched furnace-crowns 223 
 
 95. Furnace of marine boiler built by J. & W. Dudgeon, England 223 
 
 96. Furnace of boiler of U. S. tug Glance 224 
 
 97. Furnace of marine boiler built by Laird & Son, England 225 
 
 98. Mud-drum in dry-bottom boiler 226 
 
 99. Adamson joint for furnace-flues 227 
 
 100. T-iron ring for furnace-flues 227 
 
 101. Bowling-hoop for furnace-flues 227 
 
 102. Angle-iron strengthening-hoop for flues 228 
 
 103. Attachment of braces in water-legs of boilers 236 
 
LIST OF ILLUSTRATIONS INSERTED IN TEXT. 
 
 FIGURE* SUBJECT. PAGE 
 
 104 and 105. Forms of branch-braces , 236 
 
 106. Bracing rectangular boilers of French naval vessels. 237 
 
 107. Bent socket-bolt 244 
 
 108 to 111. Stays for narrow water-spaces 245 
 
 112. Diagram of forked stay . 246 
 
 113. Triangular stay for water-bottom 246 
 
 114. Brace for ends of cylindrical boiler built by R Napier & Co., Glasgow 247 
 
 115. Branch-brace used in French naval boilers. 248 
 
 116 and 117. Triangular stays for attachment of braces to furnace-crowns 248 
 
 118. Simple form of branch-brace 249 
 
 119. Split pins 249 
 
 120. Branch-brace secured by nut 249 
 
 121. Forked end of rod-brace 250 
 
 122. Flat brace 250 
 
 123. Girder-stay 251 
 
 124 and 125. Stay-plates for top of back-connection 251 
 
 126 and 127. Diagrams showing proportions of eve-bars 254 
 
 128. Raymond's recessed tube-sheet 272 
 
 129. Expanded tube-end 273 
 
 130. Expanding-roller 273 
 
 131. Beading tube-end with boot- tool 273 
 
 132. Beading tube-end with round-headed hammer 273 
 
 133. Grate-bar 320 
 
 134. Murphy shaking-grate 322 
 
 135. Martin furnace-door and grate 326 
 
 136. Manhole-plate and frame for cylindrical shell 330 
 
 137. Wrought-iron manhole-plate for boiler built by Mandslay, Sons & Field, Eng- 
 
 land 330 
 
 138. Water-gauge-cock 337 
 
 139. Diagram of lever safety-valve 346 
 
 140. Practical method of weighting lever safety-valve 347 
 
 141. Top of steam-escape pipe 348 
 
 142 and 143. Justice's quieting-chambers 349 
 
 144 and 145. Shaw's noise-quieting nozzles 350 
 
 146. Fixed-nozzle injector 355 
 
 147. Hydrometer 394 
 
 147a. Manner of securing zinc slabs in boilers 430 
 
STEAM BOILEBS: 
 
 THEIR 
 
 DESIGN, CONSTRUCTION, AND MANAGEMENT. 
 
 CHAPTER I. 
 
 INTRODUCTORY REMARKS. 
 
 Ix the early days of the steam-engine, when the working pressure was low, boilers 
 were made of various materials, many of which were soon discarded ; cast-iron, and, 
 in particular instances, even granite and wood, were used. 
 
 Later, copper became a favorite material for the construction of boilers, and it re- 
 mained in use in the United States Navy up to 1858. Its discontinuance was caused by 
 its greater first cost, greater weight in the vessel, and greater difficulty of keeping its 
 seams tight, than in the case of plate-iron, but principally the former. It went out of 
 use everywhere long before the employment of steam-pressures too high for its tensile 
 strength ; and, for a long period before its total disappearance, it was used in national 
 navies only, on account of its durability, as it exceeded iron threefold in that respect. 
 The greater cheapness of plate-iron superseded it at once in merchant-steamers as soon 
 as the manufacture of that material was sufficiently perfected. A serious objection to 
 copper for steamers, independently of its greater cost, was that boilers constructed of 
 it had a much greater weight than plate-iron boilers of the same dimensions and 
 strength, because of its greater specific gravity, and of its greater cross-sections of 
 metal required by its less tensile strength to support equal tensile strains. 
 
 The difficxilty in keeping the joints of copper boilers water-tight was an important 
 practical defect, and arose from the fact that the oxide resulting from copper corrosion 
 has scarcely any adhesion to the metal, so that a leak once commenced continually 
 increased by the washing away of the material around it ; whereas the oxide resulting 
 from the corrosion of iron boilers has so strong an adhesion to its metal, and is so 
 
 31 
 
32 STEAM BOILERS. CHAP. I. 
 
 bulky in proportion to its metallic constituent, that small leaks are soon stopped by 
 the very corrosion they produce. 
 
 The introduction of the compound engine, necessitating high boiler- pressures for the 
 development of its best economy, at once doubled and trebled the steam-pressures 
 previously employed with simple engines supplied with steam from boilers having 
 rectangular shells, compelling thereby the abandonment of that form and the substi- 
 tution of cylindrical shells. With steam-pressures still increasing, and with the 
 necessity continually pressing for lighter weights of machinery with increased powers 
 for steamers, the joints of cylindrical shells have been changed from single to double 
 and to treble riveted ; and the tendency now is to" the substitution of steel as a material 
 in place of plate-iron, of the welded joint with covering-plates in place of riveted joints, 
 and of boilers formed of tubes of small diameters variously arranged. 
 
 The essential parts of a steam-boiler are : 
 
 1st. The ashpit or chamber lying beneath the grate. 
 
 2d. The grate lying between and separating the ashpit from the furnace. 
 
 3d. The furnace or chamber lying immediately above the grate. 
 
 4th. The flues or tubes, together with their connecting chambers, extending from 
 the furnace to the chimney. 
 
 5th. The chimney. 
 
 6th. The water-room enclosing the furnace, tubes, flues, and connecting chambers. 
 
 7th. The steam-room lying above the water-room. 
 
 From the ashpit air is supplied through the grate to the fuel lying upon it, the ash 
 from this fuel falling into the pit. The grate supports the fuel, which is evenly spread 
 over it in such a manner that the air passing through its interstices may be uniformly 
 distributed. The furnace contains the fuel whose constituents are burnt by combina- 
 tion with the oxygen of the air entering through the grate. The portion of the furnace 
 above the fuel serves in part as a combustion-chamber where the uncombined gases of 
 the air and fuel may be brought by mixture into contact while still at a sufficiently 
 high temperature for combustion. 
 
 The flues, the tubes, and their connecting chambers, together with the furnace, con- 
 stitute the heating surface by means of which the heat in the gaseous products of 
 combustion is transferred to the water enveloping those surfaces. The chimney de- 
 livers into the atmosphere the gases of combustion after their heat has been extracted 
 to the desired temperature by the heating surfaces. It also causes the "draught" by 
 means of which a constant supply of new air is furnished to the fuel, thereby render- 
 ing the combustion continuous. The water-room contains the water to be vaporized, 
 
CHAP. I. INTRODUCTORY REMARKS. 33 
 
 and the steam-room contains the steam after it has been evaporated. The first may be 
 only of sufficient capacity to form barely an envelope to the heating surfaces ; but the 
 latter must be at least large enough to prevent the pressure of the steam in it from 
 sensibly varying with the intermittent draughts of the steam-cylinder. 
 
 It is quite evident that an ingenious engineer could form of the, elementary parts of 
 a boiler just enumerated an almost infinite number of combinations ; those which have 
 actually been devised and executed are so numerous that a large space would be 
 required to describe them, and their description for the most part would be as useless 
 as tedious. As they can be found extensively illustrated in Patent-Office reports and 
 in existing engineering literature, the present essay will be restricted to a consideration 
 of only such as have been found by long experience to meet the requirements of prac- 
 tice, and chiefly of those best adapted for use on board of war and ocean merchant 
 steamers. 
 
 The general conditions which determine the peculiar features of marine boilers are 
 the following : the weight of, and the space occupied by, the boilers in a vessel are 
 necessarily limited ; economy in fuel is required not only on account of its cost r but on 
 account of its weight and the space occupied by it ; the height of the chimney is limited, 
 and the location of the boilers in the hold of the vessel interferes with the draught ; the 
 great liability of marine boilers to the evil effects of scale and corrosion makes it impe- 
 rative that their interior should be easily accessible for cleaning and repairs ; the roll- 
 ing and pitching of the vessel strains the boilers and keeps their water in constant 
 motion ; special precautions have to be taken to guard against fire ; the boilers have to 
 remain in action often for weeks, day and night, without interruption. 
 
 In a man-of-war the boilers must be placed as low as possible, in order to protect 
 them against the chances of penetration by shot ; the duty required of them is very 
 unequal : they may lie idle for years ; at other times steam may have to be kept up in 
 them for months continuously ; they will be required to develop their full power only 
 on exceptional occasions ; to be prepared for all emergencies, they must be able to gen- 
 erate steam rapidly, and preserve their efficiency under greatly varying conditions. 
 
CHAPTER H. 
 
 COMBUSTION. 
 
 1. Elementary Constituents of Fuels. Chemical combination is always accom- 
 panied by the development of heat, and when the latter is sufficiently intense to pro- 
 duce light the combination is called combustion, and the combining substances are 
 called combustibles. 
 
 The chief combustible constituents of fuel are carbon and hydrogen, and their 
 chemical combination with the oxygen of the atmosphere is the source of the heat 
 used in steam boilers. Most fuel contains also sulphur and nitrogen, whose combina- 
 tion with oxygen likewise produces heat, but the amount is too insignificant for consid- 
 eration in a treatise like this. 
 
 TABLE I. 
 
 Name. 
 
 Symbol. 
 
 Proportions of 
 elements by 
 weight. 
 
 Chemical 
 equiva- 
 lent by 
 weight. 
 
 Proportions of 
 elements by 
 volume. 
 
 Chemical Equiva- 
 lent by volume. 
 
 Specific 
 Heat 
 at 
 constant 
 pressure. 
 
 Weight of a 
 cubic foot in 
 pounds at 32 
 and atmos- 
 pheric pressure. 
 
 Volume in cubic 
 feet of one 
 pound at 32 
 and atmos- 
 pheric pressure. 
 
 
 o 
 
 
 16 
 
 
 I 
 
 0.218 
 
 O.O8O7 
 
 11.204 
 
 
 N 
 
 
 
 
 I 
 
 O.24.4 
 
 0.0784 
 
 12.753 
 
 
 H 
 
 
 i 
 
 
 I 
 
 7.4OE; 
 
 o. 00=56 
 
 178.83 
 
 Carbon 
 
 c 
 
 
 
 
 
 
 
 
 Sulphur ... . . 
 
 s 
 
 
 
 
 
 
 
 
 Air 
 
 
 N?7-r-O2t 
 
 o* 
 
 N79-J-O2I 
 
 
 0.2^8 
 
 0.0807 
 
 12.387 
 
 Water 
 
 H O 
 
 ** 1 1 1 *"<,} 
 
 H2 -)-Oi6 
 
 18 
 
 
 
 1*000 
 
 62.421; * 
 
 0.016* 
 
 
 
 
 
 H2 + O 
 
 2 
 
 to. j.8o 
 
 0.0502 
 
 IQ.QI'? 
 
 Carbonic oxide 
 
 CO 
 
 Ci2 +Oi6 
 
 28 
 
 C + O 
 
 2 
 
 0.241; 
 
 0.0781 
 
 12.820 
 
 Carbonic acid 
 
 CO 
 
 Cl2 4-Ol2 
 
 AA 
 
 C +O2 
 
 2 
 
 O.2I7 
 
 o. 12^4 
 
 8.IOI 
 
 Olefiant gas 
 
 CH 
 
 CI2+H2 
 
 1A 
 
 C -f-H2 
 
 2 
 
 o ^60 
 
 O.O7Q 1 ? 
 
 12.58 
 
 Marsh gas 
 
 CH 
 
 Ci2+H4 
 
 16 
 
 C -4- HA 
 
 2 
 
 O.^O"? 
 
 0.0447 
 
 22.388 
 
 Sulphuretted hydrogen 
 
 SH 
 
 S-22 -1- Ha 
 
 
 
 2 
 
 
 
 
 Sulphurous acid 
 
 SO 
 
 S?2 -\-Ql2 
 
 34 
 64 
 
 
 2 
 
 0.1^4. 
 
 0.1814 
 
 1 5- 1 5i4 
 
 Bisulphuret of carbon . 
 
 S C 
 
 864 +Cl2 
 
 76 
 
 
 2 
 
 o.T=;8 
 
 o-2i-?7 
 
 4.679 
 
 Ammonia 
 
 NH 
 
 NIA 4- H? 
 
 T 7 
 
 
 2 
 
 
 
 
 
 
 
 *7 
 
 
 
 
 
 
 * The weight and volume of water are given for the temperature at which it attains its greatest density viz., 39. I Fahr. 
 t For aqueous vapor in the gaseous state, not saturated vapor. 
 
 34 
 
SEC. 3. COMBUSTION. 35 
 
 " Substances combine chemically in certain proportions only. To each, substance 
 known in chemistry a certain number can be assigned, called its ' cTiemical equivalent,'' 
 and having these properties : I. That the proportions by weight in which substances 
 combine chemically can all be expressed by their chemical equivalents, or by simple 
 multiples of their chemical equivalents ; II. That the chemical equivalent of a com- 
 pound is the sum of the chemical equivalents of its constituents." (Rankine.) 
 
 The elementary substances entering into the composition of the atmospheric air, and 
 of the combustible portion of different fuels, are : oxygen, nitrogen, hydrogen, carbon, 
 and sulphur. The preceding table contains the principal chemical and physical proper- 
 ties of these substances and of the most important compounds formed by them, as 
 found in different fuels and in the products resulting from their combustion. 
 
 The chemical equivalents are given in round numbers, omitting fractions too small 
 to be of consequence in calculations connected with the subject of the present treatise. 
 
 It must be borne in mind that atmospheric air is not a chemical compound, but a 
 mechanical mixture of nitrogen and oxygen. 
 
 2. Temperature of Ignition. At a temperature of about 750 Fahr. solid bodies 
 become luminous, emitting a dull red light ; the intensity of the light increases with 
 the temperature till a dazzling white heat is attained. Gases become luminous only 
 during the process of combustion, forming what is called flame. The flame of gases 
 has little brilliancy, unless intensified by the presence of small particles of incandescent 
 solid matter. Combustion cannot be maintained at a temperature lower than 800 
 degrees. 
 
 3. Combustion of the Constituents of Fuels. The action, during combustion 
 in the furnace, of the principal constituents of the various kinds of fuel commonly used 
 is described by Rankine as follows : (I.) Fixed or free carbon, which is left in the form 
 of charcoal or coke after the volatile ingredients of the fuel have passed off by distilla- 
 tion. It burns without flame ; when raised to a state of incandescence, each equivalent 
 by weight < >f carbon combines with two equivalents of oxygen, forming an invisible gas 
 called carbonic acid. If the carbonic acid gas remains in contact with incandescent 
 carbon it dissolves an additional equivalent of carbon, forming with it a compound 
 called carbon ic oxi'le. which contains only one equivalent of oxygen for each equivalent 
 of carbon. When, at a sufficiently high temperature, the carbonic oxide comes in con- 
 tact with oxygen, it absorbs a sufficient quantity to form carbonic acid, burning during 
 this process with a blue flame. 
 
 (II.) Hydrocarbons, such as olefiant gas, pitch, tar, naphtha, etc., all of which must 
 pass into the gaseous state before being burned. 
 
36 STEAM BOILERS. CHAP. U. 
 
 If mixed, on their first issuing from amongst the burning carbon, with a sufficient 
 quantity of air, these inflammable gases are completely burned with a transparent blue 
 flame, producing carbonic acid and steam. When raised to a red heat, or thereabouts, 
 before being mixed with a sufficient quantity of air for perfect combustion, they disen- 
 gage carbon in fine powder, and pass to the condition partly of marsh gas, and partly 
 of free hydrogen ; and the higher the temperature the greater is the proportion of solid 
 carbon thus disengaged. 
 
 If the disengaged carbon is cooled below the temperature of ignition before coming 
 in contact with oxygen, it constitutes, while floating in the gas, smoke, and, when 
 deposited, soot. 
 
 But if the disengaged carbon is maintained at the temperature of ignition, and sup- 
 plied with oxygen sufficient for its combustion, it burns while floating in the inflam- 
 mable gas, and forms red, yellow, or white flame. 
 
 (III.) Oxygen and hydrogen, either actually forming water, or existing in combina- 
 tion with the other constituents in the proportions which form water. The presence of 
 water, or the constituents forming it, in fuel promotes the formation of smoke or of 
 the carbonaceous flame, as the case may be ; probably by mechanically sweeping along 
 fine particles of carbon. 
 
 The absorption of the heat required for the vaporization of the water contained in 
 the fuel, or produced by its combustion, may reduce the temperature of the products 
 of combustion below the ignition-point of carbon, but not below the dissociation-point 
 of the hydrocarbons of great molecular condensation. 
 
 (IV.) Nitrogen, either free or in combination with other constituents. This sub- 
 stance is nearly inert, and the bulk of it passes off uncombined. 
 
 (V.) The sulphur of the sulphurets of iron and of copper contained in many coals 
 forms sulphuric acid when hydrated, which corrodes the metal of the boiler. 
 
 (VI.) Other mineral compounds of various kinds form the ash left after the com- 
 plete combustion of the fuel ; and also the clinker, or glassy material produced by 
 fusion of the ash. 
 
 4. Total Heat of Combustion. Carbon and hydrogen are the only constituents 
 of fuel, the combustion of which is of practical value for the generation of heat in the 
 steam boiler. The following table contains the quantities of heat developed by the 
 combustion of one pound of these elements and of their compounds, as determined by 
 the careful experiments of Favre and Silbermann, together with the weight of oxygen 
 necessary for their combustion, and likewise the weight of air required to furnish this 
 amount of oxygen. 
 
Ssc.1 
 
 COMBUSTION. 
 
 37 
 
 The thermal unit, commonly employed by scientists who use British measures, is 
 the quantity of heat required to raise the temperature of one pound avoirdupois of pure 
 water of 39. 1 one degree on Fahrenheit's scale, the barometer standing at 29.922 
 inches of mercury at 32 Fahrenheit, at the level of the sea, in latitude 45 degrees. 
 
 TABLE II. 
 
 Combustible. 
 
 Pounds of oxy- 
 gen per Ib. of 
 combustible. 
 
 Pounds of air 
 per Ib. of com- 
 bustible. 
 
 Total heat in 
 thermal 
 units. 
 
 Pounds of water 
 that can be 
 evaporated un- 
 der atmospheric 
 pressure from 
 
 212. . 
 
 Hydrogen 
 
 8. 
 
 *4 8 
 
 62.0^2 
 
 64 2 
 
 Carbon (combustion producing carbonic acid). . . . 
 Carbon (combustion producing carbonic oxide) . . . 
 Carbonic oxide 
 
 2.666 
 1-333 
 
 O.C7I 
 
 n.6 
 
 5-8 
 
 2 48* 
 
 14,5 
 4,400 
 
 J..32Q 
 
 15-0 
 
 4-55 
 4.48 
 
 Marsh gas 
 
 4. 
 
 17 4 
 
 2-?,88* 
 
 24 1 
 
 Olefiant gas 
 
 34? 
 
 14 O 
 
 21.144 
 
 22.1 
 
 
 
 
 
 
 The production of 3.66 pounds of carbonic acid gas by the complete combustion of 
 one pound of carbon is accompanied by a development of heat more than three times 
 greater than the amount of heat generated by the incomplete combustion of the same 
 weight of carbon, prodiicing 2.33 pounds of carbonic oxide. When, however, these 
 2.33 pounds of carbonic oxide combine with a sufficient quantity of oxygen to form 3.66 
 pounds of carbonic acid, 10,100 additional units of heat are generated, so that the total 
 amount of heat produced by this twofold process is equal to the heat produced by the 
 c< inversion of one pound of carbon into carbonic acid. 
 
 The total heat of combustion of a compound of hydrogen and carbon (with the 
 exception of marsh gas) is generally assumed to be nearly the sum of the quantities of 
 heat which the hydrogen and carbon contained in it would produce separately by their 
 combustion. 
 
 In many such compounds, however, the quantity of heat is greater than this sum, 
 depending on the degree of condensation of the constituents in the molecule. In the 
 case of perfectly dry wood, in which oxygen exists in addition to carbon and hydrogen, 
 it is only the excess of the latter over the oxygen constituent in the proportion neces- 
 sary to form water which produces a heating effect ; but this fact cannot be extended 
 inferentially to other compounds of carbon, hydrogen, and oxygen such, for example, 
 as coal for in their cases their heating powers have been experimentally shown to even 
 exceed that of the sum of their full constituents. The heating power of any hydro- 
 
38 STEAM BOILERS. CHAP. II. 
 
 carbon can only be known by direct experiment upon it, bnt a sufficiently close 
 approximation for practice can be made by employing the old law of Dulong based 
 on experiments with wood. This law is that, when hydrogen and oxygen exist in a 
 'compound, only the surplus of hydrogen, above the amount required for combination 
 with the oxygen present in the fuel, will be effective in raising the total heat of com- 
 bustion. 
 
 In computing the total heat of combustion of a compound it is convenient to substi- 
 tute for the hydrogen a quantity of carbon which would give the same quantity of 
 
 (62 032 \ 
 ' j=4.28. 
 
 On these principles are based the following general formulae for computing the theo- 
 retical calorific power of any compound of which the principal constituents are carbon, 
 hydrogen, and oxygen. 
 
 C, ff, and represent the fractions of one pound of the compound, which are, 
 respectively, carbon, Tiydrogen, and oxygen ; V is the total heat of combustion of the 
 compound, expressed in British thermal units ; and E denotes the theoretical vaporific 
 power of one pound of the compound, expressed in pounds of water vaporized from 
 212 under atmospheric pressure ; then 
 
 U= 14,500 <7+4.28j7- - .[I.] 
 
 The actual calorific power of coals cannot be determined with exactness by the 
 above method, for the following reasons : 
 
 1st. Different forms of pure carbon differ considerably in calorific power. According 
 to Favre and Silbermann, wood charcoal has the highest calorific power, equal to 14,544 
 units, and diamond has the lowest, equal to 13,986 units. 
 
 2d. The quantity of heat which becomes latent in the decomposition of the various 
 chemical compounds entering into the composition of coals before combustion takes 
 place, varies with the nature of these compounds. The recent experiments of Scheurer- 
 Kestner and C. Meunier developed in many instances great discrepancies between the 
 actual and the calculated calorific powers of coals. In the case of two coals, the one 
 from Ronchamp and the other from Creusot, which contained almost precisely the 
 same proportions of carbon, hydrogen, and oxygen, the calorific powers, instead of 
 being, in accordance with calculation, identical, were 16,411 and 17,320 respectively. 
 The difference between the real and calculated calorific powers amounted in some 
 instances to as much as 15 per cent. 
 
SEC. 5, COMBUSTION. 39 
 
 The quantity of heat that may be generated by the complete combustion of a fuel 
 is not the measure of its vaporific power in a steam boiler ; the latter depends in a great 
 measure on the temperature of combustion, and the completeness of the combustion of 
 the fuel in the boiler, and can be determined only by experiment under conditions of 
 actual practice (see Tables V., VI., and VI. a). The utilization of the large quantity 
 of heat generated by the combustion of hydrogen presents great practical difficulties, 
 and Johnson's experiments on the vaporific power of American and English coals, made 
 in 1842-43, established the fact that, when the weight of fixed carbon is less than four 
 times the weight of the volatile combustible matter in a coal, its vaporific power in a 
 steam boiler decreases perceptibly. 
 
 5. Fuel as a Source of Power. By multiplying the units of heat representing 
 the calorific power of a fuel by "Joule's equivalent" we find the amount of energy 
 stored up in the fuel and set free by its combustion. 
 
 Dr. Joule, of Manchester, found, by carefully-conducted experiments, the result of 
 which he finally communicated to the Royal Society in 1849, that each British unit of 
 heat was produced by the expenditure of 772.69 foot-pounds of work. Other observers 
 who have since tried to determine the mechanical equivalent of heat by various meth- 
 ods have obtained results differing more or less from that of Joule's experiments. The 
 mean of sixteen of the most accurate of these determinations gives 786 as the value of 
 the mechanical equivalent of heat. At a meeting of the Royal Society, held January 
 24, 1878, Joule read a paper in which he gives an account of the experiments he had 
 recently made, with a view to increase the accuracy of the results given in his former 
 paper. The result he has now arrived at from the thermal effects of the friction of 
 water is that, taking the unit of heat as that wjiich can raise a pound of water, weighed 
 in vacuo, from 60 to 61 Fahr. of the mercurial thermometer, its mechanical equivalent 
 reduced to the sea-level, at the latitude of Greenwich, is 772.55 foot-pounds. 
 
 For calculations the value of "Joule's equivalent'' 1 is generally taken as 772, in 
 round numbers. 
 
 A recent and careful determination of the mechanical equivalent of heat was made 
 by a Commission of the French Academy, which found that 789J foot-pounds of work 
 were equivalent to the heat required to raise the temperature of one pound of water at 
 32 to 33 under the standard atmospheric pressure. Joule's determination, however, 
 is still generally employed in scientific works. 
 
 Taking 14.000 units of heat as representing the average calorific power of good coal, 
 we find that the energy developed by the combustion of one pound of such coal is equal 
 to 14,000 x 772, or 10.808,000 foot-pounds. 
 
40 STEAM BOILERS. CHAP. H. 
 
 6. Air required for Combustion. To ensure the perfect combustion of a fuel 
 it is necessary 
 
 First, to maintain the combustible matter at such a temperature as is required for 
 its chemical combination with the oxygen of the air. 
 
 Secondly, the quantity of air admitted to the furnace must contain a sufficient 
 amount of oxygen. If C, H, and represent the same quantities as in equation [I.], 
 and A represents the number of pounds of air containing the quantity of oxygen 
 required for the complete combustion of one pound of combustible, then 
 
 A = 11.6 (7+34.8 ill- ~Y[IIL] 
 
 For all practical purposes it is sufficiently accurate to assume the quantity of air chemi- 
 cally required for every kind of coal as 12 pounds per pound of coal. 
 
 Thirdly, the air must be thoroughly mixed and brought into actual contact with 
 each particle of the incandescent solid and gaseous matter. In the furnace of a steam 
 boiler this is effected in two ways, viz. : First, by admitting the air partly below the 
 fuel, through the evenly-distributed interstices of the grate, and partly, in the shape 
 of numerous small jets, above the solid fuel among the evolved gases in the furnace ; 
 and, secondly, by admitting a quantity of air in excess of the theoretical quantity 
 required by formula [III]. The amount of this excess varies with different coals and 
 with the manner of introducing the air ; but numeroiis experiments have proved that 
 in ordinary boiler-furnaces, where the draught is produced by means of a chimney, 
 the total weight of air admitted should be, on an average, twice the amount theoreti- 
 cally required for the oxidation of the fuel, or 24 pounds per pound of coal burned ; 
 when, however, the draught is produced by artificial means, either by a steam- jet or by 
 a fan-blower, one and a half times the theoretical amoiint, or 18 pounds per pound of 
 coal, appear to be sufficient. 
 
 While an admission of air in excess of the amount actually required for the com- 
 plete oxidation of the fuel always entails a loss of heat, since the temperature of the 
 uncombined air has to be raised at the expense of the heat of combustion, the loss of 
 heat in consequence of incomplete combustion is generally far greater. 
 
 7. Temperature of Combustion. The elevation of the temperature of the pro- 
 ducts of combustion, above the temperature at which the air and the fuel are supplied 
 to the furnace, which would be obtained if the combustion was complete, and the whole 
 heat of combustion was spent in raising the temperature of the products of combustion, 
 is called the theoretical calorific intensity of the fuel, and is computed by dividing the 
 total heat of combustion of one pound of fuel by the sum of the products of the weight 
 
SEC. 7. 
 
 COMBUSTION. 
 
 and specific heat of the several products of combustion, under constant pressure. 
 When steam is present among the products of combustion, resulting either from the 
 combustion of hydrogen or from water mechanically combined with the fuel, the pro- 
 duct of its weight and latent heat must be deducted from the total heat of combustion 
 in the first place. The specific heat of varioiis products of combustion has been given 
 in Table I.; that of ashes is probably about 0.200. 
 
 The latent heat of steam at ordinary atmospheric pressure is 966. 1. 
 
 TABLE III. 
 
 CONTAINING THE THEORETICAL TEMPERATURES PRODUCED BY THE PERFECT COMBUSTION OF 
 
 VARIOUS SUBSTANCES. 
 
 Name of substance. 
 
 Pure oxygen supplied 
 sufficient for 
 complete combustion. 
 
 Atmospheric air supplied. 
 
 Sufficient for 
 complete oxidation. 
 
 One and a half times 
 the quantity necessary 
 for complete oxidation . 
 
 Twice the quantity 
 necessary for 
 complete oxidation. 
 
 Hydrogen 
 
 I2,346 
 
 18,257 
 12,695 
 
 15,475 
 
 4,9" 
 4,866 
 
 5,358: 
 4-897 
 
 3,556 
 
 33*6 
 3.921 
 3,420 
 
 2.787 
 2,526 
 
 3,094 
 2,627 
 
 Carbon 
 
 Carbonic oxide 
 
 Olefiant gas 
 
 
 Experiments on the combustion of hydrogen and carbonic oxide by Bunsen indicate 
 that the temperature of combustion cannot exceed a certain limit, owing to the pheno- 
 menon of "dissociation" that is to say, when the temperature of combustion reaches 
 this limit the elementary bodies no longer combine. When, for instance, hydrogen is 
 burnt in presence of the exact quantity of oxygen necessary for complete combustion, 
 the heat produced by the combustion of a part of the hydrogen is sufficient to raise the 
 temperature of the mixture to such an extent that no further union of the elements can 
 take place. As soon as the temperature begins to fall fresh quantities of hydrogen are 
 burnt, and this process continues until the whole is consumed. 
 
 Under the conditions obtaining in the furnace of a steam boiler the temperature of 
 the products of combustion are necessarily always much lower than the preceding table 
 indicates, owing to the more or less incomplete oxidation of the gases, the presence of 
 incombustible matter and of moisture in the fuel and in the air, and the cooling by ra- 
 diation and conduction. Ledieu states that, under conditions of ordinary practice, the 
 temperature of the gases in the furnace of a marine boiler will hardly exceed 1500, 
 with a combustion of about 19 pounds of semi-bituminous coal per square foot of grate 
 per hour. 
 
STEAM BOILERS. 
 
 CHAP. 1L 
 
 8. Volume of Products of Combustion. An inspection of Table I. will show 
 that, at equal temperatures, the volume of carbonic acid gas is the same as that of the 
 oxygen entering into its composition ; but that the volume of steam is double the vol- 
 ume of oxygen required for the combustion of the hydrogen entering into its composi- 
 tion. Since, in the coals used in marine boilers, hydrogen bears only a small proportion 
 to the whole weight, the volume of the gaseous products of combustion may be treated 
 as practically equal to that of the air supplied to the furnace. The volume of air at 32 
 may be taken, in round numbers, as 12^ cubic feet for each pound of air. Neglecting 
 the variations in density due to the slight deviations of the pressure of the furnace- 
 gases from the mean atmospheric pressure, as of trifling importance in calculations for 
 practical purposes, the volume of the furnace-gases at any temperature may be calcu- 
 lated by the formula : 
 
 y T4- 461 2 
 _ A __L FTV i 
 
 F ' 493.2 
 
 which expresses the general law " that the volumes of gases vary directly as their ab- 
 solute temperatures." V and F represent the volume of the gas in question at the 
 temperatures T and 32 respectively. 
 
 The following table, given by Rankine, is based on the foregoing assumptions : 
 
 TABLE IV. 
 
 
 Volume of gases in cubic feet, per pound of fuel. 
 
 
 Supply of air, 
 
 Temperature in 
 
 
 degrees Fahrenheit. 
 
 
 
 
 
 12 pounds 
 per pound of fuel. 
 
 18 pounds 
 per pound of fuel. 
 
 24 pounds 
 per pound of fuel. 
 
 2500 
 
 906 
 
 1359 
 
 1812 
 
 l8 3 2 
 
 697 
 
 1046 
 
 1395 
 
 1472 
 
 5 88 
 
 882 
 
 1 176 
 
 III2 
 
 479 
 
 718 
 
 957 
 
 752 
 
 369 
 
 553 
 
 738 
 
 572 
 
 3i4 
 
 47i 
 
 628 
 
 392 
 
 2 59 
 
 389 
 
 5i9 
 
 212 
 
 205 
 
 37 
 
 409 
 
 104 
 
 172 
 
 258 
 
 344 
 
 68 
 
 161 
 
 241 
 
 322 
 
 3 2 
 
 J 5 
 
 225 
 
 300 
 
 
 
 
 j 
 
 9. Rate of Combustion. The rate of combiistion in a furnace is measured by the 
 number of pounds of fuel burned on a square foot of grate per hour. The weight of 
 fuel which can be burned depends on the quantity of air which can be made to pass 
 through the furnace and part with its oxygen to the combustible matter. In practice, 
 
SBC. 10. COMBUSTION. 43 
 
 with natural chimney dranght, the rates of combnstion vary from 7 to 16 pounds for an- 
 thracite coals, and from 12 to 27 pounds for bituminous coals, in different types of 
 marine boilers. With artificial draught, produced by a steam-blast or fan-blowers, the 
 rate of combustion may be raised to 120 pounds of coke in certain types of boilers. 
 
 The high rate of combustion which can be attained with bituminous coals is owing 
 to the fact that these coals, on being heated in the furnace, part readily with their 
 hydrocarbons in the form of gas, the solid portion of the coal being either left behind 
 as a .spongy, porous mass (coke), or showing numerous cracks all over the surfaces 
 which divide the lump into a great number of loosely cohering particles. In this man- 
 ner the air gets access to the interior of the solid coal and comes in contact with a larger 
 surface. The hard anthracite coals, on the contrary, remain solid during combustion, 
 and the air comes in contact only with their exterior. 
 
 1O. Draught of Furnaces. The velocity with which air passes through the grate 
 of a furnace depends on the difference of pressures existing within the furnace and the 
 ashpit, and on the resistance offered by the layer of fuel on the grate. The pressure 
 below the grate is the atmospheric pressure, unless it is either increased by forcing air 
 into the ashpit by means of a fan-blower, or diminished by preventing the free flow of 
 air into the ashpit. The pressure above the grate in the furnace is equal to the atmos- 
 pheric pressure, less the difference in weight of a vertical column of atmospheric air 
 having a base of a unit of area and a height equal to that of the chimney and of an 
 equal column of hot chimney -gas, plus the pressure required to overcome the various 
 resistances experienced by the gases in their passage from the furnace up the chimney. 
 When by the action of a jet or fan the weight of the column of chimney-gas is partly 
 counterbalanced, the pressure in the furnace is correspondingly diminished. 
 
 Peclet expresses all the resistances encountered by the gases in their passage from 
 the ashpit to the top of the chimney in terms of the head corresponding to the velocity 
 of the air flowing to the grate, per se. 
 
 Calling this head #, 
 
 the resistance due to the grate and the bed of fuel GJi, 
 the resistance due to changes in sectional area and direction of the flues C7i, 
 and the coefficient of friction of the gases moving over the surfaces of the flues/ 1 , 
 he gets an expression of the following form for the total head ff, which produces the 
 draught of a boiler, viz. : 
 
 1 + 
 
 where v is the velocity of the air flowing to the grate in feet per second. 
 
44 . STEAM BOILERS. CHAP. II. 
 
 0" fl/t V 
 
 In the expression -~- ^ \-r-j , which represents the resistance due to friction, 
 
 I is the combined length of the flues and of the chimney in feet ; 
 
 m is the "hydraulic mean depth" of the flues and chimney that is to say, the mean 
 of the area of the smoke-passages divided by their perimeter ; 
 
 , and t are the absolute temperatures of the chimney -gas and of the external air respec- 
 tively, and ~ represents the increase of volume, and consequently of velocity, 
 ii 
 
 due to. the increase of temperature of the gases in the flues and chimney over that 
 
 of the entering air ; 
 f, the coefficient of friction, has, according to Peclet, the value 0.012 for currents of 
 
 gas moving over sooty surfaces. 
 
 The value of CTi varies according to the arrangement, form, and proportions of the 
 smoke-passages, and consists of the sum of the following resistances, which have to be 
 calculated according to the laws governing the flow of fluids : 
 
 1. On entering the tubes or flues the gases experience a loss of head due to the "con- 
 tracted vein." 
 
 2. Sudden enlargements of the sectional area of the passages produce a loss of head. 
 3. Each change in the direction of a current produces a loss of head ; this loss 
 increases with the angle which the two directions make with each other, and is far 
 greater for sudden sharp bends than for bends with easy curves. 
 
 4. Several currents entering a common channel, and moving either with different velo- 
 cities or in different directions, produce a loss of head. 
 
 The value of G varies with the kind of fuel, the thickness of the bed of fuel on the 
 grate, and the velocity of the air passing through the grate. Peclet estimates that 
 when bituminous coal is burnt at the rate of about 22 pounds per square foot of grate 
 per hour, the resistance of the grate is 8 h, and that this resistance varies as the square 
 of the number of pounds of coal burnt per square foot of grate per hour. Coke offers 
 much less resistance than coal. 
 
 In coke-burning locomotives Peclet found the value of G- to vary from 5.20 to 6.26, 
 and he gives 7.14 7i as the mean value of H^ in locomotives, viz.: 1.93 Ji for the resist- 
 ance of the tubes, and 5.21 7i for the resistance of the grate. He remarks, however, 
 that these figures are merely rough approximations which may serve to give an idea of 
 the relative value of the different kinds of resistance. 
 
 In marine boilers located in the holds of vessels there is an additional loss of head, 
 due to the work expended in drawing the air through the narrow openings in the deck 
 
SBC 11. COMBUSTION. 45 
 
 to the ashpits. Isherwood states that, owing to this cause, the rate of combustion of 
 horizontal, return fire-tube boilers falls from 24 pounds of anthracite consumed when 
 the boiler stands in an open shed, to 16 pounds of anthracite when the boiler stands in 
 the hold of a vessel, natural chimney-draught being used ; and with the vertical water- 
 tube boiler of the Martin type the rate of combustion falls, under like conditions, from 
 16 pounds to 12 pounds of anthracite coal. 
 
 11. Chimney-draught "The head produced by the draught of a chimney is 
 equivalent to the excess of the weight of a vertical column of cool air outside the chim- 
 ney, and of the same height, above that of a vertical column, of equal base, of the hot 
 gas within the chimney." (RanJcine.) 
 
 The weight in pounds of a cubic foot of air at atmospheric pressure at any tempera- 
 
 ture is given by the formula : - ^ x 0.0807 [VI.], 
 
 t 
 
 where t is the absolute temperature of the air. 
 
 The weight in pounds of a cubic foot of the gas discharged by the chimney is very 
 nearly ^ ^ (m7 + ^ [yn ^ 
 
 and varies ordinarily from 0.084 x 493 ' 2 to 0.087 X . 
 
 t, t l 
 
 In this formula V, is the volume at 32 of the air supplied to the furnace per pound 
 of fuel ; 
 t, is the absolute temperature of the gas within the chimney. 
 
 If H denotes the height of the chimney, the unbalanced pressure producing the flow 
 of air to the grate is equal to 
 
 ff**2 (0.0807) - ff^ 2 Co. 0807 + - T \ ) : 
 
 or in case 300 cubic feet of air are supplied for each pound of fuel burned, 
 
 H ^- 2 (0. 0807) -H*^- (0.084), [VIII.] 
 39.80124 4 
 
 The head, expressed in feet of the external air, corresponding to this pressure is 
 found by dividing the foregoing expression by the weight of a cubic foot of air : 
 
 (0.0807) 
 
 \ 
 
 =// (l- L0409 f); [IX.] 
 
 I 
 
 / 
 
46 STEAM BOILERS. CHAP. II. 
 
 Substituting this value of If, in equation [V.], we get the following expression for the 
 velocity with which the air flows to the grate of a furnace : 
 
 1-1.0409-?- ~" 1 
 
 [X.] 
 
 The following conclusions may be drawn from this equation, viz.: Under otherwise 
 equal conditions, the velocity of the air flowing to the grate, and, consequently, the 
 rate of combustion, varies very nearly as the square root of the height of the chimney ; 
 strictly speaking, it is slightly less, because the value of I in the denominator increases 
 likewise with the height of the chimney. 
 
 With a fixed value of t, or absolute temperature of the external air, the value of the 
 
 numerator, y \ 1.0409 -r , increases with the temperature of the chimney, and becomes 
 
 v j 
 
 equal to unity when t, is infinite ; but this increase is very slow with high temperatures. 
 Forexample, the temperature of the external air being 50 Fahr., the expression 
 
 V 
 
 1 1.0409 - becomes equal to 
 * 
 
 .5486, - - .7059, - .7804, - 1.000, 
 when the temperature of the chimney-gas, in degrees Fahr., is 
 
 300 600 900 infinity. 
 
 f 7 If \ 3 
 Since the resistance due to friction, represented by the expression J l^\ in the de- 
 
 f 
 
 nominator, increases as the square of the absolute temperature -of the hot gases, there 
 must be a certain chimney-temperature for which the value of v becomes a maximiim ; 
 but this temperature varies with the resistances represented by G and C and with the 
 
 factor . 
 m 
 
 The value of v is further diminished by the cooling of the chimney -gases by the ra- 
 diation and conduction of heat from the smoke-pipe. This loss increases likewise with 
 the height of the chimney and the temperature of the escaping gases. 
 
 Data are wanting to assign exact values to the various resistances under different 
 conditions, but equation [X.] may be used to find the limit of the influence which a 
 change of conditions can have on the efficiency of a boiler. 
 
SBC. 12. COMBUSTION. 47 
 
 The height of the chimneys of marine boilers is limited by practical considerations, 
 and it seldom exceeds 65 feet. 
 
 The expenditure of heat to produce an increase in the rate of combustion augments 
 so rapidly after a certain limit has been reached that it is not advantageous to increase 
 the chimney-temperature to the point at which the rate of combustion becomes a maxi- 
 mum. It is generally assumed that the chimney-temperature of marine boilers should 
 not exceed 600 Fahr. 
 
 12. Artificial Draught. "The head produced by a blast-pipe is equivalent to 
 that part of the atmospheric pressure which is balanced by means of the impact of the 
 jet of steam against the column of gas in the chimney." 
 
 " The work which a fan or other blowing-machine must perform in a given time in 
 blowing aii- into a furnace so as to produce a given head, is found by multiplying the 
 
 pressure equivalent to that head, in pounds on the square foot H^ ^ (0.0807) into 
 
 the number of cubic feet of air blown in, taken at the temperature at which it quits 
 the blowing-machine." 
 
 If t t is the temperature on the absolute scale at which the air leaves the blowing- 
 machine, the net or useful effect of the machine per second will be 
 
 WF^ 5; (0.0807) [XI.]; 
 
 when w denotes the weight of fuel burned in the furnace per second, 
 and F the volume at 32 of the air supplied per pound of fuel. 
 
 "The gross power or energy required to drive a blowing-fan is greater than the use- 
 ful work in a proportion which varies much in different machines and is very uncer- 
 tain." (RanJcine.) 
 
 13. Efficiency of Furnace. Under otherwise equal conditions the rate of com- 
 bustion varies very nearly as the square root of the height of the chimney. 
 
 This is true on the supposition that the mean temperatures of the chimney -gases 
 are the same from base to top for the chimneys of different heights, and that the gases 
 receive no frictional resistance from the sides of the chimney, neither of which is prac- 
 tically the case. In practice, when gases of the same temperature enter the base of 
 chimneys of different heights the temperature at the top is different, owing to external 
 refrigeration, etc., being less as the chimney is higher, so that the mean temperature in 
 the higher chimney will be less than in the lower one ; and the velocity of the draught 
 of the higher chimney will be less comparably with the velocity of the draught of the 
 lower chimney than it should be according to the law of the square roots of the heights. 
 
48 STEAM BOILERS. CHAP. II. 
 
 Again, as the Motional resistances of the sides of the chimney are as the square of the 
 velocity of the gases, the velocity in a chimney sufficiently high would become uni- 
 form, after which no fiirther increase of height would increase the draught. 
 
 An increase of the furnace-temperature of a boiler of given proportions causes an 
 increase of the chimney-temperature, and, consequently, of the draught ; and, vice 
 versa, a diminution of the furnace-temperature, in consequence of an excessive amount 
 of air admitted to the furnace or of incomplete combustion, causes a decrease of the 
 chimney-temperature and of the rate of combustion. 
 
 In a boiler of given proportions and dimensions the rate of combustion may be 
 varied by aiding the chimney-draught by a jet or fan-blower, or by impeding it by 
 means of a damper in the chimney or flues ; by regulating the flow of air to the grate 
 by means of the ashpit-doors ; or by increasing or decreasing the resistance of the grate 
 by varying the depth of the bed of fuel. 
 
 The draught can be regulated much more readily by means of dampers placed in the 
 flues or chimney than by closing the ashpit-doors. 
 
 When the damper is closed the furnace is kept filled with the gases of combustion, 
 which, by enveloping the fuel, effectually prevent its combustion, notwithstanding their 
 leakage out past the damper accompanied by a corresponding entrance of air, for com- 
 bustion will not take place with the atmospheric oxygen diluted beyond a certain point. 
 But when the ashpit-door is closed the damper being open or absent there is a free 
 escape for the gases of combustion, and all the atmospheric oxygen leaked in is available 
 for combustion. Were damper and ashpit-door perfectly tight both would be equally 
 efficacious. 
 
 The thickness of the bed of fuel can be varied only between certain limits ; for, when 
 the bed is too thin, too large a quantity of air will rush through the grate ; when it is 
 too thick, the combustion will be incomplete for want of sufficient air. 
 
 The principal causes which reduce the efficiency of the furnace of a boiler, by dimin- 
 ishing the temperature of the products of combustion and the total heat produced by 
 the combustion of fuel, are the following : 
 
 I. The absorption of heat by incombustible solid matter and by moisture contained 
 in the fuel. The proportion of incombustible matter in coal varies from If to 26 per 
 cent. In the better classes of English semi-bituminous coal used in marine boilers it 
 forms from 6 to 12 per cent, of the fuel. The average quantity of incombustible matter 
 contained in Pennsylvania anthracite is 16f per cent. 
 
 Although the refuse matter has a high temperature when it is removed from the fur- 
 nace, the quantity of heat thus lost is small, amounting in the worst cases to barely one 
 
SEC. 13. COMBUSTION. 49 
 
 per cent. The incombustible matter produces a more injurious effect, especially when 
 it fuses easily and forms clinker, by preventing the free access of air to the combustible 
 portion of the fuel. The principal losses due to the presence of incombustible matter 
 are connected with the process of cleaning the fires, and will be referred to later. 
 
 The moisture present in fuels not only makes latent a relatively large quantity of 
 heat during evaporation, but prevents often the complete oxidation of the combustible 
 portion of the fuel. 
 
 Wood, when newly felled, contains, on an average, 40 per cent, of moisture ; after 
 eight or twelve months' ordinary drying in air the proportion of moisture is from 20 to 
 25 per cent. (Itankine.) 
 
 Coke, being of a porous texture, readily attracts and retains water from the atmos- 
 phere ; and sometimes, if it is kept without proper shelter, from 0.15 to 0.20 of its 
 gross weight consists of moisture. (Rarikine.) 
 
 The quantity of moisture absorbed by coal varies with its texture, and with the 
 duration and manner of its exposure to dampness. Hard anthracites absorb only a 
 very small quantity of moisture. 
 
 II. Waste ofunburnt combustible matter in the solid state. This waste depends on 
 the behavior of the fuel during the process of combustion and on the care and skill of 
 the fireman. Anthracite coal, when suddenly heated, splits into small pieces, and dry 
 bituminous coal is converted during combustion into a loosely cohering mass of small 
 particles, which are liable to fall through the grate when the fire is stirred for the pur- 
 pose of removing the ashes. In cleaning the fire a small quantity of coal is also un- 
 avoidably hauled from the furnace with the refuse. "It is impossible to estimate 
 the greatest amount of this kind of waste which may arise from careless firing ; but the 
 amount which is unavoidable with good firing has, in some cases, been ascertained by 
 experiment and found to range from nothing up to about 2 per cent." (Rankine.} 
 
 III. Losses arising from an admission of excessive quantities of air to the fur- 
 nace. By admitting too large a quantity of air to the furnace the temperature of the 
 products of combustion is decreased and their mass and volume are increased ; the effect 
 of this is a reduction of the rate of combustion, the loss of a larger quantity of heat 
 present in the escaping chimney-gases, and sometimes, when the furnace temperature 
 is greatly reduced, the incomplete combustion of certain gaseous products of the fuel. 
 Too large a calorimeter, or cross-area of the passages over the bridge-wall or through 
 the flues relatively to the rate of combustion, favors the admission of an excessive quan- 
 tity of air. The proper proportions of the parts of a boiler will be considered later. 
 
 The air should pass in thin, evenly-distributed streamlets through the bed of fuel, 
 
50 STEAM BOILERS. CHAP. II. 
 
 or it should enter the furnace above the grate in the form of numerous fine jets. The 
 admission of air in large masses is always injurious, but is to a certain extent un- 
 avoidable with the ordinary methods of firing ; when the bed of fuel is too thin or not 
 evenly distributed over the grate, and whenever the door is opened for the purpose of 
 throwing fuel on the grate, or levelling, slicing, and cleaning the fire, large masses of air 
 rush into the furnace. 
 
 When the fuel is not evenly distributed over the grate, the air rushes with great 
 violence and in large masses through the places left uncovered or insufficiently covered, 
 causing the phenomenon of "back-draught." 
 
 When the combustion is forced by artificial draught the lumps of coal must be 
 smaller and the bed of coal must be thicker than with natural draught, so as to make 
 the interstices between the lumps smaller and the route of the air through the bed of 
 coal more tortuous. 
 
 The following is the result of "experiments showing the effect on the economic 
 vaporization of admitting an increased air-supply through the grates to the incan- 
 descent coal upon them by carrying thinner fires." "In experiment A the fires 
 were 7 inches thick, and in experiment D they were less than half that thickness, 
 the grates being kept just covered. The rate of combustion in both experiments was 
 sensibly the same, and very slow, being 6.498 pounds of the combustible portion of 
 anthracite per square foot of grate-surface per hour in experiment A, and 6.149 pounds 
 in experiment D. 
 
 "The economic vaporization in experiment A was 11.8976 pounds of water from the 
 temperature of 212 Pahr. and under the atmospheric pressure, per pound of the com- 
 bustible portion of the anthracite, and in experiment D 10.2716 pounds. Hence the 
 admission of the increased air-supply through the grates, due to maintaining a very 
 thin fire upon them, decreased the economic vaporization by the fuel 
 
 /11.8976- JHX2716\ IQO = 13.667 per centum." (Report of Board of United States 
 \ 11.8976 / 
 
 Naval Engineers on " The AsJicroft Furnace-doors and Grate-bars" March 27, 
 1878.) 
 
 The loss resulting from opening the furnace-door is avoided or lessened by using a 
 moving grate, to which the fuel is supplied by mechanical means ; or by using vibrating 
 or rocking grate-bars for removing the ash and clinker from the fire. The former have 
 not come into use for marine boilers ; examples of the latter will be found in Chapter 
 XIII. 
 
 The exact amount of the loss in vaporific efficiency of coal burnt in a boiler-furnace, 
 
SEC. 13. 
 
 COMBUSTION. 
 
 51 
 
 due to the opening of the door for the purpose of removing the ash and clinker, may be 
 deduced from the experiments made by the Board of United States Naval Engineers 
 convened to determine the relative value of the Murphy shaking-grate and of the com- 
 mon grate : " The Murphy apparatus kept the fires clean and free of holes, crushed the 
 clinker and removed all the refuse from the furnaces into the ashpits, without opening 
 the furnace-doors for such purposes and without using fire-tools"; and it was found 
 "that the economic gain in fuel due to the Murphy grate was in direct proportion to 
 the per centum of refuse removed through the furnace-door, as appears from the fol- 
 lowing exhibit": 
 
 Kind of coal burnt. 
 
 Per centum of the coal 
 consumed on the com- 
 mon grate, removed as 
 refuse through the 
 furnace-door. 
 
 Economic gain in fuel 
 by the Murphy grate 
 in per centum of the 
 fuel consumed on the 
 common grate. 
 
 Bituminous coal lumps 
 
 IO -9475 
 I3-4338 
 . I9-3I47 
 
 4-1529 
 5-7417 
 7.2972 
 
 Bituminous coal dust. 
 
 Anthracite 
 
 
 "The mean of the three determinations gives for the economic loss in fuel when 
 burned on a common grate 0.3935 per centum of that fuel for every one per centum it 
 contains in refuse removed through the furnace-door." (Report on the Murphy Grate- 
 bar by a Board of United States Natal Engineers, June 25, 1878.) 
 
 IV. Waste of unburntfuel in the gaseous and smoky states. The complete com- 
 bustion of coke, hard anthracite, and coals containing only a small proportion of hydro- 
 carbons presents no difficulty as long as the thickness of the bed of fuel is properly 
 proportioned to the draught, and a small quantity of air enters the furnace in jets 
 through the perforated furnace-door in addition to the quantity passing through the 
 grate. 
 
 The combustion of highly bituminous coal presents greater difficulties. Special care 
 has to be taken, in the design of the boiler and in the management of the fire, that the 
 hydrocarbon gases distilled from the coal are thoroughly mixed, at a sufficiently high 
 temperature, with the proper quantity of air. A greater or less quantity of the carbon 
 contained in the hydrocarbon gases remains frequently iinburnt, forming soot or 
 smoke ; and in extreme cases the fixed carbon contained in the coal is alone completely 
 burnt. 
 
 "If smoke is mixed with carbonic-acid gas at a red heat the solid carbonaceous par- 
 ticles are dissolved in the gas and carbonic oxide is produced. This is the mode of 
 
52 
 
 STEAM BOILERS. 
 
 CHAP. II. 
 
 operation of contrivances for destroying smoke by keeping it at a high temperature 
 without providing a sufficient supply of air ; and the result is a waste instead of a sav- 
 ing of fuel." (Rankine.) 
 
 TABLE V. 
 
 GENERAL SYNOPTICAL TABLE OF THE CHARACTER AND EFFICIENCY OF AMERICAN COALS, BY 
 
 W. R. JOHNSON. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 
 
 12 
 
 '3 
 
 Designation of coal. 
 
 Location of mine. 
 
 Specific gravity. 
 
 Cubic feet of space required to 
 stow a ton. 
 
 Volatile combustible matter in 
 100 parts. 
 
 Fixed carbon in 100 parts. 
 
 uj 
 
 1 
 
 8 
 
 'fe 
 1 
 
 1 
 
 Moisture in fuel in 100 parts. 
 
 Ratio of fixed to volatile com- 
 bustible matter. 
 
 Rate of combustion in Ibs. of coal 
 per square ft. of grate per hour. 
 
 Percentage of waste in ashes 
 and clinker. 
 
 Pounds of steam from water at 
 212 per pound of coal. 
 
 Steam from 212 from one Ib. 
 of combustible. 
 
 Beaver Meadow, Slope No. 3. 
 Beaver Meadow, Slope No. 5. 
 
 Pa. 
 Pa. 
 Pa. 
 
 I.6IO 
 
 I-55I 
 .477 
 
 40.78 
 39-86 
 j.1 71: 
 
 2.38 
 
 2.66 
 \ 07 
 
 88.94 
 91.47 
 OO 75 
 
 7.11 
 
 5-15 
 
 4 41 
 
 i-57 
 0.72 
 
 1.77 
 
 37-37 
 
 34-39 
 20.56 
 
 6.69 
 6.27 
 6.52 
 
 11.96 
 6.74 
 6.Q7 
 
 9.21 
 
 9.88 
 
 10.06 
 
 10.462 
 10.592 
 10.807 
 
 
 Pa. 
 
 .464 
 
 41.64 
 
 2.06 
 
 8o.O2 
 
 6.13 
 
 1.89 
 
 30.00 
 
 6.69 
 
 6.97 
 
 IO.II 
 
 10.871 
 
 
 Pa. 
 
 .500 
 
 JXX^O 
 
 5.28 
 
 8Q.I5 
 
 5.56 
 
 O.OI 
 
 16.88 
 
 6.95 
 
 7.22 
 
 8.93 
 
 9.626 
 
 
 Pa. 
 
 .421 
 
 45.82 
 
 S.QI 
 
 87.74 
 
 
 2 OO 
 
 22.44 
 
 6.45 
 
 8.93 
 
 9-79 
 
 10.764 
 
 Lykens Valley 
 
 Pa. 
 
 .389 
 
 46.13 
 
 6.SS 
 
 83.84 
 
 25 
 
 OO3 
 
 12.19 
 
 6.92 
 
 12.24 
 
 9.46 
 
 10.788 
 
 New York and Maryland) 
 
 Md. 
 
 -431 
 
 41.71 
 
 12.31 
 
 73-50 
 
 I2.4O 
 
 1.79 
 
 5-97 
 
 6.28 
 
 12.71 
 
 9.78 
 
 1 1. 2O8 
 
 NefFs Cumberland 
 
 Md. 
 
 337 
 
 41.26 
 
 12.67 
 
 74-53 
 
 10.34 
 
 2.46 
 
 5-88 
 
 7.86 
 
 10.96 
 
 9-44 
 
 10.604 
 
 Dauphin and Susquehanna . 
 
 Pa. 
 
 Pa. 
 
 -443 
 .324 
 
 44-32 
 
 42.22 
 
 13.82 
 14.78 
 
 74.24 
 73.11 
 
 11.49 
 10.77 
 
 0.45 
 1.34 
 
 5-37 
 4-95 
 
 6.86 
 7-77 
 
 16.36 
 1 1. 20 
 
 9-34 
 9.72 
 
 II.I7I 
 10.956 
 
 
 Pa. 
 
 .388 
 
 4O.45 
 
 13.84 
 
 71.53 
 
 13. 06 
 
 0.67 
 
 5.16 
 
 6.33 
 
 16.92 
 
 8.91 
 
 10.724 
 
 Cambria County 
 
 Pa. 
 
 .407 
 
 41.90 
 
 20.52 
 
 69.37 
 
 9.15 
 
 0.96 
 
 
 6.68 
 
 9-75 
 
 9.24 
 
 10.239 
 
 
 Va. 
 
 .294 
 
 41.45 
 
 29.86 
 
 53.01 
 
 14.74 
 
 2.39 
 
 1.78 
 
 6.68 
 
 14-83 
 
 8.29 
 
 9.741 
 
 
 Pa. 
 
 .252 
 
 47.85 
 
 36.76 
 
 54. Q3 
 
 7.07 
 
 1.24 
 
 1.49 
 
 
 8.25 
 
 8.20 
 
 8.942 
 
 
 Ind. 
 
 .273 
 
 47.OI 
 
 33. QO 
 
 58.44 
 
 4-97 
 
 2.60 
 
 1.72 
 
 11.09 
 
 5.12 
 
 7-34 
 
 7-734 
 
 
 
 
 IO662 
 
 
 
 O.3O7 
 
 
 
 15.87 
 
 0.307 
 
 4.60 
 
 4.707 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
SEC. 13. 
 
 COMBUSTION. 
 
 53 
 
 TABLE VI. 
 
 GENERAL SYNOPTICAL TABLE OF THE CHARACTER AND EFFICIENCY OF ENGLISH COALS, SHOWING 
 THE RESULTS OF THE INVESTIGATIONS OF DE LA BECHE AND PLAYFAIR. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 IO 
 
 II 
 
 12 
 
 3 
 
 M 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 1 
 
 it 
 
 n 
 
 
 
 3 
 
 
 
 
 
 
 
 
 5 
 
 
 
 o. 2 
 
 c t 
 
 
 
 i 
 
 
 
 
 
 
 
 
 u 
 
 1 
 
 v " 
 s'o 
 
 1 - 
 
 Locality or name of coal. 
 
 $ 
 
 i 
 
 
 
 
 
 
 
 1 
 
 1 
 
 O 
 
 o 
 
 II 
 
 1- 
 
 
 1 
 
 o e 
 S2 
 J> 
 
 
 1 
 
 . 
 
 
 
 
 O 
 
 1 
 
 1 
 
 U 
 X 
 
 *~ o 
 
 ill 
 
 H ; T 
 
 
 jj 
 
 "5 
 
 
 
 
 E 
 
 1 
 
 3 
 
 
 
 C 
 
 o 
 
 g 
 
 J 
 
 ill 
 
 - . - 
 
 
 ! 
 
 O * 
 
 1 
 
 
 
 
 "3 
 
 in 
 
 5 
 
 ^ 
 
 e 
 
 ^ 
 
 1 
 
 I J 
 
 M 
 
 i" 
 
 Welsh Graigola 
 
 
 
 Rt R-J 
 
 
 
 
 
 
 
 
 
 
 
 " Anthracite (Jones, 
 Aubrev & Co.).. ' 
 
 1-375 
 
 38.45 
 
 04.07 
 9144 
 
 346 
 
 0.21 
 
 0.45 
 0.79 
 
 7.19 
 
 2.58 
 
 3-24 
 1.52 
 
 85.5 
 92.9 
 
 91-38 
 
 6.12 
 
 19.60 
 
 13-563 
 14.593 
 
 9-35 
 946 
 
 " Old Castle Fierv , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Vein ..]' 
 
 1.289 
 
 43-99 
 
 87.68 
 
 4.89 
 
 I-3I 
 
 0.09 
 
 3-39 
 
 2.64 
 
 79-8 
 
 76.16 
 
 6.02 
 
 14.936 
 
 8.01 
 
 Binea Coal ... 
 
 
 
 RR f*h 
 
 
 
 
 
 
 
 
 
 
 
 Llangenneck 
 
 
 
 Rs ifi 
 
 4.03 
 
 M3 
 
 0.33 
 
 1.03 
 
 i't 
 
 
 84.14 
 
 7.65 
 
 15-093 
 
 9-94 
 
 Pentrepoth 
 Pentrefelin 
 
 1-31 
 
 
 88.72 
 
 4.^0 
 4.50 
 
 1.O7 
 
 0.18 
 
 
 2.43 
 3-24 
 
 0.54 
 3-36 
 
 S3.O9 
 
 82.5 
 
 77-15 
 79- !4 
 
 4-73 
 
 14.260 
 
 14.838 
 
 8.86 
 8.72 
 
 Powel's DufTrvn . . . 
 Mynydd Xewvdd . . 
 Cwm Frood Rock ) 
 
 1.326 
 
 1.31 
 
 42.09 
 39-76 
 
 .:- 
 .88.26 
 '84.71 
 
 3.72 
 4.66 
 5-76 
 
 1.45 
 
 1-77 
 
 I.2I 
 
 4-55 
 0.60 
 3-52 
 
 3.26 
 3.24 
 
 84-3 
 74-8 
 
 81.04 
 71.56 
 
 5-52 
 5-56 
 2.91 
 
 13.787 
 15-092 
 14.904 
 
 6.36 
 10.15 
 9-52 
 
 Vein \ 
 
 1-255 
 
 40.52 
 
 82.25 
 
 5-^4 
 
 i. ii 
 
 1.22 
 
 3.S8 
 
 6.00 
 
 68.8 
 
 62.80 
 
 2.08 
 
 14.788 
 
 8.70 
 
 Cwm Nanty-gros. . . 
 Pontv Pool . . . 
 
 1.28 
 
 40.00 
 
 78.36 
 
 5-59 
 
 C ftf* 
 
 1.86 
 
 3-01 
 
 3-58 
 
 5.60 
 
 65.6 
 
 60.00 
 
 1.79 
 
 13.932 
 
 8.42 
 
 Ebbw Vale 
 
 I 27> 
 
 
 . . -, 
 
 5.00 
 
 2 16 
 
 2 -39 
 
 4-3 
 
 5-5 2 
 
 
 59.28 
 
 1-73 
 
 14.295 
 
 747 
 
 Porthmawr Rock 
 
 
 
 
 
 
 
 0.39 
 
 1.50 
 
 77-5 
 
 
 3-59 
 
 15-635 
 
 IO.2I 
 
 Vein f 
 
 1.39 
 
 42.02 
 
 74-70 
 
 4-79 
 
 1.28 
 
 0.91 
 
 3.60 
 
 14.72 
 
 63.1 
 
 48-18 
 
 1-37 
 
 12.811 
 
 7-53 
 
 Scotch Dalkeith Jewel ) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Seam i' 
 
 1.277 
 
 44.98 
 
 74-55 
 
 5.14 
 
 O.IO 
 
 0.33 
 
 I5-5I 
 
 4-37 
 
 49- 8 
 
 4543 
 
 1.24 
 
 12.313 
 
 7.08 
 
 " Dalkeith Corona- ; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 tion Seam i' 
 
 1.316 
 
 43-36 
 
 76.94 
 
 5.20 
 
 Trace 
 
 0.38 
 
 14-37 
 
 3-io 
 
 53-5 
 
 50.40 
 
 1.14 
 
 12.772 
 
 7-71 
 
 " Fordel Splint 
 
 1 25 
 
 
 7n cfi 
 
 
 
 T tf, 
 
 Q a 
 
 
 
 
 
 
 
 
 
 
 
 
 1.13 
 
 
 O-33 
 
 
 52.O3 
 
 45.O3 
 
 
 13.817 
 
 7.56 
 
STEAM BOILERS. 
 
 CHAP. IL 
 
 w 
 j 
 
 M 
 
 
 
 S 
 
 iff 
 
 
 
 00 
 
 ^ 
 n w 
 
 tn 
 
 O S 
 
 W ^ 
 
 U W 
 
 < o 
 
 H 
 
 W J 
 
 U U 
 
 S S 
 
 o > 
 
 as, > 
 
 - < 
 
 i 1 
 
 oa 5 
 
 < 
 
 <J ^ 
 
 s i 
 
 H ^ 
 
 O pd 
 
 tn ~ 
 
 * H 
 
 o t: 
 
 u * 
 
 i 
 
 S- 
 
 5 
 
 0. 
 
 U 
 
 h 
 O 
 tn 
 
 iJ 
 
 7. 
 
 X 
 
 3 
 
 S ^ 
 
 PH W 
 
 s s 
 
 o 
 
 O a 
 
 f5 
 
 O 
 
 a 
 c 
 
 o )ooj ajBnbs jad JIKHJ 
 ja aiqijsnquiOD jo spunod ui 
 uoijsnqiuoD jo 3)BJ uiniuixu'i 
 
 ^. y t 
 
 o 
 
 P- 
 
 II; 
 
 Bigj 
 I 
 
 - 
 
 - 
 
 S o B a. 
 .H^-a U 3 
 
 iijhS| 
 
 K a p,.^ 
 
 iSii 1 ! 
 
 liSfSl 
 
 i 
 
 Q. Q. 
 
 uoijsnquioD jo 
 ajEJ AVOIS auj iv 
 
 -snqraoo jo S 
 
 umjpaui 
 
 
 -snquioo jo 3jej 
 uiniuixEUiaqi jy 
 
 'uoijsnquioo jo 
 ajEJ A\O[S auj jy 
 
 -snqraoD joa 
 
 flop 
 
 =41 'V 
 
 -snqtuoa jo SJBJ 
 uinuaixELU aqi )y 
 
 ]3Hj jo uoiuod 3|qilsnaiuOD 
 jo *q| j.nt oanss.ud ouaqasoui 
 -IE 43pun -j O tlt jo'j3je\\ jo 
 
 'Sq[ UI 
 
 qsy 
 
 5 1 & 
 : rt .j c 
 
 * - ** *fi. 
 
 NO CJ) 
 
 p r^ efl 
 
 l/> M O 
 
 n 
 
 CTl >O S M 
 
 M d d 
 
 11 
 odd 
 
 . 
 
 M B B* 19 10 & 
 
 T ^ *O IO 1O OS 
 
 CTi CTl O O 0\ 00 
 
 a 1 f S 
 
 \d od od t^ 
 
 in o 
 
 CO t% 
 
 * 
 
 
 "S f J 5 
 
 D, * 
 
 H 
 
 K R if & 8 
 i s 
 
 n A & <fi d 
 
 d d M d d 
 
 aw QO M o 
 
 d d - H S 
 
 ^2 
 S 
 
 K S 
 
 a s 
 
 ~ 8, 8. g 
 
 3, 
 
 a 
 
 s a ft !? 
 
 eft * -* 
 
 S 
 
 u_ 
 
 7T 
 
 O CO 
 
 \is S 
 
 * * 
 
CHAPTER in. 
 
 TRANSMISSION OF HEAT AND EVAPORATION. 
 
 1. Laws of Transmission of Heat. All bodies transfer heat to the surrounding 
 cooler bodies till their temperature becomes equalized. 
 
 The quantity of heat transferred from one body to another depends on the difference 
 of their temperatures and their respective specific heat, and, in case the molecular con- 
 dition of one or both bodies is changed, on their respective latent heat. 
 
 The rate at which the transfer takes place between two bodies depends, first, on the 
 degree and the difference of their temperatures ; secondly, on the extent of surface 
 transmitting and receiving heat ; thirdly, on the nature of the material of the bodies, 
 and on the condition of their surfaces ; fourthly, on the nature and thickness of the in- 
 tervening substances. 
 
 There are three processes by which heat is transferred, called respectively radiation, 
 conduction, and convection. 
 
 Radiation, is the exchange of heat by direct rays between bodies not in contact, and 
 takes place at all temperatures and at any distance through space. 
 
 These rays produce the phenomena of heat only in bodies which intercept them more 
 or less completely. Gases intercept and emit heat-rays very feebly ; most solids and 
 fluids, on the contrary, intercept and emit these rays, which are either wholly absorbed 
 or partly reflected by them. The rate of transmission of heat by radiation between 
 two bodies is increased by making their surfaces dark and rough, and diminished by 
 smoothing and polishing them. In the steam boiler the radiation from the solid incan- 
 descent fuel is far greater than from the carbonaceous flame, while the transparent hot 
 gases scarcely radiate any heat at all. 
 
 Conduction is the transfer of heat between bodies in contact, and is distinguished as 
 external and internal conduction, accordingly as it takes place between two distinct 
 bodies, or between the parts of one continuous body. 
 
 Convection is the diffusion of heat in a fluid mass by means of the circulation and 
 mixture of the particles of that mass. The most rapid convection of heat is effected by 
 means of cloudy vapor, which combines the mobility of a gas with the comparatively 
 
 55 
 
56 STEAM BOILERS. 
 
 CHAP. III. 
 
 greater conducting power of a liquid. Convection is the only process that can be de- 
 pended upon for the rapid distribution of heat throughout a mass of water. 
 
 The rate of internal conduction in a solid body depends, first, on the variation of 
 temperature along a line perpendicular to the section through which the heat is trans- 
 ferred, and, secondly, on the coefficient of internal conductivity of the substance in 
 question. Although this coefficient varies slightly with different temperatures, it may 
 be considered practically as constant for the same substance. Designating the number 
 of thermal units transmitted through a square foot of area per hour by Q, 
 the temperatures of the two sides of a substance by T and T a 
 
 the rate of conduction through a plate of the thickness t, in inches, may be ex- 
 pressed by the equation : 
 
 where p may be called, according to Rankine, the coefficient of thermal resistance. 
 The value of this coefficient for various substances is as follows : copper = 0.0018 ; 
 iron = 0.0043* ; zinc = 0.0045 ; lead = 0.0090 ; marble and calcareous deposits = 
 0.0716; brick = 0.1500. 
 
 When a plate consists of layers of different substances, its total internal thermal 
 resistance may be found by adding together the resistances of the several layers ; in 
 
 rp _ m 
 
 such a case equation [I.] assumes this form: Q = ~ fi. [la.] 
 
 -^ p t 
 
 * " Principal Forbes . . . has likewise determined the absolute conductivity of wrought-iron. In his experi- 
 ments conductivity was expressed in terms of the amount of heat as unity which is required to raise the temperature 
 of one cubic foot of water by one degree centigrade. He expresses the amount of heat reckoned in such units which 
 would traverse in one minute across an area of one square foot a plate of iron one foot thick with the two surfaces 
 maintained at temperatures differing by one degree centigrade. According to these experiments, the conductivity at 
 C. of one of his bars was .01337, while that of another bar was only .00992. This discordance was probably due to 
 a difference in the quality of the iron of the two bars." (Balfour Stewart, "An Elementary Treatise on Heat.") 
 
 The larger result was obtained with a square bar 1} ins. thick, the smaller one with a square bar 1 inch thick. 
 In either case the conductivity diminished as the temperature increased. In the larger bar, 
 the conductivity at 100 C. was 24.3 per cent, less than at C. 
 
 " 200 C. " 13.4 " " 100 C. 
 
 " 275 C. " 8.6 " " 200 C. 
 
 In the smaller bar " 100 C. " 15.8 " " C. 
 
 " 200 C. " 8.5 " " 100 C. 
 
 " 275 C. " 5.2 " " 200 C. 
 
 Professor Tait has given reasons for believing that the thermal conductivity of metals may be inversely proportional 
 to their absolute temperature. 
 
SBC. 8. TRANSMISSION OF HEAT AND EVAPORATION. 57 
 
 The rede of external conduction between a solid body and a fluid may be expressed 
 by dividing the difference of their temperatures by a coefficient of external resistance, 
 depending on the nature of the substances, the condition of their surfaces, and also on 
 their temperatures. Designating the value of these coefficients for the two surfaces 
 of a given plate by a and a l , the quantity of heat-units transmitted per square foot 
 of surface per hour, from one fluid to another, through a plate t inches thick, may 
 
 T T 
 
 be expressed by the formula : Q = * . [II. ] 
 
 G -j- ff l -r pi 
 
 2. Experiments on the Transmission of Heat by Isherwood. Careful ex- 
 periments on the heat-conducting power of different metals were made by Isherwood in 
 1867 at the Washington Navy- Yard. The metals experimented upon were pure, refined 
 copper, brass consisting of 60 parts of copper and 40 parts of zinc, rolled wrought-iron 
 of the best quality, and ordinary cast-iron which had had several remeltings. Cylin- 
 drical pots of these metals, 10 inches in inside diameter and 21i inches in inside height, 
 were immersed in a common cylinder, which was well protected by a non-conducting 
 material and supplied with steam of a certain temperature and pressure from a large 
 boiler. The pots were kept filled with water of 212 temperature by a steady supply 
 from separate tanks, and the quantity of this water evaporated by the heat supplied 
 by the steam-bath measured the thermal conductivity of the respective metals ; hence, 
 "exactly the same temperature was upon the entire exterior surface of all four pots, 
 while the water inside the pots was vaporized under identical hygrometrical and baro- 
 metrical conditions of the atmosphere. Each experiment consisted of from 54 to 108 
 hours, nine hours of each day being devoted to it." Great care was exercised to 
 maintain the temperatures equal throughout each experiment, and all conditions were 
 noted every 15 minutes in a tabular log. " Each series of experiments was exactly 
 repeated on pots of three thicknesses of metal viz., , J, and f inch with the view 
 of ascertaining whether the vaporizing efficiency of the metals was affected by their 
 thickness. 
 
 " With each thickness of metal twenty experiments were made, the temperature of 
 the steam increasing by 5 Fahr. from 220 Fahr. for the first to 320 Fahr. for the last 
 experiment. All the pots were turned and bored to exactly the same dimensions. 
 
 "For each experiment there was calculated the number of pounds of water vaporized, 
 from the temperature of 212 Fahr. and under the standard atmospheric pressure of 
 29.92 inches of mercury, by each pot ; and the heat-conducting powers of the metals 
 were considered to be in the ratio of those weights. The following are the general 
 results : 
 
58 
 
 STEAM BOILERS. 
 
 CHAP. III. 
 
 "1ST. All other things equal, the weight of water vaporized in a given time was in 
 the direct ratio of the difference of the temperatures, inside and outside of the pots. 
 
 "2r>. All other things equal, the weight of water vaporized in a given time was not 
 affected by the thickness of the metal. The rate of vaporization was exactly as great 
 from the f -inch thick metal as from the f-inch thick metal. 
 
 "3D. The following are the fractions of a pound of water vaporized per hour from each 
 square foot of the interior surface of the pots, from the temperature of 212 Fahr. and 
 under the standard atmospheric pressure of 29.92 inches of mercury, by a difference of 
 temperature of one degree Fahrenheit between the inside and the outside of the pots. 
 
 " These are the absolute heat-conducting powers of the metals named" viz. : 
 
 
 Thermal conductivity m 
 terms of fractions of a 
 pound of water of 212 
 vaporized under 
 atmospheric pressure. 
 
 Thermal conductivity in 
 terms of heat-units 
 transmitted per hour 
 through one square foot 
 of material by difference 
 of temperature of 
 i Fahr. 
 
 Relative thermal conduc- 
 tivity. 
 
 CoDDer. . 
 
 0.665365 
 0.576610 
 0.386895 
 0.326956 
 
 642.543 
 556.832 
 373-625 
 3I5-74I 
 
 I.OOOOOO 
 0.866607 
 0.581478 
 0.49*393 
 
 
 Wrought-iron 
 
 Ccist-iron 
 
 
 3. Experiments on the Transmission of Heat by P6clet. The results of 
 the foregoing experiments agree pretty closely with those obtained by former investi- 
 gators, as far as the relative thermal conductivity of these metals is concerned ; the 
 absolute values obtained are, however, much smaller than those found by Peclet. The 
 latter' s experiments prove that the rate at which fluids transmit heat to, and absorb 
 heat from, solid bodies, depends other conditions being equal greatly on the more or 
 less perfect circulation of the fluids, so that each particle of fluid is at once replaced by 
 other particles as soon as it has absorbed or parted with some heat by contact with the 
 solid. On this account Peclet used water instead of steam as the source of heat in his 
 experiments, because, by the condensation of the latter, a film of water is deposited on, 
 and clings tenaciously to, the walls of the experimental vessel ; and he produced rapid 
 circulation in the heat-absorbing, as well as in the heating, medium by mechanical means. 
 By observing such precautions Peclet obtained results which are in accordance with 
 the law that the quantity of heat transmitted through a solid body in a unit of time 
 diminishes in the direct ratio of the increase of thickness. 
 
 Peclet' s experiments on the cooling of vessels when exposed to the air, the circula- 
 tion of the latter being produced simply by the effect of the transmitted heat, proved 
 
SBC. 4. TRANSMISSION OP HEAT AND EVAPORATION. 59 
 
 that, when the walls of the experimental vessel were covered by pulverulent deposits, 
 the quantity of heat transmitted in a unit of time was independent of the internal ther- 
 mal conductivity of the metal and, within the limits of ordinary practice, of the thick- 
 ness of the walls, but depended greatly on the form of the vessel and increased in a 
 certain ratio with the difference of temperature ; both these latter elements affecting the 
 rapidity of the circulation of the air. 
 
 4. Transmission of Heat in a Steam Boiler. Since in a steam boiler the heat- 
 ing-surfaces become soon covered with deposits of scale, rust, and soot, while the cir- 
 culation of the hot gases on one side and of the water or steam on the other side is 
 more or less imperfect, the evaporative power of these surfaces may be considered, 
 within the limits of ordinary practice, as independent of the thickness and kind of the 
 metal used, but depending principally on their form and position, on the condition of 
 their surfaces, and on the difference of the temperatures to which the opposite sides 
 are exposed. 
 
 Rankine expresses the total thermal resistance of the plates and tubes of a steam 
 
 boiler by ffl ; substituting this expression for the divisor a -f ff t + P t, equation 
 
 *- * 
 [II. ] assumes the form : 
 
 a 
 
 The numerical value of a lies between 160 and 200. This formtila is not intend- 
 ed to give more than a rough approximation ; in fact, the varying conditions obtaining 
 in a steam boiler preclude the possibility of an accurate theoretical determination of its 
 coefficient of thermal resistance. 
 
 To ensure the proper circulation of the water to which heat is to be transmitted two 
 conditions must be observed viz., first, the heat must be applied to the bottom of the 
 vessel containing the water, so that the latter, as it becomes lighter by being heated, 
 may ascend, being displaced by a descending column of heavier, colder water ; second- 
 ly, the heating-surfaces must have such a shape and position as to permit the free 
 escape of the heated water and steam. 
 
 As long as the clean surface of a boiler-plate is in contact with solid water the most 
 intense heat of the furnace may be applied to the other side without overheating the 
 plate ; when, on the contrary, a plate is in contact with steam, it will soon assume the 
 temperature of the hot gases to which the other side is exposed. This great heat- 
 absorbing capacity of water is owing to three catises : first, its thermal conductivity is 
 greater than that of gases ; secondly, its specific heat is more than twice as great as 
 
 OF THE 
 
 UNIVERSITY 
 
 OF 
 
60 STEAM BOILERS. CHAP. III. 
 
 that of steam ; and, thirdly, a large quantity of heat becomes latent during evapo- 
 ration. 
 
 5. Efficiency of Heating-surfaces in a Steam Boiler. The efficiency of a 
 heating-surface may be measured by the ratio borne by the amount of heat transmitted 
 by it to the total amount of heat available for transmission. This efficiency in a steam 
 boiler depends on the following conditions : first, the proportion which the extent of 
 the surfaces receiving and transmitting heat bear to the volume of hot gas bounded by 
 them ; secondly, the difference of temperatures of the hot gas on the one side of the 
 plates, and of the water or steam on the other side ; thirdly, the time allowed for the 
 transmission of heat ; fourthly, the nature, condition, and thickness of the plates 
 forming the heating-surfaces ; fifthly, the position and shape of the plates ; sixthly, 
 the nature of the heating and heat-absorbing media. 
 
 The following interesting experiment on the influence of the position of heating- 
 surfaces on their efficiency is recorded by Tredgold : " Mr. Armstrong found that a 
 cubical metallic box, submerged in water and heated from within, generated steam from 
 its upper surface more than twice as fast per unit of area than it did from the sides 
 when vertical, and that the bottom yielded none at all.. These remarkable differences 
 are owing to the difficulty with which steam separates from a vertical surface to give 
 place to fresh charges of water, and to the impossibility of leaving the inverted surface 
 at all. By slightly inclining the box the elevated side much more easily parted with 
 the steam, and the rate of evaporation was increased ; while on the depressed side the 
 steam hung so sluggishly as to lead to an overheating of the metal." 
 
 In the marine steam boiler the temperature of the gases in the furnace ranges pro- 
 bably between 1,500 and 2,500 ; and when these gases enter the chimney their tempe- 
 rature has been reduced to from 450 to 650. 
 
 On the other hand, the temperature of the steam, and consequently of the water, 
 ranges between 250 and 350, according to the steam-pressure used. Since the water is 
 introduced into the boiler at a much lower temperature, varying ordinarily between 
 100 and 120, there would be theoretically a decided gain in the heating efficiency if 
 the water entered the boiler at the point where the hot gases have the lowest tempera- 
 ture. Such an arrangement is, however, rarely made in marine boilers, on account of 
 mechanical difficulties connected with it. 
 
 The plates forming the furnace of a marine boiler transmit, relatively, by far the 
 greatest amount of heat to the water ; for, in addition to the effect produced by the 
 high temperature of the evolved gases in contact with the sides and the top of the fur- 
 nace, the radiation of heat from the incandescent solid fuel is of considerable impor- 
 
SEC. 5. TRANSMISSION OP HEAT AND EVAPOEATION. 61 
 
 tance. Peclet states that the quantity of heat radiated from incandescent carbon, freely 
 suspended, is at least one-half of its total heat of combustion. In the furnace, how- 
 ever, only the upper surface of the fuel radiates heat directly to the water-heating sur- 
 faces. The rays of heat emitted through the open spaces of the grate are mostly ab- 
 sorbed by the ashes ; but this heat, as well as that received by the furnace-door, is to a 
 great extent reabsorbed by the entering air. 
 
 In the combustion-chamber or back-connection the temperature of the products of 
 combustion is probably fully as high as in the furnace, since the thorough mixing of 
 the hot gases with the air completes their combustion. The radiation of heat from the 
 carbonaceous flame is likewise frequently of much importance at this part of the 
 boiler. 
 
 Isherwood states, as the result of experiments on marine boilers, that of the evapo- 
 ration of water in well-proportioned tubular boilers about 55 per cent, is due to the 
 furnace and back-connection, while the heating-surface contained in those parts is only 
 about 20 per cent, of the total heating-surface of the boiler. 
 
 On leaving the combustion-chamber or back-connection the gases pass generally be- 
 tween or into numerous tubes. By subdividing in this manner the gaseous mass into a 
 great number of streamlets the proportion of heating-surface to volume of gas for a 
 given length is greatly increased, and the absorption of heat takes place rapidly. 
 
 When the hot gases pass through horizontal tubes the upper side of them is most 
 effective as a heating-surface, since on the inside it is kept relatively clean of sooty de- 
 posits and the hottest gases come in contact with it, while at the same time the steam- 
 bubbles escape most freely from that portion of the tubes. Only such portions of the 
 hot gases as come in actual contact with the heating-surfaces impart their heat to them. 
 In internally-heated horizontal tubes the outer film of the hot gases descends as soon 
 as it has parted with some of its heat ; in this manner the temperature of the current 
 is equalized to some extent by convection of heat. Some writers ascribe the smaller 
 efficiency of internally-heated vertical tubes to the fact that the outer film of gases in 
 contact with the metal has no tendency to mingle with the central portion of the cur- 
 rent, while at the same time the steam-bubbles generated at the lower end of the tube 
 continue to envelop the tube as they rise. 
 
 On account of the great difference in temperature of the gases as they enter and 
 leave the tubes, the heating power of the tubes decreases rapidly toward the chimney- 
 end. According to formula [III.] of this chapter the quantity of heat transmitted varies 
 directly as the square of the difference of temperature at the two sides of a plate : in 
 practice the efficiency of the tube-surface diminishes at a still greater rate, on account 
 
62 STEAM BOILERS. CHAP. III. 
 
 of the greater accumulation of soot and scale at the chimney-end of the tubes. On the 
 other hand, it must be observed that the same cause which produces the greater deposi- 
 tion of soot viz., the diminished velocity of the gases (the volume of which decreases 
 proportionally with the temperature, while the area of the tubes remains constant) 
 allows the gases also to remain a longer time in contact with the heating-surfaces. 
 
 Isherwood suspended various metals, the melting-points of which are well known, 
 at different points in the chimney-ends of horizontal fire-tubes, and found that the tem- 
 perature of the discharged gases is considerably higher at the upper than at the lower 
 rows of tubes in the same boiler ; he estimates that this difference of temperature is at 
 times as high as 300. This difference is probably due partly to the tendency of the hot- 
 test gases to rise to the highest point, partly to the fact that the mass of rising steam- 
 bubbles envelops the upper tubes to a greater or less extent, while the lower tubes are 
 surrounded by a more solid body of water. 
 
 When the gases enter the front smoke-connection, or uptake, in their passage to the 
 chimney, their temperature must be reduced to a sufficiently low degree to cause no in- 
 jury to the metal plates of the boiler, as these are no longer in contact with water, but 
 with steam or air. Sometimes special provisions are made to utilize some of the heat 
 of the escaping gases in superheating or drying the steam, by keeping them in contact 
 with extensive surfaces surrounded by the steam. 
 
 6. Loss of Efficiency of Boilers by External Radiation and Conduction. 
 The heat radiated from the incandescent coal through the openings of the furnace-door 
 and through the interstices of the grate is almost completely reabsorbed by the enter- 
 ing air. The temperature of the gases in the chimney is reduced to some extent by the 
 radiation and conduction of heat from the smoke-pipe, and the draught of the boiler is 
 correspondingly diminished ; but the loss due to this cause is trifling in large boilers. 
 
 The loss of heat due to radiation and conduction from the shell of marine boilers 
 may be reduced to a small amount by covering the shell with non-conductive materials, 
 and by forming a dead-air space between the shell of the boiler and its covering. 
 
 The loss of heat by radiation and conduction from steam boilers, pipes, etc., has 
 been determined by experiments made in the years 1863-65, under the direction of the 
 Bureau of Steam Engineering of the United States Navy Department, which have been 
 described and analyzed by Chief -Engineer Isherwood, U.S.N., in the Journal of the 
 Franklin Institute, March, 1878. 
 
 The radiator used in these experiments was a flat box constructed of plate-iron ^ 
 inch thick. " The covering employed was the ordinary cow-hair felt, manufactiared for 
 clothing steam boilers, weighing one pound per square foot when 1 inches thick. It 
 
SEC. 6. 
 
 TRANSMISSION OF HEAT AND EVAPORATION. 
 
 63 
 
 was stitched tightly over the radiator so as to be in contact at all points, thereby pre- 
 venting air-spaces, or air circulation, between the felt and the radiator." The thickness 
 of the felt covering used in the experiments varied from J inch to 7 inches. 
 
 The experiments were made with steam-pressures varying between 10 and 60 pounds 
 per square inch above the atmosphere, and the results showed that in still air, "ceteris 
 paribus, within the limits of the experimental temperatures, the quantity of heat 
 radiated in equal times from the same surface with different temperatures on its op- 
 posite sides was in the direct ratio of their difference." 
 
 By plotting the mean final results for each set of experiments it was shown that the 
 units of heat radiated per hour per square foot of surface per degree Fahrenheit diffe- 
 rence of temperature on the opposite sides of the surface, varied almost exactly "in the 
 inverse ratio of the square roots of the thicknesses of felt employed, from the thickness 
 of 7| inches up to the thickness of 1 inch, from which latter thickness up to naked 
 metal the curve, though a fair one, followed no regular law." 
 
 Thickness in inches of 
 the cow-hair felt on the 
 air side of the boiler-plate 
 iron. 
 
 Number of Fahrenheit units of heat lost per hour per 
 square foot of boiler-plate iron, 5-16 inch thick, per degree 
 Fahrenheit difference of temperature between that of the 
 steam on one side of the metal and that of the still air upon 
 the opposite side. 
 
 Naked 
 0.25 
 0.50 
 
 o-7S 
 
 I.OO 
 
 1-25 
 1-50 
 
 2.9330672000 
 1.0540710250 
 0.5728646875 
 0.4124625750 
 
 0-3070554725 
 0.2746387609 
 0.2507097171 
 
 Two experiments, each lasting 72 consecutive hours, were made to determine the 
 effect of covering the boiler with felt on the economic evaporation, at the Navy- Yard, 
 New York, in October, 1863, and are described by Isherwood in "Experimental Re- 
 searches," Vol. II. The experimental boiler was of the locomotive type, and had 5.3066 
 square feet of grate-surface, 22.1 cubic feet of water-room, and 12.2 cubic feet of steam- 
 room. The boiler stood in a shed of rough boards with one end open ; the circulation 
 of air around it was consequently considerably greater than would have been under 
 the decks of a vessel. The experiments consisted in determining the economic evapo- 
 ration of the boiler when covered with thick felt and when not covered, the condi- 
 tions of the trials being otherwise as nearly as possible alike. The results of these ex- 
 periments are summed up by Isherwood as follows : 
 
(J4 STEAM BOILERS. CHAP. III. 
 
 " The number of pounds of water evaporated per hour during the experiment with 
 the boiler not covered with felt was (- ^-^ ?= 1 687.565 ; and as we have seen that the 
 
 addition of felt effected a saving of^22.05 per centum of this quantity, we have (687.565 
 X 0.2205 ) 151.608 pounds of steam condensed per hour by its omission. The exter- 
 nal surface of the boiler from which heat was radiated was 94.09 square feet, conse- 
 quently (- ' '--^= \ 1.6113 pounds of steam were condensed per hour per square foot 
 
 \ */4. Ut/ / 
 
 of unfelted surface. The temperature of the water and steam within was 267 Fahr., 
 and of the external atmosphere 53.5. The thickness of the boiler-plate was one-quar- 
 ter inch." 
 
 The writer also calls attention to the fact that the per centum of condensation due to 
 radiation from the external surface of the boiler will be greatly less for large boilers 
 " and in proportion to size, because, while for similar boilers the external surface in- 
 creases as the square of any dimension, the contents increase as the cube of the same 
 dimension, and the steam -producing capability is as the contents." 
 
 7. Efficiency of Boilers. The efficiency of a boiler is measured by the ratio 
 borne by the quantity of heat expended in heating and vaporizing the water to the 
 quantity of heat representing the calorific power of the fuel consumed. 
 
 The quantity of heat usefully expended in raising the temperature and vaporizing 
 the water is the difference between the quantity of heat generated in the furnace and 
 the quantity of heat present in the gases discharged from the chimney, less the quan- 
 tity of heat lost by external radiation and conduction. 
 
 The efficiency of the furnace determines the total quantity of heat generated, and 
 the weight and temperature of the products of combustion. (See section 13, chap- 
 tern.) 
 
 The efficiency of the heating-surface determines the temperature of the gases dis- 
 charged from the chimney. (See section 5 of the present chapter.) 
 
 The loss of efficiency due to external radiation and conduction has been discussed 
 in section 6 of the present chapter. 
 
 Ordinarily from 20 to 33 per cent, of the total heat of combustion is expended in 
 the production of chimney-draught in marine boilers. The additional losses of heat by 
 radiation, by the incomplete combustion of the solid or gaseous parts of the fuel, and 
 by the dilution of the gases of combustion with an excess of air reduce the amount of 
 heat available for heating the water to about 60 per cent, of the total heat of combus- 
 tion in average practice with marine boilers. 
 
SEC. 7. TRANSMISSION OP HEAT AND EVAPORATION. 65 
 
 " When the draught is produced by means of a blast-pipe or of a blowing-machine 
 no elevation of temperature above that of the external air is necessary in the chimney ; 
 therefore furnaces in which the draught is so produced are capable of greater economy 
 than those in which the draught is produced by means of a chimney. It appears, fur- 
 ther, that with a forced draught there is less air required for dilution, consequently a 
 higher temperature of the fire, consequently a more rapid conduction of heat through 
 the heating-surface, consequently a better economy of heat than there is with a chim- 
 ney-draught." (Rankine.) 
 
 The following formula has been devised by Rankine to express " to an approxi- 
 mate degree of accuracy" the efficiency of a boiler: 
 
 - Bs nv i 
 
 E ~S+AF L V-J 
 
 E denotes the theoretical evaporative power, and E 1 the actual evaporative power of 
 one pound of a given sort of fuel consumed in a boiler ; B and A are constants, which 
 are found empirically ; "the value of A is probably proportional approximately to the 
 square of the quantity of air supplied per pound of fuel" ; 8 denotes the number of 
 square feet of heating-surface per square foot of grate, and fthe number of pounds of 
 fuel burned per hour per square foot of grate. 
 
 " The following are the values of the constants B and A which have been found to 
 agree best with experiment, so far as the practical performance of boilers has hitherto 
 been compared with the formula : 
 
 Boiler Class I. The convection taking place in the best manner, either 
 by introducing the water at the coolest part of the boiler and 
 making it travel gradually to the hottest, or by heating the feed- 
 water in a set of tubes in the uptake ; the draught produced by B A 
 
 a chimney, 1 o.5 
 
 Boiler Class II. Ordinary convection and chimney-draught, .... ft 0.5 
 
 Boiler Class III. Best convection and forced draught, 1 0.3 
 
 Boiler Class IV. Ordinary convection and forced draught, .... 0.3 
 
 " When there is a feed-water heater its surface should be included in computing 
 " . . . " The formula is framed on the supposition that the admission of air and the 
 management of fire are such that no appreciable loss occurs, either from imperfect com- 
 bustion or from excess of air, the construction and proportions of the furnace, and the 
 mode of using it, being the best possible for each kind of coal." (Rankine.} 
 
STEAM BOILERS. 
 
 CHAP. III. 
 
 8. Influence of the Rate of Combustion on the Evaporative Efficiency of 
 Boilers (Isherwood, ' Experimental Researches ,' vol. ii.) " The economic and poten- 
 tial evaporations of boilers, other things equal, are greatly affected by the rate of combus- 
 tion. With each increase in that rate above about 5 pounds of combustible per square 
 foot of grate per hour, the economic evaporation decreases and the potential evaporation 
 increases." . . . "In the following table will be found (for the horizontal fire-tube 
 
 TABLE VII. 
 
 SHOWING THE ECONOMICAL AND POTENTIAL EVAPORATION OF THE HORIZONTAL FIRE-TUBE 
 BOILER WITH ANTHRACITE CONSUMED WITH DIFFERENT RATES OF COMBUSTION. 
 
 Pounds of 
 anthracite con- 
 sumed per hour 
 per square foot 
 of grate-surface. 
 
 Pounds of 
 water evaporat- 
 ed under atmo- 
 spheric pressure 
 from 212 Fahr. 
 by one pound of 
 anthracite. 
 
 Per centum of 
 the total heat 
 developed by 
 the combustion, 
 utilized evapo- 
 rative ly. 
 
 Temperature 
 in degrees Fahr. 
 of the products 
 of combustion 
 when leaving 
 the boiler. 
 
 Weights of 
 steam furnished 
 by the boiler in 
 equal time, 
 expressed pro- 
 portionally. 
 
 Weights of 
 steam furnished 
 by equal 
 weights of 
 anthracite, ex- 
 pressed propor- 
 tionally. 
 
 Weights and 
 bulks of anthra- 
 cite required to 
 furnish equal 
 weights of 
 steam, express- 
 ed proportion- 
 ally. 
 
 6 
 
 10.49 
 
 84.42 
 
 444-7 
 
 .OOOO 
 
 I. OOOO 
 
 I. OOOO 
 
 7 
 
 10.44 
 
 84.01 
 
 454-8 
 
 .l6ll 
 
 0.9952 
 
 1.0048 
 
 8 
 
 io-3S 
 
 83-59 
 
 472.6 
 
 '3124 
 
 0.9867 
 
 I.OI35 
 
 9 
 
 10.23 
 
 82.33 
 
 49 6 -3 
 
 .4628 
 
 0.9752 
 
 1.0254 
 
 10 
 
 10.05 
 
 80.88 
 
 532.o 
 
 .5967 
 
 0.9580 
 
 1.0438 
 
 ii 
 
 9.81 
 
 78-95 
 
 5796 
 
 7M5 
 
 0.9352 
 
 1.0693 
 
 12 
 
 9-53 
 
 76.69 
 
 6 35-5 
 
 .8169 
 
 0.9085 
 
 I.IOO7 
 
 13 
 
 9.21 
 
 74.12 
 
 699.0 
 
 1.9023 
 
 0.8780 
 
 1.1389 
 
 14 
 
 8.87 
 
 71.38 
 
 766.6 
 
 1.9730 
 
 0.8456 
 
 I.I826 
 
 15 
 
 8.52 
 
 68.56 
 
 836.3 
 
 2.0305 
 
 O.8I22 
 
 I.23I2 
 
 16 
 
 8.21 
 
 66.07 
 
 897.7 
 
 2.0871 
 
 0.7826 
 
 1.2778 
 
 17 
 
 7-95 
 
 63.98 
 
 949-3 
 
 2-1473 
 
 0-7579 
 
 1.3I94 
 
 18 
 
 7.70 
 
 61.97 
 
 999.0 
 
 2.2021 
 
 0.7340 
 
 1.3624 
 
 *9 
 
 7.48 
 
 60.19 
 
 1042.9 
 
 2.2580 
 
 0.7131 
 
 1.4023 
 
 20 
 
 7-3 2 
 
 58-91 
 
 i74-5 
 
 2.3260 
 
 0.6978 
 
 I-433I 
 
 21 
 
 7.16 
 
 57.62 
 
 1 106.4 
 
 2.3890 
 
 0.6825 
 
 1.4652 
 
 22 
 
 7.04 
 
 56-65 
 
 "30.3 
 
 2.4608 
 
 0.67II 
 
 I.490I 
 
 23 
 
 6.92 
 
 55-69 
 
 1154.0 
 
 2.5288 
 
 0.6597 
 
 I-5I58 
 
 2 4 
 
 6.82 
 
 54.88 
 
 1174.0 
 
 2.6oo6 
 
 0.6501 
 
 I-5382 
 
 boiler, with the tubes above the furnace) the principal results due to different rates of 
 combustion, varying from 6 to 24 pounds of anthracite per square foot of .grate-surface 
 per hour, supposing its refuse to be one-sixth." . . . "The calorimeter is taken at 
 one-eighth of the grate-surface, and the heating-surface at 25 times the grate-surface. 
 The economic evaporation is given in pounds of water evaporated under atmospheric 
 pressure from 212 Fahr. per pound of anthracite. The economic evaporation is also 
 
SEC. 9. TRANSMISSION OP HEAT AND EVAPORATION. 67 
 
 expressed in ' Per centum of the Total Heat developed by the Combustion, utilized 
 evaporatively.' This is calculated on the supposition that the theoretical economic 
 evaporation of the pound of anthracite with one-sixth of refuse is 12.4263 pounds of 
 water under atmospheric pressure from 212 Fahr. . . . From these per centum and 
 assuming the temperature of the products of combustion in the furnace, at the moment 
 of their formation, to be 2,469 Fahr. above that of the atmosphere (taken at 60 Fahr.), 
 due to an air-supply of twice that which is chemically necessary for perfect combustion 
 the temperature of the products of combustion, when leaving the boiler, is easily cal- 
 culated." ..." The quantities in the second column of the table namely, the 
 poiinds of water evaporated under atmospheric pressure from 212 Fahr. by one pound 
 of combustible are the means given by a careful collation of the results of all the expe- 
 riments with the respective boilers." 
 
 9. Superheated. Steam. When the saturated steam generated in a boiler is 
 brought into contact with heating-surfaces the absorbed heat will, in the first place, 
 vaporize the particles of water held in suspension in the mass of steam, or, in other 
 words, dry the steam ; and any additional heat absorbed by the dry steam will raise its 
 temperature above the boiling-point corresponding to its pressure, or, in other words, 
 superheat the steam. 
 
 The experiments made by Tate and Fairbairn on the density of superheated steam 
 showed "that, for temperatures within about 10 Fahr. of the saturation-point, the rate 
 of expansion [of superheated steam] very greatly exceeds that of air ; whereas at higher 
 temperatures the rate of expansion very nearly approaches that of air. Hence it would 
 appear that for some degrees above the saturation-point the steam is not decidedly in 
 an aeriform state, or, in other words, that it is watery, containing floating vesicles of 
 un vaporized water." 
 
 By drying and superheating steam its dynamic efficiency in the engine is increased, 
 and fuel is economized in consequence. Comparing the efficiency of superheated steam 
 with that of dry saturated steam, Rankine calculates that, in an engine using steam of 
 an initial pressure of 34 pounds on the square inch and expanding it to five times its 
 original volume, by superheating the steam so as to raise its temperature from 257. 5 to 
 428 a saving of about 15 per cent, would be effected ; and, in case the whole of the 
 superheating is effected by heat which would otherwise have been wasted, the saving 
 would be about 23 per cent. In practice a still greater saving has frequently been 
 effected by the introduction of superheaters, even when a more moderate degree of 
 superheating was employed probably in consequence of the additional increase in effi- 
 ciency due to drying the steam. 
 
68 STEAM BOILERS. CHAP. HI. 
 
 The following are some of the results of Isherwood's experiments with superheated 
 steam recorded in ' Experimental Researches,' vol. ii. : 
 
 In the U. S. S. Mackinaw the superheating was effected by carrying the water-level 
 from 4 to 6 inches below the upper tube-plate in the "Martin" boiler. When the 
 steam was cut off at 0.70 and 0.21 of the stroke of the piston from the commencement, 
 the gain by superheating was 34.02 per centum and 38.85 per centum of the cost with 
 saturated steam, respectively. 
 
 In the U. S. S. Eutaw the superheating apparatus described in section 3, chapter 
 xii., was used. With an expansion due to cutting off the steam at 0.32 of the stroke of 
 the piston from the commencement the cost of the indicated horse-power was 8.67 per 
 centum less with steam superheated 79.7 Fahr. than with saturated steam. With an 
 expansion due to cutting off the steam at 0.58 of the stroke of the piston from the com- 
 mencement the cost of the indicated horse-power was 14.76 per centum less with steam 
 superheated 123.2 Fahr. than with saturated steam. 
 
 In the steamer Qeorgeanna the superheating was effected partly in the steam-drum 
 surrounding the uptake, and partly in a system of pipes within the uptake. The steam- 
 superheating surface in the steam-drum alone, amounting to 299 square feet, was suffi- 
 cient to impart enough additional heat to the steam not only to prevent all condensa- 
 tion of it in the cylinder due to any cause whatever, but to enable it to reach the end 
 of the stroke of the piston in a superheated state ; and this amount of superheating in- 
 creased the economic efficiency of the steam two-eighths above that due to it as satu- 
 rated steam. 
 
 The superheating-pipes and the steam-drum combined, containing an aggregate 
 heating-surface of 900 square feet, raised the temperature of the steam as it entered the 
 cylinder to 335 Fahr., while its pressure varied between 20 and 30 pounds per square 
 inch above the atmosphere ; and the economic efficiency of the steam was increased 
 thereby three-eighths above that due to it as saturated steam. 
 
 " The greater the degree of superheating the greater will be the gain, but not pro 
 rata. . . . The benefits derived from superheating, for a given number of degrees, 
 are much greater at the commencement of the superheating than at the end, because the 
 rate of expansion is higher near the saturation-point, because the prevention of conden- 
 sation in the cylinder is completed with a very moderate amount of superheating, and 
 because the loss from radiation is less. Practically, too, more steam is lost by leakage, 
 with the same pressure, past the cylinder-piston and valves the more it is rarefied by the 
 superheating ; and after a certain temperature is reached the friction of the piston 
 is materially increased by additional increments, and by tighter packing rendered 
 
SBC. 10. TRANSMISSION OF HEAT AND EVAPORATION. 69 
 
 necessary to prevent excessive leakage." (Isherwood, ' Experimental Researches,' 1 
 
 vol. ii.) 
 
 On account of these practical difficulties it is not considered advisable to let the tem- 
 perature of superheated steam exceed greatly 300 Fahr. When the temperature of 
 saturated steam exceeds that point it is better to add only enough heat to the steam 
 after it is generated to dry it, and to prevent condensation in the cylinders by steam- 
 jackets. 
 
 1O. Efficiency of Superheating-surfaces. The term " superheating-surface" 
 is applied to all heating-surfaces passing through the steam-room of a boiler. 
 
 According to section 5 of the present chapter a superheating apparatus would be 
 most efficient that is to say, transmit the greatest proportion of the total available heat 
 per unit of surface if placed at the hottest part of the furnace. The superheating 
 apparatus is, however, usually placed in the uptake of the boiler, where it absorbs a 
 portion of the heat of the escaping gases. In some cases special furnaces have been 
 used for superheating the steam. 
 
 Two methods of superheating steam in boilers may be distinguished viz., either all 
 the steam which is generated passes through the superheating apparatus placed be- 
 tween the boilers and the engines, or only a portion of the steam which is generated 
 enters the superheater, and is mixed in greater or less proportions with the saturated 
 steam coming directly from the boilers before it passes to the engines. The latter 
 method was introduced by Wethered, and the degree of superheating can be regulated 
 by it to a great nicety. 
 
 In calculating the gain in the efficiency of a boiler by the addition of superheating- 
 surface the heat absorbed in drying the steam has to be regarded as expended in evapo- 
 rating an additional amount of water. The addition of superheating-surface in the 
 uptake frequently increases the economic evaporative efficiency of the boiler at the ex- 
 pense of its potential evaporative efficiency (because the chimney-temperature is low- 
 ered, and the additional heating-surface offers an additional amount of resistance to 
 the escaping gases), unless without the superheating-surface the temperature would ex- 
 ceed the limit given in section 11, chapter ii. 
 
 Isherwood found that by properly proportioned water-heating surfaces the evapora- 
 tion in the boilers of the Georgeanna (see section 9 of the present chapter) could have 
 been 16.5 per centum greater for the same amount of coal. "Now, we have seen the 
 gain by superheating in the case of the Georgeanna to be three-eighths of the effect of 
 saturated steam from her boiler ; consequently we find that it is more economical to ex- 
 pend heat in superheating steam after it is generated than in generating its equivalent 
 
70 . STEAM BOILERS. CHAP. III. 
 
 of saturated steam, by (37.5 16.5) = 21 per centum. It is therefore advantageous, by 
 this amount, to provide a separate superheating apparatus, and superheat the steam in 
 it by the direct expenditure of fuel, in those cases in which it is not allowable to place 
 the superheater in the uptake on account of the height required very objectionable in 
 war-steamers taking care at the same time to employ a type and proportion of boiler 
 that will give the maximum evaporation." (IsTierwood, ''Experimental ResearctiesJ 
 vol. ii.) 
 
 Practically considered, the value of superheaters depends, as far as the boilers are 
 concerned, not only on the economic and potential evaporative efficiency of the latter, 
 but on the additional bulk, weight, and cost of the superheating apparatus, on the 
 labor and expense of keeping it in working order, and on its liability to derangement. 
 
SEC. 10. 
 
 TRANSMISSION OF HEAT AND EVAPORATION. 
 
 71 
 
 TABLE VIII. 
 
 SHOWING PROPERTIES OF WATER AND OF STEAM. 
 {From Isherwootfs ' Experimental Researches] vol. '.) 
 
 Total pres- 
 sure of 
 steam in !bs. 
 per sq. inch. 
 
 Temperature in 
 degrees Fahren- 
 heit. 
 7. 
 
 Total units of 
 heat per pound 
 of water from 
 32 to T Fah- 
 renheit. 
 
 Total units of 
 heat per pound 
 of steam from 
 32 to T Fah- 
 renheit. 
 
 Latent units of 
 heat per pound 
 of steam. 
 
 Weight of steam 
 per cubic foot in 
 fractions of a 
 pound. 
 
 Cubic feet of 
 steam per 
 pound. 
 
 Volume of 
 steam, water at 
 39. i being 
 unity. 
 
 I 
 
 101.36 
 
 69-4305 
 
 1112.8548 
 
 1043.4243 
 
 .0034692 
 
 288.2475 
 
 17983.200 
 
 5 
 
 162.51 
 
 130.8914 
 
 1131.5055 IOO0.6l4I 
 
 .0136650 
 
 73.1806 
 
 4565.597 
 
 10 
 
 193.20 
 
 161.8704 
 
 II4O.8660 978.9956 
 
 .0262900 
 
 38.0373 
 
 2373.075 
 
 1469 
 
 2I2.0O 
 
 180.9000 
 
 1146.6000 
 
 965.7000 
 
 .0380071 
 
 26.3109 
 
 1641.424 
 
 15 
 
 213.04 
 
 181.9540 
 
 1146.9172 
 
 964.9633 
 
 0387839 
 
 25-7839 
 
 1608.607 
 
 20 
 
 227.95 
 
 197.0789 
 
 1151.4647 
 
 954-3858 
 
 .0511491 
 
 I9-5507 
 
 1219.729 
 
 2 5 
 
 240.07 
 
 209.3952 
 
 n55- l6 i3 
 
 945.7661 
 
 .0633879 
 
 15.7760 
 
 984.236 
 
 3 
 
 250.26 
 
 219.7656 
 
 1158.2693 
 
 938.5037 
 
 0755004 
 
 13-245 
 
 826.327 
 
 35 
 
 259.22 
 
 228.8964 
 
 1161.0021 
 
 932.1057 
 
 .0874913 
 
 11.4298 
 
 713.084 
 
 40 
 
 267.17 
 
 237.0076 
 
 1163.4268 
 
 926.4192 
 
 0993593 
 
 10.0645 
 
 627.903 
 
 45 
 
 274-33 
 
 244.3208 
 
 1165.6106 
 
 921.2898 
 
 .1111084 
 
 9.0002 
 
 561.506 
 
 5 
 
 280.89 
 
 251.0279 
 
 1167.6114 
 
 916.5835 
 
 .1227283 
 
 8-1473 
 
 508.292 
 
 55 
 
 286.96 
 
 257.2400 
 
 1169.4628 
 
 912.2228 
 
 .1342847 
 
 7-4469 
 
 464.695 
 
 60 
 
 292.58 
 
 262.9967 
 
 1171.1769 
 
 908.1802 
 
 .1456581 
 
 6.8654 
 
 428.319 
 
 65 
 
 297.84 
 
 268.3892 
 
 1172.7812 
 
 904.3920 
 
 1569473 
 
 6.3716 
 
 397 5 10 
 
 70 
 
 302.77 
 
 273-4473 
 
 1174.2848 
 
 900.8375 
 
 .1681290 
 
 5-9479 
 
 371.076 
 
 75 
 
 307-42 
 
 278.2219 
 
 'I75-7946 
 
 897-5727 
 
 .1791958 
 
 5-58o5 
 
 348.156 
 
 80 
 
 311.86 
 
 282.7841 
 
 H77-0573 
 
 894.2732 
 
 .1901576 
 
 5-2588 
 
 328.086 
 
 85 
 
 316.08 
 
 287.1234 
 
 1178.3444 
 
 891.2210 
 
 .2OIOII2 
 
 4.9748 
 
 310.368 
 
 90 
 
 320.10 
 
 291.2597 
 
 II79-5705 
 
 888.3108 
 
 .2118493 
 
 4.7222 
 
 294.610 
 
 95 
 
 32394 
 
 295- 2I 34 
 
 1180.7417 
 
 885.5283 
 
 .2224IO7 
 
 4.4961 
 
 280.506 
 
 100 
 
 327-63 
 
 299.0150 
 
 1181.8671 
 
 882.8521 
 
 2329599 
 
 4.2926 
 
 267.806 
 
 I0 5 
 
 33I-I8 
 
 302.6750 
 
 1182.9499 
 
 880.2748 
 
 .2434062 
 
 4.1084 
 
 256-313 
 
 no 
 
 334-59 
 
 306.1920 
 
 1183.9899 
 
 877.7979 
 
 2537549 
 
 3.9408 i 245.860 
 
 "5 
 
 337-89 
 
 309-5979 
 
 1184.9964 
 
 875-3985 
 
 .2640060 
 
 3.7878 ' 236.313 
 
 I2O 
 
 341.06 
 
 312.8713 
 
 "85.9633 
 
 873.0920 
 
 .2742161 
 
 36475 
 
 2.27.560 
 
 I2 5 
 
 344-13 
 
 3i6-o433 
 
 1186.8996 
 
 870.8563 
 
 .2842203 3-5 l8 4 
 
 219.506 
 
 130 
 
 347-" 
 
 319.1238 
 
 1187.8085 
 
 868.6847 
 
 .2941889 3-3992 
 
 212.068 
 
 T 35 
 
 35- 2 
 
 322.1355 
 
 1188.6961 
 
 866.5606 
 
 .3040635 
 
 3.2881 
 
 205.181 
 
CHAPTER IV. 
 
 MATERIALS. 
 
 1. Relative Value of Materials for Boiler-construction. In selecting mate- 
 rials for boiler-construction it is necessary to consider their strength under different con- 
 ditions of stress, form, and temperature ; their adaptability for being manufactured into 
 the required forms ; their behavior when exposed to heat, their cost, their durability, 
 and their weight. 
 
 I. Regarding the strength of materials, it is not sufficient to consider the tensile or 
 compressive strain which a small bar or block of regular section can bear in the testing- 
 machine ; in boilers the materials are chiefly employed in the form of thin plates, pre- 
 senting either flat surfaces of great extent or curves and angles of every form, and from 
 the manner of their connection and the conditions obtaining in the use of boilers they 
 are exposed to a great variety of irregular strains. The vast amount of potential energy 
 stored up in the steam and heated water of a boiler demands the employment of a mate- 
 rial possessing a high degree of toughness, in order to mitigate the disastrous conse- 
 quences of a rupture ; for, while a tough material simply tears, brittleness would allow 
 large pieces to be detached and hurled with great violence. 
 
 In this connection it must be observed that most metals, like copper, wrought-iron, 
 and steel, possessing a high degree of toughness, may become brittle by drawing, roll- 
 ing, and hammering, in consequence of a structural change undergone during such pro- 
 cess ; the original toughness can be restored, however, by annealing. "The strains 
 produced by unequal expansion, in consequence of the great difference of temperature 
 existing at different parts of the boiler, require special attention. 
 
 II. The fitness of different materials for casting, rolling, and welding, as well as their 
 behavior when subjected to reheating for bending, to punching, drilling, and riveting, 
 require consideration. It is very important that the material should be of a reliable 
 character, sound throughout, and of uniform strength. 
 
 III. The thermal conductivity of different materials, as well as their strength at the 
 temperatures obtaining in steam boilers, varies greatly. 
 
 72 
 
SEC. 2. 
 
 MATERIALS. 73 
 
 IV. The cost of materials is not measured merely by the market price of the raw 
 pigs or ingots, sheets, and bars, but it is enhanced, for different materials in a widely 
 different degree, by the difficulties encountered during the processes which convert 
 them into constituent parts of a boiler. The final cost of a boiler is modified consider- 
 ably by its durability as well as by the value of the old material. 
 
 V. Durability, the power of resisting the destructive effects of the various stresses 
 and chemical agencies to which boilers are exposed, is of importance not only in so far 
 as it affects their cost but also their efficiency ; especially in the boilers of war-vessels 
 that have to maintain their efficiency while engaged, for many months continuously, on 
 distant stations without adequate facilities for repairs. 
 
 VI. The problem of producing the maximum effect with the minimum weight of 
 material is of increasing importance in modern marine engineering, especially for war- 
 vessels. 
 
 There is no substance that will pre-eminently satisfy all the foregoing requirements, 
 and the selection of the proper material for different parts of a boiler has to be governed 
 by experience as to their special fitness, and by the exercise of judgment as to the rela- 
 tive value of their properties in each case. Since wrought-iron has been for many years 
 the principal material employed in boiler-construction, it is convenient to use it as a 
 standard for measuring the relative fitness of other metals for this purpose. 
 
 2. Copper. Copper was largely used for the shell of boilers when steam of compa- 
 ratively low pressure was employed. Wilson says that the advantages possessed by 
 copper for boiler-making consist "in the uniformity and homogeneity of its texture, in 
 its freedom from lamination and blisters, and in its general trustworthy character when 
 well selected ; in the manner in which it resists the tenacious adhesion of most kinds of 
 incrustation ; in its great ductility and malleability, which render it capable of being 
 worked with great ease and of bearing sudden as well as oft-repeated racking strains ; 
 in its being a better conductor of heat, which not only tends to give it a higher evapora- 
 tive power under favorable conditions, but also enables it to last longer when exposed 
 to a fierce wasting heat in a boiler-furnace." An explosion generally results in a simple 
 tearing, and not in a violent projection of fragments ; small leaks are easily calked ; 
 finally, as old material, copper is worth about two-thirds as much as when new. 
 
 On the other hand, copper deteriorates more rapidly than iron when coal containing 
 sulphur or ammonia is burned in contact with it ; its strength is inferior to that of iron 
 nearly in the proportion of three to five at ordinary temperatures, but decreases rapidly 
 with each increase of temperature. Experiments made by a committee of the Franklin 
 Institute in 1837 developed the fact that at a temperature of 550 it loses one-fourth 
 
74 STEAM BOILERS. CHAP. IV. 
 
 of its tenacity at ordinary temperatures, at 817 precisely one-half, and at 1,000 two- 
 thirds of its strength are destroyed. Its first cost is very high, being nearly five times 
 that of iron. (See also chapter i.) 
 
 At present the use of copper in boilers is almost entirely limited to the crown-sheets 
 of English locomotives, and, in marine boilers, to superheating-tubes, to feed, steam, 
 blow, and escape pipes, and similar appendages. 
 
 3. Composition. Composition is a general term for certain alloys of copper and 
 tin or zinc, mixed in various proportions ; they are harder and generally more tenacious 
 than copper, and less subject to oxidation ; they are more fusible than copper, make 
 better castings, and are more easily worked with cutting tools, while at the same time 
 their cost is less. 
 
 Brass is an alloy of copper and zinc. The quantity of copper varies from 60 to 92 
 per cent. ; the best proportion for fine yellow brass appears to be, copper two parts and 
 zinc one part. The addition of a little lead from f to 3 per cent. causes the brass to 
 be more ductile and better adapted for turning in a lathe, but reduces the tenacity of 
 the metal ; a larger addition of lead renders the metal brittle. Instead of lead tin is 
 sometimes added, which renders the metal more homogeneous and increases its fluidity 
 without impairing its tenacity. Brass is more malleable than copper when cold, but 
 cannot be forged at a red heat, on account of the low melting-point of zinc. It is very 
 extensively used for castings and for brazed and drawn pipes and tubes. 
 
 Drawn seamless boiler-tubes have almost entirely stiperseded iron tubes in the ves- 
 sels of the United States Navy. Their thermal conductivity is greater than that of iron 
 in the proportion of 1.4 to 1, and their superior ductility makes it possible to render their 
 joints tight without straining the metal to an undue degree ; at the same time brass 
 tubes can be made about 20 per cent, thinner than iron ones, while their freedom from 
 corrosion gives them a reliability and durability which far outweigh their original in- 
 creased cost over that of iron. 
 
 Bronze. Such parts of the machinery of United States naval vessels as require a 
 high degree of tenacity combined with freedom from corrosion and lightness are made of 
 a composition consisting of 88 parts of copper, 10 parts of tin, and 2 parts of zinc. 
 
 By adding a small portion of dry phosphorus to these alloys in the crucible, before 
 running them into the mould, they are rendered so liquid that very thin and sound 
 castings may be obtained, while at the same time the hardness and the tenacity of the 
 metal are considerably increased. 
 
 Where expense is a secondary consideration, and where it is specially desirable to 
 avoid corrosion and lessen weight, the use of composition is extended to val\;e-cham- 
 
SEC. 4. MATERIALS. 75 
 
 bers, pipes, screw-swivels for long braces intended to be removable, hinges of connec- 
 tion-doors, manhole-plates, etc. When exposed to a temperature approaching a red 
 heat this metal loses all tenacity and crumbles to pieces. 
 
 4. Tenacity of Metals at various Temperatures (abridged from an article in 
 the Engineer, Oct. 5, 1877). The accompanying table gives the results of a series of 
 experiments made in Portsmouth Dockyard to ascertain the loss of strength and duc- 
 tility in gun-metal compositions when raised in temperature. The object was to see 
 whether gun-metal would be more or less suitable than cast-iron for stop and safety 
 valve boxes, steam-pipe connections, fastenings, etc., etc., which might be subjected to 
 high temperatures, either from superheated steam or proximity to hot uptakes or fun- 
 nels. The result shows that it is desirable to make further investigations. The speci- 
 mens and dies for griping them were heated in an oil bath, and the temperatures 
 recorded are those of the oil. In the case of gun-metal three or more tests were made 
 at each temperature, the table giving the mean, except when there were defects in the 
 metal. All the composition specimens were run in a horizontal position with a head of 
 2 inches, excepting those in columns 1 and 2. Those in No. 2 were stronger at atmos- 
 pheric temperature than No. 1, and they suffered sooner by increases of temperature. 
 All the varieties of gun-metal suffer gradual but not serious loss of strength and duc- 
 tility up to a certain temperature, at which, within a few degrees, a great change takes 
 place ; the strength falls to about one-half, and the ductility is wholly gone. Above 
 this point, up to 500, there is little if any further loss of strength. The precise tem- 
 perature at which the change and loss of strength take place, although uniform in the 
 specimens cast from the same pot, varies about 100 in the same composition at different 
 temperatures or with some varying conditions in the foundry process. In No. 1 series 
 this temperature was about 370, and in No. 2 a little over 250. The cause is not 
 known, but the fact is certain and important. Phosphor-bronze, the only metal in the 
 series which, from its strength and hardness, could be used as a substitute, was less 
 affected by temperature, and at 500 retains more than two-thirds of its strength and 
 one-third its ductility ; but the difference arising from variations in the process of cast- 
 ing or quality of material used should be tested, also whether the other compositions 
 may be hardened without loss of strength. Eolled Muntz metal and copper are satis- 
 factory up to 500, and may safely be used as securing-bolts. Wrought-irons increase 
 in strength up to 500, but lose slightly in ductility up to 300, where an increase begins 
 and continues up to 500, where it is still less than at the ordinary temperature of the 
 atmosphere. The strength of Landore steel is not affected by temperature up to 500, 
 but its ductility is reduced more than one-half. 
 
76 
 
 STEAM BOILERS. 
 
 CHAP. IV. 
 
 8 
 
 10 
 
 O 
 H 
 
 Bi 
 
 M 
 P. 
 
 ti S 
 
 s ^ 
 
 hJ W 
 
 M 
 
 < s 
 
 EH b 
 
 S 
 
 O 
 
 2 
 
 Jl 
 
 S-j-Jf 
 
 ^a^ ' 
 
 1 
 
 ii! 
 
 R 
 
 1J 
 
 Ss^j 
 
 jlj 
 
 4 4 i 
 
 f 
 
 f 
 
 111 
 
 e | $ S, 
 
 If! 11? 
 
 1 4 
 
 ; 
 
 *****;& 
 
 3 
 I 
 
 . 8 
 
 O -- J3 
 
 U H fu 
 
 MMllllliit 
 
 'XjlJIJOtlQ 
 
 I 
 
 *S 
 
 & 
 
 u P 
 
 h 
 
 'g if S 
 
 S S S 4 
 
 Xjt|tjonci 
 
 B i . 
 
 8 s ff i 
 
 if R j _ 
 
 a' s s s s s * ' '2 '= 
 
 4 8 & 8 a a 
 
 fe rt a M| - | " M 
 
 I 
 
 i 
 
 I 
 
 3 
 
 I 
 
 .S 8 
 
 S-g 
 
 3 
 
 rt C 
 
SKC. 5. MATERIALS. 77 
 
 5. Cast-iron. Cast-iron is used for grate-bars, ashpans, furnace and uptake doors, 
 manhole and handhole plates, valve-chambers, steam, feed, blow, and dry pipes, and 
 other boiler appendages. Its low first cost and the ease with which it can be worked 
 give it great advantages for such purposes ; but the uncertainty of strength caused by 
 defective moulding, its brittleness and low tensile strength, render it unfit for extensive 
 use for parts of the boiler proper. Its unyielding nature unfits it specially for parts 
 subjected to unequal expansion from differences of temperature. Several kinds of 
 modern sectional boilers consist, however, almost entirely of small cast-iron spheres or 
 tubes. 
 
 6. Wrought-iron. Wrought-iron has been for many years the material most ex- 
 tensively used in boiler-making. Wilson sums Tip the advantages possessed by it for 
 this purpose in the following words : "The great tensile strength of good wrought-iron, 
 together with its ductility, power of bearing sudden and trying strains, and general 
 trustworthy nature, its moderate facilities for working, the ease with which it can be 
 welded, riveted, patched, or mended, its moderate first cost compared with copper, are 
 all important advantages which contribute to its value." 
 
 It enters into the construction of boilers as plates, varying from \" to 1 J* in thickness, 
 as rivets, round bar-iron and flat iron, J and angle iron, and drawn seamless and lap- 
 welded tubes. In order to reduce the weight of boilers as much as possible wrought- 
 iron is frequently used instead of cast-iron for furnace, ashpit, and uptake doors, for 
 ashpans, manhole-covers, and sometimes for grate-bars. 
 
 The wrought-iron used in the construction of the boiler proper should be of the best 
 quality; it must possess not only great tensile strength but a high degree of duc- 
 tility, and it must bear, without receiving serious injury, the severe treatment to which 
 it is subjected by repeated heating, welding, hammering, punching, and drilling. The 
 quality of wrought-iron depends upon the chemical composition of the cast-iron or the 
 ore from which it has been made, and of the fuel used during the process of conver- 
 sion ; on the thoroughness with which the slag, cinders, and other deleterious sub- 
 stances have been removed ; and on the general care and labor expended on it during 
 the processes of refining, squeezing, hammering, and rolling. The most common defect 
 of boiler-plates arises from the imperfect welding of the several layers of metal forming 
 the plate, owing to the interposition of sand or cinders ; the laminations and blisters 
 produced in this manner are sometimes difficult to detect, but will appear sooner or 
 later when the plate is exposed to the intense heat of the furnace. The presence of a 
 small trace of sulphur produces hot-shortness in iron, which means that it will not 
 work advantageously and becomes brittle when red-hot, but is strong and pliable when 
 
78 STEAM BOILERS. CHAP. IV. 
 
 cold. On the other hand, the presence of phosphor and silicon produces cold-short- 
 ness, which means that the iron will not stand bending, twisting, or punching near the 
 edges when cold. An increase in the percentage of carbon renders the iron harder, less 
 ductile, and more difficult to weld. 
 
 During the process of rolling into bars or plates iron assumes a fibrous texture, and 
 boiler-plates, as a rule, are stronger when the tension takes place in the direction of the 
 fibre than when it is applied at right angles to the fibre. Increase of temperature does 
 not diminish the strength of iron till it reaches about 400. 
 
 7. Brands of Plate-iron used in Boiler-making. The use of mineral fuel is 
 generally dispensed with in the manufacture of boiler-iron, and wood charcoal is em- 
 ployed instead, in order to avoid the admixture of deleterious foreign substances. Such 
 iron is designated by the name of "charcoal-Iron." 
 
 The following are the various brands of American plate-iron commonly used for ma- 
 rine boilers, arranged in the order of their excellence and cost : Charcoal No. 1 Iron 
 (C. No. 1) will bear 40,000 Ibs. tensile strain in the direction of the fibre ; it is pretty 
 hard and is never flanged ; it is often used for the interior parts of a boiler, but breaks 
 or furrows easily when exposed to bending strains. Another brand (C. No. 1 R. H.\ 
 Charcoal No. 1 Reheated- Iron, is a very durable iron for fire-boxes, since, on account 
 of its hardness, it resists the oxidizing influence of the flame ; but it breaks easily under 
 alternating bending strains. 
 
 Charcoal Hammered No. 1 Shell-iron (C. H. No. I S.), though not necessarily ham- 
 mered, has been worked more thoroughly than the previous brands before it is rolled 
 into plates. It will bear in the testing-machine from 50,000 Ibs. to 54,000 Ibs. of tensile 
 strain per square inch in the direction of the fibre, and from 34,000 to 44,000 Ibs. across 
 the fibre. It is a rather hard iron and cannot be flanged ; at least it should not be bent 
 along the fibre, but always across the fibre and with a pretty large radius. It is used 
 especially for the outside shell of boilers. 
 
 C. II. No. I F. (Flange} Iron is a soft material, which can be flanged in every direc- 
 tion. It will bear from 50,000 to 54,000 Ibs. of tensile strain per square inch along the 
 fibre, and the best qualities do not show a much smaller strength when the strain is 
 applied across the fibre. 
 
 C. H. No. 1 F. B. (Fire-box) Iron is a harder quality, designed to withstand the de- 
 structive effect of the impinging flame ; it will generally bear flanging. There is, how- 
 ever, another quality, marked C. H. No. I F. F. B. (Flange Fire-box) Iron, or Extra 
 Fire-box or Excelsior Fire-box Iron, which is generally used for flanged fire-box 
 plates. 
 
SEC. 8. MATERIALS. 79 
 
 There are special brands, as Sligo, N. P. U. Iron (made at the Lukens Rolling-mills, 
 Coatesville, Chester Co., Pa.), Eureka (manufactured by C. E. Pennock & Co., Valley 
 Iron-works, Coatesville, Pa.), and Pine Iron, which command high prices, and are 
 specially fitted to withstand the destructive effect of heat or to be worked into the most 
 difficult forms. 
 
 These " charcoal-irons' 1 '' are manufactured principally in Pennsylvania. 
 
 The greatest width of boiler-plates made at different establishments varies between 
 48 inches and 96 inches, and the extreme weight of the trimmed plate between 1,000 
 and 1,800 Ibs. The length of the plate varies frequently with the width ; with some 
 manufacturers it is the rule to deduct one inch in width for each additional foot in 
 length : for instance, if the dimensions of the widest plates are 76" x 76", narrower 
 plates would have the following dimensions : 75" X 88", 74" X 100", 73" x 112", etc. 
 
 The best English boiler-iron is the Yorkshire iron, which seems to owe its superi- 
 ority partly to the fact that the coal used in the reduction of the ore and the manufac- 
 ture of the plates is remarkably free from sulphur and phosphorus. Wilson says: 
 " The most prominent makers of boiler-plates are the so-called ' Best Yorkshire ' houses 
 (viz., The Low Moor Iron- works, near Bradford ; Taylor Brothers & Co., Leeds ; Bowl- 
 ing Iron Co., near Bradford ; Farnley Iron Co., near Leeds ; S. T. Cooper & Co., Leeds ; 
 and The Monk Bridge Iron Co., Leeds; the other firms who make only 'Best York- 
 shire ' iron do not roll plates), who only turn out one class of iron, and that the very 
 best, if we except some of the Swedish and Russian brands." 
 
 8. Steel. The use of steel in boiler-construction has attracted much attention of 
 late, especially since the introduction of improved processes of manufacture, which 
 have resulted in the production of much more uniform qualities of steel than were 
 formerly obtainable, and in a reduction of its cost which allows it to compete in many 
 cases with wrought-iron in economical respects. 
 
 The steel used for boiler-plates is sometimes crucible steel, but is generally manu- 
 factured by the Bessemer or the Siemens-Martin process. It must be of a very mild 
 quality, containing about one-quarter or one-third per cent, of carbon. Steels contain- 
 ing a larger amount of carbon possess greater tensile strength, but are brittle, hard to 
 work, and untrustworthy for use in boiler-making. It is very important that steel 
 boiler-plates do not temper when suddenly cooled down from a red heat. Steel loses 
 this characteristic of taking a temper when the percentage of carbon is reduced below 
 a certain amount, while at the same time its welding qualities are improved. In fact, 
 the milder kinds of steel possess all the good qualities of wrought-iron, only in a 
 higher degree, and differ from it principally by their greater purity and their perfectly 
 
80 STEAM BOILERS. CHAP. IV. 
 
 homogeneous structure. While iron boiler-plate is produced by the piling of a number 
 of slabs, separately manufactured and undergoing repeated reheatings, steel plates are 
 rolled from single ingots, often at one heat. Each charge of the hearth or converter 
 produces from five to eight tons of metal, which can be carefully examined and tested, 
 and carburized or decarburized and refined to any desired degree, before it is finally 
 drawn off. In this manner a uniformity in the character of the metal is produced 
 which is unattainable in the manufacture of iron. Wilson says that, "in order to 
 ensure freedom from brittleness, from 33 to 36 tons per square inch appears to be the 
 maximum tensile strength that can be allowed. Steel plates of this strength can be 
 made sufficiently tough and ductile to render them safe and also tolerably easily 
 worked." But the surveyors to Lloyd's Registry recommend that steel used in boiler- 
 making should have an ultimate tensile strength of not less than 26 tons and not more 
 than 30 tons per square inch of section, and the same limits have been adopted by the 
 English Admiralty. 
 
 Even the mildest steels seem to be affected differently from iron by the severe strains 
 of hammering, punching, and shearing ; to restore to the plate its original character it 
 is considered necessary to anneal it. 
 
 With regard to the effect of corrosion on steel the opinions are much divided ; while 
 the greater density and homogeneity of steel would lead one to suppose that it would 
 suffer less than iron in that respect, it is asserted that steel boilers have shown an un- 
 usual amount of corrosion after short use. This question can only be settled after fur- 
 ther experience with the qualities of steel now in use. In this connection it must be 
 remembered that, since thin plates deteriorate more rapidly relatively than thicker 
 plates under the influence of corrosion, the reduction in the thickness of steel boiler- 
 plates cannot be made in the ratio of the increased strength of steel over iron. 
 
 The following extracts are taken from a report of the surveyors to Lloyd's Registry, 
 made in the early part of 1878, which discusses the advantages and difficulties incident 
 to the use of steel for boilers in a very thorough manner : 
 
 "The methods adopted in the manufacture of mild steel by the Bessemer and Sie- 
 mens-Martin processes are such as practically to ensure the production of a material 
 perfectly reliable so far as regards its uniformity in tensile strength and its power to 
 withstand certain bending tests. The limit of elasticity of this material bears about the 
 same relation to its ultimate strength as in ordinary wrought-iron ; but the elongation 
 or stretch under stresses proportional to the ultimate strengths is greater with steel than 
 with iron a fact which should not be lost sight of in forming an estimation of the 
 strength of boilers. At first sight it recommends itself by its tensile strength, mallea- 
 
SEC. 8. MATERIALS. 81 
 
 bility, and ductility, and also by its freedom from laminations and blisters, as eminently 
 suited for the construction not only of the shells and stays but also of the furnaces and 
 combustion-chambers of marine boilers. But while as a material it possesses in a special 
 degree these high qualities, it is found that they become seriously impaired by its being 
 subjected to the processes usually occurring in boiler-making, and it is necessary to 
 exercise the greatest care in the working of it to ensure these qualities being retained in 
 the structure, while in some instances it is even requisite to subsequently employ spe- 
 cial means in order to restore them. The simple process of shearing affects to some 
 extent the tensile strength of the plate operated upon, and a considerable portion of its 
 strength is lost by punching. It is contended, however, and indeed it may be said to 
 be placed beyond doubt, that the loss thus occasioned is fully recovered by the plates 
 being annealed after they have been sheared or punched, and it is the practice at almost 
 all the steel manufactories we have visited to anneal every plate after it is sheared and 
 before being sent out of the works. It may be well here to remark that this annealing 
 is not, as is frequently supposed, a process of some difficulty, requiring great care and 
 considerable time in the operation. It consists simply of heating the plates to a low 
 red heat which allows the particles that have been strained or disturbed by the work- 
 ing of the material to resume their normal condition and then cooling them uniformly. 
 
 "At some works the holes are drilled, and at others punched ; but in all cases in 
 which they are punched the plates are afterwards annealed. It is the practice at all the 
 works, and it is considered by these firms to be of the utmost importance to perform 
 the operation of flanging in only one heat, if possible, and to have the plates uniformly 
 heated throughout ; but when this is not practicable, and the operation is extended over 
 several heats the plates being heated locally piece by piece, as is usually done in flang- 
 ing iron plates care is taken that the plates so flanged are afterwards annealed. 
 
 "The opinions of those who maybe regarded as authorities on the matter differ 
 greatly with regard to the limits of tensile strength which should be adopted for this 
 material when intended for boiler-making purposes. . . . Taking into consideration 
 the fact that the milder material is more easily worked and less likely to be injured by 
 careless manipulation than that of higher strength and more brittle nature, ... we are 
 of the opinion that it would not be prudent, at least until further experience is gained, 
 to raise the limits ; while at the same time it might be advisable to recommend that 
 plates used in the construction of the furnaces and combustion-chambers be specified to 
 withstand not more than from 26 to 28 tons per square inch. 
 
 " With regard to the question of steel rivets, it has been conclusively shown, by the 
 results of some of the experiments made on the Tyne, that they may be used with as 
 
82 STEAM BOILERS. CHAP. IV. 
 
 much reliability as steel plates, but that, like the latter, they require greater care and 
 discrimination to be exercised in the working of them than those made of iron. In the 
 opinion of Dr. Siemens and other authorities the material of which the rivets are made 
 should be very mild steel, the tensile strength not exceeding 26 tons per square inch. 
 It is also needful to heat them uniformly throughout their entire length, and not to 
 raise the points to a higher temperature than the heads, as is the usual practice with 
 iron rivets, and they should not be heated beyond a bright-red heat. When these pre- 
 cautions are taken steel rivets will be found to resist steady strains and also jars and 
 concussions much better than iron rivets. 
 
 " In conclusion, we would remark that in the construction of steel boilers greater 
 care and attention must be exercised with the workmanship than is required in the case 
 of iron boilers ; and the difference between the two materials, and the consequent diffe- 
 rent manipulation required in each case, must be realized not only by the manager, but 
 by the workman who will have to use the material ; for if steel plates are drifted 
 heavily and knocked about as iron plates usually are in boiler-making, the material will 
 be injured. We may expect to see steel boilers extensively used in preference to those 
 made of iron, where lightness or increased pressure is an object, while if they are made 
 with the care which this material requires, and eventually prove to be as durable as iron 
 boilers, it will be a question whether a considerable reduction in the factor of safety 
 may not be found quite compatible with perfect safety and efficiency . 
 
 "After having given all the circumstances in connection with the whole matter our 
 most careful consideration, we would respectfully submit that, where it is proposed to 
 use steel boilers in vessels intended for classification in this society's Register Book, the 
 requirements of the case would be met by sanctioning a reduction from the scantlings 
 prescribed by the rules for iron boilers, in the shell-plates and stays to the extent of 25 
 per cent., and in the flat plates not subject to the action of heat to the extent of 12 per 
 cent., under the following conditions : 
 
 "I. The material to have an ultimate tensile strength of not less than 26 tons nor 
 more than 30 tons per square inch of section. 
 
 "II. A strip cut from every plate used in the construction of the furnaces and com- 
 bustion-chambers, and strips cut from other plates taken indiscriminately, heated uni- 
 formly to a low cherry-red heat, and quenched in water of 82 Fahr., must stand bending 
 to a curve of which the inner radius is not greater than one and one-half times the thick- 
 ness of the plate tested. 
 
 "III. All holes to be drilled, or, if they are punched, the plates to be afterwards 
 annealed. 
 
SBC. 8. 
 
 MATERIALS. 
 
 83 
 
 " IV. All plates, except those that are in compression, that are dished or flanged, or 
 in any way worked in the fire, to be annealed after the operations are completed. 
 
 " V. The boilers upon completion to be tested in the presence of one of the society's 
 engineer-surveyors to not less than twice the working pressure." 
 
 TABLE X. 
 EXHIBITING CERTAIN PHYSICAL AND MECHANICAL PROPERTIES OF VARIOUS METALS. 
 
 Material. 
 
 Weight in 
 pounds 
 per cubic foot. 
 
 Expansion of 
 unity of 
 length from 
 32 to 2 1 2 
 Fahr. 
 
 Ultimate ten- 
 sile strength. 
 
 Elongation in 
 per cent, of 
 
 length. 
 
 Shearing 
 strength. 
 
 Crushing 
 strength. 
 
 
 \\\ R. 
 
 .001 1 R. 
 
 
 
 
 
 
 
 
 20 ooo W 
 
 
 
 
 
 
 
 16,500 H. 
 
 
 27,700 R. 
 
 112 OOO R. 
 
 
 480 R. 
 
 .0012 R. 
 
 
 
 
 
 Bar-iron average .. . 
 
 481 K. 
 
 
 en COO 
 
 
 50 ooo R 
 
 38 ooo R 
 
 
 
 
 51 ooo 
 
 
 
 
 
 481 K. 
 
 
 
 
 
 
 " in direction of fibre . 
 
 
 
 ec \'i'i K 
 
 I'l A 
 
 
 
 " across the fibre 
 
 
 
 50,462 K. 
 
 60 
 
 
 
 Yorkshire bar-iron 
 
 484 K. 
 
 
 61 0";=; K. 
 
 2J. J.6 
 
 
 
 
 
 
 61 469 K 
 
 2O 46 
 
 *lfi Tl.l B 
 
 
 
 486 K. 
 
 
 d.6 7J.7 K 
 
 22 1 
 
 
 
 Steel 
 
 490 R. 
 
 .0012 R. 
 
 
 
 
 
 Landore steel boiler-plate 
 
 
 
 63 ooo 
 
 2 A 2$ 
 
 
 
 Fagersta steel boiler-pi., unannealed. 
 annealed. . . 
 
 
 
 
 
 51, 528 K. 
 
 J.7 7^O K 
 
 14-05 K. 
 16 ni K 
 
 
 
 
 
 " hammered bars, soft.. 
 Cast-steel rivets, English (Moss & 
 Gamble). . , 
 Krupp's bolt-steel 
 
 Copper 
 
 
 
 .00184 R. 
 
 61,312 K. 
 
 107,286 K. 
 92,015 K. 
 
 16.5 K. 
 
 12.4 K. 
 15.3 K. 
 
 
 
 121,333 K. 
 
 Sheets 
 
 S<1Q R. 
 
 
 20 ooo 
 
 
 
 
 
 
 
 -i-i OOO A 
 
 
 
 
 Cast bars 
 
 548.6 T.B. 
 
 
 27, 800 T.B. 
 
 6.47 T.B. 
 
 
 42 oooT B 
 
 Bronze 
 
 
 .00181 R. 
 
 
 
 
 
 96.06C, 3-76T 
 
 539.7 T.B. 
 
 
 32,000 T.B. 
 
 14.29 T.B. 
 
 
 42 ooo T B 
 
 92.nC, 7.80 T 
 90.27C, 9-58T 
 8 7 .i 5 C, I2. 73 T 
 80.95 C, 18.847 
 8 parts copper, i part tin 
 
 542.5 T.B. 
 540.9 T.B. 
 541-7 T.B. 
 545-6 T.B. 
 524 R. 
 
 
 
 28,540 T.B. 
 26, 860 T.B. 
 29,430 T.B. 
 32,980 T.B. 
 36,000 A. 
 
 5-53 T.B. 
 3.66 T.B. 
 3-33 T.B. 
 0.40 T.B. 
 
 
 
 42,000 T.B. 
 38,000 T.B. 
 53,000 T.B. 
 78,000 T.B. 
 
 Brass (rolled) : 3 copper, 2 zinc 
 
 525.4 
 
 
 49,280 A. 
 
 
 
 
 
 486.5 
 
 .00216 R. 
 
 28,900 A. 
 
 
 
 
 
 
 
 
 
 
 
 A = Anderson ; H = Hodgkinson ; K = Kirkaldy ; R = Rankine ; W = Wade ; B = Brunei ; T.B = U. S. 
 Test Board ; * = one experiment. 
 
STEAM BOILERS. 
 
 CHAP. IV. 
 
 TABLE XI. 
 WEIGHT OF WROUGHT-IRON PLATES AND BARS. (Square and Round.) 
 
 TRAUTWINE. 
 
 Thickness or Diameter. 
 Inches. 
 
 Weight of plates per square 
 foot, in pounds. 
 
 Weight per foot of square 
 bars, in pounds. 
 
 Weight per foot of round 
 oars, in pounds. 
 
 i 
 
 tt 
 
 IO.IO 
 
 n-37 
 12.63 
 13.89 
 
 .2105 
 .2665 
 .3290 
 .3980 
 
 1653 
 .2093 
 
 -2583 
 .3126 
 
 1 
 U 
 
 15.16 
 16.42 
 17.68 
 18.95 
 
 -4736 
 -5558 
 .6446 
 .7400 
 
 .3720 
 4365 
 5063 
 -5813 
 
 ! 
 
 H 
 
 20. 21 
 22.73 
 25.26 
 27.79 
 
 .8420 
 1. 066 
 1.316 
 
 1-592 
 
 .6613 
 
 .8370 
 
 1-033 
 1.250 
 
 if 
 
 A 
 
 3-3I 
 32.84 
 
 35-37 
 37-89 
 
 I.8 9S 
 2.223 
 
 2-579 
 2.960 
 
 1.488 
 1.746 
 2.025 
 
 2-325 
 
 i 
 
 TV 
 
 J 
 
 40.42 
 42.94 
 
 45-47 
 48.00 
 
 3-368 
 3-803 
 4.263 
 
 4-75 
 
 2.645 
 2.986 
 
 3-348 
 
 3-73 
 
 TV 
 1 
 TV 
 
 5-5 2 
 53-5 
 55-57 
 5810 
 
 5-263 
 5.802 
 6.368 
 6.960 
 
 4.133 
 
 4-557 
 5.001 
 5.466 
 
 1 
 3 
 
 60.63 
 65.68 
 70-73 
 75-78 
 
 7-578 
 8.893 
 10.31 
 11.84 
 
 5-952 
 6.985 
 8.101 
 9.300 
 
 2 
 i 
 
 1 
 
 80.83 
 85.89 
 9-94 
 95-99 
 
 13-47 
 15.21 
 
 17-05 
 19.00 
 
 10.58 
 "95 
 
 13-39 
 14.92 
 
 1 
 * 
 
 IOI.OO 
 
 106.10 
 
 III. 20 
 
 116.20 
 
 21.05 
 23.21 
 
 25-47 
 27-84 
 
 i6-53 
 18.23 
 20. 01 
 21.87 
 
 3 
 
 121.30 
 
 30-31 
 
 23.81 
 
MATERIALS. 
 
 85 
 
 TABLE XII. 
 WEIGHT OF FLAT BAR-IRON PER FOOT. 
 
 
 Thickness in inches. 
 
 Width in 
 
 
 inches. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 TV 
 
 i 
 
 A 
 
 i 
 
 A 
 
 1 
 
 A 
 
 i 
 
 * 
 
 i 
 
 i 
 
 i 
 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 Ibs. 
 
 
 .21 
 
 42 
 
 63 
 
 .84 
 
 5 
 
 1.26 
 
 i-47 
 
 1.68 
 
 2. II 
 
 2-53 
 
 2-95 
 
 3-37 
 
 -J- 
 
 .24 
 
 47 
 
 7i 
 
 95 
 
 .18 
 
 1.42 
 
 1.66 
 
 1.90 
 
 2-37 
 
 2.84 
 
 3-32 
 
 3-79 
 
 J 
 
 .26 
 
 53 
 
 79 
 
 05 
 
 32 
 
 1.58 
 
 1.84 
 
 2.I.I 
 
 2.63 
 
 3.16 
 
 3.68 
 
 4.21 
 
 1 
 
 .29 
 
 5 
 
 .87 
 
 .16 
 
 45 
 
 i-74 
 
 2.03 
 
 2.32 
 
 2.89 
 
 3-47 
 
 4-05 
 
 4-63 
 
 
 
 32 
 
 63 
 
 95 
 
 .26 
 
 58 
 
 1.90 
 
 2.21 
 
 2-53 
 
 3.16 
 
 3-79 
 
 4.42 
 
 5-05 
 
 | 
 
 34 
 
 .68 
 
 3 
 
 37 
 
 7i 
 
 2.05 
 
 2-39 
 
 2-74 
 
 3-42 
 
 4.11 
 
 4-79 
 
 5-47 
 
 1 
 
 37 
 
 74 
 
 .11 
 
 47 
 
 .84 
 
 2.21 
 
 2.58 
 
 2-95 
 
 3.68 
 
 4.42 
 
 5-i6 
 
 5-89 
 
 I* 
 
 .40 
 
 79 
 
 .18 
 
 58 
 
 97 
 
 2-37 
 
 2.76 
 
 3.16 
 
 3-95 
 
 4-74 
 
 5-53 
 
 6.32 
 
 2 
 
 .42 
 
 .84 
 
 .26 
 
 .68 
 
 2. II 
 
 2-53 
 
 2.95 
 
 3-37 
 
 4.21 
 
 5-5 
 
 5-8 9 
 
 6-74 
 
 4 
 
 45 
 
 .90 
 
 34 
 
 79 
 
 2.24 
 
 2.68 
 
 3- J 3 
 
 3-58 
 
 4-47 
 
 5-37 
 
 6.26 
 
 7.16 
 
 2 
 
 47 
 
 95 
 
 .42 
 
 .90 
 
 2-37 
 
 2.84 
 
 3-32 
 
 3-79 
 
 4-74 
 
 5.68 
 
 6.83 
 
 758 
 
 H 
 
 So 
 
 I.OO 
 
 5 
 
 2.OO 
 
 2.50 
 
 3.00 
 
 3-5 
 
 4.00 
 
 5.00 
 
 6.00 
 
 7.00 
 
 8.00 
 
 ,i 
 
 53 
 
 1-05 
 
 58 
 
 2. II 
 
 2.63 
 
 3-i6 
 
 3.68 
 
 4.21 
 
 5.26 
 
 6.32 
 
 7-37 
 
 8-42 
 
 4 
 
 55 
 
 I. II 
 
 .66 
 
 2.21 
 
 2.76 
 
 3-32 
 
 3-87 
 
 4.42 
 
 5-53 
 
 6.63 
 
 7-74 
 
 8.84 
 
 i 
 
 58 
 
 1.16 
 
 74 
 
 2.32 
 
 2.89 
 
 3-47 
 
 4-5 
 
 4-63 
 
 5-79 
 
 6 -95 
 
 8.10 
 
 9.26 
 
 4 
 
 .61 
 
 1. 21 
 
 .82 
 
 2.42 
 
 3-3 
 
 3.63 
 
 4.24 
 
 4.84 
 
 6.05 
 
 7.26 
 
 8.47 
 
 9.68 
 
 3 
 
 63 
 
 1.26 
 
 1.90 
 
 2-53 
 
 3.16 
 
 3-79 
 
 4.42 
 
 5-5 
 
 6.32 
 
 7,S8 
 
 8.84 
 
 IO.IO 
 
 3f 
 
 .68 
 
 i-37 
 
 2.05 
 
 2-74 
 
 3-42 
 
 4.11 
 
 4-79 
 
 5-47 
 
 6.84 
 
 8.21 
 
 9-58 
 
 10.95 
 
 4 
 
 74 
 
 i-47 
 
 2.21 
 
 2-95 
 
 3.68 
 
 4.42 
 
 S-i6 
 
 5-89 
 
 7-37 
 
 8.84 
 
 10.32 
 
 11.79 
 
 3* 
 
 79 
 
 1.58 
 
 2-37 
 
 3.16 
 
 3-95 
 
 4-74 
 
 5-53 
 
 6.32 
 
 7.89 
 
 9-47 
 
 11.05 
 
 12.63 
 
 4 
 
 .84 
 
 1.68 
 
 2-53 
 
 3-37 
 
 4.21 
 
 5-5 
 
 5-89 
 
 6.74 
 
 8.42 
 
 IO.IO 
 
 11.79 
 
 13-47 
 
 4* 
 
 .90 
 
 1.79 
 
 2.68 
 
 3.S8 
 
 4-47 
 
 5-37 
 
 6.26 
 
 7.16 
 
 8-95 
 
 10.74 
 
 12-53 
 
 I4-3 1 
 
 4 
 
 95 
 
 1.90 
 
 2.84 
 
 3-79 
 
 4-74 
 
 5.68 
 
 6.63 
 
 7-.S8 
 
 9-47 
 
 11.38 
 
 13.26 
 
 15.16 
 
 4 
 
 I.OO 
 
 2.00 
 
 3.00 
 
 4.00 
 
 5.00 
 
 6 oo 
 
 7.00 
 
 8.00 
 
 10.00 
 
 I2.0O 
 
 14.00 
 
 16.00 
 
 5 
 
 1-05 
 
 2. II 
 
 3.16 
 
 4.21 
 
 5.26 
 
 6.32 
 
 7-37 
 
 8.42 
 
 10-53 
 
 12.63 
 
 14-74 
 
 16.84 
 
 5f 
 
 i. ii 
 
 2.21 
 
 3-32 
 
 4-42 
 
 5-53 
 
 6.63 
 
 7-74 
 
 8.84 
 
 11.05 
 
 13.26 
 
 15-47 
 
 17.68 
 
 Si 
 
 1.16 
 
 2.32 
 
 3-47 
 
 4-63 
 
 5-79 
 
 6-95 
 
 8.10 
 
 9.26 
 
 11.58 
 
 1389 
 
 16.21 
 
 18.52 
 
 si 
 
 1. 21 
 
 2.42 
 
 3- 6 3 
 
 4-84 
 
 6.05 
 
 7.26 
 
 8.47 
 
 9.68 
 
 12.10 
 
 14-53 
 
 16.95 
 
 19-37 
 
 6 
 
 1.26 
 
 2 -53 
 
 3-79 
 
 5-5 
 
 6.32 
 
 7.58 
 
 8.84 
 
 10.10 
 
 12.63 
 
 I5.l6 
 
 17.68 
 
 20. 21 
 
STEAM BOILERS. 
 
 CHAP. IV. 
 
 TABLE XIII. 
 
 WEIGHT OF SHEET AND PLATE IRON. 
 
 Thickness. 
 
 Weight. 
 
 Thickness. 
 
 Weight. 
 
 B. W. gauge. 
 
 Fractions of an 
 inch. 
 
 Pounds per square 
 foot. 
 
 B. W. gauge. 
 
 Fractions of an 
 inch. 
 
 Pounds per square 
 foot. 
 
 36 
 
 .004 
 
 .126 
 
 II 
 
 .120 
 
 4.48 
 
 35 
 
 .005 
 
 .202 
 
 
 i or. 125 
 
 5^54 
 
 34 
 
 .007 
 
 .283 
 
 10 
 
 134 
 
 5.426 
 
 33 
 
 .008 
 
 .322 
 
 9 
 
 .148 
 
 5-98 
 
 3 2 
 
 .009 
 
 3 6 4 
 
 
 & or. 1562 
 
 6.305 
 
 3 1 
 
 .OIO 
 
 .405 
 
 8 
 
 .165 
 
 6.605 
 
 3 
 
 .OI2 
 
 485 
 
 7 
 
 .180 
 
 7.27 
 
 29 
 
 .013 
 
 .526 
 
 
 -fa or. 1875 
 
 7.578 
 
 28 
 
 .OI4 
 
 595 
 
 6 
 
 .203 
 
 8.005 
 
 27 
 
 .Ol6 
 
 .677 
 
 
 -fa or. 2187 
 
 8-79 
 
 26 
 
 .018 
 
 755 
 
 5 
 
 .22 
 
 8.912 
 
 25 
 
 .020 
 
 .811 
 
 4 
 
 .238 
 
 9.62 
 
 24 
 
 .022 
 
 .912 
 
 
 ior.25 
 
 10.09 
 
 23 
 
 .025 
 
 .018 
 
 3 
 
 259 
 
 10.37 
 
 22 
 
 .028 
 
 137 
 
 
 -^ or .2812 
 
 11.38 
 
 
 S^or -03125 
 
 259 
 
 2 
 
 .284 
 
 11-525 
 
 21 
 
 .032 
 
 3i 
 
 I 
 
 3 
 
 12.15 
 
 20 
 
 35 
 
 .416 
 
 
 A or .3125 
 
 12.58 
 
 r 9 
 
 .042 
 
 6 95 
 
 O 
 
 340 
 
 13-75 
 
 18 
 
 .049 
 
 075 
 
 
 tt or .3437 
 
 13.875 
 
 17 
 
 .058 
 
 2-35 
 
 
 1 or .375 
 
 15.10 
 
 
 TV or .0625 
 
 2.518 
 
 oo 
 
 .380 
 
 15.26 
 
 16 
 
 .065 
 
 2.637 
 
 
 || or .4062 
 
 16.34 
 
 15 
 
 .072 
 
 2.92 
 
 ooo 
 
 425 
 
 17.125 
 
 U 
 
 .083 
 
 3-35 
 
 
 A or -4375 
 
 17-65 
 
 
 T&or .0937 
 
 3-78 
 
 oooo 
 
 454 
 
 18.30 
 
 13 
 
 9S 
 
 3-85 
 
 
 H or .4687 
 
 18.90 
 
 12 
 
 .109 
 
 4-4 
 
 ooooo 
 
 ior. 5 o 
 
 20.00 
 
 For steel 'plates multiply the tabular number of any size by i.oi. 
 
MATERIALS. 
 
 87 
 
 TABLE XIV. 
 WEIGHT OF WROUGHT ANGLE-IRON. (C. W. &> H. W. Middleton, Philadelphia) 
 
 Dimension! 
 
 in inches. 
 
 Weight per foot 
 
 Width. 
 
 Thickness. 
 
 in pounds. 
 
 fx | 
 
 i 
 
 t 
 
 i X i 
 
 4x4 
 if xij 
 
 4x 4 
 
 if X if 
 
 A 
 
 i to A 
 
 A 
 A to i 
 
 i to ^ 
 
 i to 4 
 ii to if 
 
 4 tO 2 
 if tO 2 
 
 2} to 4^ 
 
 2X2 
 2X 4 
 2* X 2^ 
 2\ X 2 
 2f X 2f 
 
 i- to i 
 
 A to f 
 
 A to I 
 
 A- to | 
 A to | 
 
 3 to 5 
 2ito 3 
 4 to 6 
 5 to 10 
 6 to 8 
 
 3 X2 
 3 X 2\ 
 3X3 
 
 3iX 2 
 3iX 3 
 
 f to* 
 
 A to i 
 
 1 to i 
 1 to A 
 
 4 to 5 
 5i to 8i 
 7^ to 10 
 6 to 8 
 7f to ii 
 
 3*X3i 
 4X3 
 4 X 3i 
 4X4 
 4iX 3 
 
 i to | 
 
 1 
 I A 
 
 1 to | 
 
 f_^A 
 
 9 to 12 
 8 to 12 
 9 to 13 
 ii to 15 
 9 to 13 
 
 5X3 
 
 
 13 
 
 5 X 3i 
 
 
 n4 
 
 5X4 
 
 
 144 to 20 
 
 6 X 3 i 
 6X4 
 
 I 
 
 i to | 
 
 22 
 
 14 to 27 
 
 6^X4 
 
 A to | 
 
 I4| tO 20| 
 
STEAM BOILERS. 
 
 CHAP. IV. 
 
 TABLE XV. 
 WEIGHT OF WROUGHT T- IRON - (C- W. <&* H. W. Middleton, Philadelphia.} 
 
 Dimensions in inches. 
 
 Weight per foot in pounds. 
 
 i X i 
 
 ftoi 
 
 itX it 
 
 l tO 2 
 
 i X it 
 
 3 
 
 ii x 4 
 
 2 tO Z\ 
 
 if X it 
 
 3 
 
 if X if 
 
 2f 
 
 2X2 
 
 3: 
 
 2| X 2; 
 
 3: 
 
 2\ X 2 
 
 5-' 
 
 *\ X 2J- 
 
 6; 
 
 2| X I: 
 
 6; 
 
 3X3 
 
 6^ to 9 
 
 3 X 3^ 
 
 i of to ii 
 
 3X4 
 
 II tO 12 
 
 3T X 3 
 
 10 to ii 
 
 3^ X 3^ 
 
 10 to 12 
 
 3t X 4 
 
 i3i 
 
 4 X 2 
 
 6| to 8 
 
 4X4 
 
 12 
 
 4 X 4^ 
 
 13 
 
 4X5 
 
 14 
 
 5 X 2^ 
 
 10 tO 12 
 
 5 X 3 
 
 II tO 14 
 
MATERIALS. 
 
 TABLE XVI. 
 
 WROUGHT-IRON BOLTS WITH SQUARE HEADS AND NUTS. 
 
 (C. W. & H. W. Middlcton, Philadelphia.) 
 Weight in pounds of 100 of the enumerated sizes. 
 
 Lengths. 
 
 Jin. 
 
 1 in. 
 
 i in. 
 
 |in. 
 
 Jin. 
 
 |in. 
 
 i in. 
 
 i|in. 
 
 Inches. 
 ii 
 
 
 10 62 
 
 21 87 
 
 2Q ^T 
 
 
 
 
 
 1 ? 
 i 
 
 
 
 'O- / 
 
 2*5 06 
 
 oy-j* 
 
 41.^8 
 
 
 
 
 
 X T 
 
 .44 
 
 12 "IS 
 
 
 AC 6O 
 
 71 62 
 
 
 
 
 24- 
 
 75 
 
 
 28 62 
 
 ^o-"y 
 
 AO. CO 
 
 />>"* 
 76. 
 
 
 
 
 2 f 
 2! 
 
 34 
 5 .07 
 
 IA 60 
 
 2n CO 
 
 ^y-o" 
 
 CI 21 
 
 /" 
 
 7O.7C 
 
 
 
 
 *X 
 2$ 
 
 y/ 
 
 6 t\o 
 
 I 6 A.*J 
 
 ^y-j" 
 
 7T l6 
 
 C1 
 
 8l. 
 
 
 
 
 * 
 
 3 
 3i 
 
 4 
 
 4 
 
 5 
 5* 
 6 
 64 
 
 
 
 17.87 
 18.94 
 20.59 
 
 21.69 
 
 23.62 
 25.81 
 26.87 
 
 3 2 -44 
 39-75 
 42.50 
 44.87 
 48.81 
 5I-38 
 S3-3i 
 e6 87 
 
 oo- 
 56. 
 63.12 
 
 74.87 
 79.62 
 
 83- 
 87.88 
 
 92-38 
 96 88 
 
 85-38 
 
 93-44 
 108.12 
 113.12 
 
 122. 
 128.62 
 
 I 3'-75 
 
 I 7O c6 
 
 127.25 
 140.56 
 
 148-37 
 158.76 
 167.25 
 174.88 
 204.25 
 
 214 60 
 
 228 
 
 2 39 
 
 250 
 
 261 
 272 
 281 
 
 
 
 296 
 310 
 324 
 338 
 35 2 
 166 
 
 7 
 
 
 
 CO 12 
 
 00.87 
 
 ^oy';)" 
 
 228.44 
 
 2O4 
 
 "?7O 
 
 7 1 
 
 
 
 oy- 1 * 
 61 87 
 
 yy*i 
 
 IOC 7C 
 
 *w*y 
 
 150 88 
 
 21C 11 
 
 oc 
 
 184 
 
 / S 
 
 8 
 
 
 
 64.44 
 
 iu 5-/ j 
 IOQ CO 
 
 IC7.I2 
 
 *o!3-J L 
 210.88 
 
 <J U J 
 
 116 
 
 108 
 
 
 
 
 7O.CO 
 
 118 12 
 
 160.62 
 
 258 12 
 
 ^?8 
 
 426 
 
 10 
 
 
 
 77 
 
 i '8 ii 
 
 184. 
 
 276 18 
 
 oo" 
 
 16O 
 
 4C.4 
 
 ii 
 
 
 
 82 88 
 
 1 16 IO 
 
 TOC.T 7 
 
 
 -?82 
 
 482 
 
 12 
 
 
 
 86 -17 
 
 1-1/1.87 
 
 1 yo* 1 o 
 
 2OO. 7C 
 
 ^yo- u y 
 1 1 r 04 
 
 0^ 
 AO4 
 
 Clo 
 
 H 
 
 
 
 02 
 
 I CC CO 
 
 ^"y'/D 
 
 iic 81 
 
 426 
 
 J 1 " 
 
 Cl8 
 
 id. 
 
 
 
 y- 
 
 O7 71 
 
 1 J J'j" 
 
 161 c8 
 
 *aro/ 
 
 2 ?7 CO 
 
 oo5' 01 
 
 ici 88 
 
 AA& 
 
 JO" 
 
 c66 
 
 I d 
 
 
 
 y/-/ 3 
 
 IO1 2C 
 
 iu o-3" 
 
 I7O 7C 
 
 O/'J 1 -' 
 
 OO 1 - 00 
 1OT 7C 
 
 47O 
 
 0"" 
 CO4 
 
 
 
 
 
 1 /"-/3 
 
 
 oy 1 -/^ 
 
 
 oyt 
 
90 
 
 STEAM BOILERS. 
 
 CHAP. IV. 
 
 TABLE XVII. 
 
 STANDARD SIZES OF WASHERS. 
 
 (C. W. & H. W. Middleton, Philadelphia) 
 
 Size of bolt. 
 
 Diameter of washer. 
 
 Size of hole. 
 
 Thickness, wire gauge. 
 
 Number in too Ibs. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 No. 
 
 
 i 
 
 * 
 
 A 
 
 16 
 
 29,300 
 
 A 
 
 $ 
 
 -I 
 
 16 
 
 18,000 
 
 i 
 
 I 
 
 A 
 
 14 
 
 7,600 
 
 * 
 
 i 
 
 A 
 
 ii 
 
 2,100 
 
 
 ,1 
 
 1 
 
 ii 
 
 2,180 
 
 
 x i 
 
 H 
 
 ii 
 
 2,35 
 
 
 if 
 
 if 
 
 ii 
 
 i, 680 
 
 
 2 
 
 ft 
 
 10 
 
 1,140 
 
 i 
 
 t 
 
 ,i 
 
 8 
 
 580 
 
 
 2j 
 
 1} 
 
 8 
 
 47 
 
 
 3 
 
 
 
 7 
 
 360 
 
 
 3 
 
 X * 
 
 6 
 
 360 
 
or TMC 
 UNIVERSITY 
 
 OF 
 
 MATERIALS. 
 
 91 
 
 TABLE XVIII. 
 
 SHOWING NUMBER OF "BURDEN'S" RIVETS IN roo POUNDS. 
 
 Length ir 
 
 inches. 
 
 % inch diameter. 
 
 % inch diameter. 
 
 11-16 inch diameter. 
 
 % inch diameter. 
 
 i 
 
 S 
 
 1,092 
 1,027 
 
 665 
 
 597 
 
 
 
 
 
 
 
 940 
 
 538 
 
 45 
 
 .... 
 
 
 
 840 
 
 5'2 
 
 415 
 
 .... 
 
 
 
 797 
 
 487 
 
 389 
 
 356 
 
 
 
 760 
 
 460 
 
 370 
 
 329 
 
 
 
 730 
 
 440 
 
 357 
 
 28o 
 
 
 
 711 
 
 420 
 
 34 
 
 271 
 
 
 
 693 
 
 39 
 
 325 
 
 262 
 
 
 
 648 
 
 375 
 
 3'2 
 
 257 
 
 2 
 
 
 6c8 
 
 360 
 
 297 
 
 243 
 
 2- 
 
 
 573 
 
 354 
 
 289 
 
 237 
 
 2; 
 
 
 555 
 
 347 
 
 280 
 
 232 
 
 2; 
 
 
 525 
 
 335 
 
 260 
 
 220 
 
 2i 
 
 
 500 
 
 32 
 
 242 
 
 208 
 
 3 
 
 
 460 
 
 290 
 
 224 
 
 I 97 
 
 z\ 
 
 t 
 
 433 
 
 267 
 
 212 
 
 180 
 
 3: 
 
 
 
 4i3 
 
 248 
 
 201 
 
 169 
 
 IJ 
 
 i 
 
 395 
 
 241 
 
 I 9 2 
 
 160 
 
 4 
 
 
 
 23 
 
 184 
 
 158 
 
 4i 
 
 r 
 
 .... 
 
 220 
 
 '77 
 
 '5 
 
 4: 
 
 
 
 
 210 
 
 '7i 
 
 146 
 
 j 
 
 
 .... 
 
 2OO 
 
 1 66 
 
 138 
 
 5 
 
 
 .... 
 
 190 
 
 161 
 
 '35 
 
 5^ 
 
 t 
 
 .... 
 
 180 
 
 156 
 
 130 
 
 SI 
 
 ! 
 
 .... 
 
 172 
 
 164 
 
 '5? 
 '45 
 
 124 
 
 120 
 
 6 
 
 
 .... 
 
 '57 
 
 140 
 
 "5 
 
 <H 
 fr 
 
 I 
 
 .... 
 
 '5 
 146 
 
 138 
 '34 
 
 in 
 107 
 
 61 
 
 \ 
 
 .... 
 
 X 43 
 
 129 
 
 104 
 
 7 
 
 
 
 
 140 
 
 125 
 
 too 
 
 L 
 
 
 
 
 
 
CHAPTER Y. 
 
 TESTING THE MATERIALS. 
 
 1. General Character of Tests. The tests to be applied to materials used in 
 the construction of boilers are of three different kinds viz., first, the investigation of 
 the mechanical properties of the materials as developed in the testing-machine ; second- 
 ly, the trial of the quality of the materials affecting their fitness to undergo the vari- 
 ous mechanical processes employed in boiler-making ; thirdly, the search for imperfec- 
 tions in the structure of the materials developed in the process of manufacture. 
 
 2. United States Laws and Regulations regarding the Tests of Boiler- 
 plates. The ' Revised Statutes of the United States ' prescribe the following tests 
 for boiler-plates: "Section 4430. Every iron or steel plate used in the construction of 
 steamboat boilers, and which shall be subject to a tensile strain, shall be inspected in 
 such manner as shall be prescribed by the Board of Supervising Inspectors and ap- 
 proved by the Secretary of the Treasury, so as to enable the inspectors to ascertain its 
 tensile strength, homogeneousness, toughness, and ability to withstand the effect of 
 repeated heating and cooling ; and no iron or steel plate shall be used in the construc- 
 tion of such boilers which has not been inspected and approved under those rules. 
 
 "Section 4431. Every plate of boiler-iron or steel made for use in the construction of 
 steamboat boilers shall be distinctly and permanently stamped by the manufacturer 
 thereof, and, if practicable, in such places that the marks shall be left visible when such 
 plates are worked into boilers, with the name of the manufacturer, the place where 
 manufactured, and the number of pounds tensile strain it will bear to the sectional 
 square inch." 
 
 ' The General Rules and Regulations prescribed by the Board of Supervising In- 
 spectors of Steam-vessels,' January, 1879, contain the following instructions regarding 
 the tests to be applied to boiler-plates : 
 
 "JRuleS. Every iron or steel plate intended for the construction of boilers to be 
 used on steam-vessels shall be stamped by the manufacturer in the following manner 
 viz., at the diagonal corners, at a distance of about four inches from the edges, and also 
 at or near the centre of the plate, with the name of the manufacturer, the place where 
 
SEC. 3. TESTING THE MATERIALS. 93 
 
 manufactured, and the number of pounds tensile strain it will bear to the sectional 
 square inch. 
 
 " When a sheet of boiler-iron is found by the inspector with one or more stamps 
 upon the same, the inspector shall in every such case be governed by and rate the ten- 
 sile strain of iron in accordance with the lowest stamp found upon the same. 
 
 " Rule 4. The manner of inspecting and testing boiler-plates intended to be used in 
 the construction of marine boilers, by the United States inspectors, shall be as follows, 
 viz.: 
 
 "The inspector shall visit places where marine boilers are being constructed, as 
 often as possible, for the purpose of ascertaining and making a record of the stamps 
 upon the material, its thickness, and other qualities. To ascertain the tensile strain 
 [strength] of the plates the inspector shall cause a piece to be taken from each sheet 
 to be tested, the area of which shall equal one-quarter of one square inch on all 
 iron ^ inch thick and under ; and on all iron over ^ inch thick the area shall equal 
 the square of its thickness, and the force at which the piece can be parted in the direc- 
 tion of the fibre or grain, represented in pounds avoirdupois the former multiplied by 
 four, the latter in proportion to the ratio of its area shall be deemed the tensile strain 
 per square inch of the plate from which the sample was taken ; and should the tensile 
 strength ascertained by the test equal that marked on the plates from which the test- 
 pieces were taken, the said plates must be allowed to be used in the construction of ma- 
 rine boilers, provided always that the said plates possess the other qualities required by 
 law viz., homogeneousness, toughness, and ability to withstand the effect of repeated 
 heating and cooling ; but should these tests prove the marks on the said plates to be 
 overstamped, the lots from which the test-plates were taken must be rejected as failing 
 to have the strength stamped thereon. But nothing herein shall be so construed as to 
 prevent the manufacturers from restamping such iron at the lowest tensile strain indi- 
 cated by the samples, provided such restamping is done previous to the use of the plates 
 in the manufacture of marine boilers. 
 
 "To ascertain the ductility and other lawful qualities, iron under 45,000 pounds 
 should show a contraction of area of 12 per cent. ; 45,000 pounds and under 50,000 
 should show 15 per cent. ; 50,000 pounds tensile strength and over should show 25 per 
 cent, at point of rapture." 
 
 3. The Rodman Testing-machine. The Eodman Testing-machine at the Ord- 
 nance Department of the Washington Navy-Yard has been used in making several of 
 the experiments recorded in this volume. An illustration of it is given on Plate I. It 
 consists essentially of a system of three levers, A, B, and C, the arms of which have 
 
94 STEAM BOILERS. CHAP. V. 
 
 the following proportions : A = 10 : 1 ; B = 20:l; = 10:1; consequently the total 
 leverage of the machine is 2,000 : 1. The machine is capable of measuring a stress of 
 one hundred thousand pounds, and is arranged to determine tensile, compressive, trans- 
 verse, and torsional stresses, as well as indenting forces and bursting pressures applied 
 to hollow vessels. 
 
 The main lever A acts directly upon the specimen by means of the stirrup D. The 
 fulcrum F of the main lever is supported by a pair of heavy cast-iron stanchions, S S, 
 secured by bolts to the bed-frame T. The stirrup D bears upon the lever A by means 
 of the knife-edge a, and has at its lower end a shackle, d, to which the various con- 
 trivances for transmitting the stress to the specimen iinder test are attached. A strap, 
 E, connects the main lever with the intermediate lever B, which, in turn, is connected 
 by means of the adjustable rod G with the small lever C. On the latter are placed the 
 weights producing the stress to which the specimen is to be exposed. The lever C is 
 graduated from to 10 and carries a small sliding-weight, p ; by moving this weight 
 one division of the scale out from the fulcrum a stress of 100 pounds is exerted on the 
 specimen attached to the stirrup D. At the point c, corresponding to the tenth divi- 
 sion, the lever C carries a rod, H, on which additional weights can be placed ; of these 
 there are two kinds viz., ten representing a stress of 1,000 Ibs. each, and nine repre- 
 senting a stress of 10,000 Ibs. each at the point a. 
 
 In order to counterbalance exactly the weight of the levers and their attachments 
 the weights V and V can be moved in or out on the rods passing through their centre ; 
 V being attached directly to the lever C, while V acts by means of a lever and strap on 
 the intermediate lever B. In order to reduce the friction to a minimum case-hardened 
 steel knife-edges are used for the bearings of the levers and their connections. 
 
 It is evident that any movement of the point d on the lever A, caused by the elonga- 
 tion, compression, or deflection of the specimen, is increased two-thousandfold at the 
 point c of the lever C ; it is, however, essential to maintain the levers in an exactly 
 horizontal and parallel position, since otherwise the leverage of the machine would be 
 altered, and its balance would be disturbed by shifting the centre of gravity of the 
 levers. On this account the fulcrums /and ./' of the levers B and C are attached to 
 the sliding-blocks g and g', which can be moved up or down by suitable gearing. The 
 hand-crank N turns, through the intervention of a train of wheels and pinions, the 
 horizontal wheel O, which, in turn, gives motion to the screw R, the central boss of 
 the wheel forming a nut for the screw ; the lower end of the latter is attached to the 
 sliding-block g, which is guided between the frames Z Z. To this block is attached 
 the rack 7i, gearing with a pinion fastened on the shaft J which carries, on its other end. 
 
SEC. 3. 
 
 TESTING THE MATERIALS. 95 
 
 a corresponding pinion, i ', gearing with the rack Ji ' ; the latter is attached to the slid- 
 ing-block g', which is guided between brackets forming part of the stanchions S S. 
 In this manner the vertical motion of the fulcrum / of the lever B is communicated to 
 the fulcrum /' of the lever C, and the horizontally and perfect parallelism of the two 
 levers can be maintained. No means are provided to keep the main lever in a horizon- 
 tal position during the operation of the machine, but the error arising from this source 
 is trifling, since this lever has always only a comparatively slight motion. In order 
 that the disturbance of the balance of the machine in consequence of the motion of the 
 centre of gravity of the lever A be as small as possible, the lever has received such a 
 shape that its centre of gravity lies near its centre of motion. 
 
 The large screw U, placed directly under the point a, passes through the bed-frame 
 and can be adjusted to a proper height by means of a hand-wheel. Its upper end is 
 provided with a suitable arrangement for holding the specimen subjected to a tensile 
 strain or the straps used in applying compressive and indenting forces. 
 
 When transverse strains are to be produced the two movable pedestals, X X, at- 
 tached to the bed-frame are used. They are fitted with knife-edges, against which the 
 test-bar bears, the strain being applied, by a strap connected to the stirrup D, to the 
 lower side of the bar. 
 
 Torsional strains are measured by means of the lever M, which works between the 
 two pillow-blocks, Y Y, secured to the bed-frame. The hollow axis of this lever receives 
 the specimen, the projecting ends of which are firmly secured by cotters to the blocks. 
 The lever M is connected by means of a chain and link to the intermediate lever B. A 
 graduated arc is attached to the pillow-block Y, and a pointer connected with the axis 
 of the lever indicates the number of degrees through which the specimen has been 
 twisted. 
 
 Before beginning an experiment the machine is balanced by a proper adjustment 
 of the weights V and V, so that the weights placed on the lever C measure accurately 
 the stress applied to the specimen. When a specimen has been put in place to test its 
 tensile or compressive strength care is taken that all lost motion is taken up by 
 moving the screw U up or down, while the lever C is kept in a horizontal position. 
 Then the hand-crank is turned with regularity in a direction which raises the screw R, 
 and through it the fulcrums / and /', while at the same time the sliding- weight p is 
 moved so as to keep the lever C in an exactly horizontal position. Whenever the 
 sliding- weight p registers a strain of 1,000 Ibs., one of the large weights, P, represent- 
 ing an equivalent strain, is substituted for it on the rod H, and the sliding- weight is 
 moved back to the zero-point. In the same manner a weight, P', representing a strain 
 
96 
 
 STEAM BOILERS. 
 
 CHAP. V. 
 
 of 10,000 Ibs., is placed on the rod H in the place of ten weights of 1,000 Ibs. each. By 
 repeating these operations the effects of increments of 100 Ibs. each, tip to the limit of 
 100,000 Ibs., can be observed and measured. 
 
 4. Form and Dimensions of Test-specimens. The sectional area of specimens 
 that are to be tested is limited by the strength of the machine. It is very convenient 
 to give such dimensions to the specimen that its sectional area is a simple multiple or 
 fraction of a square inch. 
 
 It is found that long bars generally stretch more, in proportion to their length, than 
 short ones of the same material, sectional area, and form. In order to secure greater 
 uniformity in the conditions of tests the French Government established the rule that 
 the length of all specimens subjected to tests of tensile strength should be 200 milli- 
 metres (7.875 inches) between the shoulders ; and, following this precedent, the English 
 Admiralty has adopted a standard length of 8 inches. 
 
 Fig. 1. 
 
 
 
 
 Figures 1 and 2 show the forms usually given to specimens. The middle portion, 
 which is the part to be tested, must be carefully turned or planed, as the case may be, 
 to the exact dimensions required, so as to have the sides perfectly parallel. It is 
 essential that the strain be applied exactly in the direction of the axis of the specimen. 
 In order to measure the elongation of the specimen under strain its length is divided 
 by lines or centre-punch marks into equal spaces. Kirkaldy described on the surface 
 of some specimens a network of diagonal lines or circles, the distortion of which repre- 
 sented the relative elongation of different portions of the specimen. Specimens of very 
 thin plates may have side-plates riveted to their ends to give them longer bearings on 
 the pins of the shackles. Specimens subjected to a crushing strain are seldom made 
 longer than twice their diameter. 
 
 The influence of length and form on the apparent strength of test-specimens is illus- 
 trated by " the results of a series of trials of samples of steel cut from the same plate 
 and reduced in width as shown" (see figures 3 and 4), given by Keed in ' Shipbuilding 
 in Iron and Steel,' pp. 401, 402. "Eight cases were taken, each case being tested by 
 three experiments for each mode of reduction. The breaking section was in each case 
 
SBC. 5. TESTING THE MATERIALS. 97 
 
 reduced to a breadth of one inch, the reduction from the extreme breadth of the 
 samples (2J inches) being made in one set of plates by circular arcs, and in the other 
 
 Fig. 3. 
 
 Fig. 4. 
 
 set by parallel reductions. The lengths of the reduced parts were varied from 8 inches 
 to 1 inch by successive deductions of 1 inch, and the sketches (figures 3 and 4) show 
 the extreme cases of the longest and shortest reductions respectively. The forms of the 
 circular arcs are shown in ticked lines, and those of the parallel reduction by drawn 
 lines." . . . "Throughout these experiments, which were conducted with extreme 
 care, the same material broke at a less strain when trimmed down to a parallel breadth 
 for a considerable distance than when reduced to the same breadth at one place only. 
 By comparing the samples reduced to a parallel breadth for a length of 8 inches with 
 those similarly reduced for a length of 1 inch only, it will be seen that the apparent 
 strength rose from an average of 19J tons to an average of 21f tons ; or if we compare 
 the former with the case of the shortest circular reduction, we find it increasing from 
 19 J tons to an average of 23f tons." 
 
 Numerous experiments were also made by the United States Test Board to deter- 
 mine the proper form and dimensions of test-pieces. The results of these experiments 
 are summed up in the report of the Board in the following words : " In conclusion, 
 our results lead us to the decision that in testing [wrought] iron no test-piece should be 
 less than f inch diameter, as inaccuracy is more probable with a small than with a large 
 piece, and the errors are more increased by reduction to the square inch ; that the 
 length should not be less than four times the diameter in any case, and that with soft, 
 ductile metal five or six diameters would be preferable." 
 
 In preparing the specimens care must be taken that the metal does not receive in- 
 jury impairing its strength. On this account they should be cut from the plate by saw- 
 ing or drilling, and then filed down to the exact dimensions. In case punching is re- 
 sorted to it is necessary to leave a siirplus of metal for planing or filing down, in order 
 that the specimen may not be weakened by the injury which the metal receives in the 
 vicinity of a punched hole. 
 
 5. Effects produced by Stress. Under a gradually increasing tensile stress the 
 elongation of metals may be assumed to be in the direct ratio of the stress, within the 
 limit of elasticity, and whenever the stress is discontinued the specimen will resume 
 
98 STEAM BOILERS. CHAP. V. 
 
 very nearly its original dimensions. The elongation of wrought-iron is about -nnb 7 part 
 of its length for every ton of tensile stress per square inch of sectional area, and its limit 
 of elasticity lies between 8 and 12 tons of tensile strain per square inch. A bar which 
 has been strained beyond the limit of elasticity, and has in consequence taken a perma- 
 nent set after the stress had been removed, will show a much smaller rate of elongation 
 for all stresses less than that which produced the set. 
 
 Anderson says : "A bar of wrought-iron, when it leaves the rolls, is in a condition of 
 great restraint ; the exterior is not in perfect equilibrium with the interior of the bar, 
 which at the commencement of an experiment affects the elongation and permanent set. 
 The first effect of the application of a load is to liberate the constrained surface, and 
 true conditions on which to form an opinion do not exist until equilibrium is estab- 
 lished in the bar itself. Previous to that the result is deceptive ; hence the advantage 
 of carefully-turned specimens." 
 
 Soft and ductile materials are drawn out to a much smaller diameter at and near the 
 point of fracture before the final rupture takes place. Kirkaldy was the first to point 
 out the importance of calculating the ultimate breaking stress per square inch for the 
 contracted area at the point of fracture as a measure of the quality of the material. 
 
 Fibrous wrought-iron is stronger and stretches more in the direction of the fibres 
 than across them ; hence the importance of cutting test-specimens of boiler-plate in both 
 directions from the sheet. 
 
 The molecular changes of a ductile material, when strained, may be compared to 
 those occurring during the flow of liquids. 
 
 Anderson says: "When a bar is drawn out the principal flow of the apparently 
 solid metal is in the middle of the stream, and hence the peculiar sectional form which 
 is assumed either by a round or square bar, or one of any other shape, showing that 
 the farther the molecules of the material are removed from the centre of the flowing 
 current, so much the less are they affected by the influence of the general movement. 
 This unequal flowing of the molecules may partly account for the apparent weakness 
 of thin plates as compared with round bars of the same sectional area." 
 
 " With a flowing, malleable, or ductile metal the round bar, when under tension, is 
 drawn out to a small diameter uniformly all around, but the metal goes in the middle 
 chiefly, and the outside is shrivelled ; while with a rectangular or square bar the flat 
 surfaces are slightly hollowed to an extent proportionate to their distances from the 
 centre of the flow." 
 
 6. Experiments on the E.ffects of Hammering and Rolling 011 the 
 Strength of Bars. " Early in the course of the mechanical tests [of the United States 
 
SEC. 6. 
 
 TESTING THE MATERIALS. 
 
 Test Board] it became evident that, although, each set of nine bars (1 inch to 2 inches 
 diameter) from any maker was made of the same material and as uniformly as ordinary 
 processes would allow, yet there was a notable variation in the physical characteristics 
 of the different-sized bars. The tenacity, elastic limit, and ductility increased as the 
 diameter decreased. In fourteen sets of bars the strength per square inch of the 1-inch 
 over the 2-inch ran from 4,000 to 7,000 Ibs. ; and in bars known to have had uniform 
 treatment it averaged 5,600 Ibs. But the increase of strength was not uniform. In 
 eight sets of bars the strength fell off at the 1^-inch size. 
 
 " An investigation of the method of manufacture revealed the causes of these phe- 
 nomena. The piles from which the 2-inch, If-inch, If-inch, and If-inch bars were 
 rolled had the same cross-section, differing only in length. The piles from the 1^-inch, 
 If -inch, IJ-inch, and 1^-inch, and sometimes the 1-inch, were of the same area, al- 
 though smaller than the piles above mentioned. The areas of the piles remaining con- 
 stant with each set, while those of the bars decreased, the smaller bars received the most 
 work in the rolls. It was then found by numerous experiments that the tenacity and 
 elastic limit of the various bars of a set increased just in proportion to the decrease of 
 the percentage of the area of the bar to that of the pile. 
 
 TABLE XIX. 
 
 Size of bar. 
 
 Iron AT, showing decrease of strength by decrease of 
 reduction. 
 
 Iron /lr, showing uniformity of strength with uniformity 
 of reduction. 
 
 Area of bar in per 
 cent, of area 
 of pile. 
 
 Tensile strength. 
 
 Elastic limit. 
 
 Area of bar in per 
 cent, of area 
 of pile. 
 
 Tensile strength. 
 
 Elastic limit. 
 
 Inches. 
 
 Per cent. 
 
 Pounds per square 
 inch. 
 
 Pounds per square 
 inch. 
 
 Per cent. 
 
 Pounds per square 
 inch. 
 
 Pounds per square 
 inch. 
 
 
 (Pile 6" X 4%".) 
 
 
 
 
 
 
 2 
 
 4 
 if 
 
 if 
 
 11.36 
 10.22 
 8-90 
 7.68 
 
 51,848 
 54,034 
 55'OlS 
 
 5 6 -344 
 
 32,461 
 33.610 
 34-283 
 
 35>88 9 
 
 3-9 2 
 
 3-45 
 3-34 
 3-24 
 
 50,763 
 53-361 
 
 53.!54 
 53,329 
 
 33,258 
 35-032 
 35-323 
 33-520 
 
 
 (Pile 4" X 3%".) 
 
 
 
 
 
 
 0+-'*f-Qo)WSf-' 
 
 11.78 
 9.90 
 8.18 
 6.62 
 
 53-550 
 54.277 
 56,478 
 5 6 .543 
 
 34.690 
 33-622 
 
 33.251 
 32.267 
 
 i 
 
 3-2? 
 3-53 
 3-4i 
 3-3i 
 3,4 
 
 52,819 
 52,733 
 53-248 
 54,645 
 53,9 r 5 
 
 34,840 
 34,606 
 
 33-520 
 
 34-695 
 36,287 
 
100 
 
 STEAM BOILERS. 
 
 CHAP. V. 
 
 " In order to determine if the converse is true another set of experiments was under- 
 taken, and it proved that, by preserving a uniform proportion of bar to pile, all the bars 
 of the series have substantially the same strength per square inch. 
 
 "Table XIX. gives two typical examples, selected from the records of the Board. 
 That of iron N shows the effect of variation in the percentages of pile to bar ; that of 
 iron Fx the effect of uniformity." (A. L. Holley, ' The Strength of Wrought-iron as 
 affected by its Composition and by its Reduction in Rolling?} 
 
 In the Report of the Committee on Chain-cables, Malleable Iron, etc., of the 
 United States Test Board, it is stated that, although the strength of the entire bar was 
 increased by the extra work in rolling the bar from a larger pile, yet that of the core of 
 the iron was not, as shown by the following tests : 
 
 TENSILE STRENGTH AND ELASTIC LIMIT OF THREE BARS, IRON Fx, AS FOUND BY RUPTURE OF 
 
 ENTIRE BARS AND OF TURNED CYLINDERS. 
 
 Size. 
 Inches. 
 
 Area of entire bars in 
 per cent, of area of 
 piles. 
 
 Tensile strength. 
 
 Elastic limit. 
 
 Entire bar. 
 
 Cylinder. 
 
 Entire bar. 
 
 Cylinder. 
 
 
 
 Pounds per sq. inch. 
 
 Pounds per sq. iirch. 
 
 Pounds per sq inch. 
 
 Pounds per sq. inch. 
 
 2 
 
 3-93 
 
 52,011 
 
 45,9 6 4 
 
 34,702 
 
 31,830 
 
 '* 
 
 I 
 
 3-45 
 
 2.62 
 
 53,537 
 
 55,77 
 
 47,i 2 4 
 49,656 
 
 34,235 
 34,279 
 
 32,070 
 35,714 
 
 Plate II. contains the record of experiments conducted by Chief -Engineer Wm. H. 
 Shock, U. S. N., which are instructive in showing the effect of an increased amount of 
 hammering and rolling on the tenacity and ductility of a bar of wrought-iron. 
 
 The test-specimens were all cut from the same bar, which was of good marketable 
 American iron, 4 inches square. One end of the bar was drawn down by hammering to 
 a size of 2J inches square. The diagrams on Plate II. show from which particular por- 
 tion of the bar each specimen was cut. The specimens were carefully turned to the 
 exact size given on the plate, and were subjected to a tensile strain in the Rodman Test- 
 ing-machine represented on Plate I. 
 
 Comparing the results obtained with specimens cut from different portions of the 
 bar, we find : First, that the mean tensile strength of the eight specimens marked E, 
 which were cut from the sides of the bar, was 2. 13 per cent, greater than that of the four 
 specimens marked D, which were cut from the centre of the bar ; and that the mean 
 elongation of the former was 16.1 per cent, greater than that of the latter. Secondly, 
 
SBC. 7. TESTING THE MATERIALS. 101 
 
 that the mean tensile strength of the six specimens marked A, which were cut from the 
 reduced end of the bar, was 5.63 per cent, greater than that of the specimens marked 
 D ; and that the mean elongation of the former was 9.79 per cent, greater than that of 
 the latter. Thirdly, that the mean tensile strength of the eight specimens marked B 
 and C, which were cut across the fibre of the bar, was 40.85 per cent, less than that 
 of the specimens marked D ; and that they broke without showing any elongation. 
 Fourthly, that the several specimens marked B and C broke under widely different 
 strains : the lowest result obtained (specimen Bl) was 33.73 per cent, less than the mean 
 result obtained with the eight specimens of classes B and C ; and the highest result ob- 
 tained (specimen B2) was 35.51 per cent, greater than the mean result. Specimens Bl 
 and B2 were cut from immediately adjoining portions of the bar. The highest and low- 
 est breaking strains obtained with specimens of class D differ less than one per cent, 
 from the mean result ; and with specimens of class E this difference is only a fraction 
 more than one per cent. The difference of the results obtained with specimens of class 
 A is equally slight, with the exception of specimen A6, which broke under a strain 
 4.08 per cent, higher than the mean breaking strain of class A. The portion of the bar 
 from which A6 was cut had prpbably received a greater amount of hammering at a 
 lower temperature ; this assumption appears to be confirmed by the smaller amount of 
 elongation of this specimen before rupture occurred. 
 
 7. Appearance of Fractures. The fracture of iron and steel should present a 
 close, uniform grain of a bright gray color, or a silky fibre ; when of a dull, earthy as- 
 pect, with a loose texture, an unequal grain, or a blackish fibre, it indicates an inferior 
 quality or a badly-refined material. 
 
 Fractures of steel may be classified as granular, crystalline, and fibrous, and those 
 of wrought-iron as crystalline and fibrous, with intermediate gradations. Hard, un- 
 yielding materials always present a granular or crystalline fracture. 
 
 With respect to wrought-iron Kirkaldy deduces the following results from Ms expe- 
 riments : "1st. "Whenever wrought-iron breaks suddenly a crystalline appearance is 
 the invariable result ; when gradually, invariably a fibrous appearance. 2d. Whether, 
 on the one hand, it is finely or coarsely crystalline, or, on the other, the fibre be fine or 
 close, or coarse and open, depends upon the quality of the iron. 3d. When there is a 
 combination in the same bar or plate of two kinds the one harder or less ductile than 
 the other the appearance will be partly crystalline and partly fibrous. . . . 5th. The 
 relative qualities of various irons may be pretty accurately judged of by comparing their 
 fractures, provided they have all been treated in precisely the same way, and all bro- 
 ken under the same sort of strains similarly applied. 6th. By varying either the 
 
102 STEAM BOILERS. CHAP. V. 
 
 shape, the treatment, the kind of strain, or its application, pieces cut off the same bar 
 will be made to present vastly different appearances in some kinds of iron, whilst in 
 others little or no difference will result. . . . 
 
 "The appearance of the same bar may be completely changed from wholly fibrous to 
 wholly crystalline, . . . 1st, by altering the shape of the specimen so as to render it 
 more liable to snap ; 2d, by treatment making it harder ; and, 3d, by applying the strain 
 so suddenly as to render it more liable to snap from having less time to stretch. . . . 
 
 " In the case of the fibrous fracture the threads are drawn out and are viewed exter- 
 nally ; in the case of the crystalline fracture the threads in clusters are snapped across, 
 and are viewed internally or sectionally. . . , 
 
 "The conclusions respecting wrought-iron are equally appropriate to steel viz.: 
 Whenever rupture occurs slowly a silky fibrous, and when suddenly a granular, ap- 
 pearance are invariably the result ; both kinds varying in fineness according to quality. 
 The surface in the latter case is even, and always at right angles with the length ; in 
 the former angular and irregular in outline. The color is a light pearl gray, slightly 
 varying in shade with the quality ; the granular fractures are almost entirely free of 
 lustre, and consequently totally unlike the brilliant crystalline appearance of wrought- 
 iron." 
 
 8. Hot and Cold Forge-tests. The forge-tests applied to boiler-plates used in 
 the United States naval service vary according to the judgment of the superintending or 
 inspecting officers. The specifications generally require that the plates must be able to 
 bear the severest flanging-tests to which they may be subjected. These consist in turn- 
 ing flanges on two adjacent sides of a plate, thus forming a corner ; in cutting a hole in 
 the middle of a plate and enlarging it by turning a flange up around it ; in giving a 
 dished shape to the plate by forcing it, by means of a jack and a spherical die, into a cor- 
 responding form. The plates are further tested cold by punching holes near the edges, 
 and by bending them to angles of different degrees, according to the thickness of the 
 plates. They must bear these several tests without showing any signs of cracks or lami- 
 nations. 
 
 In the English service the iron supplied by the manufacturers has to iindergo the 
 tests described in section 11, chapter vii., which are carried out in the same manner, as 
 nearly as possible, in the various establishments, both private and public, under the su- 
 pervision of Admiralty officers. In applying the bending-tests it is prescribed that 
 plates, both hot and cold, should be tested on a cast-iron slab, having a fair surface, 
 with an edge at right angles, the corner being rounded off with a radius of inch. 
 The plate should be bent at a distance of from 3 to 6 inches from the edge. 
 
SEC. 9. TESTING THE MATERIALS. 103 
 
 Wilson says: "All plates of the very best quality, having a longitudinal tenacity 
 of 24 tons per square inch of section and an ultimate elongation of about 12 per cent., 
 and not exceeding one inch in thickness, should bend double along or across the fibre 
 when red-hot. 
 
 "For the cold forge-test plates of the very best quality ^ inch thick and under 
 should bend double without fracture. . . . 
 
 " The angle to which the plates can be bent without fracture will depend greatly 
 upon the skill of the smith who heats and operates upon them. A plate that will bear 
 the test with a number of sharp, light blows will often fail when a heavy hammer is 
 used. By striking the plate along its surface it can be successfully bent to a much 
 greater angle than when the blows are dealt perpendicularly to the surface. 
 
 "The plate will also stand the bending much better if it is performed uniformly 
 along its whole width. . . . The manner in which a plate will bear flanging out- 
 wardly, whereby the fibres are either stretched or separated, as the plate is flanged 
 across or along the grain, is generally considered the best test of its soundness and 
 quality. . . . 
 
 " Rivets and bars for boiler- work are seldom tested for their tensile strength, but 
 their quality is usually ascertained by the forge-tests. A good rivet, cold, will bend 
 double without fracture. The head of a good rivet should flatten out, by hammering 
 when hot, to about * thick, without fracture or fraying at the edge. A hot rivet-shank 
 or bar of iron, when flattened down to a thickness equal to about one-half its diameter, 
 should bear a punch driven through it without fracture at the hole." 
 
 9. Directions for testing Bar-iron. "Cut a notch on one side with a cold- 
 chisel, then bend the bar over the edge of an anvil at sharp angles. If the fracture ex- 
 hibits long, silky fibres of a leaden gray color, cohering together, and twisting or pull- 
 ing apart before breaking, it denotes tough, soft iron, easy to work and hard to break. 
 In general a short, blackish fibre indicates iron badly refined. A very fine, close grain 
 denotes a hard, steely iron, which is apt to be cold-short, but working easily when 
 heated, and making a good weld. Numerous cracks on the edges of the bar generally 
 indicate a hot-short iron, which cracks or breaks when punched or worked at a red 
 heat, and will not weld. Blisters, flaws, and cinder-holes are caused by imperfect 
 welding at a low heat, or by iron not being properly worked, and do not always indicate 
 inferior quality. 
 
 " To test iron when hot, draw a piece out, bend and twist it, split it, and turn back 
 the two parts to see if the split extends up ; finally weld it, and observe if cracks 
 or flaws weld easily. Good iron is frequently injured by being unskilfully worked ; 
 
104 STEAM BOILERS. CHAP. V. 
 
 defects caused by this may be in part remedied. If, for example, it has been injured 
 by cold-hammering moderate annealing heat will restore it." (King.} 
 
 10. Testing Steel Boiler-plates. For steel boiler-plates the following tests are 
 prescribed by the French Government: "Hot tests: These tests will be made with 
 sample plates of suitable dimensions, and consist in stamping a dished cavity, the side 
 of the plates preserving its original plane. The diameter of this cavity is to be equal to 
 forty times the thickness of the plate, and the depth will be ten times this thickness ; 
 the flat edge to be joined to the cavity by a curve the radius of which is not to be 
 greater than the thickness of the plate. Moreover, plates more than .197 inch will be 
 stamped with a flat-bottomed depression with square angles and straight sides, the 
 diameter of the bottom to be thirty times the thickness of the plate, and the depth ten 
 times the same thickness. The bottom of this cavity will be pierced with a round hole, 
 with the metal forced perpendicularly beyond the bottom of the recess ; the diameter 
 of the hole to be twenty times the thickness of the plate, and the height of the sides 
 five times the same thickness. All the corners will be rounded with a curve not of 
 greater radius than the thickness of the plate. The pieces thus tested, with every pre- 
 caution which the working of steel requires, must show no signs of yielding or cracking, 
 even when cooled in a brisk current of air. 
 
 " Tempering tests : For these tests bars 10.24 inches long by 1.58 inches wide will be 
 cut from the plate longitudinally as well as transversely. These strips will be heated 
 uniformly to a slightly dull cherry-red, and then tempered in water at a temperature of 
 82 Fahr. Thus treated they must be bent in the testing-machine to a curve of which 
 the minimum radius is not greater than the thickness of the bars. These same bars, 
 when the corresponding plates are to be used for boilers, will be bent double in the 
 press without showing any traces of fracture, and in such a way that the halves of the 
 plate may be in contact. The sides of the bars thus tested, if round, can be squared 
 up with a file. Plates not coming up to these tests will be rejected." 
 
 11. Tests for Plate, Beam, Angle, Bulb, and Bar Steel used in building 
 Ships for Her Majesty's Navy. Admiralty, 9th January, 1879. (The same direc- 
 tions are followed for steel used for boiler-shells.) 
 
 " Strips cut lengthwise or crosswise to have an ultimate tensile strength of not less 
 than 26, and not exceeding 30, tons per square inch of section, with an elongation of 20 
 per cent, in a length of 8 inches. The beam, angle, bulb, and bar steel to stand such 
 forge-tests, both hot and cold, as may be sufficient, in the opinion of the receiving 
 officer, to prove soundness of material and fitness for the service. 
 
 " Strips cut crosswise or lengthwise 1 inches wide, heated uniformly to a low cherry- 
 
SEC. 12. TESTING THE MATERIALS. 105 
 
 red and cooled in water of 82 Fahr., must stand bending in a press to a curve of 
 which the inner radius is one and a half times the thickness of the steel tested. 
 
 " The strips are all to be cut in a planing-machine, and to have the sharp edges taken 
 off. 
 
 "The ductility of every plate, beam, angle, etc., is to be ascertained by the applica- 
 tion of one or both of these tests to the shearings, or by bending them cold by the 
 hammer. 
 
 " All steel to be free from lamination and injurious surface-defects. 
 
 " One plate, beam, or angle, etc., to be taken for testing from every invoice, provided 
 the number of plates, beams, or angles, etc., does not exceed fifty. If above that num- 
 ber, one for every additional fifty or portion of fifty. Steel may be received or 
 rejected without a trial of every thickness on the invoice. 
 
 "The pieces of plate, beam, or angle, etc., cut out for testing are to be of parallel 
 width from end to end, or for at least eight inches of length. 
 
 " Plates may be ordered either by weight per superficial foot or by thickness. In 
 the former case the weight named will always be the greatest that will be allowed, and 
 a latitude of 5 per cent, below this will be allowed for rolling in plates of J inch in 
 thickness and upwards, and 10 per cent, in thinner plates. When the plates are 
 ordered by thickness their weight is to be estimated at the rate of 40 pounds per 
 square foot for plates of one inch thick, and in proportion for plates of all other thick- 
 nesses. In this case, also, the weight due to the thickness by this calculation is not 
 to be exceeded, but the same latitude as above will be allowed below the weight for 
 rolling. The average weight per foot of the plates ordered is to be ascertained by 
 weighing not less than 10 tons at a time when larger parcels than 10 tons are delivered ; 
 if these 10 tons exceed the due weight (calculated as stated above), or are more than 
 the before-mentioned percentage below it, the whole may be rejected. In smaller de- 
 liveries than 10 tons the average is to be ascertained by weighing the whole parcel. 
 The same conditions as to latitude and mode of ascertaining weight apply also to 
 other descriptions of steel in the contract." 
 
 12. Examining Boiler-plates. The method pursued in the English dockyards 
 for examining iron plates as to imperfections developed in the process of manufacture 
 is described by Reed as follows: "When parcels or lots of iron plates are delivered 
 into the building-yard they are spread out and examined with the object of first ascer- 
 taining if the manufacturer's name and the brand of quality are duly stamped upon 
 each plate, and then of searching for surface-defects, such as blisters, flaws, lamina- 
 tions, or bad places caused by dirt or cinders getting between the rolls during the 
 
106 STEAM BOILERS. CHAP. V. 
 
 rolling of the plates, any one of which, if considerable, would cause the overseer to 
 reject the plate. This surface-examination being completed, each plate is raised from 
 the ground, and, being either hung by one edge or otherwise suitably supported, is 
 tapped over with a small hammer. If it everywhere gives out a clear, ringing sound 
 the plate is considered to be solid ; but if a heavy and dull sound be given out it is 
 presumed that the plate is laminated or otherwise defective. If this test is audibly 
 decisive against the plate it is at once rejected ; but if the quality of the plate appears 
 doubtful a further test is resorted to. This consists in supporting the plate at its four 
 corners, strewing the upper surface with sand, and lightly tapping over the under side. 
 Wherever the plate is sound the sand will be driven up off the plate by each tap of the 
 hammer, but if it is blistered or laminated at any place the sand will not there be 
 moved. The plates which the foregoing tests show to be satisfactory are next carefully 
 measured and weighed, their actual weights being compared with those due to their 
 dimensions and specified thickness." 
 
 It is convenient to mark the plate off with a chalked line into squares of four or six 
 inches to make sure that each part of the plate is tapped with the hammer ; a few 
 blows in each square will be sufficient. Both sides of the plate should be thus tested ; 
 sometimes a plate will appear to be quite sound on one side, but will, nevertheless, be 
 found defective on the other side. 
 
 All these tests may, however, fail to reveal internal defects, which may become 
 apparent as soon as the plate is heated at the forge, or perhaps not until it has been for 
 some time in use in the boiler. 
 
CHAPTEK VI. 
 
 PRINCIPLES OF THE STRENGTH OF BOILERS. 
 
 1. Resistance of Spherical Shells to an Internal Fluid Pressure. An elas- 
 tic fluid contained in a closed vessel presses each unit of area of the surrounding walls 
 with equal force. The resistance offered by the walls depends on their superficial area, 
 their form, their thickness, and the coefficient of resistance of the material. 
 
 The hollow sphere encloses the largest space in proportion to the superficial area of 
 its shell, and all vessels that are not spherical, exposed to an internal fluid pressure, 
 experience distortion on account of their tendency to assume the spherical form. A 
 hollow sphere, having a shell of uniform thickness composed of a homogeneous ma- 
 terial, experiences the same tension at all sections of metal formed by diametrical 
 planes. 
 
 The area of a diametrical section, S, of a thin spherical shell is very nearly given 
 by formula : 
 
 S = 27i r t, 
 
 when t represents the thickness, and 
 r the inner radius of the shell, 
 and t is supposed to be very small compared with r. 
 
 The whole force, F, to be resisted by the tenacity of section S is equal to the excess 
 of the internal fluid pressure per unit of area over the external pressure, into the area 
 of the plane passing through this section, or 
 
 Assuming that every portion of section 8 is equally strained by F, and designating 
 by k the coefficient of the ultimate tenacity of the material of which the shell is com- 
 posed, the bursting pressure will be found from the equation : nr t p = 2nrtJc; hence 
 
 2/ 1 K rT T 
 
 P^-^i M 
 
 and the proper ratio of the thickness to the radius of a thin hollow sphere is given by 
 the formula : 
 
 A .P. 
 
 r ~2&' 
 It is to be borne in mind that this proportion applies to all spherical segments. 
 
 107 
 
108 STEAM BOILERS. CHAP.VL 
 
 The assumption that all portions of the section S are equally strained by an internal 
 fluid pressure is not strictly correct, but comes sufficiently near the truth in the case of 
 thin shells of such dimensions as are generally used in connection with steam boilers. 
 Considering the shell to be composed of a number of concentric layers, the tension ex- 
 perienced by each will decrease with its distance from the centre, on account of the 
 different ratios in which the surface and the mass of a sphere increase with an increase 
 of radius. 
 
 2. Resistance of Cylindrical Shells to an Internal Fluid Pressure. The 
 tension produced in a cylindrical shell by an internal fluid pressure may be considered 
 as being of two different kinds viz., first, a tension acting in a longitudinal direction, 
 tending to pull the ends of the cylinder apart ; and, secondly, a tension acting in a dia- 
 metrical direction, tending to split the cylinder from end to end. 
 
 The force, F, producing the first-named tension is represented by the formula : 
 
 F = r* rrp / 
 
 and the sectional area, S, of a thin shell resisting this force may be represented with 
 sufficient accuracy, as in the case of thin spherical shells, by the formula : 
 
 The value of p, when it becomes the bursting pressure, is found from the equation, 
 
 r* np = %nrtlc ; 
 
 hence p -'. [II. ] 
 
 It follows that the strength of a cylindrical shell in resisting a bursting stress applied 
 in a longitudinal direction is the same as that of a spherical shell of equal radius and 
 thickness. 
 
 To find the value of p which would split the cylinder from end to end, let the shell 
 be considered as divided lengthwise into rings, of which the length is unity. The force 
 tending to rupture such a ring at the sections formed by any diametrical plane is given 
 by formula : 
 
 and the area of these sections by 
 
 8=2t. 
 
SEC. 3. PRINCIPLES OF THE STRENGTH OF BOILERS. 109 
 
 The bursting pressure is, therefore, found from the equation : 
 
 2rp = 2tk; 
 
 tic 
 hence p = . [III.] 
 
 The value of p as found by equation [III.] is only half as great as the value found 
 by equation [II.] ; hence the tendency of a thin cylindrical shell to split from end to 
 end under an internal fluid pressure is twice as great as its tendency to rupture circum- 
 ferentially. 
 
 Representing the length of the cylinder by I, the equation between the total force F, 
 acting in a diametrical direction, and the resistance of the sections of the whole shell 
 formed by diametrical planes, assumes the form : 
 
 Since this equation gives the same value for p as is given by formula [III.] for a 
 cylindrical ring of which the length is unity, it follows that the strength of hollow 
 cylinders exposed to an internal fluid pressure is independent of their length. 
 
 Cylindrical shells would always preserve their true cylindrical form under an internal 
 fluid pressure, if their ends were not necessarily rigidly attached, and thereby prevented 
 from expanding equally with the rest of the shell in proportion to the elasticity of the 
 material with each increase of pressure ; on this account the middle portion expands 
 more freely than the ends, and the cylindrical tube has a tendency to become barrel- 
 shaped under such conditions. This effect, however, is appreciable only in the case of 
 very soft and elastic materials, and no account need be taken of it in the case of tubes 
 used in connection with boilers. 
 
 The foregoing remarks about thick hollow spheres apply equally to thick hollow 
 cylinders. 
 
 The question has been raised whether cylindrical shells, when strained longitudinally 
 and diametrically at the same time, did not offer a diminished resistance to rupture in 
 any one direction. Direct experiments by Navier on wrought-iron spheres have, how- 
 ever, proved that the resistance of the metal, when the stress is applied simultaneously 
 in all directions, is the same as when the tension acts in one direction only. 
 
 3. Resistance of Cylindrical Shells to an External Fluid Pressure. Thin 
 hollow cylinders exposed to an external fluid pressure never give way by direct crush- 
 ing, but by collapsing ; it may be assumed that, other things equal, the resistance of 
 ttibes to collapsing is greater as their form is more truly cylindrical and their shell more 
 perfectly homogeneous. 
 
 Fairbairn has deduced the following formula from experiments made mostly on very 
 
110 
 
 STEAM BOILERS. 
 
 CHAP. VI. 
 
 thin cylindrical tubes of various lengths and diameters viz., for wrought-iron cylin- 
 drical tubes let 
 
 I = the length, 
 
 d the diameter, and 
 
 I the thickness of the shell, all expressed in the same unit of measure, and let 
 p = the collapsing pressure in pounds per unit of area ; then 
 
 p = 9,672,000 ^. [IV.] 
 
 In case a tube is stiffened by T-iron rings or by flanges, I represents the distance be- 
 tween two such adjacent rings or flanges. 
 
 Fairbairn finds that the collapsing pressure of a flue of an elliptic form of cross- 
 section is found approximately by substituting, in the preceding formula, for d the 
 diameter of the osculating circle at the flattest part of the ellipse that is, let a be the 
 greater and b the less semi-axis of the ellipse ; then we are to make 
 
 In order to facilitate calculations by this formula the 2.19th power is given in the 
 following table for such mimbers as represent, in fractions of an inch, the thicknesses of 
 boiler-plates most frequently used in the construction of boiler-flues : 
 
 TABLE XX. 
 
 
 
 a.igth power of number. 
 
 Logarithm of 2.i9th power. 
 
 A 
 
 A 
 A 
 
 .18750 
 .21875 
 .25000 
 .28125 
 
 025578 + 
 
 35 8 5o 
 048027 + 
 .062160 + 
 
 .4078728 2 
 .5544863 2 
 .6814886 2 
 .7935126 2 
 
 A 
 H 
 
 1 
 U 
 
 .3125 
 
 34375 
 375 
 .40625 
 
 .078293- 
 .096465 + 
 .116715 
 .139078 
 
 .89372152 
 
 98437152 
 .0671285 I, 
 
 M32575 "I 
 
 A 
 H 
 i 
 H 
 
 -4375 
 .46875 
 .50000 
 S3I2S 
 
 163584 + 
 .190266 + 
 .219151 + 
 .250267 + 
 
 .2137420 I 
 .2793614 I 
 
 3407443 I 
 .3984046 I 
 
 A 
 
 1 
 H 
 
 1 
 
 56250 
 .62500 
 
 .68750 
 .75000 
 
 .283641 - 
 
 357254 + 
 .440177 
 
 532579 + 
 
 45 2 7683 I 
 5529772! 
 .6436272 1 
 .7263842 I 
 
 In an article in the Annales du Genie civil, March, 1879, on the " Resistance of 
 
SET. 3. PRINCIPLES OF THE STRENGTH OP BOILERS. Ill 
 
 Tubes subjected to an External Pressure," by Theodore Belpaire, an attempt has been 
 made to deduce a new formula for the collapsing strength of tubes. The following is a 
 brief synopsis of the contents of this paper : 
 
 A tube having a perfectly cylindrical form and a circular cross-section, and being 
 made of a homogeneous material, would remain cylindrical under increasing pressures 
 till the material gave way by crushing. But the circular cross-section is a form of un- 
 stable equilibrium for resisting external pressure, because a slight alteration of shape 
 suffices to destroy the equilibrium under sufficiently great pressures. Want of homo- 
 geneity of the material has likewise a great influence on the resistance. It would be 
 imprudent to rely under such circumstances upon the increase of resistance due to the 
 homogeneity of the material and to the circular form, since it is impossible to know in 
 advance to what extent they will be realized in construction. Fairbairn's formula 
 represents the results of experiments where this increase of resistance obtained to a 
 greater or less extent, and for this reason it does not appear to be reliable for determin- 
 ing the dimensions which should be given to internal flues. Such flues derive their 
 strength mainly from the fastening of their ends, which may be considered as absolutely 
 rigid, and sometimes from rigid rings by means of which the intermediate sections of 
 the flues are joined together. The collapsing of a cylinder is always preceded by dis- 
 tortion ; it is evident that collapse cannot take place when this distortion is not allowed 
 to exceed a certain limit and the material is not unduly strained. 
 
 The writer then considers the case of a tube with ends rigidly fixed, and supposes 
 that under an external pressure it changes its form in such a manner that its generatrix 
 becomes the arc of a circle, the centre of which lies on a perpendicular erected in the 
 centre of the generatrix ; and, neglecting the elastic forces due to flexure or elongation 
 of the fibres which are very small as long as the curvature is slight he investigates 
 the shearing stresses ; these attain their greatest value at the fixed ends. 
 
 Calling S the greatest shearing stress, 
 
 p the pressure in pounds per square inch, 
 t the thickness of the tube in inches, 
 L the length of the tube in inches, 
 
 he deduces the following approximate formula for the external pressure which a given 
 tube can bear with a degree of safety depending on the value attributed to S viz. : 
 
 ^ t *b rTTT T 
 
 p= L [>!] 
 
 The writer deduces then a general value S from two experiments made by Fairbairn 
 with elliptical tubes, because the uncertain and variable elements of strength due to the 
 
112 
 
 STEAM BOILERS. 
 
 CHAP. VI. 
 
 cylindrical form and to homogeneity of the material do not enter here. When the 
 factor of safety in the foregoing equation is to be four, the value of S becomes 
 
 8= 428,394 ^ - 7,111,550 (^j ; 
 
 where t is the thickness and D the diameter, both expressed in inches. 
 
 Since externally-pressed tubes always derive a large increase of strength from their 
 circular form, a factor of safety of four is considered sufficient in using these formulas. 
 
 Applying these formulas to 34 cases where tubes collapsed either in an experimental 
 apparatus or during actual use in a steam boiler, the factor is found to vary actually 
 from 4.2 to 16.8. With reference to those cases where the factor of safety exceeded 
 four greatly, the writer claims that the high pressures necessary to produce collapse 
 indicate merely the great increase of strength derived in the particular instances from 
 the uncertain element of circular form. 
 
 A greater number of experiments are required from which to deduce an expression 
 for 8, so that the influence of thickness and diameter of the tube, and of the eccentricity 
 of its elliptical cross-section, on the value of $ may be fully determined. 
 
 For the purpose of applying the foregoing formulas in practice the following table 
 
 has been calculated, giving the values of for different values of ^- and for different 
 
 factors of safety : 
 
 TABLE XXI. 
 
 Values of 
 
 Values of ^ for a factor of safety of 
 
 Values of 
 
 Values of -S 1 for a factor of safety of 
 
 t 
 
 
 
 
 t 
 
 
 
 
 D 
 
 4 
 
 5 
 
 6 
 
 D 
 
 4 
 
 5 
 
 6 
 
 0.003 
 
 1221 
 
 1526 
 
 1832 
 
 O.OI2 
 
 4117 
 
 5^6 
 
 6175 
 
 0.004 
 
 l6oO 
 
 2OOO 
 
 2400 
 
 0.013 
 
 4367 
 
 5459 
 
 6 55 
 
 0.005 
 
 1964 
 
 2 4S5 
 
 2946 
 
 0.014 
 
 4604 
 
 5755 
 
 6906 
 
 0.006 
 
 23M 
 
 2892 
 
 3471 
 
 0.015 
 
 4826 
 
 6032 
 
 7239 
 
 O.OO7 
 
 2650 
 
 3312 
 
 3975 
 
 0.016 
 
 534 
 
 6292 
 
 755i 
 
 O.OO8 
 
 2972 
 
 3715 
 
 445 s 
 
 0.017 
 
 5227 
 
 6534 
 
 7840 
 
 O.OOp 
 
 3280 
 
 4100 
 
 4920 
 
 0.018 
 
 5407 
 
 6 759 
 
 8no 
 
 O.OIO 
 
 3573 
 
 4466 
 
 536o 
 
 0.019 
 
 5572 
 
 6965 
 
 8358 
 
 O.OII 
 
 3852 
 
 48IS 
 
 5778 
 
 O.O2O 
 
 5723 
 
 7i54 
 
 8585 
 
 Professor Grashof has derived from Fairbairn's experiments the following empirical 
 
 formula viz. : 
 
 ft.tn 
 
 p = 1,057,180 i . B8 . 
 
SEC. 4. 
 
 PRINCIPLES OP THE STRENGTH OP BOILERS. 
 
 113 
 
 In which d = diameter of the tube in inches ; 
 I = length of the tube in inches ; 
 t = thickness of the tube in inches ; 
 p = pressure in pounds per square inch. 
 
 4. Experiments made on the Resistance of Cylindrical Flues to an Ex- 
 ternal Fluid Pressure. In the year 1874 experiments were made at the Washington 
 Navy- Yard to determine the resistance of large cylindrical boiler-flues to collapse. The 
 apparatus used for this purpose is represented in figure 5. 
 
 Fig. 5. 
 
 
 
 
 i 
 
 4- 
 
 It consisted of a cylindrical shell of 63 inches diameter, constructed of boiler-iron 
 f inch thick. A cylindrical flue, 77| inches long and 54 inches in inside diameter, was 
 securely riveted to flanges within this shell. This inner cylindrical flue was con- 
 structed of i-inch boiler-iron, and consisted of two rings connected by an interior butt- 
 strap 7f inches wide and J inch thick. Each ring was formed of two plates with butt- 
 joints, the butt-straps, 7| inches wide and inch thick, being placed on the inside. 
 
114 
 
 STEAM BOILERS. 
 
 CHAP. VI. 
 
 The longitudinal seams of the rings broke joint as shown in the drawing. All seams 
 were double-riveted .and carefully calked. The unsupported length of the internal flue 
 (measured between the inner edges of rivet-holes) was 71J inches. A 2f-inch pipe 
 screwed into the outer shell connected with a force-pump, by means of which a hydro- 
 static pressure was produced during the trials within the annular space formed by the 
 two cylinders. Carefully-tested spring-gauges indicated the pressure. One of the 
 rings of the inner shell collapsed under a pressure applied to try the tightness of the 
 joints and rivets, the two gauges used indicating a pressure of 100 Ibs. and 110 Ibs. re- 
 spectively. The mean of these two readings was probably the true pressure, as, on com- 
 parison with a standard gauge, the one gauge was found to indicate less and the other 
 more than the true reading near that pressure. The bulged portion of the shell of the 
 flue was pressed out and shored up from the opposite side of the flue, and the pressure 
 was again applied. This time collapse took place in the other ring ; then the opera- 
 tion of forcing out and shoring up was repeated. It is evident that by this shoring up 
 the flue became stiffer each time, and the results of the successive trials indicate this. 
 The following are the results of the tests made with this flue : 
 
 1st collapsed in testing joints and rivets, at 105 Ibs. 
 2d with one bulge shored up, at 120 " 
 
 3d " with two bulges shored up, at 148 " 
 
 4th " with three bulges shored up, at 155 " 
 5th with four bulges shored up, at 186 " 
 
 Careful measurement after construction had revealed the fact that the flue was slightly 
 oval, the larger diameter being 54^ inches and the smaller diameter 53| inches ; and 
 on removing the shell and gauging the sheets composing the flue, the one which had 
 collapsed first was found to be slightly less in thickness than J inch. 
 
 Another flue of J-inch boiler-iron was made, care being taken to gauge the sheets 
 accurately before fitting them. This flue consisted likewise of two courses, 38 inches 
 and 39 inches long respectively, but they were connected by flanges, with a ring H inch 
 thick between them (see figure 6). 
 
 Fig. 6 
 
 This flue was found to be perfectly cylindrical, having an internal diameter of exactly 
 54 inches. 
 
SEC. 5. PRINCIPLES OP THE STRENGTH OP BOILERS. 115 
 
 Three spring-gauges, that had been carefully compared with a standard gauge, 
 were used to measure the pressure. After the first trial the bulged portion of one 
 course was shored up as before described, and a second trial made, when the other 
 course collapsed. 
 
 Fiist trial. Second trial. 
 
 Ashcroft's gauge ..................................... 132 Ibs. 130 Ibs. 
 
 Post's gauge ......................................... 134 " 131 " 
 
 Utica gauge .......................................... 134 " 131 " 
 
 Applying to the first experimental flue Fairbairn's formula, modified for elliptical 
 
 / .'t i 
 
 tubes, viz., p 9,672,000 - i -, -f, we get for the collapsing pressure p = 118.42 Ibs. 
 
 <& a l 
 
 This result shows a close agreement with the first two results of the experiment, when 
 the reduced thickness of one of the sheets is taken into consideration. 
 
 The collapsing pressure of the second flue should have been nearly 240 Ibs. according 
 to Fairbairn's formula, and 225 Ibs. according to Grashof 's formula. 
 
 Belpaire's formula gives for the collapsing pressure of the second flue 101.2 Ibs. per 
 square inch ; the actual higher collapsing pressure indicating the increase of strength 
 due to the circular form. 
 
 5. Strength of Flat Plates The theory of flexure for loaded flat plates leads to 
 very complicated expressions. The following approximate method is employed by 
 Weisbach (' Manual of the Mechanics of Engineering,' vol. ii.) for finding the thickness 
 of such plates : 
 
 Let m the length, and n = the width of a rectangular flat plate secured at the cir- 
 cumference to a solid frame or by a row of rivets, and p = the pressure which it has to 
 sustain per unit of surface. Let us imagine this sheet to be cut into parallel strips in 
 the direction of the length, the ends of which are secured to the frame ; and let us 
 assume that the portion p, of the pressure p causes the tension of these strips ; then, if 
 we indicate the breadth of each strip by b, the thickness by t, and the coefficient of re- 
 sistance to rupture by I; we have the equation : 
 
 m 
 or m t = 
 
 hence t = *,Vj [VII.] 
 
 - /*' 
 
 If, on the other hand, we imagine the plate cut up into similar strips in the direction 
 
116 STEAM BOILERS. CHAP. VI. 
 
 of its breadth, and assume that the tension of these strips is caused by the pressure 
 p t = p p l} we find in the same manner 
 
 *???* ) *7?* T) 
 
 As the deflection in the first case decreases as if- 1 , and in the other case as ., , and 
 as it is as great in the one case as in the other, we can put 
 
 hence Pi = f Pi an d p = p 
 
 4 
 
 consequently p, = 4 jf. . 
 
 By introducing these values of p t and p, into equations [VII.] and [VII. a] we get for 
 the thickness of the plate in the first case 
 
 , 
 
 
 
 and in the second case 
 
 p 
 
 If n > m, we must find the thickness of the plate according to formula [VIII.] ; if 
 m > n, we must use formula [VIII. a]. 
 
 For square sheets we have m = n, therefore 
 
 * = [IX-] 
 
 The following formulas are given by Rankine for calculating the strength of un- 
 stayed flat surfaces secured at the edges, in which p, t, r, and Tc denote the same quan- 
 tities as in sections 1 and 2 of the present chapter : 
 
 m is the length of a rectangular plate, or the side of a square plate, in inches ; 
 
 n is the breadth of a rectangular plate in inches ; 
 
 m being greater than n in the case of rectangular plates : 
 
 Flat circular plates : P = i [X.] t = r\\ [X.a] 
 
 Flat rectangular 4P (m' + Qfr. -, z- 866m ' X7 V P rX T a\ 
 
 plates: Sm'xn* Y F(^T^)' 
 
 Flat square plates : p = i|-*; [XII.] t = .612 m y -; [Xlla.] 
 
 3 Tfl K 
 
SBC. 6. PRINCIPLES OP THE STRENGTH OP BOILERS. 117 
 
 Comparing the thickness required for a flat circular plate, as given by formula [Xa.], 
 with the thickness of a cylindrical shell of equal radius and of equal strength, as given 
 
 by formula [III.], section 2 of the present chapter, viz., T = ^-, we find 
 
 For a boiler 3 feet in diameter and having a cylindrical shell f inch thick, single- 
 riveted, the solid unstayed flat end-plate would have to be about 2 inches thick to 
 make it as strong as the cylindrical shell. It is, however, impracticable to use such 
 heavy plates in boiler-construction ; extensive flat surfaces of boilers are therefore sup- 
 ported by stays or stiffened by various contrivances, so that they may be formed of 
 relatively thin plates. Various methods of staying flat surfaces will be described, and 
 rules for proportioning braces will be given, in chapter x. 
 
 Experiments on the strength of flat ends of cylindrical vessels, described by Robert 
 Wilson in Engineering, September 28, 1877, indicate that the actual breaking strength 
 of flat plates is much greater than that given by the above formulae, and that the flat 
 end-plates of boilers receive a great access of strength and stiffness by flanging their 
 edges. But flat plates begin to bulge out with very low pressures, and the springing of 
 the plates as the pressure is alternately applied and relieved destroys them inevitably 
 in the course of time by grooving or channelling. 
 
 The stiffness of plates is greatly increased by buckling ; and when the surfaces are 
 not too large buckled plates are sometimes used in boilers instead of stayed flat plates. 
 Rankine gives the following rule for calculating "the load, uniformly distributed over a 
 buckled plate, which will crush it, the plate being square and fastened all around the 
 edges : Multiply the depth to which the plate is bucTcled by the square of the thickness, 
 both in inches, and by 165 ; the product will be the crushing load in tons, nearly. Cen- 
 tral load which crushes a buckled plate about one-third of uniformly-distributed load." 
 
 6. Strains on Braces and their Attachments. When a boiler is composed of 
 thin, flat plates offering little resistance to bending, it may be assumed without serious 
 error that the stress experienced by a brace is the resultant of the whole pressure act- 
 ing perpendicularly to the portion of the plate supported by the brace. When the 
 plates are increased in thickness, or are strengthened by angle or T-irons, their increased 
 resistance to bending causes a corresponding diminution of the stress on the braces. 
 
 A brace standing perpendicular to a thin, flat plate experiences a tensile stress equal 
 
118 
 
 STEAM BOILERS. 
 
 CHAP. VI. 
 
 to the total pressure borne by the supported surface. This may be expressed by the 
 equation : 
 
 when S the area of the supported surface in square inches ; 
 p the steam-pressure in pounds per square inch ; 
 and T = the total tension, in pounds, of the brace. 
 
 The tension of an oblique brace is equal to the tension which a perpendicular brace 
 supporting the same surface would experience, divided by the cosine of the angle which 
 the oblique brace forms with a perpendicular to the supported surface (see figure 7), or 
 
 T l = p8 . [XIV.] 
 cos. oc 
 
 When T ' is resolved into a pair of rectangular components acting respectively in a 
 
 Fig. 7. 
 
 perpendicular and parallel direction to the supported surface, 
 
 the latter component is equal to 
 
 T 1 sin. a = p S tang, a, [XIV. a] 
 
 and produces a bending strain on the brace when its ends are 
 
 rigidly fastened so that its angular position is fixed. When 
 the ends of the oblique brace are attached by movable joints for instance, by an eye 
 and pin offering little or no resistance to a change in the angular position of the brace, 
 the component T 1 sin. a exerts a thrust on the plate to which the brace is attached, 
 tending to produce buckling. 
 
 In order to investigate the various strains obtaining in a system of oblique bracing, 
 and the conditions required for the establishment of equilibrium, the braces shown on 
 Plate (VII.), which tie the top of the boiler to the 
 sides of the tube-boxes, are taken as an example. 
 The top of the boiler is stiffened by T-irons, to 
 which the branch-braces are attached ; the latter 
 being spaced so that each one supports an equal 
 area of plate. The points of attachment, A, B, and 
 D, are given, and the main brace, E D, is to have 
 such a direction that the resultant of the stresses 
 on E A and B E does not produce flexure, but 
 simply tension on E D. 
 
 Let P P represent the resultants of the forces 
 acting at right angles to the supported plate at the point of attachment of the oblique 
 branches, A E and B E, to the T-iron ; 
 
SBC. 7. 
 
 PRINCIPLES OP THE STRENGTH OF BOILEB8. 
 
 119 
 
 R = the corresponding stress on the brace D E ; and 
 T, r' the stresses on A E and B E respectively. (See figure 8.) 
 In order to balance the equal forces P P, the components of r and r', normal to 
 A B, must each be equal in amount and opposite in direction to P P ; this condition is 
 
 P P 
 
 fulfilled when r 
 
 and r' = 
 
 The components of r and r', pax- 
 
 cos. A JSf" ~cos.BEF' 
 
 allel to A B and equal to P x tan. A E F and P x tan. B E F respectively, are 
 resisted by the stiffness of the T-iron. 
 
 The forces r and r' being each resolved into two rectangular components, respec- 
 tively parallel and normal to E D, the sum of the former gives the tension of the brace 
 E D, while the normal components L E and K E tend to deflect E D. Equilibrium 
 requires that these normal components of the forces r and r' should be equal in amount 
 and opposite in direction, so that they balance each other ; and that the component of 
 
 2P 
 
 the tension R, normal to A B, should be equal to 2 P, or R 
 
 These con- 
 
 cos. FEC" 
 
 ditions are fulfilled when the prolongation of E D intersects the line A B at its centre, 
 or when A C = B C. The following proportion then exists between the force P and the 
 tension of the main brace and of the branches P :H:r : r' : :EF:2EC:AE:BE. 
 
 When the direction of E D is normal to A B the foregoing conditions give the equa- 
 tions R = 2 P, and r r'. 
 
 7. Strains on Circular Arcs. An internal normal pressure on any point of an 
 arc produces at such a point a tension acting in a tangential direction, equal to the nor- 
 mal pressure multiplied by the radius of curvature at the point in question ; conse- 
 quently, when the flat sides of the shell of a boiler are connected by a circular arch 
 
 FFg. 9. 
 
 forming tangential planes to the cylindrical 
 surface they experience, per unit of length, 
 a tension equal to the product of the pressure 
 per unit of area and the radius of the arch. 
 
 When the shell is formed by combining 
 several cylindrical arches, as in "Emery's 
 Connected- Arc Boiler" (see figure 92), the 
 strains on the braces which keep the system 
 in equilibrium under pressure may be found in 
 the following manner : 
 
 If the semicircle M P represents the cross- 
 
 section of a cylindrical arch subjected to internal fluid pressure (see figure 9) we may 
 
120 
 
 STEAM BOILERS. 
 
 CHAP. VL 
 
 represent the resultant tangential force at any point of the circumference M,, for 
 instance by a tangent line M, V made equal to the radius M a O. Resolving this tan- 
 gential force M, V into two rectangular components, represented in magnitude and 
 direction by the lines V W and M 3 W, respectively parallel and perpendicular to the 
 diameter M P, we find that, since the triangle M, V W is similar and equal to the 
 triangle O M, N,, line M, N, (equal to the sine of arc M M,) and line O N, (equal to the 
 cosine of arc M M,) represent the intensity of the rectangular forces which balance the 
 tangential force at the point M 5 , each of said rectangular components acting, however, 
 in the direction of the other. 
 
 "If we consider M P and M 8 O rectangular co-ordinate axes, passing through the 
 centre O of the circle, then the two forces required to hold in equilibrium the end of 
 any arc forming part of the quadrant M M 8 will be measured respectively by the pro- 
 jections of the arc and of its complement upon the co-ordinate axes, and equal the 
 length of such projections multiplied by the pressure per superficial unit. For instance, 
 the horizontal component required to balance the tangential force at M a is measured by 
 M., N, = sin. M, O M = O T a , or the projection of the arc M a M on axis M 8 O, and the 
 strain equals O T, multiplied by the pressure per superficial unit. The vertical compo- 
 nent is similarly measured by N, O = cos. M, O M, or the projection of the arc M 2 M ft , 
 which is the complement of arc M M 2 , on the axis M P ; and the strain equals N, O 
 multiplied by the pressure per superficial unit." 
 
 If figure 10 represents a section of a boiler consist- 
 ing of a series of connected circular arcs of eqiial 
 radii the centres, O,, O,, O,, and O 4 , of the circles all 
 being located in the horizontal line x y then, accord- 
 ing to the foregoing demonstration, the horizontal 
 component of the tangential force at U due to the 
 pressure on arc x U may be represented by U Y, and 
 the horizontal component of the tangential force at the same point U, due to the pres- 
 sure on arc U U a , may likewise be represented by U Y, but it acts in the opposite direc- 
 tion ; consequently these two horizontal components of the tangential forces of the 
 adjoining arcs balance each other. The vertical components of the tangential forces 
 acting at the same point may be represented by the horizontal projections of the 
 arcs W U and U W 5 viz., O, Y and Y 0, both acting in the same direction. The 
 same reasoning applies to the points U,, U,, V, V,, V,. Supposing the above figure 
 to be symmetrical with regard to the axis x y, the system of connected arcs will be 
 held in equilibrium by the vertical ties U V, U, V,, U, V,, each one of which has to 
 
 Fig. 10. 
 
SEC. 7. 
 
 PRINCIPLES OF THE STRENGTH OP BOILERS. 
 
 121 
 
 sustain a pull, for each unit of length supported, equal to the pressure per unit of 
 area multiplied by the distance between the centres O, O,, O, 0,, and O, O, respec- 
 tively. 
 
 Fig. 11. 
 
 The same rule applies to the cases illustrated in figures 11, 12, where the connected 
 arcs have different radii, but have their centres on the same horizontal line x y. 
 
 When the centres of the connected arcs do not lie on the same horizontal line x y, 
 as in figures 13, 14, the strains on the vertical ties are measured as before by the hori- 
 zontal distance between the centres viz., O Y 4- Y 6 O,. But in figure 13 the horizontal 
 strain of the arc W U is measured by U Y, while that of the arc U U, is measured by 
 
 Fig. 14. 
 
 U Y s , less than U Y ; hence the arc W U tends to straighten the arc U U, by a force 
 measured by U Y - - U Y, = Y Y, ; this action has to be prevented by a tie-rod, U U,. 
 
 In figure 14 the horizontal strains of the middle arc U U, being greater than those 
 of the outer arcs by an amount Y Y 6 , the outer arcs will fail of themselves to furnish 
 sufficient support, and a strut must be placed from U to U,, which has to sustain a com- 
 pressive strain measured by Y Y 6 . (See paper on ' Connected- Arc Marine Boilers,' by 
 C. E. Emery, read before American Society of Civil Engineers, December, 1876.) 
 
CHAPTER VII. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 1. General Considerations governing the Design of Marine Boilers. The 
 
 designing of a marine boiler, especially for a war-vessel, involves the fulfilment of 
 many conditions which are to some extent antagonistic ; hence compromises have to be 
 made, and some advantages with regard to economic and potential efficiency have to be 
 sacrificed to other essential conditions. The principal conditions to be satisfied in the 
 design of a boiler may be considered under the following heads : (1) The boiler must 
 be able to furnish the power required ; (2) its parts must be proportioned and arranged 
 with regard to economic efficiency, durability, and economy in construction ; (3) every 
 part of the boiler must possess the necessary strength. 
 
 The principal restriction imposed on the designer of marine boilers of a given power 
 is the limitation with regard to the weight of, and the space allotted to, the boilers 
 proper and their attachments, the fire-room and the fuel. In a man-of-war, where it is 
 especially important that all parts of the machinery should be placed as low as possi- 
 ble in the vessel, it is generally stipulated that no part connected with the steam-space 
 of the boilers shall reach above the water-line. In marine boilers of the ordinary type, 
 having the tubes or flues arranged over the furnaces, the area of the grate-surface is 
 the principal element which determines the space occupied by the boilers in the length 
 and breadth of the vessel. It is, therefore, convenient to determine the grate-surface 
 in the first place, and to proportion and arrange the other parts of the boiler afterward 
 according to the conditions imposed. 
 
 The width of the fire-room must exceed the length of the grate by at least two feet, 
 in order that the tools used in the management of the fires may be manipulated with- 
 out hindrance. In general the most economical disposition of the room is made by 
 arranging the boilers in pairs, facing each other, with the fire-room between them. 
 
 2. Boiler-power. The power of a boiler is measured by the weight of steam which 
 it can generate in a unit of time. It is customary to measure the relative evaporative 
 efficiency of boilers by the number of pounds of water of 212 that can be evaporated 
 under atmospheric pressure in a unit of time ; but the actual power of a boiler must 
 be calculated from the weight of steam of the working pressure that can be generated 
 
 128 
 
SEC. 3. DESIGN, DRAWINGS, AND SPECIFICATIONS. 123 
 
 in a unit of time from the feed- water as delivered into the boiler. The average tem- 
 perature of the feed-water of marine boilers may be taken as 115 when no heaters are 
 used. 
 
 The number of heat-units that can be made available and the weight of water of a 
 given temperature that can be evaporated under a given pressure in a unit of time in a 
 given type of boiler, under different conditions with regard to fuel, draught, rate of 
 combustion, and proportion of heating-surfaces, are to be calculated according to the 
 laws governing combustion and evaporation, and from the experimental data given in 
 chapters ii. and iii. of this treatise. 
 
 Marine engines consume from 20 to 30 Ibs. of steam per indicated horse-power per 
 hour, the latter quantity being consumed by engines using saturated steam of about 
 35 Ibs. pressure above the atmosphere, with a moderate rate of expansion, the cylinders 
 having no steam-jacket ; the former quantity is required for the best types of engine 
 using dry steam of from 60 to 80 Ibs. pressure above the atmosphere, working with a 
 high rate of expansion, the cylinders being provided with a steam-jacket. . A marine 
 boiler of ordinary type and proportions, using natural chimney-draught, produces 
 under these conditions with anthracite coal from 3.5 to 5.5 indicated horse-powers per 
 square foot of grate, with a free-burning, semi-bituminous coal from 4.5 to 7.5 indicated 
 horse-powers per square foot of grate. With forced draught as many as 10 indicated 
 horse-powers per square foot of grate have been developed by several large English 
 naval vessels of recent construction, during their full-power trial for six consecutive 
 hours at sea, by using from 25 to 30 Ibs. of a carefully-selected free-burning coal per 
 square foot of grate per hour. 
 
 In a large number of locomotive boilers, containing from 52 to 90 square feet of 
 heating-surface to each square foot of grate-surface, the rate of combustion increasing 
 from 43 to 126 Ibs. of coke with the above proportions of heating-surface, the average 
 evaporation was, according to D. K. Clark, 9 Ibs. of water, at the ordinary tempera- 
 tures and pressures, per pound of coke. 
 
 3. Various Types of Marine Boilers. Various types of marine boilers are dis- 
 tinguished either by the form of their shell or by the arrangement and position of their 
 flues or tubes. The class of "sectional" or "titbuloits boilers" will be considered 
 separately in section 10, chapter xi. 
 
 Boiler-shells are made rectangular or cylindrical, or present various combinations of 
 rectangular and cylindrical figures. 
 
 The rectangular shell possesses the great advantage that it is easily adapted to any 
 arrangement of the internal parts and to the space assigned to the boiler in the vessel. 
 
124 STEAM BOILERS. CHAP. VII. 
 
 The furnaces, connections, tubes, and the steam and water spaces may be arranged in 
 the most advantageous manner, as well with respect to evaporative efficiency as with 
 respect to accessibility and economy in weight and space. 
 
 The cylindrical shell has the advantages of strength and of simplicity of construction ; 
 but the circular form restricts within narrow limits the choice of the form, arrangement, 
 and proportions of the internal parts ; the steam-space is small relatively to the height 
 of the boiler, and the water-level is contracted ; much space in the vessel is wasted in 
 the spandrels formed by the shells of adjacent boilers. All these objectionable features 
 become more exaggerated as the diameter of the boiler decreases ; besides, with a dimin- 
 ished diameter the number of the boilers has to be increased, and consequently the 
 number of separate attachments, thus increasing the cost, the weight, and the liability 
 of the boilers to derangement. 
 
 The form given to the shell of boilers presents often a combination of rectangular 
 and cylindrical figures ; in this manner a compromise is effected between the respective 
 advantages and disadvantages of rectangular and cylindrical boilers. 
 
 Shells having an approximately oval cross-section have been extensively used of late 
 in English naval vessels, with steam-pressures of 60 or 70 Ibs. above the atmosphere. 
 The furnaces in these boilers are cylindrical, and when the larger diameter of the oval 
 shell is placed horizontally furnaces of larger diameter can be used than with circular 
 boilers of equal height. When the larger diameter of the oval shell is placed vertically 
 a larger and higher steam-space is obtained for the same amount of grate-surface. 
 With the oval shell the principal advantage of the circular shell is sacrificed viz., the 
 absence of bracing and the uniform distribution of the strain. 
 
 Boiler-shells have been formed of a number of circular arcs joined to one another 
 and tied together at their junction by braces, forming chords of these arcs. This sys- 
 tem, as developed by Charles E. Emery, is illustrated in figure 92, and the proper mode 
 of bracing such structures is discussed in section 7, chapter vi. This system enables us 
 to extend the boiler in the direction of its length and height indefinitely as in the 
 rectangular boiler, and independently of the radius of the circular arcs ; but the advan- 
 tages of the simple circular shell viz., absence of bracing, simplicity of construction, 
 and accessibility are completely sacrificed. (See section 1, chapter ix.) 
 
 The characteristic features of flue and tubular boilers will be discussed in sections 1 
 and 2, chapter xi., where illustrations of several kinds of marine flue-boilers will be 
 given. 
 
 Tubular boilers have superseded entirely flue-boilers for marine purposes ; but flues 
 are sometimes used in them in combination with tubes. 
 
SBC. 3. DESIGN, DRAWINGS, AND SPECIFICATIONS. 125 
 
 Tubular boilers may be divided into two grand groups viz., 1st, the water-tube type, 
 embracing all those boilers in which the larger portion of the heating-surface is arranged 
 in tubes containing the water and having their exterior surfaces acted upon by the 
 gaseous products of combustion ; 2d, the fire-tube type, embracing all those boilers in 
 which the larger portion of the heating-surface is arranged in tubes having their exterior 
 surfaces surrounded by the water and their interior surfaces acted upon by the gaseous 
 products of combustion, which pass through them on their way from the furnace to the 
 chimney. Each one of these groups may be subdivided according to the position of 
 the tubes, whether vertical, .inclined, or horizontal, and according to their location 
 viz., whether they are placed above, behind, or at the sides of the furnaces. The con- 
 siderations governing the choice of location and position of the tubes in marine boilers 
 are briefly stated in section 3, chapter xi. 
 
 " In the present state of marine steam-engineering the choice of boilers may be con- 
 sidered as restricted to the vertical water-tube and the horizontal fire-tube boilers, both 
 having their tubes arranged above their furnaces." ... " Recourse will be had to 
 other arrangements of tubes relatively to the furnaces only when considerations quite 
 independent of boiler-construction control ; as, for example, in light-draught vessels of 
 war, where it is a sine qua non that the entire boiler and its dependencies be placed 
 below the water-level. But wherever the choice is not thus trammelled one of the 
 above types will certainly be selected ; because, in a given space, both on the vessel's 
 floor and cubically, they allow the proper distribution and proportion of parts, and the 
 obtaining of the maximum economic and potential evaporation with the least weight, 
 cost, and external surface for radiation ; also, with the least weight of contained 
 water, and, when placed in pairs facing each other the fire-room being in common 
 with the least space for fire-room. These types are the most convenient, too, for repair, 
 examination, and sweeping, all of which can be done, without trouble or special provi- 
 sion, from the fire-room, whence access is easily had to the interior." (IsTierwood, 
 ' Experimental Researches," 1 vol. ii.) 
 
 Plates VI., VII., and XVII. represent the two kinds of boilers which were in gen- 
 eral use in United States naval vessels while the working pressure of steam did not 
 exceed 45 Ibs. per square inch above the atmosphere viz., the vertical water-tube 
 boiler of the Martin type, and the horizontal fire-tube boiler, both having a rectangular 
 shell and the tubes placed over the furnaces. The relative evaporative efficiency of 
 these two types of boiler, as determined by numerous experiments conducted under the 
 direction of the Bureau of Steam-Engineering of the United States Navy Department, is 
 as follows : When each boiler has 25 square feet of heating-surface per square foot of 
 
126 STEAM BOILERS. CHAP. VII. 
 
 grate-surface, a calorimeter equal to one-eighth of the grate-surface, and a chimney 60 
 feet high, the boiler being placed in the hold of the vessel and the air having to reach 
 the ashpits through restricted hatches from the upper deck, the maximum rate of com- 
 bustion with natural draught is for the horizontal fire-tube boiler 16 Ibs. of anthracite 
 per square foot of grate, and for the vertical water-tube boiler 12 Ibs. of anthracite per 
 square foot of grate ; and with these rates of combustion the former will furnish 4.36 
 per cent, more steam, but at the expense of 28 per cent, more fuel. With the above 
 proportions the space occupied by, and the weight of, the boiler proper and the watei 
 contained in it will each be a few per cent, less with the vertical water-tube than with 
 the horizontal fire-tube boiler. The latter has the practical advantages that the tubes 
 are more easily swept and that leaky tubes are easily plugged, even while the boiler is 
 in operation, and can be taken out and replaced without disturbing the bracing of the 
 boiler. 
 
 The boiler of the U. S. S. LacJcawanna (Plates VI., VII.), containing seven furnaces 
 with an aggregate grate-surface of 136.5 square feet, and having a length of 25 feet, is 
 as large as it is convenient to build rectangular boilers, on account of the limit imposed 
 by the size of the boiler-hatches. 
 
 Rectangular boilers have been built in some instances with two tiers of furnaces, 
 placed directly over one another ; each pair of furnaces discharging their gases into a 
 common back-connection and through a common set of tubes. This arrangement has 
 been adopted to augment the area of grate-surface contained within a single shell with- 
 out increasing the length and breadth of the boiler, when an increase of height was ad- 
 missible. This arrangement has, however, not given satisfactory results ; the upper 
 tier of furnaces obstructs the free escape of steam generated on the lower furnace- 
 crowns ; since a platform has to be built to fire the upper furnaces, fan-blowers are 
 required to furnish a sufficient air-supply to the lower furnaces and to ventilate the 
 lower fire-room. 
 
 With cylindrical shells vertical water-tubes cannot be used without increasing greatly 
 the amount of stayed surfaces. They may, however, be used advantageously in oval 
 shells in which the larger diameter forms the vertical axis, as the flat portions of the 
 shell may be tied directly to the flat vertical sides of the tube-boxes. 
 
 Fire-tubes, placed in the direction of the axis of the cylinder, can be arranged in the 
 most convenient manner, and are used almost exclusively, in cylindrical marine boilers. 
 
 In cylindrical boilers the diameter of the shell limits the number and the diameter 
 of the furnaces. In the type of boilers represented on Plates VIII., XI., and XII., 
 having a cylindrical shell and cylindrical furnaces, and horizontal fire-tubes arranged 
 
SEC. 3. DESIGN, DRAWINGS, AND SPECIFICATIONS. 127 
 
 above the furnaces, when the shell is from seven to eleven feet in diameter the number 
 of furnaces is generally two, and when the shell is from eleven to fourteen feet in dia- 
 meter the number of furnaces is generally three. When the working pressure of steam 
 is as high as 80 Ibs. above the atmospheric pressure the diameter of boilers seldom ex- 
 ceeds fourteen feet, on account of the difficulty of working the heavy plates required in 
 the construction of the shell. When the shell is less than seven feet in diameter two 
 cylindrical furnaces may be used with the tubes arranged behind the furnaces ; but with 
 return-tubes a single cylindrical furnace has to be used, and the tubes have to be ar- 
 ranged partly at the sides of the furnace. With the latter arrangement it is found that 
 the draught is relatively sluggish (see "Boiler Experiments," Franklin Institute Jour- 
 nal, March, 1879), and the furnace-crowns are inaccessible for cleaning unless the tubes 
 are removed ; in case the furnace is of large dimensions and the tubes are closely spaced 
 the circulation of the water is imperfect and the steam escapes with difficulty from the 
 furnace-crowns. This arrangement should be avoided in marine boilers where salt 
 water has to be used ; but it is often used to advantage in boilers of steam-launches and 
 similar small craft, supplied with fresh water from tanks and using a steam-blast, 
 because with this arrangement the bulk of the boiler and the weight of water contained 
 in it may be reduced considerably. (See Plate XVI.) 
 
 The boiler represented on Plate XV., having furnaces at both ends, illustrates a 
 method of increasing the grate-surface within a shell of a given diameter without af- 
 fecting the proportions of heating-surface, calorimeter, and steam-room to grate-sur- 
 face. Compared with two single-end boilers of equal diameter, grate-surface, and pro- 
 portion of internal parts, the bulk, weight, and cost of construction of the double-end 
 boiler is less, owing to the omission of the two back-heads and the attachment of braces 
 to them ; with the number of boilers the number of valves, pipes, and other apparatus 
 required for each separate boiler is likewise diminished; the total space in the vessel 
 occupied by double-end boilers is somewhat greater, because each end requires a sepa- 
 rate fire-room. 
 
 In order to reduce still more the length and the weight of double-end boilers the 
 water-space separating the back-connections of each pair of opposite furnaces is some- 
 times omitted, so that the two furnaces discharge their gases into a common back-con- 
 nection. But with this arrangement the action of the two currents of gas entering the 
 back-connection from opposite sides is prejudicial to -an active and reliable draught. 
 
 The boiler of the U. S. S. Daylight (figure 1, Plate III.) illustrates a type of boiler 
 frequently used on American steam-vessels. The front portion of the shell is made 
 rectangular in plan, with flat sides and a semi-cylindrical top ; the back portion is cylin- 
 
 Of THE 
 
 UNIVERSITY 
 
 OF 
 
128 STEAM BOILERS. CHAP. VII. 
 
 drical, with its top a horizontal continuation of the top of the semi-cylindrical front por- 
 tion. The cylindrical portions, which are of an oval cross-section in the present exam- 
 ple, are more frequently of a circular cross-section. The rectangular front admits of 
 the most advantageous proportions of the furnaces, and the use of the cylindrical form 
 for the rest of the shell simplifies the construction. A high steam-drum surrounding 
 the uptake affords additional steam-room and superheating surface. Flues of large 
 diameter extend from the combustion-chambers at the back of the furnaces to the back- 
 connections, and the horizontal return-tubes are likewise relatively of a large diameter 
 and of great length. This arrangement is favorable to a high rate of combustion ; the 
 draught is frequently forced by fan-blowers in this type of boiler. 
 
 Figures 2 and 3, Plate III., illustrate arrangements of the tubes adopted in marine 
 boilers where the height of the boilers has to be reduced at the expense of the room 
 occupied on the floor of the vessel. The lower flue in the boilers of the U. S. S. Ma- 
 TiasTta is sometimes omitted, the vertical water-tubes being arranged directly behind 
 the furnaces ; in other boilers the vertical water-tubes have been arranged at the sides 
 of the furnaces, the upper tube-sheet being on a level with the furnace-crown. A simi- 
 lar arrangement is sometimes adopted with horizontal fire-tubes. All these arrange- 
 ments present the advantages that the water-surface from which the steam escapes is 
 about twice as great as when the tubes are placed above the furnaces, and that the 
 steam generated on the furnace-crowns escapes freely, not having to pass through the 
 narrow water-spaces in or between the tubes. These advantages are of great importance 
 when a high rate of combustion is employed. 
 
 The type of boiler having horizontal fire-tubes placed directly behind the furnaces 
 is called the "locomotive type" because this arrangement is always adopted in the 
 boilers of locomotives. Figure 2, Plate III., illustrates this arrangement adapted to a 
 marine boiler with a rectangular shell, while Plates IV. and V. illustrate the same type 
 of boiler built for railroad purposes. In the latter the front portion of the shell is 
 always rectangular in plan, with flat sides, in order to get a roomy furnace, and the top 
 is generally made semi-cylindrical ; the back portion of the shell is cylindrical. The 
 tubes are of small diameter and great length, and are closely spaced in order to get a 
 large heating-surface within a small space. All the water-spaces are narrow ; the whole 
 boiler is made long but low ; its bulk and weight are reduced as much as possible. It 
 is designed to generate steam rapidly, and to have a high economic and potential evapo- 
 ration ; it is always worked with a steam-blast. 
 
 The same type of boiler is frequently used for marine purposes, in steam-launches 
 and similar small craft, when fresh water is used to feed the boiler. In case salt water 
 
SEC. 3. DESIGN, DRAWINGS, AND SPECIFICATIONS. 129 
 
 has to be used at times it is better to make the whole shell cylindrical, in order to re- 
 duce the stays and braces required to a minimum. 
 
 The vertical fire-tube boiler having a cylindrical shell (see figure 1, Plate XXVIII.) 
 is capable of a high rate of combustion, but its economic evaporative efficiency is 
 small, unless the tubes are made very long and the proportion of heating to grate sur- 
 face is unusually large. The water-level is carried some distance below the upper tube- 
 sheet, and the upper ends of the tubes and the uptake furnish superheating-surface. 
 The bulk and weight of this boiler are relatively small. This type is frequently used 
 for boilers of steam-launches and road-engines. It has also been used for large marine 
 boilers, but is not well suited for them, because the arrangement of the tubes makes the 
 interior of the boilers almost inaccessible, and the great" length of the tubes makes it 
 difficult to clean them or remove and replace them. 
 
 Dicker son's marine boiler had a rectangular shell ; inclined water- tubes were placed 
 directly over the furnace, and the products of combustion rising from the grate envel- 
 oped these tubes and then passed through vertical fire-tubes, placed above them in the 
 steam-space, into the uptake. The front and back of the boiler had large open- 
 ings, closed by cast-iron covers secured by bolts, to make the water-tubes accessible. 
 This boiler may be regarded as a connecting-link between the preceding types of 
 tubular marine boilers and the sectional water -tube boilers described in section 10, 
 chapter xi. With the rate of combustion usual in marine boilers the steam generated 
 in the inclined water-tubes escaped with difficulty, and the gaseous products of com- 
 bustion entered the uptake at a high temperature ; in consequence the tubes were soon 
 burnt, and the boilers have gone out of use after a short trial 
 
130 
 
 STEAM BOILERS. 
 
 CHAP. VII. 
 
 TABLE XXIL, 
 SHOWING THE DIMENSIONS, PROPORTIONS, AND WEIGHTS OF BOILERS OF VARIOUS TYPES. 
 
 
 
 z 
 
 3 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 Name of vessel. 
 
 Number of plate. 
 
 Description of boiler. 
 
 Number of boilers. 
 
 Total number of furnaces. 
 
 Total area of grate-surface in 
 square feet. 
 
 Square feet of water-heating 
 surface per square foot of 
 grate-surface. 
 
 Square feet of steam-super- 
 heating surface per square 
 foot of grate-surface. 
 
 Square ft. of grate-surface per 
 sq. ft. of cross-area through 
 tubes or flues for draught. 
 
 Square feet of grate-surface 
 per square foot of cross-area 
 of smoke-pipe. 
 
 U. S. S. Lackawanna 
 
 ( VI. ) 
 
 ^and \ 
 
 Rectangular shell ; vertical water-tubes ; 
 
 
 
 
 
 
 
 
 
 VII. 
 Ill 
 
 
 
 
 
 
 
 
 5- I' 
 
 
 
 
 
 o 
 
 
 
 
 7 018 
 
 
 " Plymouth . . 
 
 XVII 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5.870 
 
 
 
 " Daylight 
 
 III 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Kansas 
 
 in. 
 
 XXI. 
 
 Rectangular shell ; locomotive type 
 
 2 
 
 6 
 
 108.00 
 64 58 
 
 28.500 
 
 0.741 
 
 5-455 
 8.772 
 
 12.401 
 
 
 XXI. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " Nipsic 
 
 XII. 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 
 
 
 
 il Miantonomoh and class. . 
 
 VIII. 
 
 Cylindrical shell ; horizontal return fire- 
 tubes 
 
 6 
 
 18 
 
 
 
 
 
 
 
 XI. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ooo 
 
 600 
 
 
 u Lord of the Isles 
 
 XV. 
 
 
 
 
 3* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " Estelle 
 
 XXVII 
 
 
 
 
 18 i8 
 
 ^j 
 
 -_^-, 
 
 
 7 SAT 
 
 
 
 
 
 
 
 
 
 
 
 * Two telescopic smoke-pipes. 
 
 t Water-level 9 inches above tubes. 
 
Szc. 3. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 131 
 
 TABLE XXII. (Continued.) 
 SHOWING THE DIMENSIONS, PROPORTIONS, AND WEIGHTS OF BOILERS OF VARIOUS TYPES. 
 
 9 
 
 10 
 
 II 
 
 12 
 
 13 
 
 M 
 
 15 
 
 16 
 
 17 
 
 18 
 
 19 
 
 20 
 
 21 
 
 23 
 
 23 
 
 S 
 
 .5 
 
 M4 
 
 So 
 
 1 
 
 t* r w 
 
 c 
 
 . 
 
 .s 
 
 c 
 
 *o 
 
 s 
 
 C-s 
 
 o 
 
 6-S 
 
 Height of smoke-pipe abo 
 'level of grate, in feet. 
 
 Capacity of steam-room 
 boiler-shell, in cubic feet 
 
 Capacity of steam-room 
 steam-drums, in cubic fci 
 
 II 
 II 
 
 S 
 
 *i 
 
 MJ 
 
 a si 
 
 l rf 
 
 Total capacity of steam-roo 
 in cubic feet. 
 1 
 
 Weight of boilers, includl 
 doors, uptakes, plates, et 
 but excluding smoke-pi 
 and grate-bars. 
 
 1 
 
 ! 
 
 (ft 
 
 *o . 
 
 at 
 
 _-a 
 
 If 
 fl 
 
 "i 
 1 
 
 "3 
 .4 
 
 .-I! 
 
 a 
 
 1 
 
 1 
 
 o 
 
 2 
 
 fj 
 
 *\ 
 
 si 
 1* 
 
 Weight of water in boilers 
 pounds. 
 
 o. 
 2 
 
 o 
 
 V 
 
 Sj 
 
 "o.e 
 ll 
 
 * 
 
 1 
 
 2 
 S=* 
 
 S 
 
 ^"V 
 
 e 
 
 }] 
 
 u " 
 
 Length occupied by boiU 
 and fire-room in the vess 
 
 Width occupied by boilers a 
 fire-room in the vessel. 
 
 Working pressure of stea 
 in pounds per square in 
 above the atmosphere. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 59.50 
 
 866 
 
 
 .... 
 
 866 
 
 93,000 
 
 5,30 
 
 8,590 
 
 106,890 
 
 68,500 
 
 8.00 
 
 
 29.00 
 
 20.166 
 
 40 
 
 56.00 
 
 I loot 
 
 .... 
 
 K4 
 
 324 
 
 265,000 
 
 .3fa* 
 
 20,000 
 
 306,362 
 
 H5,ooot 
 
 9.00 
 
 .... 
 
 38-50 
 
 27.500 
 
 40 
 
 43.50 
 
 670 
 
 130 
 
 
 800 
 
 
 
 
 
 
 10.66 
 
 18.00 
 
 
 
 25 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 653 
 
 
 
 873 
 
 
 
 
 
 
 
 
 
 
 
 
 760 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 53.50 
 
 740* 
 
 140 
 
 20 
 
 900 
 
 185.000 
 
 14,150 
 
 IO,OOO 
 
 209,350 
 
 78,360$ 
 
 9.00 
 
 9.00 
 
 3I-50 
 
 28.25 
 
 so 
 
 .... 
 
 '5*) 
 
 300 
 
 135 
 
 2024 
 
 374.000 
 
 7,800 
 
 24,000 
 
 405,800 
 
 igi^ooot 
 
 12.13 
 
 12.75 
 
 38.00 
 
 34-00 
 
 so 
 
 
 35 
 
 60 
 
 
 
 
 
 
 
 8 740 
 
 
 
 15.6 
 
 8.00 
 
 So 
 
 
 680 
 
 
 
 
 
 
 
 
 
 
 27.75 
 
 
 28.500 
 
 
 28.00 
 
 i7- 2 4i 
 
 9-4 
 
 
 36.64 
 
 
 
 
 
 
 16,500 
 
 1. 100 
 
 11.25 
 
 
 18.00 
 
 8.250 
 
 200 
 
 Water-level 6 inches above tubes. 
 
 { Half capacity of coil. 
 
132 
 
 STEAM BOILERS. 
 
 CHAP. VII. 
 
 TABLE XXIII. 
 
 SHOWING THE ECONOMIC EVAPORATION OF BOILERS OF VARIOUS TYPES, WITH DIFFERENT RATES 
 OF COMBUSTION, AND WITH DIFFERENT PROPORTIONS OF CALORIMETER, HEATING-SURFACE, 
 AND GRATE-SURFACE. 
 
 
 I 
 
 a 
 
 3 
 
 4 
 
 S 
 
 6 
 
 7 
 
 8 
 
 9 
 
 IO 
 
 
 V 
 
 
 i"o 
 
 sil 
 
 E " v 
 
 ES" 
 
 a ii^jj 
 
 "3.5 
 
 5 
 
 
 
 I. 
 
 
 </] o 
 
 If 
 
 rt A 
 
 rt J3 fO 
 3 C ** 
 
 III 
 
 C E'o 
 
 v! 
 
 Ha 
 
 lai-j 
 
 1| 
 3 
 
 ll 
 || 
 
 C 
 
 
 u rt 
 
 
 V 3 
 
 ||j 
 
 2 |- 
 
 |12 
 
 4)-C o 
 
 t. a.0^ y 
 
 2 = 
 
 & 
 
 1 
 
 Name of vessel. 
 
 * t!i 
 
 Jl 
 
 Type of boiler. 
 
 -sj 
 
 *- ii * bo 
 
 S o 
 
 C o- 
 rt o 
 
 
 fill 
 
 ||| 
 
 *o 
 
 II 
 
 1 
 
 
 O.l 
 
 i? 
 
 
 *2i o w 
 
 t) " 4> 
 
 |!|| 
 
 fll 
 
 llil 
 
 1"= =-2 
 
 111 
 
 u > 
 
 2j 
 
 E o 
 
 IM 
 
 O 
 
 ii 
 
 
 3.0 
 
 
 
 
 
 
 g 3 S 
 
 
 V U 
 
 a 
 
 
 
 
 
 V) 
 
 t2 
 
 0. 
 
 
 
 
 (2 
 
 H 
 
 Q 
 
 * U. S. S. Mahaska. 
 
 III. 
 
 Verti'l water-tubes 
 
 29.275 
 
 7.018 
 
 5-505 
 
 0.228 
 
 13.873 
 
 93.04 
 
 298.2 
 
 I86 3 
 
 
 
 (Bartol's patent). 
 
 
 
 
 
 
 
 
 
 * " Daylight... 
 
 III. 
 
 Horizontal return 
 
 30.355 
 
 7.870 
 
 6.5II 
 
 0.214 
 
 11.199 
 
 75-10 
 
 280.1 
 
 I86 3 
 
 
 
 fire-tubes. 
 
 
 
 
 
 
 
 
 
 * " " ... 
 
 11 
 
 41 
 
 30.355 
 
 7.870 
 
 8.270 
 
 0.272 
 
 11.879 
 
 79-66 
 
 261.0 
 
 1864 
 
 * " Kansas... 
 
 III. 
 
 Locomotive type. . 
 
 28.500 
 
 5-455 
 
 14.103 
 
 0.596 
 
 10.196 
 
 68.38 
 
 495-7 
 
 1863 
 
 * ii ii 
 
 " 
 
 " 
 
 28.500 
 
 6.601 
 
 13.124 
 
 0.552 
 
 10.790 
 
 72.36 
 
 466.9 
 
 1863 
 
 * ii ii 
 
 " 
 
 " 
 
 28.500 
 
 8.148 
 
 13.176 
 
 0.565 
 
 10.789 
 
 72-35 
 
 535-6 
 
 1863 
 
 * ii ii 
 
 " 
 
 " 
 
 28.500 
 
 10.313 
 
 10.754 
 
 0.464 
 
 11.570 
 
 77-59 
 
 497-5 
 
 1863 
 
 * " Shockokon 
 
 XXI. 
 
 
 21.058 
 
 8.772 
 
 15 209 
 
 O.721 
 
 8673 
 
 58.16 
 
 
 1864 
 
 * ii ii 
 
 
 
 21.058 
 
 "//** 
 
 8.772 
 
 8.843 
 
 " i " j 
 .0.420 
 
 8.660 
 
 jw. j 
 
 58.08 
 
 .... 
 
 * ****^ 
 
 1864 
 
 * " Morse 
 
 XXI. 
 
 Double-return 
 
 21.901 
 
 9-357 
 
 9-372 
 
 0.428 
 
 10.257 
 
 68.78 
 
 
 1863 
 
 
 
 drop-flues. 
 
 
 
 
 
 
 
 
 
 f S. S. Estelk .... 
 
 XXVII. 
 
 Herreshoff coil- 
 
 13.283 
 
 1.511$ 
 
 10.698 
 
 0.805 
 
 9-938 
 
 66.64 
 
 .... 
 
 1877 
 
 
 
 boiler. 
 
 
 
 
 
 
 
 
 
 * Isherwood, 'Experimental Researches/ vol. ii. t * Report of Board of United States Naval Engineers,' 1878. 
 
 $ Proportion of grate-surface to cross-area of chimney = 7.811 to i. 
 
Sue. 4. DESIGN, DRAWINGS, AND SPECIFICATIONS. 133 
 
 4. Space and Weight required for Boilers of a given Power (Isfterwood, 
 'Experimental Researches ,' vol. ii.) 
 
 " In a steamer a certain space is to be allotted to the boiler, fire-room, and coal for a 
 given speed to be maintained a given time ; and the problem is so to apportion them 
 that this space shall be a minimum. To fulfil the speed condition a certain weight of 
 steam per hour must be furnished. Now, if a boiler with a high rate of combustion be 
 employed a less space will be required for it and its fire-room, but more space will be 
 required for the coal, as the economic evaporation will be less ; and, vice versa, if a 
 boiler with a low rate of combustion be employed a greater space will be required for 
 it and its fire-room, but less space will be required for the coal, on account of its greater 
 economic evaporation. It is evident there is a point where the aggregate space is a 
 minimum. If the cost of fuel be considered an item of importance, it extends the prob- 
 lem, in a commercial vessel, to whether it is not advantageous to employ a boiler of still 
 lower rate of combustion and higher economic efficiency, and take the additional space 
 required for it from the portion allowed to cargo. In a war-steamer this additional 
 space would be at the expense of its military power. 
 
 " In the following table will be found the solution of the above problem for the hori- 
 zontal fire-tube boiler having the tubes above the furnaces. The determination is made 
 for the case in which the vessel is to carry a supply of fuel sufficient for 200 hours' 
 maximum steaming. . . . Further, the determination is made both for aggregate 
 space occupied and weight carried, for the fuel, and the boilers and fire-room. 
 
 "In making these calculations the boiler is assumed to be 10 feet 3 inches long at 
 the level of the grates, 9 feet 9 inches high, and of sufficient width or frontage to give 
 the requisite grate-surface with furnaces 3 feet wide, containing grates 6 feet 6 inches 
 long. The fire-room is assumed to extend in the fore and aft direction of the vessel, 
 and to be 8 feet 6 inches wide, the boilers being arranged on each side of it in oppo- 
 site pairs. 
 
 "The ratio of the heating to the grate surface is taken at 25 to 1, and the calori- 
 meter at one-eighth of the grate-surface. 
 
 "The thickness of the plate for the bottom of the boiler, and for the furnaces and 
 the ashpits, is taken at f of an inch, for the tube-plates at inch, and for all other 
 parts at ^ of an inch ; all seams are taken to be double-riveted, and the boiler to 
 be braced for a working pressure of 40 pounds per square inch above the atmos- 
 phere. . . . 
 
 " The weights taken for the boiler are the actual weights as determined by weighing 
 exactly such boilers after completion. These weights include not only the boiler pro- 
 
134 STEAM BOILERS. CHAP. VH. 
 
 per, but everything appertaining to it, as grate-bars, smoke-pipe, doors, plates, valves, 
 pipes, felt and sheet-lead covering, floor-plates of fire-room, and water in boilers. . . . 
 
 "The fuel is to be anthracite, with one-sixth of refuse, and the space occupied by 
 every 53 pounds of it to be 1 cubic foot, which is the average of bunker-stowage with 
 this coal. . . . The maximum weight of steam to be furnished per hour is taken at 
 60,000 pounds. . . . 
 
 " We find, on examining the column of space occupied by the aggregate boiler, fire- 
 room, and anthracite, . . . that, leaving wholly out of view the economic evaporation 
 by the anthracite, the best rate of combustion for obtaining in a given space the great- 
 est quantity of steam per hour during 200 hours is 13 pounds of anthracite per square 
 foot of grate-surface per hour. ... If some importance be given to the economy of the 
 
 fuel we perceive that, by reducing the rate of combustion to 10 pounds of anthracite, 
 
 (Q oca _ g 23g \ 
 
 g OQQ X 100 ) = 9.14 per centum 
 
 . /59001 - 55826 \ 
 
 by increasing the space occupied ( -- 55326 ~ x / = P er centum tne 
 
 minimum aggregate weights of boiler and its appurtenances, and of the anthracite, cor- 
 respond to a rate of combustion of 11 pounds of anthracite per square foot of grate- 
 surface per hour." It is evident that when the maximum supply of steam is to be 
 furnished for a greater or less length of time, the weight and space occupied by the 
 boiler and fire-room remains the same for the respective rates of combustion, while that 
 of the anthracite increases or decreases, as the case may be. 
 
SEC. 4. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 135 
 
 TABLE XXIV. 
 
 EXHIBITING THE SPACE AND WEIGHT REQUIRED WITH THE HORIZONTAL FIRE-TUBE BOILER 
 HAVING A RECTANGULAR SHELL AND THE TUBES ARRANGED ABOVE THE FURNACES, AND 
 WITH ANTHRACITE WITH ONE-SIXTH REFUSE, TO FURNISH A GIVEN SUPPLY OF STEAM PER HOUR 
 FOR 200 HOURS WITH DIFFERENT RATES OF COMBUSTION. 
 
 "So |jSJj i 
 
 s- -. 3 
 
 Space. 
 
 Weight. 
 
 3O 3 * 
 
 
 
 
 S.2 
 
 O 
 
 ~S_0 . 
 
 
 iASfe 
 
 g-si." 
 
 I j "T-i r- * 
 
 3 * i~ r * 
 
 "3' -IE" 
 
 " > 
 
 S"g 
 
 ,s 
 
 ft 
 
 Mil 
 
 > 5JJ= 
 
 1 
 
 slfl 
 
 1*5 
 
 2 = 3 3 
 
 y rt i;- rt 
 
 1 *-^ J 7.-= 
 
 lis-s 
 
 a-S a s 
 
 ?!! 
 
 il< 
 
 - 3^ 
 
 iN 
 
 S2-=^ 
 * 3 K-o 
 t~"c rt = 
 
 e ii 
 
 || 
 
 *2 
 
 ffi 
 
 i a il 
 
 SiJll 
 
 ""- c e '"!! - Z O.3 
 
 ssis^5-s -si -a 
 lltflj'll fls' 5 - 
 
 ill 
 
 M 
 
 111 
 
 *s*i u 
 
 till ii 
 
 lilli 
 
 lilli 
 
 illlll 
 
 3 ccn a^5-o 
 
 l^illi] 
 
 w ^ o 
 
 |ij 
 
 - 
 
 0, 
 
 z 
 
 u 
 
 u 
 
 U 
 
 n 
 
 * 
 
 ^ 
 
 6 
 
 9.383 
 
 7 
 1,065.72 
 
 54,240 
 
 23,979 
 
 78,219 
 
 1,231,866 
 
 1,278,880 
 
 2,510,746 
 
 7 
 
 9.338 
 
 917.91 
 
 46,717 
 
 24,095 
 
 70,8l2 
 
 1,061,012 
 
 1,285,067 
 
 2,346,079 
 
 8 9.258 
 
 8io.ii 
 
 41,230 
 
 24,303 
 
 65,533 
 
 936,406 
 
 1,296,160 
 
 2,232,566 
 
 9 9-I50 
 
 728.60 
 
 37,082 
 
 24.590 
 
 61,672 i 842,189 
 
 1,311,467 
 
 2,153,656 
 
 10 
 
 8.989 
 
 667.48 
 
 33.971 
 
 25,030 
 
 59,001 
 
 771,540 
 
 1,334.933 
 
 2,106,473 
 
 ii 
 
 8.775 
 
 620.32 
 
 31,627 
 
 25,641 
 
 57,268 
 
 717,028 
 
 1,367,520 
 
 2,084,548 
 
 12 
 
 8.524 
 
 586.58 
 
 29.907 
 
 26,396 
 
 56,303 
 
 678,028 
 
 1,407,787 
 
 2,085,815 
 
 13 
 
 8.238 
 
 560.25 
 
 28,514 
 
 27,312 
 
 55,826 
 
 647,593 
 
 1,456,640 
 
 2,104,233 
 
 14 
 
 7-934 
 
 540.17 
 
 27,492 
 
 28,359 
 
 55,851 
 
 624,382 
 
 1,512,480 
 
 2,136,862 
 
 15 
 
 7.621 
 
 524-87 
 
 26,713 
 
 29,524 
 
 56,237 
 
 606,697 
 
 1,574,613 
 
 2,181,310 
 
 16 
 
 7-343 
 
 510.69 
 
 25,991 
 
 30,641 
 
 56,632 
 
 590,306 
 
 1,634,187 
 
 2,224,493 
 
 17 
 
 7.111 
 
 496.33 
 
 25,261 
 
 31,641 
 
 56,902 
 
 573,708 
 
 1,687,520 
 
 2,261,228 
 
 18 
 
 6.887 
 
 484.00 
 
 24,633 
 
 32,670 
 
 57,303 
 
 559,456 1,742,400 
 
 2,301,856 
 
 19 
 
 6.690 
 
 472-03 
 
 24,024 
 
 33.632 
 
 57,656 
 
 545,619 1,793,707 
 
 2,339,326 
 
 20 
 
 6-547 
 
 458.23 
 
 23,322 
 
 34,367 
 
 57,689 
 
 529,668 
 
 1,832,907 
 
 2,362,575 
 
 21 
 
 6404 
 
 446.15 
 
 22,707 
 
 35.134 
 
 57,841 
 
 515,705 
 
 1,873,813 
 
 2,389.518 
 
 22 
 
 6.297 
 
 433.17 22,046 
 
 35.731 
 
 57,777 
 
 500,701 
 
 1,905,653 
 
 2,406,354 
 
 23 
 
 6.190 
 
 42144 21,449 
 
 36,349 
 
 57,798 
 
 487,142 1,938,613 
 
 2,425,755 
 
 24 
 
 6.100 
 
 409.84 20,859 
 
 36,885 
 
 57,744 
 
 473,734 
 
 1,967,200 
 
 2,440,934 
 
136 STEAM BOILERS. CHAP. VII. 
 
 5. Proportioning the Parts of a Boiler. In the following special regard has 
 been had to the ordinary type of marine boilers, with horizontal fire-tubes or vertical 
 water-tubes arranged over the furnaces, and burning anthracite coal with natural chim- 
 ney-draught, unless the conditions are otherwise specified. 
 
 The length and the width of the grate are limited to such dimensions as will permit 
 the proper management of the fire, especially the cleaning of the back and of the 
 front corners ; on this account the length should never exceed 7 feet nor the width 42 
 inches ; the grate-surface in each furnace ranges generally between 18 and 21 square 
 feet. The grate slopes downward from the front to the back, at the rate of one inch or 
 1J inch to the foot in the length of the grate ; this arrangement facilitates the firing at 
 the back, and makes the furnace roomier at the same time. 
 
 The ashpit, even when partly obstructed by ashes, must admit a sufficient quantity 
 of air, moving at a low velocity, to every part of the grate. This condition will gene- 
 rally be fulfilled when the ashpit is made roomy enough to permit the working of the 
 fire from below the grate. With dimensions of the grate as given above, the height of 
 the ashpit-opening in front varies from 15" to 18", while its width is made equal to that 
 of the furnace. 
 
 The furnace must have sufficient height above the grate to afford room for the gases 
 to mingle thoroughly and to permit a proper working of the fire. The height of the 
 furnace-crown above the grate in marine boilers burning anthracite coal, with natural 
 draught, averages from 18 to 24 inches. High rates of combustion necessitate an in- 
 crease in the height of the furnace ; in locomotive boilers the furnace is often made 48 
 inches high. Bituminous coals require a larger combustion-chamber than anthracite. 
 
 The calorimeter oner the bridge-wall at the back end of the grate should be 
 made as small as is consistent with the desired rate of combustion, in order to in- 
 crease the velocity of the gases to a maximum at this point, and cause them to mingle 
 thoroughly as they emerge into the combustion-chamber or back-connection. The area 
 of this passage is made from | to ^ of the area of the grate in marine boilers using natu- 
 ral draught ; when forced draught is used, and the rate of combustion is increased, the 
 area of this passage is likewise to be increased. The opening should extend, if possible, 
 the whole width of the grate, the area being regulated by the height. 
 
 The ~back smoTce-connection should be as spacious as possible to afford the gases 
 room and time to complete their combustion before entering the tubes. This condition 
 will generally be fulfilled when sufficient room is provided to admit a man for making 
 examinations and repairs, expanding the tube-ends, calking, etc. ; the width of this 
 chamber ranges generally between 18 and 30 inches. 
 
SEC. 5. DESIGN, DRAWINGS, AND SPECIFICATIONS. 137 
 
 The calorimeter through or between the tubes varies from \ to \ of the area of the 
 grate-surface, the larger area being used for the higher rates of combustion ; the dia- 
 meter of the tubes depends to a great extent on the amount of heating-surface that is 
 to be placed into the allotted space. The diameter of horizontal fire-tubes varies from 
 2 to 4 inches ; smaller tubes are liable to become choked with soot and ashes, unless 
 a very strong draught is produced by a blast. Vertical water-tubes are commonly made 
 2 inches in diameter, and are spaced so as to allow from 1 inches to If inches of clear 
 space for the passage of the gases. 
 
 The calorimeter of the smoke-pipe varies from \ to of the area of the grate. When 
 several furnaces empty their gases into a common chimney, care must be taken that the 
 calorimeter of the uptake at different points is proportional to the volumes of gas that 
 have to pass such points. 
 
 The total heating-surface of marine boilers using natural draught is made from 25 
 to 35 times the area of the grate-surface for rates of combustion ranging between 12 Ibs. 
 and 22 Ibs. of coal per square foot of grate per hour. When forced draught is em- 
 ployed the heating-surface must be increased in proportion to the increased rate of com- 
 bustion ; in locomotives, burning as much as 120 Ibs. of coke per square foot of grate 
 per hour, the heating-surface is about 90 times the area of the grate-surface. The eva- 
 porative efficiency of the different heating-surfaces in a boiler varies greatly, as has 
 been shown in chapter iii. The heating-surfaces must be arranged in such a manner as 
 to allow the steam to escape from them as soon as formed ; horizontal water-tubes are 
 therefore to be avoided, as well as extensive flat surfaces for the bottom of flues and 
 smoke-connections. By rounding the corners of internal square passages with a large 
 radius the free escape of the steam and the circulation of the water will be greatly faci- 
 litated, while the strength of those parts is increased at the same time. A relatively 
 small portion of the heating-surface in the uptake of every boiler passes through the 
 steam-space ; it is often found advantageous to increase this superheating-surface by 
 means of various contrivances that will be described in a subsequent chapter. (See 
 section 3, chapter xiii.) 
 
 The water-spaces surrounding the furnace and the smoke-connections should never 
 be less than 4 inches in the clear, and, if possible, should be made 5 inches wide in 
 marine boilers. Sufficient room must be left between the furnace and the tubes to ad- 
 mit a man to the interior of the boiler to scale the crown-sheet of the furnace and to 
 make repairs. Manholes provided for this purpose in the front of the boiler are oval 
 in shape, of 15* X 12* diameter ; the smallest admissible size is 13* x 11*. 
 
 The clear space left for this purpose above the furnace is also necessary on account 
 
138 STEAM BOILERS. CHAP. VII. 
 
 of the rapid formation of steam on the furnace-crown, as it facilitates the free circulation 
 of the water ; for the same reason it is not advisable to make the furnaces of marine boil- 
 ers very wide, especially when the tubes are placed rather close to the furnace-crown. 
 
 Horizontal fire- tubes returning over the furnace should be spaced with at least 1 inch 
 of clear space between them in a horizontal direction, and must be placed in vertical 
 rows, in order to offer the least obstruction to the rising steam-bubbles. It is advisable 
 to leave larger spaces, from 5 to 7 inches wide, between the tube-rows at intervals, for 
 the passage of the descending currents of water. 
 
 The water must be carried at such a height above the back-connection and the tubes 
 that they are not bared too readily through irregularities in the admission of the feed- 
 water, excessive foaming, or the rolling of the vessel. With horizontal fire-tubes the 
 water-level should be carried not less than 6 inches above the back-connection. The 
 area of the water-level should always be as large as possible, to prevent foaming and 
 the lifting of water. 
 
 The capacity of the steam-space depends on the number of cubic feet of steam re- 
 quired by the engine in a unit of time. It should be of sufficient height to prevent the 
 lifting of the water into the steam-pipe. (See section 1, chapter xiii.) 
 
 6. Relative Value of various Forms for Boiler-construction. Although the 
 spherical shell possesses greatly superior strength over all other forms of structure, 
 and has the additional advantage of accommodating itself to the expansive action of 
 heat without distortion, its employment in boiler-construction is very limited on account 
 of the comparative difficulty of giving this form to the material used. The property of 
 the sphere of presenting less surface than any other body in proportion to its content is 
 an objection to its use in connection with the heating-surfaces of a boiler, and its use 
 for external shells would entail a great loss of room in waste spaces. The ends of 
 cylindrical shells of land boilers and the top of steam-drums are often made of spheri- 
 cal segments ; spherical strengthening-domes are sometimes attached to flat surfaces 
 where staying by rods or gussets is inexpedient. The Harrison boiler is composed 
 almost entirely of small cast-iron spheres connected by short tubes. 
 
 The cylindrical form has the most extensive application in boiler-making. It pos- 
 sesses next to the sphere the greatest strength ; it is readily produced of any required 
 size and of all materials used in boiler-making. Since with cylindrical forms stays and 
 braces can be dispensed with, their use makes a boiler accessible and cheapens its 
 construction. 
 
 The use of flat surfaces is in many cases unavoidable for an economic utilization of 
 the space allowed and for the proper arrangement of the interior parts of a boiler. The 
 
SBC. 7. DESIGN, DRAWINGS, AND SPECIFICATIONS. 139 
 
 heads of the shell, the tube-sheets, and the back of the back-connection are always 
 made flat in cylindrical marine boilers of the usual type. When the steam-pressure is 
 not to exceed 45 Ibs. per square inch above the atmosphere the outer shell of marine 
 boilers is generally made box-shaped, with square sides and the corners more or less 
 rounded ; by using this form of boiler the space in a vessel allotted to it can be most 
 fully utilized, and the proportions of the furnaces can also be more satisfactorily 
 arranged than with the cylindrical boiler. On this account, in boilers designed to carry 
 steam of high pressure, especially in locomotive boilers, the front part, containing the 
 furnace, is often made square below and semi-cylindrical on top, while the body of the 
 boiler, containing the flues or tubes, is made a complete cylinder ; or sometimes the 
 whole boiler is made oval in shape. 
 
 7. Factor of Safety. All parts of a boiler must possess equal strength. The 
 dimensions of the various parts of a boiler must not be calculated by applying a uni- 
 form factor of safety to the formulas expressing the strength of the respective shapes, 
 but account must be taken of the continual waste taking place in all parts of a boiler in 
 consequence of external and internal corrosion and the wasting effect of the intense 
 heat produced in the furnace. Since the deterioration resulting from these causes 
 affects different parts and forms in a different degree, the allowance of additional thick- 
 ness of metal to be made on this account must vary accordingly. Care must also be 
 taken that such parts as cannot be readily replaced or repaired be proportioned with 
 an extra large margin of strength. 
 
 The plates of boilers must be proportioned and stayed not only with regard to 
 strength, but sufficient stiffness must be given to them to prevent changes of shape 
 under pressure. A change of shape in one direction by pressure, and returning again 
 to its original position when the pressure is released, will sooner or later result in a 
 crack. The same is true when braces are attached in such a way that the sheet is 
 drawn from its true position. 
 
 The tensile stress exerted by the maximum steam-pressure on any part of a boiler 
 should not exceed one-sixth of its ultimate strength. This factor of safety is usually 
 employed for parts of machinery subjected to alternating stresses acting in opposite 
 directions. The steam-pressure in a boiler cannot be considered as a quiescent load, on 
 account of the constantly occurring, and sometimes considerable, fluctuations of pressure 
 due to various causes ; besides, the different parts of a boiler are subjected to continual 
 expansions and contractions owing to changes of temperature, the effect of which can- 
 not be calculated, but is very marked under certain conditions. The force exerted by 
 expansion or contraction as the effect of change of temperature is equal to that which 
 
140 STEAM BOILERS. 
 
 CHAP. VII. 
 
 would be required to elongate or compress the material to the same extent by mechani- 
 cal means. The linear expansion of ordinary wrought-iron plates is .0000064 of their 
 length for each degree Fahrenheit of increase of temperature, and the same elongation 
 is produced by a stress of about 150 Ibs. per square inch of section of metal. It must 
 be observed that this stress produced by increase of temperature is independent of the 
 sectional area of the plate, and if the expansion or contraction of the plate is not 
 allowed to go on freely a corresponding stress will be exerted on the metal by a change 
 of temperature. 
 
 In case the substitution of mild, ductile cast-steel for piled wrought-iron plates for 
 boilers should be warranted by further practice, it may be considered safe to decrease 
 the factor of safety to four, on account of the greater homogeneousness and uniformity 
 of quality of the steel plates rolled from single ingots. 
 
 It must be remembered that the strength of any structure is to be measured by that 
 of its weakest part, which in the case of boilers is the joint where the sheets are con- 
 nected. The strength of various forms of joint employed in boiler-making will be dis- 
 cussed in the next chapter. 
 
 The ' Eevised Statutes of the United States ' prescribe the following rule regarding 
 the factor of safety to be employed in determining the strength of marine boilers : 
 " Section 4433. The working steam -pressure allowable on boilers constructed of plates 
 inspected as required by this title, when single-riveted, shall not produce a strain to 
 exceed one-sixth of the tensile strength of the iron or steel plates of which such boilers 
 are constructed ; but where the longitudinal laps of the cylindrical parts of such boilers 
 are double-riveted, and the rivet-holes for such boilers have been fairly drilled instead 
 of punched, an addition of twenty per centum to the working-pressure provided for 
 single-riveting may be allowed : Provided, That all other parts of such boilers shall cor- 
 respond in strength to the additional allowances so made, and no split-calking shall in 
 any case be permitted." 
 
 In the case of large cylindrical flues subjected to compression the factor of safety 
 should be increased to eight at least when Fairbairn' s formula is employed ; in addition 
 to this the thickness of the metal must be increased tt or i i ncn to allow for corrosion 
 and other wasting influences. To allow for corrosive and other destructive influences 
 in the case of the rectangular boiler, the lower parts of the outer shell, the water- 
 legs, and the bottom of the back-connections are generally made from \ to $ inch 
 thicker than the other parts of the shell ; the parts exposed to an intense heat are less 
 increased in thickness, on account of the liability of thick plates to blistering ; there- 
 fore the furnaces and the sides, tops, and fronts of the back-connections receive an in- 
 
SEC. 8. DESIGN, DRAWINGS, AND SPECIFICATIONS. 141 
 
 crease of ^ *&<& O10 ^7 5 tne tube-sheets, on the contrary, are still further increased in 
 thickness, in order to allow for the loss of stiffness in consequence of the many large 
 holes drilled in them, and to give a sufficient bearing-surface to the tubes. 
 
 Stays and braces are generally proportioned to bear a strain of from 4,000 to 5,000 
 Ibs. per square inch of section ; a more rational method of proportioning them, how- 
 ever, is to calculate the required cross-section, according to the pressure on the surface 
 which they have to support, using six as a factor of safety, and adding a certain amount 
 to the thickness or diameter thus found to allow for corrosion. Generally it will be 
 sufficient to add i inch to the thickness required for strength ; but near the water-level 
 and in narrow water-spaces the increase in thickness has to be greater. Welded parts 
 must likewise receive an additional increase, since the metal is there more easily at- 
 tacked by corrosion and the strength of welded joints varies greatly. 
 
 8. Drawings and Specifications. After the dimensions and the general shape 
 and arrangement of a boiler have been decided upon, the exact forms and proportions 
 of the various parts can be most advantageously determined by making a drawing of 
 the boiler, showing front and side elevations, and a plan, including such full or partial 
 sectional views as are required for a complete illustration of every part of the boiler. 
 On the drawing all dimensions necessary to guide the boiler-maker in the construction 
 of the boiler must be plainly marked, including the thickness of the sheets, the size and 
 position of all openings for steam, feed and blow-off pipe connections, and for man 
 and handhole plates, safety-valves, etc., the position and attachment of the braces and 
 stays, and the location and form of the principal seams. This drawing is generally 
 made to a scale of I inch or \\ inches to the foot. Drawings on a larger scale are made 
 to show in detail the devices adopted for bracing different parts of the boiler, and the 
 various attachments of the boiler, as well as the arrangement of the uptake and chim- 
 ney. In order to determine all these details intelligently it is advisable to lay down a 
 general plan on a scale of 1 inch or f inch to a foot, showing the location of the boilers 
 in the vessel with regard to floors, keelsons, decks, beams, masts, and bulkheads, the 
 mode of securing the boilers, their connections and attachments, the position of valves 
 and pipes, also the ventilators, doors, stairs, ash-hoisters, and other arrangements con- 
 nected with the fire-room and used in the manipulation of the boilers. 
 
 When the boilers are to be built by contract specifications are written, giving such 
 instructions regarding Ihe construction of the boiler as have not been illustrated with 
 sufficient clearness in the drawings ; it is necessary to specify distinctly the quality of 
 the material to be used foi different parts of the boiler, and the workmanship. Speci- 
 mens of specifications are appended to this chapter. - 
 
142 STEAM BOILERS. CHAP. VII. 
 
 The sheets are ordered from the maimfacturer as nearly as possible of the shape and 
 size required, proper allowances being made for laps and flanging and for planing the 
 sides and ends. 
 
 A list is made of all sheets required, showing their dimensions, their weight, and the 
 part of the boiler for which they are to be used ; a sketch, showing the exact shape and 
 dimensions of all sheets that are not rectangular or circular, accompanies this list. In 
 laying off the size of the sheets care must be taken to get as few seams as possible in 
 contact with the fire, and to get them in the most convenient position for riveting and 
 accessible for calking. The greatest tensile stress on all sheets must come in the direc- 
 tion of the fibre ; therefore cylindrical shells must be made with the fibre of the sheets 
 running around them. Since the joints are the weakest and most troublesome parts of 
 a boiler, it is advantageous, with regard to economy and strength, to make the sheets 
 as large as possible. 
 
 9. Specifications for Boilers for U. S. Steam-sloop " Lackawanna," refe- 
 rence being had to the accompanying Drawings (Plates VI. and VII.) 
 There are to be three vertical tubular boilers, containing an aggregate grate-surface of 
 273 square feet and a total heating-surface of 8, 980 .square feet. The two main boilers 
 are to be made right and left ; the one for the starboard side of the ship is to be 25 feet 
 long and to have seven furnaces ; the opposite boiler is to be 21 feet 6 inches long and 
 to have six fxirnaces ; the other, or auxiliary, boiler is to have one furnace, and is to be 
 placed aft of the six-furnace boiler, and is to be 4 feet long. 
 
 Extreme height of boilers, exclusive of steam-drums, to be 9 feet 5 inches. 
 
 Extreme depth of boilers at furnaces, 10 feet 3 inches ; at top of shell, 11 feet 9 
 inches. 
 
 Each main boiler is to have a steam-drum extending 2 feet 9 inches above the top of 
 shell. Each furnace is to be made with an independent combustion-chamber, which 
 communicates through a tube-box containing 310 tubes with a front-connection, and 
 unites with the others in a common uptake which is to have a calorimeter equal to that 
 of all the tube-boxes. 
 
 Material. All the parts of the boilers, except the tubes, are to be made of the very 
 best American charcoal-iron. 
 
 Water-legs. The boilers are to be made with water-legs, which are to enclose water- 
 spaces 6 inclies wide, including the thickness of metal ; they are to be made of the best 
 flange iron & inch thick ; the bottom plate of each to be of one sheet, and to be made 
 to lap outside the vertical plates forming the leg. 
 
 Shell of Boilers. The bottom of shell at back part of boilers to be ^ inch thick 
 
SEC. 9. DESIGN, DRAWINGS, AND SPECIFICATIONS. 143 
 
 and to extend above the turn to the sides. All other parts of the shell to be ^ inch 
 thick. 
 
 Furnaces. Each furnace is to be 3 feet wide and 6 feet 6 inches long on the grates. 
 The crown of each furnace is to be f inch thick, and made of one sheet extending on 
 each side below the grate-bars. 
 
 Connections and Tube-boxes. The front-heads of furnaces and back-heads of 
 back-connections are to be made of the best flange-iron f inch thick ; sides and top 
 of back-connections f inch thick, and bottom -^ inch thick; the sides of tube- 
 boxes f inch thick ; front-connections and uptake to be f inch thick, to be of the best 
 flange-iron. 
 
 Tube-sheets to be f inch thick, of the best flange-iron. Water-spaces between the 
 tube-boxes, front and back connections, to be 6 inches wide, including the thickness of 
 metal. 
 
 The tubes are to be made of drawn brass, to be 2 inches external diameter and 33J 
 inches long, and to be of No. 13 wire-gauge thickness. 
 
 Bracing. The top and side of the shell of boiler above the tube-boxes to be stiffened 
 by T-iron 3" x 3J" X f *, placed every 12 inches, to which the braces are to be attached. 
 Each flange of the T-iron is to be riveted to the shell every four (4) inches, and the rivets 
 so placed as to alternate with each other. The braces are to be not over 12 inches apart, 
 and to be coupled to diagonals, which are to be attached to the T-iron by bolts and nuts 
 placed not over 12 inches between centres ; braces to be 1* diameter ; diagonals to 
 be double, of rectangular section 2* X f ", and secured by bolts 1J* diameter. All flat 
 spaces not stiffened by T-iron to be braced every 8 inches by socket-bolts of 1" diameter. 
 Crown of furnaces to be thoroughly braced. Care is to be taken that the strength of 
 bracing herein specified is to be carried out in all the diagonals, angle and T-irons, and 
 their rivets and attachments, and in the welding, so that no parts are left more weakly 
 braced than herein specified. 
 
 Riveting. All seams not in contact with the fire to be double-riveted, and the rivets 
 to be staggered. 
 
 Seams. All seams to be calked on both sides when practicable, and no acids are to 
 be used in forming them, nor any fitting pieces to be inserted. 
 
 Manholes. Each boiler is to be provided with two manholes, 12* X 15* diameter, 
 opening into the steam-room, one on each end of the main boilers, placed on the front 
 side, and one in the front side of the auxiliary boiler. 
 
 Manholes, 12' x 15" diameter, will also be placed in the spandrels of the furnaces. 
 Each manhole is to have around it on the inside a wrought-iron band, I inch thick and 
 
144 STEAM BOILERS. CHAP. VEL 
 
 4 inches wide, double-riveted to the shell ; inner row of rivets countersunk flush on 
 both sides. 
 
 Manhole-cotters. The manholes to be provided with cast-iron covers, the joints of 
 which are to be faced. The plates to be secured in place by wrought-iron cross-bars 
 and bolts. 
 
 Handholes. Each boiler is to have a handhole near the bottom of each leg at each 
 end. Openings to be elliptical, with diameters of 3" and 5", and to be fitted with cast- 
 iron plates, the joints of which are to be faced ; bars, bolts, and rivets to be of wrought- 
 iron. 
 
 The Furnace-doors and Grate-bars are to be furnished by the Government and to 
 be properly fitted in place by the contractors. 
 
 Uptake-doors to be of wrought-iron, double-shell, and fitted with the proper hinges 
 and latches, and to be filled with some approved non-conducting substance. 
 
 Ashpit-doors to be made of wrought-iron, \" thick, with a flange f" deep all around ; 
 to be well fitted for closing the ashpit, and arranged to hang by proper hooks to the 
 uptake-doors when not in use. 
 
 1O. Extract from Specifications for Engines of U. S. S. " Miantonomoh." 
 (Plate VIII.) 
 
 Boilers and Attachments. There are to be six return horizontal tubular boilers, 
 placed forward of the engines, three on each side of the vessel, with the fire-room run- 
 ning fore and aft between them. 
 
 There is to be one chimney in vertical line over the keel, connecting with an uptake 
 common to all the boilers. 
 
 The boilers are to be constructed of the best American charcoal flange-iron ; all 
 seams not in contact with the fire to be double-riveted. All the plates to be planed on 
 the edges, the seams to be butt-jointed and covered with butt-straps of the same thick- 
 ness as the plates with which they are in contact ; all to be calked perfectly tight. 
 
 Each boiler is to be 12 feet external diameter and 10 feet in length, to have three 
 cylindrical furnaces, 36 inches internal diameter, projecting 6 inches from front of boiler 
 and extending to the back-connections. 
 
 Grate-surface. Each furnace is to have a grate 36 inches wide and 84 inches long, 
 or 21 square feet, aggregating 378 square feet in the six boilers. 
 
 Grate-bars to be double, in two lengths, f inch apart, f inch thick at top, ^ inch at 
 bottom, 4 inches deep at centre and 2 inches at ends. 
 
 Tubes. Each boiler is to contain 197 drawn-brass tubes, 3 inches external diameter 
 and 7 feet 3 inches long, No. 12 wire-gauge thickness. 
 
SEC. 10. DESIGN, DRAWINGS, AND SPECIFICATIONS. 145 
 
 
 
 Shell to be formed of plates f inch thick ; joints to be butted and strapped, and to 
 be double-riveted each side of seams. 
 
 Heads. The front and back heads to be inch thick, each composed of three plates 
 making two horizontal joints, planed and butted ; the straps to be inch thick and 
 9 inches wide, double-riveted each side. 
 
 Bracing. The heads will be braced by rods, If inches in diameter, placed 12 inches 
 between centres ; the ends of rods to be made with jaws and coupled to stay -plates 
 at each end by a wrought -iron bolt If inches in diameter. Stay-plates to be of the best 
 iron, f inch thick, and secured to boiler by a flange 2f inches wide and an angle-iron 3 
 by 3 inches. The stay-plates to be made with lugs 6 inches in diameter, leaving open- 
 ings for the removal of the braces. 
 
 Furnaces. The furnaces are to be made of the very best iron, inch thick. Each 
 furnace is to be formed of three cylindrical sections, 36 inches internal diameter, butted 
 and strapped, flanged on the ends, and riveted to each other with a welt, f inch thick, 
 between them. The furnaces to be double-riveted at their junctions with the front- 
 heads. 
 
 The Back-connections are to be 27 inches deep and made as shown in the drawings ; 
 side and back plates to be of -inch iron. The side and back heads to be stayed by 
 socket-rivets 1 inch in diameter and spaced not over 7 inches from centre to centre. 
 
 The Tube-sheets are to be inch thick. The centre tube-sheets are to be accur- 
 ately drilled for 63 tubes ; the two outer tube-sheets to be accurately drilled for 67 
 tubes in each. The tubes are to be spaced horizontally and vertically 4 inches between 
 centres. 
 
 Manholes. There are to be manholes 9 inches by 13 inches in the front-head of 
 each boiler in the outer spandrels above the furnaces, and in the lower spandrels be- 
 tween the furnaces, and a manhole 11 by 15 inches in the space above the centre furnace. 
 Each manhole to have around it on the outside a wrought-iron band, 1 inch thick and 
 4 inches wide, double-riveted to shell ; inner row of rivets countersunk flush on both 
 sides. Manholes to be closed with cast-iron plates and secured with double wrought- 
 iron cross-bars and bolts. 
 
 The Front-connections and Uptakes are to be made with double shells of wrought- 
 iron, built on angle-iron frames ; the angle-iron to be 2J inches by If inches ; the inside 
 and outside shells to be made of iron weighing respectively 5 Ibs. and 3^ Ibs. per square 
 foot. The space between shells to be filled with some non-conducting material. The 
 uptake at the connection with the smoke-pipe to be 8 feet 3 inches in diameter. The 
 doors are to be made of wrought-iron, double shell, and fitted with the proper hinges 
 
146 STEAM BOILERS. CHAP.VH. 
 
 and catches ; outside shell i inch thick, flanged 1 inch deep ; inside shell J inch thick, 
 flanges 2J inches deep. 
 
 Furnace-fronts to be made with a cast-iron frame covered with wrought-iron plates 
 f inch thick, and having openings 20 by 13i inches for furnace-doors. The outside 
 plates to have twelve air-holes 1 inches in diameter, and the inside plates to be per- 
 forated with 250 holes ^ inch in diameter. 
 
 The Furnace-doors to be made with wrought-iron fronts i inch thick and flanged 1 
 inch deep ; each to be fitted with a perforated wrought-iron back and the necessary 
 hinges and latches of wrought-iron. 
 
 The Ashpit-doors to be made of wrought-iron \ inch thick, flanged 1 inch deep ; to 
 be well fitted to close the ashpits, and arranged to hang by proper hooks on uptake- 
 doors when not in use. 
 
 Saddles. Each boiler is to rest on two saddles made of wrought-iron and properly 
 secured to the ship. The boilers to be secured to the saddles by suitable wrought-iron 
 straps and bolts. The bolts for straps (which pass through the shell) to be turned and 
 snugly fitted into reamed holes. 
 
 Steam-drums. There are to be two steam-drums .on each side, placed in the span- 
 drels above the boilers, each to be 42 inches in diameter and 8 feet 6 inches long. 
 Shells to be inch thick, heads inch thick. Each head to be braced by ten gussets 
 equally divided on the shell ; gussets to be of f -inch iron, extending 30 inches on the 
 shell and 11 inches on the head, and securely riveted to shell and head by angle-iron 2J 
 by 2| inches. 
 
 Two Superheating Steam-pipes, 15 inches in diameter, made of the best boiler-plate 
 ^oV inch thick, are to be placed within each front-connection, and united with each other 
 at forward ends and to steam-drums placed in the spandrels above the boilers. All to 
 be of the best American charcoal-iron. 
 
 Safety-valves. Each boiler and superheating-pipe is to have a safety-valve 5 inches 
 in diameter, fitted with the proper weights and levers for a pressure of 80 Ibs. of steam. 
 The chests to be of cast-iron, the valves and seats of composition ; the valves to be con- 
 nected to copper pipes leading to the chimney. 
 
 Dry-pipes. Each boiler is to have a sheet-brass dry-pipe thoroughly tinned and of 
 an internal diameter of 7 inches ; the pipe to be placed as high as possible and extend 
 nearly the length of the boiler ; the top, for a distance of 3| feet on either side of the 
 centre of its length, to have holes f inch in diameter drilled equally distant ; the aggre- 
 gate area to be double the cross-section of the pipe. 
 
 Stop-valves. Each boiler is to have a composition stop-valve placed on the back- 
 
 
SEC. 10. DESIGN, DRAWINGS, AND SPECIFICATIONS. 147 
 
 head near the top, and united with the boiler and dry-pipe by flanges llf inches in 
 diameter and |f inch thick ; the chest to be not less than ^ inch thick. The valve is to 
 be 6 inches in diameter, and fitted with a screw-stem made to torn independently of the 
 valve, and work in a composition nut supported by wrought-iron studs on the covers ; 
 the valve to be operated by a composition hand-wheel 12 inches in diameter. 
 
 Pipes. The stop-valves are to be connected with the steam-drums by copper pipes 
 6 inches internal diameter and No. 16 Birmingham-gauge thickness ; the pipes to have 
 composition flanges, 11 inches in diameter and }$ inch thick, properly riveted and 
 brazed on. 
 
 Main Stop-valves. There are to be two main steam stop- valves placed between the 
 boilers and the engines ; they are to be connected with the superheating-pipes and each 
 other by pipes 12 inches in diameter, and arranged to close the steam from either the 
 port or starboard boilers while the others are in use. The chests are to be of cast-iron 
 | inch thick, flanges If inches thick. They are to be fitted with composition valves and 
 seats, the valves to be 12 inches in diameter, and are to have screw-stems made to turn 
 independently of the valves and work in composition nuts supported by wrought-iron 
 studs on the covers ; the valves to be operated by composition hand-wheels 20 inches in 
 diameter. 
 
 The Main Steam-pipes are to be of copper ; three sections to be 12 inches in diameter 
 and of No. 11 Birmingham-gauge thickness, two sections of 10 inches diameter and No. 
 14 Birmingham-gauge thickness ; the sections connecting with the throttle- valves to be 
 9 inches in diameter and No. 14 Birmingham-gauge thickness ; the several sections to 
 be united to each other and the valves by composition flanges inch thick, properly 
 riveted and brazed to the pipes. All the steam-pipes to be heavily tinned inside and 
 out. 
 
 Bleeding-valve and Pipe. There is to be a copper pipe, with stop-valve, of 4 inches 
 in diameter, leading from the steam-pipe to the top of the condenser for bleeder. 
 
 Check and Slow Valves. Each boiler is to have a check feed- valve, 2f inches in dia- 
 meter, enclosed in a chest having two stop- valves of the same diameter, that may be 
 closed from the boiler and feed-pipe ; the valve and chest to be of brass and made with 
 flanges 6J inches in diameter, inch thick, for connecting with feed-pipes and boilers. 
 Each boiler is to have a bottom blow- valve of brass 2 inches in diameter, also a sur- 
 face blow- valve 2 inches in diameter, all connected by suitable pipes to the sea-valves 
 on the ship. 
 
 The Main Feed and Slow Pipes to be made in sections not exceeding 12 feet in 
 length, and are to be of drawn-brass tubes, 3f inches inside diameter, of No. 8 Binning- 
 
148 STEAM BOILERS. CHAP. V1L 
 
 ham-gauge thickness ; the branches to be 3 inches inside diameter, of No. 9 Birmingham 
 gauge ; the pipes to be fitted with composition flanges and elbows for uniting the 
 sections, and are to be expanded in and sweated to them ; the interior of pipes to be 
 well tinned. 
 
 Gauge-cocks. Each boiler is to have a combination gauge, which shall include a 
 glass tube of 18 inches exposed length, and four cocks placed 6 inches apart, also drip- 
 pan and pipe ; the lowest cock to line with the bottom of glass tube, and placed to show 
 the level of the water at the highest heating-surface. 
 
 Salinometers. There are to be six of Fithian's salinometer-pots, fitted in such a 
 manner as to be easily accessible. 
 
 Steam-whistle. A large, finished steam- whistle of brass is to be conveniently placed 
 above deck, with copper pipe and cock connecting to boilers. 
 
 Test. Before being placed in the vessel all the boilers are to be subjected to a pres- 
 sure of 120 Ibs. to the square inch, which is to be obtained by filling the boilers quite 
 full of water and lighting a fire in the furnaces, producing the pressure by the expan- 
 sion of the water. The boilers, after completion, to be painted inside and out with two 
 coats of brown zinc-paint. 
 
 Covering for Boilers. After the boilers are in the vessel the entire shell and backs 
 of the same are to be covered with a casing of galvanized iron, enclosing an air-space of 
 H inches between the boilers and casing ; the sections of the casing to be substantially 
 connected with each other, and in such a manner that they may easily be removed and 
 replaced, and the joints made perfectly air-tight. The casing to be made with suitable 
 openings and covers over man and handholes and braces required to be removed from 
 the outside. The casing to be covered with two coats of brown zinc-paint. 
 
 Ventilators 18 inches in diameter are to fitted for the fire-room ; they will extend 
 below the deck to within 8 feet of the fire-room floor, and to be bell-mouthed at lower 
 end. The portion above deck will be secured to deck by screwed eye-bolts passing 
 through a flange and tapped into the rings around the holes through deck-plating ; they 
 are to have movable hoods, capable of being worked from fire-room, and will be made of 
 iron No. 11 wire-gauge thick. The forward ventilators are to be arranged for hoisting 
 ashes through them from deck by means of blocks and pulleys ; to be strengthened with 
 six strips of bar-iron placed vertically in the interior. These ventilators to have a side- 
 door for passing the ash-buckets through. 
 
 11. Specification of Boilers (Iron Shells) for Vessels of the English 
 Navy. Specification of certain particulars to be strictly observed in the construction 
 of set of marine boilers, with superheaters, of the collective indicated power 
 
SEC. 11. DESIGN, DRAWINGS, AND SPECIFICATIONS. 149 
 
 of horses, suitable for vessels of the class. They are to be delivered 
 
 complete by the , 18 .. 
 
 Boilers. The boilers to be tubular, capable of carrying steam of Ibs. to the 
 
 square inch, and to be proved by water-pressure to . to the square inch. 
 
 They are to be constructed in separate parts, in accordance with the form and 
 
 dimensions shown upon the accompanying tracing. 
 
 Furnaces. There are to be . furnaces of the dimensions given on the tracing. 
 
 To admit of bituminous coal as well as Welsh coal being burned effectively, perforated 
 bridges are to be fitted, with means of regulating the supply of air through them to any 
 degree of opening. The aggregate area of the perforations in the bridges to be not less 
 than 3 square inches per square foot of fire-grate for admission of air to the combustion- 
 chambers. The furnace-doors to be fitted with internal and external screen-plates, per- 
 forated with a few holes to keep the doors cool, and means to be provided for keeping 
 them open in a sea-way. 
 
 The furnace-bars to be of wrought-iron, 3 inches deep by 1 inches wide, and to be 
 made in lengths. The length of fire-grate to be feet inches. 
 
 Boiler-plates, etc. The tube-plates, the uptakes, the furnaces, and the combustion- 
 chambers, with the angle-iron and rivets in these parts, and all screwed stays, are to be 
 of Low Moor, Bowling, or Farnley iron, and all other parts of BB Staffordshire or other 
 iron of equal quality. The minimum thickness of the boiler-plates to be as follows : 
 
 tube-plates; inch ; uptakes and bottom of shells, inch ; bottoms of furnaces, 
 
 inch ; upper and lower parts of fronts, inch ; and all other parts, _ inch. 
 
 The bottom plates of the shells to be double-riveted throughout. All the plates of the 
 boilers to be lap- jointed, excepting the lower parts of the fronts, which are to be lap- 
 welded. 
 
 Tests of Plates. All plates (with the exception of Low Moor, Bowling, or Farnley 
 plates, which will not be tested) must be capable of standing the following tests : 
 
 Tensile strain per square inch. 
 
 Lengthways 21 tons. 
 
 Crossways 18 " 
 
 Forge-test (hot). 
 
 Plates to admit of being bent hot, without fracture, to the following angles : 
 
 Lengthways of the grain 125 degrees. 
 
 Across . . . 100 " 
 
150 
 
 STEAM BOILERS. 
 
 CHAP. VII. 
 
 Forge-test (cold). 
 Plates to admit of being bent cold, without fracture, to the following angles : 
 
 Thickness of plate. 
 
 With the grain. 
 
 Across the grain. 
 
 
 Through an angle of 
 
 Through an angle of 
 
 i inch. 
 
 15 degrees. 
 
 5 degrees. 
 
 H 
 
 
 15 
 
 
 5 
 
 
 I 
 
 
 20 
 
 
 7* 
 
 
 H 
 
 
 20 
 
 
 7i 
 
 
 f 
 
 
 22^ 
 
 
 10 
 
 
 H 
 
 
 2 5 
 
 
 10 
 
 
 1 
 
 
 27^ 
 
 
 12* 
 
 
 * 
 
 
 3 
 
 
 * 
 
 
 i 
 
 
 35 
 
 
 15 
 
 
 7 
 
 
 44 
 
 
 i7* 
 
 
 1 
 
 
 5 
 
 
 20 
 
 
 A 
 
 
 60 
 
 
 2 S 
 
 < ' 
 
 i ' 
 
 
 70 
 
 
 30 
 
 Tests of Tee-irons and Angle-irons. Tensile strain per square inch with the grain, 
 for every description, 21 tons. 
 
 The ductility and other qualities of the iron should be such as to admit of its being 
 bent hot and cold in the following manner, without fracture : 
 
 ANGLE-IKON. 
 
 Forge-test (hot). 
 Angle-iron should be tested hot by being bent thus : 
 
 Fig. lO.o 
 
 Fig. 11. a 
 
 and also by being flattened thus : 
 and the end bent over thus : 
 
 Fig. 12. a 
 
 V 
 
 Fig.13. 
 
 Fig. 14. 
 
SEC. 11. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 151 
 
 Forge-test (cold). 
 
 Angle-iron should also be notched and broken across cold to show . 
 the quality of the iron ; and one flange should be cut off and bent cold, 
 thus: Fig.ts.a 
 
 TEE-IKON. 
 
 Forge-test (hot). 
 
 Tee-iron should be tested hot by being bent thus 
 Forge-test (cold). 
 
 The cold test for tee-iron should be similar to that for 
 angle-iron. 
 
 Tests of Stays and Rivets. Samples of angle, tee, and 
 bar iron for testing are to be selected from quantities of two tons or portion of two 
 tons weight. 
 
 Bar-stays, Yorkshire iron screwed stays, and rivets are to be capable of standing a 
 tensile strain per square inch of 21 tons. 
 
 Boiler --tubes. The tubes are to be of brass, solid-drawn (with the exception of the 
 stay tubes, which are to be of Low Moor, Bowling, or Farnley iron), and are to con- 
 tain not less than 68 per cent, of best selected copper. 
 
 The total number of tubes (including stay -tubes) is to be , their mean thickness 
 
 to be not less than No. wire gauge, their external diameter to be not less than 
 
 inches, and their length, outside the tube-plates, to be feet inches. 
 
 The stay-tubes are to be not less than _ inch in thickness. 
 
 Test of Tubes. Samples of tubes, weighing at least 10 Ibs., selected by the boiler 
 overseer, are to be forwarded by the contractors to Portsmouth Dockyard, there to be 
 subjected to such tests as their Lordships may direct. Each of the tubes is to be tested 
 by water-pressure separately to 300 Ibs. per square inch in the presence of the boiler 
 overseer. 
 
 Manholes and Mudholes of Boilers. In the manufacture of the boilers care is to 
 be taken to have sufficient room for manholes for the purpose of cleaning and repairing 
 the furnaces. All manholes and mudholes of the boilers to have stiffening-rings. The 
 doors to be of wrought-iron, and to be placed on the inside of the boilers. The manhole- 
 frames on the tops of the boilers to be raised sufficiently to clear the lagging of the 
 boilers. 
 
 Stays of Boilers. The stays are to be arranged on an approved plan so as to admit 
 
152 STEAM BOILERS. CHAP. VII. 
 
 of easy access to the internal parts of the boilers. Tee-irons to be attached to the shells 
 of the boilers for the purpose of securing the long stays ; the rivets or bolts for securing 
 the stays to be at least 25 per cent, stronger than the stays, and all holes for such rivets 
 or bolts to be drilled. Palm-stays are to be forged from the solid where practicable. 
 The screwed stays to be nutted on all flat surfaces. The maximum strain on the stays 
 at the working pressure must not exceed 5,000 Ibs. per square inch of section at the 
 bottom of the thread. 
 
 Circulating -plates. Circulating-plates are to be placed in proper positions to aid in 
 the circulation of the water, for which detail drawings will be supplied to the contractor. 
 
 Zinc Blocks. Zinc blocks as required are to be supplied and suspended in each 
 boiler for the purpose of preventing corrosion. A tracing showing the proposed ar- 
 rangement of the blocks to be submitted for approval. 
 
 Superheaters. The superheaters to be of the form and dimensions shown upon the 
 
 tracing, and to be proved to Ibs. to the square inch by water-pressure. The tubes 
 
 are to be of wrought-iron, Low Moor, Bowling, or Farnley, and not less than _ inch 
 
 thick. Their length to be feet inches, their diameter to be inches, and 
 
 their total number 
 
 Drawings. Before the work is put in hand detail drawings with figured dimensions 
 of the boilers and superheaters are to be submitted for approval. The drawings are to 
 be made on tracing-cloth, to a scale of not less than 1 inch to the foot, and are to give 
 full particulars of the mode of staying the boilers and superheaters, and also full details 
 of the furnace frames and doors, perforated bridges, etc. A duplicate of the approved 
 drawings to be furnished for the guidance of the boiler overseer. 
 
 In General. No holes, with the exception of the manholes and mudholes, are to 
 be cut, but the positions of the stop and safety valves, as indicated on the tracing, are 
 to be kept clear of stays and seams. 
 
 Dampers are to be fitted to the mouth of every ashpit, and external plates to be 
 fixed to the fronts of the smoke-box doors. The threads of all nuts, screws, studs, etc., 
 used in the construction of the boilers and superheaters are to agree with the threads 
 used in Her Majesty's service. Particulars of these, together with any additional par- 
 ticulars relating to the perforated bridges, etc., will, on application, be furnished to the 
 contractors. 
 
 Spare Gear. The following articles of spare gear to be supplied : 
 
 Boiler-tubes one-tenth of the whole number. 
 
 Ferrules for boiler- tubes one set for the fire-box end of each boiler. 
 
 Stay-tubes with nuts complete one set for each boiler. 
 
SBC. 11. DESIGN, DRAWINGS, AND SPECIFICATIONS. 153 
 
 Tap and die with suitable wrenches for stay-tubes one each. 
 
 Superheater-tubes (including stay-tubes). 
 
 Furnace-bars one-half set for each furnace. 
 
 Bearing-bars sets for furnaces. 
 
 Screwed stays for boilers (one size larger than required when new) . . .the whole number. 
 Tap and die for the above, with suitable wrenches one each. 
 
 Supervision. The boilers and superheaters will be subject to the supervision of an 
 overseer, who will be directed to attend on the premises of the contractors during the 
 progress of the work, to examine the materials and workmanship used in their construc- 
 tion and to witness the prescribed tests. The extent of supervision is described on the 
 attached paper extracted from Admiralty instructions to overseers. 
 
 Inspection. The contract is to be executed in every respect to the satisfaction of 
 the Controller of the Navy, who will, as he may see fit, appoint officers to inspect the 
 work while in progress. The boilers and superheaters are to be proved by water-pres- 
 sure in the presence of the inspecting officer, and means are to be provided for ascer- 
 taining their correct weight with fittings complete as supplied by the contractors. The 
 weight of the water in the boilers, at a height of nine inches above the top of the tubes, 
 also to be correctly ascertained. The boilers are not to be painted until they have been 
 proved to the satisfaction of the inspecting officer. 
 
 Delivery. After the boilers and superheaters have been proved they are to be well 
 painted with red lead, and are then to be delivered complete, with fittings and spare 
 gear, at . 
 
154 
 
 STEAM BOILERS. 
 
 CHAP. VII. 
 
 13. Material for six Boilers of U. S. S. " Nipsic." 
 
 (Plate X.) 
 
 November, 1877. 
 
 Letter of 
 reference. 
 
 Number of 
 sheets. 
 
 Dimensions in inches. 
 
 Weight in Ibs. 
 
 Parts of boilers. 
 
 
 
 Boiler-iron. 
 
 
 
 
 24 
 
 170^ X 52 X H 
 
 41,568 
 
 Shell of boiler. 
 
 A 
 
 6 
 
 119 X 87 X i 
 
 7,040 
 
 Back-head. 
 
 B 
 
 6 
 
 1 08 X 73 X A 
 
 6,500 
 
 Back tube sheet. 
 
 C 
 
 6 
 
 119 X 86 X A 
 
 7,624 
 
 Front tube-sheet. 
 
 D 
 
 6 
 
 1 10 X 37i X A 
 
 2,813 
 
 Lower part of front head. 
 
 
 12 
 
 no X 33! X A 
 
 5,417 
 
 Furnaces. 
 
 
 12 
 
 no X 31 X A 
 
 5,012 
 
 * 
 
 
 12 
 
 1 10 x 28 xA 
 
 4,527 
 
 tt 
 
 
 12 
 
 120 X 9i X H 
 
 2,564 
 
 Transverse shell butt-straps. 
 
 E 
 
 6 
 
 !9i ) 
 and V X H 
 
 1,548 
 
 11 
 
 
 24 
 
 57 X 9f X i) 
 
 
 
 
 12 
 
 54 X 9f X iV 
 
 3,220 
 
 Longitudinal shell butt-straps. 
 
 
 12 
 
 30 X 91- X i) 
 
 
 
 F 
 
 6 
 
 no X 55 X i 
 
 3,99* 
 
 Bottom of back-head. 
 
 
 6 
 
 100 X 40 X ^ 
 
 
 Back-connection. 
 
 
 6 
 
 106 X 58 X i 
 
 5, l6 4 
 
 (t 
 
 
 12 
 
 75 X 27^ X i 
 
 3,465 
 
 U It 
 
 
 6 
 
 12 
 
 64 X 42 X 1 
 44 X 36 X | f 
 
 3,600 
 
 Butt-plates for front-head. 
 
 
 24 
 
 128 X 34} X i 
 
 14,847 
 
 Gussets. 
 
 
 6 
 
 106 X 16 X i 
 
 1,425 
 
 " 
 
 
 6 
 
 94 X 13^ X } 
 
 i, 066 
 
 " 
 
 
 12 
 
 96 X 22 X \ 
 
 3,548 
 
 <( 
 
 G 
 
 12 
 
 3 6J X 2o X 1 
 
 758 
 
 Furnace-front. 
 
 H 
 
 12 
 
 34 X 19} X 4 
 
 667 
 
 " " 
 
 I 
 
 12 
 
 36 X 16 X A 
 
 274 
 
 Ashpit-doors. 
 
 J 
 
 2 4 
 
 23} X 16} X } 
 
 57o 
 
 Furnace-doors. 
 
 
 IO 
 
 86 X 60 X A 
 
 4,5 
 
 Shell of steam- drums, gussets, etc. 
 
 a 
 
 8 
 
 37 (diam.) X f 
 
 920 
 
 Heads of steam-drums. 
 
 
 
 Total 
 
 T 7 C 088 
 
 
 
 
 
 tjj ty 
 
 
SBC. 12. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 155 
 
 MATERIAL FOR six BOILERS OF U. S. S. " NIPSIC." (Continued.) 
 
 Number of 
 pieces. 
 
 Description. 
 
 Dimensions in inches. 
 
 Weight in 
 
 E 
 
 Parts of boilers. 
 
 45 
 
 Round bar-iron 
 
 i\" diam. X 80* long 
 
 1,000 
 
 Stays. 
 
 5 
 
 a 
 
 if" diam. X 6o f long 
 
 M53 
 
 M 
 
 38 
 
 u '< 
 
 if" diam. x 80' long 
 
 i>774 
 
 
 
 
 Flat bar-iron . . . 
 
 1' X 2f 
 
 1,000 
 
 Braces. 
 
 
 
 
 i"X3' 
 
 1,000 
 
 Furnaces. 
 
 
 
 2i' X 2*' 
 
 825 
 
 Gussets. 
 
 
 
 
 Total 
 
 7.352 
 
 
 
 1,440 
 
 Rivets. 
 
 1" diam. X 3%' ! n g 
 
 i ? 340 
 
 Shell double straps. 
 
 2.535 
 
 N 
 
 |* diam. X 3!" ! n g 
 
 2,200 
 
 <* 
 
 ' J tJJ 
 
 9,112 
 600 
 
 H 
 H 
 
 I' diam. X 2J* long 
 |' diam. X a long 
 
 7,100 
 45 
 
 Single straps and heads. 
 Gussets and back-connections. 
 
 262 
 
 
 
 |" diam. X 2\" long 
 
 1 80 
 
 Manhole-frames. 
 
 4,125 
 
 
 
 1* diam. X 2f " long 
 
 2,100 
 
 Front-head strap, gussets, and fur- 
 
 
 
 
 
 nace-rings. 
 
 2,250 
 
 10,000 
 
 
 11 
 
 J* diam. X 2^' long 
 |" diam. X 2\" long 
 
 i, 080 
 4,650 
 
 Gussets and furnace-front. 
 Furnaces, back-connections, back- 
 
 
 
 
 
 heads, manholes, and gussets. 
 
 3,5 2 5 
 
 M 
 
 J" diam. X 2" long 
 
 1,480 
 
 Furnace-front and angle-iron on 
 
 
 
 
 
 gussets. 
 
 
 
 Total 
 
 20,580 
 
 
 
 NOTB. Twenty-five per cent, is added to the amount actually required of each description of rivets. 
 
156 
 
 STEAM BOILERS. 
 
 CHAP. VII. 
 
 13. List of Steel Plates for Boiler of Steamer " Lookout." (Plate X.) 
 
 Letter 
 of reference. 
 
 Number of 
 sheets. 
 
 Dimensions in inches. 
 
 Tensile strength 
 in Ibs. per sq. in. 
 
 Parts of boiler. 
 
 A 
 
 
 95 X 67 X i 
 
 6o,OOO 
 
 Front-head, upper section. 
 
 a 
 
 
 95 X-6 7 X A 
 
 7O,OOO 
 
 Back-head, " 
 
 B 
 
 
 89 X 31 X i 
 
 6o,OOO 
 
 Front-head, lower section. 
 
 b 
 
 
 89 X 31 X T \ 
 
 7O,OOO 
 
 Back head, " 
 
 c 
 
 
 85 X 52 X i 
 
 6o,OOO 
 
 Back tube-sheet. 
 
 D 
 
 I 
 
 85 X 71 X A 
 
 6o,OOO 
 
 Back- head, back connection. 
 
 
 I 
 
 192 X 20 X iV 
 
 6o,OOO 
 
 Bottom and sides, back-connect'n. 
 
 
 2 
 
 265 X 47i X A 
 
 70,OOO 
 
 Shell of boiler. 
 
 E 
 
 I 
 
 60 X 8 X 42 X i 
 
 7O,OOO 
 
 Longit'al outside strap, manhole. 
 
 F 
 
 I 
 
 60 X 8 X 25 X i 
 
 7O,OOQ 
 
 inside " 
 
 
 I 
 
 244 X 8 X T \ 
 
 7O,OOO 
 
 Circular strap. 
 
 
 I 
 
 92 X 8 X T V 
 
 7O,OOO 
 
 Butt-strap, back-head. 
 
 
 2 
 
 92 X 45 X T V 
 
 6o,OOO 
 
 Furnaces, front section. 
 
 
 2 
 
 92 X 41 X T V 
 
 6o,OOO 
 
 back section. 
 
 
 4 
 
 45 X 5 X T V 
 
 6o,OOO 
 
 straps. 
 
 
 I 
 
 55 X 25 x T V 
 
 70,OOO 
 
 Plate around manhole, front-head. 
 
 
 2 
 
 26 x 24 x T \ 
 
 70,000 
 
 a tt 
 
 
 4 
 
 24 x 30 X T \ 
 
 7O,OOO 
 
 Gussets and stay-plates. 
 
 
 2 
 
 18 x 26 X T V 
 
 7O,000 
 
 H 11 11 
 
 
 6 
 
 2? X 27 X A 
 
 70,000 
 
 it it 
 
 G 
 
 
 65 (diam.) X T \ 
 
 7O,000 
 
 Head of steam-drum. 
 
 g 
 
 
 65 X 5i X i 
 
 6o,000 
 
 Bottom 
 
 
 
 75 X 73 X A 
 
 6o,OOO 
 
 Uptake- pipe. 
 
 
 
 73 X 5 x A 
 
 6o,OOO 
 
 strap. 
 
 
 
 171 X 80 x 1 
 
 7O,OOO 
 
 Shell of drum. 
 
 
 2 
 
 80 X 6 X A 
 
 7O,OOO 
 
 Straps for drum. 
 
SEC. 13. 
 
 DESIGN, DRAWINGS, AND SPECIFICATIONS. 
 
 157 
 
 LIST OF IRON PLATES, RIVETS, TUBES, ETC., FOR BOILER OF STEAMER " LOOKOUT." 
 
 Letter 
 of reference. 
 
 Number. 
 
 Dimensions in inches. 
 
 
 
 i sheet. 
 
 9 X 8 X | 
 
 
 H 
 
 2 sheets. 
 
 31 x i?i x 4 
 
 
 I 
 
 2 " 
 
 28 x 16 X 1 
 
 
 K 
 
 2 " 
 
 2I i X isJ X i 
 
 
 L 
 
 2 " 
 
 32 X 14 X A 
 
 
 
 2 " 
 
 82 X 27 X A 
 
 
 
 2 " 
 
 72 X 24 X i 
 
 
 
 2 " 
 
 54 X 24 X i 
 
 
 
 2 " 
 
 106 X 3 X i 
 
 
 
 2 " 
 
 124 X Si X i 
 
 
 
 i sheet. 
 
 72 X si X i 
 
 
 
 2 angle-irons 
 
 96 X ii X ii 
 
 
 
 140 stay-bolts. 
 
 i' diam. x 8i" long. 
 
 
 
 40 " 
 
 i' " X 7i' " 
 
 
 
 30 rivets. 
 
 W " X 3i' " 
 
 
 
 170 " 
 
 U' " X 3' " 
 
 
 
 280 " 
 
 1 3 ff *< \/ ,,3* ** 
 15" ^ *T 
 
 
 
 120 " 
 
 if " X 2f ' " 
 
 
 
 5 ;; 
 
 1 ! X 2i" 
 
 
 
 53 
 
 X 2^' 
 
 
 
 1 1 60 " 
 
 ir " x 2 r 
 
 
 
 500 " 
 
 ii; ;; x *\\ ;; 
 
 
 
 600 " 
 
 4 -^ *"T 
 
 
 
 400 " 
 
 r " . x ? r " 
 
 
 
 1 20 brass tubes. 
 
 2i* outside diameter by 
 
 
 
 
 6 feet 2$ inches long. 
 
 
 
 
 No. 12 Birmingham 
 
 
 
 gauge. 
 
 
CHAPTER VIII. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 1. Laying-off. The boiler-plates having been tested and examined as described 
 in chapter v., they are made to pass, first with one side then with the other side up- 
 wards, between two cylindrical rollers, in order to level and flatten them and to smooth 
 their surfaces. 
 
 The exact lines to which the plates are to be trimmed and cut and the flanges are 
 to be turned, and the centres of all holes that are to be punched or drilled before the 
 flanges are turned, are now laid-off on the plates and plainly marked by means of a 
 centre-punch. The lines of curved surfaces are developed and laid down in full size on 
 paper or on a board ; in case several boilers are to be constructed from the same draw- 
 ing it is well to make wooden or sheet-iron templates of irregularly-shaped plates, and 
 
 Fig. 15. 
 
 I o o 
 
 JX" 
 
 3 > ^ ( 
 
 / \ 
 
 / \ .x-^ "-^ \ / " / ' 
 
 ^.( E/C >^-r'^- > /\i ;y 
 
 VSn^/ /V ?^ \ / /V /\ Vco^/ 
 
 v 
 
 V 
 
 
 to mark the positions of holes and points which lie in the same straight lines on sepa- 
 rate sticks, and to transfer them thence to the plates ; the rivet-holes of seams are laid- 
 
 158 
 
SEC. 1. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 159 
 
 off by means of thin board templates having holes of the proper diameter, spaced as re- 
 quired. An intelligent and careful boiler-maker will lay-off in this manner the position 
 of all lines and holes, except on flanges that are to be turned ; frequently, however, the 
 holes for stay-bolts are first laid-off and drilled only on one plate, and the correspond- 
 ing holes in the opposite plate are marked off from these holes when the plates are 
 fitted in position. Instead of using the centre-punch, the position of holes that are to 
 be punched is often marked on the plate by means of a small round stick, called the 
 marker, which, being dipped into some liquid whiting and passed through the holes of 
 the template, leaves white circles on the plate. 
 
 Figure 15 represents the front-head and tube-sheet of the boilers of U. S. S. Nipsic 
 (see Plate XII.), with the rivet-holes and centres of tube-holes punched, manholes and 
 furnace-openings marked for cutting out, and the flanges for securing the shell and the 
 furnace-tubes marked for turning. 
 
 In bending plates to the circular form the inner side is slightly compressed and the 
 outer side is elongated, the neutral axis passing through the centre of the plate. There- 
 fore the length of the plates for the shell of a cylindrical boiler with butt-joints must 
 be equal to the outside diameter of the shell minus the thickness of the plate, multi- 
 plied by 3.1416, or (d t) n. 
 
 When a cylindrical shell or flue is formed of alternate inner and outer rings with 
 overlapping joints (see figure 16), the circumferential length of the plates forming the 
 
 inner rings is found by multiplying double the 
 thickness of the plates by 3.1416 and subtract- 
 ing the product from the length of the outer 
 rings. In laying-off the rivet-holes in the cir- 
 cumferential seams of these plates space the 
 holes on the outer plates as required ; the dis- 
 Flg ' 16> tance between the end holes of each seam of 
 
 the inner plates must be equal to the distance between the corresponding holes of the 
 outer plates, less the product of double the thickness of the plates multiplied by 3.1416 ; 
 mark the end holes on the inner plates accordingly, and divide the distance between 
 them equally, according to the number of rivets required for the seam. 
 
 When the longitudinal seams of such cylinders are made with lap-joints the forego- 
 ing rales do not give the whole length of the plates, but the distance between the centre 
 lines of the rivet-holes of the longitudinal seam, and a proper amount has to be added 
 to the length of the plate as an allowance for lap. 
 
 Figure 17 represents the method of finding the form of plates for conical tubes. 
 
160 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 Draw an elevation of the cone, abed; continue the side lines till they meet the cen- 
 tre line in o ; from o as a centre draw 
 with radii o c and o a two circles, 
 and make the arcs e Ji and f g re- 
 spectively equal to the required cir- 
 cumferential lengths of the two ends 
 of the tube ; draw the radial lines 
 ef and g Ji. The figure efgh rep- 
 resents the form of the plate re- 
 quired when the tube is to be butt- 
 jointed. In case the tube is to be 
 lap-jointed add the amount neces- 
 
 Fig. 17. 
 
 sary for lap at each end of the plate, as indicated by the dotted lines, i Jc and I m, 
 drawn parallel to ef and g h respectively. 
 
 When a cylindrical shell or flue is formed of rings lapping telescopically, as in 
 figure 18, each ring is slightly conical ; the taper is, however, so small that the method 
 
 Fig. 18. 
 
 Fig! 19. 
 
 of finding the shape of the plate which was illustrated in figure 17 cannot be used prac- 
 tically, because the radius o c would be too long. The following convenient and suffi- 
 ciently accurate method of finding the shape of such plates is given by Sexton (see 
 figure 19) : 
 
 Draw a centre line and mark on it the width of the plate, g h ; draw perpendicular 
 lines through the points g and h ; find the circumferential lengths of the two sides of 
 the plate by the rule given above ; and from 7i and g respectively lay off one-half of 
 these lengths on the perpendiculars on each side of the centre line, marking the points 
 thus found a, b, c, d, and draw lines a c and b d ; erect a perpendicular, a e, to line a c 
 
 in a, cutting the centre line in// hi = * is very nearly the versed sine of the arc 
 
 & 
 
 forming the upper edge of the plate ; draw a curve by means of a flexible batten 
 
SEC. 1. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 161 
 
 through the points a, i, b, and a similar one through c, d; then aibdk c will be the 
 form of the plate. The number of plates in the circumference of the tube does not 
 affect this rule ; the camber is constructed the same way, whether for one or for several 
 plates. 
 
 When a wedge-shaped portion, or ungula, is cut off from a cylindrical shell, as in the 
 boilers of U. S. S. Nipsic (see Plate XII.), the plates forming the cylindrical shell have 
 
 Fig. 21. 
 
 Fig. 22. 
 
 U 
 
 1 _a 6 c d 
 
 _6a 1 
 
 to be cut to a line 1, 9, shown in figure 21. The method of laying-off this line is illus- 
 trated in figures 20 and 21. Mark any convenient number of divisions on the arc repre- 
 senting the portion of the shell cut away ; project the points of these divisions on the 
 side elevation of the shell, and draw the parallel horizontal 
 lines 1 ! 2 2,, 3 3,, etc., through these divisions. Lay 
 off the position of these divisions on the developed plate 
 by measuring their distance on the arc from 9 and laying 
 them off from the corresponding point 9 on the plate 
 (figure 21), and erect perpendiculars at these points of divi- 
 sion ; measure the length of the lines 1 1,, 2 2,, etc., in 
 the side elevation (figure 20), and lay them off on the perpendiculars erected in the cor- 
 responding points of figure 21. With a flexible batten draw a fair curve through the 
 points thus found, which will be the required line. 
 
 To find the shape of the plate forming the slanting part of the back-head project the 
 divisions 2, 3, 4, etc., of the arc in figure 20 on the horizontal line 1h, and mark the 
 points thus found a, b, c, etc. Lay off the distances hi, ha, Ji 5, etc., both ways 
 from a centre line on a horizontal in figure 22, and mark the points thus found 1, a, b, c, 
 etc. ; draw perpendiculars through these points, making Ji 9, g 8, f 7, e 6, d 5, 
 c 4, 63, a 2, in figure 22, respectively equal. in length to the slant lines 1, 9,, 1, 8,, 
 
162 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 l i _7 ]) i j _ 6,, 1, 5,, 1, 4,, 1, 3,, 1, 2, in the side elevation of figure 20. With a 
 flexible batten draw a fair line through the points 1, 2, 3, 4, 5, 6, 7, 8, 9, 8, 7, 6, 5, 4, 3, 
 2, 1, in figure 22 ; then draw, at a distance depending on the width of the lap, the line 
 x y parallel to 1 1, and the curve x z y equidistant from the curve 1 9 1. 
 
 Figures 23 and 24 represent the method of laying-off a plate for the cylindrical shell 
 of a steam-drum which is to be placed on the top of a cylindrical boiler. Figure 23 
 shows an elevation of the steam-drum and a portion of the boiler. Divide one-quarter 
 of the circumference of the drum into a convenient number of equal parts, and from 
 these divisions draw lines parallel with the sides of the steam-drum and touching the 
 top of the boiler, as in figure 23. Find the length of the plate required by the rule 
 given above for the development of cylindrical shells, and divide this length, exclusive 
 of lap, into four equal parts (see figure 24), and each of these parts into the same num- 
 ber of parts as the quadrant in figure 23, numbering the corresponding points alike to 
 avoid confusion. Along the plate (figure 24) draw a line, C C, representing the distance 
 from the top of the boiler in the centre to the top of the steam-drum, and correspond- 
 ing with the line C D in figure 23. Now mark on each subdivision the distance, corre- 
 sponding to its number, from the line C D to the top of the boiler, as shown in figure 
 24, and through these points, by hand or by means of a flexible batten, draw a fair 
 curve ; parallel to this curve draw another at a distance corresponding to the width 
 required for the flange. 
 
 When the diameter of the steam-drum does not exceed one-half of the diameter of 
 the boiler the following shorter method is sufficiently accurate : Divide the plate into 
 four equal parts and draw a line corresponding to the top of the boiler, as in the pre- 
 vious case ; from this line draw one short line in the centre of each of the four divisions, 
 corresponding to the points 4, 4, 4, 4 in figure 24. Call the centre of the plate and the 
 two ends 0, and the remaining two lines.8 ; mark on the lines 8 the distance fro,m the 
 
Ssc. 2. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 163 
 
 level of the top of the boiler to the bottom of the side of the steam-drum, and on the 
 lines 4 half that distance. Extend the trammel to such a radius that, by taking a con- 
 tinuation of the lines 8 as a centre, it will touch the marks on lines 8 and 4 ; and with 
 the same radius describe arcs passing through the marks on lines 4 and the points on 
 the line C C. 
 
 Similar methods are employed in developing the lines of intersection of other curved 
 surfaces. 
 
 2. Shearing and Planing. After the holes and lines are laid-off and marked on 
 the plate, the next operations are to punch or drill the holes and to cut the plates to 
 the exact shape and size required. Manholes and similar openings and curved outlines 
 are generally formed by punching a series of holes, running into each other, close to the 
 line, the ragged edges being trimmed afterwards with a chisel. When the outline of 
 the plate is straight it is either sheared or planed. 
 
 The shearing-machine commonly used for this purpose has a stationary and a 
 movable steel cutter, the edges of which form an acute angle with each other, so that 
 during the process of shearing the action is rendered gradual. The motion of the cutter 
 is produced by means of an eccentric. 
 
 The process of shearing recommends itself through its simplicity, but it distresses 
 the metal greatly, especially in the case of steel plates, and such edges as have to be 
 calked have to be trimmed afterwards by hand to the proper bevel. On this account 
 the edges of plates should be planed where careful work is required. Planing does not 
 distress the metal ; it produces a smooth edge, which can be cut at once to any bevel 
 required for calking. For butt-joints the edges must be cut square and should always 
 be planed. 
 
 3. Bending. Sheets are bent to cylindrical shapes by passing them through the 
 bending-rollers. The primitive bending-machine consists of two cast-iron rolls laid side 
 by side, and a third roll, which is adjustable vertically, placed immediately over the 
 hollow between the two lower rolls (see figure 25). In hand-power bending-machines 
 set up on this plan the levers to turn the rolls are usually attached to one end of one 
 of the bottom rolls and to the opposite end of the top roll. 
 
 In modern bending-machines the rolls are arranged as shown in figure 26. Two 
 pinching-rolls are placed one directly over the other and geared together. The 
 upper roll is adjusted to the thickness of the plate to be bent by two strong set-screws, 
 and may be lifted out of the frame for the purpose of removing the bent plate. The 
 third or bending roll is placed to one side of the lower roll, and may be moved, by 
 means of a double hand-crank, bevel- wheels, and set-screws, past the lower roll toward 
 
164 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 the upper one ; this roll revolves by the friction of the plate against it. When this roll 
 is down far enough to have its top level with the top of the lower pinching-roll a 
 
 Fig. 26. 
 
 \ 
 
 \ 
 
 \ 
 
 Fig. 26. 
 
 plate passing between the pinching-rolls will be flattened and levelled ; but when the 
 bending-roll is raised towards the upper roll the plate is bent to a circle, since each 
 portion receives an equal curvature. By raising one end of the bending-roll higher 
 than the other different degrees of curvature may be given to the two ends of the plate. 
 Sometimes two bending-rolls are used, one on each side of the pinching-rolls ; by this 
 arrangement the plates can be bent nearer to the edge of the sheet than can be done 
 with three rolls. A template is applied to the sheet from time to time during the pro- 
 cess of bending, which is a gradual one, to see whether the proper curvature has been 
 produced. The diameter of the rolls varies from eight to twelve inches. 
 
 4. Flanging. When a plate is to be bent in a straight line to an obtuse angle it 
 may be done cold by means of a set and a sledge-hammer. This process, however, as a 
 rule, is objectionable, and should be avoided except in extreme cases. To produce 
 other shapes cast-iron moulds of the required form are used. The outline of the bent 
 portion is marked on the plate with centre-punch marks ; the iron is then heated to 
 redness, laid on the mould, and bent to shape by hammering. This process of bending 
 the edge of a plate at an angle with the plate is called flanging. In flanging steel plates 
 it is recommended to bend them as nearly as possible evenly along the whole length, as 
 much as can be done in each heat. When a short circular bend is wanted in the middle 
 of a long piece it is conveniently and accurately done by heating the piece in the middle 
 and bending it over a ridge, the ends serving as levers. 
 
 Sexton gives the following practical instructions regarding flanging: "A plate 
 
SEC. 5. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 165 
 
 during the process of flanging will gain twice its thickness in length in each flange. 
 Thus, suppose you want to flange a circular plate to have a 3-inch flange all round, 
 and to be 3 feet diameter after being flanged and f inch thick, yon 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| 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. The proper radius of the bend or root 
 of a flange is twice and one-eighth the thickness of the plate for the outside curve. In 
 heating the plate to be flanged, especially if it is a difficult one, confine your heat to as 
 little as possible over the width of the flange, but take as long a heat as you can. To 
 do this as it should be done make your fire the shape of the plate to be flanged. This 
 is easily accomplished by bending a piece of thin iron and packing the coal (well wetted 
 and mixed like mortar) up to it on the forge, or by adjusting a few fire-bricks to the 
 required shape. Keep the centre of your fire always clean; do not allow dust or 
 clinker to accumulate there, but let it consist of clean coke, broken small, and the 
 harder the better. If possible, have a block or an anvil to fit the flange, and do not, on 
 any account, flange on a block with sharp edges." 
 
 A flange turned at the circumference of a circular plate will be slightly thicker than 
 the plate ; but in forming flanges in the middle of a plate, like those for attaching the 
 furnace-tubes of cylindrical boilers to the front-head and to the back tube-sheet, the 
 metal has to be spread, and at the edges the flanges are much thinner than the plate. 
 Such flanging tests the ductility of the iron severely. When the front-heads of cylin- 
 drical boilers are too large to be made of one plate it is well to let the seam run through 
 the furnaces, as it is easier to form these partial circular flanges in the two plates than 
 to turn a complete circular flange in one plate. In such a case an extra allowance of 
 metal must be left at the corners of the flanges, or they must be spread out, as shown in 
 figure 15, before the flanges are turned ; because the metal will be drawn away from the 
 edges in the process of flanging, and the flanges of the two plates would not meet in 
 their whole width otherwise. 
 
 5. Punching. The rivet-holes are either punched or drilled. During the former 
 operation the plate rests on the table of the punching-machine, or it is slung in a chain 
 and suspended from a crane, being held in position by several men, who shift it. after 
 each hole is formed, so as to bring the stations of the successive holes under the punch. 
 Slight deviations from the correct positions of the holes are almost unavoidable with 
 this process, and. in order to avoid this source of error and secure greater rapidity in the 
 
166 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 execution of the work, some machines are arranged to punch several holes at the same 
 time, and are provided with a travelling-table, which moves after each stroke of the 
 punches automatically through the proper distance. Devices for spacing the holes me- 
 chanically are of special value for cylindrical boiler- work, since the holes are punched 
 before the sheets are bent by the rollers, and it is necessary to make an accurate allow- 
 ance for the difference of circumference of the inner and outer sheets. 
 
 Punches are made of steel, and are generally cylindrical with a flat end (see figure 
 
 Fig. 28. 
 
 Fig. 27. 
 
 Fig. 29. 
 
 Fig. 30. 
 
 27). When a centre-punch is used to mark the stations of the holes a point is formed 
 at the centre of the end of the punch, as shown in figure 28, in order to feel for the 
 puncture. Reed says that punches distress the iron less when the ends are formed as 
 shown in figure 29, instead of being flat. Others claim an equal advantage for punches 
 with a slightly concave face, especially for punching large holes. 
 
 Figure 30 represents Kennedy's helical punch. "Its form may be explained by 
 imagining the upper cutter of a shearing-machine being rolled upon itself so as to form 
 a cylinder of which its long edge is the axis. The die being quite flat, it follows that 
 the shearing action proceeds from the centre to the circumference, just as in a shearing- 
 machine it travels from the deeper to the shallower end of the upper cutter." Re- 
 sults of experiments made at Crewe, England, on the tensile strength of samples of the 
 same plate punched with Kennedy's spiral and ordinary punches respectively, showed 
 an average of nine per cent, in favor of the former. Plates punched with both punches 
 broke in every case through the hole of the ordinary punch. 
 
 It is usual to have the holes ^ inch larger than the rivets, for f -inch rivets, in order 
 to allow for their expansion when hot ; it is evident, however, that the difference be- 
 tween the diameters of the hole and of the rivet should vary with the size of the rivet. 
 
 The hole in the die is made larger than the punch ; for ordinary work the propor- 
 tion of their respective diameters varies from 1 : 1.15 to 1 : 1.2. William Sellers & Co., 
 
SBC. 5. LAYING-OFP, FLANGING, RIVETING, WELDING, ETC. 167 
 
 Philadelphia, use the following rule for proportioning the size of the die-hole : The 
 diameter of the die-hole is equal to the diameter of the punch plus two- tenths the thick- 
 ness of the plate (D = d -\- 0.2 f). By making the die-hole larger than the punch a 
 taper hole is produced in the plate, and the punching can be done with less expenditure 
 of power and with less strain on the plate. 
 
 Daniel Adamson states that "the power required to punch a hole through a steel 
 plate equal to a sectional inch of detruding area may be found by multiplying the 
 maximum tensile strength per square inch by 0.74 of the same metal the detruding 
 area meaning the circumference of the punch multiplied by the thickness of the plate. 
 This law may be depended upon both for soft and hard steels." 
 
 The sheets must always be punched from the "faying" surfaces i.e., the surfaces 
 in contact ; thus no burr or roughness is left around the holes to keep the sheets apart 
 or necessitating time and labor for its removal ; the holes are also better filled by ham- 
 mering down the hot rivet, which assumes the form of two frusta of a cone joined at 
 the small ends, and brings the sheets in closer contact by contraction in cooling (see 
 figure 31). 
 
 Special attention has to be paid to this point in punching the rivet-holes in sheets 
 Fig. 31. that are intended for cylindrical shells and tubes, 
 
 since the holes at the ends have to be punched 
 from different sides, according to the style of 
 joint used. 
 
 Numerous experiments have established the 
 fact that the strength of plates is materially im- 
 paired by punching. A. C. Kirk, in a paper read 
 before the Institution of Naval Architects in 1877, 
 says: "The effect of the punch is clearly shown in the fracture by a portion highly 
 crystalline on each side of the hole. This crystalline fracture around the hole is due to 
 the bursting pressure exerted by the piece punched out, as it tends to spread out in 
 diameter under the intense pressure of the punch. Under this strain the metal is com- 
 pressed for a certain distance round the hole, and thus weakened." This zone of 
 injured metal does not extend farther than inch from the hole, and the injurious 
 effect of punching may be entirely removed by punching the holes somewhat smaller 
 than the size required and reaming or drilling them out afterwards. Annealing after 
 punching restores, likewise, to the strained metal its former strength and elasticity. In 
 many establishments steel boiler-plates are always annealed after punching. The inju- 
 rious effect is diminished by increasing the proportion of the diameter of the die-hole 
 
168 STEAM BOILERS. CHAP. VIII. 
 
 to that of the punch ; and it is much greater for steel, especially the harder qualities, 
 than for iron. 
 
 The results arrived at by different experimenters vary greatly as to the amount of 
 injury produced by punching, owing, no doubt, to differences of condition in the process 
 and in the quality of the material. C. H. Haswell concludes from recent experiments 
 that the resistance of riveted steel plates with the holes drilled is, according to the tem- 
 per of the metal, from 18 to 25 per cent, greater than when they are punched. A. C. 
 Kirk states that when the diameter of the holes is three times the thickness of the plate, 
 or greater, the injurious effect of punching is inappreciable. 
 
 6. Drilling. The practice of drilling the holes instead of punching them comes 
 more and more into favor, because the metal is not injured by this process and the holes 
 are more readily correctly spaced. It is necessary to remove the sharp edge or burr of 
 the drilled hole carefully by slightly countersinking it ; and when the holes have been 
 drilled through two or more plates together the latter must always be taken apart for 
 the purpose of removing the burr. 
 
 Fairbairn considered the exactly cylindrical, parallel, and smooth drilled holes not 
 well adapted to rivets. Besides, punching saves about one-fourth the time and labor as 
 compared with drilling as ordinarily practised. 
 
 Special machinery has been introduced of late for drilling the holes in the various 
 parts of boilers. In Harvey 's Boiler-drilling Machine the cylindrical shell of a boiler 
 is placed vertically upon a central circular turn-table. The machine is provided with 
 three drilling head-stocks, which travel on an outer rail around the boiler, and each one 
 of which can drill one-sixth of the circumference of the boiler without moving the 
 latter. Each head-stock is provided with a vertical traverse for drilling the longitudinal 
 seams. 
 
 A very convenient arrangement for drilling the holes of boilers was designed by 
 Chief-Engineer H. Newell, U.S.N., who describes the method pursued in building a 
 set of cylindrical boilers, 8 feet in diameter, for the U. S. S. Galena, at the Navy- Yard, 
 Norfolk, Va., in a report transmitted to the U. S. Navy Department, in the following 
 words : 
 
 "Figure 2 [Plate XIV.] represents the method of drilling the circumferential butt- 
 straps, which join the two sections of the boiler-shell. A is a shaft of wrought-iron, on 
 which slides and revolves easily the arm B, on one end of which is attached a Laubach 
 Patent Portable Drill, C, the weight of which is counterbalanced at the other end by 
 the weight E, made in halves, secured to the arm by a bolt and nut. The shaft is ad- 
 justed and maintained central to the shell by the tripod D, D, D, clamped tightly to the 
 
SEC. 6. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 169 
 
 shaft by three bolts near the centre, and to the shell by set-screws, as shown. The 
 back end of the shaft is supported by a block of wood bolted to the back-head through 
 the holes intended for socket-bolts. The power is communicated to the drill from any 
 convenient distance and at any angle, through a telescopic shaft and universal couplings, 
 from a counter-shaft (furnished with the patent drill) driven from the fly-wheel of a 
 small Sewell steam-pump, which is secured on blocks and can be transported to any 
 part of the boiler-shop, steam being led to it through a rubber hose and the exhaust 
 steam carried away through another hose. 
 
 " The arm B is moved from hole to hole radially and fore and aft, and secured in 
 position each time by a set-screw and gib in the boss. With this machine the entire 
 circumference is drilled without once moving the boiler, and as many as 360 holes f" 
 diameter have been drilled by it, through iron ^" thick, in one working day of ten 
 hours. The drills are driven at about 250 revolutions per minute. This speed is ren- 
 dered possible by, and the great efficiency of the machine is dependent on, the use of a 
 fine stream of soapy water directed with considerable f orce against the point of the drill, 
 which is kept cool by the rapid evaporation of the spray directed upon it. The water is 
 supplied from a barrel placed about 30 or 40 feet above the work, and conducted to the 
 drill by a rubber hose with a fine nozzle at the end. This height gives a sufficient head 
 to cause the fine stream of water to strike the point of the drill with considerable force 
 and to reach it in a hole of any depth, and it is immaterial whether the drilling is in an 
 upward or downward direction. With the use of oil it would be necessary to move the 
 boiler-shell around so as to bring the drill always in a downward direction. 
 
 " Figure 1 shows the arrangement for drilling the furnace from the inside, and also 
 the arrangement for drilling the flanges of the boiler-heads from the outside. In the 
 former the shaft F, having a flange on one end, is supported at each end by a cast- 
 iron tripod, H, H, H, adjusted in the furnace by means of a set-screw at the end of each 
 arm. The shaft is at liberty to revolve and to move fore and aft in the bosses of the 
 tripods in order to adjust the point of the drill to the holes in the template, and is held 
 securely, while drilling, by a set-screw in the boss. The Laubach Patent Drill is bolted 
 directly to the flange at the end of the shaft F, and is the same as in the last machine, 
 excepting that the limited space makes it necessary to use a shorter feed-screw. With 
 the outside drilling-machine the same shaft A, tripod D, D, D, and centre part of arm 
 B are used as in figure 2 [Plate XIV.] ; one end of the arm B is lengthened by bolting 
 to it a bar of T-iron, bent to the form shown, in order to bring the drill over the edge of 
 the boiler, and also to clear the projection of the furnace-ends. The tripod, being in 
 three pieces, is passed through the manhole separately and bolted together in place. 
 
170 STEAM BOILERS. CHAP. VIIL 
 
 The back end of the shaft is supported by a small casting, G, held in position by a bolt 
 passed through holes in the back-connections intended for securing the bracket to 
 which one of the fore-and-aft braces to the front-head is attached. For drilling the 
 flange of the back-head a bracket is used to carry the arm B in place of the shaft A 
 used for the front-head. The holes intended for socket-bolts are utilized for securing 
 this bracket. 
 
 "A template of iron $" thick is used in the furnaces for guiding the drill and to 
 avoid the loss of time that would be occasioned by the necessity of stopping to ' draw ' 
 the holes if a template were not used. This template is laid out to the proper pitch of 
 rivets, and the holes carefully punched -fa" smaller than the finished size of hole. It is 
 then put in position and the drill run through it and the furnace-sheet, reaming the 
 holes out to the proper size as it goes through. After drilling one furnace the holes in 
 the template are the proper size for the next. This device is found to answer admi- 
 rably, as the twist-drills used follow the punched holes and ream them out equally on 
 all sides. 
 
 " To avoid the expense of making templates for drilling the shells the following 
 method is being adopted for the two new boilers now building here : The outside longi- 
 tudinal straps and the circumferential edges of the shell-sheets are carefully laid out 
 and the rivet-holes punched -fa* smaller than the finished size before bending. The 
 sheets are then bent and fitted together in sections, with the outside and inside longitu- 
 dinal straps held in place by tack-bolts. Each section is then put on a rough turn- 
 table mounted on a car-truck in the machine-shop, and drilled by means of a patent 
 drill capable of being moved up and down on an upright shaft erected for the purpose, 
 while each longitudinal seam is brought to the drill by turning the shell round on the 
 turn-table. The punched holes in the outside straps serve as templates to guide 
 the drill through the shell-sheet and inside strap. This machine averaged about 240 
 holes in a day of ten hours. 
 
 " The two sections are then set up on end, one on top of the other, and the circum- 
 ferential butt-strap carefully fitted and securely tacked. The longitudinal seams of the 
 two sections are riveted up, the shell placed horizontally on rollers, the back-head fitted 
 in, and the circumferential strap is then drilled by the machine represented by figure 2, 
 the f holes previously punched in the edges of the shell-sheets acting as templates to 
 guide the drill, which reams them out to their full size, |f ", an ^ then drills through the 
 solid plate. The rivet-holes for the heads, having been previously punched in the 
 shell-sheet, are used as templates for drilling the flanges of the heads, using the 
 machine represented in figure 1 and previously described. This machine has drilled 
 
SBC. 7. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 171 
 
 
 two rows of holes around the entire head (from 180 to 200) in eight hours, including the 
 
 time expended in moving the shell around to get at the holes at the bottom of the 
 boiler. After the furnaces have been secured in place the front-head is fitted and 
 drilled in the same manner." 
 
 One man and two apprentices are required to run these machines. The man attends 
 to the feed of the drill, one apprentice directs the nozzle of the hose, and the other ap- 
 prentice stops and starts the engine, which is provided with a brake. 
 
 It is necessary 'to be careful, especially in drilling deep holes, to let the jet of the 
 lubricant strike the end of the drill constantly. 
 
 7. Riveting. The holes having been drilled or punched, the sheets are fixed in 
 position and temporarily secured by bolts and nuts. When the holes have not been 
 accurately spaced, so as to be "half -blind," or not perfectly coincident in the joint, 
 Fig. 32. it is a common practice to resort to " drifting" ' i.e., to drive 
 
 through both a tapered steel pin or drift, thus drawing the sheets 
 up and enlarging the holes by main force. This practice pro- 
 duces highly injurious strains on the sheets ; it causes the edges 
 of the holes to bulge and gives to the holes irregular shapes, diffi- 
 cult to fill with the rivet (see figure 32). Sometimes, when the want of coincidence is 
 slight, a smaller rivet is inserted. Both practices are to be condemned ; such imper- 
 fect holes should be reamed or drilled out, and a larger rivet should be used which will 
 fill and cover the enlarged hole completely. 
 
 The operation of riveting by hand requires the services of two "riveters" and one 
 "helper" in the gang, besides the boy who heats the rivets in a forge. The shank of 
 the rivet is brought to a white heat, but the head is not made quite so hot ; care has to 
 be taken not to burn the rivet. The boy passes the rivet to the helper, who places it 
 in the hole, drives the head close up to the plate, and holds a heavy hammer or other 
 mass of iron firmly against the head of the rivet while the riveters beat the protruding 
 end of the shank into the required shape. First, however, they strike a few blows 
 arcmnd the rivet-hole on the plate to bring the sheets into close contact. The first blows 
 on the rivet must fall squarely on the point, so that the rivet is upset throughout its 
 whole length and fills the hole completely before a shoulder is formed. According to 
 the form to be given to the point, the rivet is either beaten down roughly to shape and 
 then finished by a " set," or cup-shaped die, held by one riveter and struck with a heavy 
 hammer by the other ; or it is beaten to a conical shape with light hammers. The ham- 
 mers used by riveters vary from 2 to 7 Ibs. in weight, according to the character of the 
 work and the size of the rivets ; and the holding-up hammers weigh from 10 to 40 Ibs. 
 
172 STEAM BOILERS. CHAP. VIII. 
 
 To drive each f -inch rivet an average of 250 blows of the hammer is needed. The 
 largest rivets that can be worked by hand are 1J inches in diameter. For the operation 
 of riveting expert and skilful workmen are required, that the rivets may be fixed sound 
 and firm and that all unnecessary hammering may be avoided. The conical points of 
 the rivets become brittle and are liable to crack or drop off altogether when the ham- 
 mering is continued after they have grown cold. 
 
 Daniel Adamson, in a paper read before the Iron and Steel Institute in 1878, 
 states as his experience that "nearly all ordinary bar or boiler irons and mild steels 
 will endure considerable percussive force when cold and up to 450 Fahr., after which, 
 as the heat is increased, probably to near 700, they are all more or less treacherous and 
 liable to break up suddenly by percussive action." 
 
 The specimens experimented upon stood the bending test perfectly when cold and 
 at a red heat. 
 
 Modern direct-acting, steam or hydraulic riveting-machines have a cup-shaped die 
 on the end of the piston-rod, which presses against a fixed die. The work is brought 
 into position for riveting by cranes ; the rivets are placed in the holes by hand, the 
 pressure is admitted to the cylinder, and the die on. the piston-rod is pressed forward 
 upon the hot rivet and squeezes it into shape. 
 
 The riveting-machine accomplishes this work with great rapidity and regularity, 
 without the disagreeable noise of hand-riveting ; the sheets are pressed close together 
 during the operation, and the rivets are acted upon while at the proper temperature ; 
 the steady pressure compresses them evenly throughout their length, till the plastic 
 metal flows into every irregularity of the rivet-hole, and the surplus metal may be 
 formed into heads of any size and form. 
 
 It is found that rivets driven by hand fill up the hole very well immediately under 
 the points formed by the hammer, but that the same effect is not produced at every 
 point in the length of the rivet, especially when the holes are irregular. So great is this 
 difficulty that in hand-riveting shorter rivets must be used, because it is impossible to 
 work effectively so large a mass with a hammer as with a machine. The heads of the 
 machine-rivets are, therefore, larger and stronger, and will hold the plates together 
 more firmly than the smaller hand-riveted heads. There are, however, many parts of 
 a boiler that are inaccessible to the machine and must always be riveted by hand. 
 
 Figure 33 represents the steam-riveting machine built by the Providence Steam- 
 engine Company, Providence, H. I. It has an annular die, A, surrounding the 
 cupped die which acts on the rivet, both dies having an independent horizontal motion. 
 The middle lever, B, acts on the cupped die, and the two levers C C act on the annu- 
 
SEC. 7. 
 
 LAYENG-OFP, FLANGING, RIVETING, WELDING, ETC. 
 
 173 
 
 lar die. The long arm of each lever is connected by means of a forked rod to the piston 
 of the single-acting steam-cylinders, D and E. When the work has been moved into 
 position between the standard F, carrying a stationary cupped die which bears against 
 
 FIG.33 
 
 the head of the rivet, and the movable dies, the slotted sliding-bar L is moved by 
 means of the handle H. This sliding-bar operates the steam-valve of each cylinder, 
 regulating the admission and exhaust of the steam, and is adjusted in such a manner 
 as to admit steam to the cylinder E first, so that the annular die strikes a blow around 
 the rivet-hole, forcing the plates into close contact, and while it is held in that position 
 a further motion of the handle H admits steam to the cylinder D which works the 
 cupped die. The latter is made to deliver one or two powerful blows on the point of 
 the rivet. 
 
 In TweddeWs hydraulic machine-tools the power is furnished by an hydraulic ac- 
 
174 STEAM BOILERS. CHAP. VIII. 
 
 cumulator, consisting of a cast-iron cylinder in which a plunger moves which carries 
 the weight producing the pressure on the water pumped into the cylinder. For punch- 
 ing and shearing machines Tweddell uses a water-pressure equal to about fifty atmos- 
 pheres. 
 
 His riveting-machines are worked by special accumulators capable of exerting a 
 pressure of about 100 atmospheres. They consist of a vertical cylinder, loaded with 
 weights, which moves along a plunger fixed at the lower end and having a channel 
 through the centre which establishes communication between the pumps and the annu- 
 lar space between the cylinder and the plunger. The volume of water contained in the 
 accumulator is small, and consequently the fall of the cylinder during the operation 
 of' the machine is relatively great ; the vis viva of the falling counter- weights being de- 
 signed to increase the effect of the water-pressure as the ram of the riveting-machine is 
 arrested. 
 
 The use of hydraulic power has special advantages for riveting-machines : violent 
 shocks are avoided, and the pressure on the ram may be varied at will for different 
 kinds of work by changing the weights on the acciimulator. Each rivet, whether long 
 or short, is driven with a single progressive movement, controlled at will by the operator. 
 
 These riveting-machines are made either stationary or portable. In stationary 
 machines the hydraulic cylinder, made of bronze, is firmly attached, in a horizontal 
 position, on the top of a heavy cast-iron frame ; the die fixed to the end of the ram 
 is made of wrought-iron. 
 
 In the portable riveters the hydraulic ram acts on a lever, the arms of which have 
 the proportions of two to one ; the die is fixed to the short end of the lever, the fulcrum 
 being at the long end, but provision being sometimes made to interchange the position 
 of the die and the fulcrum. A fixed die is attached in a corresponding position to the 
 casting of the hydraulic cylinder. In the different sizes of these portable hydraulic riv- 
 eters manufactured by Wm. Sellers & Co., Philadelphia, the levers are made 6 inches 
 and 12 inches long, 9 inches and 18 inches long, and 12 inches and 24 inches long respec- 
 tively. The portable riveter rests in a frame having the form of a quadrantal arc, by 
 which it is suspended from a hoisting-machine on an overhead-carriage travelling on 
 rails. By this means the riveting-machine can be placed in any position required for 
 the work to be done, and moved over a large area ; the work rests on trestles and the 
 riveting-machine is moved along or around it. The water is carried from the accumu- 
 lator to the riveting-machine through jointed or flexible pipes. 
 
 The operation of the machine is described by the manufacturers as follows : 
 
 " One man raises and lowers the riveter, adjusts it to the rivets, and then closes the 
 
SEC. 7. 
 
 LAYING-OFP, FLANGING, RIVETING, WELDING, ETC. 
 
 175 
 
 dies on the rivets. Boys drop the red-hot rivets into place, with the head of the rivet 
 uppermost in horizontal work. With a skilful operator as many as 6 to 10 red-hot 
 rivets may be put in place ahead of him, and he can, on beam- work, drive from 10 to 
 16 rivets per minute. 
 
 "In using the hydraulic riveting-machine to advantage the rivets should be heated 
 rapidly and uniformly." 
 
 The weight of a portable riveter capable of driving rivets inch in diameter is about 
 450 Ibs. 
 
 The number of rivets put in for a day's work depends upon the diameter of the 
 rivets, their position with regard to greater or less accessibility, the description of the 
 points, and the care taken during the operation. Reed furnishes the following table as 
 representing the practice of hand-riveting at a large English private shipyard (a day's 
 work is taken at ten hours) : 
 
 Position of rivets. 
 
 Diameter of rivets and description of points. 
 
 Number of rivets 
 put in by a set of 
 riveters for a 
 day's work. 
 
 In outside plating 
 
 i ~mch countersunk 
 
 3c to QO 
 
 In bulkheads, etc 
 
 f-inch, snap 
 
 I 80 
 
 In made beams, etc 
 
 f-inch, snap 
 
 2OO 
 
 In beam ends 
 
 i~mch to 4-^-inch hammered 
 
 CO 
 
 In deck-plating 
 
 J-inch countersunk 
 
 5 
 
 
 
 
 Fairbairn states that, with two men and two boys attending to the plates and rivets, 
 his machine could fix 8 rivets of f diameter per minute, while three men and one boy, 
 by hand-riveting, could only fix 40 rivets per hour ; hence the quantity of work done 
 in the two cases was as 12 to 1, and one man's labor was saved. 
 
 Grantham states that with Garforth's riveting-machine 6 rivets can be put in per 
 minute, while 20 rivets per hour is the work of a set of riveters. 
 
 At Pittsburgh, Louisville, and other places west of the AHeghanies rivets are driven 
 cold in boiler- work. It is evident that none but the very best material can be nsed for 
 such rivets, and this is claimed as an advantage for this process ; besides, these rivets 
 are free from the danger of being burnt, and cannot become loose in the hole by con- 
 tracting diametrically in cooling after being hammered down. The extensive use of 
 cold-hammered rivets in the high-pressure boilers of the Western river-steamers proves 
 conclusively that tight joints can be made with cold rivets. When the total thickness 
 
176 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 of the plates is more than 4 inches it is better not to employ hot-riveting, because the 
 contraction in cooling might spring up or shear off the head of the rivet ; instead of 
 using cold rivets of such great length it is often better to use turned screw-bolts of very 
 slight taper. 
 
 In using steel rivets care must be taken to heat them uniformly, not above a cherry- 
 red heat, and to work them down and finish them off as quickly as possible. Reed 
 says on this point: "When the [steel] rivet has not been sufficiently heated, or the 
 riveters have not been expert, they have had great trouble in cutting off the burr, and 
 in doing so have often broken away part of the countersunk point with the burr. On 
 the other hand, if the rivets are heated much above a cherry-red heat they cannot be 
 properly knocked down, as they waste away under the blows of the hammer. If great 
 care is not taken the rivet may be overheated to an extent not sufficient to prevent its 
 being knocked down, but sufficient to greatly deteriorate the quality of the finished 
 rivet. It is advantageous to have plain knock-down or conical points to steel rivets in 
 preference to snap-points, as a burnt or overheated rivet is then more easily detected by 
 the crack round the edges." 
 
 8. Forms of Rivets. Rivet heads and points are of various shapes. Figure 34 
 shows the dimensions of a f-inch rivet of the form most commonly used in boiler- 
 making. In hand-riveted boilers the rivet-points are generally made conical (figure 35). 
 
 Fig. 34. 
 
 Fig. 35. 
 
 Fig. 36. 
 
 Fig. 37. 
 
 The riveting-machine and the hand-set make hemispherical heads and points, also 
 called snap-points (figure 36). In places where it is essential to preserve a smooth sur- 
 face the rivets are countersunk (figure 37). 
 
 The lengths of shank required to form these different rivet-points are given in the 
 following table : 
 
SBC. 9. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 177 
 
 TABLE XXV. 
 
 
 Kind of point. 
 
 Length cf shank required. 
 
 Countersunk point 
 
 for 2 thicknesses of sheets 
 
 i diameter 
 
 *i U 
 
 for 3 thicknesses of sheets 
 
 i diam -|- i inch 
 
 Snap-points 
 
 
 
 
 
 
 
 
 
 Snap-points present generally a greater area of metal to resist shearing than conical 
 points ; they are also considered stronger, because in forming them by means of a die 
 the metal is compressed, while in the hand-hammered conical point the metal is spread. 
 Snap-points cannot well be formed on hand-hammered rivets over inch diameter, 
 because too heavy a hammer would be required. Snap-points are extensively used for 
 interior work in shipbuilding. 
 
 Conical points are considered to make a tighter joint, since they cover a larger sur- 
 face. The height of the cone should be about three-quarters of the diameter of the 
 rivet ; if made too flat they are weak and waste away rapidly through corrosion. 
 
 The enlarged hole necessary for countersunk rivets weakens the plate, while the 
 rivet is correspondingly stronger. They are used necessarily for the outer plating of 
 vessels. In boiler-making they are only used on the strengthening-rings of manholes 
 and other openings, on furnace-fronts, and where it is necessary to clear a flange, etc. 
 Countersunk rivets should be avoided where the stress on them consists in a pull in the 
 direction of their length. 
 
 Sexton recommends to use a uniform angle of 60 for the countersinking tool for 
 Fig. 38. holes of all sizes, and not to countersink the hole down to a thin 
 
 5 \ / ~7 edge, but to leave a portion of it cylindrical, say about one-fourth 
 of the thickness of the sheet (figure 38). Other writers recom- 
 mend to use such an angle that the apex of the cone falls on the line where the shank 
 joins the head of the rivet. 
 
 9. Styles of Joint. Riveted joints are either lap-joints or butt-joints. In the 
 former case the edges of the plates lap one over the other a certain width called the lap, 
 and the rivets are put through both sheets. In the other case the edges butt against 
 each other and are covered by one or two narrow strips of plate called welts or butt- 
 straps, which are riveted to each plate. The rivets are placed either all in one line at 
 
178 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 an equal distance from the edge, or in several rows ; and in the latter case they are put 
 either directly behind each other (chain-riveting) or staggered i.e., in zigzag lines. 
 
 The following are the principal styles of joint used in boiler-making : 
 
 Figure 39 represents a single-riveted lap-joint. 
 
 Figure 40 represents a double-riveted lap-joint. 
 
 Figure 41 represents a single-riveted butt-joint with single butt-strap. 
 
 Figure 42 represents a single-riveted butt-joint with double butt-strap. 
 
 Figure 43 represents a double-riveted butt-joint with double butt-strap. 
 
 Figure 44 represents a chain-riveted butt-joint with double butt-strap. 
 
 Fig. 39. 
 
 Fig. 40. 
 
 Fig. 41. 
 
 Fig. 42. 
 
 Fig. 43. 
 
 Fig. 44. 
 
 In bridge-building and ship-building a greater number of rows of rivets are often 
 advantageously employed. In boiler-making the objects to be kept in view in selecting 
 a special kind of joint and proportioning it are strength, economy of material and labor, 
 and tightness. 
 
 1O. Friction in Riveted Joints. E. Clark and Reed have made experiments to 
 determine the amount of friction in riveted joints due to the force exerted on the sheets 
 
SEC. 10. 
 
 LAYING-OPF, FLANGING, RIVETING, WELDING, ETC. 
 
 179 
 
 by the contraction of the rivets in cooling. Reed describes his experiments in the fol- 
 lowing manner : " Three plates were united by what is known as a chain-joint that is, 
 the ends of the two outer plates overlapped the end of the middle plate. The connec- 
 tion of the plates was made by three rivets passing through the lap, the rivet-holes in 
 the outer plates being filled by the rivets, but the bearing-surface of the holes in the 
 middle plate being slotted out as shown in figure 45. It will thus be Fig. 45. 
 obvious that when a tensile strain was brought upon the middle plate 
 the amount of the friction could be measured by the force just able to 
 produce a sliding motion. The breadth of the lap was three diame- 
 ters, the rivets were a diameter clear of the edge of the plates, and 
 their pitch was four diameters." 
 
 Both iron and steel plates were experimented on with different 
 kinds of rivets ; the dimensions of the plates were \" X 8J", the rivets 
 being f " ; and " x 11", the rivets being 1 inch. The mean weight re- 
 quired to cause the plates to slide was 4.95 tons per rivet. Reed sums 
 up the result of his experiments in the following words: "It thus 
 appears that rivets with pan-heads and conical points have the advan- 
 tage over both the other descriptions of riveting." . . . "It also becomes evident 
 that countersunk riveting causes much less friction than the other systems. On com- 
 parison it will be seen that in nearly all cases steel plates and rivets give less friction 
 than iron." . . . "The use of larger rivets with the same pitch, etc., gives an in- 
 crease in the friction, but no law of increase appears to be conformed to." 
 
 The results of Clark's experiments did not differ much from Reed's. 
 
 In commenting on these results "Wilson remarks: "It must not, however, be con- 
 cluded that the value of a rivet is to be determined by adding to its shearing strength 
 the amount of friction between the plates produced by its contraction in cooling. Al- 
 though these two elements of strength act together in a well-filled hole, they cannot be 
 considered as acting independently." ..." The manner in which a severe tensile 
 strain affects a lap-joint by pulling it athwart the line of strain (see figure 46) must also 
 
 tend to diminish the friction of the plates. Long 
 before the ultimate resistance of the joint is 
 
 Fig. 46. 
 
 1 reached, especially with single-riveting, the fric- 
 tion of the plates must be greatly diminished, 
 and cannot be regarded as materially influencing 
 the ultimate strength of the joint. In old boilers it is probable that the tension of the 
 rivet becomes gradually eased by the continual straining and alteration of temperature, 
 
180 STEAM BOILERS. CHAP. VIII. 
 
 which will in time affect the nature of the iron." ... " There can be no doubt that 
 severe calking, as commonly practised, must tend to diminish the friction between the 
 plates, especially when they are thin." 
 
 11. Straining Action on Riveted Joints. "A riveted joint is in a certain 
 sense an imperfect part of a structure. It cannot be so designed as to be throughout 
 uniformly strained. It has always certain surfaces markedly weaker than the rest, at 
 which consequently deterioration of the material or fracture by the action of the load is 
 liable to occur. These surfaces of weakness are so related that in general the increase 
 of one involves a diminution of the other. The joint, therefore, which will carry the 
 greatest load before fracture will be that in which the stress reaches the breaking limit 
 for each of these surfaces simultaneously. Since the rivet-section can in general be in- 
 creased only at the expense of the plate-section, in the strongest joint the rivet and 
 plate will reach their breaking-point under the same load. It would seem, therefore, 
 that the proportions of a riveted joint could be determined by the ordinary rules of 
 applied mechanics without the need of experiment. That this is not so is probably 
 mainly due to a second condition of imperfection in riveted joints. To apply the 
 ordinary rules for the strength of materials to riveted joints it is necessary that 
 the distribution of the stresses on the surfaces of weakness should be known. If 
 those stresses were as uniformly distributed as in an ordinary bar tested for tension 
 or for shearing, the problem would be simple. But, in fact, the stresses are less 
 uniformly distributed and the law of their distribution is unknown. Consequently 
 the average stress on the surface of fracture of a riveted joint, when broken by a 
 load, is less than it would be if the stress were uniformly distributed, and needs 
 to be determined by special experiments. Further, it may be different for dif- 
 ferent forms of joint. This average stress, always less than the maximum stress 
 which causes fracture, is here termed the apparent breaking stress. Hence the 
 chief object of experiments on riveted joints is to determine the apparent break- 
 ing stresses 
 
 (1) for the different surfaces at which each joint may fracture, 
 
 (2) for the different forms of joint. 
 
 "In certain cases allowance may have to be made for progressive deterioration of a 
 joint, by corrosion or otherwise, which reduces the strength in certain directions more 
 than in others. No experiments showing the amount of deterioration in such cases 
 appear to have been made. 
 
 "Let a bar be broken at the plane ab of area w (figure 47) by a tension, T, acting 
 normally to the section. 
 
SEC. 11. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 181 
 
 " Then if the stress is uniformly distributed over the section at the moment of 
 
 T 
 
 fracture the ratio - - is the real tenacity of the material. But if it is not uniformly 
 
 w * 
 
 Fig. 48. Fig. 48.o pig. 49. F'g- 50 
 
 P tP e a ic 
 
 T 
 
 distributed then -- is only the apparent tena- 
 city, and this may be less than the real tena- 
 city to any extent whatever. It may be useful 
 to consider under what conditions the distri- 
 bution of stress necessarily becomes unequal. 
 
 " (1.) It will cease to be uniform if the 
 resultant P of the load does not pass through 
 
 the centre of figure of the section. Thus, in the case shown in figure 48 the stress is a 
 varying stress, which, however, varies regularly so long as the limit of elasticity is not 
 passed. Some of the discrepancies in the results of experiments on riveted joints are 
 probably due to want of care in ensuring the coincidence of the line of action of the 
 load with the centre line of the joint, in the plane parallel to the surface of the plates. 
 Figure 48a shows how unequal distribution of stress may arise from this cause. In the 
 plane at right angles to this there is probably always deviation of the load from the 
 centre of figure. In lap-joints the load has to be transmitted from one plate through 
 the rivet to the other plate ; in butt-joints from one plate through the rivets to the 
 cover-strip and back to the other plate. In both cases, and especially in the former 
 case, the eccentricity of the load appears to cause a reduction of strength. 
 
 " (2.) The stress maybe rendered unequal by the local action of contiguous material. 
 Thus, a bar with square corners (figure 49) is known to break with a low apparent 
 tenacity. The unstrained material at a prevents the elongation of the contiguous mate- 
 rial at b, which consequently gets an excessive proportion of the load, and the fracture 
 begins at the corners. 
 
 "Now, in the portion of metal between two rivet -holes a similar action probably 
 occurs. The outside fibre a b (figure 50) has less freedom of elongation than the central 
 fibre c d, because it is attached to the comparatively unstrained material behind the 
 rivet. Hence, instead of breaking simultaneously over the whole section, fracture 
 probably begins at the edges of the hole, and proceeds because the reduction of area 
 causes increase of stress in the part remaining unbroken. This is sometimes shown by 
 the fact that the parts of the plate will not fit after fracture. There appears to be a 
 slight reduction of strength in plates with a hole drilled in them as compared with solid 
 plates, and this is probably due to the cause now under consideration. It is also pro- 
 bable that this reduction of strength may really be greater than appears in these 
 
182 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 experiments. Short bars are known to give a higher average tenacity than long bars. 
 Now, a bar with a hole drilled in it is virtually a very short bar, and it ought, there- 
 fore, if there were no cause of diminution of strength, to show a higher tenacity than 
 ordinary test-bars. But, in fact, there is generaUy a loss of strength. 
 
 "In some experiments there is a curious apparent increase of strength after drilling. 
 Thus, in one of Mr. Stoney's experiments the drilled plate was 7J per cent, stronger 
 than the undrilled plate. In some experiments by Mr. Parker the plain plate carried 
 26.4 tons, while a plate punched and annealed carried 31.7 tons. See also the table of 
 treble-riveted joints given further on. (Section 15 of the present chapter.) Mr. Adam- 
 son also finds that the tenacity through a line of drilled holes is a little greater than the 
 tenacity of the plate before drilling. Discrepancies of this kind may be due to the 
 holes causing fracture at a section stronger per square inch than other parts of the 
 plate. An ordinary test-bar breaks at the weakest part of a more or less considerable 
 length of bar. 
 
 " (3.) If the material in the neighborhood of the surface of fracture is initially (before 
 the application of the load) in an irregularly-strained condition, or has in different 
 parts unequal power of elongating, then the stress will not be uniform at the moment of 
 fracture, and the apparent tenacity will be less than the real tenacity. This is the 
 cause of the loss of strength due to punching. By the action of the punch metal is 
 caused to flow laterally into the surrounding metal. This induces initial stresses in an 
 annulus of metal round the hole, and very probably also, as M. Barba thinks, alters its 
 power of elongation. If the power of elongating is diminished in part of the metal, 
 that part gets an excessive proportion of the load and breaks before the rest is fully 
 
 strained. The result of either loss of 
 tenacity or loss of ductility is to di- 
 minish the apparent tenacity of the 
 metal to an extent which certainly 
 reaches in some cases 20 to 30 per cent. 
 Figure 51 shows a possible condition of 
 the bar after punching, the ordinates 
 of the dotted curves representing the 
 stresses. Immediately round the hole 
 is an annulus in which the stress is 
 compressive, the compression being due 
 to material forced in. To balance the forces in this ring an annulus in which the stress 
 is tension must surround it. 
 
 Fig. 51. 
 
SEC. 12. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 183 
 
 "In some experiments there appears to occur a serious diminution of the apparent 
 tenacity in riveted joints when the bearing-surface of the rivets on the plates is too 
 small, and when consequently the crushing pressure between the rivet and plate is ex- 
 cessive. It is possible that this is due to an action like that which occurs in punching. 
 The pressure of the rivet may cause a lateral flow of the metal, and alter either the 
 stress or the elasticity of a ring of metal round it. The stress on the tearing section 
 being then unequal, a low apparent tenacity is found." (First Report of the Commit- 
 tee of the Institution of Mechanical Engineers on the Form of Riveted Joints.} 
 
 In multiple-riveted joints of materials of different elasticities e.g., steel andiron, or 
 cast and wrought iron the outermost row of rivets has to bear the greater stress. If 
 
 Fig. 52. 
 
 ' 
 
 the elastic rod S (figure 52) is riveted to a non-elastic body, 
 K, by several rows of rivets, the row 1 must bear the entire 
 stress B, for the part of B assigned to 2 must act by tension 
 on 1 2, tending to stretch it. Since 1 does not yield, on ac- 
 count of the deficient elasticity of K, the part of B assigned 2 
 is transferred back to 1 by compression. 
 
 If two bodies, whose elongations for the same stress are 
 nearly equal, are riveted double or triple they strive to attain 
 unequal elongations between rivets, because the forces acting 
 on the adjacent parts are not equal. Denote the total tension 
 on the joint by B, the stresses on the rivets (figure 53) by 
 I, n, III, and I', II', III', and the stresses on the interme- 
 diate portions of the plates by I H, II III, and I' II', II' HI', 
 then 
 
 i ii = T IT = B - 1 
 
 II = II' and III = HI' 
 II HI = II' III' = B - I - H 
 
 The parts I II and II' III' are therefore under the action of forces of different magni- 
 tudes viz., B I and B I II. The rivet I cannot yield to the elongation of I II, 
 and a portion of this force must act as pressure on I. The same holds true of the por- 
 tions II III, I' II', and the rivet I'. 
 
 Hence the weak point of every riveting which is more than double lies near the 
 outermost rivet in the direction of the strain. ( Weyrauch, ' Strength of Iron and Steel 
 Construction.') 
 
 12. Strength of Materials in Riveted Joints. The tensile strength of boiler- 
 
184 STEAM BOILERS. CHAP. VI1L 
 
 plates at the joints per square inch of section is generally less than that of the original 
 plates ; but this loss of tenacity varies according to the treatment received by the plates 
 in the process of construction. When the rivet-holes are drilled the strength of the 
 material is not diminished to an appreciable extent. When the holes are punched the 
 loss of tenacity varies with the form of punch used and with the quality of the mate- 
 rial ; it is greater for hard than for very ductile materials, and is generally greater for 
 steel than for iron ; it increases with the thickness of the plates, and as the diameters 
 of the punch and of the die-hole are more nearly alike. Experiments on punched iron 
 plates show a loss of tenacity varying from 5 to 20 per cent, of the original strength of 
 the plates. In steel plates punching produces a loss of tenacity varying from 8 to 35 
 per cent, of the original strength ; but the plates can be restored to their original tena- 
 city by annealing them after punching, or by reaming out the punched holes. (See sec- 
 tion 5 of the present chapter.) 
 
 In an experiment made by Adamson the strength of a perforated bar was increased 
 5.8 per cent, by driving a turned pin into the hole, so as to prevent the metal round the 
 hole from collapsing into an elliptical shape ; thus producing more nearly the same con- 
 dition as obtains in riveted joints. 
 
 It is generally assumed that the shearing strength of wrought-iron is 80 per cent. 
 of its tensile strength, if the shear is in a plane perpendicular to the direction of rolling, 
 and if the tension is applied parallel to the direction of rolling. In a paper on the 
 strength and proportions of riveted joints, by W. R. Browne, communicated to the 
 Institution of Mechanical Engineers in 1872, it is assumed that the shearing resistance 
 of iron rivets 
 
 in single-shear is 22 tons per square inch. 
 
 in double-shear is 21 " " 
 
 Some of Fairbairn's experiments on the shearing resistance of rivets gave the following 
 results : 
 
 Eivet-holes drilled, edges of holes sharp 19.23 tons per square inch. 
 
 Rivet-holes drilled, edges of holes rounded 21.52 " " 
 
 Rivet-holes punched 20.95 " " 
 
 Experiments made by David Greig and Max Eyth on Taylor's Yorkshire rivet-iron 
 and Brown & Co.'s mild rivet-steel gave for the tensile strength of the iron 22.2 tons per 
 square inch, and of the steel 28.8 tons per square inch. The shearing resistance of the 
 iron was 19 tons, and that of the steel 22.1 tons. Some plates riveted together were 
 then tested, and a somewhat higher shearing resistance was found than for bars not 
 formed into rivets. This is ascribed partly to the rivet being increased in diameter to 
 
SBC. 12. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 185 
 
 fill a hole larger than its normal size, partly to the friction of the plates. The harden- 
 ing of the rivet is a possible cause of increased resistance of rivets as compared with 
 simple bars. 
 
 The crushing pressure of the rivet on the plate is discussed by Professor W. C. 
 Unwin in the ' First Report of the Committee of the Institution of Mechanical Engi- 
 neers on the Form of Riveted Joints,' as follows: 
 
 " If F is the tension on a joint corresponding to one rivet, 
 d the diameter of the rivet, and 
 t the thickness of the plate, 
 
 then C = [I.] 
 
 may be denned as the mean crushing pressure of the rivet on the plate. 
 
 "Putting Sfor the shearing resistance of the rivet, then, the rivets being in single 
 shear, 
 
 - = .786 -, [II.] 
 
 or the crushing pressure is greater as the ratio of the rivet diameter to the thickness of 
 plates is greater. 
 
 "This is sometimes given as the reason why the rivet diameter should not exceed 
 2 to 3 times the plate thickness. It is by some writers asserted that if in any case 
 C is more than 30 or 40 tons per square inch for iron joints, then the joint gives way 
 with a very low apparent tenacity. During the application of the load the rivet-hole 
 becomes oval, the metal of the plate is crushed and its tenacity diminished. The pre- 
 cise way in which the crushing affects the tenacity has not hitherto been indicated ; 
 but it is suggested above (see section 11 of the present chapter} that it produces an 
 unequal distribution of the stress similar to that induced by punching. There are no 
 direct experiments on the crushing of iron and steel which are of any value in deter- 
 mining the proper limits of crushing pressure for riveted joints. . . . 
 
 "From the very irregular distribution of the pressure on the surface of the rivet it 
 is probable that the maximum pressure of the rivet on the plate is much greater than 
 its mean value C." 
 
 A mathematical investigation of the stresses indicates that the maximum crushing 
 pressure is 1.27 times the mean crashing pressure, but the writer is of the opinion that 
 in practice the value of the maximum crushing pressure is much greater, especially with 
 rivets in single shear. 
 
186 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 Discussing some results obtained with actual joints which have a bearing on this 
 question, he finds that there seems to be a tolerably regular increase of apparent tena- 
 city as the crushing pressure diminishes, and that the diminution of tenacity is sen- 
 sible in lap-joints where the crushing pressure exceeds 30 tons, and was very great in 
 some cases where the crushing pressure reached 40 tons. These remarks apply, how- 
 ever, only to iron lap-joints. Experiments with butt-joints show great anomalies. 
 
 "With steel joints also, even with very high crushing pressures, no regular effect 
 on the tenacity is traceable. It seems possible to the reporter that the explanation of 
 these anomalies may be found in the variation of the relative hardness of the rivets and 
 plate. If the rivet is sensibly harder than the plate, the plate will suffer ; but if the 
 rivet is sensibly softer than the plate, the rivet will suffer. With iron plates some- 
 times the rivet and sometimes the plate is the harder. With steel the rivet appears to 
 be generally softer than the plate. It must be borne in mind that this suggestion is 
 only offered as a conjectural explanation of anomalies which, unless they are due to 
 errors in the experiments, are extremely puzzling." 
 
 13. Proportioning Riveted Joints. A riveted joint subjected to tension can 
 break 
 
 (1) By the tearing of the plate ; in this case its strength is measured by the tensile 
 strength of the plate multiplied by its least sectional area, which obtains on a line pass- 
 ing through the rivet-holes, and depends upon the thickness of the plate and the diame- 
 ter and spacing of the rivets ; 
 
 (2) By the shearing of the rivets ; in this case its strength is measured by the shear- 
 ing strength multiplied by the sectional area of the rivets ; 
 
 (3) In consequence of the thrust exerted by the rivets on the plate, which may cause 
 it either to be crushed (figure 54), or to split (figure 55), or to have a portion sheared 
 
 Fig. 54. 
 
 Fig. 55. 
 
 Fig. 56. 
 
 out (figure 56) from the rivet-holes to the edge of the plate ; and in this case the 
 strength of the joint depends on the diameter of the rivets, the thickness of the plate, 
 
SEC. 13. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 187 
 
 the width of the lap, and a coefficient of resistance depending on the nature of the 
 fracture. The shearing of the plate from the rivet-holes to the edge is, however, not 
 likely to take place with the ordinary proportions of lap and rivets. 
 
 The thickness of the plates, the diameter and spacing of the rivets, and the width of 
 the lap must be proportioned in such a manner that the strength of the joint approaches 
 as nearly as possible the strength of the whole plate, and that the same liability exists 
 for the different kinds of fracture to take place ; at the same time the tightness of the 
 joint and facilities of construction have to be taken into consideration. 
 
 Assuming that the average shearing strength of iron rivets is 19 tons per square inch, 
 and that the crushing pressure of the rivets on the plates should not exceed 30 tons 
 per square inch (see sections 11 and 12 of the present chapter), we can find the proper 
 diameter of a rivet for a given thickness of plates by introducing these values into for- 
 mula [//.] of section 12 of the present chapter, viz.: 
 
 - - . 
 
 S 19 t 
 
 consequently d = 2 1, 
 
 for joints in which the rivets are in single-shear. 
 
 When the rivets are in double-shear they will bear about 90 per cent, more than the 
 same rivets in single-shear, and under these conditions equation [II. ] assumes the fol- 
 lowing form, viz. : 
 
 Odt = 1.908 ff; 
 4 
 
 hence .==!.. 
 
 consequently d = 1.05 t. 
 
 In practice, when the rivets are in single-shear, d is generally made equal to 2 t for 
 plates up to f inch thick. But this proportion is gradually decreased for thicker 
 plates, because the formula d 2t would give rivets of so large a diameter that they 
 could not be spaced close enough to make a steam-tight joint and at the same time 
 make the plates and rivets of equal strength ; and, when the thickness of the plates 
 exceeds ^ inch, the rivets would become so large that they could not be properly 
 worked down. In boiler-making rivets exceeding l inches in diameter are rarely used. 
 
 When the plates to be connected are of unequal thickness the diameter of the rivets 
 is proportioned to the thicker plate. When more than two plates are to be connected 
 the diameter of the rivets is increased by about one-eighth inch. 
 
 For steel boiler-plates it is better to use steel rivets than iron xivets, which should 
 
188 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 Fig. 57. 
 
 be of somewhat smaller diameter in proportion to the thickness of the plates, and 
 spaced correspondingly closer, than iron rivets with iron boiler-plates. 
 
 The diameter of the rivets being determined by the thickness of the plates, it is con- 
 venient to express the pitch of the rivets and the width of the lap in terms of the 
 diameter of the rivets. 
 
 The width of the lap, measured from the 
 centre of the rivet-hole to the edge of the 
 plate, has been fixed practically at 1.5 d. 
 
 (This gives ample strength to resist the thrust 
 of the rivet, and makes a proper allowance 
 for calking; too large a lap does not make 
 a tight joint, since the sheets are apt to 
 be forced apart by ' severe calking (figure 
 57). 
 
 14. Lap-joints. The single-riveted lap-joint requires less^ labor and material 
 than any other riveted joint ; but its strength, compared with that of the solid plate, is 
 small, because the sectional area of the plate in the- line of rivets is greatly reduced, 
 and on account of the unequal distribution of the stresses. In boilers this joint is used 
 especially in the furnaces and back-connections, where it is advantg,gedBMmake the 
 lap as narrow as possible, and in those parts which are not subjected 
 and where great strength and tightness are not required, as for 
 connections, etc. 
 
 Calling the total tensile force applied to a single-riveted joint 
 
 the ultimate tensile strength of the plate per unit of area .......... T, 
 
 the ultimate shearing strength of the rivets per unit of area ........ S; 
 
 and representing the number of rivets in the joint by .............. n, 
 
 
 
 their diameter by ............ .................................. d, 
 
 the thickness of the plates by .................. ............... . . t, 
 
 the pitch of the rivets by .................... * ..... .............. p, 
 
 we can express the width of the joint by ......................... np. 
 
 The dimensions of the rivets and of the plate being so proportioned that they offer 
 equal resistance to F, and supposing this stress to be borne equally by every part of the 
 plate in proportion to its sectional area, we have the equation : 
 
 F=nd* .7854 8=n(p d)t T; 
 
 ressure 
 
 giptakes, 
 
 $. . . 
 
 and 
 
 p = d? .7854 
 
 , 
 
 t 
 
 d. 
 
SEC. 14. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 189 
 
 Since, with the dimensions ordinarily used in boiler-making, the value of d varies be- 
 tween 2 1 and 1.5 1, the values of p lie between d ( 1+ 1.5708 ^ \ and d (l + 1.1781-^Y 
 
 In using these formulae for calculating the value of p we must insert for T and 8 
 the values of the apparent tensile and shearing strengths of plates and rivets in single- 
 riveted lap-joints, as found by experiment. (See section 15 of the present chapter.} 
 The sectional area of the plates in a boiler is reduced continually by corrosion, while 
 the shank of the rivet remains intact. This action must be taken into account in pro- 
 portioning a joint. 
 
 D. K. Clark says that "the shearing section of rivets should not in any case exceed 
 the net section of the plate, and that the maximum strength of joint is attainable when 
 the shearing section is from 90 to 100 per cent, of the net section of the plate." Making 
 the area of the rivets 90 per cent, of the net section of the plate, the value of p, for 
 d = 2t, becomes j? = 2.745d; and for d = 1.5 t, p 2.309 d. These values do not differ 
 much from ordinary practice. 
 
 The double-riveted lap-joint is from 20 to 33 per cent, stronger, and is more easily 
 kept tight, than the single-riveted joint. It is used most extensively for steam-tight 
 joints which do not come in contact with the fire and hot gases. 
 
 Retaining the notation given above, we have the equation : 
 
 
 
 . 78548= (p-d)tT; 
 
 & 
 
 Fig. 58. 
 
 Fig. 59. 
 
 and p = d 1 
 
 Making the area of the rivets 90 per cent, of the net section of the plate, the value 
 of p, for d = 2t, becomes 
 
 p = 4. 4907 tf; 
 and for d = 1.5 1, p = 3.618 d. 
 
 The rivets of a double-riveted joint in 
 boilers are generally placed in a zigzag line. 
 Some of Brunei's experiments show that 
 when the rows of rivets are too close the line 
 of fracture is a zigzag, running backward 
 and forward between the rows (figures 58 
 and 59), and a much greater section of 
 metal is divided than if the fracture took 
 
 place on a line passing through the centre of the rivet-holes in either row. This is 
 explained by the fact that punching weakens the sheet to some distance around the 
 
190 STEAM BOILERS. CHAP. V11I. 
 
 hole, and that the amount of this weakening effect on the area represented by the 
 zigzag line is twice as great as that on the area represented by the straight line be- 
 tween two contiguous holes in the same row, as has been illustrated by the shading 
 around the holes. Brunei found that the distance between the two rows of staggered 
 rivets should be two-thirds of the pitch of the rivets. 
 
 In chain-riveting it is safe to make the distance between the centre lines of the rows 
 of rivets equal to 2 diameters of the rivets. 
 
 Treble and quadruple riveted lap-joints are sometimes, but rarely, used for the 
 shell of cylindrical boilers. With plates of ordinary thickness, in which the diameter 
 of the rivets is from 1 to 2 times the thickness of the plates, multiple-riveting makes 
 the pitch of the rivets so large that the joint cannot well be calked steam-tight. This 
 objection does not exist in the case of thick plates, in which the diameter of the rivets 
 exceeds but little or nothing the thickness of the plates. The increase of strength ob- 
 tained by increasing the rows of rivets is, however, not proportionate to the additional 
 labor and material required for making the joint, on account of the very unequal dis- 
 tribution of the stresses. (See section 11 of the present chapter.) The dimensions of 
 multiple joints may be calculated by formulae similar to those given for single and 
 double riveted lap-joints, introducing for S and T the values given in Table XXVI. 
 
 15. Experiments on the Strength of Lap-joints. The experiments made by 
 Fairbairn in 1838 have served up to the present time as the basis for calculating the 
 strength of riveted joints. According to these experiments the strength of a double- 
 riveted joint is 70 per centum of the strength of the plate, and of a single-riveted joint 
 56 per centum. Of these experiments it is necessary to remark : 
 
 1st. That the results are only for the case in which the rivet-holes diminish the sec- 
 tion of the plate 30 per centum, while for the most part in practice, and particularly 
 for the single-riveted joint, that loss is very much greater. 
 
 2d. That the experiments were made on plates of only 0.224 inch thickness. 
 
 3d. That the experiments gave 46, and not 56, per centum for the strength of the sin- 
 gle-riveted joint ; the coefficient was arbitrarily increased by Fairbairn to cover certain 
 imperfections in the experiments. 
 
 This increase was partly made for the purpose of allowing for the increase of 
 strength given to riveted joints by arranging contiguous plates in such a manner that 
 their joints do not lie in the same line ; but the increase of strength due to this arrange- 
 ment is much greater with narrow plates, such as were formerly in general use, than 
 with wide plates, such as are nowadays manufactured. 
 
 Experiments on various plate-joints made by W. Bertram at Woolwich Dockyard 
 
SEC. 15. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 191 
 
 I 
 
 were published and discussed in 1860 by D. K. Clark. The thicknesses of the plates 
 were f inch, T \ inch, and inch, and in the single-riveted joint the net sectional area of 
 the plates in the line of rivets was 62.5 per cent, of the solid plate. The relative 
 strength of the joints of the f-inch plate is given by him as follows : 
 
 Entire plate 100 
 
 Double-riveted joint 72 
 
 Single-riveted joint 60 
 
 He found that in the f-inch plate the tensile strength per square inch of net section 
 of the single-riveted joint was nearly equal to that of the entire plate, while in the ^V- 
 inch plate it was only four-fifths, and in the -inch plate two-thirds, of the strength of 
 the entire plate, so that the joint of the thinner plates was actually stronger than that 
 of the thicker plates. This remarkable reduction of strength in thick plates is as- 
 cribed to the distorting leverage of the lap, which increases with the thickness of the 
 plates. Figure 46 shows the ultimate distortion of lap-joints by the oblique action of 
 the tension. The decrease of strength with the thickness of the plates in these experi- 
 ments is, however, much greater than has been found in more recent experiments, and 
 the strengths of the riveted joints of the f-inch plates, as given above, are likewise 
 greatly in excess of the average results of experiments. 
 
 Clark found also that countersunk riveting did not impair the strength of the joint 
 as compared with external heads, which result he explains, likewise, by the oblique 
 stress on the lap-joint. He says : " On the principle here noticed one may account for 
 the practically equal strength of the joints made with countersunk rivets, compared 
 with those having external rivet -heads, notwithstanding the greater reduction of solid 
 section by countersinking : the leverage is shortened and it may be measured from the 
 Fig. 60. centre of the cylindrical part of the 
 
 ~J rivet in the line a b (figure 60), or 
 
 I thereabouts, toward the inner side 
 of the plate. On the same princi- 
 ple the conical form of punched holes reduces the leverage and the obliquity of the 
 pulling stress." 
 
 In the above-mentioned ' First Report of the Committee of the Institution of Me- 
 chanical Engineers ' the most reliable experiments on riveted joints have been tabu- 
 lated, all experiments being omitted in which the crushing pressure of the rivets on 
 the plate was so great as probably to have affected in a considerable degree the appa- 
 rent tenacity of the joint. The ratio of the tension on the joint to the area of the sec- 
 
192 STEAM BOILERS. CHAP. VIII. 
 
 tion at the place of fracture is called the apparent tenacity of the joint, which is ren- 
 dered less than the original tenacity of the iron by any injury done in drilling or punch- 
 ing, and by the irregularity of stress due to the crushing action between the rivets and 
 plates, and by the irregular distribution of stress due to bending of the joint under the 
 action of the load, etc. 
 
 From a large number of experiments on single-riveted lap-joints of iron plates it 
 appears that the apparent tenacity of the plate in this joint is from 20 to 32 per cent, 
 less than that of the original plate, with punched holes, and about 12 per cent, less with 
 drilled holes. Since iron plates do not receive any appreciable injury in drilling, this 
 loss in tenacity of 12 per cent, has to be ascribed mainly to the irregular distribution of 
 the stress. 
 
 The mean shearing resistance of the rivets is about 6 per cent, greater in punched 
 holes than in drilled holes. With punched holes the ratio of the apparent tenacity of 
 the plates to the shearing resistance of the rivets is 85 to 100, but with drilled holes the 
 plates are stronger per unit of area than the rivets in the ratio of 107 to 100. 
 
 The mean efficiency of the single-riveted lap-joint, in per cent, of the tenacity of the 
 solid plate, is 44.6 per cent, when the holes are punched and 50 per cent, when the holes 
 are drilled. 
 
 The mean results of nine experiments with double-riveted lap-joints of iron plates 
 with punched holes give an apparent tenacity of the plate of 89.5 per cent., and an 
 efficiency of the joint equal to 59 per cent, of the tenacity of the solid plate. 
 
 Two experiments with drilled iron plates and double-riveted lap-joints, by Greig and 
 Eyth, give a mean apparent tenacity of 95 per cent, and a mean efficiency of joint of 61 
 per cent. 
 
 Several experiments with double-riveted lap-joints of punched iron plates 1 inch and 
 | inch thick, made by Kirkaldy for R. V. J. Knight, gave remarkably low results. 
 The mean apparent tenacity of the plates at the joint was only 56.4 per cent, of the 
 tenacity of the solid plate ; and the mean of four experiments with 1-inch plates gave 
 34.5 per cent., and the mean of two experiments with f-inch plates gave 42.6 per cent., 
 for the efficiency of the joint. This great reduction of strength appears to have been 
 due to the unequal distribution of the stress in consequence of the bending of the 
 joint under the action of the load, since two similar iron plates, 1 inch thick, with 
 punched holes, but forming a double-riveted butt-joint with double covering-plates, 
 which were tested by the same parties, gave an apparent tenacity of 90 per cent, of 
 the tenacity of the solid plate. 
 
 The average values of a number of experiments with double-riveted steel lap-joints 
 
SEC. 15. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 193 
 
 make the apparent tenacity of the plates at the joint from 4 to 8.5 per cent, greater 
 than the tenacity of the solid plate, and 21 per cent, greater than the shearing resistance 
 of steel rivets per square inch of sectional area. 
 
 Table XXVI. contains the results of experiments with treble and quadruple riveted 
 lap-joints made of steel plates with steel and iron rivets, and is compiled principally 
 from the ' First Report of the Committee of the Institution of Mechanical Engineers 
 on the Form of Riveted Joints.' The joints were made partly by Denny & Co., of 
 Dumbarton, and the specimens were tested by Kirkaldy. The stresses corresponding 
 to the actual mode of fracture of the joints are printed in heavy type; the other 
 stresses, printed in ordinary type, are those which obtained at the moment of frac- 
 ture, but are lower than those at which the joint would have given way by the respec- 
 tive modes of fracture. The joints marked a, b, and c in the table and their lines of 
 fracture are represented in figures 61, 62, and 63 respectively. The steel plates had a 
 
 Fig. 61, 
 
 Fig, 62. 
 
 Fig. 63. 
 
 T~ 
 
 th- 
 
 nominal thickness of | inch, and the steel rivets had a diameter of 1.13 inches. The 
 tensile strength of the rivets in these four joints was 28.9 tons per square inch, and 
 their apparent shearing strength varied consequently from 66.5 to 71.4 per cent, of their 
 tenacity. 
 
194 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 TABLE XXVI. 
 
 RESULTS OF EXPERIMENTS WITH TREBLE AND QUADRUPLE-RIVETED LAP-JOINTS. STEEL PLATES. 
 
 TESTED BY KIRKALDY. 
 
 Mode 
 of riveting. 
 
 Holes. 
 
 !r 
 
 i 
 
 gg 
 
 J 
 
 Sjs 
 
 rt rt o 
 
 J3J 
 
 Stress at moment of fracture, in 
 tons per square inch. 
 
 Apparent tenacity of 
 plate at joint in per 
 cent, of tenacity of 
 solid plate. 
 
 Efficiency of joint. 
 
 Thickness of plate. 
 
 Rivets. 
 
 Tensile. 
 
 Shearing. 
 
 Crushing. 
 
 Treble-riveted 
 
 
 31-2 
 
 28.8 
 30.9 
 30.4 
 31.2 
 
 31.6 
 
 32.7 
 28.3 
 28.2 
 31.6 
 29.1 
 28.6 
 27.7 
 27.1 
 31-7 
 29.1 
 34 
 27.5 
 27.4 
 27-3 
 27-4 
 30-7 
 32.2 
 28.8 
 28.8 
 27.6 
 28.0 
 30.0 
 26.7 
 
 23.34 
 22.47 
 36.11 
 35.00 
 35.38 
 32.75 
 31.27 
 29.88 
 30.14 
 35.88 
 33.83 
 30.47 
 29.44 
 25.34 
 32.30 
 29.89 
 25.89 
 26.07 
 
 J9-74 
 21.41 
 
 25-9 1 
 34.76 
 34-95 
 31.92 
 31.84 
 26.42 
 29.22 
 
 25-55 
 28.16 
 
 I2.O 
 12.2 
 
 16.1 
 
 15.6 
 
 18.2 
 
 17.6 
 16.5 
 15.7 
 
 15.8 
 
 23-3 
 
 22.O 
 22.2 
 
 21-5 
 18.4 
 
 25.1 
 25.5 
 23.9 
 24.1 
 19.4 
 20.6 
 19.2 
 19.1 
 19.2 
 17.4 
 17.4 
 16.7 
 15.2 
 16.5 
 15.9 
 
 17-58 
 16.85 
 24.32 
 23-57 
 26.79 
 25.76 
 
 24.59 
 23.31 
 
 23-51 
 32.96 
 
 22.IO 
 21.24 
 15.83 
 18.91 
 3&.OI 
 25.IO 
 
 23.14 
 20.54 
 19.63 
 21.25 
 
 I9.5I 
 27.03 
 27.12 
 
 23-93 
 23.88 
 2O.80 
 2O.22 
 2O. I O 
 20.12 
 
 75 
 78 
 H7 
 "5 
 "3 
 104 
 96 
 106 
 107 
 
 H4 
 
 116 
 107 
 106 
 94 
 
 52 
 54 
 77 
 76 
 
 79 
 73 
 67 
 74 
 75 
 83 
 77 
 72 
 69 
 70 
 79 
 73 
 62 
 
 67 
 54 
 59 
 71 
 79 
 76 
 
 77 
 77 
 67 
 70 
 60 
 7i 
 
 1 in. 
 
 ,i t < 
 
 i 6 
 
 i .'.' 
 
 1 ;; 
 
 * " 
 i ;; 
 
 j_ 
 
 4 
 
 1 " 
 
 i 
 
 ! " 
 I ;; 
 
 i " 
 \ ;; 
 
 M 
 
 i ;; 
 
 " 
 
 * 
 
 i " 
 
 a. < 
 
 4 
 
 1 " 
 
 Iron. 
 Steel. 
 
 II 
 
 Steel. 
 
 VI 
 
 (1 
 
 Iron, 
 it 
 
 u 
 11 
 
 M 
 
 
 
 1 t 
 
 Drilled 
 
 <i 
 
 (i 
 
 ft 
 
 < i 
 
 K 
 
 ii 
 
 
 
 i 
 
 4< 
 
 , 
 
 M 
 
 i 
 
 Treble-chain 
 
 < 
 
 (1 
 
 i 
 
 M 
 
 i 
 
 1 1 
 
 i 
 
 Quadruple-zigzag (<:). 
 
 Punched and reamed. 
 Drilled 
 
 
 
 ii 
 
 u 
 
 (i 
 
 
 
 Treble-zigzag (a) 
 Treble-zigzag (a).. . . 
 ?uadruple-zigzag (). 
 
 Drilled and reamed.. 
 
 Punched and reamed. 
 K i 
 
 Drilled 
 
 
 II 
 
 
 (4 
 
 
 I 
 
 
 I 
 
 
 I 
 
 
 I 
 
 
 1 
 
 
 
LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 195 
 
 x 
 
 tn 
 
 x 
 
 l-t 
 
 O 
 
 w 
 
 w 
 
 in 
 
 2 
 O 
 
 O 
 
 B. 
 
 O 
 B! 
 
 - 
 
 -dcus qlu 
 
 juiod spins 
 
 s.iu jo xsuoiEia 
 
 B -IAUJO 
 
 " 
 
 SO I 
 
 
 SAU jo notuna 
 
 TOAUjoqDlM S 
 
 % uci|4 .iv^jiifi lou speajj 
 
 conannp ^C urqi sss| 105; 
 
 -tip jnoj 
 
 joa 93jtji ireqi 
 
 JOJ.J 
 
 I 
 
 1 
 
 s 
 
 ; at st 
 
 MMHM 
 
 222S 
 
 3! at x 
 
 nnncoM 
 
 22222 
 
 , 5 
 
 S3U111 -t UEql 3JOU1 J3A3U puE ' J3 J3OTBIp 3q) SStnll 
 
 J! 
 I 
 
 SAU JO JSWUIElfl 
 
 ' Ja n d ww 
 
 I r * 
 
 j^ndrqnii . ;, 
 
 x>l|dt)|n]\- . 
 
 S )SAU JO m3l!3-[ 
 
 $ S 
 
 I 
 
 f 
 
 2 
 
 f : R : 
 
 n ci 
 
 jandjlinw 
 
196 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 TABLE XXVILz. 
 FRENCH PRACTICE IN SINGLE-RIVETED JOINTS. 
 
 
 From D. K. Clark's ' Manual of Rules.* 
 
 1 hickness of plate. 
 
 Diameter of rivet. 
 
 Pitch of rivets. 
 
 Lap. 
 
 li 
 
 Lj 
 ri 
 
 3 8 
 
 i-IH 
 
 'sg-s 
 
 li 
 
 v c 
 
 C';3 o 
 11- 
 
 If 
 
 
 
 fl u 
 
 c 
 
 c--* 13 
 
 i 
 
 J5 
 
 ^3 a! 
 
 E 
 5 o-g 
 
 C 
 
 rt O 
 
 I'iPI 
 
 i 
 
 o - s 
 
 
 9 
 
 
 
 ii 
 
 o - 
 
 SB'S 
 
 A 
 
 - 
 
 o-s - 
 
 3 
 
 .118 
 
 i- 
 
 8 
 
 315 
 
 A + 
 
 27 
 
 1. 06 
 
 irV 
 
 3 
 
 1.18 
 
 iA 
 
 4 
 
 .158 
 
 
 10 
 
 394 
 
 1 + 
 
 3 2 
 
 1.26 
 
 ii + 
 
 34 
 
 1-34 
 
 r ii 
 
 5 
 
 .197 
 
 rV ~4~ 
 
 12 
 
 .472 
 
 li "4" 
 
 37 
 
 1.46 
 
 T rV ~H 
 
 40 
 
 1.58 
 
 i-j^- -f- 
 
 6 
 
 .236 
 
 li 4" 
 
 14 
 
 551 
 
 A 
 
 43 
 
 1.69 
 
 i^-| -j- 
 
 44 
 
 i-73 
 
 i f 
 
 7 
 
 .276 
 
 ft 
 
 16 
 
 .630 
 
 1 + 
 
 48 
 
 1.89 
 
 i* + 
 
 5 
 
 1.97 
 
 I fi ~t~ 
 
 8 
 
 3'5 
 
 fV ~t~ 
 
 17 
 
 .669 
 
 H- 
 
 5 1 
 
 2.OI 
 
 2 + 
 
 54 
 
 2.13 
 
 2 i + 
 
 9 
 
 354 
 
 li + 
 
 19 
 
 .748 
 
 !- 
 
 54 
 
 2.13 
 
 a* + 
 
 56 
 
 2. 2O 
 
 2 T 3 T ~t~ 
 
 10 
 
 394 
 
 1 + 
 
 20 
 
 .787 
 
 1 + 
 
 56 
 
 2. 2O 
 
 2 t\ ~H 
 
 58 
 
 2.28 
 
 2 i ~t~ 
 
 ii 
 
 433 
 
 rV 
 
 21 
 
 .827 
 
 tt + 
 
 57 
 
 2.24 
 
 2 i 
 
 60 
 
 2.36 
 
 2 f 
 
 12 
 
 .472 
 
 If + 
 
 22 
 
 .866 
 
 \ 
 
 58 
 
 2.28 
 
 2 i ~H 
 
 60 
 
 2.36 
 
 2 -| 
 
 13 
 
 5 12 
 
 i + 
 
 2 3 
 
 . .906 
 
 I + 
 
 60 
 
 2.36 
 
 2 f 
 
 62 
 
 2.44 
 
 2 rV "4~ 
 
 14 
 
 55' 
 
 rV- 
 
 24 
 
 945 
 
 
 62 
 
 2.44 
 
 2 rV + 
 
 64 
 
 2.52 
 
 2i + 
 
 15 
 
 591 
 
 H- 
 
 25 
 
 .984 
 
 B+ 
 
 63 
 
 2.48 
 
 2 i 
 
 66 
 
 2.60 
 
 
 16 
 
 .630 
 
 1 + 
 
 26 
 
 1.024 
 
 
 65 
 
 2.56 
 
 2 A- 
 
 68 
 
 2.68 
 
 2H + 
 
LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 197 
 
 TABLE XXVIII. 
 PROPORTIONS OF DOUBLE-RIVETED LAP-JOINTS. 
 
 
 
 Wilson. 
 
 Fairbaim. 
 
 Shipbuilding. 
 
 
 
 
 
 Thickness 
 
 Diameter of 
 
 
 
 
 
 Lloyd's rule 
 
 Liverpool rule. 
 
 
 
 Pitch. 
 
 Lap. 
 
 Lap. Inches. 
 
 Length. 
 
 Lap. 
 
 Lap. 
 
 Butt-strap. 
 
 Inch. 
 
 
 Inches. 
 
 
 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 A 
 
 Jo 
 
 
 .. 
 
 2.084 2 
 
 
 i 
 
 t m 
 
 . . 
 
 -L > ~ 
 
 l& 
 
 > 2.<; - 
 
 O 
 
 -a 
 
 . . 
 
 
 ft 
 
 1 
 
 i 
 ft 
 
 i 
 
 "S; s ^ -j 
 
 ,_ 5 i = 
 
 4 
 
 H 
 1 
 
 diameter of ri 
 \ii from edge. 
 
 3-'34 g- 
 3-333 "a 
 
 .... CJ 
 
 3-75 ; 
 
 <^- . > 
 
 o iO 
 
 4-5 8 4 ^- 
 
 he thickness 
 sheet. 
 
 S 
 
 cn 
 * *C 
 
 3 
 3J 
 3f 
 
 4i 
 
 4 
 
 7i 
 8 
 8 
 
 10 
 10 
 10} 
 
 H wjl" 5 2 | 
 
 TS~ * t t? TJ 3l 
 
 5J ^ 
 
 ' ' ' ' T3 4> 
 
 .1 
 
 1! 
 
 cn 
 cn 
 
 si- 
 si 
 si 
 
 3 
 
 i 8-53.- 3i 
 
 rt 
 
 
 
 
 6 
 
 13 
 
 _A r- - } 
 16 r" ^ _, _ 3? 
 
 > " 
 
 .... ^ 
 
 ^ 
 
 O 
 
 6 I 
 
 3f 
 
 i .S -5 g j 3i 
 
 
 
 
 
 fc 
 
 6J 
 
 I4l 
 
198 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 16. Various Forms of Liap-joints. On account of the inequality of stress on 
 the transverse and longitudinal joints of cylindrical boilers it has been proposed to 
 arrange the joints diagonally (see figure 64). Taking the angle of the joints at 45, the 
 
 Fig. 64. 
 
 resultant of the transverse and longitudinal stresses 
 per inch run of the joint is found by calculation to be 
 nearly 80 per cent, of the greater stress, acting at an 
 angle of about 72 to the joint. 
 
 J. G. Wright gives the strength of two specimens 
 of single-riveted square lap-joints and two of diagonal 
 joints, at an angle of 45, which were tested by Kirkaldy. They were made of f-inch 
 Staffordshire plate, exactly .38" thick, 12" wide, with 2J" lap, punched holes, and 
 six |f" rivets in the square joint at 2" pitch. The diagonal joint was made with eight 
 rivets of the same size and pitch. The ultimate tensile strength of the solid plate was 
 19.69 tons per square inch with the fibre and 16.80 tons across. The sectional area 
 of the entire plate was (12 X .38) 4.56 square inches. The net sectional area of the 
 square joint was 2.71 square inches, and the shearing section of the rivets 3.11 square 
 inches, or 115 per cent, of the net section. 
 
 
 Ultimate tensile strength. 
 
 Net sectional area. 
 
 Net tensile strength per square 
 inch of sectional area 
 per cent. 
 
 Tons. 
 
 Per cent. 
 
 Square inches. 
 
 Per cent. 
 
 Entire plate 
 
 89.8 
 
 43- 
 58.0 
 
 IOO 
 4 8 
 6 4 
 
 4-s 6 
 2.71 
 
 3-98 
 
 IOO 
 
 59-4 
 8 7 .2 
 
 IOO 
 8l.2 
 
 73-4 
 
 
 Diagonal joint 
 
 
 It will be seen that the diagonal joint was one-third stronger than the square joint, 
 although per square inch of net section it opposed less resistance. 
 
 An ingenious method of increasing the strength of a single-riveted joint that of 
 Webb consists in having the rivets of oval section, and placed with their smaller dia- 
 meters in a straight line parallel with the direction of the seam, and the transverse dia- 
 meters parallel to each other and perpendicular to the direction of the seam, as in figure 
 65. The strength of this joint is claimed to be about 83 per cent, of the solid sheet ; 
 the smaller diameter of the rivets should be so large that they do not act as wedges in 
 splitting the sheet. The alleged practical difficulties of manufacture disappear when 
 the question is considered, for it would be easy to punch oval holes or to drill them 
 with milling tools ; the rivets also could be made in oval dies as well as in circular ones. 
 
SBC. 16. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 199 
 
 Still another method of strengthening a joint while retaining single-riveting is that 
 of Beattie, who has patented a single-riveted joint over which is a covering-plate, also 
 single-riveted, as seen in figure 66. 
 
 Fig. 66. 
 
 Fig. 65. 
 
 o 
 
 o 
 
 o o o o 
 
 o 
 
 o 
 
 When it is difficult to set a sufficient number of single-shear rivets a forked arrange- 
 ment, like that in figure 67, may be employed, making the riveting double-shear and of 
 half the number. 
 
 A method of double-riveting which would greatly increase the strength of the joint, 
 
 Fig. 67. 
 
 Fig. 68. 
 
 ) 
 
 Fig. 69. 
 
 and even bring it to equality with that of the solid plate, has been proposed : it is to 
 have the edges of the plates to be connected rolled thicker than the rest of the sheet, 
 
200 STEAM BOILERS. CHAP. VIII. 
 
 as shown in figure 68, and Fairbairn has proposed further to add the thickness to the 
 faying surfaces and to plane shoulders on each, so that they shall lock into each other 
 illustrated by figure 69. By these means the strength lost in the holes is restored, 
 but the difficulty and expense of manufacture would probably prevent their general 
 adoption. 
 
 17. Butt-joints. The proportions of a single-welt butt-joint are the same as those 
 of a lap-joint, as it is in effect equal to two laps in juxtaposition ; the butt-strap is 
 made equal in thickness to the sheets connected. Although single-welt butt-joints are 
 not stronger than lap-joints, and are subject to the same distortion on account of the 
 oblique action of the stress on the rivets, they are much used for large furnace-tubes 
 subjected to an external pressure, where it is essential to preserve a perfectly cylindrical 
 form ; and for the transverse joints of cylindrical shells, which experience only one-half 
 the stress borne by the longitudinal joints, and where a tight joint is more easily ob- 
 tained with a single welt at places where the longitudinal and transverse joints meet ; 
 and in the flat heads of cylindrical boilers, where with thick plates smoother work 
 can be made with butt-joints than with lap-joints, and where stiffness is of greater im- 
 portance than tensile strength, since the direction of the principal strain is at right 
 angles to the plate. The joint should be made with the butt-strap on the outside, so 
 that it is accessible for subsequent calking, and that the action of the steam-pressure 
 may assist in preventing the opening of the joint. 
 
 In the double-welt butt-joint the shearing of the rivets must occur in two places, and 
 on this account their resistance is very nearly twice as great as in other joints. This 
 joint is free from the distortion on account of the oblique action of the stress on the 
 rivets to which the lap-joints and single-welt butt-joints are subjected, and for this rea- 
 son it should be used for thick plates when practicable. Wilson says : "Besides the 
 loss of strength due to the unequal distribution of the strain through the whole thick- 
 ness of the plates in a lap-joint, very thick plates are also liable to be much reduced in 
 strength through the body of the plate by injury done in the excessive amount of set- 
 ting they require where the transverse and longitudinal seams cross each other. For 
 this reason alone butt-joints should always be used, at least for the longitudinal seams 
 with plates over inch thick. The width of the strap for double-riveting should be at 
 least nine times the diameter of the rivet, and may with thick plates be made equal to 
 ten times the diameter, the distance from the centre of the holes to the edge of the 
 plates and straps in all cases being equal to the diameter of the rivet multiplied by f." 
 
 The butt-straps are made equal in thickness to at least one-half the thickness of the 
 plates. 
 
SEC. 17. LAYTNG-OFF, FLANGING, RIVETING, WELDING, ETC. 201 
 
 Applying the same notation as used heretofore the proportions of the double-welt 
 butt-joints are found from the following equations : 
 
 SiNGLE-KIVETED DOUBLE-WELT BUTT-JOINT. 
 
 To find the pitch of the rivets. 
 
 F=2nd?. 7854 S=n(p d)tT; 
 
 1.5708 tf 8 . 
 P = --jT + d; 
 
 when d = 1.05 1 (see section 13 of the present chapter) and S = V T; 
 
 p = 2.9 d. 
 
 DOUBLE-RIVETED DOUBLE- WELT Buir-jcmr. 
 To find the pitch of the rivets. 
 
 F = 4nd? .7854 S=n(p d)tT; 
 
 3.1416 d* S . 
 P = -~tT~ + d; 
 
 and when d = 1.05 1 and 8 = V T; 
 p = 4.4 d. 
 
 The following practice prevails at the Crewe Works (England), where Bessemer 
 steel is used for locomotive boilers : "The joint of the barrel is made along the top, 
 and is a single-riveted butt-joint, with inside and outside covering-strips. The barrel- 
 plate is Jf inch thick, the cover-strips $ inch thick by 5J inches wide ; rivets J inch 
 diameter, spaced with 2 inches pitch and placed with the centres about 1 inches from 
 the edge of the plate. This joint has been found to give 71.6 per cent, of the strength 
 of the solid plate. The rivets are of steel. A noticeable feature in the proportioning of 
 the joint is the distance of the rows of rivet-holes from the edges of the plate, this dis- 
 tance having been found necessary to prevent the distortion of the holes under strain. 
 On the other hand, in the cover-strips, where there is an excess of strength, the holes 
 come at an ordinary distance from the edge, so that there is no difficulty in calking 
 properly." (Engineering, October 15, 1879.) 
 
 In practice it is convenient to have all the holes in the same sheet of equal size, and 
 to use as small a variety of rivets as possible in the same boiler ; on this account the 
 diameters of rivets as found by the above formula are modified to suit these conditions. 
 "The greatest difficulty in making a well-proportioned joint with the same-sized rivets 
 occurs when butt-joints with double strips and lap-joints come together in the same 
 plate. In such a case we must either sacrifice the advantage of having the same-sized 
 
202 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 hole throughout the plate or have a badly-proportioned joint in one seam or the other. 
 On this account, when double-fished butt-joints are used in the same plate with lap- 
 joints, the former may be single and the latter double riveted, in which case the same 
 pitch and diameter of rivet might be judiciously employed, were it not for the difficulty 
 of keeping a tight joint in the butt arrangement, which necessitates the reduction of the 
 pitch unless the workmanship is very good." ( Wilson.) 
 
 TABLE XXIX. 
 
 
 
 WILSON'S TABLE OF PROPORTIONS OF DOUBLE-RIVETED BUTT-JOINTS WITH TWO COVERING-PLATES. 
 
 'o 
 
 "3 
 
 o 
 
 
 o 
 
 3 
 
 
 3 
 
 
 3 
 
 KM 
 
 o 
 
 3 
 
 
 1 
 
 V . 
 
 1 
 
 I 
 
 
 1 
 
 V 
 
 
 g 
 
 
 1 
 
 s . 
 
 is 
 
 
 
 SI 
 
 J a 
 
 j 
 
 C v 
 
 II 
 
 1 
 
 1 
 
 If 
 
 
 
 S S 
 Jj3 
 
 
 if 
 
 - S 
 
 j| 
 
 15 ci 
 
 .2'C 
 
 IS tn 
 
 B 
 
 'Ja o. 
 
 
 
 [J 
 
 
 8 
 
 2 o- 
 
 rt -j 
 
 X 
 
 ii 
 
 H 
 
 P 
 
 H 
 
 S 
 
 H 
 
 9 
 
 
 H " 
 
 S 
 
 H 
 
 p 
 
 H 
 
 S 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inches. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inches. 
 
 Inch. 
 
 Inch. 
 
 Inch. 
 
 Inches. 
 
 1 
 
 4 
 
 J 
 
 H 
 
 I 
 
 J 
 
 4 
 
 
 1 
 
 3 
 
 , 
 
 I 
 
 t 
 
 3l 
 
 TV 
 
 4 
 
 
 4 
 
 H 
 
 i 
 
 j 
 f 
 
 
 1 
 
 3i 
 
 H 
 
 I 
 
 \ 
 
 3! 
 
 i 
 
 i 
 
 ft 
 
 2j 
 
 
 j 
 
 i 
 
 
 TV 
 
 3i 
 
 
 4 
 
 A 
 
 4 
 
 rV 
 
 tt 
 
 A 
 
 4 
 
 ft 
 
 * 
 
 TV 
 
 3i 
 
 
 
 
 
 In Kirkaldy's experiments on the comparative strength of chain and zigzag riveting 
 in double-welt butt-joints, summed up in the ' Eeport of the Chief Engineer of the 
 Manchester Boiler- Insurance Company for 1877,' the following statement occurs, tabu- 
 lated herewith, showing that with chain-riveting greater strength is obtainable with 
 smaller pitch a great advantage in calking and making a tight joint. It will also be 
 remarked that in both cases thicker plates, larger rivets, greater pitch, and drilled holes 
 gave a smaller percentage of strength compared with the plate. 
 
 TABLE XXX. 
 
 Number of 
 tests. 
 
 Thickness of 
 plate. 
 
 Diameter of 
 
 rivets. 
 
 Pitch of 
 rivets. 
 
 Rivet-holes. 
 
 Ratio of 
 strength of 
 joint to plate. 
 
 System 
 
 of riveting. 
 
 
 Inch. 
 
 Inch. 
 
 Inches. 
 
 
 Per cent. 
 
 
 2 
 2 
 2 
 2 
 
 TV 
 TV 
 
 
 
 4 
 
 3 
 3 
 
 Punched .... 
 Punched .... 
 Drilled 
 
 67.2 
 66.9 
 66.2 
 
 63.3 
 
 Chain. 
 Zigzag. 
 Chain. 
 Zigzag. 
 
 Drilled 
 
 
SEC. 17. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 203 
 
 The following table of the comparative strength of punched and drilled rivet- work, 
 containing the result of Kirkaldy's experiments, is taken from the ' Proceedings of the 
 Mechanical Engineers for 1872 ' and forms part of a paper read by W. R. Browne : 
 
 TABLE XXXI. 
 
 Description of joint. 
 
 Riveting. Rivet-holes. 
 
 Proportions. 
 
 Ratio of 
 strength of 
 joint to that 
 of plate, 
 per cent. 
 
 Diameter of 
 rivets to 
 thickness of 
 plates. 
 
 Lap or cover to diameter 
 of rivets. 
 
 Pitch to 
 
 diameter of 
 rivets. 
 
 
 
 
 Lap. 
 
 
 
 Lap 
 
 c,. , ( Punched. . 
 Smgle. | Drilled ... 
 
 2 
 2 
 
 3 
 
 3 
 
 3 
 4 
 
 55 
 62 . 
 
 
 
 
 
 Chain. 
 
 Zigzag. 
 
 
 
 
 , . ( Punched. . 
 Double] Drilled ... 
 
 2 
 2 
 
 5i 
 
 5 
 
 6 
 Si 
 
 4* 
 
 4 
 
 69 
 75 
 
 
 
 
 
 Covering- strip. 
 
 
 
 Butt, i cover. . . 
 
 ~- , ( Punched. . 
 Single. -J Drilled ... 
 
 2 
 2 
 
 6 
 6 
 
 k 
 
 55 
 62 
 
 
 
 
 Chain. 
 
 Zigzag. 
 
 
 
 Butt, i cover. . . 
 
 T-. , , ( Punched. . 
 Double] Drilled _ 
 
 2 
 
 2 
 
 ii 
 10 
 
 12 
 11 
 
 4i 
 
 4 
 
 69 
 
 75 
 
 Butt, 2 covers. . 
 
 . , ( Punched.. 
 Smgle. j orined... 
 
 1} 
 I* 
 
 6 
 6 
 
 3i 
 3 
 
 57 
 67 
 
 
 
 
 Chain. 
 
 Zigzag. 
 
 
 
 Butt, 2 covers . . 
 
 , , ( Punched. . 
 Double -j Drilled ... 
 
 1^ 
 Ii 
 
 ii 
 
 10 
 
 3 
 
 12 
 
 5f 
 
 4f 
 
 72 
 79 
 
 The experiments recorded in the following table have been selected from the ' First 
 Report of the Committee of the Institute of Mechanical Engineers on the Form of 
 Riveted Joints.' The stresses corresponding to the actual mode of fracture are printed 
 in heavy type ; the other stresses, printed in ordinary type, are those which obtained at 
 the moment of fracture, but are lower than those at which the joint would have given 
 way by the respective modes of fracture. The joints marked d, e, and / in the table, 
 and their lines of fracture, are represented in figures 70, 71, and 72 respectively. The 
 steel plates had a nominal thickness of f inch. The butt-straps were T 9 F inch thick, and 
 the steel rivets had a diameter of 1.13 inches. The tensile strength of the rivets was 
 
204 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 28.9 tons per square inch, and the apparent shearing strength of the rivets in experi- 
 
 Fig. 70. 
 
 Fig. 71 
 
 > 
 
 r 
 
 )- 
 
 
 
 
 
 
 \( 
 
 
 -eH-*-e-f--e>- 
 
 <a 
 
 
 
 zjujj^r. 
 
 i./l _ 
 
 k- 
 
 
 in - 
 
 
 
 
 
 > 
 
 > 
 
 ment/was, therefore, 68.5 per cent, of their tenacity. In experiments d and e the 
 plates broke on the line marked in figures 70 and 71. 
 
 TABLE XXXII. 
 
 RESULTS OF EXPERIMENTS WITH SINGLE AND DOUBLE-RIVETED DOUBLE-WELT BUTT-JOINTS. 
 
 
 
 . 
 
 .E 
 
 Stress at moment of frac- 
 
 li 
 
 I 
 
 
 
 
 U 
 
 0) J= 
 
 ture in tons per sq. in. 
 
 i rt 
 
 
 
 
 
 .- 
 
 n e 
 
 
 *" S 
 
 *-' 
 
 
 
 
 
 
 "^'S 
 
 
 
 
 O w 
 
 ;| 
 
 
 Mode of riveting. 
 
 Holes. 
 
 I 
 O 
 
 i! 
 
 
 
 
 rt .; " 
 
 I si 
 
 s 
 
 Remarks. 
 
 
 
 
 
 "E, 
 
 a 
 
 S a 
 
 JU 
 
 ttt 
 
 3 
 
 iS 
 
 X 
 
 u 
 
 gj 
 
 
 
 
 a 
 
 
 8 
 
 J 
 
 i 
 
 rt w 
 
 giS'S 
 
 ji 
 
 
 
 
 M 
 
 H 
 
 H 
 
 M 
 
 u 
 
 
 H 
 
 
 Single-riveted 
 
 Punched 
 
 Iron plates 
 and rivets. 
 
 25-77 
 
 24.31 
 
 16.38 
 
 43-09 
 
 94 
 
 H 
 
 iMean of three experi- 
 ments by Fairbairn. 
 
 u 
 
 Drilled 
 
 
 22.25 
 
 24.24 
 
 11.48 
 
 33-83 
 
 109 
 
 63 
 
 Greig and Eyth. 
 
 ^ 
 
 H 
 
 Steel plates 
 
 36.22 
 
 36.62 
 
 18.75 
 
 52.04 
 
 ZOI 
 
 60 
 
 Henry Sharp. 
 
 Double-riveted 
 
 Punched 
 
 and rivets. 
 Iron plates 
 and rivets. 
 
 2 S-77 
 
 21.44 
 
 10.82 
 
 25.17 
 
 83 
 
 
 Mean of two experi- 
 ments by Fairbairn. 
 
 ** 
 
 Drilled 
 
 
 22.25 
 
 20.65 
 
 8.92 
 
 25.72 
 
 93 
 
 64 
 
 Greig and Eyth. 
 
 
 
 
 
 
 
 
 
 i 
 
 Plates i inch thick. 
 
 " 
 
 Punched 
 
 H 
 
 '9-35 
 
 17.52 
 
 6.90 
 
 13.00 
 
 91 
 
 j 
 
 Two experiments, R. 
 
 
 
 
 
 
 
 
 
 54 1 
 
 V. J. Knight. 
 
 H 
 
 Drilled . . . 
 
 Steel plates 
 and rivets. 
 
 36.22 
 
 42.93 
 
 13.16 
 
 37.20 
 
 "9 
 
 70 
 
 Henry Sharp. 
 
 
 
 U 
 
 Punched . 
 
 u 
 
 36.22 
 
 39.11 
 
 11.57 
 
 33.56 
 
 108 
 
 63 i 
 
 
 
 
 
 
 
 
 
 
 
 Mean of ten experi- 
 
 H 
 
 K 
 
 u 
 
 36.20 
 
 33.75 
 
 15.30 
 
 4 1 -'39 
 
 93 
 
 64 < 
 
 ments. Plates an- 
 
 
 
 
 
 
 
 
 
 
 nealed. D. Kirkaldy. 
 
 Double-riveted (zigzag) (d) . 
 
 Drilled and reamed. 
 
 H 
 
 28.10 
 
 20.O4 
 
 16.40 
 
 21.92 
 
 71 
 
 59 " 
 
 
 (). 
 
 Punched and ream'd 
 
 it 
 
 27.10 
 
 22.73 
 
 16.80 
 
 22.66 
 
 84 
 
 63 
 
 David Kirkaldy. Plates 
 % inch thick. 
 
 Double-riveted (chain) (f).. 
 
 Drilled and reamed. 
 
 H 
 
 28.20 
 
 27.23 
 
 16.80 
 
 27.10 
 
 
 73 . 
 
 
 18. Calking. Since the surfaces of boiler-plates are always more or less rough, the 
 riveted joints require calking to make them steam and water tight. Calking consists in 
 bringing the extreme edge of one boiler-plate so close to the solid part of the other that 
 there shall be no leakage between the two. Calking should always be done on both 
 
SBC. 18. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 205 
 
 Fig. 73. 
 
 sides, where it is possible to do so. If the riveting has not entirely closed the extreme 
 edge a heavy hammer has to be used to do this. Wedges or pieces 
 of hoop-iron should never be driven between the hips, nor should 
 sal-ammoniac or any other substance be used to make the joint tight 
 by rusting. If the edge of the plate has not already been planed it 
 should be chipped smooth to an angle of about 110. The tools or- 
 dinarily in use are represented in figure 73. 
 
 The bevel of the calking-tool should be about 20, the sharp 
 corner being first used for making a slight indentation in the lap- 
 edge this operation is called splitting the lap; then the tool is 
 turned and the whole width is used for driving in or upsetting the 
 
 Fig. 74 
 
 edge against the sheet. Care must be taken not to move the calking-tool its entire 
 breadth after each blow, else small places will be left uncalked ; one hard blow at each 
 place should be sufficient. Sexton says that the proper thickness for a calking-tool for 
 plates from f inch to $ inch thick is ^ inch ; for plates more than inch thick the tool 
 should be i inch. 
 
 Cannery's calking-tool is made convex, so that it shall not cut the metal. The com- 
 parative working of the two tools is shown 
 in the sketch (figure 74), which, however, 
 is somewhat exaggerated ; the old method, 
 shown to the right, is to chip or plane the 
 edge of the lap, then to drive up the tool, 
 indenting the lower sheet and tending to curve the lap upward as shown in dots ; if 
 afterward the lower sheet be bent, either purposely or by the action of unequal expan- 
 sion, grooving ensues at the indentation caused by the tool. With the concave tool 
 a depression is made in the edge of the lap, and the lower portion of the lap is driven 
 against the other sheet without injuring the latter ; the comparative extent of compres- 
 sion in the two methods is said to be shown by the -wedge of dark shading in the two 
 cases. 
 
 The butt is calked with the tool delineated in figure 75, which makes an indentation 
 as sketched at figure 76. 
 
 Boilers should not be calked under pressure, as the jarring would probably start 
 leaks in seams elsewhere. Excessive calking of lap-joints works mischief in several 
 ways : thin plates may be forced apart when a set and heavy hammer are used this is 
 shown in figure 57 ; when the edge of the calking-tool is very thin it sometimes acts as 
 a wedge, forcing the joint wide open. In contrast with the foregoing, figure 77 is given 
 
200 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 from Burgh's 'Practical Treatise,' which he calls "an illustration of the result of pro- 
 per drilling, fitting, riveting, and calking." 
 
 Fig. 75. Fig. 77. 
 
 Fig. 76. 
 
 19. Welding. Nasmyth says of welding that it consists in inducing upon malle- 
 able iron, by means of a very high heat, a certain degree of adhesion, so that any two 
 pieces of malleable iron, when heated to the requisite degree, will, if brought into close 
 contact, adhere or stick together with a greater or less tenacity, according to the 
 amount of force applied to urge them into close contact. . . . The chief cause of 
 defective welding arises from portions of the vitreous oxide of the iron being shut up 
 between the surfaces at the part presumed to have been welded ; and since, besides the 
 impossibility of ascertaining in the majority of cases, after the process of welding has 
 been gone through, whether or not this vitreous oxide has been thoroughly expelled 
 and the surfaces at the welding brought into perfect metallic union, no after-heating 
 or hammering can dislodge the vitreous oxide when once it has effected a lodgment, our 
 best and only true security is to form the surfaces of the iron, at the part where the 
 welding is desired to take place, so that when applied to each other when at the weld- 
 ing-heat their first contact with each other shall be in the centre of each. 
 
 Much attention has been paid of late to the effect of chemical composition on the 
 welding of iron. D. Adamson states in his paper read before the Iron and Steel In- 
 stitute, September, 1878 : "After many trials and many failures in attempting to weld 
 steel boiler-plates the writer found it necessary to ascertain in all cases the composition 
 of the metal before putting any labor upon it, and from a large experience it is now 
 considered desirable that the carbon should not exceed 0.125 per cent., while the sul- 
 phur and phosphorus should, if possible, be kept as low as 0.04 per cent., silicon being 
 admissible up to the extent of 0.1 per cent." 
 
 A. L. Holley, in a paper read before the American Institute of Mining Engineers 
 in February, 1878, discusses the results obtained by the United States Test-Board in 
 experiments upon fourteen brands of wrought-iron intended for chain-cables. He comes 
 
SBC. 20. LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 207 
 
 to the following conclusions, viz. : "Phosphorus up to the limit of one-quarter per 
 cent, had not a notable effect on welding." " Carbon notably affected welding. It 
 ran in connection with regularly decreasing welding power from 0.02 to 0.35 per cent." 
 "Carbon, in a greater degree than phosphorus, promotes fluidity; hence the iron is 
 burned at the ordinary welding temperature of low-carbon irons." " Slag should theo- 
 retically improve welding, like any flux, but its effects in these experiments could not 
 be definitely traced." " The experiments prove that the strength of the link, which is 
 chiefly dependent on welding power, as compared with the bar was more decreased by 
 overworking (in reducing the pile to the bar) than by any other cause, excepting the 
 high carbon in the steely iron L and the excessive copper, phosphorus, etc., in the 
 peculiar iron M." Regarding the strength of the welded joint of the latter iron he 
 says : "Its surfaces were pretty well united by welding, but the iron about the weld 
 was weakened, especially at a high heat. Of 59 ruptures of links made of this iron, 33 
 were through the weld and the iron was little distorted. Of 303 ruptures of links made 
 of other irons, but 36 were through the weld." 
 
 He proposes the following theory regarding welding: " It is certain that perfect 
 welds are made by means of perfect contact due to fusion, and that nearly perfect welds 
 are made by means of such contacts as may be got by partial fusion in a non-oxidizing 
 atmosphere or by mechanical fitting of the surfaces, whatever the composition of the 
 iron may be within all known limits. While high temperature is thus the first cause 
 of that mobility which promotes welding, it is also the cause, in an oxidizing atmos- 
 phere, of that 'burning' which injures both the weld and the iron. Hence welding 
 in an oxidizing atmosphere must be done at a heat which gives a compromise between 
 imperfect contact due to want of mobility on the one hand and imperfect contact due to 
 oxidation on the other hand. This heat varies with each different composition of irons. 
 It varies because these compositions change the fusing-points of irons, and hence their 
 points of excessive oxidation. Hence, while ingredients such as carbon, phosphorus, 
 copper, etc., positively do not prevent welding under fusion or in a non-oxidizing at- 
 mosphere, it is probable that they impair it in an oxidizing atmosphere, not directly but 
 only by changing the susceptibility of the iron to oxidation. 
 
 "The obvious conclusions are: 1st. That any wrought-iron, of whatever ordinary 
 composition, may be welded to itself in an oxidizing atmosphere at a certain tempera- 
 ture, which may differ very largely from that one which is vaguely known as ' a weld- 
 ing-heat.' 2d. That in a non-oxidizing atmosphere heterogeneous irons, however im- 
 pure, may be soundly welded at indefinitely high temperatures." 
 
 2O. Welding Boiler-plates. The welding of boiler-plates was first successfully 
 
208 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 Fig. 78. 
 
 tried by W. Bertram, at Woolwich Dockyard, in 1857. His method is represented in 
 
 figure 78. The edges of the plates are 
 scarfed and placed together ; two non- 
 oxidizing gas-flames are obtained by the 
 combustion of coal or coke in suitable 
 furnaces, and these are directed against 
 either side of the plates until they are 
 raised to a welding-heat, when they are 
 united by pressure or hammering. For 
 this purpose a number of stampers are 
 sometimes used viz., upright rams 
 raised by four- toothed cams and falling by their weight upon the work, which is placed 
 on an anvil. The work is said to be done three or four times as expeditiously as hand- 
 riveting. 
 
 The scarf -weld, either of the form shown in figure 78 or else prepared as drawn in 
 figure 79, is much stronger than the lap-welded joint, which is shown in figure 80, since 
 
 Fig. 81. 
 Fig. 79. 
 
 Fig. 80. 
 
 i / 
 
 the strain on the scarfed joint is direct, instead of tending to spring the joint ^as with 
 the lap- welded plates shown in figure 81. The lap-weld should be used, on this account, 
 only with very thin plates. The welded joints of thicker plates might be greatly in- 
 creased in strength by a covering- plate, as shown in figure 82. 
 
 The seams of the flues of high-pressure boilers are frequently welded in the manner 
 
 Fig. 83. Fig. 84. 
 
 Fig. 82. 
 
 illustrated in figure 83. The edges of the bent sheet are kept apart, a distance of about 
 & inch, by small blocks, a a, the sheet being held together by the bands b 5, secured by 
 bolts and nuts. A narrow strip, c c, is welded over the seam to the plates at each end, 
 in order to hold them more securely in position. The two edges and a rod of a wedge- 
 shaped or round section are brought to a welding-heat in an open fire ; the tube is 
 passed over the long horn of an anvil, the rod is inserted between the edges, and all 
 
SBC. 20. 
 
 LAYING-OFF, FLANGING, K1VETING, WELDING, ETC. 
 
 209 
 
 three parts are welded by hammering for a distance of a few inches. Then the nearest 
 band and block are moved along, the rod is cut, another heat is taken, and the process 
 is repeated. In small tubes where pressure by machinery can be applied the weld can 
 be made with a butt-joint. 
 
 The following practical points in reference to welding are selected from Sexton's 
 ' Boiler-makers' Pocket-book.' Pipes or cylinders up to four feet diameter may be 
 easily welded in the following manner : Bend the pipe or cylinder in the form of figure 
 84, the edges A and B not touching by about a quarter of an inch ; place it on a clear 
 fire, throw a little sand and scale on the edges, turned downward ; place a fire-brick on 
 the part through which the greatest heat is coming. The brick is a very slow conductor 
 of heat, and will greatly assist the iron in getting hot, as it does not absorb much heat 
 itself. The blast is started moderately till the iron becomes of a pale yellow, then put 
 on strongly till the iron is white-hot, when the work is brought out and placed on the 
 mandrel as quickly as possible. The first blow should fall gently on the edge A, and 
 B is hammered down on it ; and when the iron has cooled so as not to fly to pieces 
 under the blows these are repeated as hard and quickly as possible until the edge has 
 disappeared and a smooth surface is left. There is also an arrangement of a forge on 
 wheels made to run inside the cylinder with an india-rubber blast-pipe connected to it, 
 and when the lap of the plate is sufficiently heated Fig. 85. 
 
 the forge is withdrawn and an anvil wheeled in its 
 place. 
 
 In welding the longitudinal seams of a boiler 
 one inch thick, with four plates to the circle, the 
 plates are punched for the circumferential seams, 
 but not quite to the longitudinal edges ; these last 
 are then planed to an angle of 45 ; then, after 
 being rolled and set, the plates are fastened to- 
 gether temporarily with the planed edges inside. 
 These edges are heated in a clear fire at the same 
 time that a piece of square iron is heated in 
 a separate fire, and, when sufficiently heated, the 
 boiler, suspended from a suitable crane, is brought 
 out, placed on the block, the square iron, which 
 is rather hotter than the plates, laid in cornerwise, as at C in figure 85, and hammered 
 direct on the upper corner. Six or seven inches at a heat will be sufficient to weld. 
 The weight of the ring of plates must not rest on the block, or it will become flat. 
 
210 
 
 STEAM BOILERS. 
 
 CHAP. VIII. 
 
 The bar should have its sides equal in width to the planed edge of the plate. The 
 welding should be commenced in the middle and worked towards the ends, which are 
 kept securely bolted as long as is necessary. In putting the rings together the welded 
 seams should break joint. In furnace-tubes the weld should be below the grate. 
 
 When it is desirable to have the front of a boiler in one piece but a plate of suffi- 
 cient size cannot be procured, two plates may be welded together, either by laying in a 
 square bar, as in figure 86, or by thickening both edges, splitting one and tapering the 
 
 Fig. 86. 
 
 Fig. 87. 
 
 Fig. 88. 
 
 f~ 
 
 other, as sketched in figure 87, then forcing one into the other, as shown in figure 88 ; 
 this form, however, is not favorable for the escape of the scale, and the difference in 
 thickness of the edges in the fire is not favorable for the iron being properly and uni- 
 formly heated. It is more applicable for welding bars and plates of very short length. 
 The plates are secured by several short pieces of angle-iron bolted to them, and by as 
 many turn-buckles and chains as the length of the work requires (see figure 89), then 
 placed in the fire so as to heat the edges in question. When the plate is hot the ex- 
 pansion of the metal acting against the strain of the screws will weld the plates even 
 before they are hammered on the block, which, of course, must be done when suffi- 
 ciently heated. 
 
 Fig. 90. Fig. 91. 
 
 Fig. 89. 
 
 " In welding angle-iron rings for flanging, for strengthening flues, etc., the ends are 
 upset and scarfed or bevelled on the side required ; then, a short heat being taken, about 
 six inches at each end are bent ; next, as long a heat as possible being got, the ring is 
 bent around a shaper to the required curve. The ends of an external ring shoiild have 
 the lump inside before welding, as in figure 90, since this ring must be hammered on 
 the inside during the welding process ; while in the internal ring, figure 91, the lump is 
 left on the outside, that being the side on which the ring is most hammered. 
 
 "In making up a fire for welding, as the article in question is not to be made hot all 
 
SEC. 21. 
 
 LAYING-OFF, FLAGGING, RIVETING, WELDING, ETC. 
 
 211 
 
 over, but only at the part to be welded, the size of the fire should be limited to that 
 portion. For complicated shapes the small coal may be wet and built up around the 
 object or a rude pattern of it, and, the latter being withdrawn, the hole filled with clean 
 hot coke, together with small fresh coke ; next, the object being put in position, some 
 sand is sprinkled on the edges to be welded, which are then covered with a fire-brick, 
 the blast is started gently, and wet coal built up around what soon becomes a hollow 
 fire ; this should be disturbed as little as possible, and only to watch its progress. The 
 blast is presently increased, and when the sand is melted, and the metal white and be- 
 ginning to emit brilliant white sparks, the work may be brought out, and no time must 
 then be lost in the welding, all tools being in place and preparations made beforehand." 
 
 One of the advocates of welding says : " Among the difficulties of welding plates are 
 that they heat unequally, buckle, blister, and frequently become exposed while hot to 
 the air for a space of time sufficient to provide a coat of oxide fatal to the future joint. 
 But by the adoption of this principle we could reduce the thickness of boiler-plates by 
 one-half, rendering them lighter and more easily worked and handled in every way. 
 The expense incurred should not be higher than that of the riveted joint ; the extra 
 trouble of fitting and scarfing would not equal that of punching, and the fuel consumed 
 in rivet making and heating would nearly suffice for welding. There are certain situa- 
 tions in a boiler where this process would be inapplicable and where rivets would have 
 to be used. In early experiments a seam 12 feet long in f-inch plates was welded in 
 one hour and twenty minutes with mouth-pieces or nozzles only 6 inches long." 
 
 21. Strength of Welded Plates. Wilson gives the result of some experiments 
 recorded by Kirtley on the tensile strength of strips cut across the weld and taken from 
 several boilers with welded longitudinal seams ; the strips were 7 inches long and 
 -fa inch thick : 
 
 Width of strip. 
 
 Number of 
 strips tested. 
 
 Broke in 
 weld. 
 
 Broke in 
 
 solid. 
 
 Breaking strength in tons per square inch. 
 
 Inch. 
 
 Least. 
 
 Greatest. 
 
 Mean. 
 
 I 
 
 I.I 
 
 4 
 
 IS 
 
 4 
 4 
 
 8 
 2 
 I 
 
 7 
 
 2 
 
 3 
 
 16.5 
 19.6 
 
 18.1 
 
 23.8 
 22.2 
 
 2 3-5 
 
 20.2 
 21.0 
 21.7 
 
 Total 
 
 23 
 
 II 
 
 12 
 
 16.5 
 
 2 3 .8 
 
 2O.6 
 
 1 1 strips of the same plates unwek 
 
 led 
 
 20.7 
 
 25-8 
 
 23.6 
 
 
 After giving the foregoing table he proceeds: "It appears from these results that 
 
212 STEAM BOILERS. CHAP. VIH. 
 
 half of the test-pieces broke in the solid, and not at the weld. The average loss of 
 strength of the 23 welded plates was only 12.7 per cent, compared with the strength of 
 the 11 unwelded plates ; the worst pieces, showing as defective a weld as would occur in 
 practice, had 70 per cent, of the average strength of the unwelded plates. The weld is 
 best made when the edges of the plates are upset at a red-heat, by hammering or pres- 
 sure, to nearly double their thickness, and bevelled to an angle of about 45. The 
 edges can then be heated simultaneously, and the weld made by hammering down the 
 joint to the original thickness of the plate. In some cases it has been found that the 
 plates are rapidly pitted at the weld." This is doubtless owing to the outer protecting 
 film or skin being removed by the fire and working. 
 
 " Some time ago Mr. Gillott, of Farnley, who has had great experience in smithing 
 boiler-plates, had reason to suspect that in welding boiler-tubes in successive heats and 
 in short lengths after the ordinary manner there is a straining action upon the length 
 already welded, owing to the expansion under heat of the adjacent length being 
 welded, and we are indebted to Mr. Gillott for the following results of some experi- 
 ments he made. In order to ascertain the correctness of his views he took a welded 
 and flanged tube 2 feet 8 inches in diameter, 3 feet long, of T \-inch best Yorkshire plate, 
 welded by hand-hammers, ahd in every respect a fair average job. The plate was 
 prepared for welding with a "bent scarf," so that the finished work was generally some- 
 what thicker, or at any rate not less, than the solid plate. The welding was commenced 
 at the middle of the plate and worked towards each end. The tube was put in a lathe, 
 and a length of about 5 inches was cut off one end. Six rings were then cut off about 2 
 inches wide and numbered consecutively 1 to 6, counting from the edge where the 
 5-inch length was cut off, so that the piece marked 6 would be about the original middle 
 of the plate. Strips 2 feet 6 inches long, having the weld approximately in the centre, 
 were cut from each ring, and similar strips of the solid plate also. The strips were then 
 all heated and carefully straightened." . . . " The strips were then tested by being 
 pulled asunder, with the results given in the annexed table." . . . "Not only the 
 strength but the ductility of the weld diminished progressively from the end to the 
 middle portion of the plate, or from the part last welded to that which was first welded." 
 All strips broke at the weld. 
 
 "In a furnace-tube, where the pressure tends to assist in keeping the weld together, 
 the limited tensile strength due to the unsoundness of the weld may be outweighed by 
 other advantages, as for flanging, etc., which the welding gives ; but for the longitudinal 
 seams of boiler-shells the uncertainty of making a sound and strong weld, when the 
 job is done in lengths, renders it difficult to conclude otherwise than that such welding 
 
SEC. 21. 
 
 LAYING-OFF, FLANGING, RIVETING, WELDING, ETC. 
 
 213 
 
 is not so good as riveting, notwithstanding the liability of the latter to leak. This con- 
 clusion does not apply to joints where the weld is completed at one heat by pressure 
 either with rolls or otherwise." (Engineering, January 31, 1879.) 
 
 Mark. 
 
 * 
 
 w. 
 
 w, 
 
 w. 
 
 w. 
 
 w. 
 
 Average result for 
 strips of 
 solid plate. 
 
 Breaking weight, in tons per 
 sq. in of original area. 
 
 2*2.4. 
 *-?* 
 
 IQ J.8 
 
 18 7? 
 
 17.67 
 
 16 3 
 
 
 
 Elongation per cent, in 18 
 inches. 
 
 o~_>* u ei 
 g^"2-'S. 
 
 9-375 
 
 6-597 
 
 6.25 
 
 4.166 
 
 3-472 
 
 2 3-34 
 17.71 
 
 OF THE 
 
 UNIVERSITY 
 
 Of 
 
CHAPTER IX. 
 
 SHELL, FURNACES, AND BACK-CONNECTIONS. 
 
 1. Various Forms of Shells. The forms commonly given to the shell of marine 
 boilers have been described and illustrated in section 3, chapter vii. These forms are 
 produced by combining flat and cylindrical plates in various ways. Figure 92 illus- 
 trates one form of the " connected-arc Fig. 92i 
 marine boiler" of C. E. Emery, who 
 states that " the object of the invention 
 is to construct steam boilers of great 
 strength, with the minimum amount of 
 small stays or braces, and in such form 
 as to occupy less room for a given power 
 than ordinary cylindrical boilers ; also 
 to obtain, when desirable, a large boiler 
 of moderate height." 
 
 When such a boiler is braced according to the rules given in section 7, chapter vi., 
 the plates forming the several arcs experience only a tensile strain, like the circular shell 
 of a cylindrical boiler, while the flat surfaces of boilers are subjected to bending strains. 
 
 The thickness of the plates forming the cylindrical portions of the shell is found by 
 formula [III.], section 2, chapter vi., by making the value of 7c equal to the ultimate 
 strength of the joints divided by a factor of safety ; the strengths of various joints rela- 
 tively to the strength of the entire plate, as determined by experiment, are given in 
 sections 15 and 17, chapter viii. 
 
 The strength of the flat surfaces of the shell of boilers depends almost entirely on 
 the bracing and staying, the plates being made thick enough to give them proper stiff- 
 ness. Plates less than one-quarter inch thick cannot be calked efficiently, and are 
 therefore not used for boiler-shells. Tight joints cannot well be made by riveting 
 together plates varying greatly in thickness ; on this account the thickness of the 
 flat plates of the shell does not vary generally more than one-eighth inch from that of 
 the cylindrical plates. Various methods of, and rules for, staying the flat surfaces of 
 boilers will be found in chapter x. 
 
 In places where holes are cut in the shell for handholes, manholes, etc., stiff ening- 
 
 214 
 
SBC. 2. SHELL, FURNACES, AND BACK-CONNECTIONS. 215 
 
 plates or angle-irons are riveted around the opening. In cylindrical shells, depending 
 on their form for strength, it is important that these compensating-rings do not only 
 stiffen the weakened parts, but provide at any section at least the same amount of 
 metal as has been lost in making the opening. In the boiler illustrated on Plate XII. 
 the butt-straps serve also as strengthening-plates around the manholes and where tie- 
 rods pass through the front-head. 
 
 It is recommended to place the larger axis of manholes in a transverse direction on 
 circular shells, so that the least metal is cut away in the direction in which the stress on 
 the shell is greatest. 
 
 The precautions to be taken when steam-domes are placed on cylindrical shells will 
 be discussed in section 2, chapter xiii. 
 
 For the shell of boilers an inferior sort of iron is often used, known as " shell-iron," 
 which does not flange well ; but for the boilers of United States naval vessels it is 
 always stipulated that all parts shall be constructed of the best charcoal flange-iron. 
 
 2. Rectangular Shells. The form of boilers generally used in vessels of the 
 United States Navy, in cases where the steam-pressure does not exceed 45 Ibs. above 
 the atmosphere, is illustrated on Plates III., VI., VII., and XVII. The top, bottom, 
 sides, front, and back of the shell are flat ; but the sheets joining the top to the sides, 
 front, and back, and the sides and back to the bottom, are bent to as large a radius as 
 the internal arrangement of the boiler admits. By giving to the lower part of the boiler 
 the form shown on Plate XVII. there is not only a useless space in the boiler omitted, 
 but a great additional advantage is often gained in narrow vessels, since, with the fire- 
 room running in a fore-and-aft direction between the boilers, these can be placed much 
 farther outboard. Rectangular boilers are often made without water-bottoms (see 
 Plates VI., VII.) ; this form will be described in section 5 of the present chapter. 
 
 The thickness of the plates of rectangular boiler-shells varies ordinarily from ^ to 
 -fa inch, according to their location, the steam-pressure, and the arrangement of the 
 stays and braces. The plates forming the lower portion of the shell exposed to the cor- 
 rosive action of the bilge-water are made tt or i i ncn thicker than the rest. Lap-joints 
 are generally employed throughout the shell of rectangular boilers, and these should be 
 double-riveted. The laps are sometimes slightly bent, one inward and the other out- 
 ward, so as to keep the sheets forming a flat surface in the same plane ; by giving in 
 this manner to the lap-joints the form which they tend to assume under pressure the 
 joints are less severely strained (see figure 46). Care must be taken that the laps of 
 contiguous sheets break joint, and that they are accessible for calking. In the horizon- 
 tal joints of the sides, front, and back the lap of the upper end of each sheet is placed 
 
216 STEAM BOILERS. CHAP. IX. 
 
 on the outside of the boiler ; this prevents the loose scale from lodging on the laps in- 
 side the boiler. The square corners connecting the sides, front, and back of the boiler 
 are formed by turning flanges on the sheets. The practice of forming these corners by 
 angle-irons placed inside the boiler, to which the plates are riveted, permits the use of 
 an inferior iron for the shell, but is not to be recommended. 
 
 Plates VI., VII., XVII. illustrate fully the construction of the shell of rectangular 
 boilers, showing the size and shape of the plates and the manner of forming the flanges 
 and joints. An English practice of welding the plates forming the boiler-front has been 
 described and illustrated in chapter viii. 
 
 3. Cylindrical Shells. Cylindrical shells are generally built up by joining to- 
 gether several rings or belts. The plates forming these belts are so arranged that the 
 fibre of the iron runs in a circumferential direction. When plates of moderate thick- 
 ness are used the longitudinal seams are often formed by double-riveted lap-joints ; it 
 is, however, better to use double-riveted butt-joints with an inner and outer covering- 
 plate. Sometimes these rings are formed by welding the plates together. In connect- 
 ing these several belts care must be taken that the longitudinal seams of adjoining belts 
 are placed as far apart as possible. 
 
 The transverse joints connecting these belts are either lap or butt joints. When the 
 belts are connected by lap-joints it may be advantageous to place them telescopically, 
 as shown in figure 18, in short cylindrical boilers with few widths, since that arrange- 
 ment tends to drain the mud and dampness toward the large end, where cocks, hand- 
 holes, or other provisions may be made for cleaning. When the boilers are long the 
 sheets are alternately outside and inside, or raised and sunken. When a single butt- 
 strap is used for the transverse joints it must be placed on the outside of the shell, in 
 order to make it accessible for calking ; the inner longitudinal butt-straps must lap 
 over on the adjoining sheets, so that the ends of these straps may be calked properly 
 (see Plate XIII.) Single butt-straps give ample strength for the transverse joints of 
 shells ; sometimes, however, a thin inner strap is added, in order to give facilities for calk- 
 ing the joint more thoroughly from the inside. It is well to put a rivet through the seam 
 at places where the longitudinal and transverse joints meet, in order to close them effec- 
 tually against any leakage. The ends of the transverse butt-straps are made to lap over 
 one another, the inner end being properly scarfed. The riveting of these transverse 
 straps to the belts is commenced at their middle, so that any slack may be worked into 
 the seam. Where the longitudinal straps are crossed by the transverse straps their 
 ends are scarfed, so that they all lie close together and against the shell, and the trans- 
 verse strap requires less setting. 
 
SBC . 3. SHELL, FURNACES, AND BACK-CONNECTIONS. 217 
 
 The front and back heads of large boilers are flat, and are flanged in order to connect 
 them to the shell. In small boilers these flanges are often placed outside the shell 
 when the space inside the boiler is too contracted to be accessible for riveting ; but in 
 large boilers these flanges are turned inside the boiler, so that the joint can be calked 
 inside and outside. Angle-irons are sometimes used to make this connection. The 
 heads of large boilers are necessarily composed of several plates, connected either by 
 lap-joints or by butt-joints. When lap-joints are used the laps of the same plate are 
 either placed one inside and the other outside, or the plates are placed in alternate in- 
 side and outside courses. When the plates composing the heads are thick it is best to 
 unite them by means of a double-riveted single butt-strap placed on the outside. The 
 ends of the butt-strap are bent with the flange of the head and thinned, the cylindrical 
 shell being set out, where it covers the strap, in order to make a close-fitting, tight 
 joint. 
 
 'Lloyd's Register of British and Foreign Shipping' prescribes the following 
 "Rules for Determining the Working Pressure in New Boilers": 
 
 Cylindrical Shells. The strength of circular shells to be calculated from the actual 
 strength of the longitudinal joint by the following formula : 
 
 Cx Tx B 
 
 p: - = working pressure, 
 
 where C = constant as per following table ; 
 T= thickness of plates in inches ; 
 D mean diameter of shell in inches ; 
 
 B = percentage of strength of joint found as follows the least percentage 
 to be taken : 
 
 For plate at joint B = ^^- X 100 ; 
 
 For rivets at joint B = - ^ X 100 with punched holes ; 
 
 B = X ~ X 90 with drilled holes 
 p X T 
 
 (in case of rivets being in double-shear 1.75 a is to be used instead of a) ; 
 where p = pitch of rivets ; 
 
 d diameter of rivets ; 
 a = sectional area of rivets ; 
 n = number of rows of rivets. 
 NOTE. In steel boilers it is required that the strength of the rivets used to resist 
 
21S 
 
 STEAM BOILERS. 
 
 CHAP. IX. 
 
 shearing should be shown to be at least 26 tons per square inch. If it is less than 26 
 tons per square inch the rivet-area should be proportionately increased. 
 
 TABLE OF CONSTANTS. 
 IRON BOILERS. 
 
 Description of longitudinal joint. 
 
 For plates '/, inch 
 thick and 
 under. 
 
 For plates % inch 
 thick and above 
 1/2 inch. 
 
 For plates above % 
 inch thick. 
 
 
 155 
 170 
 IJO 
 I 80 
 
 165 
 1 80 
 1 80 
 190 
 
 170 
 190 
 190 
 2OO 
 
 " drilled holes 
 
 Double butt-strap joint, punched holes.. . . 
 ' " drilled holes 
 
 
 STEEL BOILERS. 
 
 Description of longitudinal joint. 
 
 For plates % 
 inch thick 
 and under. 
 
 For plates 9-16 
 inch thick 
 and above % 
 inch. 
 
 For plates % 
 inch thick 
 and above 9-16 
 inch. 
 
 For plates 
 above % inch 
 thick. 
 
 
 2OO 
 215 
 
 215 
 230 
 
 230 
 
 250 
 
 240 
 260 
 
 Double butt-strap joints 
 
 
 NOTE. The inside butt-strap to be at least three-quarters the thickness of the plate. 
 
 For the shell-plates of superheaters or steam-chests exposed to the direct action of 
 the flame the constants should be two-thirds of those given in the above tables. 
 
 Proper deductions are to be made for openings in shell. 
 
 All manholes in circular shells to be stiffened with compensating-rings. 
 
 The shell-plates under domes in boilers so fitted to be stayed from the top of the 
 dome or otherwise stiffened. 
 
 The ' Surveyors of the Board of Trade ' (England) are guided in the inspection of 
 boilers by the following rules : 
 
 " When cylindrical boilers are made of the best material, with all the rivet-holes 
 drilled in place and all the seams fitted with double butt-straps, each of at least five- 
 eighths the thickness of the plates they cover, and all the seams at least double-riveted 
 with rivets having an allowance of not more than 50 per cent, over the single-shear, and 
 provided that the boilers have been open to inspection during the whole period of con- 
 struction, then six may be used as the factor of safety ; but the boilers must be tested by 
 hydraulic pressure to twice the working pressure in the presence and to the satisfaction 
 of the Board's surveyors. But when the above conditions are not complied with the 
 
 
 
SEC. 3. 
 
 SHELL, FURNACES, AND BACK-CONNECTIONS. 
 
 219 
 
 additions in the following scale must be added to the factor six, according to the cir- 
 cumstances of each case : 
 
 A 
 B 
 C 
 D 
 
 E* 
 F 
 
 G 
 
 H 
 I 
 
 J* 
 K 
 
 L 
 M 
 N 
 O 
 P 
 
 Q 
 
 R 
 
 S 
 T 
 U 
 
 W* 
 X* 
 
 3 
 3 
 
 5 
 
 75 
 .1 
 
 i5 
 '5 
 
 .2 
 
 .2 
 
 .2 
 
 .1 
 
 3 
 
 -15 
 .1 
 
 .1 
 
 .2 
 .1 
 
 .1 
 .2 
 
 2 5 
 
 4 
 4 
 
 1.65 
 
 To be added when all the holes are fair and gpod in the longitudinal seams, but 
 
 drilled out of place after bending. 
 To be added when all the holes are fair and good in the longitudinal seams, but 
 
 drilled out of place before bending. 
 To be added when all the holes are fair and good in the longitudinal seams, but 
 
 punched after bending, instead of drilled. 
 To be added when all the holes are fair and good in the longitudinal seams, but 
 
 punched before bending. 
 
 To be added when all the holes are not fair and good in the longitudinal seams. 
 To be added if the holes are all fair and good in the circumferential seams, but 
 
 drilled out of place after bending. 
 To be added if the holes are fair and good in the circumferential seams, but drilled 
 
 before bending. 
 To be added if the holes are fair and good in the circumferential seams, but punched 
 
 after bending. 
 To be added if the holes are fair and good in the circumferential seams, but punched 
 
 before bending. 
 
 To be added if the holes are not fair and good in the circumferential seams. 
 To be added if double butt-straps are not fitted to the longitudinal seams, and the 
 
 said seams are lap and double riveted. 
 To be added if double butt- straps are not fitted to the longitudinal seams, and the 
 
 said seams are lap and treble riveted. 
 To be added if only single butt-straps are fitted to the longitudinal seams, and the 
 
 said seams are double-riveted. 
 To be added if only single butt-straps are fitted to the longitudinal seams, and the 
 
 said seams are treble-riveted. 
 
 To be added when any description of joint in the longitudinal seams is single- 
 riveted. 
 To be added if the circumferential seams are fitted with single butt-straps and are 
 
 double-riveted. 
 To be added if the circumferential seams are fitted with single butt-straps and are 
 
 single-riveted. 
 To be added if the circumferential seams are fitted with double butt-straps and are 
 
 single-riveted. 
 
 To be added if the circumferential seams are lap-joints and are double-riveted. 
 To be added if the circumferential seams are lap-joints and are single-riveted. 
 To be added when the circumferential seams are lap and the streaks or plates are 
 
 not entirely under or over. 
 
 To be added when the boiler is of such a length as to fire from both ends, or is of un- 
 usual length, such as flue-boilers ; and the circumferential seams are fitted as de- 
 scribed opposite P, R, and S ; but, of course, when the circumferential seams are 
 
 as described opposite Q and T, V-3 will become .4. 
 To be added if the seams are not properly crossed. 
 To be added when the iron is in any way doubtful and the surveyor is not satisfied 
 
 that it is of the best quality. 
 To be added if the boiler is not open to inspection during the whole period of its 
 
 construction. 
 
220 STEAM BOILERS. CHAP. IX. 
 
 "Where marked * the allowance may be increased still further if the workmanship 
 or material is very doubtful or very unsatisfactory 
 
 " The strength of the joints is found by the following method : 
 
 (Pitch diameter of rivets) x 100 _ j Percentage of strength of plate at joint 
 Pitch ' { as compared with the solid plate. 
 
 (Area of rivets X No. of rows of rivets) x 100 _ ( Percentage of strength of rivets as 
 Pitch x thickness of plate : ( compared with the solid plate.* 
 
 "Then take iron as equal to 23 tons, and use the smallest of the two percentages as 
 the strength of the joint, and adopt the factor of safety as found from the preceding 
 scale : 
 
 (51520 X percentage of\ /twice the thickness of\ 
 strength of joint. I I the plate in inches. J Pressure to be allowed per 
 
 =F n =r, -Vrr r~ri ~e rz~ = \ square inch on the safe- 
 
 Inside diameter of the boiler in inches x factor of safety | ty-valves. 
 
 "Plates that are drilled in place must be taken apart and the burr taken off, and 
 the holes slightly countersunk from the outside. 
 
 "Butt-straps must be cut from plates and not from bars, and must be of as good a 
 quality as the shell-plates, and for the longitudinal seams must be cut across the fibre. 
 The rivet-holes may be punched or drilled when the plates are punched or drilled out 
 of place, but when drilled in place must be taken apart and the burr taken off, and 
 slightly countersunk from the outside. 
 
 " When single butt-straps are used, and the rivet-holes in them punched, they must 
 be one-eighth thicker than the plates they cover. 
 
 "The diameter of the rivets must not be less than the thickness of the plates of which 
 the shell is made ; but it will be found when the plates are thin, or when lap-joints or 
 single butt-straps are adopted, that the diameter of the rivets should be in excess of the 
 thickness of the plates. Dished ends that are not truly hemispherical must be stayed ; 
 if they are not theoretically equal in strength to the pressure needed they must be 
 stayed as flat surfaces, but if they are theoretically equal in strength 'to the pressure 
 needed the stays may have a strain of 10,000 Ibs. per effective square inch of sectional 
 area. 
 
 " Surveyors will remember that the strength of a sphere to resist internal pressure is 
 double that of a cylinder of the same diameter and thickness. 
 
 "All manholes and openings must be stiffened with compensating-rings of at least the 
 same effective sectional area as the plates cut out, and in no case should the plate-rings 
 
 * If the rivets are exposed to double-shear multiply the percentage as found by 1.5. 
 
SEC. 4 SHELL, FURNCAES, AND BACK-CONNECTIONS. 221 
 
 be less in thickness than the plates to which they are attached. The openings in the 
 shells of cylindrical boilers should have their shorter axes placed longitudinally. It is 
 very desirable that the compensating-rings round openings in flat surfaces be made of 
 L or T iron." 
 
 4. Furnaces. The plates used in the construction of the furnaces should be made 
 as thin as is consistent with proper strength and stiffness, because thick plates are 
 liable to blister when exposed to the intense heat of- the fire. For the same reason as 
 few seams as possible are used in the furnace and back-connection, and the lap-joints 
 are single-riveted. The laps should be placed in such a direction that the current of 
 the products of combustion does not strike the edge of the plate, and that there is no 
 tendency for the scale to lodge and accumulate at any place exposed to an intense heat. 
 The furnace-crown should have as little bracing as practicable, and the attachments of 
 the stays and braces should be of such a form that they interfere as little as possible 
 with the circulation of the water, and do not serve as a nucleus for an excessive accu- 
 mulation of scale. 
 
 For furnaces special brands of iron, known as "fire-box" and "extra fire-box" iron, 
 are used, which are specially fitted to withstand the oxidizing and wasting influence of 
 intense heat and the impact of flame. The furnaces of all boilers built for the English 
 Admiralty are made of "Low Moor" or "Bowling" iron. It is essential that the iron 
 used for furnaces should be free from all lamination. The use of mild steel for fur- 
 naces is advantageous especially on account of its homogeneous structure ; it is, how- 
 ever, necessary to exercise great care in the selection and use of steel for furnaces, since 
 in many instances the steel plates of furnaces and fire-boxes have cracked after having 
 been in use a short time, although the material was of a mild quality and stood the 
 tempering-test well. 
 
 Copper has been used extensively for locomotive fire-boxes, especially in Europe, on 
 account of its homogeneous structure and high thermal conductivity, but it is not used 
 for marine boilers. 
 
 Yaiious devices have been proposed for increasing the heating-surface of the fur- 
 nace, but nearly all of them interfere to so great an extent, with the removal of the 
 scale or with the cleaning of the fire, besides presenting difficulties of construction, 
 that they are almost unknown in practice. 
 
 The corrugated boiler-flues made by the Leeds forge Company (England), under 
 the patents of S. Fox, have come into extensive use of late, and have generally given 
 satisfaction. In some cases iron corrugated furnace-flues have shown signs of blister- 
 ing after having been in use a short time ; but this is ascribed to original defects in the 
 
222 STEAM BOILERS. CHAP. IX. 
 
 plates, and not. to any injurious effect of the process of manufacture. Corrugated flues 
 have borne test-pressures more than twice as great as the pressures at which plain cylin- 
 drical flues, made of the same material and of equal dimensions, collapsed. The great 
 resistance to collapse of corrugated flues of an oval cross-section, compared with that of 
 plain oval flues, is of special importance." These corrugated flues possess great longi- 
 tudinal elasticity, and thus accommodate themselves readily to differences of pressure 
 and temperature, rendering the flat plates of the boiler-front and back-connection to 
 which they are fastened less liable to grooving. This longitudinal elasticity prevents 
 thick scale from adhering firmly to the metal, and thus the surfaces of the tubes are 
 always kept clean ; by this fact the greater evaporative efficiency claimed for corru- 
 gated boiler-flues may perhaps be explained. 
 
 These flues are made either of the best Yorkshire iron or of Siemens-Martin steel of 
 a very mild quality. The plate is first bent to a cylinder and welded by machinery ; 
 then the circumferential corrugations are rolled in. It is proposed to roll solid steel 
 tubes without welds from seamless circular blooms under S. Fox's patents. 
 
 The following data in reference to tests of S. Fox's corrugated boiler-flues are given 
 in Engineering, March, 1878, and June, 1880 : .- 
 
 A flue made of welded steel, f inch thick, had thirteen corrugations in a length of 6 
 feet J inch between extreme centres, giving a mean pitch of 6.03 inches. The depth of 
 the corrugations was 1 inches, and the mean least diameter of the flue was about 33| 
 inches. There was at each end a plain part extending 12f inches beyond the end corru- 
 gations ; these plain parts were packed in the end plates of the test-cylinder by cupped 
 leather rings. In this manner the flue was quite free to move longitudinally. 
 
 The flue bore a hydraulic pressure of 550 Ibs. per square inch without showing any 
 permanent set. At a pressure of 600 Ibs. the flue began to fail, but retained a symme- 
 trical oval form, the difference of the longer and shorter diameters being about 5 inches. 
 This oval flue, being again tested, gave way at a pressure of 350 Ibs. 
 
 A welded iron flue, f inch thick, having a mean least diameter of about 35 inches, 
 and having circumferential corrugations 1^ inches deep and 6 inches pitch, was tested 
 in the same apparatus, and gave way by general distortion as a pressure of 450 Ibs. was 
 approached. 
 
 A plain cylindrical flue, made of the same material, f inch thick, and having a mean 
 horizontal diameter of 36.63 inches and a mean vertical diameter of 36.98 inches, came 
 down like a blister on the top at a pressure of 200 Ibs. while being tested in the same 
 apparatus. 
 
 Corrugated plates are also coming into use for the fire-boxes of portable and locomo- 
 
SEC. 5. 
 
 SHELL, FURNACES, AND BACK-CONNECTIONS. 
 
 223 
 
 Fig. 93. 
 
 Fig. 94. 
 
 ; 
 
 
 
 WWWWWVi 
 
 tive boilers in England. In Garrefs portable engine boilers the top of the rectangular 
 fire-box is made with several deep corrugations extending lengthwise the furnace. The 
 Leeds Forge Company are building portable and locomotive boilers the semi-cylindri- 
 cal top and flat sides of which are stayed by a new system of diagonal corrugations 
 patented by Fox and Greig. The external shell around the fire-box is corrugated in 
 the same manner, and the use of stay-bolts and sling-stays is dispensed with. 
 
 5. Furnaces of Rectangular Boilers. Furnaces of rectangular or semi-cylin- 
 drical boilers have generally flat, vertical sides joined 
 to a flat bottom by corners curved to a short radius, 
 and to a more or less arched crown. 
 
 A flat furnace-crown is not only the weakest 
 form, requiring very heavy bracing, but interferes 
 greatly with the proper circulation of the water. 
 Figures 93, 94 are intended to illustrate the circu- 
 lation of the water with flat and arched furnace- 
 crowns respectively, the dotted lines representing 
 the ascending bubbles of steam, and the arrows in- 
 dicating the direction of the currents of water flowing in to fill the vacant space. Flat 
 furnace-crowns have the advantage of making the furnaces roomy over the grate, and 
 are used generally in locomotives, in which heavier bracing can be used with safety on 
 the furnace-crown than 
 in marine boilers, since 
 they are less liable to 
 be injured by incrusta- 
 tions. In marine boil- 
 ers, carrying a mode- 
 rate steam - pressure, 
 the furnace-crown has 
 often the form of a flat 
 arch ; in this case the 
 crown-sheet is lap- 
 jointed to the flat sides; 
 sometimes the arched 
 crown-sheet is bent to 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 r 
 
 Fig. 95, 
 
 a shorter radius where it is joined to the sides, as in figure 95. 
 
 The form of furnaces commonly used in United States naval vessels is illustrated on 
 
224 
 
 STEAM BOILERS. 
 
 CHAP. IX. 
 
 Plates VI. , VII., XVII. The semi-cylindrical form of the furnace-crown gives a much 
 less roomy furnace, but it makes the inside of the boiler much more accessible for the 
 same height of boiler, and requires scarcely any bracing. In this case the whole crown 
 and the straight sides of the furnace may be made in one piece, so that no seams occur 
 in the furnace except at the ends and below the grate ; but often the crown-sheet con- 
 sists of two plates connected by a transverse seam in the middle of the furnace. When 
 it is necessary to use longitudinal seams within the furnace above the grate they should 
 never be placed near the line of fire, but (as in Plate XVII.) near the top of the crown, 
 far enough to one side to clear the foot of the braces. The fibre of the plates forming 
 the furnace runs generally circumferentially. 
 
 The flanges required to secure the furnace in the boiler at the front and back are 
 generally not turned on the plates forming the furnace ; in some English and French 
 boilers a flange is turned outward on the back end of the furnace-crown, to a radius of 
 four or five inches, to connect it with the back tube-sheet. With this method of fasten- 
 ing the movements due to the expansion and contraction of the furnace and of the tubes 
 tend to spring the joint open and cause leaks, unless the flange is made sufficiently flex- 
 ible by the large radius with which it is turned. Angle-irons are frequently used in 
 English and French practice for securing the furnace at the front and back in the boiler. 
 The cheapest way of securing the furnace to the front of the boiler is to cut out the 
 front plate equal to the cross-section of the furnace, rivet an angle-iron around this 
 opening, inside or outside the shell, and secure the furnace to this angle-iron. Cast- 
 iron frames, to which the furnace and ashpit doors are attached, are bolted to the front 
 of the boiler around the furnace-opening. These large frames are apt to warp and 
 
 crack, and give much trouble. On 
 this account the furnace-mouth is 
 contracted in many French and Eng- 
 lish boilers by flanging the furnace 
 crown-sheet downward in the manner 
 shown in figure 95. 
 
 Sometimes the crown-sheet is made 
 to slope downward to the front of the 
 boiler, as illustrated in figures 96, 97, 
 
 in order to gain room under the front- 
 
 lgl ' connection, when the boiler is neces- 
 
 sarily low. This can be done without impairing the efficiency of the furnace, since more 
 height is required over the grate at the back than at the front. 
 
SEC. 5. 
 
 SHELL, FDBNACES, AND BACK-CONNECTIONS. 
 
 225 
 
 Fig. 97. 
 
 Figure 96 represents the furnace of a tubular marine boiler built by Laird & Son, 
 Birkenhead, England ; and figure 97 represents the furnace of a boiler for the United 
 States tugboat Glance, built in 1879. 
 
 When the water-spaces are too narrow to connect the furnace to the shell by flanges 
 turned on the respective plates, or by special flanged pieces as around the furnace- 
 door opening in figure 97, the 
 connection is formed either 
 by placing a frame, formed 
 by bending and welding a 
 wrought - iron bar of square 
 cross-section to the required 
 shape, between the furnace and 
 the shell, and riveting the plates 
 to the frame by through-rivets 
 (as at the bottom of the boiler 
 in figure 96, and around the 
 furnace door-opening in figure 
 1, Plate XXVIII.), or by bend- 
 ing the plate of the furnace 
 with an easy, reverse curve 
 outward, and riveting it directly to the shell (as at the bottom of the boilers represented 
 on Plate XXVIII.) . 
 
 The usual method of securing the furnace used in boilers of United States naval 
 vessels is illustrated on Plates VI., VII., and XVII. An opening of the size and shape 
 required for the furnace-door and ashpit is formed in the boiler-front, a flange being 
 turned inward around this opening. The crown-sheet and sides of the furnace are 
 riveted to a separate flanged piece, which has an opening corresponding to the furnace- 
 door opening of the boiler-front, but is flanged outward around this opening. A narrow 
 strip is riveted inside the furnace-door opening to the flanges of the boiler-front and 
 of the furnace front -piece. 
 
 The bottom of the furnace of rectangular boilers is sometimes arched downward ; 
 but the flat bottom, illustrated on Plate XVII., is best adapted for supporting the flat 
 bottom of the boiler by stay-bolts, while the curved corners give additional room in the 
 water-space for collecting solid matter and for cleaning out the water-legs and the 
 water-bottom through the handholes. 
 
 The principal use of the water-bottom is to form a convenient ashpan, which does 
 
226 
 
 STEAM BOILERS. 
 
 CHAP. DC 
 
 not become overheated and warped by the heat radiated through the grate and by the 
 fire falling through the bars ; it serves also as a receptacle for mud, broken scale, and 
 other solid matter entering the boiler with the feed- water. The water-bottom is gene- 
 rally one of the first parts to give out in marine boilers through internal and external 
 corrosion, and it is in many cases inaccessible for thorough repair while the boilers re- 
 main in the vessel. On this account they are often omitted, and separate cast-iron or 
 wrought-iron ashpans are placed under each furnace. The weight of these so-called 
 dry-bottom boilers, including the supports and ashpans, is about the same as that of the 
 water-bottom boilers, including the additional amount of water carried by them ; but 
 the former are somewhat cheaper to construct. To prevent the over-heating and warp- 
 ing of the ashpans of dry -bottom boilers water has to be kept in them. Although they 
 deteriorate rapidly, the durability of the boiler is not impaired thereby as by the corro- 
 sion of the water-bottoms. The durability of the latter is increased by giving to the 
 plates an additional thickness in order to allow for corrosion. 
 
 To form the water-legs of the dry-bottom boiler the sides of the furnaces, carried 
 some distance below the grates, are connected at the bottom by a separate curved plate. 
 The lap of this plate is often placed inside the boiler ; by placing it on the outside the 
 joint is made more accessible for calking while the boiler remains in position in the 
 ship, and some additional room is gained in the water-leg. Sometimes the water-legs 
 are enlarged at the bottom to make them more accessible for cleaning ; they must ex- 
 tend far enough below the grate so that there is no danger of their becoming filled with 
 
 solid matter up to the line of fire. It is well to 
 let the bottom of the water-space running length- 
 wise at the back of the boiler terminate in a cylin- 
 drical drum, unobstructed by stays and accessi- 
 ble through manholes placed at the ends for 
 cleaning and repairs (see figure 98). 
 
 6. Cylindrical Furnaces. The furnaces of 
 cylindrical boilers calculated to carry steam of a 
 high pressure are made of a circular cross-sec- 
 tion. In chapter vi. the laws governing the col- 
 lapsing strength of cylindrical flues have been 
 investigated, and the importance of making their cross-section perfectly circular has 
 been pointed out. For this reason the furnace-flues are generally made with longitudi- 
 nal butt-joints, and not with lap-joints. The strap is placed inside the flue at the side 
 and below the grate, so as to be accessible for calking, not to come in contact with the 
 
 r 
 
 Fig. 98. 
 
SBC. 6. 
 
 SHELL, FURNACES, AND BACK-CONNECTIONS. 
 
 227 
 
 Fig. 99. 
 
 fire, and not to be in the way of hauling the ashes. It is still better to weld the longi- 
 tudinal seams in the manner described in section 20, chapter viii. 
 
 The necessity of stiffening long flues by flanges or by riveting bands around them at 
 intervals has been demonstrated in chapter vi. In boilers of United States naval vessels 
 the furnace-flues are generally strengthened by means of the 
 "Adamson" joint (see figure 99 and Plates VIII., XL, XII.) 
 The furnace-flue consists of two or three sections flanged outward ; 
 a wrought-iron ring, about f thick, is placed between the flanges, 
 and the sections are connected by single-riveting. There are no 
 laps or rivets in contact with the fire, and the ring or welt allows the joint to be calked 
 from the inside and outside. This flanging can be done only with very good iron ; in 
 some boiler- works it is done by machinery in one or two heats, which distresses the 
 plate less than the repeated heating with the common method an important advantage, 
 especially in the case of steel. The radius at the root of the flange should not be less 
 than f inch on the inside, or the plate will be liable to become grooved by the alternate 
 expansion and contraction. 
 
 Sometimes the several lengths of the furnace-flue are connected by T-iron rings, as 
 shown in figure 100. In order to admit of calking these 
 joints at any time from inside the furnace a clear space 
 of at least one inch should always be allowed between ^_^ 
 
 the ends of the plates ; this lessens also the liability to ^^ J- . , , . 1 ,,,,,^ 
 overheating at the seam. Accurate workmanship is re- 
 quired for this joint ; the two lengths of tube embraced by the same ring must be of ex- 
 actly the same diameter or the joint will give trouble. 
 
 In order to allow for the contraction and expansion of the furnace-flue, which often 
 cause serious trouble by grooving in long boilers, the Bowling hoop has been intro- 
 
 I Fig. 100. 
 
 Fig. 101. 
 
 duced (see figure 101). Like the T-iron hoop, it has 
 the disadvantage of placing a double thickness of 
 plate and the rivet-heads in the fire at the joints. 
 
 The report of the Chief-Engineer of the Man- 
 chester Steam-Users' Association for the year 1871 
 contains the following directions regarding the 
 
 application of strengthening-hoops to the flues of boilers originally made without 
 
 them : 
 
 "The hoops should be of angle-iron section, about 3" X 3" X i". They should be 
 
 made in halves, so that they may be passed in at the manhole and then riveted to the 
 
228 STEAM BOILERS. CHAP. IX. 
 
 furnace or flue tubes in position, thus rendering it unnecessary either to remove the 
 tubes or cut any opening in the boiler. The angle-iron should not be brought into 
 direct contact with the plates of the tube, but a clear space of not less than one inch 
 should be left between the two, the hoop for this purpose having a diameter some two 
 inches larger than the furnace-tube (see figure 102). The hoop 
 should be secured to the furnace-tube by rivets spaced about six 
 inches apart; blocking-pieces, through which the rivets should 
 
 pass, being inserted between the tube and the angle-iron, so as to 
 give a solid abutment for the riveting, while the halves of the hoop 
 should be connected by butt-straps riveted to their ends at the back. 
 . . . The blocking-pieces should be made of a strip of iron not 
 more than ^ inch thick, bent round into a circular shape, the ends being welded to- 
 gether so as to form a short tube or ferrule. These ferrules should be well fitted into 
 the space between the hoop and the plating of the furnace and flue tube, while the ends 
 of one half -hoop should be firmly butted against the ends of the other half -hoop, so 
 that the whole may be tightly drawn together, as much of the support afforded by these 
 hoops depends on their being made one with the furnace and flue tubes, and not put in 
 so as to act merely as separate hoops from which the plates are hung. . . . 
 
 "A hoop of angle-iron is preferable to one of T-iron, as the single flange of the 
 angle-iron, being narrower than the double flange of the T-iron, offers less impediment 
 to the escape of the steam generated within the annular space, and also less harborage 
 for deposit. . . . The hoops should not be allowed to touch the shell of the boiler, or the 
 furnace-tubes may become strained and leakage be induced, since furnace and flue 
 tubes rise and fall with variations of temperature, and thus grind against the sides of 
 the shell or against one another where in contact." 
 
 Hoops secured in the foregoing manner should be used only when their addition is 
 an after-consideration ; for new structures one of the joints illustrated in figures 99, 100, 
 and 101 should be used, because, as has been pointed out in section 3, chapter vi., it is 
 important that the flanges or rings supporting a flue should be attached to it rigidly 
 around its whole circumference, and not merely connected with it at detached points. 
 When a boiler contains several furnaces it is well to arrange the lengths of the sections 
 forming the flues in such a manner that the flanges or strengthening-hoops of adjoining 
 furnaces are about six inches apart, so that any loose scale dropping down from above 
 does not lodge between the flanges. 
 
 Angle-iron rings are frequently used in French and English boilers for securing the 
 furnaces to the shell and to the back-connections. In all cylindrical boilers constructed 
 
SEC. 7. SHELL, FURNACES, AND BACK-CONNECTIONS. 229 
 
 for United States naval vessels the furnaces are riveted to flanges turned on the back 
 tube-sheet and on the front-head of the boiler, double-riveting being used at the front 
 and single-riveting at the back of the furnace. 
 
 The furnaces of the boiler illustrated on Plate XV. are made of two plates, the up- 
 per one " thick and the bottom one -fa" thick, welded together 6 inches below the 
 centre line. The plate forming the furnace-crown is flanged outward at the back and 
 is riveted to the back tube-sheet ; the bottom-plate is continued straight till it meets 
 the back-plate of the back-connection, being riveted to the flange turned on the 
 latter. 
 
 Lloyd's formula for the collapsing strength of circular flues is : 
 
 89,600 x square of thickness of plate in inches __ j working pressure in 
 
 Length of flue in feet X outside diameter of flue in inches ~ ( pounds per sq. inch. 
 
 The Surveyors of the Board of Trade (England) determine the strength of circular 
 furnaces according to the following rule : 
 
 " Circular furnaces with the longitudinal joints welded or made with a butt-strap : 
 
 90,000 x the square of the thickness of the plate in inches 
 
 (Length in feet + 1) x diameter in inches = WOrkm S P ressure ^ ** m " 
 
 "Without the Board's special approval of the plans the pressure is in no case to 
 
 , 8,000 X thickness in inches , 
 
 exceed - ,. : = . The length to be measured between the rings, if 
 
 diameter m inches 
 
 the furnace is made with rings. 
 
 " If the longitudinal joints, instead of being butted, are lap-jointed in the ordinary 
 way, then 70,000 is to be used instead of 90,000, excepting only when the lap is bevelled 
 and so made as to give the flues the form of a true circle, when 80,000 may be used. 
 
 " When the material or the workmanship is not of the best quality the constants 
 given above must be reduced that is to say, the 90,000 will become 80,000, the 80,000 
 will become 70,000, the 70,000 will become 60,000 ; and when neither the material nor 
 the workmanship is of the best quality such constants will require to be further re- 
 duced, according to circumstances and the judgment of the surveyor, as in the case of 
 old boilers. 
 
 " One of the conditions of best workmanship must be that the joints are either 
 double-riveted with single butt-straps or single-riveted with double butt-straps, and the 
 holes drilled after the bending is done and when in place, and afterwards taken apart, 
 the burr on the holes taken off, and the holes slightly countersunk from the outside." 
 
 7. Combustion-chambers and Back-connections. In the usual type of loco- 
 
230 STEAM BOILERS. CHAP. IX. 
 
 motive boilers (see Plates IV. and V.) the front tube-sheet forms the back of the fur- 
 nace ; in other instances (see Plate III.) the gases enter a separate combustion-chamber 
 on leaving the furnace ; and in the usual type of return-tube boilers the back-connection 
 serves as a combustion-chamber. 
 
 The opening leading from the furnace to the combustion-chamber is contracted to 
 the smallest area that will give the required draught by the bridge which at the same 
 time serves as a support for the back of the grate. Plates VI., VII., XVII. illus- 
 trate the usual manner of forming the bridge-wall in rectangular boilers of United States 
 naval vessels. In cylindrical marine boilers, where the simplest forms and unstayed 
 surfaces are used as much as possible, the bridge is generally formed by a casting lined 
 with fire-brick above the surface of the grate. Figure 1, Plate III., illustrates a fre- 
 quent mode of forming the bridge viz., by means of a hollow wall communicating with 
 the water-space of the boiler ; this is called a water -bridge. The top of a water-bridge 
 should always slope or curve upwards towards the ends to admit of the rapid escape of 
 steam-bubbles. Sometimes a water-bridge projects downward for the purpose of de- 
 flecting the flame ; it is then called a hanging-bridge. 
 
 Return-tubular boilers are built either with one back-connection common to all the 
 furnaces or with a separate back-connection for each furnace. The former plan simpli- 
 fies the construction and lessens greatly the weight of the boiler, including water ; and 
 since it admits of placing at least one additional row of tubes over each furnace, the 
 heating-surface is about the same as when each furnace has a separate back-connection. 
 The principal advantages of the latter plan are twofold viz., first, the products of com- 
 bustion are kept separate till they enter the front-connection or uptake, consequently 
 the efficiency of any furnace is not affected by counter-currents, or currents of cold air 
 entering the back-connection through adjoining furnaces ; secondly, the water-spaces 
 between the separate back-connections and nests of tubes are of great utility in 
 facilitating the circulation of the water in the boiler. When more than two fur- 
 naces are contained in the same shell separate back-connections are generally used. 
 Cylindrical boilers calculated to bear a high pressure of steam are often made with 
 one common back-connection, in order to reduce the amount of flat stayed surface 
 as much as possible ; when they contain more than two furnaces it would be diffi- 
 cult to get plates of sufficient size for a single back tube-sheet and to flange the sheet 
 properly. 
 
 In the horizontal-tubular boiler the front of the back-connection is formed by the 
 back tube-sheet, which in the cylindrical boiler consists of one plate. In the rectangu- 
 lar boiler it generally extends down to the straight sides of the furnace. A flange 
 
SEC. 8. SHELL, FURNACES, AND BACK-CONNECTIONS. 231 
 
 encircling the furnace is turned on the tube-sheet ; the top and sides of the same are 
 likewise flanged for riveting it to the top and side plates of the back-connection. 
 
 In the vertical return-tube boiler the bottom tube-sheet is connected with the fur- 
 nace by a separate flanged piece (see Plates VI., VII.) The top of the back-connection 
 is often formed by curving the back-plate to a large radius ; in other cases the back- 
 plate is made to slope inward, thereby increasing the width of the water-space at the 
 back of the boiler, and allowing the steam-bubbles to escape from the plate readily as 
 soon as formed. It is necessary to leave sufficient room at the top of the back-connec- 
 tion for calking the seams and expanding the tubes. 
 
 8. Systems of Boiler-building. The preliminary tests applied to boiler-plates 
 to determine their strength, soundness, and quality, methods of developing the lines 
 of curved sheets and laying out the flanges and rivet-holes, and the various opera- 
 tions of bending, flanging, welding, shearing, punching, drilling, riveting, and calking, 
 have been described and discussed in chapters v. and viii. 
 
 Various systems are followed by boiler-makers in building boilers, as far as the 
 laying-off, the fitting, and the order in which the several parts are riveted together are 
 concerned. Since in bending, flanging, and forming the plates absolute accuracy in 
 accordance with prescribed dimensions is impracticable, and since slight changes in the 
 location of laps and seams often encroach upon the available space in such a manner as 
 to make other modifications absolutely necessary, it is important to verify all work by 
 fitting the plates together after they are bent and flanged, securing them temporarily 
 by bolts, placing the various parts of the boiler in position, and thus to make sure 
 before riveting the joints permanently that this can be done without straining any part 
 unduly, and that the proper clearance is maintained between the several parts. Care 
 must be had to join the parts together in such order as to have all seams accessible for 
 riveting and calking. 
 
 The rivet -holes in the shell of rectangular boilers are generally punched or drilled at 
 first in one plate only, which then serves as a template for marking the corresponding 
 holes on the other plate while it is temporarily fitted and secured in position. The 
 rivet-holes of the circumferential and longitudinal seams of cylindrical shells and of 
 their butt-straps are often all laid off and punched or drilled before the plates are bent ; 
 this work can be done more accurately, however, when the holes of each seam are 
 formed in one plate only, which then serves as a template for drilling the correspond- 
 ing holes in the other plates while secured in position, as described in section 6 of 
 chapter viii. 
 
 In building the boiler represented on Plate XII. the two rings forming the cylindri- 
 
232 STEAM BOILERS. CHAP. IX. 
 
 cal shell are fitted and secured together with their butt-straps by tack-bolts. The 
 back-head, having been flanged in the meanwhile, is then fitted in place, and the rivet- 
 holes on the circumferential flange are marked and drilled. The back-head having 
 been fitted and secured again to the back-section of the shell, the longitudinal butt- 
 strap of the latter and the transverse seam of the back-head are riveted, and the back- 
 head is riveted to the shell. The circumferential butt-strap is riveted to the back- 
 section, while the front-section is fitted and secured to it ; the longitudinal and 
 circumferential seams of the front-section are then likewise riveted. After this the 
 laps are calked from the inside, and the gusset and stay plates are fitted and bolted to 
 the back-head. In the meanwhile the furnace-flues have been constructed, the diffe- 
 rent sections having been riveted together and calked ; the front and back tube-sheets 
 have been flanged and drilled, and are being fitted to the furnace-flues. In doing this 
 the tube-sheets are kept the proper distance apart and parallel with each other by a 
 couple of wooden struts and some long bolts passing through the tube-holes ; a few 
 tubes are put through the holes to make sure that the corresponding holes in the two 
 plates come fair with each other ; thus secured, the furnaces and tube-sheets are fitted 
 in the boiler-shell. The holes in the flanges joining the tube-sheets to the furnaces and 
 the front-head to the shell are marked and drilled. The furnaces having been removed 
 from the shell, they are riveted to the back tube-sheet, and the flanges encircling the 
 furnaces are calked. The plates forming the back-connection are fitted to the back 
 tube-sheet, and the furnaces and back-connection, temporarily connected by bolts, are 
 fitted in the shell. The rivet-holes of the seams of the back-connection are then 
 marked and drilled, the plates are riveted in place, and such joints as would not be 
 accessible afterward are calked as soon as riveted. The furnaces and back-connections 
 are now placed in position in the shell ; the front-head is fitted and riveted to the fur- 
 naces and to the shell ; the stay and gusset plates are riveted to the heads and back- 
 connections ; the stay-bolts and other braces are put in place and secured. The tubes 
 are then put in and expanded, and finally all remaining seams are calked. 
 
 In building rectangular boilers the bottom and the lower part of the sides and back 
 are fitted and riveted while the furnaces and back-connections are being constructed ; 
 these are then put in position in the shell on blocks or wedges ; the front, as far as the 
 front-connection, is fitted and riveted, the holes for stays are marked and drilled, and 
 the furnaces are permanently secured within the shell by stay-bolts, and the crow-feet 
 for the attachment of the braces are riveted to the furnace-crowns. The plates forming 
 the back, sides, and top are in the meanwhile fitted and riveted on, generally one sheet 
 at a time. 
 
SEC. 8. SHELL, FURNACES, AND BACK-CONNECTIONS. 233 
 
 The tube-boxes of the vertical-tubular boiler represented on Plates VI. and VII. 
 are fitted, riveted, and calked ; the lugs and straps for the attachment of the braces are 
 riveted to them, and, when the sides of adjoining boxes are tied together by lugs and 
 pins instead of stay-rivets, the tubes are put in and expanded before the boxes are put 
 into the boiler. 
 
 The front-connection and uptake are then built in, and, before the boiler is finally 
 closed, the angle-irons and long braces, dry-pipes, surface blow-pipes, and everything 
 which could not afterwards be introduced through manholes are passed into the boiler. 
 The braces having been connected, the tubes are put in and expanded, and all the re- 
 maining seams of the boiler are calked inside and outside. 
 
CHAPTER X. 
 
 STAYS AND BRACES. 
 
 1. Systems of Bracing. Bracing has to be applied to all surfaces of a boiler 
 which are liable to alteration of form under pressure. The only figures which require 
 no bracing when subjected to an internal pressure are the cylinder and the sphere. 
 Mathematical formulae expressing the resistance of cylindrical and spherical forms to a 
 bursting pressure have been developed in sections 1 and 2, chapter vi. ; and in section 3 
 of the same chapter the resistance of cylindrical flues to an external collapsing pressure 
 has been investigated. 
 
 Surfaces are stayed, -first, by contrivances which make them self-supporting, or 
 transmit the strain due to the pressure on them to well-supported places lying in the 
 same plane. To this class belong the girder-stays used on locomotive fire-boxes and on 
 the flat top of the back-connections of many marine boilers (see Plates IV., V., XV.) ; 
 the angle-irons, T-irons, and spherical strengthening-domes sometimes used on flat sur- 
 faces of small extent (see Plate XII.) ; likewise the various devices for strengthening 
 furnace-flues described in section 6 of chapter ix. 
 
 Secondly, by tying them to other surfaces lying in different planes and experiencing 
 an equal amount and kind of stress in an opposite direction, or possessing sufficient 
 stiffness to prevent alteration of form. The latter method is the one most usually em- 
 ployed ; stay-bolts, rod-braces, direct or oblique, branch-braces, stay-tubes, and gusset- 
 plates are used for this purpose. These two methods are also frequently used in 
 combination, as when angle or T-irons or stay-plates are used to distribute the strain 
 of braces over a larger area of the stayed surface. 
 
 There are several advantages in favor of diminishing the number of braces and in- 
 creasing their sectional area correspondingly viz., the boiler is made more accessible 
 and the circulation of the water is less obstructed ; it is easier to give an equal tension 
 to the fewer braces ; larger braces are relatively less reduced in sectional area by an 
 equal rate of corrosion than smaller ones. When, however, the spacing of the braces 
 exceeds a certain limit, depending on the steam-pressure and the thickness of the plate, 
 the latter bulges excessively between the stays, and, even if rupture does not take place, 
 
SEC. 1. STAYS AND BRACES 235 
 
 leaks occur around the stays ; to prevent this action the plates have to be stiffened in 
 such cases by T or angle irons or by an additional thickness of plate. 
 
 Various devices in bracing are employed to keep the interior of the boiler accessible 
 as much as possible : the braces are made so that they can be easily removed and re- 
 placed ; they are spaced as wide apart as practicable ; surfaces are strengthened by self- 
 supporting devices, or diagonal braces and gusset-plates are used to tie adjoining sides 
 together, leaving the space midway between the ends of the boiler unobstructed. In 
 order to obstruct the interior of the boiler as little as possible without leaving too long 
 an interval between the attachment of the braces, two or more short oblique braces are 
 attached to one main-brace ; the connection of these branches with one another and with 
 the main-brace is made either by a flexible joint or by welding. These branch-braces 
 are also called "half -moon" braces from the form sometimes given to the branches. 
 
 Careful attention must be paid to see that the parts to which the braces supporting 
 a surface are attached are strong enough and sufficiently stayed to bear the strain 
 thrown on them. Difficulties may be encountered in this respect when a larger surface 
 is tied to a smaller one. When a surface is tied to a self-supporting cylindrical surface 
 the strain on the latter will no longer be uniformly distributed around its circumference, 
 and distortion will take place unless the cylindrical plate possesses a surplus of 
 stiffness. 
 
 In cylindrical marine boilers the cylindrical shell is self-supporting ; the flat ends 
 and the back-connections have to be supported by bracing ; the methods used to secure 
 flues against collapse have been described in the last chapter. Above the back-connec- 
 tions the flat ends are either tied directly together by braces extending through the 
 whole length of the boiler, or are secured to the adjoining portions of the shell by 
 oblique braces or giisset-plates. The lower parts of the flat ends are tied to the back- 
 connections. The back-head is secured to the latter by stay-bolts ; the tube-sheets are 
 held by the tubes, and for further security special stay-tubes or rod-braces are often 
 added. The furnace-flues serve to tie the lower part of the front-head to the back-con- 
 nection, and any portions of these parts remaining unsupported are either tied together 
 by rod-braces or are strengthened by gussets, angle-irons, stay-domes, or similar de- 
 vices. The sides and bottom of the back-connections are secured against collapse by 
 tying them with stay-bolts to the circular shell or to opposite flat sides. The top of the 
 back-connections is generally supported either by girder-stays or by gusset-plates 
 secured to the back-head. 
 
 Plates XVII. , X VIII. illustrate the system of bracing used in rectangular boilers of 
 United States naval vessels. These boilers are designed to carry from 35 to 40 pounds 
 
236 
 
 STEAM BOILE&S. 
 
 CHAP. X. 
 
 Fig. 103. 
 
 of steam, and the shell is made of ^-inch or f-inch iron. Opposite portions of the flat 
 bottom and sides of the furnaces and connections and of the lower part of the shell are 
 tied to one another by socket-bolts spaced from 7f to 9 inches apart. 
 The tube-sheets are sufficiently stayed by the tubes. The upper por- 
 tions of the ends and of the back and front of the shell are tied together 
 by a system of parallel horizontal branch-braces traversing the length and 
 breadth of the boiler. In the horizontal return-tube boiler the top of 
 the shell is tied directly to the arched furnace-crown, the braces passing 
 between the nests of tubes ; but in the boiler having vertical tubes the 
 braces of the top are attached to the sides of the tube-boxes, and these 
 are tied to the furnace-crowns. The tendency of these braces to distort the circular 
 furnace-crown is counteracted by a system of horizontal stays, secured near the point 
 of attachment of the braces to opposite portions of each pair of adjacent furnaces. The 
 braces are connected with the shell of the boiler by means of T-irons, generally 4" X 3^" 
 X I", spaced from 12 to 14 inches apart from centre to centre, and securely riveted to the 
 shell. These T-irons run in continuous lengths as far down the sides of the boiler as is 
 
 Fig. 104. Fig. 105. 
 
 necessary for the attachment of the braces. Lugs, called " crow-feet,'" are riveted to the 
 furnace-crowns and to the top of the back-connections for the attachment of the braces 
 by means of a bolt passing through the eye (see Plate XVIII.) Similar lugs are riveted 
 
SEC. 2. 
 
 STAYS AND BRACES. 
 
 237 
 
 Fig. 106, 
 
 to the top and bottom of the sides of the vertical tube-boxes of the boiler illustrated on 
 Plate VI. 
 
 In many English horizontal return-tube boilers the lower ends of the braces sup- 
 porting the flat top are attached in the manner shown in figure 95. By the direct 
 attachment of the braces to the bottom of the boiler the arched furnace-crown is kept 
 free from unnecessary bracing ; but the cleaning of the boiler through the handholes is 
 rendered difficult. This defect is remedied by attaching these braces in the manner 
 shown in figure 103, a method used principally in dry-bottom boilers. 
 
 In order to dispense with the braces which are attached to the centre of the 
 furnace-crown it has been proposed to subdivide each branch of the braces sup- 
 porting the top of rectangular boilers in the manner shown in figure 104; and in 
 the horizontal return-tube boilers of the U. S. S. Tippecanoe (built in the year 1864) 
 the system of bracing shown in figure 105 was applied. The main-brace passes be- 
 tween the nests of tubes and is tied to the sides of the furnaces. But ordinarily it will 
 not be found advantageous to complicate the construction 
 of the braces by such devices, since very accurate fitting 
 is required in order to ensure equal tension on all the 
 branches. 
 
 Figure 106 shows a system of bracing applied to rectan- 
 gular boilers of French naval vessels. The top of the shell 
 is tied by oblique braces to the back and front of the shell 
 and to the uptake. N"o braces are attached to the arched 
 furnace-crowns, and few braces have to be removed to 
 make the upper part of the boiler quite accessible ; but 
 the greatly-inclined oblique braces and their attachments 
 experience a great strain, and the top and sides of the 
 boiler would possess much greater stiffness if the angle- 
 irons were made in continuous lengths. 
 
 2. Rules for Proportioning Braces. In calculating the strength of braces it 
 is generally assumed, when the plates are thin relatively to the pressure of steam 
 carried in the boiler, that each brace has to bear the whole strain due to the total 
 pressure on the portion of the surface which it supports. When the thickness of the 
 plate is increased, or when the plate is otherwise stiffened, the resistance which it 
 offers to bending may be taken into account in proportioning the dimensions of the 
 braces. 
 
 When the braces are spaced evenly in parallel and perpendicular rows the portion 
 
238 STEAM BOILERS. CHAP. X. 
 
 of surface supported by each brace is represented by the rectangles into which the 
 whole surface is divided by the rows of braces. 
 
 A large factor of safety, ranging from eight to sixteen, is used in calculating the 
 strength of braces on account of the constant diminution of their sectional area by cor- 
 rosion, the uncertain strength of welded parts, and the impossibility of ensuring an 
 equal tension on all braces under the varying conditions of pressure and temperature. 
 
 The corrosion of braces takes place most rapidly at or near the water-level, where 
 the mechanical action of the water in washing from side to side, and the alternate wet- 
 ting and drying of the braces, cause the layers of oxide and scale to fall off, thus ex- 
 posing new surfaces to corrosion. The stay-bolts of water-bottoms corrode rapidly be- 
 cause dirt accumulates there, which absorbs and retains moisture and often contains 
 corrosive substances. It is also found that braces in the path of steam flowing to stop- 
 valves, etc., suffer greatly from decay, and that braces are more easily attacked at welds 
 than at points where the fibres of the rolled bar have remained undisturbed. It is re- 
 commended to protect braces by thin washes of Portland cement, or by wrapping them 
 with marline or canvas and white or red lead, and with galvanized iron wire. Rods of 
 circular section are evidently less affected by corrosion than square or flat bars and 
 gusset or stay plates, since they expose less surface in proportion to their sectional 
 area. 
 
 ' Lloyd's Register of British and Foreign Shipping' contains the following " Rules 
 for Determining the Working Pressure in New Boilers": 
 
 "Stays. The stays supporting the flat surfaces are not to be subjected to a greater 
 strain than 6,000 Ibs. per square inch of section if of iron, and 8,000 Ibs. if of steel, cal- 
 culated from the weakest part of the stay or fastening, and no steel stays are to be 
 welded. 
 
 " Flat Plates. The strength of flat plates supported by stays to be taken from the 
 following formula : 
 
 C X T* 
 = working pressure in Ibs. per square inch, [I.] 
 
 where T= thickness of plate in sixteenths of an inch ; 
 P = greatest pitch in inches ; 
 C = 90 for plates -ft thick and below, fitted with screw- stays with riveted 
 
 heads ; 
 
 C = 100 for plates above -ft-, fitted with screw-stays with riveted heads ; 
 C = 110 for plates -ft thick and under, fitted with screw-stays and nuts ; 
 C 120 for plates above -ft, fitted with screw-stays and nuts ; 
 
SBC. 2. STAYS AND BRACES. 239 
 
 C = 140 for plates fitted with stays with double nuts ; 
 
 C =100 for plates fitted with stays with double nuts, and washers, at least 
 thickness of plates and a diameter of of the pitch, riveted to the 
 plates. 
 
 "NOTE. In the case of front-plates of boilers in the steam-space these numbers 
 should be reduced 20 per cent., unless the plates are guarded from the direct action 
 of the heat." 
 
 The rules of the Board of Trade (English) limit to 5,000 Ibs. the pressure per 
 ,. effective square inch of area of stay, but allow a greater pressure when the flat surface 
 is stiffened by T-irons or angle-irons in an approved manner. 
 
 The same authority prescribes the following rules for determining the highest ad- 
 missible steam-pressure in boilers under their supervision, according to the manner in 
 which the stays are spaced and secured, viz. : 
 
 " The pressure on plates forming flat surfaces will be easily found by the following 
 formula, which is used in the Board of Trade : 
 
 "~g 1 ^ = workin g P ressure - C 11 -] 
 
 T= thickness of plate in sixteenths of an inch ; 
 
 S surface supported in square inches ; 
 
 C = constant according to the following circumstances : 
 
 C = 100 when the plates are not exposed to the impact of heat or flame, and 
 the stays are fitted with nuts and washers, the latter being at least 
 three times the diameter of the stay and two-thirds the thickness of 
 the plates they cover ; 
 
 C = 90 when the plates are not exposed to the impact of heat or flame, and 
 the stays are fitted with nuts only ; 
 
 C = 60 when the plates are exposed to the impact of heat or flame, and steam 
 in contact with the plates, and the stays fitted with nuts and washers, 
 the latter being at least three times the diameter of the stay and two- 
 thirds the thickness of the plate they cover ; 
 
 C = 54 when the plates are exposed to the impact of heat or flame, and steam 
 in contact with the plate, and the stays fitted with nuts only ; 
 
 C = 80 when the plates are exposed to the impact of heat or flame, with 
 water in contact with the plates and the stays screwed into the plate 
 and fitted with nuts ; 
 
240 STEAM BOILERS. CHAP. X. 
 
 C 60 when the plates are exposed to the impact of heat or flame, with 
 water in contact with the plate, and the stays screwed into the plate, 
 having the ends riveted over to form a substantial head ; 
 C = 36 when the plates are exposed to the impact of heat or flame, and steam 
 in contact with the plates, with the stays screwed into the plate and 
 having the ends riveted over to form a substantial head. 
 
 " When the riveted ends of screwed stays are much worn, or when the nuts are 
 burned, the constants should be reduced ; but the surveyor must act according to the 
 circumstances that present themselves at the time of survey, and it is expected that in 
 cases where the riveted ends of screwed stays in the combustion-boxes and furnaces are 
 found in this state it will be often necessary to reduce the constant 60 to about 36." 
 
 The foregoing rules of Lloyd's and of the Board of Trade are modifications of for- 
 mula [IX.], section 5, chapter vi., representing the strength of flat square plates secured 
 at the edges, viz. : 
 
 . n 
 
 = -y -- or p 
 2 k n 
 
 In using this formula for calculating the thickness of flat plates supported by stay- 
 bolts Weisbach substitutes for n the diagonal of the square formed by four adjacent 
 stay-bolts. Since this diagonal is equal to n V 2 we have the equation : 
 
 k 
 
 [III.] 
 
 With the dimensions usually employed in practice plates supported by stay-bolts do 
 not give way by rupture in the middle between the rows of stays, but by buckling and 
 stretching, thereby breaking off the riveted heads of the stays in the first place and 
 then pulling the stay through the enlarged hole, or by tearing through the holes. It 
 must be remembered that when a single stay gives way the area of the unsupported 
 surface between the adjoining stays is increased four times. 
 
 Neglecting the stiffness of the plate, the following equation must exist between the 
 strength of the stay-bolts and that of the supported flat plate, when d is the diameter of 
 the stay-bolt, and when the tensile strength and the factor of safety of the stay-bolts and 
 of the plate are supposed to be the same, viz. : 
 
 n'p=2 k V .7854 d? k ; hence 
 d = 1.8t [IV.] 
 
SEC. 2. STAYS AND BRACES. 241 
 
 In practice, to allow for corrosion and give greater holding power to the riveted 
 heads, the diameter of stay -rivets is seldom made less than twice the thickness of the 
 plate. 
 
 Rules for proportioning and for finding the strength of screw stay-bolts with riveted 
 ends or nuts, as deduced from a series of experiments, will be found in section 9 of the 
 present chapter. 
 
 The tension of gussets must be calculated by the same rule as that for oblique stay- 
 bars ; but a much larger factor of safety must be employed in the case of gussets, not 
 only because they expose relatively much more surface to corrosion, but because the 
 resultant tension of a gusset is concentrated near one edge. Rankine says : "It ap- 
 pears advisable that its sectional area should be three or four times that of a stay-bar 
 for sustaining the pressure on the same area." 
 
 In the girder-stay the plate acts as a bottom flange to the girder and is fixed at the 
 ends, while the bar forms the web and upper flange and is merely supported at both 
 ends. We may consider the stay as a rectangular beam supported at both ends and 
 loaded in the middle or at several points, according to the number of bolts sup- 
 porting the plate. When loaded in the middle its strength is determined by formula 
 
 ^ Z = *|^[V.], where 
 
 p = pressure of steam in pounds per square inch ; 
 
 S= area of plate supported in square inches ; 
 
 I = length of span of stay-bar in inches ; 
 
 b breadth of stay -bar in inches ; 
 
 d = depth of stay-bar in inches ; 
 
 Jc coefficient of resistance equal to about 50,000 Ibs. per square inch. 
 A small factor of safety, about four, may be used in calculating the dimensions of the 
 stay-bar from the foregoing formula, on account of the strength imparted by the plate 
 acting as the bottom flange of the girder. Calling the distance between two adjoining 
 girders from centre to centre D, and taking four as the factor of safety, we get for the 
 depth of the girder-bar, from the above formula, the expression : 
 
 The breadtn of girder -stays varies generally from one-third to one-fifth of the depth. 
 In v,Tought-iron bars having a depth of not less than one-tenth of their length the de- 
 flection due to a load, less than that required to overcome the limit, of elasticity, is 
 
242 ' STEAM BOILERS. CHAP. X. 
 
 trifling. The deflection varies directly as the load and the cube of the length, and in- 
 versely as the breadth and the cube of the depth. 
 
 The Board of Trade (English) prescribes the following rule for determining the 
 highest pressure of steam admissible in boilers having surfaces stayed by girder- 
 stays, viz. : 
 
 " When the tops of combustion-boxes or other parts of a boiler are supported by 
 solid rectangular girders the following formula, which is used in the Board of Trade, 
 will be useful for finding .the working pressure to be allowed on the girders, assuming 
 that they are not subjected to a greater temperature than the ordinary heat of steam, 
 and, in the case of combustion-chambers, that the ends are fitted to the edges of the 
 tube-plate and the back plate of the combustion-box : 
 
 (W-P)I)XL = workin S P ressure - IT 11 -] 
 W = width of combustion-box in inches ; 
 P = pitch of supporting-bolts in inches ; 
 
 D = distance between the girders from centre to centre in inches ; 
 L = length of girder in feet ; 
 d = depth of girder in inches ; 
 T '= thickness of girder in inches ; 
 
 C = 500 when the girder is fitted with one supporting-bolt ; 
 C 750 when the girder is fitted with two or three supporting-bolts ; 
 C = 850 when the girder is fitted with four supporting-bolts. 
 
 "The working pressure for the supporting-bolts, and for the plate between them, 
 shall be determined by the rule for ordinary stays." 
 
 Lloyd's Eegister prescribes the same rule in a slightly different form. 
 The shearing strength of bolts, pins, or rivets by which braces are connected or are 
 attached to the boiler may be considered equal to the tensile strength of the brace ; for 
 the weakening effect of welding may be considered as offsetting the excess of tensile 
 strength over shearing strength of wrought-iron. 
 
 "To find the strength of an easy -fitting fastening against shearing, multiply the 
 sectional area by the modulus of strength ; then take two-thirds of the product if the 
 fastening is rectangular in section, or three-quarters if it is circular or elliptical in 
 section. 
 
 " For & perfectly tight-fitting fastening the strength is the whole product just men- 
 tioned. Many actual fastenings are intermediate between easy and perfectly tight 
 fastenings." (Rankine.) 
 
SEC. 3. STAYS AND BRACES. 243 
 
 Experiments on the strength of wronght-iron bolts, when subjected to the action of 
 single-shear and of double-shear, are recorded in section 7 of the present chapter. 
 
 In proportioning the ends of eye-bars or braces connected by pins or bolts the bear- 
 ing-surface of the latter is an important element of strength. When the dimensions of 
 the bolt are proportioned so as to make its sectional area equal to the least sectional 
 area of the bar, but with insufficient bearing-surface, the originally round hole will be- 
 come pear-shaped under an excessive strain ; the iron around the hole which was sub- 
 jected to compression will become thickened, and the other portions of the iron around 
 the hole which were subjected to tension will become thinned, and fracture will ulti- 
 mately commence and continue at this thin part around the eye regardless of the width 
 of the head. Experiments by Charles Fox on the flat links of the chains of suspension- 
 bridges lead him to the conclusion that the area of the semi-cylindrical bearing-surface 
 should be a little more than equal to the sectional area of the smallest part of the body, 
 and, as the iron in the head is not generally as strong as that in the body, the sum of 
 the width of the iron on both sides of the hole should be ten per cent, greater than the 
 width of the body. 
 
 In boiler braces this increased area of bearing-surface is often obtained by making 
 the depth of the eye greater than the diameter or thickness of the bar. 
 
 Experiments made to determine the proper proportions of pins and eye-bars, and 
 the best method of forming them, will be found in section 8 of the present chapter. 
 
 3. Screw-stays and Socket-bolts. In narrow spaces, as at the sides and be- 
 tween the furnaces and in the water-legs and water-bottoms of rectangular boilers, at 
 the sides, back, and bottom of the back and front connections, and between the tube- 
 boxes of vertical water-tube boilers, the opposite parallel plates are tied together by 
 stay-bolts, of which two varieties maybe distinguished viz., screw-stays and socket- 
 bolts. The former are screwed into both plates, and, in addition, their ends are gene- 
 rally secured by a riveted head or by a nut ; the socket-bolts pass through the plates 
 without being screwed, being held either by a riveted head or by a nut, and derive their 
 name from the thimble or socket surrounding them and fitted between the plates. Such 
 stays resist a collapsing as well as a bursting strain, and permit the riveting to be per- 
 formed and the nuts to be set up hard without springing the plates. Stay-bolts should 
 always be spaced, when possible, in vertical and horizontal rows in large surfaces, as 
 such an arrangement facilitates the scaling of narrow water-spaces. It is necessary that 
 the holes in the two plates connected by stay-bolts be exactly opposite each other, and 
 that the stays stand perpendicular to both plates, else the varying lateral strains due to 
 differences in pressure and temperature will soon cause the bolts to leak. Stay-bolts 
 
244 STEAM BOILERS. CHAP. X. 
 
 generally fail in consequence of the excessive bulging of the plates, which causes either 
 the heads of the stays to break or the plates to crack through the holes ; the strength 
 of such stays is therefore greatly increased by every addition to the depth of their 
 heads and to the area covered by them. On this account stay-bolts secured by nuts are 
 much stronger than riveted stays ; and by the use of washers the strength of stayed 
 surfaces may be still further increased. Stay-bolts secured by nuts have the additional 
 advantage that they may be renewed in many places where rivets could not well be 
 driven without moving the boiler from its seat. Nuts should, however, not be used on 
 surfaces in direct contact with the fire or exposed to great heat, because they are more 
 liable to be burnt than riveted heads, and will soon give trouble by leaking ; in ashpits 
 the nuts interfere with the hauling of the ashes and are liable to be loosened by the 
 hoes, unless they are protected by a false ashpan. 
 
 The thimble or socket of socket-bolts (see Plate XVIII.) is either made of cast-iron 
 or of a piece of boiler-iron bent into a cylinder. Its length should be exactly equal to 
 the width of the space between the stayed plates. They are put in position by hand or 
 with tongs, and are held by wooden plugs passing through them and both plates till 
 the stay-bolts are put in. Where not otherwise accessible the sockets are wired into 
 place by reeving a wire or string through both holes, hooking the bight between the 
 plates, pulling it within reach, cutting it, passing one of the ends .through the socket, 
 fastening the cut ends together, and hauling taut on the other ends till the socket is in 
 position. When stay-rivets are used the shank of the stay should not be too long and 
 should fit the socket well when hot, else in riveting the shank may be bent, as repre- 
 sented in figure 107; in such a case, when pressure is applied, the rivet 
 will be straightened, the plates will bulge, and leakage will ensue. 
 
 Owing to the absence of sockets screw stay-bolts are less cumbersome 
 and obstruct narrow water-spaces less than socket-bolts, but more labor 
 and greater accuracy in fitting are required with the former. The screw- 
 stays should fit tight in both plates. Whenever screw stay-bolts have to 
 be replaced it will be found necessary to ream out the holes and cut the threads anew. 
 Hollow screw stay-bolts have been used to admit jets of air to the furnace and combus- 
 tion-chamber. 
 
 For plates f inch thick and less the screw-stays should have fourteen threads to the 
 inch ; for J-inch and f-inch plates twelve threads to the inch are sufficient. The thread 
 is generally cut over the whole length of the bolt, and one end is left square till the 
 stay is screwed into the plates. The middle of the shank should be turned down to the 
 bottom of the thread, for it has been found that the elasticity of bars under a tensile 
 
SEC. 4 
 
 STAYS AND BRACES. 
 
 245 
 
 strain is much impaired by narrow grooves turned in them, the elongation being appa- 
 rently confined to these reduced places. Bolts with a smooth shank seem also to suffer 
 less from corrosion than when the thread is continuous. 
 
 The holding power of screw stay-bolts is greatly increased by securing their ends by 
 nuts or by riveting them over. A very soft quality of iron is required for such stays, 
 in order that this cold-riveting may be done without injuring the screw-threads in the 
 plate, and that the riveted head may possess proper strength. A description of the 
 best shape and dimensions and of the proper manner of forming the riveted heads of 
 screw stay-bolts will be found in section 9 of the present chapter. 
 
 4. Various Forms of Stays and Modes of fastening them. In French boilers 
 the stay represented in figure 108 is often used in narrow water-spaces ; it is stiff and 
 efficient, preventing collapse as well as bursting, but its use involves more labor than 
 that of stay-bolts. 
 
 The vertical sides of tube-boxes and back-connections have been stayed by means of 
 lugs or pieces of angle-iron riveted to the opposite surfaces and connected by pins pass- 
 ing through corresponding holes (see figure 109). This method is to be condemned, for, 
 even when these lugs are accurately fitted, the strain on the pin is unequal ; and this 
 action is aggravated to a dangerous extent by the almost unavoidable inaccuracies in 
 the location of these lugs, as shown in figure 110. A less objectionable form of this 
 
 Fig. 108. 
 
 Fig. 109. Fig< 110 
 
 Fig. 111. 
 
 tStSs. 
 
 kind of stay, which was applied to the boilers of U. S. S. Lancaster, is shown in 
 figure 111. 
 
 For short stays, connecting surfaces that are not parallel for instance, on the arched 
 crowns and at the lower rounded corners of furnaces, and on the rounded top of back- 
 connections a flat bar with the ends bent to the required angle, and secured at each end 
 by a single rivet, is generally used (see Plate XVIII.) The foot should never make an 
 acute angle with the body of this stay, else there will be difficulty in driving the rivet ; 
 on this account the two ends often have to be bent in different directions, making the 
 stay "]_-shaped. Such stays have the fault mentioned in connection with oblique braces 
 
246 STEAM BOILERS. CHAP. X. 
 
 viz., that there is a bending strain tending to spring the angle which the body of the 
 stay forms with the foot, and which should be made, on that account, with a large fillet. 
 When such rigidly-fastened braces are longer it is better to make the ends T-shaped, as 
 in figure 108, and secure them by two rivets, so that there is no longer a bending strain 
 but a direct pull on the fastenings ; sometimes the ends of the brace are made forked, 
 as in figure 112, the two branches being welded to the main body. The triangular 
 
 brace shown in figure 113, used to tie the bot- 
 I ' torn of the boiler to the lower rounded corners 
 
 112< fl ^ adjacent furnaces, is very stiff, but inter- 
 
 feres with the cleaning of the water -bottoms 
 through the handholes. 
 
 The ends of braces that have to be removed 
 
 and replaced from time to time either pass 
 
 through the plates and are secured by nuts and washers, or they are connected by pins 
 or bolts with lugs, angle or T-irons riveted to the stayed surfaces inside the boiler. The 
 former method of fastening is frequently used in English rectangular boilers for secur- 
 ing the lower ends of the braces which tie the top of. the boiler to the bottom or to the 
 furnace-crown and back-connection. Such braces are generally secured by a nut on 
 both sides of the plate (see figure 95), but sometimes a shoulder forged on the brace 
 takes the place of the nut inside the boiler. The large nuts and washers inside the fur- 
 nace, exposed to an intense heat, are apt to give much trouble, and the varying lateral 
 strains to which these long braces are exposed tend not only to cause leaks but to break 
 off the ends in the thread. Rods secured by nuts and washers are commonly used to 
 stay the uptake by tying it to the surrounding steam -drum. 
 
 The flat ends of cylindrical boilers are often tied together by cylindrical rods passing 
 through the shell and secured by nuts on both sides of the plates, the ends of the brace, 
 as far as the thread is cut, being enlarged. The strain due to the tension of these braces 
 is distributed over the plates either by large washers or by riveting angle-irons or an 
 extra thickness of plate to them. This brace is simple and easily adjusted, and with 
 accurate workmanship all lateral strains may be avoided when the end plates are suffi- 
 ciently stiffened so that they do not buckle to an appreciable degree. It is, however, 
 inconvenient to remove long braces by pulling them out of the boiler, and the screw- 
 threads inside the boiler become soon coated with rust and scale, making the turning of 
 the nuts very difficult. On this account a different fastening was applied to some rod- 
 braces of the boilers of U. S. S. Terror, illustrated on Plate IX. The brace is held by 
 bolts screwed into the enlarged ends from outside the shell ; the ends of the brace bear 
 
SEC. 4. 
 
 STAYS AND BRACES. 
 
 247 
 
 against washers which can be driven out after the bolts are withdrawn, and the brace 
 can then be removed to a convenient place within the boiler. In the boilers of U. S. S. 
 AmphitrUe (see Plate IX.) the ends of the shell are stiffened by two angle-irons riveted 
 to them a small distance apart, and the brace is drawn up by nuts against a cross-bar 
 resting on these angle-irons. In the boilers of U. S. S. Miantonomoh (see Plate IX.) 
 the tap-bolt passes through a ferrule against which the brace bears ; this ferrule is to 
 be filled with red-lead, and is intended to protect the bolt against scale and corrosion. 
 Figure 114 illustrates a brace used to stay the upper portion of the flat ends of some 
 
 Fig. 114. 
 
 cylindrical boilers built by R. Napier & Co., Glasgow, in 1870. The ends of the boilers 
 are stiffened by two T-irons (4$" x 4") spaced 10 inches between centres ; bolts, 2 inches 
 in diameter and 21 inches long, are riveted to the T-irons, 20 inches apart ; each pair of 
 bolts carries a stout cross-bar secured by nuts ; the brace, 2J inches in diameter, passes 
 through the middle of this cross-bar, the ends being held by cotters ; the brace is made 
 in two parts, connected by means of an eye and pin. This brace is easily removed and 
 adjusted to the proper tension, but the work involved in its manufacture makes it 
 rather expensive. 
 
 The bracing applied to the boiler on Plate XV. is much used, the T-ends being 
 secured to the angle-irons either by rivets or bolts. This form makes a very stiff brace, 
 and the strain is well distributed over the stayed plate ; consequently the braces may 
 be spaced rather wide apart, but it is difficult to remove and replace such heavy braces 
 within the boiler. 
 
 The attachment of a brace by means of a single bolt or pin, making a flexible joint, 
 has the great advantage that it allows the brace to adjust itself to the direction of the 
 resultant of the opposing forces, so that it experiences tension only in the direction of 
 its axis ; on this account this method of fastening is to be recommended, especially 
 when the braces are very long and relatively slender, and when they are not perpen- 
 dicular to the stayed surfaces. When such braces are made with jaws or forked ends, 
 which take hold of the lugs or T-irons riveted to the plate (see U. S. S. Monadnock, 
 Plate IX.), the place where the forked end joins the body of the brace is apt to be a 
 weak spot ; a simple eye-bar is not only free from this weakness, but is cheaper to 
 
248 
 
 STEAM BOILERS. 
 
 CHAP. X. 
 
 make. The eye-bar takes hold of two angle-irons or bracket-plates, placed far enough 
 apart to let the end of the brace pass between them without jamming ; the bracket- 
 plates are often made of considerable depth, and are stiffened by tying each pair to- 
 gether by bolts or rivets, using thimbles to keep them the proper distance apart. The 
 bracket-plates are made of the form shown on Plate IX., to make them lighter and to 
 facilitate the removal of the braces. 
 
 The brace represented in figure 115 is illustrated in plates accompanying Ledieu's 
 
 Fig. 115. 
 
 ' Traite elementaire des Appareils a Vapeur de Navigation,' and, 
 although not commonly applied, may sometimes be used to ad- 
 vantage instead of branch-braces. The short horizontal link re- 
 sists the normal component of the stress on the braces. 
 
 In the branch-brace each of the oblique branches is sometimes 
 formed by a pair of separate links ; usually, however, each link 
 consists of two rigidly-connected branches, as shown on Plate 
 IX. (U. S. S. Terror}. Two links are used in order to avoid forked 
 ends. These links have been formed of solid plates, an example 
 of which may be found on Plate XVIII. ; these are .very stiff, but heavy and clumsy. 
 
 Triangular links of the form shown in figure 116 have been iised to connect the lower 
 end of braces to the arched crowns of adjacent furnaces. The horizontal bar experi- 
 ences compression and prevents distortion of the furnace-crowns ; but it renders access 
 to the interior over the furnaces through the manholes very difficult, and it is therefore 
 better to rivet a separate stay to the furnaces a little lower down, as shown on 
 Plate XVIII. 
 
 In the boilers of U. S. S. Mohongo a similar triangular frame was riveted to the fur- 
 nace-crowns (see figure 117), making the boilers almost inaccessible over the furnaces. 
 
 Fig. 117. 
 
 Frequently the oblique branches are formed by simply bending a bar and letting the 
 pin of the brace bear directly on this bar at the angle ; the ends of the bent bar are 
 
SBC. 5. 
 
 STAYS AND BRACES. 
 
 249 
 
 either made with an eye, for the purpose of attaching them by means of a pin, or they 
 are rigidly riveted to the boiler (see figure 118). Such braces are much cheaper and 
 more easily made than those illustrated on Plate XVIII., but they experience irregular 
 bending strains, and they do not take up the strain on the stayed plates as well, and 
 are consequently less reliable and not to be recommended. 
 
 The braces should be connected by well-fitting bolts with coarse threads secured by 
 nuts, or by pins secured by cotters or keys ; the nut or key does not merely hold the 
 bolt or pin in place, but, when drawn up tight, adds much to the strength of the joint 
 
 Fig. 118. 
 
 Fig. 119. Fig. 120. 
 
 by preventing the jaws from spreading or the displacement of the links, so that the pin 
 or bolt experiences simply a shearing stress and not a bending stress. The use of split 
 pins (see figure 119) should be avoided, since they rarely fit the holes as well as bolts, 
 and give little or no lateral stiffness to the joint ; besides, the split ends are liable to 
 break off when they are opened, after inserting the pin, to prevent working out, or closed 
 for the purpose of backing it out. 
 
 In order to facilitate the adjustment of long braces to the proper tension the connec- 
 tion between the branches and the brace is sometimes made in the manner illustrated in 
 figure 120 ; or the braces are made in halves, connected by a turnbuckle made of brass 
 so that it does not rust fast to the brace. (See Plate XVIII.) 
 
 5. Fitting and Adjusting Rod-braces. Welding should be avoided as much as 
 possible in boiler-braces, for the strength of the brace depends upon the soundness of 
 the weld, which is frequently an uncertain element, and the iron at the weld is much 
 more readily attacked by corrosion than at the parts where the fibre has remained un- 
 disturbed. 
 
 The ends of long braces have to be forged separately to the required shape ; and 
 after being fitted, the threads cut, the pin-holes bored, etc., they are welded to the rods. 
 It is safer to forge rod-braces at first a trifling amount too long, and to adjust them to 
 the exact length by "upsetting" the rod. "Drawing-down," even to a small extent, 
 should be avoided ; in case the brace is found to be too short the rod should be cut and 
 a piece welded in. 
 
250 
 
 STEAM BOILERS. 
 
 CHAP. X. 
 
 The pin-holes in the ends of braces should be bored accurately, so that the pins fit 
 well and are thus subject to a shearing stress and not to a bending stress. In ordinary 
 boiler- work the pin-holes are often not bored at all, but are made by bending a square 
 bar so as to form a loop, and welding the ends together. Figure 121 illustrates a com- 
 mon method of forming the forked end of a brace which takes hold of a T-iron or a 
 single "crow-foot." 
 
 The vertical braces that pass between the rows of horizontal tubes are often made of 
 
 Fig. 121. 
 
 taken into account. 
 
 Fig. 122. flat bars in order to save room. When the manner of attaching 
 the brace to the shell would place its longer side in a transverse 
 direction to the axes of the tubes a half -twist is given to the end 
 of the brace, as shown in figure 122. Sufficient clearance must be 
 allowed between the braces and the tubes so that under the vary- 
 ing strains the braces do not chafe against the tubes and cut them 
 through. In long, horizontal braces the "sagging" has to be 
 In the rectangular boiler the long, horizontal braces tying the ends 
 of the boiler together are placed slightly above the plane of the shorter, stiffer braces 
 which tie the front and back of the boiler together, and rest on these ; in other cases 
 long, horizontal braces are supported in the middle by light hooks suspended from 
 the T-irons above. 
 
 Braces should be set up before the tubes are expanded and before the external seams 
 of the shell are calked. Before setting up the braces of the large flat surfaces of rect- 
 angular boilers it is well to shore these up evenly, else there is danger of setting up the 
 braces near the middle of the surface more than those near the sides, thus giving to the 
 surface a " dished" form. 
 
 To test whether long braces are set up to the same tension, tap each brace at the 
 same distance from the support with a hammer, and note whether the sound is in the 
 same key the tauter brace vibrates quicker and gives a note of higher key than the 
 slacker one, provided the rods have the same diameter and length. 
 
 6. Girder-stays, Gusset-stays, Stay-plates, Stay-domes, etc. Girder-stays 
 are either forged solid (see Plate IX.) or they are made of two plates riveted together at 
 the ends, with distance-pieces between them and a square washer placed on top for 
 the bolt to bear against (see figure 123). A clear space of about 1 inches should be 
 kept between the girder-stay and the supported plate ; to prevent the buckling of the 
 plate in screwing up the bolts tight, ferrules surrounding the bolts are inserted between 
 the plate and the bar, or the bolts pass through short projections forged on the lower 
 side of the bar. The girder should be of such length that its ends rest OH the perpen- 
 
.-. 0. 
 
 STAYS AND BRACES. 
 
 251 
 
 Fig. 123. 
 
 dicular plates forming the sides or ends of the back-connection or fire-box, and not on 
 the supported plate. The girder-stays of locomotive fire-boxes extend sometimes 
 through the whole width of the boiler, the ends 
 being riveted or bolted to the shell. Heavy 
 T-irons are also used for girder-stays, and 
 when they are long they are supported in one 
 or several places by braces hung from the top 
 of the shell ; the stay-bolts are placed stagger- 
 ing, passing alternately through either flange 
 of the T-iron. 
 
 The front plate of the back-connection of 
 the boiler illustrated on Plates VIII. and IX. 
 is stayed by a contrivance which may be classed among the girder-stays ; the plate is 
 supported by a bolt passing through a wrought-iron frame with four branches, which 
 rest on the plate at places well supported by the furnace-flues and the sides of the back- 
 connection. 
 
 Flat surfaces of small area are sufficiently stiffened by angle or T-irons riveted to 
 them, without the use of braces. The flat ends of cylindrical sheUs have been stiffened 
 in the same manner, but when their diameter is large an awkward strain is thrown on 
 the rivets attaching the heads to the cylindrical shell ; in such cases it is better to use 
 gusset-plates. These should be secured by double flanges formed either by riveting 
 two angle-irons to the plate or by turning one flange on the plate and riveting an angle- 
 
 iron to the other side ; the rivets attaching the two flanges to the shell should be spaced 
 staggering. It is advantageous to extend the length of the gusset along the shell, and 
 
252 STEAM BOILERS. CHAP. X. 
 
 secure it also to the second belt of plates, and not to the first only, although the latter is 
 the usual practice. 
 
 The gussets which tie the heads to the cylindrical shell of a boiler are often arranged 
 radially, so that the flanges attaching the gussets to the shell may all be bent to a sight 
 angle ; but when the continuation of the gussets on the flat ends forms stay-plates for 
 the attachment of braces (see Plate XII.) it is preferable to place them parallel to each 
 other. 
 
 Where gussets or stay -plates are attached to heating-surfaces, as to the top of back- 
 connections, portions of the flanges between the rivets are cut away, as in figure 124, or 
 the plates are held by lugs riveted to them (see figure 125) ; a better plan is shown on 
 Plate VIII., where thimbles surrounding the rivets, and at least one inch long, are placed 
 between the flanges of the angle-irons and the stayed plate ; the rivets are spaced stag- 
 gering, passing alternately through either of the two flanges. 
 
 The stay -domes which strengthen the front plate of the back-connection of the boiler 
 shown on Plate XII. were formed by pressing the heated plate into a mould, using a 
 spherical shot of suitable diameter as a die, the ends of the plate being held by clamps 
 so as to form a flat flange. These domes were riveted over circular openings cut in the 
 plate of the back-connection. The furnace-tubes and side plates of the back-connection 
 give sufficient stiffness to the plate to support the thrust on the flange of the stay- 
 dome. 
 
 7. Experiments on the Shearing Strength of Wrought-iron Bolts. Ex- 
 periments on shearing wrought-iron bolts, conducted at the Washington Navy- Yard 
 in 1868, by Chief Engineer William H. Shock, U.S.N., gave the results recorded on 
 Plate XX., where the shearing attachments for single and double shear are like- 
 wise illustrated. The Rodman testing-machine was used in making these experi- 
 ments. 
 
 The bolts were of good American commercial iron, not turned. Five sizes of bolts 
 were tested, their diameters being |", f ", f", f ", and 1" ; six specimens of each size were 
 subjected to the single-shear test, and the same number to the double-shear test. The 
 bolts fitted snugly in the respective holes of the shearing attachments, but the latter 
 were made slightly oval, the larger diameter lying in the direction in which the 
 stress was applied. The nuts on the bolts were set up close, but not hard, so as to 
 prevent lateral motion of the two parts of the shearing attachment without producing 
 friction. 
 
 The smaller bolts showed, on the whole, a larger shearing strength per square inch 
 of sectional area than the larger bolts ; but the decrease in strength was not regular or 
 
SBC. 8. STAYS AND BRACES. 253 
 
 uniform. The increase of average strength per square inch of sectional area of the bolts 
 for double-shear over that of single-shear was, for the 
 
 f-inch bolts, 86.2 per cent, 
 f-inch bolts, 97.0 per cent, 
 f-inch bolts, 101.1 per cent, 
 f-inch bolts, 82.6 per cent. 
 1-inch bolts, 85.0 per cent. 
 
 Average for all sizes of bolts, 90.2 per cent. 
 
 In experiments No. 20 and 21 of the double-shear test the nuts were not screwed 
 up close, and the results show a remarkable decrease of strength in comparison with 
 the four other bolts of the same diameter subjected to double-shear. In experiment 
 No. 20 this decrease of strength amounted to 9.1 per cent, of the average result given 
 by the four other bolts, and to 7.1 per cent, of the least result given by the bolts of 
 the same series ; and in experiment No. 21 this decrease of strength amounted to 13.8 
 per cent, and 11.9 per cent, respectively. 
 
 8. Experiments made to determine the proper Dimensions of Pins, Eyes, 
 and Shanks of Boiler-braces. In the course of the years 1878-79 experiments were 
 conducted at the Navy- Yard, Washington, D. C., under the direction of the Bureau of 
 Steam-Engineering, by a board consisting of Chief Engineer Jas. P. Sprague, U.S.N., 
 and Passed Assistant Engineer George E. Tower, U.S.N., the object being to ascertain 
 the proper proportions for the ends of boiler-braces. The Rodman testing-machine 
 represented on Plate I. was used in making these experiments. 
 
 The test-specimens were made in the form of eye-bars, which were secured by accu- 
 rately-fitted iron or steel pins to jaws attached to the testing-machine, so that the pins 
 were subjected to double-shear. The proportions of the eye-bars in each series of expe- 
 riments were gradually changed till the metal at the sides and at the crown of the hole 
 and in the shank of the brace appeared to be very nearly equally strained when rup- 
 ture took place. 
 
 In the first series of experiments the specimens were made of flat iron bars $ inch 
 thick and from 1^ to 1J inches wide. The eyes were formed by drawing out the bar 
 under the hammer, bending and welding it around a mandrel inch less in diameter 
 than the finished hole, then reaming out the hole to fit the pin ; the rest being finished 
 to the proper size in a shaping-machine. The surfaces were planed and finished, and 
 careful examination did not reveal any defect in the welding. -Nevertheless three of 
 the fifteen specimens broke in the weld, and some of the eyes made from the same bar 
 
254 
 
 STEAM BOILERS. 
 
 CHAP. X. 
 
 broke at greatly different strains, indicating that the iron had been injured somewhat 
 in welding. The depth of the eye was in every case equal to the thickness of the bar. 
 
 The following proportions are submitted by the board, "as those which will give 
 nearly a uniform strength in the eye, slightly in excess of that of the shank, suppos- 
 ing the weld to be perfect and the quality not to be materially affected in welding 
 and working ; these proportions will apply until the thickness of the bar is equal to 
 its breadth ; with a steel pin of proper tensile strength its diameter can be reduced to 
 65 per cent, of the breadth of the bar, with the same results," (see figure 126 :) 
 
 Fig. 126. 
 
 Breadth of shank of iron bar = x 
 
 Diameter of iron pin = .917 x 
 
 Width of metal on each side of eye = .600 x 
 Width of metal at crown of eye = .600 x 
 Depth of eye equal to thickness of bar. 
 
 In the second series of experiments the eyes were cut from flat bars, f inch thick 
 and 2f inches wide, witJiout forging. The specimens were planed smooth on both 
 sides to bring them to the proper thickness. The holes were drilled and reamed to fit 
 the pins accurately. The specimens were then put on mandrels and cut out to the re- 
 quired form in the shaping-machine. 
 
 The following proportions are submitted by the board for eye-bars made in this man- 
 Fig. 127. 
 
 i 
 
 ner, as approximating as nearly as possible to a uniform strength in all parts, the depth 
 of the eye being equal to the thickness of the bar. These proportions will apply until 
 
SEC. 9. STAYS AND BRACES. 255 
 
 the thickness of the bar is equal to its breadth. With a steel pin the diameter can be 
 reduced to 66 per cent, of the breadth of the bar. (See figure 127.) 
 
 Breadth of shank of iron bar = x 
 
 Diameter of iron pin = .917 x 
 
 Width of metal on each side of eye = .665 x 
 Width of metal at crown of eye = .722 x 
 Depth of eye equal to thickness of bar. 
 
 A third series of experiments was made with specimens cut from flat bars and hav- 
 ing similar dimensions as the specimens tested in the second series of experiments, but 
 with iron or steel pins of increased diameters. The result showed that, by using iron 
 and steel pins having the respective proportions deduced from the first and second 
 series of experiments, the conditions of strain on the eye-bars are not materially 
 altered. 
 
 A fourth series of experiments was made with eye-bars made of round iron slightly 
 larger than the required size. "The eye was formed (solid) by upsetting the end of the 
 bar and forging to the required shape ; the eye and bar were brought as nearly as pos- 
 sible to given dimensions in the lathe and planer. The hole in the eye was drilled, and 
 the pin made to fit easily but not loose." 
 
 The following proportions are recommended by the Board for eye-bars formed in 
 this manner, the depth of the eye being equal to the diameter of the bar : 
 
 Area of cross-section of shank of round iron bar = if 
 Area of cross-section of iron pin y* 
 
 Area of cross-section of metal on each side of eye = .74 y* 
 Area of cross-section of metal at crown of eye = .90 y* 
 
 (See 'Report on Experiments to ascertain Proportions for the Ends of Boiler-braces.' 
 Washington, D. C., November 24, 1879.) 
 
 9. Experiments on Screw Stay-bolts. For the purpose of determining the 
 strength and holding power of screw stay-bolts for boilers under different conditions, 
 experiments were conducted at the Navy-Yard, Washington, D. C., in the course of the 
 year 1879, under the direction of the Bureau of Steam-Engineering, by a board consist- 
 ing of Chief Engineer James P. Sprague, U.S.N., and Passed Assistant Engineer George 
 E. Tower, U.S.N. 
 
 These experiments were of two kinds. 
 
 In the first place a series of tests was made " to determine the comparative force 
 necessary to pull screw stay-bolts of iron and copper through iron, low-steel, and cop- 
 
256 STEAM BOILERS. CHAP. X 
 
 
 
 per boiler-plates." The Rodman testing-machine illustrated on Plate I. was used in 
 making these tests. 
 
 " Three trials each were first made with " iron plates and 1" iron stay-bolts, not 
 riveted, and riveted over with the ordinary thin or low conical head, simply arranged so as 
 to show the actual strength to resist pulling through the plate, the .supports consisting 
 of heavy plates with a hole If" in diameter ; the boiler-plate resting upon the heavy 
 plate and the stay-bolt adjusted to the centre of the hole, thus allowing the bolt to 
 have a clear space around it equal to the overlapping of the riveted head on the boiler- 
 plate. The bolts not riveted drew out at an average strain of 32,785 pounds ; those 
 riveted with the low conical head, made according to general practice by leaving three 
 threads through to form the head, required an average strain of 35,033 pounds to draw 
 them through the plate ; the rivet-head giving an additional strength of 2,248 pounds in 
 a 1" stay-bolt. 
 
 " In testing those with low conical heads it was observed that the bulging of the plates 
 caused the lap of the rivet-head on the plate to commence giving way or break off 
 some time before the maximum strain was reached, thus leaving more for the threads 
 on the bolts to sustain. As the strain and bulge of the plates increased, the plate 
 around the bolt turned downward and outward until the threads in the plate almost en- 
 tirely cleared those on the bolts, so that in almost every case there were only from one 
 to two threads stripped or injured on the bolt when it drew out ; therefore it was 
 deemed advisable to form the head in a different manner, and, after several experi- 
 ments, it was decided that the rivet- head should be made as follows : First, by leaving 
 as much of the bolt through the plate as could be riveted over without injury to the 
 iron, which was, in case of the excellent iron being used, equal in length to about one- 
 half the diameter of the bolt. This was riveted over in the following manner : A few 
 quick, sharp blows were struck on the end, slightly upsetting the iron ; the head was 
 then formed to shape with a button-head set made to a spherical segment. 
 
 " It was found that this could be done in nearly the same time as that used in riveting 
 the ordinary low conical stay-bolt heads at the Washington yard, and with much less 
 injury to the iron ; also, that it only required one riveter and a helper, whereas by the 
 old method two riveters were used. 
 
 "Three trials each were then made with " iron plates and 1" iron stay-bolts: not 
 riveted ; riveted with ordinary low conical head, with three threads left through for 
 riveting ; riveted with button-head, a little over five threads left through for riveting ; 
 and with button-head, the size of stay-bolt being increased to li". 
 
 "Each end of the stay-bolt was secured, in the manner specified, in the centre of a 
 
SBC. 9. STAYS AND BRACES. 257 
 
 square plate, which was supported by four bolts, one in each corner, by means of which 
 it was held in the testing-machine. These supporting-bolts were placed, in different 
 experiments, four and five inches apart from centre to centre, equally distant from the 
 stay-bolt. 
 
 " The average ultimate strain required to pull the above bolts through the i* plate 
 was as follows : 
 
 WITH SUPPORTS 4* FROM CENTRE TO CENTRE. 
 
 Pounds. 
 
 Y bolt, not riveted 21,970 
 
 1" bolt, ordinary low conical head, three threads left through for riveting 25,147 
 
 \" bolt, button-head ; length of bolt left through for riveting equal to VV diameter 
 
 of bolt 33,791 
 
 li* bolt, button-head ; length left through for riveting equal to diameter of 
 
 bolt 38,885 
 
 WITH SUPPORTS 5* FROM CENTRE TO CENTRE. 
 
 1" bolt, ordinary low conical head 22,137 
 
 1" bolt, button-head ; length left through for riveting equal to ft diameter of 
 
 bolt 31,282 
 
 If bolt, button-head ; length left through for riveting equal to i diameter of 
 
 bolt 35,812 
 
 "The great increase of holding power given to screw stay-bolts by forming the 
 riveted head with a button-head set being demonstrated by these tests, further experi- 
 ments were made with iron, steel, and copper plates and stay-bolts secured in the same 
 manner, for the purpose of determining the best proportions of diameter of stay-bolts 
 to thickness of plate under different conditions 
 
 -" In comparing the results of three different thicknesses in each case (f, f, i* 
 plate) of iron plates and iron bolts, steel plates and iron bolts, steel plates and steel 
 bolts, the diameter of the bolts being 1", li", and If, their distance apart and condi- 
 tions of trial being the same, it was found that in the case of the iron plates and iron 
 bolts the strain required to draw the bolts through the plates was equal to 74.77 per 
 cent, of the tensile strength of the bolt, with the steel plates and iron bolts 77.36 per 
 cent., and with the steel plate and steel bolts 85.42 per cent." 
 
 In the next series of experiments the plates and stay-bolts were arranged so as to 
 represent a portion of the wall of a fire-box, and water-pressure was used to produce 
 the strain. 
 
258 STEAM BOILERS. CHAP. X. 
 
 Iron, steel, and copper plates were used, varying from J inch to \ inch in thickness. 
 The iron and steel screw stay-bolts were spaced from 4 inches to 8 inches apart, and, 
 after being screwed through the plates, their ends were secured either by riveting them 
 over with a button-head set or by means of nuts and washers. 
 
 The Board conclude their report on these experiments with the following recommen- 
 dations, viz. : 
 
 " After a careful examination of the results of these experiments in particular we 
 are satisfied that the following formulae will correctly and safely represent the working 
 strength of good material in flat surfaces, supported by screw stay-bolts with riveted 
 button-shaped heads or with nuts, when the thickness of the plates forming said sur- 
 faces and the screw stay-bolts are made in accordance with the dimensions and con- 
 ditions given in Table Y. W safe working pressure ; T thickness of plate ; 
 d distance from centre to centre of stay-bolt : 
 
 rjrt 
 
 For iron plates and iron bolts W 24,000 -- 
 
 rjn 
 
 For low-steel plates and iron bolts W 25,000 -~- 
 
 yn 
 
 For low-steel plates and low-steel bolts W= 28,000 - ~- 
 
 yn 
 
 For iron plates and iron bolts, with nuts W 40,000 -^ 
 
 JTJ 
 
 For copper plates and iron bolts W ' 14,500 -^ 
 
 " To obtain the ultimate bursting pressure multiply the results of the above formulae 
 by 8, which is the factor of safety iised. 
 
 " The rivet-heads to be a segment of a sphere, formed by first upsetting the end of 
 the bolt with a few quick, sharp blows of the hammer, then finished to shape with the 
 hammer and button-head set. Where nuts can be used instead of riveted heads they 
 should be of the standard size, suited to the diameter of the bolt, faced on the side 
 bearing on the plate, and dished out so as to form an annular bearing-surface of as 
 large a diameter as the nut will allow, and of a breadth and depth given in the table. 
 Before securing the nut in place the dished portion should be filled with red-lead putty 
 made stiff with fine iron borings." 
 
SEC. 9. 
 
 STAYS AND BKACES. 
 
 259 
 
 TABLE Y. 
 
 DIMENSIONS AND CONDITIONS FOR MAKING IRON AND LOW-STEEL SCREW STAY-BOLTS FOR FLAT 
 SURFACES SUBJECT TO INTERNAL PRESSURE FOR DISTANCES RANGING FROM FOUR TO EIGHT 
 INCHES (INCLUSIVE) FROM CENTRE TO CENTRE OF STAY-BOLTS. 
 
 Thickness of 
 plate. 
 
 Diameter of 
 bolt outside of 
 thread. 
 
 Number of 
 threads per 
 inch. 
 
 Length of bolt 
 left through for 
 riveting in 
 fractions of 
 diameter 
 of bolt. 
 
 Height of 
 rivet-head 
 when 
 finished. 
 
 Diameter of 
 base of rivet- 
 head not to 
 exceed when 
 finished 
 
 Nuts. 
 
 Breadth of 
 annular 
 bearing-surface. 
 
 Dished out to 
 a depth of 
 
 i" 
 
 i" 
 
 u 
 
 i 
 
 TV" 
 
 ITV" 
 
 TV" 
 
 TV" 
 
 i" 
 
 if 
 
 J 4 
 
 i 
 
 i" 
 
 *" 
 
 i" 
 
 TV" 
 
 }" 
 
 if" 
 
 12 
 
 i 
 
 TV" 
 
 if" 
 
 TV" 
 
 A" 
 
 i" 
 
 if" 
 
 12 
 
 i 
 
 i" 
 
 4" 
 
 i" 
 
 A" 
 
 
CHAPTER XI. 
 
 FLUES AND TUBES. 
 
 1. Flue-boilers. The flues or channels for the passage of the products of combus- 
 tion from the furnace to the chimney were at first made very large in marine boilers, so 
 as to give easy access for cleaning and repairs, and, in order to get a great amount of 
 heating-surface, these passages were often made very tortuous. While the pressures 
 of steam used in marine boilers exceeded but little the atmospheric pressure the flues 
 were frequently made with flat sides (see figure 1, Plate XXI.) ; but with increased 
 steam-pressures flues of a circular cross-section have come into general use. The longi- 
 tudinal seams of these flues are either lap-welded or riveted. Stationary flue-boilers 
 are frequently made very long, with one or two large circular or elliptical flues running 
 through the whole length of the boiler. The length of marine boilers being limited by 
 the available space, the required amount of heating-surface is obtained in them by 
 using return-flues, and by decreasing their diameter and increasing their number (see 
 figure 2, Plate XXI.) 
 
 In the drop-flue boiler (see figure 3, Plate XXI.) the products of combustion pursue 
 a downward course in their passage from the furnace to the chimney ; by this arrange- 
 ment the cooler gases are brought in contact with surfaces surrounded by the less 
 heated water at the bottom of the boiler, and consequently part more readily with 
 their heat. 
 
 The efficiency of stationary flue-boilers has been greatly increased by the introduc- 
 tion of the so-called Galloway tubes / these are conical tubes placed with the larger end 
 uppermost across the flues, sometimes slightly inclined. Besides furnishing additional 
 very efficient heating-surface they facilitate greatly the circulation of the water and act 
 as stiff stays ; in this latter capacity they are particularly useful when the flues have 
 an elliptical cross-section. The introduction of these tubes into old flue-boilers has 
 often produced a remarkable improvement in their steaming capacity ; but, on the 
 other hand, they make the flues more difficult to clean and repair, so that the accumu- 
 lation of soot and dirt may actually cause a diminution of the efficiency of the heating- 
 surface, while, at the same time, the obstructions to the draught diminish the rate of 
 combustion. Similar tubes are sometimes placed in the back-connections of marine 
 
 260 
 
SEC. 2. FLUES ASD TUBES. 261 
 
 boilers, especially when there is one back-connection common to all the furnaces of a 
 boiler. 
 
 Flue-boilers are bulky and heavy, and the large amount of water contained in them 
 makes it impossible to get up steam quickly. To remedy these defects, which are 
 often of vital importance in a marine boiler, it became necessary to reduce the length of 
 the flues, and to increase the proportion of their superficial area to their cross-area by 
 subdividing each flue into a number of narrow passages. 
 
 In the Lamb and Sumner boiler the flues returned over the furnaces and consisted 
 of a number of narrow, flat-sided passages, separated by equally narrow water-spaces, 
 from If to 2 inches wide in the clear, and from 36 to 45 inches high. The flat sides 
 were held by stay-rivets passing through the smoke-passages. These boilers were in 
 great favor some years ago ; the flat sides of the water-spaces were easily scaled, and 
 the flues were not so soon obstructed by soot as small tubes ; but the narrow passages 
 were inaccessible for repairs in case of leaks, and corrosion destroyed them rapidly. 
 These boilers have now gone out of use. 
 
 At the present day it is the nearly universal practice to get the principal quantity of 
 heating-surface in marine boilers by the use of cylindrical tubes varying from 2 to 4 
 inches in diameter. 
 
 2. Relative Advantages of Flues and Tubes for Marine Boilers. The prin 
 cipal advantages possessed by tubular over flue boilers may be shortly summed up as 
 follows : Less weight and space is required for boilers of equal economic and potential 
 evaporative efficiency ; steam can be raised rapidly after the fires are started, in conse- 
 quence of the relatively small weight of water contained in the tubular boiler in pro- 
 portion to the extent of heating-surface ; the small tubes have far greater strength than 
 flues ; the tubes are less liable to leakage from the absence of riveted joints ; they can 
 be made of material not liable to corrosion, and are easily removed and replaced ; the 
 escape of steam or water from a ruptured tube seldom produces serious effects, and the 
 leak can often be temporarily stopped without interrupting the working of the boiler. 
 
 On the other hand, crowding the numerous tubes into a narrow space, and the rapid 
 formation of steam on their surfaces, often cause foaming or priming, affecting very un- 
 favorably the economic performance of the engine ; a large portion of the heating- 
 surfaces is inaccessible for cleaning, so that the accumulation of scale and other foreign 
 matter soon impairs the evaporative efficiency of the surfaces, and causes the destruc- 
 tion of the boiler by corrosion or the burning of the metal. 
 
 3. Various Types of Tubular Boilers. There exists great variety in the arrange- 
 ment of tubes within the boiler : the hot gases pass either through them or around 
 
202 STEAM BOILERS. CHAP. XI. 
 
 them ; tubes may be horizontal, vertical, or inclined, and may be arranged above, be- 
 hind, or at the sides of the furnaces ; they are generally straight, but bent and spiral 
 tubes are employed in some types of boilers. Examples of different arrangements of 
 tubes in marine boilers have been given and their advantages and disadvantages dis- 
 cussed in chapter vii. 
 
 The considerations governing the location and position of the tubes in marine boilers 
 may be summed up as follows : When the room in the length and breadth of the vessel 
 available for the boilers is limited the tubes must be arranged directly over the fur- 
 naces ; when, on the contrary, it is essential to keep the boilers as low as possible, the 
 tubes have to be arranged behind or alongside the furnaces. Every change in the 
 direction of the current of the hot gases passing from the furnace to the chimney in- 
 volves a loss of head, or, in other words, diminishes the draught, and consequently the 
 maximum rate of combustion. The resistance to the flow of gases through a tube is 
 mainly due to friction against its inner surface ; the resistance to the flow of gases 
 'between a nest of water- tubes is relatively much greater, being due to friction against 
 the outer surfaces of the tubes, to the loss of head produced by the successive changes 
 in the cross-area of the passages, and to the counter-currents caused by the impinge- 
 ment of the gases on the tubes. The evaporative efficiency of the vertical water-tube is 
 superior to that of the fire-tube, because the hot gases impinging on the surface of the 
 former part with their heat much more readily than the gases which move in a direction 
 parallel to the axis of the fire-tube. The evaporative efficiency of vertical fire-tubes 
 is inferior to that of horizontal fire-tubes, because the steam escapes more readily from 
 the most efficient portion i.e., the top of the latter, and the tendency to an equali- 
 zation of the temperature of the mass of gases by convection is greater in the horizontal 
 than in the vertical fire-tube ; for the gases which are cooled down by contact with the 
 upper portions of the surface of the ho'rizontal tube sink by gravity and are replaced 
 by hotter gases, while in the vertical tube the gases occupying the central part of the 
 tube are likely to pass through the tube without coming in contact with its sides. 
 Horizontal water-tubes are inefficient and dangerous with a rapid evaporation, on ac- 
 count of the difficulty which the steam experiences in escaping from them. Externally- 
 heated horizontal tubes can, however, be used safely for the purpose of drying or super- 
 heating steam. Scale is easily removed from the inner surface of water-tubes, but fire- 
 tubes are more easily swept of soot and ashes. With vertical tubes the water-level can 
 be safely carried below the upper tube-sheet, while horizontal tubes are quickly de- 
 stroyed when the upper rows are bared, of water. 
 
 4. Dimensions and Spacing of Tubes. The width of the space available for 
 
SEC. 4. FLUES AND TUBES. 263 
 
 each nest of tubes is generally limited by the width of the furnace ; and its height and 
 the length of the tubes are dependent not only on the dimensions of the shell of the 
 boiler, but also on economical considerations. In proportioning the dimensions of the 
 tubes and their spacing the following conditions must be kept in view : 
 
 First. The opening through or between the tubes must be sufficient for the passage 
 of the products of combustion ; the calorimeter determines to a great extent the rate of 
 combustion in the furnace, and varies in marine boilers from | to of the area of the 
 grate-surface. 
 
 Second. The tubes must present a sufficient amount of heating-surface. The ratio 
 of the heating-surface to the rate of combustion affects the economic and potential eva- 
 porative efficiency of the boiler ; and the evaporative efficiency of the tube-surface 
 decreases rapidly from the end where the gases enter to where they are discharged into 
 the uptake. In the ordinary types of return-tubular boilers, in which the total heating- 
 surface is equal to 25 times the grate-surface, the proportion of the tube-surface to the 
 grate-surface is very nearly as 18 to 1. The heating-surface of a tube is to be calculated 
 for the side in contact with the hot gases ; therefore it depends on the outer circum- 
 ference of a water-tube, and on the inner circumference of a fire-tube. 
 
 Third. The spaces between the tubes must be arranged with regard to facility for 
 scaling and cleaning, and to the free escape of the steam as soon as formed. The upper 
 rows of horizontal fire-tubes and the upper part of vertical tubes are surrounded or 
 filled by the mass of steam-bubbles rising from the lower heating-surfaces, and are on 
 this account less efficient as heating-surfaces. Isherwood found by experiment that 
 the gases emerging from the upper rows of tubes were often nearly 300 Fahr. hotter 
 than those leaving the lower rows of tubes. For this reason it is advantageous to make 
 the nest of tubes as low as possible. 
 
 The outside diameter of the vertical water- tubes in the Martin boiler (see Plate VI.) 
 is generally 2 inches, and they are spaced from 3 inches to 3J inches apart from cen- 
 tre to centre on a line across the tube-box ; when the clear space between two adjoining 
 tubes is made less than one inch the draught of the boiler becomes seriously impaired. 
 On a line in the direction of the length of the tube-box, the closeness of the spacing of 
 the tubes is limited only by the possibility of boring the holes in the tube-plates without 
 impairing the stiffness of the plates too much ; with 2-inch tubes the distance between 
 the centres is seldom less than 2|| inches. The length of the tubes depends on the 
 amount of heating-surface required, but is limited by the height of the boiler; the 
 evaporative efficiency of short tubes is greater than that of longer tubes having the 
 same diameter and presenting an equal amount of heating-surface. 
 
264 STEAM BOILERS. CHAP. XI. 
 
 In the boiler represented on Plate VI. the tubes are 2 inches in diameter and 32 
 inches long between the tube-sheets ; they are spaced 3J inches apart from centre to 
 centre across the tube-boxes, and 2f inches apart from centre to centre in the rows 
 running lengthwise the tube-boxes. Each tube-box is of the same width as the fur- 
 naces viz., 36 inches and is 85| inches long, containing in this space 306 brass tubes 
 with an aggregate heating-surface of 427.25 square feet. 
 
 These tubes have sometimes been arranged so that the longitudinal rows formed 
 zigzag lines, in order that the gases might impinge on a greater amount of surface ; 
 but what is gained in evaporative efficiency of tube-surface in such a case is lost in 
 the rate of combustion. The most serious objection to this arrangement is the impossi- 
 bility of sweeping the spaces between the tubes properly. 
 
 The diameter of fire-tubes in marine boilers varies ordinarily between 2f inches 
 and 4 inches. Horizontal tubes of smaller diameter than 2 inches would soon become 
 choked with ashes and soot, unless forced draiight is used, as in locomotives, when the 
 diameter is reduced sometimes to 1 inches. Vertical fire-tubes may be made of smaller 
 diameter than horizontal tubes, since they are not obstructed by soot and ashes like the 
 latter. When bituminous coal is the fuel used tubes.of larger diameter become necessary 
 than when anthracite is burned, on account of the great quantity of soot produced by 
 the former coal. Large tubes offer less resistance to the flow of the gases, and admit of 
 a higher rate of combustion with natural draught, than smaller ones ; but with the latter 
 a larger amount of heating-surface can be got in a given space. When the diameter of 
 the fire-tube is increased it becomes necessary to increase its length in the same propor- 
 tion, in order to preserve the same ratio between the surface and the cross-area of the 
 tube, or, in other words, between the heating-surface and the quantity of gas passing 
 through each tube. 
 
 Experiments by Dewrance, Woods, and C. W. Williams demonstrated the rapid 
 decrease in the evaporative efficiency of each additional length of tubes. After a cer- 
 tain limit is reached the gain in the evaporative efficiency of a tube through an increase 
 of its length, and consequently of its surface, is trifling compared with the additional 
 bulk, weight, and cost of the boiler, while the additional friction retards somewhat the 
 draught. Since for equal economic evaporative efficiency the amount of heating-sur- 
 face must be proportioned to the quantity of hot gases in contact with it, the length of 
 the tube must depend on the diameter of the tube and the rate of combustion. In loco- 
 motive boilers, with forced draught, the length of tubes is made often 120 times their 
 diameter. Wilson recommends that with natural draught the length of tubes should 
 not be greater than 24 times their diameter. This is, however, less than the usual prac- 
 
SEC. 5. FLUES AND TUBES. 265 
 
 tice, and Isherwood says : " For a tube 3 inches diameter a length of 38 diameters will 
 be found a good proportion with a rate of combustion exceeding 12 Ibs. of anthracite 
 per hour in the hold of a vessel." 
 
 The clear space between adjoining horizontal fire-tubes varies between one-half and 
 one-third the diameter of the tubes. To facilitate the washing and the scaling between 
 the tubes, as well as the escape of the steam-bubbles as soon as generated, horizontal 
 tubes in marine boilers should always be arranged in vertical rows, and not in diagonal 
 or zigzag rows, which is sometimes done for the purpose of crowding a greater number 
 of tubes into a given space. In locomotive boilers, where fresh water is used and the 
 steam generated on the furnace-crown has not to pass between the rows of tubes, the 
 latter are frequently arranged in zigzag lines. 
 
 In Skime? s differential tubular boiler the horizontal fire- tubes in each successive hori- 
 zontal row, from the bottom upwards, decreased in diameter \ inch. The boilers of U. S. 
 S. Tippecanoe and class had eight horizontal rows of tubes over each furnace ; the out- 
 side diameter of the tubes was 3 inches in the bottom row and 2f inches in the top row ; 
 the spacing of the tubes in a horizontal direction was uniform viz., 4f inches from 
 centre to centre. This arrangement facilitates greatly the escape of steam, and to some 
 extent the scaling of the tubes, and was also intended to equalize the evaporative effi- 
 ciency of the heating-surface in each horizontal row of tubes ; but its practical disad- 
 vantages are great viz., to form the holes eight sizes of drills or eight adjustments of 
 the boring-cutters are required ; eight sizes of tubes are used, and three or four sizes of 
 expanding-tools ; spare tubes and expanding-tools of assorted sizes have to be carried, 
 and tube-brushes of assorted sizes are used. With these drawbacks it is not strange 
 that the system has not come into favor. 
 
 When a large amount of heating-surface is required it is often difficult to propor- 
 tion the number and dimensions of the tubes in such a manner in the given space as to 
 get at the same time the best ratio of calorimeter to grate-surface. In such a case the 
 calorimeter of fire-tubes may be reduced by driving ferrules into the ends of the tubes, 
 at the back end of return-tube boilers, and at the front end of locomotive boilers. In 
 the latter boilers, when iron tubes are used, their diameter is sometimes reduced at the 
 front end by swaging, as illustrated on Plate XXIV. When the opening through the 
 tubes is contracted by these means the sweeping of the tubes is made more difficult ; 
 the effect of the ferrules in increasing the holding power of the tubes in the tube-plates 
 will be discussed further on. 
 
 5. Iron, Steel, Brass, and Copper Tubes. The tubes of marine boilers are 
 generally made of brass or iron ; steel tubes have also been introduced of late. Copper 
 
266 STEAM BOILERS. CHAP. XI. 
 
 tubes were formerly used for locomotive boilers, but they wore out rapidly in conse- 
 quence of the mechanical action of the cinders carried through them at a great velocity 
 by the strong draught ; the great difference in the expansion of iron and copper by 
 heat produces also inconveniences in the combination of the two metals in steam 
 boilers. The use of copper tubes in marine boilers is prevented by the lively galvanic 
 action which takes place when copper is in contact with iron in the presence of salt 
 water ; they are used, however, sometimes in the steam-space as superheating-tubes, for 
 which purpose the great thermal conductivity of copper and its freedom from corrosion 
 give them great advantages. 
 
 Lap-welded wrought-iron tubes are still extensively used in boilers of merchant ves- 
 sels, but seamless drawn-brass tubes are now almost exclusively used in the boilers of 
 naval vessels. Brass tubes possess many of the advantages of copper tubes, without 
 their disadvantages : they are very ductile, their thermal conductivity is greater than 
 that of iron, and they expand less under the influence of heat than copper ; they suffer 
 little from wear, are not subject to corrosion, and do not appear to produce serious gal- 
 vanic action in marine boilers. Iron tubes have the advantage of lower first cost over 
 brass tubes. Since iron tubes are not so easily injured by biirning as brass tubes when 
 the water-level falls below the tubes, a few of them are often used as stay-tubes in hori- 
 zontal-tubular boilers to hold the tube-sheets, while the principal portion of the tubes 
 is of brass. For vertical water-tubes brass should always be used, because iron tubes 
 are rapidly corroded by the sulphuric acid which is distilled from the soot by moisture, 
 runs down the tubes, and collects at their lower end. 
 
 With vertical brass tubes the water-level can be carried safely below the upper tube- 
 plate, as was proved by several interesting experiments conducted by Chief Engineer 
 Isherwood, U.S.N., and recorded in ' Experimental Researches,' vol. ii. In the trial of 
 the vertical water-tube boiler of the U. S. S. Wyoming the water-level was carried, at 
 different times, 7f, 15, and 22 inches below the top tube-plate. " The whole length of 
 the tubes, which were of seamless brass, was 30 inches. The only damage done was 
 when the water was carried 22f inches below the top tube-plate, and then, after a trial 
 of 72 consecutive hours, burning at the rate of 15.77 Ibs. of anthracite per square foot 
 of grate per hour, the joints of only the two rows of tubes next the back smoke-con- 
 nection were loosened. Neither the brass of the tubes nor the iron of the tube-plate 
 and of the exposed portions of the back smoke-connections was in the least degree 
 injured." 
 
 Iron tubes are more apt to be injured than brass tubes in the process of securing 
 them in the tube-sheets by expanding their ends, and leakage with its attendant evils ia 
 
SEC. 5. FLUES AND TUBES. 267 
 
 more frequent with them. Iron tnbes, being relatively thin, are destroyed by corrosion 
 much sooner than the plates of boilers. The thickness of boiler-tnbes varies from No. 
 8 to 14 of the Birmingham gauge, according to their diameter, the material of which 
 they are made, the steam-pressure and the kind of pressure, bursting or collapsing, to 
 which they are exposed. 
 
 Coppered and tinned iron tubes have been tried, but they have not been long enough 
 in use to warrant an opinion regarding their durability. 
 
 Tubes made of soft steel are either lap-welded or drawn seamless. They are more 
 expensive than iron tubes, and whether they can be made lighter than the latter de- 
 pends principally on their liability to corrosion, regarding which fact opinions are much 
 divided. 
 
 When the tubes are new and clean their thickness and the thermal conductivity of 
 the metal of which they are composed exert, no doubt, an influence on their evapora- 
 tive efficiency ; but as soon as their surfaces become covered with deposits of soot and 
 scale they become of equal value in this respect. 
 
 Drawn seamless tubes are made from short cylinders through which a small hole has 
 been bored or left in casting. The cylinder, being passed through several sets of dies 
 very slightly decreasing in size, and over mandrels very gradually increasing in diame- 
 ter, is, by a succession of steps and, in the case of steel tubes, after frequently under- 
 going the annealing process, drawn out into a long, thin tube of the desired dimen- 
 sions. The process is not severe as long as the dies are in good order, but when they 
 are in the least degree rough great heat is evolved in the passage of the metal. An ob- 
 jection to the method in the case of brass is that the smallest defect is drawn out into a 
 scratch which becomes a source of weakness. Latent defects, however, can be dis- 
 covered in the testing which all tubes should undergo at the works ; this test should 
 have reference to the use to which the tubes are to be put those intended for water- 
 tubes should withstand an internal bursting pressure, while those intended for fire-tubes 
 must be able to resist collapse. 
 
 Copper and brass tubes were at first made of long, narrow sheets bent into the cylin- 
 drical form and brazed at the joint. Wrought-iron ones are now constructed of narrow 
 sheets brought to a welding-heat and passed through rollers. Drawn tubes are gene- 
 rally slightly conical to permit their being easily delivered from the mandrel ; and this 
 feature has been exaggerated by some inventors, who make the outside diameters of 
 tubes which are intended to be removable for the purpose of scaling i inch larger at the 
 front than at the back end. 
 
268 
 
 STEAM BOILERS. 
 
 CHAP. XI. 
 
 TABLE XXXIII. 
 
 SIZES AND WEIGHTS OF LAP- WELDED IRON BOILER-TUBES OF STANDARD GAUGE MANUFACTURED 
 
 BY THE NATIONAL TUBE WORKS COMPANY. 
 
 Outside diameter. 
 Inches. 
 
 Thickness. 
 
 Weight per foot. 
 Pounds. 
 
 Birmingham wire-gauge. 
 
 Inches. 
 
 4 
 
 14 
 
 0875 
 
 1-25 
 
 if 
 
 13 
 
 .1000 
 
 1. 60 
 
 2 
 
 13 
 
 .1000 
 
 2.OO 
 
 i 
 
 13 
 
 .1000 
 
 2.IO 
 
 4 
 
 12 
 
 .1125 
 
 2-75 
 
 4 
 
 12 
 
 .1125 
 
 3.00 
 
 3 
 
 12 
 
 .1125 
 
 3-33 
 
 3 
 
 II 
 
 .1256 
 
 4.00 
 
 3* 
 
 II 
 
 .1250 
 
 4-33 
 
 3f 
 
 II 
 
 .1250 
 
 4-63 
 
 4 
 
 IO 
 
 .1406 
 
 5-5 
 
 4 
 
 IO 
 
 .1406 
 
 6.00 
 
 5 
 
 9 
 
 1563 
 
 7-25 
 
 6 
 
 8 
 
 .1719 
 
 9-33 
 
 7 
 
 8 
 
 .1719 
 
 12.50 
 
 8 
 
 8 
 
 .1719 
 
 15.00 
 
 9 
 
 7 
 
 1875 
 
 i7-3i 
 
 10 
 
 6 
 
 .2031 
 
 20.80 
 
 These tubes are made of the best American charcoal hammered iron and are stamped 
 The greatest regular length of these tubes is 18 feet ; lap-welded tubes of any 
 thickness and size up to 18 inches diameter, and of a length exceeding 18 feet, manu- 
 factured to order. 
 
 Lap-welded steel boiler-tubes are made from 1 to 18 inches in diameter. 
 
SEC. 5. 
 
 FLUES AND TUBES. 
 
 269 
 
 TABLE XXXIV. 
 
 LAP-WELDED AMERICAN CHARCOAL-IRON BOILER-TUBES MANUFACTURED 
 BY MORRIS, TASKER & Co., 1877. 
 
 Table of Standard Dimensions. 
 
 External 
 
 diameter. 
 
 Inches. 
 
 Standard 
 thickness. 
 
 Inches. 
 
 Nearest 
 B. W. G. 
 
 Internal 
 circumference. 
 
 Inches. 
 
 External 
 circumference. 
 
 Inches. 
 
 Internal area of 
 cross- section. 
 
 Square inches. 
 
 External area of Weight 
 cross-section. per foot. 
 
 Square inches, j Pounds. 
 
 I 
 
 .072 
 
 15 
 
 2 689 
 
 3-142 
 
 0-575 
 
 0.785 
 
 0.708 
 
 if 
 
 .072 
 
 15 
 
 3-474 
 
 3-9 2 7 
 
 0.960 
 
 1.227 
 
 0.900 
 
 4 
 
 .083 
 
 14 
 
 4.191 
 
 4.712 
 
 1.396 
 
 1.767 
 
 1.250 
 
 if 
 
 95 
 
 13 
 
 4.901 
 
 5-49 8 
 
 1.911 
 
 2.405 
 
 1.665 
 
 2 
 
 .098 
 
 13 
 
 5.667 
 
 6.283 
 
 2-556 
 
 3-142 
 
 1.981 
 
 i 
 
 .098 
 
 13 
 
 6.484 
 
 7.069 
 
 3-3 '4 
 
 3-976 
 
 2.238 
 
 4 
 
 .IO9 
 
 12 
 
 7.172 
 
 7-854 
 
 4.094 
 
 4.909 
 
 2-755 
 
 4 
 
 .109 
 
 12 
 
 7-957 
 
 8.639 
 
 5-39 
 
 5-940 
 
 3-045 
 
 3 
 
 .109 
 
 12 
 
 8-743 
 
 9-425 
 
 6.083 
 
 7.069 
 
 3-333 
 
 3i 
 
 .II 9 
 
 II 
 
 9.462 
 
 IO.2IO 
 
 7-125 
 
 8.296 
 
 3-958 
 
 - 3i 
 
 .119 
 
 II 
 
 10.248 
 
 10-995 
 
 8-357 
 
 9.621 
 
 4-272 
 
 3* 
 
 .119 
 
 II 
 
 ".033 
 
 IIjSl 
 
 9.687 
 
 11.045 
 
 4-59 
 
 4 
 
 130 
 
 10 
 
 "-753 
 
 12.566 
 
 10.992 
 
 12.566 
 
 5-320 
 
 4 
 
 .130 
 
 10 
 
 i3-3 2 3 
 
 14-137 
 
 14.126 
 
 I5-904 
 
 6.010 
 
 5 
 
 .I4O 
 
 9i 
 
 14.818 
 
 15-708 
 
 17-497 
 
 I9-635 
 
 7.226 
 
 6 
 
 IS 1 
 
 9 
 
 17.904 
 
 18.849 
 
 25-509 
 
 28.274 
 
 9-346 
 
 7 
 
 .172 
 
 8* 
 
 20.914 
 
 21.991 
 
 34-8o5 
 
 38.484 
 
 12.435 
 
 8 
 
 .182 
 
 8 
 
 23-9 8 9 
 
 25- T 32 
 
 45-795 
 
 50.265 
 
 15.109 
 
 9 
 
 193 
 
 7f 
 
 27-055 
 
 28.274 
 
 58.291 
 
 63.617 
 
 18.002 
 
 10 
 
 .214 
 
 4 
 
 30.074 
 
 31.416 
 
 71-975 
 
 78.540 
 
 22.190 
 
 ii 
 
 .220 
 
 5 
 
 33- '75 
 
 34-557 
 
 87-479 
 
 95-033 
 
 25.489 
 
 12 
 
 .229 
 
 4* 
 
 36.260 
 
 37-699 
 
 103.749 
 
 113.097 
 
 28.516 
 
 !3 
 
 .238 
 
 4 
 
 39-345 
 
 40.840 
 
 123.187 
 
 132.732 
 
 32.208 
 
 14 
 
 .248 
 
 3i 
 
 42.414 
 
 43-982 
 
 143.189 
 
 I53.938 
 
 36.271 
 
 15 
 
 259 
 
 3 
 
 4S-49 6 
 
 47.124 
 
 164.718 
 
 176.715 
 
 40.612 
 
 16 
 
 .271 
 
 *i 
 
 48.562 
 
 50.265 
 
 187.667 
 
 201.062 
 
 45-!99 
 
 i? 
 
 .284 
 
 2 
 
 51.662 53-407 
 
 212.227 
 
 226.980 
 
 49.902 
 
 18 
 
 .292 
 
 4 
 
 54.714 56.548 238.224 
 
 254.469 
 
 54.816 
 
 J 9 
 
 .300 
 
 i 
 
 57-8o5 59- 6 9 265.903 
 
 283.529 
 
 59-479 
 
 20 
 
 .320 
 
 i 
 
 60.821 62.832 294.373 
 
 3i4- I 59 
 
 66.765 
 
 21 
 
 340 
 
 O 
 
 63-837 65.973 324.311 
 
 346.361 
 
 73-404 
 
 The thickness of tubes can be varied to order. Tubes cut to specific lengths to suit purchasers ; lengths greater 
 than eighteen feet at special rates. 
 
270 
 
 STEAM BOILERS. 
 
 CHAP. XT. 
 
 TABLE XXXV. 
 
 REGULAR SIZES AND WEIGHTS OF SEAMLESS DRAWN BRASS AND COPPER TUBES MANUFACTURED 
 
 BY AMERICAN TUBE-WORKS, BOSTON, MASS., 1879. 
 
 Outside 
 diameter. 
 
 Length. 
 
 Thickness 
 
 Weight per foot. 
 
 Outside 
 diameter. 
 
 Length. 
 
 Thickness 
 
 Weight per foot. 
 
 Inches. 
 
 Feet. 
 
 Stub's W. G 
 
 Pounds. 
 
 Inches. 
 
 Feet. 
 
 Stub's W.G 
 
 Pounds. 
 
 1 
 
 
 
 Brass. 
 
 Copper. 
 
 
 
 
 Brass. 
 
 Copper. 
 
 f 
 
 12 
 
 18 
 
 0.32 
 
 0-34 
 
 2 t 
 
 * 
 
 j 14 
 
 i-97 
 
 2.07 
 
 
 12 
 
 17 
 
 0.47 
 
 0.49 
 
 8 
 
 
 ( I2 
 
 2-55 
 
 2.68 
 
 T5 
 
 IO 
 
 17 
 
 0.50 
 
 o-53 
 
 j_ 
 
 T A 
 
 \ 14 
 
 2.08 
 
 2.19 
 
 tt 
 
 10 
 
 IO 
 IO 
 
 17 
 
 16 
 16 
 
 -55 
 0.64 
 0.70 
 
 0.58 
 0.66 
 0.74 
 
 1 
 
 4 
 
 13 
 
 ( 12 
 
 I' 4 
 / 12 
 
 2.71 
 
 2.20 
 2.86 
 
 2.85 
 2-32 
 3-oi 
 
 1 i 
 
 IO 
 
 16 
 
 0.79 
 
 0.83 
 
 2 i 
 
 1 1 
 
 . 13 
 
 2.65 
 
 2-79 
 
 1 1. 
 
 15 
 
 j 14 
 
 1. 12 
 
 1.18 
 
 
 1 o 
 
 II 
 
 3-31 
 
 3-48 
 
 
 
 ( 12 
 
 f 
 
 1.44 
 
 i-5 2 
 
 2 i 
 
 I 2 
 
 13 
 
 2. 7 8 
 
 2-93 
 
 if 
 
 12 
 
 13 
 12 
 
 14 
 
 I 12 
 
 14 
 
 1 12 
 
 14 
 
 i- 2 5 
 1.60 
 
 1.36 
 1.76 
 1.48 
 
 1.68 
 
 i-43 
 1.85 
 
 1-56 
 
 3 
 
 12 
 12 
 
 II 
 
 I' 3 
 i II 
 
 I 13 
 j II 
 
 3-49 
 2-93 
 3.66 
 3.20 
 4.01 
 
 3-08 
 3-85 
 3-37 
 4.22 
 
 
 
 ( 12 
 
 1.92 
 
 2 02 
 
 i 
 
 IO 
 
 13 
 
 3-34 
 
 3-5 1 
 
 if 
 
 13 
 13 
 12 
 
 14 
 
 ( 12 
 
 14 
 
 ( I 2 
 
 1.61 
 2.07 
 1.66 
 
 2-15 
 1.72 
 
 1.69 
 2.18 
 
 i-75 
 2.26 
 1.81 
 
 3i 
 3* 
 
 10 
 IO 
 
 I'" 
 1 II 
 
 I' 3 
 j II 
 
 4.18 
 3-48 
 4-35 
 3-75 
 4.70 
 
 4.40 
 3.66 
 4-58 
 3-95 
 4-95 
 
 iff 
 
 12 
 
 . 14 
 
 2.23 
 1.78 
 
 2-35 
 1.87 
 
 4 
 
 10 
 
 \\\ 
 
 4-3 
 5-4 
 
 4-53 
 5.68 
 
 2 
 
 15 
 
 12 
 
 2.31 
 1.84 
 
 2-43 
 i-94 
 
 5 
 
 10 
 
 (12 
 \ 10 
 
 6.18 
 7-56 
 
 6.50 
 7.96 
 
 
 
 12 
 
 2-39 
 
 2.51 
 
 6 
 
 IO 
 
 I- 
 
 7-44 
 
 7-83 
 
 
 
 
 
 
 
 
 \ 10 
 
 9.11 
 
 9-59 
 
SEC. 6. 
 
 FLUES AND TUBES. 
 
 271 
 
 These tubes are polished, both inside and outside. They are perfectly cylindrical 
 on the outside ; the bore is a gradual taper, the difference in diameter being two wire- 
 gauges in eleven feet. 
 
 TABLE XXXVI. 
 STUB'S WIRE-GAUGE. 
 
 Stub's wire-gauge. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 ii 
 
 12 
 
 13 
 
 14 
 
 15 
 
 16 
 
 17 
 
 18 
 
 19 
 
 20 
 
 Fractions of an inch, . 
 
 / 
 
 */ 
 
 */ 
 
 H/ 
 
 A/ 
 
 
 
 A* 
 
 tt* 
 
 A/ 
 
 A* 
 
 t* 
 
 ft 
 
 A/ 
 
 A/ 
 
 A* 
 
 A/ 
 
 A* 
 
 A/ 
 
 A* 
 
 A/ 
 
 NOTE. f means full ; b means bare. 
 
 6. Methods of expanding Tubes. The tubes are generally secured in the tube- 
 sheets by expanding their ends by means of a special tool, which forces the metal of 
 the tubes into close contact with the circumference of the holes in the tube-plates, and 
 in some cases forms a shoulder on the tubes inside and outside the tube-plates. Some- 
 times a ferrule is driven tightly into the tube, and the projecting end of this ferrule 
 may be riveted over the end of the tube. These ferrules add much to the tightness and 
 holding power of the tubes in the plates, but contract the opening for draught and in- 
 terfere with the sweeping of fire-tubes. The ends of the tubes are annealed before 
 undergoing the process of expansion. It is of great importance that the joints be- 
 tween the tubes and the tube-plates shall be made perfectly tight, not only to pre- 
 vent leakage with its attendant evils, but also to make the tubes act as efficient stays 
 for the tube-plates. On the other hand, when the tube-plates are kept too rigidly in 
 position, and cannot yield in obedience to the expansion of the tubes in the direction of 
 their length with an increase of temperature, the tubes must have a chance to bend 
 sidewise, which action will take place when the length of the tubes is so great in pro- 
 portion to their diameter that they sag down. The thickness of the tube-plates must be 
 sufficient to afford a good bearing to the tubes for making the joint, and to ensure 
 proper stiffness in the plates after the numerous large holes are drilled. They are gene- 
 rally made at least \ inch thick. 
 
 The centres of the holes being marked on the tube-sheets, a small hole, may be 
 punched to serve as a guide for the drill, and the tube-holes are then drilled accurately 
 circular and cylindrical to such a size that the tubes can be just passed through by 
 hand ; if the holes be too large the ends of the tubes might have to be expanded so 
 much that they might crack or be strained, causing them to split afterward. The hole, 
 
272 
 
 STEAM BOILERS. 
 
 CHAP. XL 
 
 Fig. 128. 
 
 after being drilled, is often slightly counterbored at the outer face of the sheet ; the 
 edge of the hole is then smoothed with a file to remove burrs and inequalities, which 
 would prevent contact all around and dent or cut the tube-end. The tube, which is a 
 quarter of an inch longer than the distance between the outer faces of the two sheets, is 
 then put in place so that each end protrudes one-half of that amount ; the expanding- 
 tool being then inserted at the end, the tube is expanded, sometimes inside and outside 
 the tube-plate, and sometimes riveted over slightly on the outside into the counterbore. 
 Sometimes the tube-holes are bored slightly conical for the purpose of increasing the 
 holding power of the tubes (see Plates XXIII., XXIV.) 
 
 Figure 128 represents Raymond's patent recessed tube-sheet. The tube-sheet is 
 
 made very thick to give it great stiffness, and the 
 tube-holes are countersunk, so that the tube-ends 
 may be turned over within the recess. In this man- 
 ner the tube-ends in the combustion-chamber are less 
 exposed to the impingement of the products of com- 
 bustion, which causes them to wear rapidly, thus de- 
 stroying their tightness and holding power. For fur- 
 ther protection the recess may be filled with cement. 
 
 Prosser's expanding - tool is shown on Plate 
 XXIII., and acts by percussion. It consists of a 
 circular die formed by several truncated sectors, held 
 together loosely by a ring, which, however, is large 
 enough to permit them to be driven asunder by a 
 slightly conical mandrel forced between them ; the 
 dies are provided. with two rounded shoulders, one of 
 which, when the mandrel is driven, enlarges the por- 
 
 tion of the tube within, and the other that without, the plate. The mandrel being 
 backed out by tapping it on the side or through the other end of the tube, the sectors 
 are released and turned somewhat and the mandrel driven in again, so as to subject all 
 portions of the tube-end to their action. 
 
 Dudgeorf s tube-expander (see Plate XXIII.) was invented 1867, and is very exten- 
 sively used. It consists of a hollow cylinder of less diameter than the tube, and pro- 
 vided with openings to receive three or more steel rollers, which rest inside upon a cen- 
 tral conical mandrel. A guide-sleeve, which bears against the tube-sheet, is secured to 
 the hollow cylinder by a set-screw. By shifting the position of this sleeve the ex- 
 pander is made to answer for any thickness of tube-sheet. By inserting the tool into 
 
SEC. 6. 
 
 FLUES AND TUBES. 
 
 273 
 
 Fig. 129. 
 
 the tube and pressing upon and revolving the mandrel the rollers are set in motion, and, 
 travelling around the circumference of the tube, press out the metal of the tube-end 
 gradually. After the metal of the tube has completely filled the hole it is compressed 
 in the wake of the tube-sheet, but it yields on each side of it, forming 
 shoulders there ; this action is shown exaggerated in figure 129. 
 
 These expanders are made of different sizes to fit any size of tube be- 
 tween the limits of one inch and seven inches external diameter 
 of tube. Fig. 130. 
 
 Sometimes the rollers are shaped as in figure 130 ; this 
 form, it is claimed, strains the metal of the tube less than the 
 straight rollers. 
 
 In still another tool the rollers are made conical, with the same angle as the man- 
 drel, but placed in an opposite direction, so as to set the tube out exactly parallel to 
 its axis. 
 
 TweddeWs hydraulic tube-expander, represented in figure 6, Plate XXII., is said 
 to be very efficient and rapid in its action. The water is pumped in at A directly 
 from a small hand-pump, and forces outward the ram B, which draws the hexagonal 
 wedge C through the dies D, thus expanding the tube in the hole in the tube-plate 
 E. Upwards of sixty tube-ends an hour, it is said, can be finished by this tool, using 
 a pressure of from 1 to If tons on the square inch. 
 
 After the tube has been expanded the ends are often riveted over by the boot-tool, 
 sketched in figure 131, or, in large and accessible tubes, they may be hammered out 
 with a round-headed coppersmith's hammer, drawn 
 in figure 132. It is objected to these operations that 
 the end of the tube becomes thereby a brittle ring, 
 which is burned off by the action of the fire, espe- 
 cially in the back-connection ; such a result, how- 
 ever, indicates rather that the tube-ends were not 
 sufficiently annealed, or that the operation had been 
 overdone or unskilfully done. The operation of beading should not be commenced till 
 all the tubes are fixed in place and expanded, since the stiffness of the tube-plate is 
 greatly increased in this manner, and the blows of the hammer jar the plate less. 
 
 The beading of tube-ends is done rapidly and smoothly by Selkirk's tube-beader, 
 illustrated in figure 7, Plate XXII. It is worked by means of a ratchet, which enables 
 the operator to work in very confined spaces. The beading is done by rolling over the 
 tube-ends against the tube-plate gradually with an increasing pressure ; this method 
 
 Fig. 131. 
 
 Fig. 132. 
 
274 STEAM BOILERS. CHAP. XI. 
 
 obviates the danger of splitting the tube-ends, jarring severely the expanded tubes, and 
 indenting the tube-plate by blows with the hammer, so common under the system of 
 hand-beading in ordinary use. The action of the tool is described in Engineering, 
 June, 1877, as follows: "The fixing-piece a is first secured firmly in the boiler- tube b 
 by the action of the nut <?, which draws the coned mandrel d outwards, thrusting the 
 serrated wedge-pieces e (three in number) outward against the tube b. The body-piece 
 f is then brought into position, so that the beading- rollers (three in number) bear upon 
 the edge of the tube at b'. A considerable pressure is then brcmght to bear upon the 
 beading-rollers by the action of the nut 7i, friction-rollers, i, being placed between the 
 nut h and the body-piece f in order to avoid all unnecessary labor. The body-piece f 
 is then made to revolve upon the fixing-piece a by means of a ratchet-wheel, Jc, and 
 pawl, I ; the body-piece carries with it the beading- rollers g, and a very perfect bead is 
 thus formed quickly upon the end of the tube as shown at &'." 
 
 In using the expanding-tools great care must be exercised not to carry the operation 
 farther than necessary to secure tightness. The process of expanding by means of the 
 Prosser tool, when carried too far, strains the metal severely and may cause the tube- 
 ends to split. The Dudgeon tool becomes a dangerous instrument in the hands of an 
 inexperienced or careless person, since the operation of rolling out the metal may be 
 continued till the tube-ends are entirely cut off without giving warning. To prevent 
 this the taper mandrel often carries a loose collar, which may be secured in any posi- 
 tion by means of a set-screw, and thus limit the distance which the mandrel may enter 
 the tool and force out the rollers. 
 
 In securing boiler-tubes it should be taken into consideration in which direction the 
 pressure tends to force the tube-sheets ; thus, in the vertical water-tube boiler, the ten- 
 dency being to force the tube-sheets together, it would seem that the Prosser expand- 
 ing-tool, which forms a shoulder within the sheet, would be the preferable one to be 
 used ; but in the fire-tube boiler the steam exercises pressure tending to force the tube- 
 sheets off the ends of the tubes, and here it is evident that riveting over the ends of the 
 tubes increases greatly their holding power, while the offset within the sheet adds to 
 their tightness. The holding power of expanded tubes is, however, ample without 
 riveting their ends oVer, as shown by experiment. 
 
 7. Stay-tubes. Formerly it was often thought necessary to secure the tube-plates 
 by stay-rods, designed to take the whole strain due to the pressure on the plates ; but, 
 by reference to the experiments recorded in section 9 of this chapter, it will be seen 
 that the simple process of expanding the ends of tubes gives them sufficient holding 
 power to enable them to act as efficient stays for the tube-sheets. 
 
SBC. 7. FLUES AND TUBES. 275 
 
 In the boilers of U. S. S. Nipsic, represented on Plate XII., special stay-tubes 
 are provided to support the tube-plates at the points where an additional strain is 
 thrown on them by the pressure acting on the centre manhole-plate at the front tube- 
 plate, and on the stay-dome at the back tube-plate. These stay -tubes are of brass like 
 the other tubes, and have the same diameter and thickness, but their holding power is 
 increased by beading over their ends after being expanded with the Dudgeon tool ; for 
 this purpose the tubes are made about one-quarter inch longer than the other tubes. 
 Such stay-tubes are sometimes made of iron, so as to be less easily injured by burning 
 in case the water should get low in the boiler ; but to place these iron tubes among a 
 mass of closely-spaced brass tubes seems hazardous, on account of their increased lia- 
 bility to corrosion. 
 
 The practice of providing special stay-tubes secured by screw-threads and nuts pre- 
 vails still to a great extent in boilers where the steam-pressure exceeds 45 Ibs. per 
 square inch. In English boilers these stay-tubes are generally placed not more than 18 
 inches apart from centre to centre. At the front end these tubes are often secured by 
 two nuts, one inside and one outside the tube-plate ; the other end, exposed to the in- 
 tense heat prevailing in the back-connection, is secured by a single nut and by screwing 
 the tube into the tube-plate with about eleven threads to the inch. These stay-tubes 
 are generally of the same external diameter as the other tubes, but thicker. 
 
 The boilers of the steamer Atrato, built by James Watt & Co. in 1872, and designed 
 to be worked with a steam-pressure of 60 Ibs. per square inch, had tube-plates f inch 
 thick, brass tubes 4 inches in diameter outside, spaced 5f inches from centre to centre. 
 Each tube-sheet contained four stay-tubes, spaced 16J inches apart according to the 
 following directions : 
 
 "The stay-tubes are to be screwed into the back tube-plate with a nut on the out- 
 side, and there are to be two nuts on the front end, one inside and one outside. All 
 tubes, including stay-tubes, to be ferruled at the back end. Tubes are to be one-eighth 
 inch larger in diameter at front end, to render it easier to draw them when they have a 
 scale on. Back tube-plates to have tapered holes into which the tubes will be ex- 
 panded. Back end of tubes to be beaded ; front end need not be. Both ends may be 
 put in with Dudgeon's tool, if advisable. Stay-tubes to be i incl^ thick ; others No. 8 
 B. W. G. Total number of stay-tubes, 48 ; of others, 432." 
 
 In the boilers of the steamer Lord of the Isles, represented on Plate XV., the out- 
 side diameter of the stay-tubes is 4 inches, while that of the other tubes is 3 inches. 
 The stay-tubes are made inch thick, with a thread cut into the body of the tube, 
 leaving an effective outside diameter of 3f inches, and an inside diameter of 3J inches. 
 
276 STEAM BOILERS. CHAP. XL 
 
 8. Devices for rendering Boiler-tubes Removable. Many devices have been 
 tried to secure boiler-tubes in the tube-plates in such a manner as to make it possible to 
 remove them without injuring them and to replace them easily, in order to clean their 
 surfaces thoroughly or to make the crown-sheets of the furnaces accessible for scaling 
 and repairs. But so far none of these devices has proved perfectly satisfactory, so 
 that the method of securing boiler-tubes permanently by expanding their ends con- 
 tinues to be the almost universal practice. When tubes secured in this manner have to 
 be removed the diameter of the expanded ends is reduced by closing them, and even 
 then considerable force is often required to draw the tubes. The ends of iron tubes 
 generally crack during this process, and brass tubes are so much injured about the ends 
 that these have to be cut off and new ends have to be brazed on. 
 
 To render boiler-tubes more easily removable after scale has formed on them they 
 are often made tapering, the outside diameter at the front end being sometimes inch 
 larger than at the back end. 
 
 The following method is said to have been successfully used in France with iron 
 tubes : A short piece of tube made of very soft iron, and having the fibres running cir- 
 cumferentially instead of longitudinally, is welded to. the ends of the boiler-tubes ; these 
 are secured in the tube-plates by expanding them and turning over slightly the extreme 
 projecting ends. When a tube is to be removed the ends are grasped and closed by 
 suitable nippers ; this, it is said, can be done without injuring the tube-ends perma- 
 nently, on account of their peculiar structure. 
 
 Figure 1, Plate XXII., represents a removable tube used in some boilers built at 
 the West Point Foundry, Cold Spring, N. Y., in 1878. The tubes are 4 inches in dia- 
 meter and 13 feet long ; the ends of the tubes are thickened by brazing or welding upon 
 them a coned ring ; the largest diameter at the back end is made a little smaller than 
 the smallest diameter at the front end. The tubes are driven tightly into the holes of 
 the tube-plate and slightly expanded. The tube-plates are one inch thick, and were 
 turned down at the rim to a thickness of inch to turn the flange. The holes for the 
 tubes were punched, and reamed in place after the tube-plates were riveted to the shell. 
 The tube-plates are tied together by braces of l-inch round iron, secured by nuts and 
 placed about 18 inches apart. The boilers were tested with cold water, and found to be 
 tight under a pressure of 135 pounds per square inch. These tubes are known as 
 PaukscK s boiler-tubes, and have been used to some extent in England and France. 
 
 Figures 2, 3, 4, Plate XXII., represent methods of securing removable tubes used in 
 the French navy, and known there respectively as the Systeme Infernet et Gouttes 
 (figure 2), Systeme Toscer (figure 3), and Systeme Langlois (figure 4). In the latter 
 
SEC. 9. FLUES AND TUBES. 277 
 
 the front end of the tubes is packed by a leaden washer, a small gutter being cut in 
 both collar and plate, into which the lead is squeezed by screwing up the tubes ; a 
 brass collar is brazed to the end and notched for the purpose of applying the wrench. 
 To prevent the adhesion of the tubes by rusting the threads are smeared with zinc 
 cement, and it has been found that tubes can be taken out without much difficulty after 
 having been in use two years. The back end is fixed by means of a slightly conical 
 steel or iron ferrule, tightened by means of an expanding-tool. It is claimed that a tube 
 can be removed in five minutes. 
 
 Figure o, Plate XXII., represents a method of fixing removable tubes which was 
 tried some years ago in high-pressure boilers of United States naval vessels, but proved 
 unsuccessful. The bushings which secure the brass tubes in the tube-plates were made 
 of composition metal ; those at the back end of the tubes were soon burnt by the intense 
 heat prevailing in the back-connection, and even those at the front end, which were not 
 injured by heat, could not be unscrewed after the boilers had been in use a short time, 
 but were twisted off in the attempt. Similar devices have been tried in English boilers 
 with equally unsatisfactory results. 
 
 9. Experiments on the Holding Power of Boiler-tubes secured by various 
 Methods. In January, 1877, a series of experiments were made at the Navy-Yard, 
 Washington, D. C., under the direction of Chief Engineer William H. Shock, U.S.N., 
 on the holding power of boiler-tubes fixed by various methods employed in marine and 
 locomotive engineering. Each tube subjected to trial had its ends fixed in square 
 pieces of plate resembling portions of tube-plates. The pull was applied to stirrups 
 attached by nuts to cross-heads which bore against the plates (see Plate XXV.) The 
 cross-heads were made in halves and with a circular opening in the centre, in order to 
 enclose the tube and allow the pull to be applied exactly in the direction of the axis of 
 the tube. Plates XXIII., XXIV. illustrate fully the methods of fastening the tubes 
 and the appearance of the specimens after the trial, and contain a complete tabulated 
 record of the results of these experiments. The Rodman testing-machine was used in 
 making these experiments. 
 
 Plate XXIII. exhibits the results of forty-eight experiments made with brass tubes 
 having an external diameter of 2.5 inches and 2.6 inches, and an area of metal, in cross- 
 section of 0.9 and 1.33 square inches respectively. The tube-plates were of iron, and 
 varied in thickness from f inch to f inch. Both the Prosser and the Dudgeon tool 
 were used in expanding the tubes ; the effects of beading over the ends and of driving 
 ferrules into the tube-ends were tested xinder different conditions: 
 
 A comparison of the results obtained with tubes No. 5 and No. 6 shows that the 
 
278 STEAM BOILERS. CHAP. XI. 
 
 partial turning-over of the ends of the tube effected by the Prosser expander does 
 not give such a firm hold of the tube-plate as the beading-over by hand in connec- 
 tion with the action of the Dudgeon roller-expander. The tubes secured by simple 
 expansion gave way by being drawn through the plates, while the ends which were 
 beaded over generally broke. 
 
 The effect of ferrules in increasing the holding power of tubes is very marked ; 
 they prevent the ends from collapsing and being drawn through the plates. 
 
 The tubes secured by nuts only, screwed on the outside of the plates, gave way by 
 drawing the ends through the nuts without stripping or otherwise injuring the thread. 
 When iron ferrules were used in connection with the nuts the holding power of the 
 tubes was greatly increased. In experiments Nos. 15 and 16 the tubes gave way 
 by tearing through the thread ; but in experiments Nos. 19 and 20 the tubes drew from 
 the nuts without breaking, like the unferruled tubes. 
 
 The lowest results were obtained in experiments Nos. 21 and 22, when the tubes 
 were simply expanded by the Dudgeon tool in a f -inch plate, without being beaded 
 over or secured by ferrules, the resistance being 7,650 Ibs. and 5,850 Ibs. respectively. 
 It will be seen that even in this most unfavorable case the holding power of the tube 
 was greatly in excess of any strain which would be occasioned by the pressure of steam 
 upon the portion of the tube-plate which any one tube would have to support in a 
 boiler. 
 
 The following general conclusions drawn from the results of these trials are quoted 
 from an article in Engineering, Sept. 14, 1877, where an account of these experiments 
 was first published viz. : " (1) That 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 are not 
 beaded over the hold afforded by the Dudgeon is less than that afforded by the Prosser 
 system of fixing ; (3) that with both expanders the introduction of ferrules adds very 
 materially to the holding power of the tubes ; (4) that, on the whole, the effect of fer- 
 rules 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 
 afforded by the introduction of ferrules is not materially affected by the thickness of 
 the tube-plates ; (5) that iron ferrules are more efficient than those of brass ; and (6) 
 that the employment of nuts screwed on the tubes outside the tube-plates is not of any 
 service in increasing the holding power unless the tubes are ferruled." 
 
 Plate XXIV. exhibits eighteen experiments made with iron tubes secured in iron, 
 steel, and copper tube-plates by various methods obtaining in locomotive engineering. 
 
SEC. 9. FLUES AND TUBES. 279 
 
 Dudgeon's tool is used in all cases in expanding these tubes. The outside diameter of 
 the tubes is 2$- inches, but their lower end is in every case reduced by swaging to 2f 
 inches, and is fixed in a copper or steel plate, respectively f inch and f inch thick ; the 
 upper end is fixed in every case in an iron plate ^ inch thick. 
 
 In experiments No. 1 to No. 8 the ends of the expanded tubes were beaded or 
 riveted over, and in every case fracture took place by breaking the riveted end off the 
 tube, except in experiment No. 8, when the riveted end was cracked but not completely 
 detached from the tube, and was pulled through the plate. This took place also in ex- 
 periments Nos. 9 and 10, where the ends were only partly riveted over. It is important 
 to notice that in experiments Nos. 5, 6, 7, and 8, where the lower tube-plate was of 
 copper, as well as in experiments Nos. 1 and 2, where at the lower end a thin copper 
 ring was inserted between the tube and the steel plate, the fracture took place invaria- 
 bly at the upper end ; while in experiments Nos. 3 and 4, identical with experiments 
 X( is. 1 and 2 with the omission of the copper ring, fracture took place once at the upper 
 end, fixed in an iron plate, and once at the lower end, fixed in a steel plate. 
 
 The great difference in the strain under which tube-ends fixed in precisely the same 
 manner gave way indicates that the method of riveting over iron tubes is apt to injure 
 the metal, although for these experiments all the tubes were carefully fixed by the same 
 experienced workman. The strains at which rupture took place in the first ten experi- 
 ments ranged from 29,050 in tube No. 1 to 17,300 Ibs. in tube No. 6 ; the mean breaking 
 strain for these ten tubes was 22,837.5 Ibs. 
 
 In experiments No. 11 to No. 18 the tube-ends were not riveted over, and the tubes 
 gave way invariably at the lower end, showing that their holding power was decreased 
 by diminishing their diameter and the thickness of the tube-plates ; experiments Nos. 
 8, 9, and 10 indicate, on the contrary, that when the ends are riveted over the holding 
 power of the tube is not influenced perceptibly by a slight decrease in the thickness of 
 the plate and in the diameter of the tube. 
 
 Comparing experiments Nos. 11 and 12 with Nos. 13 and 14, it will be noticed that 
 the holding power of the tubes is more than doubled by the insertion of ferrules. Ex- 
 periments Nos. 13 and 14 show also that iron tubes, simply expanded by Dudgeon's 
 process, possess more than sufficient holding power to bear any strain which may be 
 thrown on the tube-plates by the steam-pressures used in locomotive boilers. 
 
 In experiments Nos. 15, 16, 17, and 18 the holes were made tapering. In the former 
 two experiments the larger diameter of the holes was ^ inch greater than the smaller 
 diameter, and the results show a remarkable decrease in the holding power of the tubes 
 from that of the tubes Nos. 13 and 14, with simply expanded ends in cylindrical holes. 
 
280 STEAM BOILERS. CHAP. XI. 
 
 In experiments Nos. 17 and 18 the larger diameter of the taper holes was -fr inch greater 
 than the smaller diameter, and with these proportions the holding power was greatly 
 increased. 
 
 1O. Sectional or Water-tube Boilers, Hanging Tubes, Double Tubes, etc. 
 
 Since a number of years the so-called sectional or water-tube boilers have come into 
 great favor as stationary boilers for various purposes, and many devices relating to this 
 class of steam-generators, and presenting more or less novelty, have been patented and 
 introduced into the market. In general these boilers consist of an assemblage of tubes 
 connected with one another by means of elbows or branch-pipes, and placed in vertical 
 and horizontal tiers, over and surrounding a grate, and enclosed by walls built of fire- 
 brick or constructed of some other non-conducting material ; in some arrangements the 
 tubes are bent into a coil or into a siphon-shape. 
 
 The principal advantages claimed for this class of boilers are the following : (1) The 
 small diameter of the tubes of which they are composed, and the absence of riveted 
 joints, render them much stronger than the ordinary rectangular or cylindrical boilers. 
 (2) They are safer ; for even in case some tubes burst no violent explosion ensues, be- 
 cause the fractured parts present a relatively small opening, and the quantity of water 
 and steam contained in these boilers is small in proportion to their power. (3) They 
 can be cheaply built, and repaired with great facility, duplicate pieces being easily kept 
 in store for this purpose ; the separate parts of a boiler can be transported long dis- 
 tances without great expense or inconvenience ; the form and proportions of a boiler 
 are easily altered or adapted to any available space, and the power of a boiler is in- 
 creased by simply adding new tiers of tubes and grate-surface. (4) Their evaporative 
 efficiency can be made equal to that of other boilers, and, in fact, for equal proportions 
 of heating-surface and grate-surface, it is often somewhat higher. 
 
 With two or three exceptions all attempts made so far to adapt these boilers for use 
 on board of vessels have resulted unsuccessfully, and in several instances disastrously. 
 While in some instances these failures could be traced to avoidable mistakes in the 
 design of the boilers, there are several reasons why tubulous boilers are not well adapted 
 for marine purposes, unless radical changes should be introduced in the present prac- 
 tice of marine engineering. 
 
 (1) These tubulous boilers occupy as much valuable space as the ordinary types of 
 marine boilers. 
 
 (2) On account of the small quantity of water carried in them any irregularities in 
 the supply of feed-water or in the management of the fires cause sudden fluctuations of 
 pressure, and a sudden, rapid generation of steam leads to an accumulation of steam in 
 
SEC. 10. FLUBS AND TUBES. 281 
 
 the water-chambers, and to priming, loss of water, and overheated tubes ; these trou- 
 bles can be much more easily avoided with regularly and continuously working factory 
 engines than with marine engines, which have often to be operated in an intermittent 
 manner in getting under way and in coming to a wharf or to an anchorage. 
 
 (3) The horizontal or inclined water-tubes, of which these sectional boilers are 
 mainly composed, do not present a ready outlet for the generated steam. The steam- 
 bubbles, instead of being able to follow their natural tendency and rise, have generally 
 to travel in a horizontal direction the whole length of the tubes, and this they will not 
 do without being urged by an extraneous force. There exists, consequently, a liability 
 for steam to accumulate in the water-tubes and cause them to be burnt; and this 
 liability is generally greater in the case of marine boilers, where economy of space de- 
 mands a rapid combustion and evaporation, than in stationary boilers, where economy 
 of fuel is sought for by slow combustion and evaporation. 
 
 (4) This liability to overheating of the tubes is still increased by the use of water 
 which forms deposits of solid matter on the heating-surfaces of boilers. The surface- 
 condensers of marine engines are seldom so perfectly tight as to keep the boilers fully 
 supplied with distilled water for any length of time ; and, besides, it is a well-estab- 
 lished fact that with marine boilers the choice lies between the formation of a slight 
 deposit of scale and rapid corrosion. 
 
 Some inventors have relied upon the scouring action of the water circulating rapidly 
 through the tubes of their boilers as a means of preventing the deposit of scale ; others 
 thought that the expansion of the heated tubes would detach each thin deposit of scale, 
 and thus prevent it from accumulating to a dangerous extent. But these speculations 
 have not been realized in practice. Besides, in boilers of naval vessels, which are often 
 kept for many days in succession under banked fires, the circulation of the water is 
 during such time necessarily sluggish. 
 
 In PerJriwfs tubulous boiler, which, on a small scale, has been successfully applied 
 in two or three instances to marine purposes, distilled water, as nearly as possible 
 chemically pure, is used. All joints of the boiler are made perfectly tight. A surface- 
 condenser and stuffing-boxes of peculiar design are used for the engines, in order to 
 avoid all loss of steam ; no lubricant is used for the valves or pistons ; the whistle, the 
 blast, and the steam-pumps are supplied with steam by an auxiliary boiler; in this 
 manner it is made possible to use the same pure water over and over again. A quantity 
 of fresh water carried in tanks, and a distiller, are provided to replace any water acci- 
 dentally lost. 
 
 The boiler illustrated in figure 1, Plate XXVI., is described by the inventor, 
 
282 STEAM BOILERS. CHAP. XL 
 
 L. Perkins, in a lecture delivered before the Royal United Service Institute, 1877, as 
 follows : 
 
 " The horizontal tubes are 2J inches internal and 3 inches external diameter, except- 
 ing the steam-collecting tube, which is 4 inches internal and 5 inches external diame- 
 ter. The horizontal tubes, being welded up at each end one-half inch thick, are con- 
 nected by small vertical tubes | inch internal and 1^ inches external diameter. The 
 fire-box is formed of tubes bent into a rectangular shape, placed at a distance of If 
 inches apart, and connected by numerous small vertical tubes. The body of the boiler 
 is made of a number of vertical sections composed each of eleven tubes, connected at 
 either end by a vertical tube ; these sections are connected at both ends by a vertical 
 tube to the top ring of the fire-box, and by another to the steam-collecting tube. The 
 whole of the boiler is surrounded by a double casing of thin sheet-iron, filled up with 
 vegetable black to avoid loss of heat. Every tube is separately proved by hydraulic 
 pressure to 4,000 Ibs. on the square inch, and the boiler complete to 2, 000 Ibs., this 
 pressure remaining on for some hours." The connecting-tubes are screwed into the 
 main tubes, and the threads are then calked down to make the joint perfectly tight. 
 
 Some boilers of this description have been used on land for thirteen years, with pres- 
 sures varying from 250 to 300 Ibs., and tubes, cut out for examination at the end of 
 that period, have been found to be clean and to show no signs of corrosion, owing partly 
 to the rigorous use of distilled fresh water, and partly, it is supposed, to the formation 
 of the black oxide of iron by the contact of superheated steam with the highly-heated 
 surfaces. It is stated that steam is formed even in the lowest tubes ; that water exists 
 in the form of spray in the middle portion of the boiler, and that the upper tubes con- 
 tain dry steam ; and that the safety of the tubes is ensured by the great density of the 
 steam, which increases greatly its power of conducting heat. These boilers have gene- 
 rally been operated with a low rate of combustion, while, at the same time, the ratio of 
 heating-surface to grate-surface is large. 
 
 Figure 1, Plate XXVI., represents one of the four boilers of the steam-yacht Wan- 
 derer. Each boiler contains 19 square feet of grate-surface and 760 square feet of heat- 
 ing-surface ; the working pressure is 400 Ibs. per square inch. The total weight of the 
 four boilers, including water, is about 34 tons. It is claimed that the Wanderer's 
 engines developed, with 92 revolutions per minute, a maximum of 907 horse-powers. 
 In other vessels, having boilers of identical design and dimensions, a performance of 150 
 horse-powers per boiler was obtained. 
 
 Figure 2, Plate XXVI., represents a Howard water '-tube boiler designed for marine 
 purposes. The tubes forming each vertical tier communicate with each other at one 
 
SEC. 10. FLUES AND TUBES. 283 
 
 
 
 end by a stand-pipe, and near the other end, which is closed, by short corrugated con- 
 necting-tubes. The closed ends are provided with mudhole-doors. The feed-water 
 enters a horizontal tube at the back of the boiler, connected by short branches to the 
 vertical stand-pipes. A cylindrical steam-drum runs across the top of the boiler, being 
 connected to the top tube of each vertical tier. 
 
 The Belleville water-tube boilers have attracted much attention in France, and a 
 great number of them were built some years ago for the French navy. Various modifi- 
 cations have been introduced in these boilers from time to time. Figure 2, Plate 
 XXVII., represents a boiler of this class built for the despatch- vessel IS Active, of the 
 French navy, in 1868. There were three of these boilers, supplying steam to compound 
 engines of 400 indicated horse-powers. 
 
 The following information regarding them is derived from Engineer ing, March 4,1870 : 
 Each steam-generator proper consists of eleven "elements," each composed of 
 twelve tubes, disposed one above the other and coupled at their alternate ends by con- 
 necting-boxes. The eleven elements are placed side by side at a short distance apart. 
 A transverse tube, B, connects all the lower tubes and serves to distribute the feed- 
 water to the various elements. The upper tubes are all connected with a transverse 
 steam-collecting tube, C, parallel to which is placed the round pipe D, which serves to 
 lead the steam off to the engines. The pipes C and D are connected by five vertical 
 pipes of different sizes, for the purpose of equalizing the draught of steam from the 
 various sections. The tubes are of wrought-iron, lap- welded; the connecting-boxes 
 into which the tubes are screwed are of malleable cast-iron, and each is furnished with a 
 couple of mudholes opposite the ends of the tubes. All the elements are connected to- 
 gether at the front by the upper and lower connecting-tubes, and at the back by a 
 wrought-iron frame and tie-bolts, distance-pieces being provided to maintain the proper 
 intervals between the elements ; in this manner the tubes are left free to expand or con- 
 tract longitudinally. The tubes are each formed in a single piece, except the upper and 
 lower tubes of each element, which are each made in two pieces, connected by a 
 coupling, for the purpose of facilitating the putting together and taking to pieces of the 
 elements. The whole of the boiler proper is enclosed in a casing composed of wrought- 
 iron plates and angle-irons, lined with fire-bricks. A number of plates are placed in 
 the upper part of the boiler to deflect the current of the gases. The feed is regulated 
 by a self-acting, adjustable float placed in the vertical wrought-iron cylinder in front of 
 the boiler, to which also" the gauge-cocks are attached. A "separator,''' placed between 
 the boilers and the engines, frees the steam of the water and other foreign matter held 
 in suspension. The principal dimensions of the boilers of IS Active are as follows : 
 
284 STEAM BOILERS. CHAP. XI. 
 
 Number of boilers 3. 
 
 Height of boilers 7 feet 7 inches 
 
 Length of boilers 8 feet f inch. 
 
 Width across three boilers 14 feet 8 inches. 
 
 Total number of tubes 396. 
 
 Interior diameter of tubes 2f inches. 
 
 Thickness of tubes 0.24 inch. 
 
 Length of tubes 6 feet lOf inches. 
 
 Total grate-surface 75.34 square feet. 
 
 Total heating-surface 2,347 square feet. 
 
 Load on safety-valves 113 Ibs. per square inch. 
 
 Figure 1, Plate XXVII., represents the HerresJioff coil-boiler of the steam-yacht 
 Estelle. 
 
 The following data are derived from the report of the Board of Naval Engineers, who 
 conducted a trial with this vessel in December, 1877 : 
 
 This boiler consists of a single circular grate, 7 feet in diameter, surrounded by a 
 fire-brick wall 18 inches in height above the top of the grate-bars and 7 inches in thick- 
 ness. Upon the top of this brick wall there rests a single coil of continuous wrought- 
 iron pipe, which contains the steam and water, while its outside is exposed to the hot 
 gases of combustion. The outline of this coil, considered as a whole, is composed of 
 the frustums of two right cones, one superimposed upon the other. The lower frustum 
 is 7 feet in inner diameter at the base and 6 feet in inner diameter at the top, with a 
 vertical height of 7 feet and 4 inches. The upper frustum is 6 feet in inner diameter at 
 the base and 1 foot and 11 inches in inner diameter at the top, with a vertical height of 
 lOf inches. The spirals of the lower frustum are kept apart f inch by stirrup-bolts. 
 The spirals in the upper frustum touch, and the entire top is covered with sheet-iron ; 
 thus all the gases are compelled to pass between the openings between the spirals of the 
 lower frustum. The coil makes a total of thirty turns, all of which are continuous, 
 without a single joint, being made by welding together the ends of the several sections 
 of pipe composing the coil. Starting from the bottom, the pipes forming the coil 
 gradually decrease in size, the lowest section being 4| inches in outside diameter and 
 0.119 inch in thickness, and the uppermost section being If inches in outside diameter 
 and 0.069 inch in thickness. The coil is surrounded by a hollow cylindrical casing, 
 8 feet 3 inches in outside diameter, filled with fire-brick If inches thick. 
 
 The feed-water enters the coil at the extreme top, and, flowing slowly down the 
 
SEC. 10. FLUES AND TUBES. 285 
 
 
 
 spirals, becomes converted into steam. From the lower end of the coil a straight 
 wrought-iron pipe, 4| inches in outside diameter, passing up on the coil directly over 
 the grate, leads the steam to the "separator" a cylindrical vessel located outside the 
 boiler, where the water and other impurities mingled with the steam are deposited. 
 The steam is drawn off at the top of the separator, and passes through a coil of pipe 
 within the uptake of the boiler, where it is superheated before it enters the engines. It 
 is essential that a portion of the feed-water should pass in the form of spray with the 
 steam into the separator, in order to prevent the overheating of the lower coils of the 
 tube. In the older boilers of .this type, in which sea-water was used, this surplus of 
 water entered the separator as highly-concentrated brine, and was blown off. In later 
 boilers, where fresh water is used, this water is either blown into the condenser or is 
 directly fed back into the boiler by a special pump. The quantity of water which 
 passes thus into the separator is about 25 per cent of the quantity evaporated. 
 
 The following are the principal dimensions and proportions of the boiler illustrated 
 on Plate XXVII. : 
 
 Diameter of the boiler to outside of casing 8 ft. 3 ins. 
 
 Height of the boiler from bottom of ashpit to top of coil 11 ft. 3 ins. 
 
 Area of grate-surface 38.4846 sq. ft. 
 
 Total area of heating-surface, measured on the outside of the pipe 511.184 sq. ft. 
 
 Length of the axis of the coil 539.550 ft. 
 
 Capacity of the coil 34.488 cub. ft. 
 
 Ratio of heating-surface to grate-surface 13.283 to 1. 
 
 Weight of brick- work in the boiler (calculated) 7,250 Ibs. 
 
 Weight of iron in the boiler (calculated) 9,250 Ibs. 
 
 Total weight of boiler, including brick- work, grate-bars, ashpit, 
 
 uptake, separator, casing, coil, etc 16,500 Ibs. 
 
 Weight of water in coil and separator, allowing the coil to be half- 
 
 filled 1,100 Ibs. 
 
 Steam was raised with wood from water at the temperature of 44 Fahr. in sufficient 
 quantity to start the motive-engine, and maintain it in motion, in seven minutes from 
 the lighting of the fire. The feeding has to commence a few moments before the fires 
 are lighted. 
 
 During the trial, which lasted eight hours, the extreme variations of boiler-pressure 
 were between 65 and 75 Ibs. per square inch. This uniformity of pressure is ascribed to 
 be due to the use of artificial draught by means of a fan-blower, and to the uniform 
 working of the engine. 
 
286 STEAM BOILERS. CHAP. XI. 
 
 
 
 During a short full-power trial, lasting fifteen minutes, the engine developed 293.28 
 indicated horse-powers, with a steam-pressure of 106.5 Ibs. in the boiler. 
 
 The interior of the pipe is, of course, utterly inaccessible for scaling or examination. 
 The use of the boiler is limited to fresh water, supplied by tanks or surface-condensers. 
 When sea- water is used the coil gradually scales up, commencing at the lower end of 
 the coil. 
 
 The lightness of the boiler is dependent to a great extent upon the small amount of 
 heating-surface and to the thin walls of the fire-brick casing ; but, even after increasing 
 the former to the amount usual in marine boilers, the weight of the Herreshoff boiler 
 would be only about one-half of the weight of the ordinary marine boiler. 
 
 Many novel devices in the arrangement and form of the tubes of vertical boilers have 
 been introduced of late, with a view to combining lightness and compactness and rapid 
 circulation of the water in every part of the boiler with high potential and economic 
 evaporative efficiency. 
 
 The Davey-Paxman boiler, illustrated in figure 2, Plate XXVIII., is highly recom- 
 mended in these respects. The merits of this boiler are due to the bent and tapering 
 water- tubes and to the deflector inserted in the top of each tube. The water in these 
 tubes, being exposed to the full heat of the furnace, soon acquires a high temperature, 
 and as it rises rapidly it is replaced by solid water flowing in at the lower end. So 
 great is the velocity of the water through these tubes that it rises in jets up to the 
 crown-plate, unless arrested by the deflectors, which divert these water-jets downward 
 and keep the water perfectly smooth on the surface. This rapid circulation of the water 
 allows no incrustation to take place in the tubes. 
 
 In many cases boilers are fitted with so-called " hanging-tubes." These tubes hang 
 vertically over the fire, being, closed at the bottom by means of a plug or by welding, and 
 being secured at the upper end in the tube-plate which forms the crown-plate of a very 
 high furnace. A tube of smaller diameter leads down into the larger tube, leaving an 
 annular space through which the steam ascends, while solid water flows down through 
 the inner tube. The top of the latter extends a short distance above the outer tube, 
 but the lower end does not reach quite to the bottom of the outer tube, in order to leave 
 room for the passage of the water. The success of this arrangement depends on the 
 complete separation of the ascending and descending currents in the tubes. The outer 
 tubes are sometimes fitted at the top with deflecting arrangements of various forms. 
 These hanging-tubes are principally used in the boilers ^of fire-engines, road-en- 
 gines, etc. 
 
CHAPTER XII. 
 
 UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWEBS, ETC. 
 
 1. Smoke-connections and Uptake. The gases of combustion are discharged 
 by the tubes or flues into chambers called the smoke-connections, or, in the return-tube 
 boiler, the front-connections, which gradually converge to a common passage, called the 
 uptake, leading to the base of the chimney. The smoke-connections and uptake are 
 either built permanently within the shell of the boiler, forming an integral part of the 
 latter and being partly surrounded by water and steam-spaces, or they consist of a 
 separate box, constructed of angle-irons and plate-iron, and secured to the outside of 
 the shell of the boiler. In the former case the uptake may be so arranged as to present 
 valuable heating-surface for drying and superheating the steam ; this subject will be 
 considered in chapter xiii. When the front-connections and uptake are built sepa- 
 rately they form, with their linings, a considerable part of the total weight of the boil- 
 ers ; they increase greatly the temperature of the fire-room by radiation, unless they are 
 well protected by non-conductive materials ; and they give frequently trouble by 
 warping when the gases of combustion are discharged at a high temperature. 
 
 The smoke-connections must not only form a sufficiently large and unobstructed pas- 
 sage for the gases of combustion, but must allow easy access to be had to the tubes for 
 sweeping, replacing, or calking them ; for this purpose they are provided with large 
 hinged doors. The cross-area of the front-connections increases gradually from the 
 bottom to the top and from the ends of the boiler to the place where they merge in the 
 uptake, so as to preserve a uniform ratio of cross-area to the bulk of the gases dis- 
 charged into them by each additional row of tubes. All sudden enlargements should 
 be avoided as much as possible ; all bends should be made with easy curves, and the 
 irregular form of the uptake should change gradually into the regular figure represent- 
 ing the cross-section of the smoke-pipe. When several currents of a fluid, moving in 
 different directions, meet in a common passage they retard each other, and, under cer- 
 tain conditions, the one moving with the greatest speed may even obliterate entirely 
 
 287 
 
288 STEAM BOILERS. CHAP. XIL 
 
 the other currents. This well-known phenomenon is too often lost sight of in the con- 
 struction of the uptakes of boilers, and the draught of boilers may be seriously injured 
 in consequence. The current of gases issuing from one set of flues should never cross the 
 direction of other currents on entering the same passage, but partitions should be pro- 
 vided which keep the currents separate till they have assumed the same direction. 
 
 In rectangular boilers of the return-tube type the front smoke-connections and the 
 uptake are generally built permanently in the boiler (see Plates VI., VII., XVII.) In 
 this case the front tube-plates are set back a short distance from the front of the boiler, 
 and the latter is made to slope outward from the bottom of the front-connection up- 
 ward, in order to get more room at the top for the passage of the gases. The bottom of 
 the connection must be placed so as to give sufficient room for expanding and calking 
 the lower row of tubes, and for turning the flange which connects the tube-plate to the 
 bottom plate of the connection. Room is saved between the latter and the furnace- 
 crown by turning the flange on the tube-plate, and not on the bottom plate of the front- 
 connection. The top of the front-connection is generally arched, not only to give it 
 strength, but because such a form offers less resistance to the flow of the gases than a 
 square cross-section. 
 
 The front-connections form often a clear passage extending from one end of the 
 boiler to the other. In this case the construction of the boiler is greatly simplified by 
 leaving the whole front of the connections open, and forming the jambs or supports for 
 the connection-doors by bolting flat bars to the front of the boiler across this open 
 space. 
 
 In the rectangular boilers of United States naval vessels these jambs are generally 
 formed by columnar water-spaces (see Plates VI., VII.) ; they complicate the con- 
 struction of this part of the boiler, but are useful as channels for the downward course 
 of the water. 
 
 By extending these water-spaces across the front-connections, so as to form walls 
 which separate the several nests of tubes, the weight of the boiler is slightly increased, 
 but some additional heating-surface and freer channels for the circulation of the water 
 are gained ; and the draught of the end furnaces is improved, because the body of 
 gases generated by each furnace is kept separate until the different currents assume the 
 same direction on entering the uptake. 
 
 The form of the uptake depends on the arrangement of the boilers with reference to 
 one another and to the chimney. When a single boiler is used the uptake slopes in- 
 ward, so as to bring the smoke-pipe over the base of the boiler. When several boilers 
 are placed opposite to one another the uptake slopes outward, spanning the space be- 
 
SBC. 1. UPTAKE, CHIMNEY, STEAM-JETS, FAX-BLOWERS, ETC. 289 
 
 tween the boilers ; with such an arrangement each boiler contains a portion of the 
 uptake, so that when the boilers are placed in position the sides of the overhanging 
 uptake meet and are bolted together, forming thus a common passage (see Plates VI., 
 VII.) In such a case it would be better to keep the uptake of each boiler separate by 
 a partition extending to the base of the chimney. Such parts of the uptake as lie out- 
 side the shell of the boiler, and are not surrounded by steam-drums, are frequently 
 lined with fire-brick, or are coated on the outside with some incombustible, non-con- 
 ductive material to prevent the radiation of heat. 
 
 With high-pressure cylindrical boilers it is more convenient to build the boiler 
 proper complete in itself, and to add the front-connections and uptake as separate 
 structures ; this plan simplifies the construction of the cylindrical shell and avoids the 
 use of flat stayed surfaces. The bottom, top, and sides of the uptake and connections 
 are generally either lined with fire-brick or are made with a double shell, which is fre- 
 quently filled with some non-conductive substance, like plaster-of -Paris or a mixture of 
 plaster-of -Paris and ashes. 
 
 The uptakes, and the fastenings which secure them to the boilers, must be strong 
 enough to carry the weight of the smoke-pipe in addition to their own weight ; and, 
 besides, they are sometimes designed to tie the boilers together at the top. They are 
 made to rest partly on the top of the cylindrical shell, and special provisions, in the 
 shape of beams and stanchions, are often made to support their overhanging portions. 
 The required strength and stiffness of the structure should be provided for by a proper 
 arrangement of the frames, made of angle-irons, T-irons, or channel-irons. The selec- 
 tion of the thickness of the plate-iron must be governed by the following considera- 
 tions : It must be riveted together with air-tight joints ; it must not buckle under the 
 strains or warp in consequence of the heat to which it is exposed ; and it must not be 
 destroyed too rapidly by corrosion. For double shells much thinner iron may be used 
 than for single shells lined with fire-brick. The inner lining of double shells is made 
 heavier than the outer lining, because it is more exposed to warping and corrosion. 
 
 In the boilers of the U. S. S. Nipsic (see Plates XXIX., XXX.) the front-connec- 
 tions of each set of boilers on the same side of the vessel form a clear passage from one 
 extreme end to the other. They are attached to the front of the boilers by angle-irons, 
 and to the cylindrical shells by channel-irons, secured by tap-bolts. The bottom of the 
 front-connections is formed by a single thickness of plate riveted to the angle-iron, 
 which extends in one continuous length along the fronts of the boilers. The other 
 walls of the front-connections and of the uptake are double, the inner lining being 
 made of No. 10 W. G. iron and the outer lining of No. 13 W. G. iron. The inner and 
 
290 STEAM BOILERS. CHAP. XII. 
 
 outer plates of the double shell are connected by channel-irons (2J* X 2" X f "), through- 
 rivets passing through both plates and both flanges of the channel-irons. The sides of 
 the connections and uptake are connected by angle-irons. Additional supports for the 
 connections are provided in the shape of brackets resting on the flanges of the furnace- 
 tubes, and in the centre under the smoke-pipe the uptake is supported by a pair of 
 flanged beams, 12 inches deep, running in the fore-and-aft direction of the vessel. 
 These beams are placed 20 inches apart, and are supported at either end by a wrought- 
 iron stanchion resting on the keelson. Supports for the deck-beams over the boilers 
 rest likewise on these beams. 
 
 The weight of the front-connections and uptake of these boilers is, according to cal- 
 culation, about 10,000 Ibs., and the actual weight of the plaster-of-Paris used in filling 
 them was, in the dry state, 5,000 Ibs., making the total weight of the front-connections 
 and uptake of these boilers when filled about 16,350 Ibs. 
 
 In the U. S. S. Trenton, having eight three-furnace boilers twelve feet in diameter, 
 each boiler has a separate front-connection, sloping outward from a least width of 9 
 inches at the bottom to a uniform width of 30 inches at the top, where it is open to the 
 uptake. The latter forms a continuous passage along the fronts of the boilers on each 
 side of the vessel. Its cross-section is square, and its width and height increase gradu- 
 ally from the extreme ends to its junction with the base of the chimney. All the walls 
 of the connections and uptake consist of a single shell, lined with fire-brick in the up- 
 take. The side, bottom, and front plates of each front-connection are riveted to a two- 
 inch angle-iron bent to the shape required for the outline of the box and secured to the 
 front of the boiler, and to the square frames, made of 2-inch angle-irons, surrounding 
 the connection-door openings. The plates forming the uptake are of J-inch iron, con- 
 nected by 2-inch angle-irons. The uptake is fastened at the back to the cylindrical 
 shell of each boiler by means of a 3-inch angle-iron, and the overhanging part is sup- 
 ported by 4-inch T-irons, two of which are placed on the top of the front-connection box 
 of each boiler. These T-irons extend across the fire-room, and are secured at either end 
 by a strap bolted to the front of the boiler. For the support of the smoke-pipe the 
 central part of the uptake forms a heavy framework. The pipe rests on a square 
 f-inch plate, with a circular opening corresponding to the cross-section of the smoke- 
 pipe, and strengthened by a ring formed of 4-inch angle-iron and having an inner 
 diameter 2 inches larger than the outside diameter of the base of the pipe. The four 
 sides of this horizontal top plate are secured by 3-inch angle-irons to vertical plates f 
 inch thick. The vertical plates running athwartships are 12 inches deep, and those 
 running in a fore-and-aft direction are secured to the cylindrical shells of the boilers by 
 
SEC. 2. UPTAKE, CHIMNEY, STEAM-JETS, PAN-BLOWERS, ETC. 291 
 
 3-inch angle-irons. The rest of the uptake is connected to this central portion by 
 2-inch angle-irons. 
 
 It is of the utmost importance that the uptake-doors be made to fit air-tight against 
 their seats, in order to prevent the in-leakage of air, the effect of which is to decrease 
 the draught of the chimney by lowering its temperature and increasing the bulk of 
 gases to be passed through it in a given time. 
 
 The draught of the boiler measures, other things equal, its potential vaporization ; 
 and having constructed a boiler, every precaution should be taken to obtain from it the 
 utmost performance by losing none of the draught due to the temperature and bulk of 
 the gases of combustion delivered into the base of the chimney. When it is neces- 
 sary to force from a boiler the utmost quantity of steam in a given time, the uptake- 
 doors, where they meet their seat, should be luted with clay, so as to absolutely prevent 
 the ingress of the cold external air. These remarks, of course, apply only to the cases 
 where natural draught is employed alone or in conjunction with a steam-jet in the 
 chimney. When the draught is produced artificially by means of blowers delivering 
 blasts of air into closed ashpits, tight uptake-doors may still be needed to prevent the 
 gases of combustion from being driven into the fire-room. 
 
 When the uptake-doors are so large that the labor of opening and closing them be- 
 comes serious it is frequently convenient to construct in them a much smaller door, to be 
 opened when it is desired to check the combustion in the furnaces. This combustion 
 may, indeed, be checked still more promptly by opening the furnace-doors, thereby 
 allowing the inrush of a mass of cold air above the incandescent fuel and through the 
 furnaces and tubes ; but it is done at the risk of injuring the riveting by the too sudden 
 cooling of the plates, and the radiation into the fire-room from the glowing fires is so 
 great as to be a serious inconvenience, and sometimes an injury, to all the persons who 
 have duties there. 
 
 2. Forms and Dimensions of Chimneys. The chimneys of marine boilers are 
 generally cylindrical, with a circular cross-section. Sometimes the cross-section is oval, 
 with the greater diameter lying in the fore-and-aft direction of the vessel. This form 
 is used to gain room on deck athwartships for clearing certain parts of the rigging. 
 Another advantage claimed for this form viz., that it offers less resistance to a head- 
 wind is practically insignificant. The flat sides of oval chimneys are stiffened by 
 braces. A chimney of circular cross-section has not only the strongest form, requires 
 the least weight of metal, and is most easily made, but offers the least surface for fric- 
 tion and radiation. 
 
 When several boilers discharge their gases into the same chimney the latter is some- 
 
292 STEAM BOILERS. CHAP. XII. 
 
 times divided by partitions running the whole length of the pipe, so that each division 
 forms a separate chimney for each boiler. This arrangement is advantageous for war- 
 vessels, which frequently steam with only a fraction of their boiler-power. According 
 to Ledieu, this plan of subdividing the chimney is often adopted in the French navy, 
 even when hoisting-pipes are used, although in such cases it complicates their con- 
 struction greatly. 
 
 The effect of this division of a chimney by partitions extending from bottom to top 
 was tested by Chief Engineer Isherwood, of the United States navy, on board the 
 United States steam-frigate San Jacinto, in 1862. .The tubular boilers of that vessel 
 were two in number, placed opposite each other with the fire-room between and in com- 
 mon to both, the chimney being also common to both and placed over the centre of the 
 fire-room. When only one boiler was in operation the difference in its draught was 
 strongly marked, whether the partition was left out and the whole chimney cross-area 
 used, or whether the partition was put in and half the chimney cross-area used ; the 
 ashpit-doors, furnace-doors, and uptake-doors of the boiler out of operation being care- 
 fully luted in the former case so as to absolutely prevent any passage of air through it. 
 
 When the draught is produced by artificial means viz., by a steam -jet or by fan- 
 blowers there is a certain cross-area of chimney which gives the least resistance by 
 friction and the best effect of the blast, while the height of the chimney need only be 
 great enough to discharge the products of combustion without producing inconvenience. 
 In the English torpedo-vessel Vesuvius the chimney consists of a horizontal duct leading 
 aft along the sides of the vessel. 
 
 When natural draught is to be used that is to say, when the draught is to be pro- 
 duced by the difference in weight of the column of hot gas within the chimney and of 
 an equally high column of outside air the dimensions of the chimney for a given rate 
 of combustion may be calculated according to the rules and formulae given in section 
 11, chapter ii. When natural draught is used in marine boilers the cross-area of the 
 chimney varies from one-sixth to one-tenth of the area of the grate ; and the limit 
 of the height of the chimney of steamships is about 65 feet, measured from the level of 
 the grate. 
 
 In war-vessels which are intended to manoeuvre frequently and for long periods 
 under sail without the use of steam-power the chimney is made telescopic, with one or 
 two movable sections, which slide within a fixed pipe and are hoisted when the boilers 
 are in use and lowered after the fires are hauled. The chimneys of steam-launches and 
 similar small vessels are frequently provided with a hinge (see Plate XVI.), so that they 
 can be let down into a horizontal position. 
 
SEC. 2. UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWERS, ETC. 293 
 
 Chimneys must he made with air-tight joints to prevent leakage of air on account 
 of the difference of pressure inside and outside, as such leakage injures the draught. 
 For this reason, also, the use of telescopic chimneys is very objectionable, causing a 
 marked decrease in the draught of the boiler, as it is impossible to make them air-tight, 
 there being necessarily an annular space of more or less width, according to accuracy 
 of fitting, between the standing and the sliding portions. 
 
 Telescopic chimneys are employed only on board of war-steamers, and in them only 
 because a considerable portion of their cruising is done under sail alone. The position 
 of the chimney being so near the mainmast as to prevent the setting of the mainsail, 
 the inconvenience is sought to be avoided by lowering the chimney ; the height of the 
 standing portion, however, is frequently such that, even when reduced to the minimum, 
 the mainsail cannot be set. The principal benefit derived from telescoping the chimney 
 is to make the vessel look more like a sailing ship an appearance extravagantly paid 
 for in the decreased power of the boiler and consequently lessened speed of the vessel 
 under steam-power. 
 
 If it was desirable to keep the inner surface of the chimney clean, then that surface 
 should be made as smooth as possible, so as to offer the least resistance to the ascending 
 gaseous currents, which would be particularly important in the case of small chimneys 
 and rapid currents. But it is found advantageous to allow the inner surface to remain 
 coated with the soot and tarry hydrocarbons deposited from the gases of combustion, 
 as this coating efficiently prevents the radiation of heat from the outer surface, which 
 it is desirable to avoid, as such radiation reduces the draught by lowering the tempera- 
 ture of the gases within the chimney. So long as it is the surface of this coating which 
 is exposed to the gaseous currents, the smoothness or roughness of the surface to which 
 the coating adheres is of but little importance. Of course the friction-resistance of the 
 rough surface presented by the hydrocarbon coating reduces the draught more than a 
 smooth surface of metal, and to that extent loses what it gains by its less heat-conduct- 
 ing power. 
 
 To regulate the draught of a boiler, and consequently its rate of combustion, a valve, 
 called a damper, is often placed within the flues, which slides or swings across their 
 opening, and which in stationary boilers is sometimes regulated automatically by the 
 steam- pressure. In marine boilers the damper is frequently omitted ; when used it is 
 placed within" the smoke-pipe, near its lower end, and consists of a circular plate which 
 swings around a horizontal spindle. The latter projects outside the pipe, and is ope- 
 rated by hand by means of a rope or chain attached to a wheel or crank fixed to the 
 spindle. 
 
294 STEAM BOILERS. CHAP. XII. 
 
 3. Fixed Chimneys. Chimneys are built up of separate rings or courses ; the 
 length of the courses depends on the size of the plates used in their construction. Plate- 
 iron varying from No. 6 to No. 12 W. G. is used for large chimneys, the lower courses 
 being made of heavier iron than the upper courses. The upper end of the chimney 
 is stiffened by making it flaring, or by riveting a heavy wrought-iron band around it on 
 the outside. The longitudinal as well as the transverse seams of the courses are some- 
 times made with lap-joints ; the lap of each upper course must be placed on the out- 
 side, so that the ascending currents of gas do not strike against the ends of the plates. 
 A better and neater plan is to use butt-joints. For the longitudinal seams the butt- 
 straps may be placed inside the pipe ; the bands which connect the several courses at 
 the transverse seams are placed on the outside. 
 
 When the chimney is bolted rigidly to the uptake the bolts pass in some cases 
 through the flanges of angle-irons riveted around the top of the uptake and around the 
 base of the smoke-pipe, or a stout iron band is riveted around the upper end of the up- 
 take, and the lower end of the pipe fits into this band and is firmly bolted to it. With 
 such a rigid attachment of the chimney to the uptake great strains are often thrown on 
 the boiler when the ship rolls, especially when the stays are not adjusted with great 
 exactness. On this account it is better not to attach the chimney rigidly to the boiler, 
 but to let it simply rest on the uptake. In such a case the base of the chimney is 
 reinforced by a stout iron ring riveted to it, which fits, with sufficient clearance to allow 
 for expansion and irregularity of form, in an annular space formed on the top of the 
 uptake by a ring of angle-iron riveted around the mouth of the uptake ; the latter pro- 
 jects generally a few inches within the chimney. 
 
 The chimney is held in position by stays attached to lugs secured to the bands 
 which connect the upper courses, and leading to eye-bolts placed on the upper deck of 
 the vessel. When these stays are formed by chains they are attached to the eye-bolts 
 on the deck by short lengths of rope or marline, so as to make their length adjustable 
 and to give them some elasticity. The stays are sometimes provided with turn-buckles, 
 in order to make them adjustable. For appearance' sake chimneys are often made to 
 rake aft, but such an arrangement serves no practical purpose. 
 
 A cylindrical casing surrounding the lower part of the chimney, and extending from 
 the top of the uptake to a height of several feet above the upper deck, forms an annu- 
 lar space, generally from three to four inches wide, around the pipe, through which the 
 air can circulate freely, intercepting the heat radiated from the chimney. This casing 
 is made of No. 10 or No. 12 W. G. iron, and is stiffened at the top by a heavier wrought- 
 iron band. It is secured to the hatch-coaming of the upper deck. In some cases it 
 
SEC. 4. UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWERS, ETC. 295 
 
 rests on the top of the boilers, openings being cut at the lower end for the circulation 
 of the air. The top of this jacket projects within the " apron" which is a cover riveted 
 to the chimney to protect the annular space between the jacket and the pipe from rain, 
 etc., while it allows the free escape of the rising air-currents. For the further protec- 
 tion of the surrounding wood- work against the heat of the chimney an annular copper 
 tank, from 2 to 3 inches wide, filled with water, is often secured around the air-casing 
 within the hatch. 
 
 The following directions were attached to the drawing of a chimney, 72 inches in dia- 
 meter and 64 feet high, designed for the U. S. S. Algoma and class in 1866 : 
 
 "Each pipe to be made of seven sections vertically, of the best charcoal-iron. The 
 plates of the three lower sections to be made of No. 6 W. G. iron, and those of the four 
 top sections to be made of No. 8 W. G. iron. The vertical seams to be butted with 
 strips of iron on the inside ; the joints to be made very close ; rivets to have button- 
 heads on the outside. The horizontal joints to be made with neat wrought-iron bands 
 on the outside. The bands of the first and second joints from the top to have each 
 eight lugs for stays. Sixteen stays to be provided for each pipe, of suitable length, in 
 links of wrought-iron \ inch in diameter, with proper attachments to the deck." 
 
 4. Hoisting-chimneys. Hoisting or telescopic chimneys are made with one or 
 two movable sections, which generally slide within a lower fixed section. In the 
 ' Traite elementaire des Appareils a Vapeur de Navigation,' by Ledieu, an illustration is 
 given of a telescopic chimney of oval cross-section with the hoisting part larger than the 
 standing part, so that it slides outside the latter within the air-casing. This arrange- 
 ment seems to have been adopted because the standing part of the pipe is divided by a 
 partition lengthwise into two separate passages. In another example given in the same 
 work, where the standing part is divided in a similar manner, a separate semi-cylin- 
 drical pipe slides within each division of the fixed pipe. 
 
 The usual mode of constructing telescopic chimneys for United States naval vessels 
 is illustrated on Plate XVII. 
 
 The chimney of the U. S. S. Plymouth (see Plate XVII.) is circular in cross-section 
 and has one movable part. The fixed pipe has an inside diameter of 76 inches and is 
 made of No. 6 W. G. iron. The movable pipe has an inside diameter of 72 inches and 
 is made of No. 8 W. G. iron. The plates forming each course are butt-jointed, being 
 riveted to longitudinal wrought-iron bars, -fa inch thick and 3 inches wide, placed 
 inside the fixed and outside the movable pipe ; the heads of the rivets are countersunk 
 in the bars. The several courses are connected by circumferential butt-straps placed 
 outside the pipes. The top of the fixed pipe, as well as that of the movable pipe, is 
 
296 STEAM BOILERS. CHAP. XII. 
 
 stiffened by an iron band, 4 inches wide and about % inch thick, riveted to the pipe on 
 the outside. To each of these bands are secured six wrought-iron links for the attach- 
 ment of stays. A wrought-iron ring, 1J inches thick and 2 inches wide, is riveted 
 around the top of the fixed pipe on the inside. Similar rings are riveted on the outside 
 of the movable pipe, one at the bottom and one at a distance of 34 inches from the bot- 
 tom. When the movable pipe is lowered it rests with the bottom ring on the top of the 
 uptake ; when it is hoisted to its full height the second outside ring bears against the 
 inner ring around the top of the fixed pipe, forming as close a joint as practicable. The 
 longitudinal bars form ways for guiding the pipe when it is being raised or lowered. 
 To prevent jamming a clearance of ^ inch is left between the ways and the If-inch 
 rings when the movable pipe is concentric with the fixed pipe. To support the mova- 
 ble section when hoisted four steel bolts are tapped through composition sleeves 
 secured at a suitable height to the fixed pipe. When the bolts are run in after the 
 pipe is hoisted its bottom ring rests on these bolts. 
 
 The pipe is hoisted by four wire ropes, f inch in diameter, leading over pulleys, 
 7 inches in diameter, secured to the top of the fixed pipe. One end of each wire rope 
 is attached to a lug or eye-bolt fixed to the bottom ring of the movable pipe, suit- 
 able openings being cut in the upper part of the fixed pipe and in the lower part of 
 the movable pipe to let the wire ropes pass through. The other end of each rope 
 leads either directly or over guide-pulleys to a windlass carrying four drums fixed to 
 one horizontal shaft. This windlass is mounted over the hatch of the lower deck, and 
 is operated by a hand-crank attached to an endless screw which gears in a wheel secured 
 to the shaft carrying the drums. 
 
 Two windlasses, placed on opposite sides of the chimney, may be used, each carry- 
 ing two drums. These two windlasses are either operated independently of each other 
 or they are connected by means of bevel-gearing on a shaft provided with hand-cranks. 
 Steam-power may be used instead of hand -power in hoisting the pipe, without changing 
 the arrangement of the drams materially. Sometimes each rope leads to an indepen- 
 dent drum operated by a hand- crank and suitable gearing ; but in this case it is found 
 difficult to maintain an equal tension on all the ropes in hoisting the pipe. When 
 several drums are fixed to a common shaft the ropes must be provided with an arrange- 
 ment for adjusting them readily to the proper length and tension ; for this purpose the 
 shank of the eye-bolt to which each rope is attached passes through a lug fixed to the 
 pipe, and is secured by means of two nuts, one placed above and the other below 
 the lug. 
 
 The chimney of U. S. S. Nipsic has two movable sections. The fixed pipe has an 
 
SEC. 4. UPTAKE, CHIMNEY, STEAM-JETS, PAN-BLOWERS, ETC. 297 
 
 inside diameter of 79 inches and a length of 16 feet, and is made of No. 7 W. G. iron. 
 The lower movable pipe has an inside diameter of 74 inches and a length of 17 feet 3 
 inches, and is also made of No. 7 W. G. iron. The upper movable pipe has an inside 
 diameter of 69 inches and a length of 16 feet 10 inches, and is made of No. 8 W. G. 
 iron. All seams of these pipes are made with butt-joints ; the longitudinal butt-straps 
 are inch thick and 3 inches wide, placed inside the fixed pipe and outside the two 
 movable pipes ; the heads of the rivets are countersunk in the straps. Each pipe is 
 stiffened at the top by a wrought-iron band, 4 inches wide and 1 inch thick, placed on 
 the outside of the pipe. The bottom end of each pipe has on the outside a wrought- 
 iron band 2 inches wide and from 1 to If inches thick. A similar ring is placed 
 around the top of the fixed pipe and of the lower movable pipe on the inside. When 
 the pipes are hoisted to their full height the rings at the bottom of the movable sections 
 bear against similar rings secured inside the fixed and the lower movable pipes, 26 
 inches from the top of the pipes. 
 
 The upper section is hoisted simultaneously with the lower movable section by 
 means of four wire ropes. One end of each rope is secured to the top of the fixed pipe. 
 Passing over a pulley attached near the top of the lower movable section, the rope leads 
 down between the upper and lower movable pipes, and has its other end attached to the 
 band at the bottom of the upper movable pipe. 
 
 The hoisting-gear of the lower movable pipe is of a novel design. A chain, made of 
 f -inch iron, has both ends attached to a wrought-iron beam fixed across the lower end 
 of the movable pipe, and passes then over two stationary puUeys secured to the upper 
 end of a pipe 18 inches in diameter, which is fixed centrally within the standing part of 
 the chimney and leads down through the uptake into the fire-room. The bight of the 
 chain passes around a movable pulley within this central pipe. This pulley carries a 
 swivel, to which the end of a chain, made of one-inch iron, is attached, that leads down 
 through the central pipe to a drum in the fire-room. The drum is revolved by hand- 
 power and suitable gearing. 
 
 When the chimney is hoisted it has a height of 44 feet and 1 inch, measured from 
 the bottom of the standing pipe. When the chimney is lowered the lower movable sec- 
 tion rests on the top of the fixed pipe, and the upper movable section rests on the top of 
 the lower movable section ; and the total height of the pipe, measured from the bottom 
 of the fixed pipe, is 18 feet f inch. 
 
 Double-hoist chimneys have less height when lowered than single-hoist chimneys ; 
 this is, in fact, the only advantage possessed by the former over the latter. On the other 
 hand, all the disadvantages connected with the use of all hoisting-pipes are greatly in- 
 
298 STEAM BOILERS. CHAP. XII. 
 
 creased in the case of the former. Compared with fixed pipes the disadvantages of 
 hoisting-pipes are as follows : they occupy more room in the hatch and are heavier ; 
 their first cost is far greater ; they are more liable to accidents, and much labor is re- 
 quired to keep their gear in working order ; they are less efficient on account of the 
 sudden changes in their cross-sections and because air-leaks are unavoidable with them. 
 
 In the U. S. S. Quinnebaug and class the chimney is made with one movable sec- 
 tion, and the height of the pipe above deck, when lowered, is reduced by letting the 
 movable section pass through an opening in the bottom of the uptake and rest on the 
 floor of the fire-room. When the pipe is hoisted the opening in the uptake is closed by 
 a door secured by clamps. The lower part of the movable pipe has two large openings 
 opposite to each other, provided with doors, in order to form a passage between the for- 
 ward and after parts of the fire-room when the pipe is lowered. 
 
 5. Artificial Draught : Blast-pipe, Steam-jets, Fan-blowers. To attain the 
 best possible natural draught with a given height of chimney the temperature of the 
 column of hot gas within the chimney has to be so great that from 25 to 33 per cent, of 
 the total heat generated by the combustion of the fuel is expended in producing the 
 draught. (See section 11, chapter ii., and section 7, chapter iii.) When the height of 
 the chimney and the bulk and weight of the boiler are limited, artificial draught has to 
 be used for increasing the evaporative power of the boiler beyond a certain limit. 
 Artificial draught has the great advantages that, all things considered, it is cheaper 
 than natural draught for high rates of combustion, and that it can be readily adjusted 
 for the combustion of different kinds of fuel and for widely different rates of combus- 
 tion, so that a given boiler may be worked under greatly varying conditions. 
 
 The general measure of the efficiency of the mechanism for producing artificial 
 draught is the ratio of the power expended in operating the mechanism to the increase 
 of draught produced ; the increase of draught may be measured by the increase in the 
 rate of combustion. With a given boiler the value of a mechanism for producing arti- 
 ficial draught depends not only on its efficiency as measured by the foregoing rule, but 
 on the economic evaporative efficiency of the boiler with different rates of combustion ; 
 on the weight and bulk of the mechanism ; on its first cost ; and on the labor and ex- 
 pense of operating it and keeping it in working order. 
 
 Artificial draught is produced either by diminishing the resistance of the column of gas 
 within the chimney or by increasing the pressure of the atmospheric air under the grate. 
 
 To produce the first of these conditions a fan-blower has been used for exhausting 
 the gases escaping up the chimney ; but this method has not come into general use on 
 account of practical difficulties. The ordinary method is to use a jet of steam in the 
 
SEC. 5. UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWERS, ETC. 299 
 
 chimney, which balances by its impact the weight of the column of gas within the 
 chimney. The action of the class of "fluid-on-fluid impulse machines" to which the 
 jet belongs, is described by Rankine in the following words : "A stream of fluid, mov- 
 ing at first with a certain velocity, drives and carries along with it an additional stream, 
 the two streams finally mingling and moving together with a velocity less than that of 
 the driving stream." 
 
 In locomotives the blast is produced by discharging the exhaust steam from the 
 cylinders into the chimney. This plan can, of course, be used only with high-pressure, 
 non-condensing engines, and is applied to many river-boats and to tug-boats and simi- 
 lar small craft. The efficiency of such a blast is to be measured by the increase of the 
 evaporative power of the boiler relatively to the increase of back-pressure produced in 
 the steam-cylinders. 
 
 " The effect of the blast-pipe in producing a draught depends upon its own diameter 
 and position, on the diameter of the chimney, and on the dimensions of the fire-box, 
 tubes, and smoke-box. Mr. D. K. Clark has investigated the influence of these circum- 
 stances from his own experiments and from those of Messrs. Ramsbottom, Polonceau, 
 and others, and has shown that the vacuum in the smoke-box is about 0.7 of the blast- 
 pressure ; that the vacuum in the fire-box is from to \ of that in the smoke-box ; that 
 the rate of evaporation varies nearly as the square root of the vacuum in the smoke- 
 box ; that the best proportions of the chimney and other parts are those which enable 
 a given draught to be produced with the greatest diameter of the blast-pipe, because 
 the greater that diameter the less is the back-pressure produced by the resistance of the 
 orifice ; that the same proportions' are best at all rates of expansion and at all speeds ; 
 and that the following proportions are about the best known : 
 
 Sectional area of tubes within ferrules = \ area of grate. 
 
 Sectional area of chimney = ^ area of grate. 
 
 Area of blast-orifice (which should be 
 somewhat below the throat of the 
 chimney) = -fa area of grate. 
 
 Capacity of smoke-box = 3 feet X area of grate. 
 
 Length of chimney = four times its diameter. 
 
 "If the tubes are smaller the blast-orifice must be made smaller also for exam- 
 ple, if 
 
 Sectional area of tubes within ferrules = T V area of grate, 
 
 Then area of blast-orifice = T V area of grate." 
 
 (Rankine, ' Manual of the Steam-engine:} 
 
300 STEAM BOILERS. CHAP. XII 
 
 When condensing-engines are used the steam-jet is supplied directly from the 
 boiler. Its efficiency is to be measured by the increase of evaporative power of the 
 boiler relatively to the weight of steam expended in the jet. The increase of draught 
 produced by the jet depends on the velocity with which a given weight of steam strikes 
 against the column of gas within the chimney, and on the area of that column immedi- 
 ately acted upon by the jet. C. Wye Williams found by experiment that thirty jets of 
 one-tenth inch sectional area, placed three inches apart, were more effective than sixty 
 jets of one-quarter inch sectional area placed one inch apart, and, moreover, saved an 
 enormous amount of steam. 
 
 Isherwood says ('Experimental Researches,' vol. ii.): "It is found experimentally 
 that with a properly-constructed steam-jet, composed of small brass nipples or hollow 
 truncated cones inserted in concentric rings of steam-pipe, placed in the smoke-pipe 
 about two feet above its bottom, the rings being so spaced as to equally distribute the 
 area of the smoke-pipe over them, and the nipples being three inches between centres 
 on the rings, the expenditure of steam of 40 Ibs. per square inch above the atmos- 
 phere by the jet to raise the rate of combustion in the water-tube boiler from 15$- Ibs. 
 per square foot of grate per hour to 24 Ibs.., the air-supply to the ashpit being copious 
 and not brought at the expense of the draught, is 7.22 per centum of the total evapo- 
 ration." 
 
 The jet arrangement designed for the chimney of the U. S. S. Algoma and class 
 (described in section 3 of the present chapter) consisted of three rings of gas-pipe 
 arranged in a pyramidal form above one another, with a vertical distance of 16 inches 
 between them, and connected with one another by four branch-pipes screwed into 
 couplings. These rings were made of 3-inch, 2^-inch, and 2-inch pipe, and had an out- 
 side diameter of 65 inches, 48| inches, and 32 inches respectively, the larger pipe and 
 ring being at the bottom and the smaller pipe and ring at the top of the system. The 
 steam-jets issued through brass nipples screwed into the rings and placed 3 inches 
 apart. These nipples were -ff inch long ; the opening through them was conical, de- 
 creasing from ^ inch near the bottom to ^ inch near the top, and flared at the bottom 
 and at the top to f inch and J inch respectively. 
 
 When a fan-blower is used to force a supply of air directly under the grates the ash- 
 pit-doors and furnace-doors are closed tight, and the mouths of the ducts which convey 
 the air from the blower enter the ashpit either through openings at the back of the 
 boiler or in front through openings in the ashpit-doors. Each branch-duct is provided 
 with a valve to shut off the air-supply from any furnace when it is necessary to open 
 the ashpit or furnace doors. 
 
SEC. 5. UPTAKE, CHIMNEY, STEAM-JETS, PAN-BLOWERS, ETC. 301 
 
 The net power required to work a blower may be calculated by means of the formula 
 given in section 12, chapter ii. 
 
 "The quantity of steam required to work the fan-blower rapidly enough to produce 
 a combustion of even 35 Ibs. of anthracite per square foot of grate-surface per hour is 
 quite an insignificant per centum of the total evaporation. Such an apparatus is by far 
 the most economical method of producing the draught ; but as the blast must be de- 
 livered beneath the grate-bars, with air-tight ashpit-doors, the ventilation of the fire- 
 room is almost wholly destroyed by it, and the firemen, with boilers in the hold of ves- 
 sels, find the heat and dust insupportable. The fan-blower is generally worked by a 
 small independent steam-cylinder, and in a vessel the space occupied by the apparatus 
 and its weight are considerable ; also, the trouble of looking after the numerous cocks, 
 valves, etc., connected with it, and the unavoidable complexity attending additional 
 machinery, have operated to its exclusion on board of marine steamers." (Isher- 
 wood, i Experimental Researches ^ vol. ii.) 
 
 Sometimes, instead of forcing the air directly into the ashpits, it is delivered into 
 the fire-room, which is enclosed by air-tight bulkheads and decks, and has no outlet for 
 the air except through the grates. In this manner an increased barometric pressure is 
 produced within the fire-room. The boilers are worked with open ashpits, and the 
 ventilation of the fire-room is as perfect as with natural draught. When, however, the 
 space is not quite air-tight a waste of power ensues ; and this can scarcely be avoided. 
 This plan has been adopted to advantage in iron-clad batteries, torpedo-boats, and 
 similar vessels, which have to be constructed with tight decks, and are, consequently, 
 dependent upon artificial means for ventilation and a supply of air to the furnaces. 
 
 In some instances air has been forced through nozzles into the closed ashpits by 
 means of steam- jets ; and in the place of steam- jets a blast of air supplied by a fan- 
 blower has been tried in the chimney. 
 
 Koerting's jet apparatus consists of a series of short nozzles 'gradually increasing in 
 size (see figure 1, Plate XXX Y.) According to the inventor the number and dimen- 
 sions of these nozzles are determined by the following considerations : 1st, to make the 
 velocity of the motive fluid a maximum as it escapes under a varying back-pressure ; 
 2d, to produce an intimate mixture of the propelling fluid with the fluid to be set in 
 motion. 
 
 The tube admitting the propelling fluid is fixed obliquely at the side of the smallest 
 nozzle, the opening of which can be varied by a screw in order to regulate the admis- 
 sion of steam or compressed air. The annular openings between the nozzles, admitting 
 the air to the propelling jet, gradually increase in width as the distance from the open- 
 
302 STEAM BOILERS. CHAP. XII. 
 
 ing admitting the steam increases and the velocity of the jet decreases. Provision is 
 made sometimes to vary the relative position of the conical nozzles by means of a screw, 
 in order to regulate the width of the annular spaces. 
 
 In the first nozzle the steam-jet is mixed with a certain quantity of air and forms 
 with it a new jet, which becomes mixed in the successive nozzles with additional quan- 
 tities of air entering through the annular spaces. The jet gradually decreases in 
 velocity, while the volume of air set in motion increases proportionately. The velocity 
 attained in the last nozzle corresponds to the reqiiired pressure. 
 
 The apparatus represented in figure 1, Plate XXXV., is designed to force air under 
 the grate of a boiler or other furnace. With an initial steam -pressure of 40 Ibs. per 
 square inch the pressure of air produced by this apparatus is equal to a head of 2 
 inches of water. 
 
 Apparatus of similar construction are used for steam-jets in chimneys. 
 
 6. Experiments with Artificial Draught in Marine Boilers. In the years 
 1865-66 experiments were made at the United States Navy- Yard, New York, with 
 various devices for increasing by artificial means the rate of combustion in marine 
 boilers. Steam and air jets of various forms and dimensions, and located at different 
 heights within the chimney, were tried, and air was forced into closed ashpits by means 
 of fan-blowers and steam-jets of various dimensions. These experiments were further 
 varied by altering the dimensions of the grates and the ratios of calorimeter and heat- 
 ing-surface to grate-surface. The results of these experiments are tabulated in the 
 Report of the Board of Engineers convened by the United States Navy Department 
 to make experiments with the horizontal fire-tube boiler and the vertical water-tube 
 boiler of the Martin type, for the purpose of determining the relative merits of these 
 two types of marine boilers. Since the primary object of these experiments was not to 
 determine the relative efficiency of the various devices used for producing artificial 
 draught, the results of all the experiments are not comparable. The experiments with 
 the vertical water-tube boiler were much more numerous and varied than those with 
 the horizontal fire-tube boiler, therefore only the former will be considered ; and, in 
 order to eliminate as much as possible all uncertain elements, only those will be 
 selected for comparison which were made without altering the heating-surface and 
 calorimeter of the boiler, and in which the grate either had its original length of 6 
 feet 6 inches or was shortened to 6 feet so as to reduce the grate-surface to 36 square 
 feet. 
 
 The vertical water-tube boiler, which was built expressly for the purpose of these 
 experiments, had two furnaces, each 36 inches wide and containing a grate 6 feet 6 
 
SEC. 6. UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWERS, ETC. 303 
 
 inches long. It had a separate chimney 35 inches in diameter and 60 feet high above 
 the grate, and it contained 748 brass tubes of 2 inches external diameter and 28f inches 
 long. 
 
 Total grate-surface 39 sq. ft. 
 
 Total heating-surface 1264.81 sq. ft. 
 
 Ratio of grate to heating-surface 1 to 32.43 
 
 Ratio of grate to calorimeter of tubes 7.04 to 1 
 
 Ratio of grate to calorimeter of chimney 5.79 to 1 
 
 Each experiment lasted from 48 to 80 hours continuously. In the experimental 
 boiler the evaporation took place under atmospheric pressure. The steam for operating 
 the jet or fan was supplied by an independent boiler, in which the pressure ranged 
 from 25 to 40 Ibs. per square inch above the atmosphere. The water supplied to both 
 these boilers .was measured in separate tanks. The coal was Pennsylvania anthracite 
 of "egg-size," free from dust, and the amount used was accurately weighed. 
 
 The quantities given in columns b, c, d, e of the following table are taken directly 
 from the tables accompanying the above-mentioned report, except when they represent 
 the mean results of several experiments, which are calculated from the quantities given 
 in the original tables. The quantities contained in columns /, / g, g,, h, Ti t , j, ./. are 
 calculated from the results recorded in the original tables. 
 
 Column f shows the quantity of steam expended for blast in per centum of the 
 total weight of water evaporated, and column f l the weight of water of 212 Fahr. 
 evaporated under atmospheric pressure equivalent to the actual weight of steam ex- 
 pended for blast per hour. 
 
 Columns g, g a show the quantity of steam, in pounds, available for useful work, or 
 the difference of the weight of steam expended for blast and of the total weight of 
 water evaporated. 
 
 Columns h, Ji and /, j, show the cost of producing the increase in the rate of com- 
 bustion and in the evaporative power of the boiler in each experiment, expressed in 
 pounds of steam expended in blast for each additional poutid of coal burned and of 
 water evaporated respectively. 
 
 On account of the unavoidable differences in the quality of the coal, and on account 
 of the influence of the variations in the rates of combustion on the economic evapora- 
 tive efficiency of the boiler, the tabulated results indicate only approximately the rela- 
 tive value of the various methods tried. 
 
 Comparing the results of experiments Nos. 3, 4, 5, 6, and 7, it will be seen from 
 
304: STEAM BOILERS. CHAP. XH 
 
 columns h, 7i a that when the jet-coil is supplied with nozzles of proper form and dimen- 
 sions the efficiency of the steam-jet is nearly twice as great as when the steam issues 
 through plain holes drilled in the coil. Columns h, 7i indicate likewise that in ex- 
 periments Nos. 4, 6, 6, and 7 the weight of steam expended for each additional pound 
 of coal consumed did not vary greatly. 
 
 Comparing experiments Nos. 8, 9, 10 N 11, 12, and 13 with the preceding ones, it ap- 
 pears that the large single nozzle produced on the whole a better useful effect than the 
 steam-jet consisting of numerous small nozzles. This result, which apparently contra- 
 dicts other experiments, may be explained by the fact that the chimney was not of 
 large diameter, and that the coils of the jet-pipe formed a serious obstruction to the 
 passage of the gases in the smoke-pipe. To preserve the proper calorimeter the smoke- 
 pipe was enlarged to a diameter of 38 inches at the place where the jet-coil was situ- 
 ated ; a better result would probably have been obtained by giving to the coil the 
 pyramidal form described in section 5 of this chapter, since it obstructs less the cross- 
 section of the pipe. 
 
 The height of the single jets above the grate was varied in several experiments with- 
 out producing any marked difference in the results. When, however, the jet was 
 placed only 6 feet below the top of the' chimney it became ineffective. A single steam- 
 jet placed in each back -connection showed likewise no useful results. 
 
 The great efficiency of the fan-blower in increasing the rate of combustion by forc- 
 ing air into closed ashpits is illustrated in columns h, ~h m experiments Nos. 16, 17, and 
 18 ; and columns j, j\ show how far the relative economy of this system of forcing the 
 draught is maintained for high rates of combustion with decreased economic evapora- 
 tive efficiency of the boiler. In experiment No. 18 the economic evaporation falls so 
 low that the boiler actually furnishes less available steam than in experiment No. 17, 
 although the rate of combustion is nearly 33 per cent, greater in the former case than in 
 the latter. This great falling-off in economic evaporation is probably to some extent 
 owing to the fact that too large an amount of air entered the furnaces relatively to the 
 weight of fuel consumed, and a better result might have been obtained if a thicker bed 
 of fuel could have been maintained or if coal of smaller size had been used. With 
 smaller coal the interstices affording passage to the entering air are more numerous, 
 narrower, and more tortuous, and a larger amount of surface of the incandescent fuel is 
 presented to the air. A great increase in efficiency has resulted in some cases from the 
 use of smaller coal with forced draught. 
 
SEC. 6. 
 
 UPTAKE, CHIMNEY, STEAM-JETS, FAN-BLOWERS, ETC. 
 
 305 
 
 w 
 
 H 
 
 tn 
 
 W 
 
 a 
 
 w 
 Q 
 
 </) 
 
 D 
 
 o 
 
 x 
 
 h 
 
 1O 
 
 vo 
 oo 
 
 ON 
 
 a s? 
 
 o 
 
 
 S 1 & 
 
 pu re^ 
 >>EJ i s ~ 
 
 JO -q 
 JOj 
 
 dsn joj 
 
 U1C31S 
 
 JO spunoj 
 
 3JKJ 
 
 D jo 
 
 aqj 
 ' 
 
 bsjadpatunsuoo 
 IBOO jo -q| jad 
 
 *-S3Jd -SOUI1B J3p 
 
 -un -j oZie tuojj 
 
 p.dEA3 J3JE.VI JO 
 
 in ise|q Joj 
 
 jo i pspusdxi AIII i is 
 -uenb ssaj 'jno'u 
 
 jad P3JBJ3U33 ,_ 
 
 meais jo spunoj o -" 
 
 Iff; 
 
 11 j i 
 
 : '3. 
 
 i i i i 
 
 iR >o 
 
 il i 
 
 \* * f= = 
 
 rf; o- o- 
 
 "jnoij J3<1 jffBjq 
 JOj papuadxa 
 meals jo spunoj 
 
CHAPTER XIII. 
 
 STEAM-ROOM AND SUPERHEATERS. 
 
 1. Capacity of Steam-room. Too large a steam-room not only increases the 
 bulk and weight of the boiler unnecessarily, but increases its heat-radiating surface. 
 It is useless to increase the steam-space beyond a certain limit for the purpose of stor- 
 ing up steam to meet any sudden demand arising from irregular loads on the engines ; 
 for a simple calculation will show that the heated water contained in the boiler can be 
 relied upon for a far greater supply of steam in case of a sudden emergency than could 
 be obtained by any admissible increase in the capacity of the steam-room. 
 
 When the capacity of the steam-room is small relatively to the quantity of steam 
 drawn from the boiler per stroke of engine, the pressure in the boiler fluctuates greatly. 
 Any sudden fall of pressure causes a violent ebullition of the water heated to the boil- 
 ing-point, producing injurious pulsations and priming. To diminish the fluctuations of 
 pressure in the boiler the latter must carry a large amount of water relatively to the 
 amount evaporated in a unit of time, and the capacity of the steam-room must be pro- 
 portioned to the capacity of the cylinders. With engines working with a high rate of 
 expansion, and making relatively few revolutions per minute, the capacity of the steam- 
 room has to be relatively greater. 
 
 Authorities differ as to the best proportions of steam-room and water-room in boilers. 
 According to Bourne ('Handbook of the Steam-engine'), the total capacity or bulk of 
 a marine boiler, exclusive of chimney, is usually about 8 cubic feet for each cubic foot 
 of water evaporated per hour, divided in the proportion of 6.5 cubic feet devoted to the 
 water, furnaces, and tubes, and 1.5 cubic feet occupied as steam-room. The capacity of 
 the steam-room of several boilers illustrated in this book will be found in Table XXII., 
 chapter vii. 
 
 In the French navy experience has developed the fact that, with rectangular boilers 
 of ordinary dimensions of the type illustrated in figure 106 and on Plate XVII., and 
 burning about 20 Ibs. of coal per square foot of grate as a maximum, when the capacity 
 of the steam-room is equal to the volume of steam consumed by the engines during 14 
 
 306 
 
SEC. 2. STEAM-ROOM AND SUPERHEATERS. 307 
 
 seconds, no water is carried over into the cylinders ; bnt when it contains steam for 
 only 12 seconds the steam is generally very wet. With cylindrical boilers working 
 with high pressures, and having a capacity of steam-room equal to the volume of steam 
 consumed by the engines during 16 seconds, water is carried over at times ; while simi- 
 lar boilers containing steam for 20 seconds give no such trouble. 
 
 When the steam-room is contained partly in a steam-drum to which the steam -pipe 
 leading to the engines is connected, the opening by which the drum communicates with 
 the steam-space within the shell of the boiler must be arranged in such a manner that 
 the steam has no tendency to enter with a violent rush, lifting the water or carrying it 
 along in the form of spray. Height of steam-space is important especially in marine 
 boilers, in which the water is frequently greatly agitated in consequence of the motion 
 of the vessel. 
 
 Height of steam-room is also necessary in order to afford vertical space for the sepa- 
 ration from the steam of the water carried up with it mechanically. This separation 
 is effected by the greater gravity of the water enabling it to fall back after being 
 carried to a certain height, so that a definite height is absolutely necessary for the 
 operation. No amount of steam-room a few inches high will enable a boiler to fur- 
 nish dry steam, while with a considerable height a comparatively small volume of steam- 
 room will be efficient. 
 
 The greater value of steam-drums upon a boiler than their volume of steam-room 
 within the shell is due to the simple fact of their greater height. The real purpose of 
 the drum is not so much to gain increased steam-room as increased height of steam- 
 room. And the wonderful efficiency of steam-chimneys, as they are called that is, an- 
 nular steam-drums enveloping concentrically the base of the chimney arises not only 
 from the superheating which the steam obtains in them, but from the very considerable 
 height given to them, whereby the water entrained by the steam has time and space to 
 become separated very thoroughly by its greater gravity. 
 
 2. Steam-drums. In rectangular boilers of naval vessels the steam-space is gene- 
 rally contained entirely within the shell, although sometimes a low steam-drum, sur- 
 rounding the uptake, is added. Cylindrical and semi-cylindrical boilers are nearly 
 always provided with steam-drums, which generally form an annular space around 
 the base of the chimney ; with this arrangement the steam-drums occupy the most 
 convenient place in the boiler-hatch, and a considerable amount of superheating- 
 surface is gained. These drums are either built directly on the top of the boiler, their 
 bottom being open to the steam-space, or they are separate. structures which are con- 
 nected by pipes, provided with stop-valves, with the steam-space within the shell of the 
 
308 STEAM BOILEES. CHAP. XIII. 
 
 boiler. In merchant-vessels the steam-drums are generally placed vertically on the top 
 of the boilers, in order to gain additional height of steam-space. In war-vessels they 
 lie horizontally in the upper spandrels formed by the cylindrical shells of each pair of 
 adjacent boilers, and are greatly less efficient. All steam-drums should be provided 
 with manholes, and should be made roomy enough to be accessible for examination and 
 cleaning. They should be provided with drain-pipes for drawing off any water which 
 may have been carried into them by priming or formed within them by the condensation 
 of the steam. 
 
 When a steam-drum or dome is built vertically on the cylindrical shell of a boiler 
 it is the usual practice to cut a hole, corresponding in diameter to the drum, in the shell 
 of the boiler, and secure the cylindrical drum to the shell by means of an angle-iron 
 ring. When the drum is not of large diameter, as in locomotive boilers, the top is 
 often made hemispherical to avoid the use of stays. It is evident that the cylindrical 
 shell is very much weakened by the large hole, unless a heavy wrought-iron strengthen- 
 ing-ring is riveted around the opening. Sometimes, instead of cutting a large opening 
 in the shell, a great number of small holes are drilled in the shell to establish communi- 
 cation between the steam-space within the boiler and the interior of the dram. The 
 perforated plate is intended to check a rush of water into the drum and to cause less 
 reduction of strength of the boiler-shell. The only addition given by it to the strength 
 of the shell is what is due to the stiffne&s of the curved plate. In other cases the shell 
 is cut away only sufficiently to allow a man to pass from the boiler into the dome, the 
 opening being made, however, so large that the rush of steam through it does not induce 
 priming. 
 
 In these cases the portion of the shell which forms the bottom of the drum is not 
 subjected to direct tension, like the rest of the shell, by the pressure of steam. Conse- 
 quently the tangential forces due to the tension on the shell tend to straighten this por- 
 tion of the shell, and to open out the lower part of the cylindrical shell of the drum, 
 and thus throw a strain on the flange of the drum at the opposite sides. To pre- 
 vent this strain the bottom of the drum must be subjected to a tension equal to that 
 on the rest of the shell, which may be effected to a certain extent by making the top of 
 the drum flat and tying it by vertical rods to the cylindrical boiler-shell forming the 
 bottom of the drum. 
 
 Straight braces may also be placed across the opening of the drum transversely to 
 the shell, thus restoring, in a measure, the strength due to the portion of the shell cut 
 out for the drum. 
 
 In small cylindrical boilers the vertical steam-domes are sometimes made with a 
 
SEC. 3. STEAM-ROOM AND SUPERHEATERS. 309 
 
 spherical top and with a contracted neck at the bottom. This neck is made of stont 
 material and with a broad flange, which serves to compensate in a great measure for the 
 loss of strength due to cutting a hole in the shell of the boiler. 
 
 3. Superheaters. The general principles according to which the theoretical and 
 the practical efficiency of superheaters are to be determined have been stated in sections 
 9 and 10, chapter iii. 
 
 In nearly all marine boilers a portion of the uptake passes through the steam-space, 
 and, in a measure, dries or superheats the steam. In all sectional water-tube boilers 
 the tubes forming the steam-space act as efficient superheaters, and in launch boilers 
 the steam passes frequently through a coiled pipe in the uptake before it is led off to 
 the engine. In the vertical fire-tube boiler the water-level is carried some distance 
 below the upper tube-sheet, and the upper ends of the tubes and the uptake form 
 superheating-surfaces. In the vertical water-tube boiler of the Martin type, likewise, 
 the water may be earned with safety several inches below the upper tube-sheet, and 
 efficient superheating-surface will thus be gained. The vertical steam-drums of marine 
 boilers are commonly traversed by one or several large flues for the purpose of drying 
 or superheating the steam. In some cases the steam-drum surrounding the base of the 
 chimney is divided by partitions into several compartments, which communicate with 
 one another by openings at opposite ends in such a manner that the steam has to pass 
 in succession through all the different compartments before it enters the steam-pipe, 
 and thus is forced to remain a longer time in contact with the superheating-sur- 
 face. 
 
 In the high-pressure cylindrical boilers of United States naval vessels superheating- 
 surface is provided by letting the steam-pipe pass several times through the whole 
 length of the uptakes along the fronts of the boilers. 
 
 By means of flues traversing or surrounding the steam-drum or the upper part of 
 the boiler the steam may be effectually dried and its temperature may be raised to a 
 point exceeding somewhat, but not very much, the boiling-point corresponding to the 
 pressure in the boiler. When a much higher temperature is to be given to the steam 
 the superheating has to be effected in a separate chamber containing a larger amount of 
 heating-surface than can be obtained conveniently by the above-mentioned arrange- 
 ments. 
 
 The superheater of TJ. S. S. Plymouth (see Plate XYII.) consists of a box extending 
 along the front of the boiler and traversed by numerous vertical brass tubes, through 
 which the products of combustion pass from the front-connections to the uptake. At 
 one end of the box the saturated steam is admitted from the boiler through suitable 
 
310 STEAM BOILERS. CHAP. XIII. 
 
 pipes and stop- valves, and at the other end the superheated steam is carried off to the 
 main steam-pipe. 
 
 In other superheaters horizontal pipes are used, and these are often arranged in two 
 groups connected at the ends by chambers in such a manner that the steam passes 
 twice through the whole length of the superheater, entering through one group and re- 
 turning through the other, before it is discharged into the steam-pipe. 
 
 Superheaters constructed of flat plates, after the plan of Lamb and Sumner's boiler 
 (see section 1, chapter xi.), have been much used in England. Many other devices 
 have been tried with the view of gaining a large amount of efficient heating-surface in a 
 cheaply- constructed apparatus. 
 
 The superheaters are generally arranged in the uptake of the boiler, so as to utilize 
 some of the heat of the escaping gases. It is evident that when a high degree of super- 
 heating is desired considerable difference must exist between the temperatures of the 
 gases in the uptake and of the saturated steam in the boiler ; and to obtain this dif- 
 ference the water-heating surface of the boiler must be made small relatively to the 
 amount of coal burnt in a unit of time. Besides, the additional resistance offered by the 
 superheating apparatus to the escaping gases makes it necessary that the chimney tem- 
 perature should be correspondingly increased, in order to maintain the same, rate of 
 combustion as without the superheating apparatus. 
 
 A great increase of efficiency has been obtained in cases where such superheaters 
 have been added to boilers already built which were subject to priming or, being defi- 
 cient in heating-surface, discharged the gases at a higher temperature than was required 
 for the desired rate of combustion. The saving in fuel expended for a given amount of 
 work, effected by the introduction of superheating apparatus, has amounted, in a num- 
 ber of cases cited by Bourne ('A Treatise on the Steam-engine'), to from 18 to 34 per 
 centum. 
 
 Superheaters of the foregoing description, with their steam-pipe connections and 
 stop-valves, add largely to the weight and cost of boilers ; and unless they are easily 
 accessible for sweeping (which is frequently not the case), the efficiency of their heating- 
 surface is soon impaired and the draught of the boiler is often seriously affected by the 
 accumulation of soot. The most serious troubles, however (which have brought super- 
 heaters somewhat into disrepute), are due to the rapid corrosion of the iron of which 
 superheaters are constructed, and to the leakage of their tubes. The rapid corrosion of 
 superheating-surfaces has been observed for a long time, even in the case where the de- 
 gree of superheating was relatively small, as in steam-drums traversed by flues, but its 
 causes have not been definitely determined. The leakage of the tubes after short use is 
 
SEC. 3. STEAM-ROOM AND SUPEKHEATEBS. , 311 
 
 probably mainly owing to the fact that, after the fires are lighted, some time elapses 
 before steam is formed in the boiler, and the hot gases passing through the empty 
 superheater raise the metal to an unduly high temperature. To remedy this defect it is 
 proposed to keep the superheater filled with water until steam is formed in the boiler. 
 
 Some years ago many United States naval vessels were furnished with special super- 
 heating-boilers, one being provided for each pair of main boilers containing in all four- 
 teen furnaces. Each superheating-boiler contained one furnace of the usual dimen- 
 sions. The products of combustion, after passing through return-flues, situated over 
 the furnace, to a front-connection, passed through a set of horizontal iron tubes to an 
 upper back-connection, and returned thence through another set of like tubes to an 
 upper front-connection, which communicated with the uptake of the main boilers. The 
 lower flues of the boiler were kept covered with water, and the double-return tubes fur- 
 nished the superheating-surface. The superheating-boiler communicated with the 
 main steam-pipes through stop- valves, arranged in such a manner that the steam from 
 the main boilers would either pass wholly or in part through the superheaters, or go 
 directly to the engines. These boilers were efficient superheaters ; but they were 
 rapidly destroyed by corrosion, because their interior was inaccessible and the iron 
 superheating-tubes were left without the coating of scale which protects the iron water- 
 heating tubes of marine boilers effectually. In the TJ. S. S. Congress each of these 
 superheating-boilers was 3 feet 10 inches wide, 10 feet 3 inches long, and 9 feet 9 inches 
 high, and its weight complete was 19,000 Ibs. The superheating effected was about 30 
 Fahr. above the saturation temperature. 
 
 Independent superheating-boilers possess the advantage that any degree of super- 
 heating may be obtained in them with great exactness by regulating the rate of com- 
 bustion in the furnace ; furthermore, the efficiency of the main boilers is entirely inde- 
 pendent of the efficiency of the superheaters ; any derangement of the latter does not 
 affect the former. The additional space occupied by them is an important element in 
 determining their relative usefulness ; but in naval vessels it may often be advisable to 
 sacrifice room on the floor of the vessel in order to get a lower and a more reliable 
 boiler. 
 
 The superheating arrangement of the U. S. S. Eutaw (built in 1863) consisted of two 
 groups of horizontal tubes, with the ends secured in tube-plates. Each group con- 
 tained 176 iron tubes 1| inches in diameter and 21 inches long between the tube-plates. 
 At one end the two groups communicated by means of a common chamber formed by 
 an iron casting bolted to the tube-plate, and at the other end each group had a separate 
 connection formed by an iron casting provided with a nozzle and likewise bolted to the 
 
312 STEAM BOILERS. CHAP. XIII. 
 
 tube-plate. This superheater was placed in the tube-box of one of the wing furnaces 
 of each boiler, from which the vertical water-tubes had been removed with the excep- 
 tion of six rows at the back of the box, left for the purpose of reducing somewhat the 
 temperature of the gases before they impinged on the superheating-tubes. The tubes 
 of the superheater were placed across the tube-box, and the nozzles of the connections 
 projected through openings in the side of the boiler. To the nozzle nearest the front 
 of the boiler a pipe bringing the saturated steam to the superheater was bolted, and to 
 the other nozzle another pipe was bolted carrying the superheated steam to the main 
 steam-pipe. These pipes were' all controlled by stop-valves, so that the superheater 
 could be shut off when the fire was hauled from the furnace, and the engine could be 
 supplied with saturated steam, or superheated steam, or a mixture of saturated and 
 superheated steam. 
 
 Each of the two boilers contained five furnaces, and the two boilers contained in the 
 aggregate 200 square feet of grate-surface, 4,536 square feet of water-heating surface, 
 and 1,058 square feet of superheating-surface. 
 
 Using natural draught and burning 11.67 Ibs. of anthracite coal per square foot of 
 grate per hour, the temperature of the steam was raised from 270.2 Fahr. when satu- 
 rated to 365.0 Fahr. when superheated. 
 
 Using a fan-blower and burning 27 Ibs. of anthracite coal per square foot of grate per 
 hour, the temperature of the steam was raised from 295.0 Fahr. when saturated to 
 380.0 Fahr. when superheated. 
 
 One of the greatest practical objections to the use of separate tubular superheaters 
 as described is the impossibility of cleaning the steam side of the surfaces. The inte- 
 rior of the tubes or the spaces between the tubes, as the construction may be, become 
 filled with the mud and grease carried over from the boiler by priming or foaming, and 
 there is no means of removing these substances without destroying the superheater. 
 
CHAPTEK XIV. 
 
 SETTING AND ERECTION OF BOILERS. 
 
 1. Setting of Boilers. The weight of large boilers must be well distributed over 
 the floor of the vessel. When the boilers are placed so that the fire-room runs in the 
 fore-and-aft direction of the vessel they rest generally on two keelsons. To protect the 
 bottom of the boilers from the bilge- water a tight platform is often built on the keel- 
 sons for the boilers to rest on. 
 
 The following directions are given by Bourne for setting flat-bottomed boilers in 
 wooden vessels : "In the setting of marine boilers care must be taken that no copper 
 bolts or nails project above the wooden platform upon which they rest, and also that 
 no projecting copper bolts in the sides of the ship touch the boiler, as the galvanic 
 action in such a case would probably soon wear the points of contact into holes. The 
 platform may consist of three-inch planking laid across the keelsons, nailed with iron 
 nails the heads of which are well punched down, and calked and puttied like a deck. 
 The surface may then be painted over with thin putty, and fore-and-aft boards of half 
 the thickness may then be laid down and nailed securely with iron nails having the 
 heads well punched down. This platform must then be covered thinly and evenly with 
 mastic cement and the boiler be set down upon it, and the cement must be calked be- 
 neath the boiler by means of wooden calking-tools so as completely to fill every vacuity. 
 Coamings of wood sloped on the top must next be set round the boiler, and the space 
 between the coamings and the boiler must be calked full of cement, and be smoothed 
 off on the top to the slope of the coamings, so as to throw off any water that might be 
 disposed to enter between the coamings and the boiler." 
 
 Ledieu gives the following compositions for the cement used in setting boilers viz., 
 equal quantities of whale-oil, ox-blood, and powdered unslaked lime ; or, Spanish white, 
 oil, and a small quantity of cow-hair. 
 
 Hamelin's mastic cement for the setting of boilers is compounded as follows: "To 
 any given weight of sand or pulverized earthenware add two-thirds such given weight 
 of powdered Bath, Portland, or similar stone, and to every 560 Ibs. weight of the mix- 
 ture add 40 Ibs. weight of litharge, 2 Ibs. of powdered glass or flint, 1 Ib. of minium, 
 
 313 
 
314 STEAM BOILERS. CHAP. XIV. 
 
 and 2 Ibs. of gray oxide of lead ; pass the mixture through a sieve and keep it in a 
 powder for use. When wanted for use a sufficient quantity of the powder is mixed with 
 some vegetable oil upon a board or in a trough in the manner of mortar, in the propor- 
 tion of 605 Ibs. of the powder to 5 gallons of linseed, walnut, or pink oil, and the mix- 
 ture is stirred and trodden upon until it assumes the appearance of moistened sand, 
 when it is ready for use. The cement should be used on the same day that the oil is 
 added, else it will set into a solid mass." (Bourne.} 
 
 With boilers set on a platform in the above-described manner the cement is apt to 
 crack after a while and become detached from the shell of the boiler, and when leaks 
 occur in the bottom of the boiler the water spreads over the platform and corrosion 
 takes place over a large surface unnoticed. Access to the bottom of the boiler can be 
 had only by cutting away portions of the platform from below, and it is a difficult mat- 
 ter to locate a leak in the bottom of the boiler. For these reasons it is thought prefer- 
 able to omit the platform and let the boiler rest directly on keelsons, a clear passage, ex- 
 tending the whole length of the boiler, being left between the keelsons, so that the bot- 
 tom of the boiler may be examined, cleaned, painted, and repaired. The boiler should 
 be placed so that the seams connecting the front and back to the bottom are acces- 
 sible for calking, and for removing and replacing rivets. Boilers must never be set 
 directly on oak or other wood containing acids which corrode iron, but a cap-piece of 
 pine rrmst be spiked to the top of the keelson. 
 
 The water-legs of dry -bottom boilers are set on cast-iron frames or saddles. Figures 
 3 and 4, Plate XXXI., illustrate the form of saddles used in United States naval vessels. 
 The ashpan and half of the saddles for the sides and back of each furnace are fre- 
 quently cast in one piece. When they are made in separate pieces they are more 
 easily handled in case it is necessary to remove them for the purpose of examining or 
 repairing the water-legs without moving the boiler from its seat. At places where laps 
 or rivets occur on the water-legs the top flange of the saddle is cored out to clear them. 
 All spaces between the water-leg and the upper flange of the saddle are filled with 
 cement, so that no water can lodge there. When dry-bottom boilers are used in a 
 wooden vessel precautions have to be taken to prevent the keelsons and the lining of 
 the vessel being set on fire by the heat radiated through the interstices of the grates or 
 by the fire which falls in cleaning or hauling the fires. In several United States nav:il 
 vessels the wooden keelsons are protected by cast-iron cap-pieces on which the saddles 
 rest ; an air-space is formed under the cast-iron ashpan by corrugated wrought-iron 
 plates, i inch thick, resting on the keelsons and on ledges formed on the rib which 
 stiffens the bottom of the ashpan. 
 
SEC. 2. SETTING AND ERECTION OF BOILERS. 315 
 
 The arrangement shown in figure 4, Plate XXXI., was designed for U. S. S. Yantic. 
 The bottom of each ashpit is formed by a shallow tank 4 inches deep, built of f -inch 
 wrought-iron plates and stiffened by f-inch socket-rivets placed 9 inches apart. This 
 tank is filled with cement, which must be poured in from the side and not from the 
 top. This ashpan rests at its four corners on cast-iron blocks 8 inches square, which 
 are secured by a large wood-screw to the keelsons. The saddles rest on the top of the 
 ashpan on wrought-iron strips, and the sides and the back of the saddles are made in 
 separate pieces. This arrangement allows any of the saddles to be removed and re- 
 placed with ease without raising the boiler from its seat ; and after removing any of the 
 saddles the corresponding tank on which they rest may likewise be pulled out and 
 replaced. 
 
 Cylindrical boilers rest on saddles, as shown on Plate XXX. and in figure 2, Plate 
 XXXI. These saddles are either made of cast-iron or are built up of wrought-iron plates 
 and angle-irons. In wooden vessels they are secured by holding-down bolts to the keel- 
 sons ; in iron vessels they are frequently riveted to the frames of the hull. 
 
 2. Securing Boilers. To prevent the boilers from shifting or moving in conse- 
 quence of the violent motions of the vessel in a sea-way, they are securely tied to the 
 hull of the vessel, and when there are several boilers arranged in pairs they are tied at 
 the top to one another. No part of the boilers rigidly connected with the shell should 
 be attached to the decks of the vessel ; on the contrary, sufficient clearance must be left 
 between the deck-beams and hatch-framing and the boilers to allow for the working of 
 the ship when it is severely strained. The longitudinal or pitching motion of large ves- 
 sels is not sufficiently violent to necessitate special precautions for holding flat-bottomed 
 boilers in their seat. To guard against the effect of the transverse or rolling motion of 
 the vessel the boilers are tied down to the floor of the vessel by wrought-iron straps 
 passing diagonally up the sides of the boilers. These straps are secured to the shell by 
 bolts and nuts ; the bolts must fit accurately the holes in the straps and in the shell. 
 
 When boilers of the type illustrated on Plates VI., VII., and XVII., in which the 
 xiptake forms an integral part of the shell, are arranged in pairs opposite to one another, 
 they are sufficiently tied together at the top by their uptakes. When the uptakes are 
 separate structures, built on the shell of the boilers, the latter are generally tied to one 
 another at the top* by straps or braces. In U. S. S. Trenton each pair of opposite 
 boilers is tied together near the top by two wrought-iron straps, 1 inch thick and 4 
 inches wide, extending across the fire-room immediately below the uptakes. Each end 
 of these straps is secured to the cylindrical shell of the boilers by four bolts IJ inches in 
 diameter. The cylindrical boilers of U. S. S. Nipsic (see Plate XXX.) are secured to 
 
316 STEAM BOILERS. 
 
 CHAP. XIV. 
 
 their saddles by turned bolts passing through reamed holes in the shell ; the saddles are 
 held by composition bolts passing through the frames of the hull. 
 
 In other vessels each cylindrical boiler is held down in the saddles by four wrought - 
 iron braces, one being placed near the front and another near the back at either side of 
 the boiler. The lower end of each brace is bolted either directly to the hull of the ves- 
 sel or to the saddle, and the upper end is secured by a nut to a lug bolted to the cylin- 
 drical shell of the boiler. 
 
 3. Erection of Boilers in the Vessel. In preparing the bed on which the 
 boilers or their saddles are to rest, the boiler-keelsons are dubbed off to lines marked on 
 their sides representing the intersections of the plane of the bottom of the boilers or 
 their saddles with the keelsons. To find the traces of this plane stretch two lines, one 
 at the after end and one at the forward end of the boiler-bed, at a determined height 
 above the top of the main keelson, and at right angles with the horizontal centre line of 
 the main keelson and with perpendiculars drawn from the centre line of the deck to the 
 centre line of the main keelson ; then measure from the transverse lines the distance at 
 which the bottom of the boilers or their saddles should be placed below the horizontal 
 plane passing through these two transverse lines, and mark the points thus found on 
 the sides of the boiler-keelsons. 
 
 In case the boilers are to be set on a platform allowance is to be made for the thick- 
 ness of the planking and of the bed of cement, and the platform is built on the keelsons 
 after these have been dubbed off to the proper height. 
 
 The distance of the centre line of the boilers from the centre line of the engines is 
 measured along the centre line of the main keelson according to the dimensions given 
 on the drawing, and is marked on the keelsons by a transverse line perpendicular to the 
 centre line of the main keelson. The location of each boiler is then determined by 
 measuring the distances given on the drawing from the centre line of the main keelson 
 and from the common centre line of the boilers, and the points thus found are marked 
 on the boiler-keelsons or the platform. 
 
 In case saddles are used for the boilers, they are now placed in position. 
 
 After these preparations are made the boilers are put aboard by means of a crane or 
 shears. When the boilers are slung in chains the corners of the shell must be protected 
 by chafing-gear of wood and mats. The boilers are put aboard without the furnace and 
 uptake doors, grate-bars, valves, and other removable attachments ; but it is well to 
 keep the manholes and handholes and similar openings closed, in order that if the 
 slings should give way and the boiler should fall overboard it would not fill with 
 water. 
 
SEC. S. SETTING AND ERECTION OF BOILERS. 317 
 
 The decks are left open sufficiently to allow the boilers to pass through them ; and 
 the deck-framing around the boiler-hatches is arranged in such a manner that it can be 
 taken up without disturbing the deck-beams when the boilers are to be hoisted out at a 
 future time. 
 
 The first boilers to be put aboard are those which are to be situated- at the extreme 
 forward and after ends of the fire-room. As the boiler is lowered down into the hold of 
 the vessel it is placed on blocks or rollers ; the slings passing around the shell are then 
 cast off, and the boiler is moved by means of tackle, jacks, or other appliances into its 
 proper position corresponding to the marks on the keelsons, and then it is gradually 
 lowered from the blocks into its seat. The correctness of the position of boilers of the 
 type represented on Plates VI., VII. , XVII. is verified by seeing that the upper por- 
 tions of the uptakes, which form the base of the chimney, meet properly, and that the 
 centre of the circle formed by their cross-section falls on a line stretched from the centre 
 of the boiler-hatch to the centre line of the main keelson. 
 
 
 
 When the boilers are to be set in cement on a platform, they must now be raised by 
 jacks from their seat sufficiently high to allow men to crawl under them and spread the 
 cement evenly over the platform ; during this process the boilers are supported at the 
 four corners by blocks of wood. After the bed of cement has been laid the boilers are 
 lowered carefully into position. 
 
 Boilers which are to rest on saddles are first placed on blocking directly over their 
 saddles ; after the correctness of their position relatively to each other and to their 
 saddles has been verified they are lowered carefully into their seat. 
 
 The straps which are to tie the boilers to the hull of the vessel and to one another are 
 now fitted and secured to them, and the uptakes which are built in the boilers are 
 riveted together where they meet at the top. With cylindrical boilers the construction 
 of the smoke-boxes and uptakes commences now. The grates, doors, valves, and other 
 fixtures are attached to the boilers as soon as these are placed permanently in position. 
 The pipes connecting the boilers with the outboard- valves, the pumps, and the main en- 
 gines are next put up ; the exact length, shape, and position of these pipes is deter- 
 mined by making board templates after the boilers are placed in position, care being 
 taken that the cocks and valves attached to them are accessible, that one pipe may be 
 taken down without removing another one, that the bends of the pipes form easy curves, 
 and that the pipes follow the most direct course compatible with the foregoing con- 
 ditions. 
 
 Finally, the felting and other covering is placed over the boiler-shells and the 
 pipes. 
 
318 STEAM BOILERS. 
 
 CHAP. XIV. 
 
 As soon as the uptakes are constructed, and the deck around the boiler-hatch has 
 been completed, the chimney and the escape-pipe are hoisted aboard and placed in 
 position, and the hatch-gratings, plates, ventilators, etc., are put up. In slinging the 
 smoke-pipe for hoisting it must be stiffened by temporary wooden stays placed inside, 
 or by boards placed outside between the slings and the pipe. 
 
CHAPTER XY. 
 
 BOILER MOUNTINGS AND ATTACHMENTS. 
 
 1. Grate. The grate of furnaces in which coal is burnt is composed of alternate 
 bars and spaces. 
 
 In many boilers the front part of the grate is formed by a horizontal or slightly in- 
 clined iron plate without perforations, about 20 inches long ; this is called the dead- 
 plate or dumb-plate. It was introduced by Watt, and is used especially in furnaces 
 where bituminous coal is used as fuel. In firing the coal is thrown first on the dead- 
 plate, where the radiant heat of the fire volatilizes the hydrocarbons ; and after the coal 
 is thus reduced to coke it is pushed inwards and spread over the fire. In the boilers 
 of United States naval vessels the dead-place is omitted. 
 
 The grate-bars are ordinarily placed lengthwise the furnace and rest on supports at 
 the front and back of the furnace, and, when the grate is long, on one or two interme- 
 diate cross-bars. The bars must be strong enough to bear the weight of the fuel and to 
 withstand the rough usage to which they are unavoidably subjected in working the fire. 
 They must rest securely on their supports, but must be allowed to expand and contract 
 freely with the great variations of temperature to which they are exposed. 
 
 The overheating of the bars is prevented by the rapid currents of air rushing to the 
 bed of fuel through the spaces between the bars, and by a thin layer of ashes accu- 
 mulating on the top of the bars. The overheating of bars may be due to their faulty 
 form, or to obstructions in the spaces preventing the free inflow of air, or to the intensity 
 of the fire ; in such cases the bars will bend and warp and partially melt on the top. 
 
 Coals containing sulphur or forming easily-fused clinker destroy the bars rapidly ; 
 the clinker sticks between the bars and obstructs the air-passages. 
 
 The top of the grate should always form a level surface flush with the bottom of the 
 opening of the furnace-door ; the bars which project above the level of the grate are 
 liable to be burnt and to be displaced in working the fire. 
 
 Grate-bars have been made hollow to allow a current of air or water to pass through 
 them, for the purpose of increasing their durability and adding to the efficiency of the 
 
 319 
 
320 STEAM BOILERS. CHAP. XV. 
 
 furnace ; but the cost of such contrivances and the difficulty of keeping them in order 
 have caused their rejection. 
 
 In marine boilers the grate generally slopes downward from the furnace-mouth to 
 the bridge- wall. By this means the back of the grate is more easily kept covered with 
 fuel, and the coal is prevented in a measure from falling out of the furnace into the fire- 
 room when the ship rolls and the door is open. The rate of this slope varies from one 
 in ten to one in twenty, and is limited by the heights required over the top of the grate 
 at the furnace-mouth for proper firing, and below the grate at the bridge for admitting a 
 sufficient amount of air and for working the fire from the ashpit. 
 
 Grate-bars are usually made of cast-iron. Wrought-iron bars are frequently used 
 in the boilers of locomotives, and are also not unf requently used in marine boilers ; they 
 bend and warp easily, but can be straightened ; they are not so easily broken or fused 
 and burnt as cast-iron bars, and they may be made somewhat lighter than the latter. 
 For these reasons wrought-iron bars are probably cheaper in the end, although their 
 first cost is greater than that of cast-iron bars. Wrought-iron bars are most simply 
 made by riveting two plain bars together, with thimbles between them for distance- 
 pieces at the ends and in the middle, and letting the heads of the rivets determine the 
 width of the space between two adjacent double bars. 
 
 The length of grate-bars should not much exceed three feet. According to Ledieu, 
 the grate-bars of French naval boilers are usually made of wrought-iron, and 29 inches 
 or 21f inches long, according to the length of the grate. Short bars are more easily 
 handled than long ones, and are twisted less out of shape by overheating. 
 
 Fi 133 _ Various shapes have been given to grate-bars 
 
 fatiaaaaalfail mainly with a view to increase their durability. 
 nirijljliiijl Figure 133 represents the top and bottom views of 
 uuiluuujj a grate-bar in common use, designed to be free from a 
 tendency to warp on account of its peculiar shape. 
 
 The cast-iron bars generally used in United States naval boilers are illustrated on 
 Plate XIX. These grate-bars are usually made from f inch to f inch wide on the top. 
 At the bottom they should be made as thin as they can be cast, and the necessary 
 strength should be obtained by proportioning their depth to their length. Thin and 
 deep bars are less liable to warping than thicker and less deep bars, because the inflow- 
 ing air abstracts the heat more readily from them. Grate-bars are generally made 
 from J inch to f inch thick at the bottom, and from 3, inches to 3f inches deep in the 
 middle of their length. The outline of the bottom has approximately the form of a 
 parabola. 
 
SBC. 1. BOILER MOUNTINGS AND ATTACHMENTS. 321 
 
 Grate-bars have sometimes a uniform width, for a depth of about f inch from the top, 
 and below that depth a rib of diminished but uniform thickness ; it is, however, prefer- 
 able to let them taper evenly from the top to the bottom, as such a form facilitates the 
 flow of air to the fuel, the fall of refuse matter through the grate, and the pricking of 
 the fire from below. 
 
 At each side of the ends of the bars projections are formed which determine the width 
 of the spaces between them. When the length of bars exceeds 30 inches similar pro- 
 jections are formed midway between their ends to increase their lateral stiffness. Grate- 
 bars are generally made double, so that two bars with the proper space between them 
 form one piece ; this saves time in removing and replacing them, and increases greatly 
 their stiffness. A number of single bars are provided with the double bars, so that the 
 whole width of the furnaces may be filled by the bars without jamming them and with 
 the proper spaces between them. 
 
 The following considerations govern the width of the spaces between the bars : a suf- 
 ficient quantity of air must be admitted to the fuel ; the spaces must not be obstructed 
 too easily by clinkers ; the prick-bar must pass through them to free the bottom of the 
 bed of fuel from ashes ; on the other hand, the coal used as fuel must not drop through 
 them, therefore small, free-burning coal requires narrower spaces than lump coal and 
 caking coal. The width of the clear space between two bars is usually T \ inch or 
 i inch when good semi-bituminous coal is used ; with anthracite, and with coals that 
 cake much or yield large quantities of ash and clinker, the space is made inch wide, 
 and sometimes even more. 
 
 All cast-iron bars used in United States naval vessels have a shallow groove on the 
 top ; the ashes which accumulate in these grooves prevent clinkers from adhering to the 
 bars, and the latter are less easily burnt. 
 
 The grate-bars rest with their ends on cross-bearers or bearing-bars, and they are 
 allowed to expand freely in the direction of their length. The front end of each bar is 
 often provided with a lug at the bottom, which hooks on the bearing-bar and limits the 
 motion of that end of the bar ; this lessens the chance of the bar sliding off its seat 
 when it becomes shortened by warping. The space allowed for expansion at the ends of 
 bars is frequently insufficient ; it should not be less than the width of the air-spaces be- 
 tween the bars. 
 
 To prevent coal or clinker from lodging tightly between the ends of bars the ends 
 are often made slanting, either at the top or at the bottom, instead of square. Some- 
 times one end of the bars is made tapering and rests on an inclined seat (see Plate XY.) ; 
 this arrangement allows the bars to expand freely ; but the bars are apt to be raised 
 
322 
 
 STEAM BOILERS. 
 
 CHAP. XV. 
 
 above the level of the grate, and, in consequence, to be burnt or become displaced in 
 cleaning the fires. 
 
 The bearing-bars are made either of cast-iron or of wrought-iron, and rest with their 
 ends on lugs attached by bolts to the sides of the furnace (see Plate XIX.) 
 
 The middle bearing-bar is made double, with a wide space to let ashes and clinker 
 fall through. 
 
 The front bearing-bar is secured by a few large bolts to the furnace door-frame ; or, 
 when a dead-plate is provided, it serves as a support for the front end of the grate- 
 bars. 
 
 The back bearing-bar rests on lugs bolted to the water-bridge, and is provided with a 
 lug at each end in order to maintain an air-space between the bar and the bridge-wall. 
 When the bridge-wall consists of a separate iron frame supporting a wall of fire-brick 
 the back end of the grate-bars rests on a ledge formed on the frame of the bridge-wall. 
 
 2. Moving-grates. In order to diminish the labor of attending to the fires and the 
 
 Fig. 134, 
 
 loss in efficiency of boilers due to the open- 
 ing of the furnace-doors for supplying the 
 fuel and cleaning the fires, various contri- 
 vances have been made for supplying fuel 
 to furnaces evenly and continuously by 
 mechanism. Circular revolving grates 
 turning slowly about their centre, and 
 grates consisting of an endless web of short 
 bars moving on horizontal rollers, and tra- 
 velling from the furnace-mouth to the 
 bridge and returning through the ashpit, 
 have been used ; but none of these contri- 
 vances have been successfully applied to 
 marine boilers. 
 
 In other forms of moving-grates ap- 
 plied to marine boilers a short, reciprocat- 
 ing motion up and down, or from side to 
 side, may be given to the grate-bars, in 
 
 order to keep the grate clear of ashes and clinker without opening the furnace-door 
 
 and using fire-tools. 
 
 The Murphy shaking-grate (see figure 134) consists of alternate stationary and 
 
 vibrating bars placed at right angles to the length of the furnace. There are two sets of 
 
SEC. 2. BOILER MOUNTINGS AND ATTACHMENTS. 323 
 
 bars, forming two inclined planes, sloping downward from the sides of the furnace to 
 the centre. The stationary bars rest with their upper ends against the sides of the fur- 
 nace and with their lower ends on a cast-iron frame consisting of two parallel bars 
 extending through the length of the furnace, supported by suitable brackets. The 
 vibrating bars are pivoted at their upper ends to the upper ends of the stationary 
 bare, and their lower ends rest against the continuous feather of a horizontal bar lying 
 lengthwise the furnace in pillow-blocks placed at the front and back of the furnace ; 
 there are two of these horizontal bars, one for each side of the grate. By rocking the 
 horizontal bar forward and back on its axis by means of a lever attached to the end 
 protruding from the ashpit into the fire-room, the pivoted bars receive a vibrating mo- 
 tion, their lower ends being forced alternately above and below the level of the station- 
 ary bars. 
 
 At the bottom of the two inclined planes of the grate, and entirely independent of it, 
 is a horizontal bar, 3 inches in diameter, lying lengthwise the furnace in five pillow- 
 blocks attached to the central frame. This bar bristles with eight rows of projecting 
 teeth, forming cubes of one inch a side. This bar may be rocked forward and back, or 
 revolved entirely around its axis, by a lever attached to the end protruding into the 
 fire-room. This is called the clinker-crusher and refuse-remover. 
 
 In firing the coal is thrown along the upper ends of the grate-bars solely, and slides 
 downward by gravity over the grate-surface as the coal on the latter is consumed. An 
 occasional shaking of the grate by the vibrating bars accelerates the descent of the coal, 
 removes from it such refuse as can fall between the grate-bars, and prevents the occur- 
 rence of holes in the fire. Such of the refuse from the coal as cannot pass through the 
 spaces between the grate-bars slides down the grate-surface by gravity, and is broken up 
 and worked into the ashpit by the vibratory motion of the clinker-crusher. The fires 
 are thus kept clean and free of holes without the employment of fire-tools, and the fur- 
 nace-doors are only opened for throwing in the coal. 
 
 The weight of this grate is about double the weight of the ordinary grate. 
 
 Experiments made by a board of United States naval engineers to determine the 
 efficiency of the Murphy grate showed that, compared with the ordinary grate, the 
 economic gain due to its use was in direct proportion to the per centum of refuse re- 
 moved through the furnace-door with the ordinary grate in use. (See section 13, chap- 
 ter ii., and ' Report on the Murphy Grate-bar, by a Board of United States Naval 
 Engineers, June 25, 1878.') 
 
 The Martin or Ashcroft grate (see figure 135) is formed by wrought-iron bars, 1 
 inches square in cross-section, extending the whole length of the fiirnace and projecting 
 
324 STEAM BOILERS. CHAP. XV. 
 
 5 or 6 inches beyond the front of the boiler. These grate-bars rest on wrought-iron 
 bearing-bars placed about 18 inches apart and supported at the ends by lugs bolted to 
 the sides of the furnace. The upper side of the bearing-bars is either bevelled to a 
 knife-edge or it is crenated into semicircles, in which the grate-bars rest and from 
 which they cannot be displaced in turning. The bars are revolved by means of a 
 socket- wrench applied to their ends, and the fire is thus to be cleaned from ash and 
 clinker without opening the furnace-door ; but no thorough cleaning of the fires can be 
 made by merely revolving the grate-bars, which by a few turns cut out grooves in the 
 fires and leave the coherent mass above untouched. (See ' JZeport on AsTicroft Furnace- 
 doors and Grate-bars, by a Board of United States Naval Engineers, March 27, 1878.') 
 
 3. Bridge-walls. The bridge is a wall or partition near the back of the furnace 
 which limits the length of the grate and forms an abutment for the bed of fuel. The 
 height of the bridge-wall regulates the area of the passage leading from the furnace to 
 the combustion-chamber or back-connection, and affects greatly the efficiency of boilers, 
 through the influence which it exerts on the amount of air admitted to the furnace and 
 on the thorough mixing of the gases during combustion. (See section 5, chapter mi.} 
 
 The bridge is often formed by a hollow wall communicating with the water-space of 
 the boiler and forming an integral part of the latter (see section 7, chapter ix.), or it is 
 formed by a wall of fire-brick. Sometimes a solid mass of fire-brick fills the back of the 
 furnace behind the grate ; but this arrangement is objectionable, because it hides leaks 
 which may occur there. The bridge-wall consists usually of a vertical cast-iron frame 
 resting on the bottom and bolted by means of flanges or straps to the sides of the fur- 
 nace, or of a horizontal plate extending from the back of the grate to the back of the 
 furnace, and resting on brackets or angle-irons bolted to the sides of the furnace ; this 
 frame or plate supports the back end of the grate-bars and, behind them, a wall of fire- 
 brick (see Plate XII.) 
 
 The bridge-wall is often provided with openings for admitting jets of air to the com- 
 bustion-chamber or back-connection, the supply of air being regulated by a register or 
 by a hinged valve, the position of which can be adjusted by means of a rod and suitable 
 connections from the fire-room (see Plate XV.) Such openings in the bridge- wall are 
 now always omitted in the boilers of United States naval vessels, since repeated experi- 
 ments have demonstrated that an air-admission to the back-connection produces no use- 
 ful results when anthracite coal is used as fuel. In English and French boilers designed 
 to burn bituminous coal the bridge-wall is always provided with openings for air-admis- 
 sion. Specifications issued by the Admiralty for boilers for English naval vessels 
 require an aggregate area of not less than three square inches of opening in the bridge 
 
SEC. 4 BOILER MOUNTINGS AND ATTACHMENTS. 325 
 
 for the admission of air to the back-connection for each square foot of grate-surface. 
 (See section 11, chapter mi.} 
 
 4. Fumaee-doors and Door-frames. The furnace-door opening in marine boil- 
 ers having grates of the usual dimensions is ordinarily from 14 inches to 16 inches high 
 and from 18 inches to 20 inches wide, the upper part being arched and the lower part 
 square. 
 
 In the rectangular boilers of United States naval vessels the furnace-door openings 
 are constructed in the front wall of the boiler in the manner shown on Plates VI., VII., 
 and XVII. When cast-iron furnace-doors are used the opening is surrounded by a cast- 
 iron frame bolted by five or six |-inch bolts to the boiler-front. On this frame are cast 
 the catch and the hinges for the door and the sill-plate ; the frame is fitted to the boiler- 
 front by means of a narrow chipping-strip surrounding its outer circumference, and 
 around the door-opening it has another chipping-strip, against which the door is 
 fitted. 
 
 When the furnace-doors are made of wrought-iron, door-frames are dispensed with, 
 the doors being made to fit directly against the boiler-front and the sill-plate. The 
 latter is made either of wrought-iron or cast-iron, and is bolted in place as shown on 
 Plate XIX. The top of the sill-plate is level with the top of the front of the grate, and 
 it has a flange to which the front bearing-bar is bolted or riveted. 
 
 When the furnace is secured to the boiler-front in the manner shown in figure 96 
 and on Plates VIII. , XL, and XII., a frame is bolted in the furnace-mouth, to which 
 the furnace-door is fitted (see Plate XXIX.) This frame consists of a single casting 
 about 6 inches deep, to which at the front and back wrought-iron plates f inch thick 
 are bolted, thus forming a double wall which intercepts the heat radiated from the 
 incandescent fuel. The front plate is provided with a few holes about 1 inch in diame- 
 ter, and the back plate is perforated with numerous holes J inch or ^ inch in diameter. 
 The catch and hinges for the furnace-door are riveted to the front plate of the door- 
 frame. 
 
 The wrought-iron furnace-doors of boilers of United States naval vessels consist of a 
 front plate provided with about fifteen 1-inch holes, and stiffened by a lip turned up 
 around its circumference which fits against the boiler-front ; and of a back plate, which 
 has flanges aboiit 2| inches deep turned up around its circumference, and is perforated 
 with numerous J-inch or ^-inch holes. The two plates are made of J-inch boiler-iron, 
 and are tied together by four f-inch socket-rivets (see Plates XIX. and XXIX.) 
 
 Cast-iron furnace-doors have a similar box-form ; the front and sides are cast in one 
 piece about inch thick, and a wrought-iron or cast-iron screen-plate, perforated with 
 
 
326 
 
 STEAM BOILERS. 
 
 CHAP. XV. 
 
 numerous small holes, is bolted to tlie back, lugs being cast on the sides for securing it. 
 The hinges are cast on the front plate, which is also provided with a number of large 
 holes for the admission of air. The hinges of furnace-doors are made frequently of 
 composition. 
 
 The Martin or Ashcroft furnace-door (see figure 135), which has been fitted to a 
 
 number of boilers of United 
 States naval vessels, con- 
 sists of a square wrought- 
 iron plate, slightly concave 
 on the inward side, which 
 is hung from a horizontal 
 axis fixed to the upper 
 edge of the door. This 
 axis rests on brackets at- 
 tached to the cast-iron fur- 
 nace door-frame at each 
 side of the furnace-mouth, 
 and is provided with coun- 
 terbalances which are to 
 keep the door in any position in which it may be placed. After a series of competitive 
 trials with the Martin door and the ordinary furnace-door, the former was condemned 
 by a board of naval engineers as possessing none of the practical and economical advan- 
 tages claimed by the patentee, and offering, on the contrary, serious inconveniences in 
 managing the fires. (See ' Report on AsJicroft Furnace-door and Grate-bar by a Board 
 of United States Naval Engineers.'} 
 
 In order to intercept more completely the heat radiated from the bed of incan- 
 descent fuel and communicate it to the entering air, several sheets of wire gauze have 
 sometimes been placed between the perforated front and back plates of the furnace- 
 door. 
 
 "The most complete apparatus for intercepting the heat radiated to the furnace-door 
 is that of Mr. Prideaux, which consists of three gratings, each made of a series of thin 
 iron plates set edgeways, with narrow passages between them for the entering streams 
 of air. The radiant heat is completely intercepted by placing two of those sets of plates 
 with opposite obliquities, and the third parallel to the sides of the furnace mouth- 
 piece." (RanMne.} 
 
 To regulate the admission of air through the door to the furnace Prideaux made the 
 
SEC. 5. BOILER MOUNTINGS AND ATTACHMENTS. 327 
 
 gratings movable like Venetian blinds ; a self-acting mechanism opened them when 
 fresh coal was supplied, and gradually closed them as the fuel became converted into 
 coke. The openings in the ordinary furnace-door are frequently provided with a regis- 
 ter, by means of which the air-admission can be regulated by hand. 
 
 According to Rankine the total area of the perforations in the furnace-door, in recent 
 English examples, is ^ of the area of the grate when 25 pounds of bituminous coal are 
 burnt per square foot of grate per hour. In United States naval boilers, burning from 
 12 to 16 pounds of anthracite per square foot of grate per hour, the aggregate area of 
 the openings in the furnace-doors varies between square inch and 1 square inch per 
 square foot of grate. 
 
 5. Connection-doors, Ashpit-doors, and Ashpans. Cast-iron connection-doors 
 are at present superseded by wrought-iron ones, which are fitted without door-frames 
 directly to the front of the boiler. The hinges are often made of composition, and are 
 placed either at the top or at one side of the door. 
 
 To diminish the radiation of heat from the large surfaces of these doors a shield- 
 plate is secured to the door-plate by means of socket-rivets, leaving a space of two or 
 three inches between the two plates. This shield-plate is placed either on the inside or 
 the outside of the door-plate, and in some cases two shield-plates, an inner and an outer 
 one, are employed. 
 
 The connection-doors of United States naval boilers (see Plates XIX. and XXIX.) 
 are made double, of J-inch plate-iron, stayed by socket-bolts ; the outer plate is stif- 
 fened by a lip turned up around its circumference, the edge of which fits closely against 
 the front of the boiler ; the edges of the inner flanged plate form a well-fitting joint on 
 the outer plate. The space between the two plates is sometimes filled with a non-con- 
 ducting material, as plaster-of- Paris, but this makes the door heavy. On this account 
 the dead air in the space is generally relied on as a non-conductor. 
 
 Ashpit-doors are used to check the draught in order to diminish the rate of com- 
 bustion, and to prevent the inflow of cold air through furnaces which are not in use. 
 They should be made to fit close and to be easily opened and shut. 
 
 Sometimes the door consists of a simple plate placed in the mouth of the ashpit, 
 which turns on a horizontal axis passing through the middle of the plate (like a damper), 
 catches being provided to secure the plate in any desired position. 
 
 Wrought-iron or cast-iron doors, opening in halves and hinged on either side of the 
 ashpit, are frequently used. In such cases a cast-iron or angle-iron frame is secured by 
 bolts to the boiler-front around the ashpit-opening, and the doors are provided with 
 openings and a register for regulating the admission of small quantities of air to the 
 
328 STEAM BOILERS. CHAP. XV. 
 
 ashpit. Such hinged doors are generally used when air is forced into the ashpits by 
 means of a fan-blower. 
 
 In boilers of United States naval vessels the ashpit-doors are now made always of a 
 single wrought-iron plate, i inch or ^ inch thick, stiffened by a lip turned up around 
 its circumference. The edge of this lip fits directly against the front of the boiler. 
 The construction and manner of securing these doors are illustrated on Plates XIX. 
 and XXIX. When the ashpits are to be kept wide open the doors are lifted off their 
 catches and hung upon hooks permanently attached to the connection-doors. 
 
 In dry-bottom boilers cast-iron or wrought-iron asJipans are used which are remova- 
 ble and are intended to contain water. Wrought-iron ashpans are made of J-inch or 
 f-inch iron, of a single plate, with a flange turned up around the sides and back. The 
 front slopes gradually up to the height of the flange to facilitate the hauling of the 
 ashes. The bottom of the pans may be stiffened by two or three angle-irons running in 
 a longitudinal direction. 
 
 False asJipans are sometimes used to protect the iron and the stay-bolt heads of 
 water-bottoms from the corroding effect of wet ashes and from rough usage in hauling 
 ashes. They are made of wrought-iron, having a lip a couple of inches high turned up 
 on each side, and in front a lip turned down which laps over the ledge of the floor-plates. 
 
 6. Manhole and Handhole Plates. Manholes giving access to the interior of 
 the boiler are cut in the front of the boiler in the spandrels between the furnaces, be- 
 sides one or two near the top of the boiler leading into the steam-space. 
 
 Handholes or mudholes are cut in the water-legs near the bottom in the front and 
 back of the boiler, and at other convenient places, for scaling, cleaning, and washing out 
 the boiler. 
 
 Manhole and handhole plates should always be put on the inside of the boiler, so 
 that the steam-pressure tends to tighten the joint and keep the plate in position in case 
 the threads of the bolts which secure the plates should be stripped. 
 
 Manholes and handholes are generally made oval in shape, of such proportions that 
 the smallest diameter of the plate is somewhat less than the largest diameter of the 
 hole. Where practicable the largest and smallest diameters of manholes are made 
 about 15 inches and 12 inches respectively. When the space in the spandrels between 
 the furnaces does not admit of cutting oval holes of sufficiently large size the holes are 
 often triangular in shape. In cylindrical shells manholes shotild be cut in such a way 
 that their shorter axis lies in the longitudinal direction of the shell, so that the least 
 quantity of metal is removed in the line where the greatest strain obtains. 
 
 A flat, welded wrought-iron ring, about 3 inches wide, is riveted around manholes 
 
SEC. 6. BOILER MOUNTINGS AND ATTACHMENTS. 329 
 
 
 
 inside the boiler and calked tight, the rivet-heads being countersunk. The practice of 
 putting this ring outside the boiler, preTailing still to a great extent in England, is wrong ; 
 for the ring gives not only stiffness to the boiler-plate but protects it inside the boiler 
 from corrosion, which is often very active in the vicinity of manholes and mudholes. 
 Cast-iron rings are sometimes used instead of wrought-iron ones ; but especially on 
 cylindrical shells, where these rings have to restore the strength lost by cutting the 
 openings, cast-iron rings do not answer the purpose on account of the difference of elas- 
 ticity of cast-iron and wrought-iron under a tensile strain. 
 
 The following example will illustrate the manner in which the proper size of strength- 
 ening-rings of manholes in cylindrical shells may be determined : Suppose the cylin- 
 drical shell of a boiler to be inch thick, and a manhole 15 inches by 12 inches to be 
 cut in it, with the longer axis in the circumferential direction of the boiler. The weak 
 places near such a hole lie in the longitudinal axis of the boiler, and there have been 
 removed from the shell (12 X i =) 6 square inches of metal in the line of this axis. 
 To make this part of the shell as strong as the longitudinal joint of the shell the quan- 
 tity of metal added by the ring surrounding the manhole should be equal to about 65 
 per cent, of the metal cut away, and, consequently, the cross-section of the ring at 
 
 each end of the hole should be (- X ' 65 = \ 1.95 square inches. Making the ring f 
 
 inch thick, its least width should be 3f inches, if f inch is allowed for the rivet-holes. 
 
 In English and French boilers the strengthening-rings around manholes are often 
 made of angle-iron, being in such a case riveted to the outside of the boiler. This has 
 the advantage that the largest amount of metal is concentrated where the strain on the 
 plate is most severe, and rupture would commence, and the greatest stiffness is required 
 viz., at the edge of the hole. On the boiler represented on Plate XY. the rings around 
 the manholes on the ends of the boiler are made of angle-iron 3" X 3" X f. 
 
 The manhole-plates are usually made of cast-iron, the larger sizes being about li 
 inches or 1J inches thick. They have generally a dished form, the convex side being in- 
 side the boiler this form being best calculated to resist the strains on the plates with- 
 out buckling. They are secured to the boiler by one or two wrought-iron bolts passing 
 through cross-bars which straddle the hole outside the boiler. These cross-bars are now 
 also generally made of wrought-iron. The bolts are generally secured permanently to 
 the plate by a countersunk riveted head. Large plates are provided with a wrought- 
 iron handle screwed into a boss in the centre of the plate. The plate, bolts, and 
 cross-bars must be made sufficiently strong and stiff to bear with safety and without 
 springing the great strains thrown upon them in screwing up the plate. The bolts are 
 
330 
 
 STEAM BOILERS. 
 
 CHAP. XV. 
 
 made with a coarse thread, and square nuts should be used, because the corners of 
 hexagonal nuts are liable to become rounded when the wrench does not fit well. 
 
 Instead of fitting the flange of a manhole-plate directly to the cylindrical shell of a 
 boiler, as in figure 5, Plate XXXI. (which requires very careful work), the stiffening- 
 ring inside the boiler sometimes forms a plane seat for the flange of the plate ; or, when 
 the radius of the curved surface is small, a casting is riveted to the outside of the boiler 
 around the manhole, having a flange which forms a plane seat for the plate (see 
 figure 136). 
 
 In order to reduce the weight of large manhole-plates they are sometimes made of 
 wrought-iron. Figure 137 represents the wrought-iron manhole-cover of a boiler de- 
 Fig. 137. 
 
 signed for a working pressure of 70 Ibs., built by Maudslay Sons & Field (England) in 
 1873. The dimensions of the manhole are 15 inches X 10 inches, and its shape is rect- 
 angular with rounded corners. Two plates f inch thick are riveted together with 
 countersunk rivets. The outer plate is of the size of the hole, while the inner plate 
 is large enough to form the flange 1 inches wide. The bolts are screwed into the 
 plate, the ends being riveted over. The ring around the hole is 1 inch thick and 2| 
 inches wide, secured by countersunk rivets to the shell, which is i inch thick. 
 
 Wrought-iron plates made of a dished form by pressing them with a die of suitable 
 shape into a mould would be much stiffer than flat plates, and could be made propor- 
 tionately thinner. 
 
 Figures 1 and 5, Plate XXXI., represent the manhole-covers made recently for 
 United States naval boilers. They are cast of old composition metal consisting of 88 
 
SEC. 7. BOILER MOUNTINGS AND ATTACHMENTS. 331 
 
 parts of copper, 10 parts of tin, and 2 parts of zinc. The wrought-iron bolts are secured 
 in the plates by riveting over the ends, which pass through accurately-drilled holes. A 
 handle is cast on the plate. 
 
 7. Steam Stop- valves, Dry-pipes, and Steam-pipes. The stop-valves of 
 boilers must be arranged in such a manner that they are easily accessible and can be 
 opened and closed quickly. Each boiler must have a stop- valve, bolted directly to the 
 shell, to shut off all communication between the boiler and the steam-pipe connected 
 with the engines or the other boilers. In case there are separate steam-drams or super- 
 heating-chambers the stop-valves and connecting steam-pipes must be arranged in such 
 a way that any one of the boilers or steam-chambers may be shut off without the neces- 
 sity of putting any of the others out of use. 
 
 It is a safe rule to make the area of stop-valves and steam-pipes sufficiently large 
 that the velocity of the steam passing through them does not exceed 100 feet per 
 second when the speed of the engines is a maximum. 
 
 Figures 1 and 2, Plate XXXII., represent the stop-valves of the boilers of the U. S. S. 
 Nipsic, and may serve to illustrate the usual construction of these valves for United 
 States naval boilers. When the valves are large the chamber is made of cast-iron, and a 
 valve-seat, made of composition metal, is fitted in it and secured by riveting over the 
 lower end. For smaller valves the whole chamber is made of composition. The valve- 
 disc is made of composition and of a dished form to increase its stiffness, and it has a 
 conical seat. It is guided by a central spindle below, working easily in a sleeve con- 
 nected by ribs to the valve-seat. The wrought-iron stem has a square screw-thread, 
 which works in a corresponding thread cut in a cross-bar supported by wrought-iron 
 studs on the cover of the valve-chamber. The stem must not be rigidly attached to the 
 valve-disc, so that the latter does not turn with the stem and seats itself always ex- 
 actly. In the valves represented on Plate XXXII. the stem passes through the valve- 
 disc and its lower end forms the guide-spindle. In other cases the guide-spindle is cast 
 on the valve-disc, and the stem has a collar at its lower end which fits in a recess 
 formed in a projection on the top of the valve-disc ; it is held in place by an annular 
 nut screwed to this projection on the valve-disc and secured by a pin. 
 
 The stop-valve and steam-pipe must take the steam from the highest part of the 
 boiler, where it is in the driest state. When the boiler has no vertical steam-drum the 
 stop- valve is generally connected with a dry-pipe, which draws the steam evenly from a 
 large area within the boiler, and separates to some extent the water which is carried 
 along with the steam from the latter. The dry -pipe extends through the length of the 
 boiler close to the top, and in large rectangular boilers has several lateral branches. It 
 
332 STEAM BOILERS. 
 
 CHAP. XV. 
 
 is connected at one end with the stop- valve chamber by a tight joint, while its other end 
 is closed. On the top it is perforated by numerous evenly-spaced holes of about f inch 
 diameter, or has narrow, transverse slits cut into it by means of a saw. The aggregate 
 area of these openings should be at least double the area of the cross-section of the 
 pipe. 
 
 The dry-pipe is often made of cast or wrought iron, but sheet-brass is a preferable 
 material, since the pipe is much exposed to corrosion. 
 
 Dry-pipes are frequently omitted because they make the interior of the boiler less 
 accessible ; in such cases the opening in the shell is often protected by deflecting-plates 
 or by a box perforated with numerous holes, in order to throw off any water carried up 
 by foaming. When a boiler foams because the area of the stop-valve is too small, and 
 it is not convenient to fit dry-pipes within the boiler, it is better to place an additional 
 stop-valve on the boiler at some distance from the original one than to enlarge the ex- 
 isting stop-valve. 
 
 The steam-pipe should have as direct a course and as few bends as possible. Ex- 
 pansion-joints must be provided between rigid attachments of the pipe, unless there are 
 bends which will allow the pipe to spring as it expands or contracts in the direction of 
 its length. 
 
 Copper pipes, tinned inside and outside and fitted with composition flanges, are 
 generally used for the steam-pipes of United States naval boilers. Cast-iron pipes are 
 far cheaper, but are heavier, and, from the unyielding nature of the material, liable to 
 break when the ship works much or in case the boilers should move in their seats. 
 Wrought-iron pipes, either lap- welded or riveted, are used, but have the disadvantage 
 of being speedily attacked by corrosion. 
 
 Drain-pipes must be fitted to all valve-chambers and to all parts of the steam-pipes 
 where water is liable to accumulate. 
 
 The arrangement of the steam-pipes and stop-valves of the boilers of the U. S. S. 
 Nipsic is shown on Plate XXX. 
 
 A stop-valve is bolted to the outboard end of each boiler, connected with a dry -pipe 
 extending through the length of the boiler. These stop-valves (see figure 1, Plate 
 XXXII.) are operated by means of a hand-wheel from below. The stem is continued 
 through the top of the valve-chamber, and over each valve a small hole is cut in the 
 deck, which is ordinarily kept closed by a composition cap. This arrangement makes it 
 possible to operate the valves from the main deck by means of a socket-wrench in case 
 the passage at the back of the boilers should be inaccessible. 
 
 A copper pipe, 6 inches in diameter, No. 13 B. W. Gr. thick, tinned inside and out- 
 
SEC. 8. BOILER MOUNTINGS AND ATTACHMENTS. 333 
 
 side, is bolted by means of composition flanges ff inch thick to a nozzle on the stop- 
 valve, and connects each wing boiler with the nearest horizontal steam-drum above it, 
 the middle boiler being similarly connected to either drum. A composition casting, 
 having suitable nozzles for connecting with the steam-pipes, is bolted to the bottom of 
 each steam-drum, and has at its lowest point a nozzle, 3 inches in diameter, to which a 
 drain-pipe is attached which leads to a water-trap. 
 
 The inboard end of each drum is connected by means of a short wrought-iron nozzle 
 with the superheating steam-pipe passing through the length of the uptake at either 
 side of the vessel. These superheating-pipes are of wrought-iron, lap-welded. They 
 are 9 inches in diameter, 0.344 inch thick, and have wrought-iron flanges 1 inch thick 
 riveted to them. The nozzles connecting the drums with the superheating-pipes are 
 riveted to the latter. Each superheating-pipe is made in two lengths ; the flanges con- 
 necting them are surrounded by an iron casing which protects the joint from the heat of 
 the uptake. A safety-valve is attached to the forward end, and a stop-valve 9f inches 
 in diameter (see figure 2, Plate XXXII.) is attached to the after end of each superheat- 
 ing-pipe. A copper pipe, 8 inches in diameter, No. 12 B. W. G. thick, bolted by 
 means of a composition flange inch thick to a nozzle of the valve-chamber, conveys the 
 steam from the boilers to the engines. 
 
 Each forward and after wing boiler at either side of the vessel is connected by means 
 of a stop-valve, 4f inches in diameter and bolted to the cylindrical shell of the boiler, 
 with a copper pipe leading over the top of the boilers to the auxiliary pumps and to the 
 distiller. 
 
 The steam-pipes and stop-valves for the United States ironclad Miantonomoh and 
 class are described in the specifications of the boilers of these vessels in section 10, 
 chapter vii. 
 
 8. Check-valves and Feed-pipes A check-valve, consisting of a disc-valve with 
 a conical seat, is placed between the feed-pipe and the boiler. The valve is kept closed 
 by the pressure within the boiler acting on its upper surface, and rises with each stroke 
 of the pump as the water-pressure within the feed-pipe, acting on the lower surface of the 
 valve, exceeds the boiler-pressure. A detached stem with a square thread bears, when 
 it is screwed down, on the upper surface of the valve and keeps it closed, and when 
 raised regulates the lift of the valve, and consequently the supply of feed- water to the 
 boiler. For guiding the valve it is provided below its seat with three or four wings, 
 or with a central spindle working in a sleeve, and above by a spindle working in a 
 socket in the enlarged end of the detached stem. 
 
 When check-valves have much lift the hammering action of the valve causes 
 
334 STEAM BOILERS. CHAP. XV. 
 
 the rapid destruction of the valve and seat, so far as tightness is concerned, especially 
 with a quick-acting pump. The lift of a check-valve should not exceed ordinarily 
 inch ; and the area of the valve should be such that with this lift the rate of flow 
 of the feed-water through the valve-opening does not exceed 600 feet per minute. 
 
 Since much trouble is caused by check- valves leaking or not closing properly when 
 foreign matter lodges between the valve and its seat, a stop- valve is frequently placed 
 between the check- valve and the boiler, which permits the communication between the 
 boiler and check- valve to be cut off for the purpose of examining and cleaning the 
 latter. This stop-valve is also used for regulating the supply of feed-water to the 
 boiler. 
 
 In such a case the upper spindle of the check- valve is sometimes continued through 
 the chest-cover, and carries a weight on its top which is sufficiently heavy to overcome 
 the friction of the stem in the stuffing-box and ensures the prompt seating of the valve 
 after each stroke of the pump. 
 
 Plug-cocks are preferred by many engineers to screw stop-valves on feed-pipes, be- 
 cause the latter may be prevented from shutting by some solid matter getting under the 
 valve-disc ; but cocks of large dimensions are often very difficult to turn, and a feed- 
 cock which cannot be opened causes much greater inconvenience than a feed-valve 
 which cannot be shut tight. 
 
 Figure 3, Plate XXXII., represents the feed and check valves of the boilers of the 
 U. S. S. Nipsic. A stop- valve is placed between the boiler and the check-valve, and a 
 like stop-valve is placed between the check- valve and the feed-pipe. These three valves 
 have all an opening 2| inches in diameter, and are contained in a single casting made 
 of composition metal. The stem of the stop-valves is made of steel. The upper guide- 
 sleeve of the check-valve is cast on the cover of the chamber, and the cover is held 
 down by a single bolt passing through a wrought-iron bail. 
 
 The check-valve chamber is bolted directly to the shell of the boiler at such a place 
 where it is most conveniently situated for controlling the feed-supply. It is generally 
 placed on or near the front of the boiler at about the height of the furnace-crown, and 
 discharges the water directly into the boiler. It is objected to this arrangement that the 
 comparatively cold feed-water causes injury by impinging directly against the most 
 highly heated part of the boiler. On this account an internal pipe is sometimes pro- 
 vided which leads downward, discharging the feed-water near the bottom of the boiler. 
 In other instances this internal pipe leads upward, discharging the water near the 
 smoke-connection, where the temperature of the gases is least ; with the latter arrange- 
 ment the cool feed-water, sinking by gravity, promotes also the circulation of the water 
 
SEC. 9. BOILER MOUNTINGS AND ATTACHMENTS. 335 
 
 in the boiler. In some cases the internal pipe, continued horizontally across the smoke- 
 box end of the tubes, has been provided with numerous small openings throughout its 
 length, through which the feed- water is distributed over a wide space instead of being 
 discharged in a mass at one point. 
 
 The feed-pipes for United States naval boilers are generally either cast of composi- 
 tion metal or are made of drawn-brass tubes connected by composition flanges and 
 heavily tinned on the inside. Copper pipes have been abandoned on account of the 
 galvanic action produced in the boiler by the small particles of copper abraded and 
 carried along by the feed-water. Cast-iron pipes are heavy and become soon perfo- 
 rated with small holes. 
 
 The feed-pipes must be made with as few bends as possible ; when they are long 
 they must be provided with slip-joints. The feed-pipes must be placed where they 
 are easily examined and repaired. 
 
 The usual arrangement of the feed-pipes in United States naval vessels may be 
 seen on Plate XXX., illustrating the boilers of the U. S. S. Nipsic. A feed-pipe, made 
 of composition and having 3 inches internal diameter, runs along the front of the boilers 
 at either side of the vessel below the fire-room floor ; and it is connected by vertical 
 branch-pipes, having an internal diameter of 2 inches, with the check-valve chamber of 
 each boiler. 
 
 See section 10, chapter vii., for the specifications of the feed-pipes and check- valves 
 of the boilers of the United States iron-clad Miantonomoh and class. 
 
 9. Blow-valves and Pipes. While marine boilers were worked with very low 
 steam-pressure, pumps were used to withdraw continuously a certain quantity of the 
 concentrated water from the boiler, so as to maintain the saturation within the boiler 
 at a given point. But these brine-pumps have gone out of use, and the water is 
 blown from the boiler overboard directly by the steam-pressure, the quantity thus 
 blown out being regulated by the opening of the blow-valves provided for the purpose. 
 
 The blow-valves of United States naval boilers are generally disc-valves with a coni- 
 cal seat, operated by means of a screw-thread cut on the stem of the valve ; the 
 valves and valve-chambers are made of composition metal, and are similar in construc- 
 tion to the stop and feed valves represented on Plate XXXII. It is urged against the 
 iise of disc-valves that small chips, pieces of incrustation, or other solid matter are 
 liable to lodge on the seat of such valves and prevent their closing tight when to all 
 outward appearance they may seem quite shut. On this account cocks are preferred 
 by many engineers for blow-valves. 
 
 The principal drawback to the use of large cocks is their liability to stick fast in 
 
336 STEAM BOILERS. CHAP. XV. 
 
 consequence of corrosion or incrustation, of unequal expansion of the plug and shell, 
 or of other causes producing excessive friction. The tendency to stick fast is greatly 
 aggravated when the shell is made of cast-iron and the plug of brass ; both should 
 be made of composition metal, not too soft. "For pressures of 20 or 30 Ibs. a taper of 
 one in four is found to work well, but for pressures of 90 or 100 Ibs. a taper of one in 
 six is necessary to ensure tightness." ( Wilson.} 
 
 The regulations of the Board of Trade (English) prescribe that- 
 'll! blow-off cocks and sea-connections are to be fitted with a guard over the 
 plug, with a feather- way in the same, and a key on the spanner, so that the spanner 
 cannot be taken out unless the plug or cock is closed. One cock is to be fitted to the 
 boiler, and another cock on the skin of the ship or on the side of the Kingston 
 valve." 
 
 The chamber of the blow-valve is bolted directly to the shell of the boiler in a 
 convenient position on or near the front of the boiler. 
 
 The bottom blow-valve is used, while the boiler is in operation, to remove the dirt and 
 sediment which collects in the bottom of the boiler by discharging a limited quantity 
 of water at intervals, and to fill the boiler with water from the sea before starting 
 the fires, and to empty the boiler after hauling the fires. The blow-valve takes the 
 water from the bottom of the boiler through an internal pipe secured with a tight 
 joint to the shell of the boiler. The bottom blow-valve is generally made of the same 
 size as the feed-valve, so that the boiler may be filled and emptied quickly. 
 
 The surface blow-valve is generally made about one-half as large as the feed- valve. 
 It is used to remove the scum and other impurities floating near the surface of the 
 water. The valve is connected with a system of perforated pipes extending through 
 the boiler a short distance below the water-line, or with one or several perforated 
 boxes or strainers in which the water, being undisturbed by ebullition, deposits the 
 solid particles held in suspension. 
 
 The arrangement of the blow- valves and pipes of the boilers of the TJ. S. S. Nipsic 
 is shown on Plate XXX. The bottom blow- valves of the several boilers on each side 
 of the vessel are connected by means of composition branch-pipes, 2 inches in internal 
 diameter, with the main blow-pipe leading along the front of the boilers to the Kingston 
 valve in the bottom of the vessel. The main blow-pipes are made of seamless drawn 
 brass tubes, connected by flanges, and are provided with a slip-joint. The pipes of the 
 surface blow-valves are connected with a stop-valve on the side of the vessel a short 
 distance below the water-line ; and a branch-pipe connects the surface blow-pipe also 
 with the bottom blow-pipe. 
 
SEC. 10. BOILER MOUNTINGS AND ATTACHMENTS. 337 
 
 The blow-pipes are frequently exposed to violent shocks and jars when they are dis- 
 charging the hot water into the sea. On this account bends must be avoided as much 
 as possible and must be made with easy curves, and the pipes must be made of a tough 
 material. The use of cast-iron is to be condemned. Brazed copper pipes are liable to 
 split. Cast composition or seamless drawn brass pipes are nowadays generally used for 
 the blow-pipes of United States naval boilers. (See " Specifications for Boilers of 
 United States Ironclad MiantonomoJi and Class,''' section 10, chapter vii.) 
 
 1O. Instruments and Attachments for Measuring and Indicating the 
 Height and the Density of the Water, and the Pressure and Temperature 
 of the Steam. 
 
 Water-gauges. The rules and regulations of the Supervising Inspectors of 
 Steam-vessels provide that "all steamers having one or more boilers shall have three 
 suitable gauge-cocks in each boiler ; those having three or more boilers in battery shall 
 have three in each outside boiler and two in each remaining boiler in the battery ; and 
 the middle gauge-cocks in all boilers shall not be less than 4 inches above the top of the 
 flues, tubes, or crown of the fire-box." 
 
 United States naval boilers are fitted with three or four water-gauge cocks, placed 
 from 4 to 6 inches apart, the lowest gauge-cock being placed on a line with the top of 
 the back-connections. Either screw-valves with conical seats or plug-cocks are used for 
 water-gauges. To keep their opening clear of any solid matter the valves are provided 
 with feathers projecting beyond the opening of the valve-cham- Fig. 138. 
 
 ber (see Plate XXXIII.) Provision is made to clear plug-cocks 
 by means of a wire, by forming a straight passage through them, 
 which is ordinarily kept closed by a screw-plug at the front end 
 (see figure 138). The gauge-cocks discharge the steam and water 
 into a copper drip-pan provided with a drain-pipe which leads 
 down into the bilge of the vessel or to a water-trap. 
 
 In addition to the gauge-cocks boilers are generally provided with water-gauge 
 glasses, consisting of a glass tube from 12 to 18 inches long, the top and bottom of 
 which communicate by means of suitable fittings with the steam and water spaces re- 
 spectively, so that the water within the glass stands at the same level as the water 
 within the boiler. Pipes lead from the top and bottom of the gauge well up into the 
 steam-space and down into the water-space respectively, so that the indications of the 
 water-level in the glass are not affected by violent ebullitions and foaming. These 
 pipes should not be less than one inch in diameter, so as not to be clogged easily by 
 pieces of loose scale or impurities in the water. 
 
338 STEAM BOILERS. CHAP. XV. 
 
 Gauge-glasses must be made of a white, transparent glass without bubbles or other 
 defects which would impair their strength, and must be carefully annealed. Gauge- 
 glasses have been introduced of late in which the side of the glass turned toward the 
 boiler is covered with a white enamel, while the other half is left transparent ; on the 
 white background the line of the water-level is more plainly seen. 
 
 At the top and bottom the glass fits in a stuffing-box with a screw-gland, and is 
 packed with soft rubber or cotton-wick. Cocks or screw-valves are provided for shut- 
 ting off the communication between the gauge and the boiler at the top and bottom, and 
 for opening a passage between the gauge-glass and the drain-pipe. All the passages of 
 the gauge must be of such size and form that they are not clogged easily by dirt, and 
 so arranged that they can be cleared while the boiler is in operation. 
 
 The water-gauge cocks and glasses are either placed directly on the front of the 
 boiler or they are attached to a tube made of composition or cast-iron and having an 
 internal diameter of 2J or 3 inches. This tube is placed close to the boiler, with its 
 upper and lower ends communicating with the steam and water spaces as described 
 above. 
 
 Plate XXXIII. illustrates the type of water-gauge recently constructed for the boil- 
 ers of the U. S. S. Nipsic and other United States naval vessels. A glass gauge and 
 four gauge-cocks are attached to a composition casting, the general outline of which is 
 cylindrical, so that its whole exterior can be finished by turning. This casting contains 
 several small compartments and passages and one large chamber extending through 
 the length of the casting and connected at the top and bottom with the boiler by means 
 of composition pipes 1J inches in internal diameter. The communication between these 
 pipes and the boiler may be shut off by means of plug-cocks placed on the shell of the 
 boiler (see Plate XXX.) Four gauge-cocks, consisting of screw-valves with conical 
 seats, and placed 6 inches apart, are screwed into a recessed plane face of the casting, 
 and communicate directly with the large chamber. The cylindrical shell of the casting 
 forms a shield enclosing the discharge-nozzles of the gauge-cocks and leading the 
 steam and water blown out to a waste-passage at the bottom of the casting. A gauge- 
 glass, having an external diameter of f inch and an exposed length of 19 inches, enters 
 at the top and bottom into small separate compartments, which communicate by means 
 of screw- valves with the large chamber and with a waste-passage, to the lower end of 
 which a copper drain-pipe f inch in diameter is attached. Behind the gauge-glass a 
 lamp is placed from the side, which can be moved up or down and clamped in any 
 position. 
 
 These gauges are made right and left. All the fittings are of composition. The 
 
SEC. 10. BOILER MOUNTINGS AND ATTACHMENTS. 339 
 
 handles of the gauge-cocks are all placed at the same angle when the cocks are 
 shut. 
 
 Some years ago percussion water-gauges were frequently used on United States 
 naval boilers in addition to the gauge cocks and glasses. They are especially intended 
 to indicate the height of solid water when the boilers foam badly. They consist of a 
 cylinder, made of composition metal, about 4 inches in diameter and about 20 inches 
 long, connected at the top and bottom by means of pipes with the steam and water 
 spaces of the boiler, and placed at about the same height on the front of the boiler as 
 the water-gauge cocks and glass. This cylinder contains an easily-fitting piston, with 
 a rod passing through a stuffing-box on the top, to which a handle is attached which 
 leads downwards and carries a pointer at the same level as the piston. The latter 
 having been raised clear of the water, it is easy to feel when it strikes the water on 
 being pulled down suddenly, and the position of the pointer relatively to rings formed 
 on the outside of the cylinder shows the height of the water in the boiler. 
 
 In another class of devices a float indicates the height of the water in the boiler. 
 This float is formed frequently by a large hollow sphere which floats in an upright 
 cylindrical vessel in which the water stands at the same level as in the boiler. The 
 position of the water-level is indicated by a pointer, which is moved by a rod attached 
 to the float. 
 
 In the Belleville boiler such a float is used to regulate the quantity of feed-water 
 admitted by adjusting the opening of the feed-valve. In other cases the float admits 
 steam to an alarm-whistle when the water falls below a certain height. 
 
 Floats are, however, seldom used on marine boilers ; they are more applicable to 
 stationary and steamboat boilers which are always fed with fresh water, and where 
 their indications are not affected by violent motions of the vessel. 
 
 Fusible Plugs. Another safeguard against the dangers arising from low water in 
 the boiler is the fusible plug which closes a small hole in the water-heating surface 
 of the boiler at a height below which the water cannot be allowed to fall without immi- 
 nent danger. The plug is made of some metal or alloy which will melt before the 
 iron is overheated to a dangerous degree. The discharge of steam through the hole 
 thus formed gives warning of the danger and at the same time relieves the pressure 
 within the boiler and retards the combustion. 
 
 The rules and regulations of the Supervising Inspectors of Steam-vessels re- 
 quire that all fire-box boilers shall have one plug of Banca tin 1 inch in diameter 
 inserted in the crown of the back-connection. These fusible plugs are never used in 
 United States naval boilers. 
 
340 STEAM BOILERS. CHAP. XV. 
 
 Salinometer-pots. All boilers of United States naval vessels are provided with per- 
 manently-attached salinometer-pots for testing the density of the water in the boiler. 
 The water-pipes through which the pots communicate with the boilers, as well as the 
 drain-pipes, are made of copper or brass about inch in diameter, and are connected 
 by means of screw-couplings, so as to be easily detached for cleaning. Stop-cocks are 
 provided to open and shut off communication between the pipe and the pots and boil- 
 ers. The water-pipe must be connected to the boiler at some place above the furnace- 
 crowns where the temperature of the water is equal to that of the steam, and not in the 
 vicinity of the feed- valve. 
 
 The object sought to be obtained in the construction of salinometer-pots is to main- 
 tain in an open vessel a constant flow of the water drawn from the boiler while testing 
 its density, and to reduce this water to a fixed temperature below the boiling-point 
 under atmospheric pressure, so as to avoid ebullition and the formation of clouds of 
 vapor. 
 
 Long's salinometer-pot, which has been in use on United States naval boilers for 
 many years, consists of two brass cylindrical vessels placed side by side and communi- 
 cating at the bottom. The water enters one of these vessels, which is kept closed by a 
 cover perforated with a few small holes for the escape of steam, through a central pipe 
 closed at the top and perforated near its upper end. The rate of flow of the water from 
 the boiler is regulated by a stop-cock. The water rises simultaneously in the two ves- 
 sels. A central tube reaching nearly to the top of the second vessel, which is kept 
 open, serves as an overflow. This overflow-pipe passes through the bottom of the pot 
 and is coupled to a drain-pipe ; another drain-pipe communicates with the bottom of 
 the vessels to draw off the water from them. The hydrometer is placed in the open 
 vessel, which is provided at the side with clamps for holding a thermometer which indi- 
 cates the temperature of the water during the test. 
 
 fitTiian's salinometer-pots have been used on several United States naval boilers 
 since the introduction of higher steam-pressures. The hot water drawn from the boiler 
 passes, before it enters the open vessel, through a coiled pipe immersed in a stream of 
 cold water. By regulating the flow of the cooling water the temperature of the water 
 which is to be tested can be regulated quickly and exactly. 
 
 Steam-gauges. Each boiler must be connected with a separate steam-gauge, which 
 must communicate directly with the boiler and not with the steam-pipe, so that the clos- 
 ing of a stop-valve does not put the gauge out of use. The gauge is located at a conve- 
 nient place in the fire-room, and is connected with the top of the boiler by a copper or 
 brass pipe about inch in diameter, to which a downward ciirve is given close to the 
 
SBC. 11. BOILER MOUNTINGS AND ATTACHMENTS. 341 
 
 gauge, so that the water accumulating at this point prevents the hot steam from coming 
 in contact with the spring of the gauge. Plug-cocks are placed on the boiler and on the 
 gauge, and the pipe is connected to them by screw-couplings. It is advisable to use a 
 soft lead washer as a packing in the coupling, as rubber is apt to swell and be squeezed 
 out till it closes the opening of the small pipe. 
 
 United States naval vessels have generally, in addition to the spring-gauges attached 
 to each boiler, one standard mercurial gauge, which is connected with the main steam- 
 pipe of the boilers. 
 
 Thermometers. All separate superheating-chambers should be fitted with thermo- 
 meters, especially when the steam is superheated to a high degree. The thermometer 
 must be immersed in the steam as far as possible, leaving only such a length of the stem 
 exposed as is necessary to read the instrument. The part of the instrument which is 
 immersed in the steam is surrounded by a perforated pipe ; the projecting stem is pro- 
 tected by a shield-plate, or by a brass case fitted with a sliding-plate. 
 
 11. Safety-valves. Each boiler must be provided with a safety-valve, arranged in 
 such a manner that the communication between the valve and the boiler cannot be shut 
 off. A safety-valve must also be provided for each separate superheating-chamber and 
 feed- water heater. The safety-valve should be placed on the top of the boiler or be con- 
 nected by an internal pipe with the highest part of the steam-space. 
 
 Safety-valves must be so arranged that they may be opened by hand in order to 
 relieve the boiler of steam-pressure at any time and to try whether the valve moves 
 freely in its seat. The rules of Government inspectors require that, in addition, each 
 boiler shall carry a lock-up safety-valve of sufficient size, which, being set to blow off at 
 the pressure allowed, is entirely beyond the control of the persons manipulating the 
 machinery. 
 
 Safety-valves are weighted by applying the load to them either directly or by means 
 of a lever. Springs are used almost universally instead of weights for directly-loaded 
 valves on marine and locomotive boilers ; they are also frequently used for lever safety- 
 valves. Spring-loaded valves come more and more into use in sea-going steamers 
 where steam of high pressure is used, on account of the difficulties incident to the use of 
 heavy weights in consequence of the violent motions of vessels in rough weather. When 
 the lever of safety-valves is loaded by dead weights it is placed in the fore-and-aft 
 direction of the vessel. 
 
 In a spring-loaded valve the tension or compression of the spring increases with the 
 lift of the valve, while the weight of a dead load is the same for every lift. But with a 
 properly-proportioned, directly-loaded valve the increase of resistance of the compressed 
 
342 STEAM BOILERS. CHAP. XV. 
 
 or extended spring is trifling. When the spring acts on a lever some compensating 
 arrangement should be adopted to counteract the effect of the increased resistance. 
 
 A safety-valve acting automatically must fulfil the following essential conditions 
 viz. : 
 
 It must be capable of discharging at a given pressure the greatest weight of steam 
 which the boiler is capable of generating in a unit of time. 
 
 It must not allow the pressure within the boiler to rise above a fixed limit, and it 
 must close quickly when the pressure falls below that at which the valve is set to open. 
 
 It must be reliable in its action under continued use ; it must be simple in its con- 
 struction and easily adjusted and managed. 
 
 The size of the valve must be proportioned to the greatest weight of steam which may 
 be generated in a unit of time. Rules which determine the size of safety-valves by the 
 dimensions of the grate or heating surface, or by the weight of coal consumed in a unit 
 of time, are based on the supposition that under the given conditions a fixed rate of 
 evaporation obtains, and apply consequently only to special classes of boilers. A gene- 
 ral rule must determine the area of the valve by the weight of steam to be discharged 
 in a unit of time and by the pressure at which it is to be discharged. 
 
 The weight of steam, in pounds, discharged into the atmosphere per second through 
 an orifice having an area of 1 square inch, is approximately equal to the absolute pres- 
 sure of steam in pounds per square inch divided by 70 when the steam -pressure is equal 
 to or greater than 25 pounds per square inch above zero. (Rarikine, ' Manual of the 
 Steam-engine. ' ) 
 
 The effective opening of a safety-valve lifting automatically is very small relatively 
 to the area of its disc, because the lift of the valve is always small. With the ordinary 
 disc-valve a greater lift than -^ inch should not be counted upon. This small lift is due 
 to the rapid diminution of the force exerted by the steam-pressure on the valve as it 
 rises from its seat. Various methods have been tried of increasing the lift of the valve 
 by making the escaping steam impinge on a lip turned down around the rim of the 
 valve, or by otherwise obstructing the passage of the escaping steam (see figure 3, Plate 
 XXXIV., representing Ashcroft's safety-valve). But, in general, such arrangements 
 tend also to produce an excess of pressure within the boiler over the pressure at 
 which the valve begins to lift, or to make the action of the valve irregular or inter- 
 mittent. 
 
 A further diminution of the effective opening of valves is due to the conical form 
 ordinarily given to the seat. 
 
 The safety-valve must discharge the steam, when the evaporation is a maximum, so 
 
SEC. 11. BOILER MOUNTINGS AND ATTACHMENTS. 343 
 
 rapidly tha,t the greatest increase of pressure within the boiler does not exceed 10 or 12 
 per cent, of the pressure at which the valve begins to lift. 
 
 When the diameter has to exceed 5 inches in order to get sufficient area, it is better 
 to increase the number than the size of the valves. 
 
 Thurston proposes the following formula for determining the area of safety- 
 valves : 
 
 0.5 w 
 
 A = 
 
 IO 
 
 when A = area of safety-valve in square inches ; 
 
 p pressure of steam in pounds per square inch above the atmosphere ; 
 w = weight of water, in pounds, evaporated per hour as a maximum. 
 
 Rankine proposed the following rule for determining the area of safety-valves : 
 Multiply the number of pounds of water evaporated per hour by 0.006 ; tJie product 
 will be the area of the valve in inches. 
 
 The rules of the United States Supervising Inspectors of Steam-vessels, of the Board 
 of Trade (English), and of Lloyd's Register require that the safety-valves of marine 
 boilers shall have an area of not less than half a square inch to each square foot of 
 grate-surface when the ordinary safety-valve is employed. But when a safety-valve of 
 an approved pattern is used which gives a greater lift than the common safety-valve 
 the size of the valve may be diminished. 
 
 In the report on safety-valve tests, made in 1875 at the United States Navy- 
 Yard, Washington, D. C., by a special committee of the Board of Supervising Inspec- 
 tors of Steam-vessels, it is stated that an ordinary disc-valve with a bevelled seat, 
 having an area of ten square inches, will discharge two thousand pounds of steam in an 
 hour at all pressures from 20 to 100 Ibs. per square inch. The following rule for deter- 
 mining the size of safety-valves is deduced from these experiments : 
 
 Multiply the weight of water in pounds evaporated in one hour by 0.005 ; the re- 
 sult is the area of the valve-disc in square inches. 
 
 It is likewise recommended that the area of safety-valves should not exceed ten 
 square inches, and that several valves be employed when a larger area is required for a 
 boiler. 
 
 Numerous forms of safety-valves and arrangements for loading them have been de- 
 vised. Annular valves and double poppet-valves have been used for the purpose of 
 obtaining a large area of opening with a given lift. In other cases an auxiliary valve 
 or piston has been used in combination with the safety-valve, with a view to increasing 
 
 r THE 
 UNIVERSITY 
 
 OF 
 
34.4 STEAM BOILERS. CHAP. XV. 
 
 the lift of the valve and ensuring its prompt seating when the pressure falls below the 
 point for which the valve is set. But, generally speaking, the ordinary disc- valve is 
 not only the simplest in construction, but most reliable in its action and least liable to 
 derangement. 
 
 Disc- valves are made either with a conical or with a flat-faced seat. It is claimed for 
 the latter that they are less liable to stick and that they present a larger opening with the 
 same lift than the former. On the other hand, it is objected that it is more difficult to 
 keep them tight, and that the steam escapes with greater difficulty through their open- 
 ing, since it has to make two abrupt changes in direction. When a wide bearing-surface 
 is given to flat-faced valves they are apt to have a trembling, vibratory motion when they 
 are discharging steam. 
 
 Conical valves are most usually employed, the bevel of their seat forming an angle 
 of 45. With a narrow face the valve is more easily kept tight and the steam escapes 
 with greater ease than with a wide face ; some authorities claim that a width of ^ i ncn 
 is sufficient for the seat of a conical valve 4 inches in diameter. The valve and seat are 
 generally made of composition metal for marine boilers, but in Ashcroft's valve (see 
 figure 3, Plate XXXIV.) the bearing-surfaces are formed by nickel rings let into the 
 valve and seat. 
 
 The valve is guided either by wings attached to the disc below its seat or by a cen- 
 tral spindle Working in a sleeve. The opinions of competent engineers differ as to which 
 of these two devices for guiding the valve is preferable ; both are liable to cause the 
 valve to stick when they are fitted too close, in consequence of the lodgment of dirt or 
 scale, or of unequal expansion when steam is raised quickly. 
 
 To prevent the canting of the valve in consequence of the oblique thrust thrown 
 upon it by the lever as it lifts, the central spindle upon which the lever rests is often 
 detached from the valve-disc ; the latter is hollowed out on the top so that the point 
 where the spindle rests upon the valve lies below the valve-seat. When this spindle is 
 rigidly attached to the valve-disc it is connected with the lever by means of a short 
 link. The pins or bolts of the articulated joints of the lever and link are often made of 
 composition to lessen the liability of their becoming fast by rusting or by getting 
 clogged with grease and dirt. To reduce the friction as much as possible it is best to 
 make the lever turn on knife-edges, case-hardened. 
 
 The lever safety-valve recommended by the Board of Supervising Inspectors of 
 Steam-vessels is represented in figure 2, Plate XXXIV. The following directions are 
 given regarding its construction : 
 
 "All the points of bearing on lever must be in the same plane. 
 
SEC. 11. BOILER MOUNTINGS AND ATTACHMENTS. 345 
 
 " The distance of the fulcrum must in no case be less than the diameter of the valve- 
 opening. 
 
 "The length of the lever should not exceed the distance of the fulcrum multiplied 
 by ten. 
 
 " The width of the bearings of the fulcrum must not be less than three-fourths () 
 of one inch. 
 
 " The length of the fulcrum-link should not be less than four (4) inches. 
 
 "The lever and fulcrum-link must be made of wrought-iron or steel, and the knife- 
 edged fulcrum-points and bearings for the points must be made of steel and hardened. 
 
 " The valve, valve-seat, and bushings for the stem or spindle must be made of com- 
 position (gun-metal) when the valve is intended to be attached to a boiler using salt 
 water ; but when the valve is to be attached to a boiler using fresh water and generating 
 steam of a high pressure, the parts named, with the exception of the bushings for the 
 spindle, may be made of cast-iron. 
 
 "The valve must be guided by its spindle, both above and below the ground seat 
 and above the lever, through supports either made of composition (gun-metal) or bushed 
 with it. 
 
 " The spindle should fit loosely in the bearings or supports. 
 
 " When the valve is intended to be* applied to the boilers of steamers navigating 
 rough waters the fulcrum-link may be connected directly with the spindle of the valve, 
 providing always that the knife-edged fulcrum-points are made of steel and hardened, 
 and that the object sought by the link is obtained viz., the vertical movement of the 
 valve unobstructed by any lateral movement. 
 
 "In all cases the weight must be adjusted on the lever to the pressure of steam re- 
 quired in each case by a correct steam-gauge attached to the boiler. The weight must 
 then be securely fastened in its position and the lever marked for the purpose of facili- 
 tating the replacing of the weight, should it be necessary to remove the same." 
 
 Figure 1, Plate XXXIV., represents the form of safety-valve used on the boilers of 
 the U. S. S. Nipsic represented on Plates XII. and XXX. 
 
 Figure 3, Plate XXXIV., represents the Ashcroft spring safety-valve of the boilers 
 of U. S. S. Adams and class. One valve of this description, 3 inches in diameter, was 
 used on each boiler of this vessel. In addition there was one lever safety-valve of ordi- 
 nary construction, having a diameter of 10 inches, which was connected with one of the 
 main steam-drums and had an arrangement for lifting it from the fire-room. Each 
 boiler had 24 square feet of grate-surface and 598.6 square feet of heating-surface, and 
 was designed for a working pressure of 80 Ibs. per square inch above the atmosphere. 
 
346 STEAM BOILERS. CHAP. XV. 
 
 The Board of Trade (England) prescribes that, in case spring safety-valves are used 
 in passenger-steamers, there must be fitted to each boiler at least two separate valves ; 
 the spring and valve must be so cased in that they cannot be tampered with ; provision 
 must be made to prevent the valve flying off in case of the spring breaking ; screw lift- 
 ing-gear must be provided to ease all valves, if necessary, when steam is up ; the springs 
 must be protected from the steam and impurities issuing from the valves"-' when the 
 valves are loaded by direct springs the compressing-screw must abut against a metal 
 stop or washer when the load sanctioned by the surveyor is on the valve ; the size of 
 the steel of which the spring is made is found by the following formula : 
 
 D = diameter or side of square of the wire, in inches ; 
 
 d = diameter of the spring from centre to centre of wire, in inches ; 
 
 S = load on the spring, in pounds. 
 
 Ic = constant = 8,000 for round and 11,000 for square steel. 
 
 The accumulation of pressure is not to exceed 10 per cent, of the loaded pressure. 
 
 
 
 
 Fig. 139. 
 
 
 
 
 
 
 H j 
 
 
 
 
 
 I Q 
 
 
 
 
 t 
 
 
 
 When p = pressure of steam in pounds per square inch above the atmosphere, 
 A area of valve in square inches, 
 W = weight of the load applied to the lever, in pounds, 
 w = weight of the lever and its attachments, in pounds, 
 GC = distance of centre of gravity of lever from fulcrum C, 
 10^ weight of valve and spindle, in pounds. 
 DC = distance of axis of valve from fulcrum C, 
 BC = distance of centre of gravity of load from fulcrum (7, 
 
 the steam-pressure at which a safety-valve loaded by means of a lever (see figure 139) 
 will open is found by the formula : 
 
 i v flC< 
 P = (" "^^ *^ +*0-^. P.] 
 
SEC. 12. 
 
 BOILER MOUNTINGS AND ATTACHMENTS. 
 
 347 
 
 The weight of the load required with a lever of a given length for a given steam- 
 pressure is found by the formula : 
 
 BG 
 
 The length of lever required with a given weight of the load and a given steam-pres- 
 sure is found by the formula : 
 
 (pX A-w^DC-w&C 
 
 W 
 
 [III.] 
 
 The results given by the foregoing formulae are modified in practice by the friction 
 of the articulations and of the stem, and the position of the weight on the lever must be 
 finally adjusted when steam is on the boiler, so that the valve lifts at the required 
 pressure. 
 
 Fig. 140. 
 
 Figure 140 represents a practical method of weighting a safety-valve lever, taking 
 into account the load due to the weight of lever and valve and to friction, when it is not 
 convenient to adjust the valve under steam viz., calculate the total effective pressure, 
 p X A, acting on the valve ; apply at the end of the lever B an ordinary weigh-beam, 
 which tends to raise the lever-arm ; adjust the load P on the weigh-beam so that it will 
 
 fn \/ A "V* 7~)fy 
 
 balance a weight equal to fify - then adjust the weight TFon the valve-lever 
 
 till it brings the lever and the weigh-beam into a horizontal position. 
 12. Miscellaneous Attachments of Boilers. 
 
 ^Escape-pipes. The steam discharged by the safety-valves and by the exhaust of 
 the steam-pumps and other auxiliary engines is conducted to the escape-pipe and from 
 it discharged into the atmosphere. There are one or several escape-pipes, which are car- 
 ried up alongside the smoke-pipe to a greater or less height above the upper deck. 
 
348 STEAM BOILERS. CHAP. XV. 
 
 With a hoisting chimney the escape-pipe reaches either only to the top of the 
 stationary section of the chimney, or it is made telescopic, having a movable section 
 attached to the movable section of the chimney, which slides within the lower station- 
 ary section, the latter being provided with a stuffing-box at its upper end. 
 
 The escape-pipe is generally made of copper. Its upper end is either simply made 
 flaring or it is provided with an arrangement for intercepting the water which is carried 
 up with the steam (see figure 141). 
 
 Fig. HI. ^ e U- S. S. Nlpsio has a single escape-pipe, 9 inches in diameter, 
 
 which is carried up at the forward side of the chimney to the top of 
 the stationary section. The safety-valve chambers of the boilers on 
 each side of the vessel are connected with each other by pipes 5f 
 inches in diameter, and communicate with the escape-pipe by means 
 of a branch-pipe 7 inches in diameter. These pipes are made of cop- 
 per, No. 14 B. W. G. thick, and are connected by composition flanges. 
 The exhaust-pipes of auxiliary engines should be connected directly 
 with the escape-pipe, and not with any of the safety-valve chambers, 
 as is often done in order to effect a saving in the length of the pipes, 
 because the steam and condensed water leaking through the safety-valve will keep the 
 boiler damp when it is not in use. 
 
 Justice's quieting-chamber is designed to prevent the deafening noise produced by 
 steam issuing from safety-valves, from the exhaust of high-pressure engines, etc. The 
 chamber, which is either cylindrical or of any other convenient shape, is filled with 
 balls of suitable material and of proper size, confined compactly between copper grat- 
 ings ; for low pressures balls or beads of annealed glass are found best, while for high 
 pressures hollow copper or brass balls of small size are used. The current of steam 
 flowing through this chamber is broken up into numerous streamlets in its passage 
 through the tortuous interstices formed by the balls. The vibrations produced in the 
 several balls by the impact and friction of the steam are not uniform and interfere with 
 one another, and thus do not produce sound. By making the total area of the open- 
 ings sufficiently large the steam is allowed to escape without an appreciable increase of 
 back-pressure. The area of the exit-opening is always in excess of that of the inlet- 
 pipe. The cross-area of the chamber is proportioned to the pressure and volume of the 
 steam ; the diameter of the chamber is from five to six times the diameter of the escape- 
 pipe. The depth of the chamber is about 8 inches for all sizes. 
 
 Figure 142 shows a sectional elevation of a quieting-chamber, which communicates 
 below with a safety-valve and above with the escape-pipe. The safety-valve, which is 
 
SEC. 12. 
 
 BOILER MOUNTINGS AND ATTACHMENTS. 
 
 349 
 
 placed on the main steam-pipe between the boilers and a stop-valve, can be opened by 
 means of a hand-lever, and discharges the steam through the quieting-chamber into the 
 escape-pipe. 
 
 Figure 143 is a sectional elevation of another form given to the quieting-chamber. 
 The latter has a central pipe passing through it, which is provided with a valve, so that 
 the steam may be discharged either through the chamber or directly into the escape- 
 pipe. The valve may be loaded to discharge the steam automatically, or it may be 
 lifted by a hand-lever. 
 
 Similar quieting-chambers are introduced between the blast-pipes and nozzles of 
 locomotives, steam-launches, torpedo-boats, etc. 
 
 Fig. 143. 
 
 In Shaw's noise-quieting nozzle the steam escapes through cylindrical coils of wire, 
 the diameter and length of the coils being such that, when compressed nearly to contact, 
 the spaces between the turns of the coils will make a total area of opening much greater 
 than that of the steam escape-pipe. As the individual turns of the wire coils cannot 
 vibrate without coming in contact with the adjacent ones, interference and consequent 
 silence results in the same way as two vibrating piano- wires, if brought together, will 
 immediately destroy each other's sound. The spirals are opened wider by any increase 
 of steam passing through, and this action, as well as the tremulous motion of the spirals 
 produced by the issuing steam, prevents their clogging by an accumulation of foreign 
 
350 
 
 STEAM BOILERS. 
 
 CHAP. XV. 
 
 matter. The spiral coils are made of brass wire about f- inch thick, and are from 4 to 5 
 inches long. A greater or less number of coils are arranged in various ways for each 
 escape-pipe. 
 
 Figure 144 represents Shaw's 
 
 noise-quieting nozzles arranged in 
 
 clusters for large steamships. P is 
 
 the escape-pipe from safety-valve, on 
 
 which the casing D, containing the 
 
 cluster of nozzles, is bolted. G are 
 
 nozzles of spiral wire, having a solid 
 
 top, and secured to elbows, J, that 
 
 are connected with the escape-pipe 
 
 P. The steam escapes through the 
 
 wire coils into the casing D, whence 
 
 it is led to the escape pipe F. 
 
 Figure 145 represents a sectional 
 
 elevation and top view of another ar- 
 rangement of the noise-quieting noz- 
 zle. The nozzle is made entirely of 
 
 brass and copper. The bottom flange 
 
 connects with the escape-pipe close to 
 
 the safety-valve. The steam escaping 
 
 from the safety-valve is conducted 
 
 into the base B and into a central 
 
 tube, C, where it is distributed to 
 
 numerous coils of brass wire, F, secured in the top plate 
 of base B and in the central tube C. Escaping between 
 the turns of these coils, the steam enters the copper casing 
 P, from which it is conducted through the regular escape- 
 pipe to the outer air. A valve, Gr, is sometimes provided at 
 the top of the central pipe C, said valve being loaded with 
 about 2 Ibs. pressure, which guarantees a sufficient outlet 
 without reference to the coils F ; but this valve is not a 
 necessity, as abundant area of outlet exists in the brass 
 coils F. 
 
 oooUooo 
 ooooooo 
 
 30OOOOO 
 IOOOOOO 
 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 ooooooo 
 
 Fig. 145. 
 
 Bleeding valves and pipes are intended to pass waste steam into the condenser in- 
 
SBC. 13. BOILER MOUNTINGS AND ATTACHMENTS. 351 
 
 stead of blowing it off through the safety-valve. For this purpose a copper pipe, hav- 
 ing a diameter of about 4 inches, leads from the main steam-pipe to a stop-valve on the 
 top of the condenser, another stop- valve being placed between the main steam-pipe and 
 the bleeding-pipe. 
 
 The reverse or vacuum valve is a small safety-valve opening inwards, designed to 
 open and thus prevent the collapse of a boiler in case the pressure within the boiler 
 falls below the atmospheric pressure. These vacuum-valves were generally attached to 
 boilers as long as steam of very low pressure was used and the shell of boilers was pro- 
 portionately weak, but they are seldom used nowadays. 
 
 Stop-valves and steam-pipes are attached directly to the boilers to supply the steam- 
 pumps and other auxiliary engines, the distiller, steam-whistle, and steam-blast with 
 steam when communication between the main steam-pipe and the boilers is shut off. 
 
 Drain-cocks are fitted to the bottom of the boilers, to superheating-chambers and 
 steam-drums, and to all valve-chambers and pipes where water is liable to lodge after 
 the boilers are emptied. 
 
 In the U. S. S. Nipsic the drain-pipes discharge the waste water into a cylindrical 
 wrought-iron tank placed in the spandrel under the after boiler. The several drain- 
 pipes communicate with a common pipe, which is connected to a stop- valve placed on 
 the top of the tank. Another pipe, which is likewise provided with a stop-valve, takes 
 the water from the bottom of the tank to the feed-pumps. A gauge-glass shows the 
 height of the water in the tank. 
 
 13. Covering for Boilers. The saving in fuel which may be effected by prevent- 
 ing the loss of heat by radiation and convection from the shell of boilers, steam-pipes, 
 etc., has been discussed in section 6, chapter iii. Besides the economic advantage 
 resiilting from surrounding boilers with non-conducting materials, the reduction of the 
 temperature in the confined spaces around the engines and boilers of a vessel is of great 
 importance. It is also necessary, especially in single-deck vessels, to provide a tight 
 covering for the top of boilers to protect them from the water which may leak through 
 the deck. 
 
 The principal methods used for protecting the shells of marine boilere are the follow- 
 ing : the boilers are surrounded with an air-tight casing enclosing an air-space ; or they 
 are covered with hair-felt or other loose, fibrous material, held in place by an outer cas- 
 ing ; or they receive a thick coating of some cement applied like plaster to their surfaces. 
 Sometimes several of these methods are used in combination. These different methods 
 of covering can be made equally effective, as far as the prevention of loss of heat from 
 the covered surface is concerned, provided the casing is fitted with sufficient exactness 
 
352 STEAM BOILERS. CHAP. XV. 
 
 and the covering or coating is applied in sufficient thickness. Their relative value is, 
 therefore, to be measured by the weight, the first cost and the durability of the cover- 
 ing, and by the facility with which it can be removed and replaced for the purpose of 
 examining and repairing the covered parts. 
 
 The entire cylindrical shell and the back of the boilers of the U. S. S. MiantonomoTi 
 and class are covered with an air-tight casing of galvanized iron, enclosing an air-space 
 of 1J inches between the boiler and the casing. (See Specifications of Boilers of U. S. S. 
 MiantonomoTi and class, section 10, chapter vii.) 
 
 Cow-hair felt, stitched on canvas, weighing 1 pound per square foot when \\ inches 
 thick, has usually been employed for covering the boilers of United States naval ves- 
 sels. The specifications for the cylindrical boilers of the U. S. S. Adams and class pre- 
 scribe that "after the boilers are in the vessel, have been tried with steam, and all 
 leaks have been made tight, the boilers are to be covered with felting \\ inches thick, 
 strongly stitched to No. 1 canvas, and secured by four hoops, 2 inches wide, encircling 
 the boiler ; and over this is to be placed sheet-lead of No. 14 wire-gauge, securely sol- 
 dered at all edges." 
 
 Rectangular boilers are covered in this manner on the top, and on the back and sides 
 for some distance down ; the temperature of the water in the lower part of the boiler 
 is generally so much less than the temperature of the steam that it is not necessary 
 to extend the covering to the bottom of the boiler. On the felting is frequently placed 
 a covering made of wooden staves, tongued and grooved. The manner of securing the 
 wood casing to launch-boilers is shown on Plate XVI. On steam-pipes these staves 
 are generally held together by brass hoops drawn together by a single bolt. 
 
 When steam of more than 45 pounds pressure is used its temperature is sufficient 
 to char the felt when it comes in immediate contact with the metal of the boiler. On 
 this account various contrivances have been designed for maintaining a narrow air-space 
 between the felt and the boiler, and asbestos boards or other mineral substances of 
 low thermal conductivity are sometimes placed between the boiler and the felt. 
 
 Various mastic compositions of clayey material, and cements containing an admix- 
 ture of asbestos in greater or less proportions, are applied like plaster either directly 
 to the surface of the boiler-shell or to a wire netting stretched over the latter. The 
 weight of such a covering is considerable. The most serious objection to its use on 
 steam-chambers is the difficulty of removing and replacing it for the purpose of examin- 
 ing the covered parts. 
 
 ''Mineral wool" is a loosely-cohering, fibrous substance resembling coarse wool, 
 formed by blowing a jet of steam into a stream of fluid slag. It is used as a non-con- 
 
SEC. 14. BOILER MOUNTINGS AND ATTACHMENTS. 353 
 
 ductor by packing it in a space formed around the protected vessel by an outer casing. 
 It sometimes contains foreign substances which attack the iron under the influence of 
 heat and moisture. 
 
 14. Feed-water Heaters and Filters. Feed-water heaters are designed to utilize 
 waste heat and to lessen the difference of the temperatures of the steam and water in 
 the boiler. The latter is an important consideration ; the introduction of a mass of 
 cold water in a highly -heated boiler causes injurious local contractions, and the diffe- 
 rence of the temperatures in the top and bottom of cylindrical boilers, amounting often 
 to from 100 to 200 Fahr., produces often far greater strains than those due to the 
 steam-pressure. 
 
 The saving of heat effected by the use of heaters, in per centum of the total heat 
 expended, may be expressed by the formula : 
 
 .*.-* 
 " T t' 
 
 where t, and T are the temperatures of the feed-water before and after it passes 
 through the heater, and the total heat of an equal weight of steam, respectively. 
 This saving may be in many cases to a great extent counterbalanced by the cost and 
 weight of, and the space occupied by, the heater, and by the additional cost and labor 
 required to keep it in order. 
 
 When the temperature of the escaping gases in the boiler-uptake exceeds the limit 
 given in section 11, chapter ii., the economic and potential evaporative efficiency of the 
 boiler will be increased by utilizing the excess of heat in raising the temperature of the 
 feed- water. With a well-designed boiler this should not be necessary, and the arrange- 
 ment of the heater- pipes in the uptake of a marine boiler presents many difficulties 
 and inconveniences. 
 
 When the engines are fitted with a jet-condenser, and the boilers are fed with salt 
 water, heaters are used to advantage to impart to the feed- water a portion of the heat 
 contained in the supersalted water which is blown off to reduce the saturation of the 
 boiler. In the U. S. S. Wabash each boiler, containing 83.5 square feet of grate-sur- 
 face, was provided with a heater lying beneath the floor-plates of the fire-room. This 
 heater was composed of a cast-iron cylindrical shell, 12 inches in external diameter, 
 containing 31 brass tubes \\ inches in external diameter and 13 feet long. The super- 
 salted water of the boiler was blown off continuously from the surface by a cock and 
 pipe, and passed around the tubes on its way to the sea, while the continuous feed 
 passed through these tubes on its way from the hot- well to the boiler. When the 
 
354 STEAM BOILERS. CHAP. XV. 
 
 water of the boiler was kept at a density of If thirty-seconds the feed-water received 
 an accession of temperature of about 30 Fahr. 
 
 With non-condensing engines the exhaust steam may be used to heat the feed-water. 
 The steam is either blown into an open tank, where it mingles with and is condensed by 
 a shower of water, or the heater is constructed on the principle of a surface-conden- 
 ser the steam being condensed as it passes through a nest of tubes around which 
 the feed-water circulates. 
 
 The Berryman heater consists of a closed cylindrical wrought-iron tank, the bottom 
 of which is bolted to a casting divided by a partition into two compartments. The inte- 
 rior of the tank is occupied by a number of siphon-shaped brass tubes, which are se- 
 cured with both ends in the bottom plate of the tank in such a manner that their ends 
 communicate with either compartment of the lower casting. The exhaust steam enters 
 one of these compartments, and, after parting with its heat in passing through the tubes, 
 it is discharged from the other compartment. The feed-water is forced by the pump 
 into the tank near the bottom, and passes out through an opening at the top. With 
 this arrangement the warmest water rises naturally to the top and passes off to the 
 boiler, and foreign matter held in suspension in the feed-water has a chance to settle 
 in the bottom of the tank. 
 
 Filters. United States naval boilers using high pressures of steam have been fitted 
 with feed-water filters consisting of a tank divided by screens into several compartments, 
 which are filled with various substances for filtering the water or neutralizing the fatty 
 acids contained in it. 
 
 In Selderi* s filter, as fitted to the U. S. S. Miantonomo7t, the water coming from the 
 hot- well enters at the top of a tank, and passes through a vertical partition formed by a 
 sheet of Burlap cloth placed between two wire screens into an upper compartment which 
 is filled with coke. The plate forming the bottom of this compartment is perforated at 
 one end with 62 f -inch holes, through which the water passes into the lower compart- 
 ment, which is filled with sponge. The bottom plate of this compartment has an equal 
 number of holes near the opposite side of the tank, through which the water flows 
 into the channel-way, whence it is withdrawn by the feed-pump. Doors are provided 
 for removing the screens, the coke, and the sponge for the purpose of cleaning or renew- 
 ing them. 
 
 15. Feed-pumps and Injectors. When the boilers are fed with fresh water the 
 weight of water evaporated as a maximum is the least quantity which the feed-pump 
 must be capable of delivering in a given time ; when sea-water is used the weight of 
 water to be blown off to maintain the water in the boiler at the proper density has to be 
 
SEC. 15. BOILER MOUNTINGS AND ATTACHMENTS. 355 
 
 added to the weight of water evaporated in a given time. The feed-pump should, how- 
 ever, be capable of discharging about twice the quantity of water evaporated and blown 
 off in a unit of time, in order to supply losses due to priming and leakage and to make 
 allowances for irregularities in the working of the pump. 
 
 In calculating the dimensions of a pump it may be assumed that the volume of the 
 water discharged is ordinarily about 75 per cent, of the space displaced by the plunger 
 per stroke of pump. 
 
 According to the foregoing rule the capacity of the main feed-pump connected with 
 the engines is to be calculated, as well as the least capacity of the auxiliary steam- 
 pump. But it is advantageous to increase the dimensions of the latter pump so that it 
 may be worked at a low speed ; because with a slow- working pump shocks in the feed- 
 pipes are avoided, and the hammering action of the check-valves is lessened, and the 
 volume of water discharged relatively to the space-displacement of the piston is in- 
 creased. 
 
 The feed-water Injector was invented by Giffard, a French engineer, in the year 
 1858. In its simplest form (see figure 146) the injector consists of a pipe, A, for the ad- 
 mission of steam, which, escaping through 
 the conical nozzle of the receiving-tube C 
 at a high velocity, is joined by water 
 which, flowing in through the pipe B, 
 mingles with and condenses the steam in 
 the conical combining -tube D. The con- 
 densed steam gives an impulse to this water, which is driven in a continuous stream 
 through the delivery-tube H and the check- valve J into the boiler, provided it possesses 
 sufficient velocity. During the passage of the water from D to H it is driven across the 
 space F, called the overflow, which communicates by means of the over/low-nozzle G 
 with the outside air. If too much water is supplied to the steam some water may 
 escape at this point and flow out through the overflow-nozzle ; if there be too little 
 water air will be drawn in at G and carried into the boiler with the water. 
 
 The fact that a mass of steam should be capable of imparting to a much larger mass 
 of water sufficient velocity to overcome even a higher pressure than that which caused 
 the original motion of the relatively small mass of steam, has frequently been looked 
 upon as a paradox. This action of the injector depends on the following prin- 
 ciples : 
 
 The change in the molecular condition of the steam by condensation does not affect 
 the motion of its particles. The mass of condensed steam, moving with its original high 
 
856 STEAM BOILERS. CHAP. XV. 
 
 velocity, produces a concentrated effect by its impact on the mass of the condensing 
 water ; and the resultant momentum of the two unelastic fluids is equal to the sum 
 of the momenta of their masses before the impact. (See equation [II.]) 
 
 An injector made as shown in figure 146 is known as a fixed-nozzle injector. With 
 a given steam-pressure it will give a constant feed of a given quantity. To adapt the 
 instrument to variations in the steam-pressure, and to effect variations in the quantity 
 of feed- water delivered, the areas of the openings of the conical nozzles have to be 
 altered in order to diminish or increase the steam or water supply. This adjustment 
 was an essential feature of Giffard's injector. It is usually effected by means of a taper- 
 ing spindle which can be raised or lowered by means of a screw within the receiving- 
 tube, and by making either the receiving-tube or the combining-tube movable, so that 
 by raising or lowering the same the annular space between the receiving-nozzle and the 
 combining-tube is either enlarged or contracted. In the fixed-nozzle injectors the ad- 
 mission of steam and water is regulated by stop-valves in the supply -pipes ; but the 
 range of these instruments, as far as steam-pressure, temperature, and quantity of 
 feed-water are concerned, is much more limited than that of the adjustable in- 
 jectors. 
 
 The first action of the steam-jet issuing from C (see figure 146) is to drive the air out 
 of the tube D, thus forming a more or less perfect vacuum in the chamber surrounding 
 the nozzle C, in consequence of which the water will be lifted to a greater or less 
 height in the supply-pipe. This lifting power of the injector may be greatly improved 
 by giving to the openings suitable forms and dimensions. The water is frequently 
 lifted from 6 to 8 feet ; and it is claimed that with some large injectors of improved form 
 a lift of 18 feet has been obtained. The water to be lifted must be free from air, and 
 its initial temperature must be less than the boiling-point of water under the dimin- 
 ished pressure existing in the chamber surrounding C. Therefore, when a feed-water 
 heater is to be used in connection with an injector, it is better to place it between the 
 latter and the boiler. 
 
 The quantities of steam and water admitted must be so regulated that the jet of 
 steam is completely condensed ; otherwise a certain quantity of vapor will enter the 
 chamber F and escape into the atmosphere, proving a complete loss. The temperature 
 of the water-jet issuing from D must be less than 212 Fahr., otherwise the water will 
 vaporize as it is brought into communication with the atmosphere in passing from D to 
 H. The steam must be perfectly dry to give the best results. 
 
 The relation existing between the quantities of water and steam admitted, and their 
 respective temperatures, is expressed by the following equation : 
 
SEC. 15. BOILER MOUNTINGS AND ATTACHMENTS. 357 
 
 Calling t final temperature of the feed-water discharged at H, 
 S initial temperature of the feed- water entering at B, 
 T total heat contained in a pound of steam, 
 Q = weight of steam in pounds expended in a unit of time, 
 q = weight of water entering at B in a unit of time, 
 
 we have (Q+q) t = Q T+qS; 
 
 hence Q(T-f) = q(t - S) ; 
 
 tL-T^L [I] 
 Q " t-* 
 
 As the quantity Ovaries little with the pressure of the steam, the value of the pro- 
 portion ^~ depends principally upon t and $ that is, the final and initial temperatures 
 
 V 
 
 of the feed-water. 
 
 Neglecting the effect of friction and other disturbing influences occurring in practice, 
 the relation existing between the velocities of the steam-jet and water-jet is expressed 
 by the following equation, in which 
 
 v = the velocity of the mass of steam Q issuing from the receiving-tube ; 
 
 U = the initial velocity of the mass of water q entering at B ; 
 
 W = the velocity of the water-jet issuing from the combining-tube : 
 
 o (Q + 4) = Qv-\-qu; 
 Qv + qu n] 
 
 Calling F the cross-area of the nozzle of the receiving-tube, 
 
 FI = the cross-area of the larger orifice of the combining-tube, 
 
 G = the cross-area of the nozzle of the combining-tube, 
 
 (?, = the cross-area of the feed-pipe, 
 
 w, = the velocity of the water passing through G H 
 
 m = the specific volume of the steam, 
 
 we can represent the mass of water delivered in a unit of time by the following ex- 
 pressions : 
 
 Q + q = Gw = G l w l = ^ + F l u. [III.] 
 
 The water-jet entering the boiler must perform the same amount of work as an equal 
 mass of water issuing from the boiler under the pressures existing within the boiler 
 
358 STEAM BOILERS. CHAP. XV. 
 
 and the injector would be capable of doing. This is expressed by the following 
 equation : 
 
 where h = the height of a column of water representing the absolute steam-pres- 
 
 sure in the boiler ; 
 h t = the height of the nozzle of the receiving-tube above the water-level 
 
 within the boiler ; 
 x = the height of a column of water representing the absolute pressure of 
 
 the water at the mouth of the combining- tube. 
 
 Calling Tc = the height of a column of water representing the atmospheric pressure, 
 A, = the height to which the water is lifted, 
 
 u* 
 
 we have Jc h, x = -^ . [V.] 
 
 (See Weisbach, ' I/eTirbuch der Ingenieur und MascMnen-MecJianiTcJ dritter The/I 
 II. Abtheilung.) 
 
 By means of the foregoing equations the duty of an injector under given conditions, 
 and the cross-area of the openings of the various parts of an injector for a given duty, 
 may be calculated. 
 
 The causes decreasing the efficiency of injectors are friction, the shocks experienced 
 by the water in the passages, the incomplete condensation of the steam and the admix- 
 ture of air with the water- jet, and the waste of water at the overflow. 
 
 The least quantity of water which can be delivered by an injector is generally not 
 less than 60 per cent, of the largest quantity delivered under the same conditions of 
 temperature and pressure. 
 
 Experiments made in the year 1879 on Irwin's injector (see Franklin Institute Jour- 
 nal, February, 1880) indicate that on the whole the ratio of the weight of water de- 
 
 livered to the weight of steam used (or -$-. of equation [I.]) decreased as the pressure of 
 
 Tb/ 
 
 steam increased from 15 to 105 pounds per sq. inch above the atmosphere ; that, on the 
 contrary, the work done, expressed in foot-pounds per pound of steam, increased under 
 the same conditions, the water being delivered in every case against a pressure equal to 
 
 that of the steam used ; that the ratio Q- decreased likewise when the delivery of water 
 
 V 
 
 was less than the maximum for the pressures and temperatures of steam and water. 
 The largest amount of work was done when the steam and water pressures were 90 
 
SBC. 15. BOILER MOUNTINGS AND ATTACHMENTS. 359 
 
 pounds per square inch above the atmosphere, and the injector delivered 15.71 Ibs. of 
 water, supplied under a head of 6 inches, per pound of steam ; the work done in this 
 case being equal to 3449.94 foot-pounds per pound of steam used. Assuming that a 
 pump uses from to 1 Ibs. of steam for every 33,000 Ibs. of water lifted one foot high, 
 the highest efficiency of the injector in the above experiments was nearly from 13 to 
 6.5 times less than the efficiency of a steam-pump. Since, however, in the injector 
 nearly all the heat of the steam which is not converted into mechanical work is utilized 
 in raising the temperature of the feed- water, the injector compares favorably with a 
 steam-pump as a feed apparatus. 
 
 There are certain inconveniences connected with the use of injectors which have pre- 
 vented its general adoption as a feed apparatus for marine boilers viz., the rolling of 
 the ship is apt to cause a break in the water- jet ; the foaming of the boilers interferes 
 with its action ; it is easily disarranged by particles of salt or other solid matter en- 
 trained by the steam or the feed- water ; any air entering through the overflow spoils the 
 vacuum of the condenser. Besides, all steam- vessels must carry steam-pumps for vari- 
 ous purposes, so that the addition of injectors is unnecessary. 
 
 The difficulty of making by hand the proper adjustments regulating the admission 
 of steam and water, which become necessary whenever the steam-pressure changes, led 
 to the introduction of the self-adjusting injector, manufactured by William Sellers & 
 Co., Philadelphia (see figure 2, Plate XXXV.) The upper end of the combining-tube 
 G is made in the form of a piston, which slides freely in the exterior case ; the lower 
 part of the combining-tube is guided by a sleeve on the upper end of the delivery-tube 
 D. The overflow- valve is closed as soon as the apparatus begins to work ; if now the 
 water-supply becomes too great a portion of the water escapes by the opening O in the 
 upper part of the delivery-tube, and, accumulating in the chamber surrounding the 
 combining-tube, presses under the piston and raises the combining-tube ; on the other 
 hand, when the feed-supply is insufficient a partial vacuum is formed under the piston, 
 and the combining-tiibe is forced down till the increased feed-supply establishes equili- 
 brium on both sides of the piston. In this manner the instrument regulates automati- 
 cally the water-supply so as to give always the best result with the pressure and weight 
 of steam used, and the indraught of air and waste of water at the overflow is avoided. 
 
 The conical spindle which regulates the flow of steam in the receiving-tube is per- 
 forated by a narrow passage bored along its axis, which communicates with the steam- 
 space through grooves at the screwed end, when the valve W, formed by an enlargement 
 of the spindle, is raised. The valve W seats on the upper side of a second valve, X, 
 which in turn seats on the receiving-tube A. The small jet of steam which escapes 
 
360 
 
 STEAM BOILERS. 
 
 CHAP. XV. 
 
 through the passage in the spindle before the valve X is raised is more effective in ex- 
 hausting the air and lifting the water in the supply-pipe than a jet escaping through 
 the narrow annular space between the taper plug and the receiving-nozzle. 
 
 With this central jet water is raised from 10 to 18 feet in the supply-pipe, according 
 to the size of the instrument. 
 
 To start the instrument the lever K is drawn back a short distance till the collar on 
 the spindle comes in contact with the lower side of the valve X. As soon as water ap- 
 pears at the overflow the lever is drawn entirely back and the valve X lifted from its 
 seat, admitting a free flow of steam through the receiving-tube. Then, after closing the 
 overflow- valve by rod L and lowering latch V into teeth of ratchet, the lever K may be 
 pushed in to any required point between the stops on the rod L so as to obtain the de- 
 sired water-supply. 
 
 This instrument has a greater range than the ordinary adjustable injector : the mini- 
 mum water-supply is 40 per cent, of the maximum supply, a larger quantity of water 
 is discharged by instruments of the same size, and the self-adjusting injector is capable 
 of working with hotter water. Experiments made by the manufacturers with this in- 
 strument gave the following results : 
 
 Pressure of steam in pounds per square inch 
 
 20 
 
 40 
 
 60 
 
 80 
 
 IOO 
 
 120 
 
 140 
 
 ISO 
 
 Admissible temperature of feed-water before enter- 
 
 
 
 
 
 
 
 
 
 
 i?8 
 
 iis 
 
 T 1O 
 
 1 1Q 
 
 132 
 
 T 2 ? 
 
 127 
 
 128 
 
 
 
 OJ 
 
 J 3^ 
 
 *a w 
 
 
 A oo 
 
 l * 1 
 
 
 The self-adjusting injector works to the best advantage when it is lifting water, and 
 in no case must the water be fed to it under pressure. 
 
 The numerical size of an injector is the diameter of the smallest part of the delivery- 
 tube expressed in millimetres. 
 
 Numerous modifications have been made in the form of injectors by different 
 makers with a view to simplifying their construction and manipulation and extending 
 the range of their action. 
 
 Figure 3, Plate XXXV., represents Koerting 1 s universal lifting -injector, which 
 consists of two injectors combined in the same chamber in such a manner that the de- 
 livery-tube of the first injector communicates by means of lateral passages with the 
 combining- tube of the second injector. The two steam-valves V, V, and the overflow- 
 cock E, are connected with the lever A in such a manner that the same movement of 
 this lever sets the apparatus in operation. By moving the lever in the direction of the 
 
SEC. 15. BOILER MOUNTINGS AND ATTACHMENTS. 361 
 
 arrow the valve V is first raised slightly from its seat, and as the steam rashes out 
 through the open overflow-cock E the water is lifted and enters J. By the continued 
 movement of the lever the valve V is opened wide, and as soon as the first injector is in 
 operation its communication with the overflow by means of E is closed, and the water 
 is forced into the combining-tube of the second injector. At this moment the steam- 
 valve V of the second injector begins to lift, and when the second injector is in full 
 operation its communication with the overflow by means of E is likewise closed, and 
 the water is forced through the check-valve into the boiler. These several operations 
 take place in such rapid succession that, practically, it is sufficient to raise the lever to 
 start the apparatus. 
 
 It is claimed that with this injector water having an initial temperature of 156 
 Fahr. can be lifted, and that the temperature of the water is raised to 190 Fahr. in the 
 combining-tube of the first injector, the pressure produced in the passages between the 
 first and second injectors being considerable, so that the boiling-point of the water is 
 raised and the condensation of a greater quantity of steam is made possible. 
 
CHAPTER XVI. 
 
 TESTS, INSPECTIONS, AND TRIALS OF STEAM BOILEES. 
 
 1. Testing Boilers. All new boilers and all boilers that have been extensively re- 
 paired must be subjected to a hydraulic pressure in excess of the highest working pres- 
 sure, in order to test the tightness of the seams and rivets, the soundness of the plates, 
 and the structural strength of the boilers. Such tests must be repeated periodically 
 during the lifetime of the boiler. 
 
 A test-pressure equal to three times the working pressure was formerly held neces- 
 sary by many authorities, but nowadays it is not considered prudent to subject marine 
 boilers of the ordinary form to so severe a test. An excessive pressure may produce 
 injuries which do not become apparent during the short test, but which continue to 
 increase under the ordinary working pressure when the boiler is put into regular use. 
 The test-pressure must in no case strain any part of the boiler beyond the limit of elas- 
 ticity of the metal. 
 
 Section 4418 of the 'Eevised Statutes of the United States' provides that "all boilers 
 used on steam -vessels, and constructed of iron and steel plates, inspected under the 
 provisions of section 4430 (see section 2, chapter v.), shall be subjected to a hydrostatic 
 test in the ratio of 150 Ibs. to the square inch to 100 Ibs. to the square inch of the work- 
 ing steam-power [pressure] allowed." 
 
 United States naval boilers, when new or extensively repaired, are also subjected to 
 a test-pressure equal to one and a half times the highest working pressure above the 
 atmospheric pressure. 
 
 French laws require that tubular boilers of merchant-vessels are to be tested to 
 double the working pressure above the atmosphere at least once a year and whenever 
 repairs or alterations have been made on them. The boilers of French naval vessels are 
 subjected, when new, to a test-pressure equal to twice the working pressure, and annu- 
 ally thereafter to a test of one and a half times the actual working pressure above the 
 atmosphere ; but this pressure is to be kept on the boilers not longer than five minutes. 
 (Ledieu, ' Traite des Appareils d Vapeur de Navigation," 1 vol. ii.) 
 
SEC. 1. TESTS, INSPECTIONS, AND TRIALS OF STEAM BOILERS. 363 
 
 The 'Steam Manual,' issued by the English Admiralty (1879), contains the follow- 
 ing instructions regarding " Periodical testing by water-pressure of the boilers of Her 
 Majesty's ships and vessels in commission": 
 
 "In the case of ships having new boilers, or boilers repaired for a commission of 
 four years, the boilers are to be tested by water-pressure at the end of two years' 
 service, and subsequently at half-yearly intervals during the remainder of the 
 commission. 
 
 " As regards ships whose boilers have been repaired for shorter periods of service 
 the boilers are to be tested by water-pressure at the end of six months' service, and 
 subsequently at half-yearly intervals. 
 
 "During the application of water-pressure the boilers are to be carefully examined, 
 and proper gauges are to be used to detect any change in the form of the furnaces, 
 combustion-chambers, etc. 
 
 "The water-pressure is to be double the working pressure, provided that during the 
 examination no indications of weakness are observed. Should, however, any indica- 
 tions of probable permanent deformation be observed^the test is to cease, and the work- 
 ing pressure is then to be limited to one-third that of the test-pressure arrived at before 
 such indications were seen. 
 
 " The water-pressure is intended to supplement, not to supersede, the occasional 
 drill-testing. Should the latter test reveal unusual thinness of any plates the water- 
 pressure is to be very carefully applied, in order that injury may not be caused by 
 overpressure." 
 
 The Board of Trade (English) regulations for marine boilers provide as follows: 
 "All new boilers, and boilers that have been taken out of ships for thorough repair, 
 must be tested by hydraulic pressure up to at least double the working pressure that 
 will be allowed, previous to the boilers being replaced in position, to test the workman- 
 ship, etc. ; but the working pressure is to be determined by the stay -power, thickness of 
 plates, and strength of riveting, etc." 
 
 Anderson states that the boilers belonging to the British War Department are sub- 
 jected periodically, after about every 500 hours of actual work, to a hydraulic pressure 
 equal to double the pressure to which the safety-valve is ordinarily loaded, or to one- 
 third of their ultimate strength. 
 
 The usual method of testing boilers is to fill them with water and produce a pres- 
 sure within them by means of a hand force-pump. All the openings of the boiler are 
 securely closed. The safety-valve, which is loaded to the required test-pressure, is 
 kept raised till the boiler is completely filled with water. Then, after closing the 
 
364 STEAM BOILERS. CHAP. XVI. 
 
 safety-valve, the pump is worked till the steam-gauge indicates the test-pressure. The 
 pump should deliver only a small quantity of water at each stroke, and must be worked 
 carefully as the pressure rises, in order to avoid jarring the strained boiler and pro- 
 ducing a sudden rise of pressure beyond the limit of the test-pressure. Some engineers 
 close the safety-valve before the boiler is quite full of water, and so retain a quantity of 
 air to act as a cushion when the pressure is applied by the pump ; but when this en- 
 closed air escapes through leaky seams and rivets no marks indicating such leaks are 
 left on the plates. 
 
 When a boiler is connected with a high steam-drum the difference of the pressures 
 at the top of the steam-drum and at the bottom of the boiler, equal to the weight of a 
 column of water of corresponding height, may be a considerable quantity. The rules 
 and regulations of the Board of Supervising Inspectors of Steam -vessels provide that, 
 "in applying the hydrostatic test to boilers with a steam-chimney, the test-gauge should 
 be applied to the water -line of such boilers." 
 
 The test should be applied before the boiler is painted and while every part is ex- 
 posed to view. New boilers should be tested before they leave the boiler-shop. The 
 boiler is placed on blocks so that the bottom may be inspected, and the furnace and 
 connection doors are kept wide open. Every part of the boiler is watched and care- 
 fully examined while the pressure continues, to discover any leaks in rivets, seams, or 
 tubes, or through cracks in the plates, and any signs of bulging of stayed surfaces or 
 of collapse of flues. Leaky tubes, rivets, and seams are marked, and are calked after 
 the boiler has been relieved of the pressure. Flat stayed surfaces and flues should be 
 accurately gauged before and during the test. 
 
 "After the test-pressure has been maintained some time the measurements pre- 
 viously obtained should be checked, and any extension, distortion, bulging, etc., care- 
 fully noted. Then again, when the pressure is relaxed, which may be done suddenly, 
 it should be ascertained whether any changes of shape that may have been found are 
 permanent or not. If there be any permanent enlargement or distortion, even of the 
 slightest degree, it should be satisfactorily examined to decide whether it is due to the 
 elastic limit of the material having been exceeded or to malconstruction. There are 
 cases, as, for instance, with flat surfaces, where a permanent set might take place, and 
 which would be quite safe at the ordinary working pressure. This is especially the case 
 with stayed surfaces, for it seldom happens that each stay in a series takes its due pro- 
 portion of load until the stays have been stretched or the plates distorted by the pres- 
 sure. 
 
 "But cases of a permanent flue-tube distortion or flattening must always be treated 
 
SEC. 1. TESTS, INSPECTIONS, AND TRIALS OP STEAM BOILERS. 365 
 
 with the greatest caution, since the change of shape is liable to become aggravated on a 
 subsequent application of the same or even a less pressure. In all cases where a perma- 
 nent set is discovered the test should be repeated again and again, if necessary, to as- 
 certain if the set becomes increased." ( Wilson.) 
 
 Time is an important element in boiler-tests. A boiler which bears a momentary 
 pressure without apparent injury may burst with the same pressure continued through 
 half an hour. No boiler should be considered safe if unable to bear the test-pressure 
 for a considerable length of time. The test-pressure should always be maintained at 
 least long enough to enable the inspector to make a thorough examination of all parts 
 of the boiler. 
 
 "Want of tightness in the joints is often revealed by leakage only after the pres- 
 sure has been applied for some time. In explanation it may be stated that the steam 
 or water leaking from a joint does not always find its way between the plates imme- 
 diately opposite the point of issue, but the actual source of the leakage, as we may call 
 it, is at some point perhaps several inches distant, whence it requires a considerable time 
 to force its way to the point where it makes its appearance. There can be no doubt 
 that, from the manner in which boilers are usually put together, the internal pressure is 
 not equally resisted by all parts of the shell, and produces an undue and often very 
 severe strain on one plate or portion of a plate. This is probably the cause of many 
 leakages that occur, and which only ' take up ' after the plate becomes stretched and 
 relieved of the extra strain, and it is, therefore, advisable in testing to allow the pres- 
 sure to act long enough to stretch such weak portions. . . . 
 
 " It is often much more difficult to keep a boiler perfectly tight and free from oozing 
 at the rivets, plate-edges, stays, and tube-ends under a very high water-pressure than 
 under an equal pressure of steam. This is probably owing to the fact that the high 
 temperature in the latter case tends to close the joints, and with certain kinds of water 
 any slight oozing is found to take up by the opening becoming closed with deposit or 
 corrosion which is induced by the high temperature." ( Wilson.) 
 
 Cold water is generally used in testing boilers. Some engineers advocate the use of 
 hot water, because the expansion of the metal due to the higher temperature brings the 
 different parts of the boiler more nearly under the conditions of stress which obtain 
 when the boiler is in actual use, and because at low temperatures iron is more easily 
 injured by strains. The water should, however, not be so hot as to be liable to cause 
 injury by scalding in case of serious leaks, or to interfere with a thorough examination 
 of the boiler within the furnaces and connections as well as outside. The effects pro- 
 duced by the uniform expansion of the whole boiler when hot water is used are, how- 
 
366 STEAM BOILERS. CHAP. XVI. 
 
 ever, very unlike the effects produced by the local expansion of the parts in immediate 
 contact with the fire and hot gases. 
 
 Boilers have been tested by filling them completely with water and lighting a fire in 
 the furnaces, the pressure being produced by the expansion of the water. (See Specifi- 
 cations of Boilers of U. S. S. Miantonomoh and class, section 10, chapter vii.) It 
 is claimed that with this method the increase of pressure is much more gradual than 
 that produced by a pump, and that the conditions of actual practice, as far as diffe- 
 rences of temperature are concerned, are at least approximately obtained. But a care- 
 ful examination of the furnaces and back-connections is not possible with this 
 method. 
 
 According to Wilson, it is not an unfrequent practice in England to test new boil- 
 ers by steam under a pressure one and a quarter or one and a half times greater than 
 the working pressure. It is argued that this is the only method by which the same 
 conditions of strain can be produced as obtain when the boiler is worked. But this 
 practice is to be condemned, not only because it is dangerous, but because it renders a 
 careful examination of the furnaces and back-connections, while the pressure lasts, im- 
 possible. 
 
 Boilers that have been tested with water-pressure should be tested under steam to 
 their working pressure, in order to prove their tightness after they have been located 
 and connected in the vessel, and before their shell has been covered with felt or other 
 non-conductive material. Every leaky seam, rivet, or tube should be made tight before 
 the boiler is finally accepted for service. 
 
 2. Inspection of Boilers. The testing of boilers by hydraulic pressure has to be 
 regarded merely as an auxiliary means for ascertaining the strength and workmanship 
 of a boiler ; it should never be considered as making a careful examination of every 
 accessible part unnecessary. Boilers which are faulty in design, or built of inferior 
 material, or have bad workmanship put on them may stand the hydraulic test, but, 
 under the varying and continued strains of actual practice, they will sooner or later 
 develop weaknesses which seriously impair their life and safety. Grave defects may 
 be hidden from view after a boiler is built so that they cannot be discovered by 
 the closest scrutiny ; therefore the inspection of boilers should commence with the 
 process of construction, and should be repeated frequently during the lifetime of the 
 boiler. 
 
 Section 4418 of the 'Revised Statutes- of the United States' provides that "the local 
 inspectors shall inspect the boilers of all steam- vessels before the same shall be used, 
 and once at least in every year thereafter. They shall subject aD boilers to the 
 
SBC. 8. TESTS, INSPECTIONS, AND TRIALS OF STEAM BOILERS. 367 
 
 hydrostatic pressure, and shall satisfy themselves by thorough examination that the 
 boilers are well made, of good and suitable material, etc." 
 
 The regulations of the Board of Trade (English) for the survey of marine boilers pro- 
 vide that, when a boiler is not open to inspection during the whole period of its con- 
 struction, the factor of safety for cylindrical boilers is to be increased 27.5 per cent, (see 
 section 3, chapter ix.) ; that special attention should be paid to the survey of super- 
 heaters, which must be inspected inside and out ; that the hammer-test should not 
 be relied on entirely for superheaters, but that the plates should be drilled occa- 
 sionally. 
 
 When boilers for United States naval vessels are built under contract at private 
 establishments inspecting engineer officers are detailed to watch their construction, and 
 to see that they are built in strict conformity to the drawings, and that the mate- 
 rial and workmanship are of the best quality 'and in accordance with the speci- 
 fications. 
 
 The attention of inspecting officers is to be directed especially to the following 
 points : 
 
 All the material must be of the proper quality, without flaws, and of the prescribed 
 dimensions. It is not an uncommon practice with boilermakers to use plate-iron of an 
 inferior quality for the internal parts of a boiler, which are hidden from view when the 
 boiler is finished. 
 
 Flanged plates must show no cracks or signs of laminations. Cracks extending from 
 punched holes to the edge of the plate, or from hole to hole, are dangerous sources of 
 weakness, and frequently indicate an inferior iron. Cracks are often produced in a row 
 of rivets by drifting. 
 
 When the rivet-holes in a seam do not come fair they should not be corrected by 
 drifting ; nor is the use of smaller rivets in half-blind holes to be permitted. When the 
 half -blind holes of a seam are corrected by punching or drilling so much of the metal 
 may be cut away that the strength of the joint is seriously impaired, or that the holes 
 can be closed only imperfectly by the rivets. 
 
 All plates drilled in place must be taken apart, and the burr must be removed from 
 all drilled holes. 
 
 When plates are cut too small the boilermaker often tries to correct the mistake by 
 punching the holes of the seams closer to the edge of the plates. All laps must be of 
 the proper width, and the joints of contiguous plates must be placed as far as possible 
 apart. 
 
 When the laps of plates do not lie close together the boilermaker often tries to cor- 
 
368 STEAM BOILERS. CHAP. XVL 
 
 rect or hide the evil by excessive calking, or by the insertion of pieces of hoop-iron, or 
 by filling the open space with cement of cast-iron borings mixed with sal-ammoniac. 
 By these means the bad workmanship may be concealed during the cold-water test, but 
 it will cause trouble sooner or later under steam. 
 
 The edges of all plates should be planed or chipped fair before calking. 
 
 See that the proper width of water-spaces is maintained between the shell of the 
 boiler and the back-connections, and that there is sufficient clearance between the 
 flanges or strengthening-hoops of cylindrical furnace-flues and adjoining parts. 
 
 The tube-holes must be bored of such a size that the tubes fit them exactly. The 
 tube-ends must not be expanded excessively, and must show no cracks after being 
 expanded. The tube-ends should project at least inch beyond the tube-plates. 
 
 In the boilers built by contract for a United States naval vessel several tubes had 
 been cut too short. To hide this defect short pieces of tube, turned down to a thin 
 edge at one end, had been inserted in the back end of these tubes ; the tubes and fer- 
 rules had then been expanded together and the projecting end of the ferrules beaded 
 over. This work was so neatly done that the piecing of the tubes could not be detected 
 by the eye, and the tubes showed no leaks under the cold-water pressure, but the 
 continued leaking of the tubes under steam led to the discovery of their dangerous 
 condition. 
 
 In the same boilers the pin-holes in the T-ends of several braces did not come fair 
 with the holes in the angle-irons riveted to the shell of the boilers. Smaller bolts had 
 been used to connect the braces to the angle-irons, and several bolts had been omitted 
 entirely. 
 
 The bolts or pins of braces must fit the holes exactly, and must be secured by nuts 
 or cotters. The T or angle irons to which the braces are attached must be securely 
 riveted to the shell. Their rivet-holes must show no cracks, and their bolt-holes must 
 not come too close to the edge of the flange. The long braces must be of equal tension ; 
 they must be straight, not bent to clear anything. Examine especially the stays run- 
 ning across the boiler between horizontal tubes to see that there is no danger of their 
 bearing against the tubes. The holes of stay-bolts must come exactly opposite each 
 other in both plates. 
 
 The explosion of the boiler of the TL S. S. Chenango, in 1864, was due to the omis- 
 sion of several braces. 
 
 See that no pieces of wood or iron have been left inside the boiler under and between 
 the furnaces ; that the valves open and close freely ; that no pipes are closed by 
 blank flanges ; that the gauge-pipes are not closed by putty or rubber packing. 
 
SEC. 3. TESTS, INSPECTIONS, AND TRIALS OP STEAM BOILERS. 369 
 
 After the hydraulic test the boiler should always be examined inside to see whether 
 the braces or their fastenings show any signs of having been unduly strained. 
 
 The periodical examination of boilers which have been in use is directed to the dis- 
 covery of leaky tubes, seams, rivets, and stay-bolts ; of cracks and blisters, and of the 
 distortion of plates by overheating ; of the accumulation of scale between the tubes and 
 on the furnace-crowns, and of loose scale and dirt in the water-bottoms. The extent of 
 coiTosion of rivet-heads, of braces and their fastenings, and of plates must be carefully 
 investigated. 
 
 The hammer -test is much relied on in examining old boilers. The plates are tapped 
 lightly with a hand-hammer, and from the sound given out and the rebound of the ham- 
 mer conclusions are drawn as to the thickness and soundness of the plates. In making 
 this test the influence of the more or less close proximity of stays, angle-irons, or gusset- 
 plates on the vibrations and springiness of plates must be taken into consideration. 
 
 When the thickness of plates appears doubtful a small hole is drilled through 
 them. 
 
 3. Trials of Boilers. Experiments on the evaporative power of boilers are, in 
 general, of two kinds, being designed to determine either the greatest weight of water 
 which the boilers are capable of evaporating in a unit of time, or the weight of fuel re- 
 quired for the evaporation of a given weight of water. 
 
 Numerous experiments, made under the direction of the Bureau of Steam-engineer- 
 ing of the United States Navy Department, to determine the relative evaporative effi- 
 ciency of different types of boilers ; the influence of changes in the proportions of grate- 
 surface, heating-surface, and calorimeter, and in the rate of combustion, on the evapora- 
 tive efficiency of boilers ; and the value of different kinds of fuel for marine boilers, have 
 been described by Isherwood in ' Experimental Kesearches ' and in the various reports 
 submitted by the boards of United States naval engineers charged with the conduct of 
 these experiments. 
 
 In all such experiments the quantity of water fed into the boiler, the weight of fuel 
 actually burnt and the weight of refuse matter of the fuel remaining unconsumed, the 
 pressures of steam and of the outside air, and the temperatures of the feed- water, steam, 
 external air, and chimney -gases should be carefully measured with accurately -tested 
 instruments, and the observations noted at regular intervals. The firing must be done 
 by experienced men and in a uniform manner. The fuel must be of a known and uni- 
 form quality. All conditions affecting combustion and evaporation in any manner must 
 be carefully recorded. Foaming must be guarded against. To prevent leakage of steam 
 or water the boiler must be tested under steam and water pressures before the experi- 
 
370 STEAM BOILERS. CHAP. XVI. 
 
 ment commences, and the boiler itself and the joints of all its pipes and valves must be 
 made perfectly tight. 
 
 Sometimes the steam generated in a boiler experiment is utilized in working an en- 
 gine, and the weight of water evaporated is deduced from indicator diagrams taken at 
 intervals during the trial. This method gives, however, no reliable results, since the 
 loss of steam by condensation and leakage in the cylinders, valves, and pipes varies 
 greatly under different conditions of the mechanism and with the manner of work- 
 ing the engines. 
 
 Errors resulting from leakage, foaming, and radiation will generally be diminished 
 by evaporating the water under atmospheric pressure. 
 
 The longer the time during which the experiment is continued, the less is the final 
 result affected by accidental disturbing elements and by inaccuracies of measurement 
 and errors of observation. 
 
 Each boiler experiment conducted under the direction of the Bureau of Steam- 
 engineering lasts, if possible, from 24 to 72 consecutive hours. The shorter the duration 
 of the experiment, the shorter should be the intervals of time between the recorded 
 observations. 
 
 For the sake of comparison the resiilts of experiments on the evaporative efficiency 
 of boilers are to be reduced to a uniform standard. For this piirpose it is convenient to 
 calculate the weight of water of a fixed temperature (either 100 or 212 Fahr.) which 
 would be evaporated under a fixed barometrical pressure of the atmosphere, provided 
 the same number of units of heat were communicated to the water under these condi- 
 tions as were transmitted to the water in the boiler per pound of fuel, or of combus- 
 tible matter of the fuel, consumed per hour. 
 
 The following extract from the 'Report on the Murphy Grate-bar,' by a board of 
 United States naval engineers, June 25, 1878, gives a description of the usual method 
 pursued in making the numerous boiler experiments which have been carried on under 
 the direction of the Bureau of Steam-engineering : 
 
 " Before commencing the experiments the blow-off pipe was removed and a plate 
 bolted across the aperture. This pipe was the only means through which water could 
 escape from the boiler. A temporary steam-escape pipe of 1\ inches inside diameter 
 was bolted on the top of the steam-drum, giving a straight discharge to the steam. The 
 safety-valve, of 5 inches diameter, was removed from its chamber, and the permanent 
 escape-pipe attached to it was used in addition to the temporary escape-pipe. . . . 
 
 " The feed- water, previous to entering the boiler, was accurately measured in two 
 covered tanks placed on the hurricane or upper deck of the vessel. One of these tanks 
 
SBC. 3. TESTS, INSPECTIONS, AND TRIALS OF STEAM B01LEES. 371 
 
 discharged, by measurement, 54.67535 cubic feet of water, and the other discharged 
 47.59028 cubic feet of water at each delivery. The two tanks were connected by a pipe 
 at their bottom, and in this pipe were two stop-cocks, one at each tank. From the 
 centre of the pipe connecting the tanks another pipe was carried vertically downward 
 15 feet to the check- valve near the bottom of the boiler, so that the feeding of the boiler 
 was effected by gravity, the quantity of water entering being regulated by the stop- 
 cocks in the connecting-pipe. The tanks were supplied with lake-water by a small 
 steam-pump worked with steam from a donkey-boiler. 
 
 "All the coal consumed was carefully weighed on a tested platform-scales in quan- 
 tities of 182 pounds at each weighing. The refuse from this coal in ash, clinker, soot, 
 etc., was similarly weighed and in the dry state. 
 
 "At the end of each experiment the furnaces, smoke-connections, flues, and tubes 
 were swept clean of soot and ash, which were weighed, and their weight added to that 
 of the refuse withdrawn from the furnaces and ashpits. 
 
 " In commencing an experiment the water in the boiler was brought to the boiling- 
 point under the atmospheric pressure by wood alone, which was then allowed to burn 
 down to the embers required for igniting the coal. No account was taken of the weight 
 of wood thus consumed. The water-level in the boiler was now adjusted to the proper 
 height in the glass water-gauge, the time noted, the coal fired, and the experiment held 
 to commence. Each experiment was ended with the water in the boiler at the same 
 level as at the commencement, and with the fires entirely burned out. It was intended 
 that each experiment should last twenty -four consecutive hours, and from this duration 
 none varied more than a few minutes. . . . 
 
 " One machinist was stationed at the tanks to note the time each was discharged and 
 to report it to the engineer of the watch. Another was stationed in the fire-room to see 
 the firing properly performed. A third was stationed at the scales for weighing the coal 
 and its refuse. The watches were four in number, of six hours' duration each, and were 
 superintended by the members of the board and the two engineer officers of the Michi- 
 gan, who personally weighed the coal and its refuse, and kept a log, or tabular record, 
 in the columns of which were entered hourly the kind of breeze blowing, the height of 
 the barometer, the steam-pressure in the boiler, the pounds of coal thrown into the fur- 
 naces, the pounds of refuse in ash, clinker, etc., withdrawn from the furnaces and ash- 
 pits, the temperature of the air on deck and in the fire-room, the temperature of the 
 feed-water in the tanks, and the temperature of the gases of combustion in the chimney. 
 This last temperature was obtained by means of a metallic pyrometer placed perma- 
 nently in the base of the chimney, with its index outside." 
 
372 STEAM BOILERS. CHAP. XVI. 
 
 The following description of an approximate method for determining the tempera- 
 tures of the gases in the uptake of a boiler is taken from the ' Report on the Ashcroft 
 Furnace-doors and Grate-bars,' by a board of United States naval engineers, March 27, 
 1878: 
 
 " The best approximation to the temperature of the gases of combustion in the boiler- 
 uptakes that could be made was to place on little wire tripods small fragments of tin, 
 lead, zinc, and antimony, and then insert these tripods into the mouths of the tubes at 
 their uptake ends ; the pieces of metal being at about the axes of the tubes and wholly 
 surrounded by the escaping hot gases of combustion, the tripods touching the tubes at 
 only three points. The melting-points of these metals may be taken approximately at 
 450, 650, 750, and 850 Fahr. ; and it is obvious that if one of them were found 
 melted, and the next not melted, the temperature of the gases of combustion passing 
 over them must have been somewhere between the respective melting-points. In this 
 manner two limits of temperature are found at about 200 Fahr. apart, as an extreme, the 
 mean perhaps not varying too widely from the truth for practical purposes. Three tri- 
 pods were placed in the top row of tubes, three in the middle, and three in the bottom row 
 of the tubes of each furnace one tripod in the next to the corner tubes of each of these 
 rows, and the third in the middle tube of the row ; but the mean of all the approximate 
 temperatures thus found cannot be assumed as the mean temperature of the whole mass 
 of escaping gas, because the velocity of this gas varies much through the different tubes, 
 being greatest through the top row, least through the bottom row, and intermediate 
 in the rows between. The melting-points of the metals, though they furnish only indi- 
 cations of the temperature, prove the enormous difference in the temperature of the 
 gases of combustion escaping from the top and bottom rows of tubes separated by a 
 vertical distance of but a few inches a difference sometimes as great as 300 Fahr." 
 
 To eliminate the errors due to inaccuracies in observing the steam-pressure, to super- 
 heating of the steam, and to the presence of Tinvaporized water in the steam escaping 
 from the boiler, the whole of this steam may be led to a surface-condenser, the weights 
 and temperatures of the water of condensation and of the condensing water before enter- 
 ing and after leaving the condenser measured, and, from these data, the units of heat 
 actually present in the steam may be calculated. This method was used in the trials 
 of steam boilers at the Fair of the American Institute in 1871. 
 
 In other instances the quality of the steam has been determined at regular intervals 
 during the trial by introducing a portion of the steam into a calorimeter, where it was 
 employed to heat a known quantity of water. This method was used in the boiler-tests 
 at the International Exhibition, Philadelphia, 1876. A tank containing a known 
 
Sic. 3. TESTS, INSPECTIONS, AND TRIALS OP STEAM BOILERS. 373 
 
 weight of water of a known temperature was set on scales. Into this a sufficient quan- 
 tity of steam was admitted to raise the temperature of the water a certain number of 
 degrees. From the differences of the weights and temperatures of the water in the tank 
 before and after the admission of steam the number of units of heat in, and the weight 
 of, the steam were found. 
 
CHAPTER XYII. 
 
 MANAGEMENT OF BOILERS. 
 
 1. Getting up Steam. When the order is given to get up steam commence 
 closing the boilers as soon as possible, so that, in case the joints of the manhole or 
 handhole plates should be found to leak after the water is run up in the boiler, the 
 latter can be emptied and the joints remade without delaying the starting of the fires. 
 Before closing the boiler satisfy yourself that all the braces are secured, and that no 
 articles used in repairing or cleaning the boiler have been left inside. 
 
 Rubber gaskets, manufactured in continuous rings of the size and shape required, 
 are used almost universally for making the joints of manhole and handhole plates. 
 When gaskets are cut out of sheet-rubber they are generally made in several pieces 
 with dovetailed ends, in order to economize material. The gasket must fit accurately 
 around the projecting rim of the plate and lie perfectly flat on the flange. When the 
 flange of the plate, and the ring around the manhole on which it seats, are smooth and 
 level the joint can be made tight without using white or red lead, which makes the rub- 
 ber hard and brittle. In removing the plate the rubber gasket is apt to stick partly to 
 the boiler and partly to the plate, and thus become injured. This may be prevented by 
 coating the gasket with black lead on the side in contact with the boiler. A mixture of 
 black lead and tallow is also used for this purpose and to soften the gasket, but the tal- 
 low rots the rubber. A coating of white lead is frequently put on the flange of the 
 plate, so that the gasket may stick to the plate in preference to sticking to the boiler. 
 When a manhole-plate is found to leak after steam has been raised it may often be 
 made tight by driving thin, flat wedges of soft pine wood between the projecting rim of 
 the plate and the edge of the manhole. The sides of these wedges should be slightly 
 bevelled, and adjoining wedges should overlap one another. 
 
 All the valves and cocks connected with the boiler should be examined before 
 getting up steam, to make sure that they can be operated freely. The steam stop- 
 valves are closed, but it is well to ease them off their seat slightly, else it may happen 
 that, in consequence of the uneqiial expansion of the valve-disc and chamber when 
 steam first begins to form, the valve be jammed in its seat. The safety-valves are 
 
 374 
 
SEC. 1. MANAGEMENT OP BOILERS. 375 
 
 raised and kept open till steam begins to form, to allow the escape of air from the 
 boiler. 
 
 The boilers are filled either by opening the bottom blow- valves and letting the water 
 run in from the sea, or by pumping water in through the check-valves. When the water 
 is taken from a tank or hydrant on shore it may be run in by means of a hose through a 
 manhole on top of the boiler or through the safety-valve. Boilers should be filled with 
 warm water when practicable. The height of the water within the boiler before it has 
 risen to the level of the water-gauges may be found, when the temperature of the 
 entering water is different from that of the boiler-shell, by applying the hand to the 
 shell and judging by the feeling. Or it may be found by tapping the shell with a ham- 
 mer and judging by the different sounds produced at places where the boiler is filled 
 and empty. To know whether the water is rising in the boiler when it enters through 
 the bottom blow-valve, open the water-gauge cocks and apply the hand or a lighted 
 lamp to the opening ; the rising water will expel the air through the opening. 
 
 Before charging the furnaces with fuel see that the grates, bridge-walls, and ashpits 
 are quite clear of ashes, clinker, etc., that the tubes are unobstructed, and that no arti- 
 cles are left in the front or back connections. Then close the uptake-doors. Remove 
 the hood from the chimney and open the damper. Hoist the chimney and secure it. 
 Leave the stays slack, and defer their adjustment till after the fires are well started and 
 the pipe has become hot ; but never set them up quite rigid. 
 
 In charging the furnaces the back of the grate is covered evenly with a thin layer of 
 small coal ; on the front of the grate some billets of split wood are placed side by side, 
 the ends of which are supported by a couple of pieces laid crosswise the furnace. A 
 few shovelfuls of coal are thrown on the wood, and some small kindling-wood, shav- 
 ings, oily rags, or other inflammable substances are placed at the furnace-mouth below 
 the layer of wood. 
 
 When the water has risen to the proper height in the boiler the kindling-wood in 
 the furnaces is lighted ; the furnace-doors are kept slightly open and the ashpit-doors 
 partly closed till the whole mass of wood has caught fire. More coal, broken up in 
 small pieces, is thrown on the burning mass, and the furnace-doors are closed and the 
 ashpit-doors opened. When the heap of coal on the front of the grate has become in- 
 candescent it is partly pushed back ; more coal is added, which is likewise pushed 
 gently back as soon as it becomes incandescent. This operation is repeated till there is 
 a sufficient mass of burning coal on the grate. 
 
 Sometimes, especially in very damp weather or when the air meets with many ob- 
 structions in flowing to the ashpits, it may be necessary to produce an artificial draught 
 
376 STEAM BOILERS. CHAP. XVII. 
 
 in the chimney in order to start the fires. By placing some burning shavings in the 
 uptake, and then closing the uptake-doors quickly, the column of air in the chimney is 
 heated and an ascending current is produced. 
 
 When it is necessary to raise steam with all possible despatch for a great emergency 
 the fires may be lighted while the water is still rising in the boiler, as soon as the heat- 
 ing-surfaces are barely covered with water ? and the water is then allowed to rise only to its 
 lowest admissible level in the gauge-glass. By using bituminous coal or greasy or tarry 
 matter in starting the fires the time required for getting up steam may be greatly short- 
 ened. By this means steam may be raised from cold water in large marine boilers in a 
 comparatively short time. The unequal expansion of the parts in contact with the fire 
 and the products of combustion, and of the boiler-shell, causes very injurious strains, 
 especially when the fires are urged from the beginning while the water is still cold, 
 which should, therefore, be avoided except in cases of great emergency. The fires 
 should be allowed to burn up slowly by being kept banked while the water is being 
 heated. Under ordinary circumstances not less than three hours should be allowed for 
 getting up steam. With very long cylindrical boilers, like the double-end boiler repre- 
 sented on Plate XV., it is advisable to allow even six hours for this purpose. 
 
 A great saving of wood may be effected, when there is no hurry in getting up steam, 
 by starting fires with wood at first only in the alternate furnaces of the boiler, and then 
 transferring some of the incandescent coal to the other furnaces, the grates of which 
 have been previously covered with a thin layer of coal. 
 
 As soon as a light column of steam rises from the escape-pipe or issues from the 
 open gauge-cocks the safety-valves are closed, the stop- valves opened, and the boilers 
 put in communication with each other. 
 
 2. Firing. The thickness of the bed of fuel on the grate must be regulated, accord- 
 ing to the kind, quality, and size of the fuel and to the force of the draught, in such a 
 manner as to ensure the passage of the proper quantity of air evenly distributed through 
 the grate. 
 
 With ordinary chimney-draught, giving a rate of combustion of from 12 to 16 Ibs. 
 of coal per square foot of grate per hour, a fire of anthracite coal of egg-size may be 
 carried from 5 to 6 inches thick. A thin fire, under otherwise equal conditions, offers 
 less resistance to the passage of the air through the grate, and is thus favorable to a 
 rapid rate of combustion. When the lumps of coal are of smaller size the fire may be 
 carried relatively thinner ; but when the fire is less than four inches thick a large grate 
 cannot be kept evenly covered, and too much air will pass through the grate. When 
 bituminous coal is used with a rapid rate of combustion the fire must be carried thicker 
 
SEC. 2. MANAGEMENT OP BOILERS. 377 
 
 than with anthracite coal, else the grate cannot be kept covered evenly ; this is espe- 
 cially the case with free-burning semi-bituminous coals. 
 
 With a forced draught the thickness of the fire is to be increased, and the size of the 
 lumps of anthracite coal should be diminished at the same time. 
 
 The furnaces should be fired regularly with moderate charges. When much coal is 
 thrown into the furnace it will take a long time to kindle ; but when the charges are 
 very small and oft repeated the frequent opening of the door causes a waste of heat. 
 Anthracite coal may be fired at intervals of from 15 to 20 minutes, according to the rate 
 of combustion. Bituminous coal, especially when it is of small size, should be fired 
 more frequently, every 10 or 15 minutes, because the evolution of the hydrocarbon 
 gases as each charge is thrown into the furnace makes a large quantity of heat latent, 
 and the temperature of the furnace would vary greatly if a large mass of coal was in- 
 troduced into the furnace at one time. Each charge of coal should be spread in an even 
 layer over the grate, and, as the bed of fuel burns away irregularly, it has to be 
 levelled. 
 
 When the coal cakes much the fire has to be broken up from time to time to afford 
 a passage to the air through the grate. The several furnaces of a boiler or set of boilers 
 should be fired and worked in rotation, so that, if possible, no two furnace-doors are 
 open at the same time. The coaling and working of the fires must be done as rapidly as 
 possible to limit the inflow of cold air in a mass through the open doors to a minimum. 
 
 When the fires are kept thin while the draught is active, or when the fires are not 
 kept properly levelled, the air rushes sometimes with great violence through the un- 
 covered parts of the grate, producing a roaring noise and severe concussions of the 
 boiler. This phenomenon is called back-draught. By partly opening the furnace- 
 doors or closing the ashpit-doors the draught is generally checked sufficiently to stop 
 the violent rush of air, but to remedy the evil the fire must be levelled or carried a little 
 thicker. 
 
 When the incandescent fuel throws a uniform bright light below the grate it indi- 
 cates that the fires are clean and burning actively. When the ashpits are dark, either 
 totally or in parts, it indicates an accumulation of ashes or clinker on the grates. To 
 clear the air-passages the hook-bar is run through the interstices of the grate from the 
 ashpit, and when much difficulty is experienced in moving the hook-bar back and forth 
 between the grate-bars it indicates the formation of clinker. Clinkers adhering to the 
 top of the grate-bars are detached by means of the slice-bar introduced through the 
 furnace-door, and are then removed from the furnace. Ashes and cinders are apt to 
 accumulate in the corners of the furnace ; such places must be cleaned out and the 
 
378 STEAM BOILERS. CHAP. XVII. 
 
 whole grate must be kept covered with live coal. When the coal is friable it should be 
 disturbed as little as possible with the slice-bar, to avoid the loss of small coal falling 
 unburnt through the grate. 
 
 When the accumulation of ashes, cinders, and clinker on the grate becomes so great 
 that they cannot be removed by pricking and slicing, the fire must be thoroughly 
 cleaned by hauling all refuse matter from the furnace, leaving a clean bed of incandes- 
 cent coal on the grate. This cleaning should be done at regular intervals, generally not 
 exceeding twelve hours, but depending on the amount and kind of the refuse matter 
 contained in the coal. That the cleaning may be done quickly and thoroughly the fire 
 should not be very heavy, but it should have a sufficiently thick layer of incandescent 
 coal on top to cover the grate completely after the refuse has been removed. The fire- 
 man pushes back the top layer of incandescent coal from the front half of the grate 
 and hauls the mass of refuse below it from the furnace, cleaning the grate entirely. 
 Then he hauls all the clean coal from the back of the furnace to the front of the grate ; 
 he works the mass of ashes and cinders at the back through between the grate-bars, 
 and hauls the larger of pieces of slate and clinker out of the furnace over the heap of 
 coal in front. The clean coal is then spread evenly over the grate and covered at once 
 with a thin layer of fresh fuel. 
 
 Some firemen, instead of cleaning first the front and then the back of the fire, clean 
 the two sides of the furnace in succession, using the slice-bar to move the top layer of 
 clean coal from the side to be cleaned to the other side. 
 
 The slack of anthracite coal can be burnt only when mixed with a certain proportion 
 of lump-coal. The lumps should never be larger than a cube of three inches a side. 
 
 The dust of bituminous coal may be burnt by mixing it with water so as to form a 
 cohering mass or thick paste. The evaporation of the water mixed with the coal ab- 
 sorbs much heat, but without it the dust would fall through the grate or be carried 
 into the flues by an active draught, proving thus a total loss. 
 
 The friable, free-burning semi-bituminous coals are frequently mixed in equal pro- 
 portions with caking coals, in order to bind the small particles of the former together 
 and prevent their falling unconsumed through the grate. 
 
 Many kinds of bituminous coal require special methods of firing, in order that their 
 gaseous products may be completely consumed and that the production of smoke may 
 be prevented. When a fresh charge of coal is thrown upon the fire the first effect pro- 
 duced, as it becomes heated, is the evolution of a mass of hydrocarbon gases, which 
 make latent a large quantity of heat and require a larger quantity of air for their com- 
 plete combustion than the remaining solid portion of the coal or coke. On this account 
 
SBC. 3. MANAGEMENT OP BOILERS. 379 
 
 the air-admission through the door into the furnace, or through the bridge-wall into the 
 combustion-chamber, may be regulated by registers, which are opened after a fresh 
 charge of coal has been thrown upon the fire, and closed when the evolution of hydro- 
 carbon gases ceases that is to say, when the coal burns without flame. 
 
 When the furnaces are wide side-firing may be employed, which consists in throw- 
 ing each new charge of coal alternately on either side of the fire, so that the evolution 
 of hydrocarbon gases takes place only on one half of the grate, while on the other half 
 the coked coal of the previous charge is burnt. 
 
 Some kinds of free-burning semi-bituminous coal are burnt to best advantage by 
 piling each fresh charge up on the dead-plate at the front of the grate, where the vola- 
 tile ingredients of the coal are expelled by the heat radiated from the incandescent fuel. 
 As these gases pass through the furnace and mix with a sufficient quantity of air, 
 which passes in excess through the grate, they ignite and are completely consumed. 
 As soon as the mass of coal on the dead- plate becomes converted into coke it is pushed 
 back and spread over the grate. 
 
 When wood is to be used as fuel in a furnace designed to burn coal the grate has to 
 be lowered to increase the capacity of the furnace, and the spaces between the bars 
 should be increased in width by omitting a number of grate-bars. 
 
 3. Management of Boilers under Steam. That the boilers may furnish a uni- 
 form supply of steam the fires must be supplied with fuel and cleaned in regular rota- 
 tion, as described in the preceding section, and the water in the boiler should be kept, 
 at nearly a uniform height. When the opening of the check and blow valves is pro- 
 perly adjusted the feeding and blowing may be kept up continuously. The water- 
 gauge glass is generally located in relation to the tubes and back-connections so that a 
 proper level of water is maintained when the glass is kept about half-full of water. 
 Previous to cleaning a fire the water may be allowed to rise a few inches above the usual 
 level, so that during the process of cleaning the fires, and till they have been brought 
 again to their normal state, the supply of feed- water may be diminished. 
 
 When bituminous coal is used as fuel the tubes must be swept at regular intervals, 
 about once in twenty-four hours or less frequently, according to the greater or less ten- 
 dency of the coal to form deposits of soot. 
 
 In order to increase the evaporation to a maximum diminish the quantity of water 
 blown out and increase the temperature of the feed- water as much as practicable ; secure 
 an ample air-supply by turning the fire-room ventilators and windsails to the wind ; 
 keep the ashpits wide open and clear of ashes, and the fires clean ; use coal of about 
 egg-size and free of slack, when the fuel is anthracite ; and regulate the thickness of the 
 
380 STEAM BOILERS. CHAP. XVII. 
 
 fires according to the force of the draught and the kind, quality, and size of the fuel. 
 The boilers of naval vessels are generally provided with a steam-jet in the chimney for 
 the purpose of increasing the draught. It should be borne in mind that with a given 
 boiler the rate of combustion cannot be increased advantageously beyond a certain 
 limit, on account of the decrease of the economic evaporative efficiency of the boiler ; 
 that the efficiency of the boiler may be greatly diminished by foaming with an increased 
 evaporation ; and that the available quantity of steam furnished by a boiler may be 
 actually diminished when the rate of combustion is increased by forcing the draught. 
 (See section 6, chapter xii.) 
 
 When only a fraction of the boiler-power is used all the openings of the uptake, 
 furnace and ashpit doors of such furnaces as are not in use must be closed tight to pre- 
 vent the inflow of cold air. Often it is advantageous to keep all the furnaces in use, 
 but to diminish the effective grate-surface by covering the back of the grate either with 
 a thick layer of ashes or by a wall of fire-brick built up to the height of the bridge-wall. 
 The latter plan should be adopted only in case the reduced boiler-power is to be 
 used for long periods during which no necessity will arise for using full boiler- 
 power. 
 
 When the boilers are not required to work up to their full power fuel may be econo- 
 mized by burning a greater proportion of slack or coal-dust, and by sifting the ashes 
 and burning a portion of them containing particles of unconsumed coal. 
 
 Banking a fire means to pull the coal together in a heap, leaving a part of the grate 
 uncovered. By this means the combustion is greatly retarded, and the cold air flowing 
 in through the uncovered part of the grate checks the evaporation. When the boilers 
 are not to be used for a certain length of time, but fires are to be kept in them to keep 
 the water hot, the fires should be banked and covered with a layer of ashes or slack, and 
 the ashpit-doors and the damper should be nearly closed. 
 
 When the steam-supply is to be temporarily diminished while the boiler is in opera- 
 tion some fires may be banked, and the draught may be checked by closing the ashpit- 
 doors and the damper and by opening the uptake and furnace doors. The opening of 
 the furnace-doors should be avoided, if possible, because the cold air, rushing into the 
 furnaces and striking the highly-heated plates, causes sudden and unequal variations of 
 temperature, which produce local strains resulting frequently in permanent injuries to 
 the boiler. 
 
 When the engines are stopped suddenly, but the boilers have to be kept ready 
 for starting again shortly, the formation of steam has to be checked as quickly as pos- 
 sible ; to this end open the furnace and uptake doors, close the ashpit-doors and dam- 
 
SKC. 3. MANAGEMENT OF BOILERS. 381 
 
 pers, uncover a p"art of the grate, increase the feed- water supply and the quantity of 
 water blown out ; utilize the steam for distilling or other useful purposes. When the 
 steam-pressure rises to the limit allowed the safety-valve has to be raised. United 
 States naval boilers are frequently provided with a bleeding-valve, by which an excess 
 of steam may be discharged from the boiler into the condenser, instead of being allowed 
 to escape through the safety-valve. 
 
 The safety-valve should always be opened gradually. When it is suddenly opened 
 wide the steam, rushing from it with great violence, is apt to carry a mass of water with 
 it, which falls in a shower from the top of the escape-pipe on the deck. When the 
 water-capacity of the boiler is small this sudden loss of water may even cause some 
 parts of the heating-surfaces of the boiler to become uncovered. 
 
 In the indications of the water-level in the boiler by the gauge-glass proper allow- 
 ance must be made for the oscillations and the list of the vessel, and the gauge, must be 
 tried from time to time to see that the narrow passages are not choked with sediment or 
 scale. When the water-passage is clear, but the steam-passage is closed, the gauge- 
 glass will remain completely filled with water. When, on the contrary, the steam-pas- 
 sage is clear and the water-passage is choked the water remains at a constant level in 
 the glass without oscillating. The gauge-glass must be blown through from time to 
 time by opening the waste-cock and shutting off the water and steam cocks alternately. 
 If this does not clear the passages it is necessary to run a wire through them. 
 
 The indications of the water-gauges are frequently very uncertain and deceptive 
 when the boiler foams ; the water in the gauge-glasses rises and falls rapidly and in an 
 irregular manner, and on opening the gauge-cocks a mixture of steam and water issues, 
 producing a sputtering sound. 
 
 All boilers will foam to some extent when the rate of combustion exceeds a certain 
 limit. But boilers with insufficient or low steam-room, contracted water-surface, and 
 defective circulation are especially liable to foaming. In boilers with narrow water- 
 spaces and a high rate of combustion the water is frequently lifted in a mass, so that 
 the water-gauges indicate a steady level of solid water while the engines are in opera- 
 tion ; but as soon as the supply of steam drawn from the boiler and the rate of evapora- 
 tion are diminished the water falls suddenly to its true level, disappearing sometimes 
 entirely from the gauges. In all such cases it is necessary to check the evaporation in 
 order to stop the foaming ; and it is frequently necessary to slow down the engines or 
 open the furnace-doors from time to time for the purpose of finding the true level of 
 the water in the boiler. 
 
 Vertical fire-tube boilers and horizontal cylindrical boilers of small diameter foam 
 
382 STEAM BOILERS. CHAP. XVII. 
 
 frequently because there is too much water in them, in consequence of which their 
 steam-room and the area of the water-level are simultaneously reduced. 
 
 Another cause of foaming is the presence of mud or dirt of a mucilaginous nature in 
 the water, which may be recognized by the appearance of the water in the gauge-glass. 
 Also, when a vessel enters a river in coming from the sea, or vice versa, the boilers are 
 liable to foam when they are fed directly with water from overboard. In all such cases 
 it is advisable to change the water in the boilers as rapidly as possible by opening the 
 surface blow- valves wide and feeding strongly. 
 
 In reducing the saturation of boilers the surface blow- valves should be used in pre- 
 ference to the bottom blow- valves, unless the vessel rolls so much that the former would 
 frequently discharge steam instead of water. 
 
 When the water falls below its usual level in the boiler examine the feed and check 
 valves to see whether they are open and in operation. The latter will be indicated by 
 the clicking noise made by the check-valve as it rises and falls in its seat ; also by the 
 temperature of the pipe immediately below the valve, which should be comparatively 
 cool to the touch. Close the blow-valves, and if the water continues to fall in the 
 boiler it must be owing to a leak in the boiler. If no water enters through the check- 
 valve, although the feed-pump is throwing water, the latter may escape through a leak 
 in the feed-pipe, or all the water may enter some of the other boilers, or the relief- valve 
 of the pump may be gagged or insufficiently loaded. If the pump does not throw any 
 water it may be owing to air-leaks in the stuffing-box or in the suction-pipe, or because 
 the valves or the piston are leaking, or because the feed-water is too hot. The feed- 
 pump may get hot because the check- valves are leaking. When the valve is kept open 
 by being jammed a slight jar produced by tapping the valve-chamber with a hammer is 
 sometimes sufficient to seat it. When some matter has lodged under the valve and 
 prevents its seating it may be washed away by a strong feed with the donkey-pump. 
 
 In case the water should suddenly disappear from the gauge-glass on stopping the 
 engines, the safety-valve may be opened wide in order to cause the boiler to foam again 
 and the water to be lifted sufficiently to cover the heating-surfaces ; at the same time 
 close the blow- valves and check the evaporation by opening the furnace and connection 
 doors and by covering the fires. Do not put on the feed unless you are sure that the 
 water has not fallen low enough to cause the plates to become overheated. 
 
 In case any part of the boiler should be discovered to have become red-hot in conse- 
 quence of low water, or the furnace-crowns to be collapsing, do not open the safety- 
 valve or change the working of the engines, and especially do not open the feed- 
 valve, but haul the fires from the furnaces at once or cover them with wet ashes, and 
 
SEC. 3. MANAGEMENT OP BOILEE& 383 
 
 then successively close the stop- valve, blow the water from the boiler through the bot- 
 tom blow-valve, and open the safety-valve. 
 
 Leaks in a boiler under steam become manifest by a hissing sound and by the ap- 
 pearance of the issuing steam or water. Serious leaks in the bottom of a boiler may be- 
 come known only by the falling of the water-level in the boiler, and by an increase in 
 the quantity and in the temperature of the water in the bilge. All leaks should be 
 closely watched, and, in case they are found to increase or cause serious inconvenience, 
 the pressure in the boiler should be diminished, if the leaks cannot be stopped other- 
 wise. Small leaks of water frequently stop of themselves by the gradual accumulation 
 of salt deposited by the issuing water. 
 
 Leaks in the joints of manhole and handhole plates, and leaks caused by the blow- 
 ing-out of a bolt or rivet in the boiler, or by small holes in the feed and blow pipes, 
 may be temporarily stopped by driving into the holes causing the leaks slightly taper- 
 ing plugs or wedges cut from dry, soft pine wood. 
 
 When a feed or blow pipe is split or cracked it must be tightly wrapped with stout 
 cotton canvas painted with red lead on the side nearest the pipe, and closely bound 
 with marline. Cut the canvas in long strips about 3 inches wide, and let each turn 
 overlap the preceding one half its width. Wind the marline around it in such a direc- 
 tion as to tighten the canvas wrapping, and lay each turn close to the other, pulling 
 hard to prevent its stretching or shifting afterwards. 
 
 Leaks in the boiler may sometimes be stopped temporarily by covering the defective 
 place with a piece of plate-iron fitting closely to the surface and held firmly in position 
 by means of wedges driven against the bottom or side of the vessel, or against the op- 
 posite wall of the boiler or of an adjoining boiler. To make the joint of the patch 
 tight use either canvas painted with red lead, or putty made of white and red lead and 
 stiffened by mixing it either with fine iron borings or with hemp chopped very fine. 
 
 When leaks appear in the. furnace-crowns, the water issuing either from cracks in 
 the plates or from the seams, it is best to haul the fires from the respective furnaces 
 when the leaks are found to be increasing with continued use. 
 
 The leaks of tubes may be caused either by the defective joints of their ends or by 
 holes or rents in the tubes. In the latter case the leak may be stopped by plugging the 
 defective tube with a turned soft pine plug of very slight taper. These plugs are about 
 5 or 6 inches long, and are wrapped with canvas painted with white or red lead. For 
 the front end of the tubes the diameter of the larger end of the plug is slightly greater 
 than the internal diameter of the tubes, and the small end is swelled out by the water 
 which penetrates the pores of the wood, and the plug is thus held tightly in the tube. 
 
384: STEAM BOILERS. CHAP XVIL 
 
 When the leak is not very large it is sufficient to plug the uptake end of return-tube 
 boilers, so that the water does not run out through the uptake-doors and interfere with 
 the working of the fires, but runs into the back-connections. A small leak may some- 
 times be stopped completely by securing a plug within the tube over the defective 
 place. When the leak is so large that the water issuing from it would greatly impair 
 the efficiency of the fire and cause a large accumulation of salt, which is not only diffi- 
 cult to remove, but, by closing a number of tubes, may seriously impede the draught of 
 the furnace, both ends of the leaky tube have to be plugged. The back-connection end 
 of the horizontal fire-tubes may be plugged without hauling the fires, by introducing 
 from the front end of the boiler a plug with a long wedge inserted in a cross-cut at its 
 outboard end, which butts against the back wall of the back-connection when the plug 
 is in place, so that when the latter is driven home the wedge forces it tightly against 
 the tube. Sometimes the plug itself is made long enough to reach to the back wall of 
 the back-connection when it is home, and the wedge is inserted and driven in the end 
 of the plug projecting within the tube. The part of the plug or wedge which projects 
 within the back-connection soon burns away, but within the tube the plug is protected 
 by the water which penetrates its pores. After the plug has been driven in the back 
 end another phig is driven in the front end, as described above. It is, however, neces- 
 sary to lower the steam-pressure greatly before commencing this operation. 
 
 Sometimes it may be necessary to enter the back-connection in order to plug a tube 
 effectually. In such a case haul the fire from the furnace which is to be entered, and 
 bank and cover up with ashes the fires in the other furnaces of the boiler. Open the 
 surface-blow and pump water into the boiler from the sea, so as to reduce the tempera- 
 ture of the water in the boiler. Then open the safety-valve and give to the vessel a list 
 to let the water run out of the front end of the leaky tube. After covering the grate, 
 bridge- wall, and back-connection of the empty furnace with boards, bags, or old canvas, 
 a man can safely enter the back-connection. Then the tube may be stopped up with a 
 wooden plug in the same manner as described above for the front end of the tubes. Or 
 both ends of the tube are covered with cup-shaped washers filled with stiff putty, and, to 
 hold them in place, a stout iron rod is put through the tube, so that its ends, on which 
 screw-threads are cut, project through the washers ; the latter are then screwed up 
 tight with nuts. 
 
 When a boiler is to be put out of use it is best to let it cool off gradually. To this 
 end, after the fires are hauled from the furnaces, all the doors should be kept closed and 
 the steam should be allowed to condense gradually in the boiler. When the pressure 
 has fallen nearly to that of the atmosphere the safety-valve is to be raised and kept 
 
SBC. 4. MANAGEMENT OP BOILEBS. 385 
 
 open. When the temperature of the water in the boiler has decreased sufficiently the 
 water may be pumped out of the boiler when the latter is provided with a valve and 
 pipe-connection for this purpose, or the water may be allowed to run out into the bilge 
 through one of the mudholes in the bottom of the boiler. 
 
 When the boiler has to be cooled off quickly for the purpose of cleaning and repair- 
 ing it in an emergency, the water is blown out through the bottom blow- valve as soon as 
 the fires are hauled. Previously the steam should be raised to the highest working 
 pressure, so that the water may be blown out as completely as possible. A peculiar 
 crackling noise in the blow-pipe, produced by the condensation of the steam as it comes 
 in contact with the cold water, indicates that the boiler is empty. When the blow- 
 valve is kept open after the steam-pressure has fallen so low that it no longer balances 
 a column of water equal in height to the difference between the levels of the water in 
 the boiler and of the sea- water, the latter will enter the boiler. This would be indicated 
 by the sudden cooling of the blow-pipe. The height to which the water in the boiler 
 has fallen may be found by sounding the boiler with a hammer. As soon as the blow- 
 valve is closed the safety-valve is raised, and all the furnace and connection doors are 
 opened and the manhole and mudhole plates are taken off to dry and cool the boiler. 
 In case of urgency the boiler may be cooled quickly by filling it with cold sea- water 
 after blowing the water out, afterwards running the sea- water out into the bilge. 
 
 Blowing a boiler down jars it severely, especially when the blow-valve is opened 
 wide ; and this is often the cause of leaks in the boiler and m the pipes connected with 
 it. It should never be done except in cases of great emergency. 
 
 4. Foaming : its Causes, Effects, and Prevention. Foaming or priming 
 means that the water in the boiler is in a state of violent agitation, rising and falling 
 rapidly in the form of waves, or that the steam is mixed with water in the form of 
 spray. Foaming is a source of great inconvenience, and not unfrequently of danger, on 
 account of the uncertain and wrong indications of the water-level given by the gauges ; 
 and, as the water is carried with the steam into the cylinders, it causes a serious loss of 
 efficiency and may cause a breaking-down of the engines. 
 
 Foaming is made evident by the boiling-up or the rapid and irregular oscillations of 
 the water in the gauge-glass, and by the sputtering sound produced as the mixture of 
 steam and water issues from the gauge-cocks. When the water is carried over into the 
 cylinders its presence is made known by a clicking noise caused by the partial collapse 
 of the piston-rings, and, when the water is present in large quantities, by the thumping 
 of the piston at each end of the stroke. 
 
 All boilers are apt to foam when the water contains much mud or dirt of a mucil- 
 
386 STEAM BOILERS. CHAP. XVII. 
 
 aginous nature. Soda, introduced into the boiler to neutralize the fatty acids contained 
 in the feed-water, often produces foaming. The various organic substances introduced 
 into boilers to prevent the formation of scale are apt to produce the same effect. (See 
 section 11, cJiapter xviii.) The engines of the English naval vessel Hecate were broken 
 down by excessive foaming caused by the lime placed in her boilers to preserve them 
 and not removed before getting up steam. When a vessel coming from the sea enters 
 fresh water, or from a river enters the sea, the boilers foam frequently. In all such 
 cases it is advisable to change the water in the boiler as rapidly as possible by opening 
 the surface blow-valves wide and putting on a strong feed. 
 
 The plan of stopping foaming by covering the surface of the water in the boiler with 
 a layer of oil or molten tallow injected through the feed-pumps is not to be recommended. 
 It is not only an expensive remedy, but the decomposition of the animal or vegetable fats 
 at high temperatures, and in contact with metals, produces fatty acids which are very de- 
 structive to boilers. 
 
 Boilers are liable to foam when they have an insufficient and low steam-room, a con- 
 tracted water-surface, and such an arrangement of the internal parts as to render the 
 circulation of the water defective. It may be assumed that any boiler will foam more 
 or less when its evaporation exceeds a- certain limit, so that the steam-bubbles rise so 
 rapidly as to carry some of the water through which they pass along with them. For 
 this reason some water-tube boilers are provided with deflecting-plates at the upper ends 
 of the tubes, without which the water would be thrown in jets from the tubes into the 
 steam-space. (See figure 2, Plate XXVIII., and section 10, chapter xi.) 
 
 When the steam, as it is generated, has to escape in large masses through very nar- 
 row water-passages separate channels must be provided for the descending water-cur- 
 rents, else the meeting of the two currents moving in opposite directions is very apt to 
 result in foaming, or sometimes in lifting the water. The latter expression means that 
 the steam does not rise, as it is generated, through the overlying mass of water, but ac- 
 cumulates on the heating-surfaces, so that water appears at a greater height in the 
 boiler than would be the case if the steam and water occupied their natural positions. 
 Under these circumstances the heating-surfaces which are kept bare of water are liable 
 to become overheated, and the water-gauges give wrong indications of the quantity of 
 water in the boiler. Whenever the evaporation is checked the water falls to its true 
 level. Thus it frequently happens, in small boilers worked with a high rate of combus- 
 tion, that the water disappears suddenly from the gauges which had indicated a moment 
 before an ample supply of water in the boilers. The overheating of the metallic sur- 
 faces may cause the water, as it comes in contact with them, to assume the spheroidal 
 
SEC. 4. MANAGEMENT OP BOILERS. 387 
 
 state, and its evaporation to take place in an intermittent, explosive manner, thereby 
 producing rapid oscillations of the water-level or projections of water into the steam- 
 space. 
 
 In hanging water-tubes the separation of the ascending and descending currents is 
 effected by an inner tube. (See section 10, chapter xi.) 
 
 In the vertical fire-tube boiler represented in figure 1, Plate XXVIII., the internal 
 annular tank, which serves as a steam-reservoir, separates also the ascending and de- 
 scending currents; the steam rises from the crown of the furnace between the tubes, and 
 the water flows downward in the annular space between the tank and the shell of the 
 boiler. 
 
 In the ordinary marine boiler with the tubes arranged over the furnaces the great 
 mass of steam generated on the furnace-crowns, in addition to the steam formed on the 
 tube-surfaces, has to rise through the narrow spaces between the tubes ; while in the 
 locomotive type of boiler the steam escapes from the furnace-crowns and from the tube- 
 surfaces directly to the steam-room. This difference in the arrangement of the heating- 
 surfaces is the principal reason why the rate of combustion in marine boilers cannot 
 approach remotely the rate of combustion common in locomotive boilers without pro- 
 ducing violent foaming. To lessen this evil the circulation of the water in marine 
 boilers should be facilitated by leaving wide, unobstructed passages between the nests 
 of tubes of adjoining furnaces. Good results have been obtained in English boilers by 
 separating these spaces, by means of removable plates, from the nests of tubes so that 
 they may serve exclusively as passages for the descending currents of water flowing to 
 the furnace-crowns. In the boiler designed for the U. S. S. Polos similar plates, reach- 
 ing from the upper row of tubes to the furnace-crowns, were attached by means of 
 socket-bolts to the cylindrical shell, leaving a space several inches wide between the 
 shell and the plate. The introduction of circulating-tubes in the back-connections has 
 also had the effect of lessening foaming in some cases. 
 
 In horizontal cylindrical boilers the water-surface and the steam-room both diminish 
 rapidly as the height of the water-level is increased, and this circumstance renders this 
 form of boiler especially liable to foaming. This effect is frequently produced when the 
 water is allowed by accident to rise beyond a certain height in the boiler. In some cases 
 it has been found advantageous to cut out the upper row of tubes and plug up the holes 
 in the tube-sheets, so as to be able to carry the water lower in the boiler, the loss in 
 evaporative power by the diminution of the heating-surface and calorimeter of the tubes 
 being far less than the gain in efficiency due to lessened foaming. This plan was adopted 
 in the boiler of the U. S. S. Triana, and, to make it possible to carry the water below the 
 
388 STEAM BOILERS. 
 
 CHAP. XVII. 
 
 top of the back-connection, a false top was built in the connection, and the space thus 
 formed filled with plaster-of -Paris. 
 
 Whenever the steam-pressure is suddenly diminished by withdrawing a large quan- 
 tity of steam from the boiler, the temperature of the water will be so much higher than 
 the boiling-point corresponding to the reduced steam-pressure that a sudden evolution 
 of steam, accompanied by violent ebullition, takes place. This effect is produced when 
 the steam-room in the boiler is too small, or when the engines are suddenly started at a 
 high speed or are racing, or when the safety-valve is all at once thrown wide open. In 
 such cases the steam rushing violently to the opening may carry along with it a large 
 mass of water, which rises as a wave over the surface of the water, and, when the steam- 
 room is low, may be carried into the steam-pipe and flood the cylinders, or be projected 
 from the escape-pipe. It is of little use to surround the opening of the stop-valves with 
 deflecting-plates which are to throw off the mass of water carried up by the steam-cur- 
 rent. The evil may frequently be corrected by drawing the steam equally from a large 
 area by the use of perforated dry-pipes ; or, if these cannot be applied, a second 
 stop- valve may be placed on the boiler which takes steam at some distance from the 
 original stop-valve. This plan is far more effective than merely enlarging the 
 original stop-valve in order to diminish the velocity of the steam passing through its 
 orifice. 
 
 In many cases it will be found necessary to increase the steam-room by the addition 
 of a steam-drum. High vertical steam-drums have the advantage that the water held 
 suspended in the steam is separated to a great extent from it before it can reach the top 
 of the drum, from where the stop-valve should take the steam. (See also section 1, 
 chapter xiii.) 
 
 Superheaters are efficient in correcting some of the evils of foaming by increasing the 
 steam-room and by evaporating the water carried along with the steam. (See section 10, 
 chapter iii.) It is, however, found that foaming causes frequently the rapid destruction 
 of superheaters. (See section 3, chapter xiii.) 
 
 5. Constituents of Saline Matter in Sea-water. The proportion of saline 
 matter contained in the waters of different seas varies greatly. According to Dr. Mar- 
 cet, the ocean of the southern hemisphere contains more salt than that of the northern 
 hemisphere, the mean specific gravity of sea- water near the equator being 1.0277, or 
 intermediate between that of the waters of the northern and southern oceans. 
 
 Dr. Ure gives the following proportions of saline matter in 1,000 parts (by weight) of 
 sea- water from different localities : " The largest proportion of salt held in solution in 
 the open sea is 38, and the smallest 32. The Red Sea contains 43 ; the Mediterranean, 
 
SEC. 5. MANAGEMENT OP B01LEBS. 389 
 
 38 ; the British Channel, 35.5 ; the Arctic Ocean, 28.5 ; the Black Sea, about 21 ; and 
 the Baltic, only 6.6." 
 
 An analysis of the water of the English Channel at Brighton, made by Dr. Schweit- 
 zer, gave the following results viz : its specific gravity was 1.0274, and 1,000 parts, by 
 weight, contained 
 
 Water 964.74372 
 
 Chloride of sodium 27.05948 
 
 Chloride of magnesium 3.66658 
 
 Chloride of potassium 0.76552 
 
 Bromide of magnesium 0.02929 
 
 Sulphate of magnesia 2.29578 
 
 Sulphate of lime. . .-. 1.40662 
 
 Carbonate of lime. . 0.03301 
 
 Total weight 1000.00000 
 
 An analysis of the water of the Mediterranean, made by Dr. Laurens, gave the fol- 
 lowing results viz. : its specific gravity was 1.0293, and 1,000 parts, by weight, con- 
 tained 
 
 Water 959.06 
 
 Chloride of sodium 27.22 
 
 Chloride of magnesium 6.14 
 
 Sulphate of magnesia 7.02 
 
 Sulphate of lime 0.15 
 
 Carbonate of lime 0.09 
 
 Carbonate of magnesia 0.11 
 
 Carbonic acid 0.20 
 
 Potash.. 0.01 
 
 Total weight 1000.00 
 
 The amount of carbonate of lime contained in sea- water is insignificant. According 
 to Bucholz, 100 parts of cold water are capable of holding in solution from .00416 to 
 .00625 parts of carbonate of lime. The French chemist Couste states that water loses 
 entirely its power of holding this salt in solution when its temperature reaches a point 
 lying between 285 and 300 Fahr. After being once precipitated the carbonate of lime 
 is not redissolved when the temperature of the water is lowered. 
 
390 
 
 STEAM BOILERS. 
 
 CHAP. XVIL 
 
 When water contains an excess of carbonic acid the carbonate of lime is converted 
 into an acid carbonate of lime, which is much more soluble in water. On the applica- 
 tion of heat the acid carbonate loses a portion of its carbonic acid, and the neutral 
 carbonate of lime is deposited in the form of a powder. 
 
 The solubility of sulphate of lime is, according to Regnault, a maximum at 95 
 Fahr., when 100 parts of water dissolve 0.254 parts of this salt. At 212 Fahr. 100 
 parts of water dissolve only 0.217 parts of this salt ; and, according to Couste, water 
 loses completely its power of holding in solution the sulphate of lime when its tempe- 
 rature reaches a point lying between 285 and 300 Fahr. This salt is more soluble in 
 dilute solutions of chloride of sodium, and it is insoluble in a saturated solution of the 
 same salt. The precipitated sulphate of lime is redissolved when the water cools down ; 
 but this process is the slower the higher the temperature at which it was precipitated. 
 When the deposit is formed at 300 Fahr. it takes several days before it is redissolved 
 by the water, even if the quantity is small relatively to the water. 
 
 The chloride of calcium, formed by a reaction between the carbonate of lime and 
 the chloride of magnesium, is soluble in water, and undergoes no decomposition in the 
 presence of water at the temperatures obtaining in steam boilers. 
 
 The chloride of sodium, or common salt, which forms by far the greatest proportion 
 of the saline matter of sea- water, undergoes no decomposition by heat. Its solubility in 
 water is nearly the same at all temperatures : 
 
 100 parts of water at 57 Fahr. dissolve 36 parts of this salt. 
 " " " 140 Fahr. " 37 " " 
 
 " " " 212 Fahr. " 40 " " 
 
 The chloride of magnesium is very soluble in water. At 60 Fahr. 100 parts of 
 water dissolve 200 parts of this salt. It is decomposed at 212 Fahr., forming hydro- 
 chloric acid and magnesia ; the latter substance is deposited in the form of a white 
 powder, while the former enters into combinations with the iron of the boiler and with 
 the lime. (See sections 7 and 11, chapter xviii.) 
 
 Sulphate of magnesia. 100 parts of water at 207 Fahr. dissolve 644 parts of the 
 crystallized salt ; 100 parts of water at 58 Fahr. dissolve 104 parts. 
 6. Composition of Boiler-scale. 
 
 Station. 
 
 Fracture. 
 
 Sulphate of lime. 
 
 Carbonate of magn. 
 
 Magnesia. 
 
 Water. 
 
 Hamburg 
 
 Partly crystallized 
 
 8s 20 
 
 ? ?C 
 
 5nc 
 
 6 e 
 
 Mediterranean . 
 
 Amorphous 
 
 8/1. QA 
 
 Z -- J 5 
 
 2. 1A. 
 
 V3 
 7.66 
 
 *5 
 
 A 6c 
 
 
 
 
 ^O^r 
 
 
 I 
 
SKC. 7. 
 
 MANAGEMENT OF BOILERS. 
 
 391 
 
 The water was present in mechanical combination. (Engineering, 1866.) 
 In the ' Third Report of the Admiralty Committee on Boilers ' the composition of 
 scale from the boilers of various ships is given viz. : 
 
 
 H. M. S. 
 
 A metkyst. 
 
 H. M. S. 
 
 Malabar. 
 
 H M.S. 
 Fox. 
 
 S. S. 
 Patrxlus. 
 
 S.S. 
 Vtlindra. 
 
 Sulphate of lime 
 Magnesia 
 
 94.64 
 
 2.88 
 
 95-93 
 ?. 10 
 
 77-3 
 10. SS 
 
 9 6 37 
 1.62 
 
 91.38 
 2.-?! 
 
 Silica 
 
 Traces. 
 
 ) j 
 
 Traces. 
 
 I j 
 
 I.4O 
 
 Peroxide of iron .... 
 
 Traces. 
 
 0.40 j 
 
 5.80 
 
 \ - 6 \ 
 
 1. 60 
 
 Water 
 
 2.4 1 ? 
 
 
 6.00 
 
 i-75 
 
 T..T.Q 
 
 
 
 
 
 
 
 In the boiler of the S. S. Deccan the water had been raised to ten times the density 
 of sea-water, so that the brine contained in 100 parts- 
 Chloride of sodium 27.76 
 
 Chloride of magnesium 3.72 
 
 Sulphate of magnesia 1.65 
 
 Sulphate of potassa 0.87 
 
 Water 66.00 
 
 and had a specific gravity of 1.224 at 60 Fahr. The thick saline deposit from the fur- 
 nace-crowns of this boiler contained in 100 parts 
 
 Chloride of sodium 97.83 
 
 Chloride of magnesium 0.59 
 
 Sulphate of lime 0.56 
 
 Peroxide of iron 0.02 
 
 Water.. 1.00 
 
 100.00 
 
 7. Cousin's Theory of the Formation of Deposits in Steam Boilers. When 
 
 sea- water in its natural state is evaporated in a boiler the following phenomena are 
 observed : 
 
 I. A few moments after ebullition commences the water in the boiler grows turbid, 
 and holds in suspension first free magnesia, then carbonate of magnesia. These two sub- 
 stances are present in small quantities, and are light, flaky, and have no tendency to 
 agglomerate. They form with the organic and earthy substances which the water holds 
 
392 STEAM BOILERS. CHAP. XVII. 
 
 in suspension the muddy deposits which are found in the bottom of boilers and on 
 horizontal heating-surfaces. 
 
 II. By the continuance of ebullition the water arrives soon at the point of saturation 
 with regard to the sulphate of lime, and from this moment, if the degree of saturation 
 is allowed to pass the point where the motion of the molecules of water are capable of 
 keeping the particles of sulphate of lime mechanically in suspension, these particles are 
 deposited as a crystalline crust on all surfaces in contact with water. 
 
 III. The heating-surfaces impart to the water in contact with them a sufficiently 
 high temperature to make it supersaturated as far as the sulphate of lime which it con- 
 tains is concerned. This sulphate is then deposited on the plates constituting the 
 heating-surfaces, forming there at once a thin layer of incrustation, whatever may be 
 the degree of concentration of the mass of water. Afterwards, when the degree of 
 concentration rises above the point mentioned above (see II.), the particles of sulphate 
 of lime spoken of in the same paragraph cling to this layer and increase the thickness 
 of the scale. It appears that -in case those particles which are precipitated without 
 being in contact with the heating-surfaces did not cling to such a layer of scale, they 
 would not adhere to the metal of the heating-surfaces, and would form merely a de- 
 posit and not scale. 
 
 IV. When the fires are hauled and the water in the boiler has cooled down, that por- 
 tion of the muddy deposit which the water held in suspension by the motion of its par- 
 ticles produced by ebullition falls down on the surfaces of the boiler, or rather on the 
 scale which covers them. This extremely thin layer of mud, lodged in the depressions 
 of the rough surface of the scale, remains there mostly when vaporization recommences. 
 Then, as soon as the water reaches again the above-mentioned point of concentration, 
 a second layer of sulphate of lime is formed on top of the first one, but separated 
 from it by a film of magnesia and carbonate of magnesia combined with a little oxide 
 of iron and organic matter which give a yellowish color to this film. Since the eva- 
 poration is very active on the crown-sheets of the furnaces, the incrustation forms there 
 very rapidly ; hence the scale should be much thicker there than elsewhere. But the 
 contrary is the case, because as soon as scale is formed it is detached by the con- 
 tractions and expansions of the metal occurring every time the intensity of the fire 
 varies. Indeed, it is found that the scale which covers this portion of the heating- 
 surfaces consists generally of a single layer, a fracture showing no intermediate film. 
 For the contrary reason the scale reaches greater thickness at points of less intense 
 heat, where it is composed of distinct layers separated by films, each layer being gen- 
 erally less thick than the single layer on the furnace-crown. 
 
SBC. 9. MANAGEMENT OF BOILERS. 393 
 
 V. The fracture of the incrustations shows an amorphous structure throughout 
 nearly their whole thickness, except at the side opposite to that in contact with the 
 metal, where it has an appearance of crystallization. On the other hand, the greater 
 part of the isolated concretions which are found at the bottom of the boiler consist 
 each of an amorphgus core enveloped by crystalline layers. These facts indicate that, 
 at the moment when a layer is deposited, it has a crystalline character due to the pre- 
 sence of a certain proportion of water ; but after having been in contact with the metal 
 for a certain length of time, this water of crystallization is set free and the scale be- 
 comes amorphous. Contact with the heating-surfaces suffices to produce this effect, 
 for their temperature exceeds always 570 Fahr., and it is known that 390 are suffi- 
 cient to deprive the sulphate of lime of its water of crystallization. Besides, it is easily 
 conceived that this calcination is less complete as the portions of scale are farther 
 removed from the metal, and is nothing at the face in contact with the liquid. (Ledieu.) 
 
 8. Prevention of the Formation of Scale in Boilers. Three methods are 
 employed to prevent the accumulation of scale in marine boilers viz. : I. Blowing-off 
 a portion of the water in the boiler when it has reached a degree of concentration at 
 which deposits would be formed, and replacing it by ordinary sea-water. II. Feeding 
 the boiler with water from a surface-condenser or distilled by a special apparatus, or 
 with sea-water which has been deprived of a portion of its salts by heating it to a 
 high temperature in a separate vessel. III. Mixing various substances with the water 
 of the boilers, which prevent the formation of scale either by mechanical or by chemical 
 action. (See section 11, chapter xviii.) 
 
 9. The Hydrometer. The concentration of the water in the boiler is determined 
 by means of an instrument called "hydrometer" or " salinometer," which is a float 
 having a constant weight and measuring by the depth of its immersion the relative bulk 
 of fluids of different densities having the same weight. 
 
 The hydrometer used in the United States Navy is a graduated narrow, cylindrical 
 tube closed at the top. The lower part is enlarged to give buoyancy to the instrument, 
 and terminates in a small globe filled with shot, which serves to keep, the hydrometer 
 floating in an upright position (see figure 147). 
 
 These instruments are generally made of glass, and the scale is marked on a slip of 
 paper, which is secured within the narrow tube. Each vessel is also furnished with a 
 standard copper hydrometer of similar shape, having the scale engraved on the narrow 
 stem. It is necessary to handle these latter very carefully, since any indentation would 
 alter the relation existing between the bulk and the weight of the instrument, and con- 
 sequently destroy the correctness of the scale. 
 
394 
 
 STEAM BOILERS. 
 
 CHAP. XVII. 
 
 Fig. 147. 
 
 
 32 
 
 In order to graduate this instrument it is first placed into a vessel containing dis- 
 tilled water of a fixed temperature, and the point to which it is immersed is marked 
 zero. Then sea-salt is dissolved in the water in the successive proportions of one, two, 
 three pounds, etc., of salt to thirty- two pounds of water, and the respective points to 
 
 n which the instrument, floating in the solu- 
 
 tion, is immersed at a uniform temperature 
 are marked ^, fa, ^, etc. Each of these 
 divisions is further subdivided into halves 
 and quarters. The temperature for which 
 the hydrometer is graduated must be mark- 
 ed on the scale. Hydrometers have some- 
 times three different scales, corresponding 
 respectively to temperatures of 190, 200, 
 and 210 Pahr. 
 
 Approximately, for an increase of 10 in 
 temperature the density of the solution, 
 as indicated by the hydrometer, decreases 
 by one-eighth of a division in other words, 
 the hydrometer will indicate a concentra- 
 tion which is one-eighth of a thirty-sec- 
 ond less than the true one when the tem- 
 perature of the solution is 10 higher 
 than the temperature for which the scale 
 was constructed. It is, therefore, necessary 
 to keep a thermometer immersed in the 
 water which is drawn from the boiler into 
 a suitable vessel for the purpose of testing 
 its concentration. (See section 10, chap- 
 ter xv.) 
 
 The divisions of the scale constructed 
 in the aforesaid manner are not uniform, 
 but decrease in length as the degree of con- 
 centration increases. The following inves- 
 tigation will show the relation which would exist between the lengths of successive 
 divisions of the scale in case the density of the solution was proportionate to the sum 
 of the densities of its separate constituents : 
 
 SCALE DEVELOPED 
 FULL SIZE 
 
 <f& 
 
 190 
 
 
 <& r< * 
 200 
 
 
 
 210 
 
 
 1 
 
 1 
 
 1 
 
 82 
 
 Ht 
 
 V 
 
 32 
 
 :? 
 
 32 
 
 I 1 
 
 3' 
 
 82 
 
 HJ 
 
 3 
 
 33 
 3 
 
 32 
 
 -1 
 8 
 
 32 
 
 33 
 
 33 
 
Sue. 10. 
 
 MANAGEMENT OP BOILERS. 395 
 
 If w = the weight of water contained in the solution, 
 v = its corresponding volume, 
 
 10 
 d = the volume of salt, the weight of which is go j 
 
 and if x, # #, # 1U , etc., represent the volumes of the solution displaced by the hy- 
 drometer at successive degrees of concentration corresponding to 0, -?, -fa, -fo, etc., of 
 the scale, then, since the weight of these volumes is equal, we have the equations : 
 w 33 w 34 w 35w , 
 
 ' := 320 + 2 d) " - "" 
 
 consequently 
 
 32 (v + d) . _ 32 (v + 2 d) 32fr + 3d) . . 
 
 ** = ~~33V~ ' " ~ ~34~ '" ~35~ 
 
 and the length of each division is proportionate to 
 
 x x -x x 
 
 l ~ 
 
 33 v 33 
 
 . 32 Q + d) 32(t>+2d) _ 32 . tt-32^ 
 
 *' " 33 34 - 33 ~34~~ 
 
 oo /^. i q rf\ qo . qo /7 
 
 O^ ^C j O U/j & V O^ C*' 
 
 The density of the brine, however, does not increase proportionately with the weight 
 of salt held in solution, but a condensation takes place when salt is dissolved in water, 
 which appears to be greatest for very much diluted brines ; when 100 ounces of water 
 are mixed with 34 ounces of common salt the decrease in volume is equal to 4 per cent, 
 of the sum of their respective volumes. 
 
 The presence of saline matter in water has the effect of raising its boiling point, and 
 this property has been made use of to determine the concentration of the solution. 
 For water containing the usual proportions of the salts of sea- water the boiling tem- 
 perature under mean atmospheric pressure is raised very nearly one degree for each ad- 
 ditional 2.58 per cent, of saline matter. 
 
 The specific heat of sea- water is .82, that of fresh water being 1.00. 
 
 1O. Influence of Temperature and Pressure on the Limit of the Satura- 
 tion of Water in a Boiler. The hydrometer indicates the density of the water in 
 the boiler due to the saline matter held in solution, but gives no indication of the rela- 
 tive quantities of the different salts constituting this saline matter. The formation of 
 scale depends chiefly on the quantity of sulphate of lime present in the water, and this 
 salt will be deposited when the temperature reaches a certain point, whatever the con- 
 
396 
 
 STEAM BOILERS. 
 
 CHAP. XVII. 
 
 centration of the water, as indicated by the hydrometer, may be. By reference to 
 Table XXXVIII. it will be seen that when the temperature of the water becomes 264 
 it can no longer hold the whole quantity of sulphate of lime present in ordinary sea- 
 water, and when the temperature rises to 280 it cannot hold any of this salt in solu- 
 tion. Under these circumstances the amount of deposit formed is in the direct ratio of 
 the quantity of this salt introduced into the boiler with the feed- water ; and since 
 blowing-off necessitates an increase in the quantity of feed- water, it increases instead of 
 diminishing the formation of a deposit of sulphate of lime and magnesia, besides wast- 
 ing the heat imparted to the water which is blown off, and introducing additional quan- 
 tities of air and of the destructive chloride of magnesium into the boiler. The only 
 advantageous effect of blowing-off under these conditions is the removal of the particles 
 of lime and magnesia kept in suspension mechanically by the currents of water in the 
 boiler. 
 
 Supposing that by feeding from the sea the concentration of the water in the boiler 
 is raised to -fa on the scale of the hydrometer, and when this point is reached the sea- 
 feed is shut off and the supply of distilled water from the surface-condenser is substi- 
 tuted, no more lime-salts will be introduced and no more scale will be deposited, but the 
 hydrometer will still indicate &. 
 
 TABLE XXXVIII. 
 
 Absolute pres- 
 sures in 
 atmospheres. 
 
 Temperatures of 
 sleam, correspond- 
 ing to pressures, 
 in degrees Fahr. 
 
 Density of water 
 at a temperature 
 of 59 Fahr. 
 
 Weight of sulphate 
 of lime contained in 
 100 parts of the 
 water in the boiler. 
 
 Total weight of sa- 
 line matter con- 
 tained in 100 parts of 
 the water in boiler. 
 
 Degrees of concen- 
 tration indicated 
 by hydrometer. 
 
 Temperatures at 
 which the water 
 would boil under 
 mean atmospheric 
 pressure. 
 
 I.OO 
 
 212 
 
 1.0990 
 
 O.6OOO 
 
 13979 
 
 si 
 
 216.32 
 
 1-25 
 
 224 
 
 1.0768 
 
 0.4683 
 
 IO.9I2 
 
 3f 
 
 215.42 
 
 i-5 
 
 234 
 
 1.0589 
 
 -3834 
 
 8-935 
 
 3| 
 
 214.52 
 
 i-75 
 
 243 
 
 I.04S9 
 
 0-3055 
 
 7.120 
 
 2 i 
 
 214.16 
 
 2.0O 
 
 25 1 
 
 1.0344 
 
 0-2335 
 
 5-442 
 
 f 
 
 213.62 
 
 2.25 
 
 258 
 
 1.0245 
 
 0.1688 
 
 3-934 
 
 if 
 
 213-39 
 
 2.50 
 
 264 
 
 I.Ol62 
 
 0.1132 
 
 2.639 
 
 1 
 
 213-35 
 
 2-75 
 
 270 
 
 1.0078 
 
 0.0551 
 
 1.285 
 
 | 
 
 213-34 
 
 3.00 
 
 275 
 
 I.OOO2 
 
 O.OOI2 
 
 0.030 
 
 i 
 
 213-33 
 
 3-25 
 
 3-5 
 
 280 
 285 
 
 |- the water cannot hold any portion of sulphate of lime in solution. 
 
 The foregoing table contains the conditions accompanying the saturation of sea- 
 water with regard to the sulphate of lime at various pressures and temperatures. The 
 data, adapted to our hydrometric scale and to Fahrenheit's thermometric scale, are 
 selected from a table prepared by Couste. It is based on the assumption that sea- water 
 
SBC. 11. MANAGEMENT OP BOILERS. 397 
 
 in its natural state, and hav-ing a density of 1.026, contains 0.15 per cent, of sulphate of 
 lime, 2.65 per cent, of chloride of sodium, and 0.70 per cent, of other substances, or in 
 all 3.50 per cent, of saline matter and 96.50 per cent, of fresh water. 
 
 11. Calculation of the Quantities of Water and Heat lost by Bio wing-off. 
 
 In order to maintain the water in the boiler at a certain concentration the quantity of 
 water extracted by blowing must bear to the quantity of water fed into the boiler the 
 same ratio as the number indicating the concentration of the feed- water to the number 
 indicating the concentration of the water in the boiler. 
 Let x represent the number on the hydrometer indicating the density of the feed- water ; 
 
 " y " the number on the hydrometer indicating the density of the water in 
 the boiler ; 
 
 " s " the number of pounds of water evaporated in a unit of time ; 
 
 " b " the number of pounds of water in the brine which is blown off. 
 
 Supposing the quantity of water in the boiler to remain the same, then the quantity 
 of salt blown off in a unit of time must be equal to the quantity of salt introduced with 
 the feed-water in a unit of time, in order to maintain the concentration at a fixed point. 
 This relation may be expressed by the following equation : 
 
 x (s -f- b) = y 5 / hence (s -j- b) : b : : y : x, 
 which proves the above rule. 
 
 In calculating the loss of heat by blowing-off absolute accuracy cannot be attained, 
 because the specific heat of sea- water at different densities has not been ascertained. It 
 has been taken as being the same for all densities as ordinary sea- water viz., 0.82 
 although for greater densities it is probably smaller. Further, since the boiling-point 
 of sea- water at different concentrations has not been ascertained for pressures higher 
 than the atmospheric pressure, no account has been taken in the following example of 
 any increase of temperature due to the higher degree of concentration. For simplicity's 
 sake the temperature of the water may be taken as representing the units of heat con- 
 tained in each unit of weight. The error due to these causes is slight and practically 
 unimportant. The method of calculating the loss of heat by blowing-off will be ex- 
 plained by the following example : 
 
 Supposing the temperature of the feed-water to be 110, its concentration 7 V on the 
 scale of the hydrometer, the steam-pressure in the boiler 35 Ibs. per square inch above 
 the atmosphere, what percentage of the total heat imparted to the water in the boiler is 
 
 lost by blowing-off when the density of the water is kept at - - on the scale of the 
 
 oH 
 
 hydrometer 1 
 
398 STEAM BOILERS. CHAP. XVII. 
 
 Regarding the weight of water evaporated as unity, the weight of fresh water 
 contained in the brine blown off, represented by b, bears to the weight of fresh 
 water contained in the feed the following proportion viz.: 6 : (1 + 6) :: 1 : 1.76 ; hence 
 
 b = A = 1.333. 
 
 1 Wi V 1 7^ 
 
 The total weight of brine blown off is consequently = 1.333 -f - - = 1.406 ; 
 
 o& 
 and the total weight of feed- water for each pound of water evaporated = 2.406. 
 
 The temperature of steam at a pressure of 35 Ibs. above the atmosphere is 280 in 
 round numbers. The units of heat present in the feed required for each pound of 
 water evaporated are (2.406 X 110 x .82) = 217.02. 
 
 The units of heat required to raise the amount of water blown off from 110 to 280 
 are equal to 1.406 x (280 - 110) .82 = 196.00. 
 
 Total units of heat present in each pound of steam = 1199.4 
 
 " " " " in water blown off = (1.406 x 280 x .82)= 322.8 
 
 Total units of heat in water and steam = 1522.2 
 
 " " " present in feed-water = 217.0 
 
 Total units of heat communicated to the water = 1305.2 
 
 Percentage of heat in water blown off in terms of total heat expended = - ^ K ^ = 14.94 
 
 per cent. 
 
 12. Cleaning and Scaling Boilers. The boilers should be cleaned of soot and 
 ashes as soon as possible after the fires are hauled. The tubes are to be swept first, 
 commencing with the top row of each tube-box. Before commencing to sweep the tubes 
 turn the ventilators away from the wind and close the furnace and ashpit doors, so that 
 as little dust as possible may fly about the fire-room ; but open all the connection-doors 
 to let the draught carry the soot or ashes up the chimney. 
 
 Tube-brushes are made either of wire or of coir. They should fit the tubes snugly 
 so that it requires some force to push them through, because they should not merely 
 remove the loose soot, but detach also the hard carbonaceous scale which covers the fire- 
 surfaces of the tubes. This scale frequently clings so tenaciously to the surfaces that it 
 has to be scraped off. 
 
 Tube- scrapers consist of steel strips secured to circular end-pieces and bent so as to 
 form ridges running at an angle with the axis of the tool. An efficient scraper is formed 
 
SBC. 12. MANAGEMENT OF BOILERa 399 
 
 by long, thin steel strips bent to a spiral shape, which are secured to short, cylindrical 
 end-pieces, and can be adjusted to a greater or less diameter by moving a nut in the 
 direction of the axis of the tool. 
 
 Deposits of salt from leaks which may have accumulated in or around the tubes 
 must be carefully removed with a scraper or long chisel. In the narrow spaces between 
 vertical water-tubes such deposits are frequently very difficult to remove, unless they 
 are loosened by soaking them for a long time in fresh water. For this purpose a dam 
 is built up at both ends of the tube-box, and water is kept in the latter at such a height 
 as to cover the deposits of salt. Such deposits may also be loosened by directing 
 against them, by means of a hose, a strong stream of water thrown by a force-pump. 
 But in no case should water be used in the flues or connections unless the tubes, con- 
 nections, and furnaces have been thoroughly cleaned of soot and ashes. 
 
 After the tubes have been cleaned remove the soot from the uptake and back-con- 
 nections ; scrape the plates clean of scale and salt, and sweep them off with a stiff 
 broom. The furnaces and ashpits are to be cleaned in the same manner after the grate- 
 bars have been removed. The grate-bars are to be cleaned by knocking clinkers and 
 cinders off with a scaling-hammer. Cast-iron bars which are much broken, burnt, or 
 bent have to be replaced with new ones ; wrought-iron bars have to be repaired or 
 straightened. 
 
 The chimney is to be swept and scaled periodically on the inside. When bitumi- 
 nous coal is burnt the accumulation of soot in the chimney becomes so great that it 
 frequently catches fire and causes serious injury to the chimney unless it is swept off 
 periodically. 
 
 The outside of boilers must be cleaned of salt deposited from leaks after every run. 
 When the leak cannot be repaired before the boilers are put in use again the deposits 
 which close a leak should not be disturbed. Special attention must be paid to the 
 cleaning of the front of boilers, where salt accumulates in consequence of the water 
 leaking from the water-gauges. The lower part of the boiler-front is generally covered 
 by a coating of ashes, which cling to the boiler when they are wetted after hauling them 
 from the furnaces and ashpans. The bottom of boilers is frequently found covered with 
 salt and dirt deposited by the bilge- water which is washed up against the boiler by the 
 rolling of the vessel. Dry ashes and dust accumulate on the top of boilers, and must 
 be swept off from time to time. 
 
 The water-gauges, salinometer-pots and their pipes, and other attachments must be 
 cleaned not only on the outside, but must be disconnected so as to clear their passages 
 of scale and sediment. Valves and cocks must be reground and packed, if necessary. 
 
400 STEAM BOILERS. CHAP. XV1L 
 
 Valve-stems and plug-cocks must be greased with oil or tallow, and the articulations of 
 safety-valve levers must be cleaned and oiled. 
 
 The outside of boilers must always be kept covered with a heavy coat of paint, 
 which will require frequent renewal at the front and bottom, where the leakage of water 
 from the gauges and the manholes and mudholes, the wetting of ashes, and the wash- 
 ing-up of the bilge-water render the boiler especially liable to corrosion, while the 
 frequent cleaning and scraping soon wears off the coat of paint. For painting the 
 shell of boilers either red lead or a brown metallic paint prepared from the brown 
 oxide of iron is used. The furnaces, connections, uptakes, and the inside of chimneys 
 should receive a coat of the same paint when the boilers are to be put out of use for 
 some time. The inside of steam-drums, if accessible, should be kept covered with a 
 heavy coat of lead or brown metallic paint, especially the lower part of horizontal 
 drums, where water is apt to collect. The cast-iron of connection, furnace, and ash- 
 pit doors may be painted with lamp-black and oil, or may be simply oiled. Some engi- 
 neers paint the connection-doors white, in order to make the fire-room brighter. In 
 no case should the iron of boilers be whitewashed, unless it has first received a heavy 
 coat of paint, because the lime absorbs moisture, and thus would promote corrosion. 
 The outside of chimneys, the hatch-gratings, ventilators, etc., may be painted with 
 asphaltum. In no case should paint be applied to any part of the boiler before it is 
 thoroughly dried, cleaned, and scaled, so that the paint may cover the clean body of 
 the metal. 
 
 The scaling of boilers should be commenced as soon as they have cooled off suffi- 
 ciently to be entered, for when the scale is still damp it can frequently be removed much 
 more easily than after it has got quite dry. As long as the thickness of the scale does 
 not exceed the thickness of ordinary writing-paper it should not be disturbed, as it 
 forms the best protection of the iron against corrosion ; but as it grows thicker its 
 low thermal conductivity produces a perceptible diminution of the evaporative effi- 
 ciency of the boiler, and exposes the plates to the danger of being burnt. The expan- 
 sion and contraction of the plates cause thick scale to crack and to become partially 
 detached from the plates, and in this condition the scale favors the corrosion of boilers 
 by admitting and retaining moisture in contact with the plates. 
 
 It has been recommended to loosen the scale, before commencing the operation of 
 cleaning the boiler, by the sudden expansion or contraction of the plates or tubes. This 
 might be done by admitting cold water into the boiler immediately after hauling the 
 fires and blowing the hot water out of the boiler ; or by admitting a large mass of steam 
 into the cooled-off boiler and allowing it to condense there ; or by making a light fire of 
 
SEC. 12. MANAGEMENT OP BOILERS. 401 
 
 wood-shavings in the furnaces or connections. The latter plan has resulted in several 
 instances in burning the boiler. But all these methods should be avoided, as the sudden 
 expansion and contraction of the plates is always injurious to the boiler. 
 
 The scale on the plates must be chipped off with scaling-hammers or wedged off with 
 scaling-bars. These tools must not be ground to a sharp edge, and the chipping must 
 be done carefully to avoid making indentations in the surface of the iron and injuring 
 the rivet-heads ; where such indentations are made the scale which is subsequently 
 formed adheres more tenaciously, and the iron is more readily attacked by corrosion. 
 Special care must be taken in scaling the tube-sheets to avoid injuring the tube- 
 ends. The crown-sheets of furnaces are generally scaled first. The scale should be re- 
 moved completely from the plates wherever they are accessible, especially between the 
 rivet-heads and at the edges of laps, and around the crow-feet or the heels of braces. 
 The narrow water-spaces around the back-connections are frequently not accessible for 
 thorough scaling unless handholes are cut at suitable places in the shell. 
 
 The fire-tubes of marine boilers are generally spaced so closely that the greater por- 
 tion of their surfaces is not accessible for scaling. The vertical spaces between these 
 tubes must be kept clear by running a scaling-bar through them, by which means the 
 scale may be detached from the tubes in flakes. Various devices for making the tubes 
 readily removable for scaling have been tried, but have not given satisfactory results. 
 When the accumulation of scale on the tubes becomes very great it will be necessary to 
 remove a sufficient number of tubes from the boiler to make the remaining ones accessi- 
 ble for scaling. 
 
 Water-tubes can be scaled by running through them a steel bar with a cutting edge 
 at the end, or by boring out the scale with Garviri's scaling-tool, consisting of a revolv- 
 ing cutter worked by a wrench, and fed by turning a screw which passes through the 
 tube and is secured and centred at both ends. Care is required in working and adjust- 
 ing these tools so as not to cut into the metal of the tube. 
 
 The shell of the boiler in the water and steam spaces must be scraped clean of mud 
 and rust, all the loose scale must be knocked off the braces, and the whole mass of 
 scale and dirt must be swept down through the water-spaces to the bottom of the 
 boiler, taking special care that no dirt lodges between the tubes or on the furnace- 
 crowns. The dirt is then raked with small hoes out of the boiler through the mudholes 
 in the bottom. 
 
 Finally the boiler is to be washed out. By means of a hose, connected to a force- 
 pump, direct a heavy stream of water at every portion of the interior of the boiler, 
 especially at such parts as are not accessible for scraping. By this means much loose 
 
402 STEAM BOILERS. CHAP. XVII. 
 
 scale and mud will be washed down. The latter contains frequently grease, partly de- 
 composed, various salts, particles of copper, and other substances which are very inju- 
 rious to the boiler. Commence the washing-out in the steam-space, directing the jet of 
 water especially into the steam-drum, between the tubes, and into the water-spaces 
 around the back-connections. Then take the hose in succession into each manhole over 
 the furnace-crowns, and finally into the mudholes at the bottom ; commencing at one 
 end of the boiler, wash all the dirt to the other end. During this operation continue 
 to rake the dirt and scale out through the mudholes in the bottom. Give a slight list 
 to the vessel while washing out the boilers, so that the water will naturally flow to the 
 front of the boilers, carrying the dirt along with it. The fire-room floor-plates in front 
 of the boiler are removed to let the water run directly into the bilge of the vessel. 
 
 After the boiler has been washed out, and all the water removed from the bottom by 
 raking and swabbing, the boiler must be kept open till it is perfectly dry in every part. 
 Pans with burning charcoal may be placed in the furnaces and connections to dry the 
 boiler more rapidly. (See section 14 of the present chapter.') 
 
 13. Repairing Boilers. All leaks, however small, should be stopped as soon as 
 possible, and all temporary repairs, such as have been described in section 3 of the pre- 
 sent chapter, should be replaced, after the boilers are emptied, by substantial work 
 which will correct the evil permanently and strengthen the defective part. The mass 
 of steam evolved from a leak frequently prevents its exact location while the boiler is 
 under steam, and the accumulation of salt deposited by the water issuing from a leak 
 is often so great that its source cannot be traced after the boiler is emptied, except by 
 filling it again with cold water and producing a pressure in it by means of a force- 
 pump. 
 
 Leaky seams and rivets must be calked. When a rivet leaks after repeated calking 
 it must be cut o\it and a new rivet put in its place. The same must be done with rivets 
 the heads of which are so much corroded that their holding power is seriously impaired. 
 Such rivet-heads are found especially at the bottom of boilers on the outside of the 
 shell and in the connections and uptakes, and their condition is frequently not indi- 
 cated by any leaks, and may not be discovered, since the corroded heads often retain 
 their original shape, unless they are tested by striking them with a hammer. In places 
 where new rivets cannot be driven properly bolts secured by nuts may be substituted 
 for them. To make a tight joint with them wrap a little hemp or cotton wick covered 
 with putty around their shank close under the head, and put a washer covered with 
 stiff putty under the nut. Such bolts must likewise be used when the boiler is old and 
 weak, as the jars produced by riveting and calking would be likely to start new leaks. 
 
SBC. 13. MANAGEMENT OP BOILERS. 403 
 
 These bolts have frequently to be wired in place by the same method as is applied to 
 sockets, described in section 3, chapter x. 
 
 In replacing a socket-bolt keep the socket in place by inserting the new bolt or a 
 temporary plug as the old bolt or rivet is being withdrawn. After removing screw stay- 
 bolts the threads in the plates are generally found so much injured that the holes have 
 to be reamed out and new threads cut, necessitating the use of larger bolts. When the 
 holes have become much enlarged or the plates much weakened by corrosion replace 
 socket-rivets and screw stay-bolts with socket-bolts, and put large washers under their 
 heads and nuts. 
 
 Defective parts of a boiler have often to be patched with plate-iron, in order to 
 strengthen them or to stop leaks. 
 
 When a patch is riveted on and made tight by calking it is called a hard patch. 
 When a patch is bolted over the defective place, the joint being made with putty, it is 
 called a soft patch. The bolts are generally secured by nuts, but sometimes tap-bolts 
 have to be used. The putty is prepared by kneading together white lead, ground in oil, 
 with dry red lead, and mixing with it some fine iron borings or filings to make it stiff. 
 
 A soft patch should not be applied to a surface in contact with fire or hot gases, 
 except in case of necessity as a temporary expedient to stop a bad leak. When a soft 
 patch has been applied to a furnace-crown it is best not to use the furnace, because the 
 heat would cause the patch to warp sooner or later. On the bridge-wall or on the bot- 
 tom of the back-connection a soft patch may be protected by covering it with fire-clay or 
 ashes. When a boiler is old and weak a soft patch is often to be preferred to a hard 
 patch to avoid jarring the boiler by riveting and calking. Leaky seams which have 
 been repeatedly calked, so that the lap is much reduced in width, have to be made tight 
 by covering them with a soft patch. It must be observed that a soft patch generally 
 does not add to the strength of the defective plate, but is frequently the cause of greater 
 weakness on account of the metal cut away in drilling the bolt-holes, and because corro- 
 sion may continue unseen under the patch. 
 
 When a hard patch is to be applied to the outer shell it may be placed directly over 
 the defective plate. But in places which come in contact with the fire or hot gases it is 
 necessary to reduce the thickness of the metal as much as possible ; therefore the defec- 
 tive part has to be cut out. The hole thus formed in the plate should be rounded, since 
 sharp corners favor the formation of cracks. 
 
 Cut the patch from boiler-plate of good quality and of the same thickness as the 
 plate which is to be repaired, unless the latter is much worn, when the patch may be 
 reduced in thickness. It is advisable to calk the edges of the patch rather than those oi 
 
404 STEAM BOILERS. CHAP. XVII. 
 
 the old plate ; for this reason, in repairing a crown-sheet, put the patch inside the fur- 
 nace ; in this position it can also be fitted better than when it is placed inside the boiler. 
 After cutting away the defective portion of the plate with a cape-chisel, lay off and drill 
 around the opening rivet-holes of the required size and spaced according to the rules 
 given in sections 13 and 14, chapter viii. Pit the patch so that it lies close on the plate, 
 and, holding it in position, mark on it the rivet-holes and the opening cut in the plate. 
 Then drill the holes and trim the patch to the proper size and form. Bolt it securely in 
 place while it is being riveted, and finally calk it. 
 
 When the patch is to cover a curved or uneven surface make a template for it of a 
 piece of sheet-lead hammered into shape while holding it against the surface to be cov- 
 ered. After fitting and drilling the patch heat it to a dull-red heat, put it in position, 
 and draw it tight up against the plate by means of bolts, so as to make a close fit. 
 
 Blisters occur frequently in iron plates exposed to an intense heat. When they are 
 thin it is sufficient to chip them off as far as the lamination extends ; but when a blister 
 is thick the plate has to be cut out and patched. Some engineers recommend to rivet 
 blisters down when they first appear, and thus prevent their extension. 
 
 Cracks formed by unequal expansion in the middle of plates, or starting from rivet- 
 holes, are sure to increase in length unless promptly stopped. When a crack does not 
 exceed three inches in length it may be closed with rivets. Drill a small hole at each 
 end of the crack, taking care that it does not extend beyond the holes, then drill one or 
 two more holes through the crack and countersink them. Put rivets through these 
 holes and spread their heads well, hammering them down pretty flat so that they cover 
 the crack completely. 
 
 When the crack is of considerable length the corresponding portion of the plate 
 must be cut away and patched. The same must be done when cracks run from hole 
 to hole in a seam, or when a number of cracks appear in close vicinity, running from 
 the rivet-holes of a seam to the edge of the plate. 
 
 When a furnace-crown has come down, or partly collapsed, but the injury is not very 
 extensive, it may be restored to its original shape. Bring the defective place to a dull- 
 red heat by means of a charcoal-fire in a portable furnace, then force it up with a screw- 
 jack, placed on a foundation of blocks in the ashpit, and acting directly on a block the 
 upper surface of which has the shape of the furnace-crown. The adjoining unin- 
 jured parts of the crown-sheet should be firmly wedged to prevent their distortion dur- 
 ing this operation. It is to be observed that the injured part never regains its original 
 strength and stiffness. It may be strengthened by additional braces, or by angle-iron 
 rings, as described in section 6, chapter ix. The safest plan is to renew the plate, either 
 
SEC. 13. MANAGEMENT OP BOILERS. 405 
 
 in part or wholly, according to the extent of the injury. When these precautions can- 
 not be taken before the boiler has to be used again, the defective furnace may be put 
 out of use and the injured crown-sheet shored up. 
 
 When a crown-sheet is defective in several places it is better to cut out the whole 
 plate, or at least the portion containing the several defects, than to put on a number of 
 small patches. Arrange the patches so that their seams do not come close to the fire, 
 and that they clear the crow-feet and other attachments of the braces. 
 
 When tubes leak in the joints at their ends, and cannot be made tight by expanding 
 or calking them, the leak may be stopped by driving ferrules in their ends. Several 
 methods of plugging fire-tubes have been described in section 3 of the present chapter. 
 On account of the expansion of the rod exposed to the hot gases, water-tubes cannot be 
 plugged effectually in the same manner as fire-tubes by passing through them a long 
 rod with a nut at each end which holds the cup- washers that cover the ends of the tube. 
 Such tubes must be cut out, and the holes in the tube-sheets must be closed by cup- 
 washers, each held by a short bolt having a T-head or passing through an iron bar 
 which straddles the tube-hole. Tubes which have been plugged should be replaced as 
 soon as possible with new ones. The manner of securing tubes has been described in 
 section 6, chapter xi. 
 
 To remove a ferrule from a tube split it by cutting a groove through it with a cape- 
 chisel. A tube in one of the outside rows may be removed by cutting it off with a tube- 
 cutter or with a chisel inside the tube-sheets. To remove a tube from one of the inner 
 rows bend its ends inward, pass a rod through it, and put a nut and washer at one end ; 
 hook a tackle at the other end and pull the tube out, driving it at the same time at the 
 other end. The accumulation of scale on a tube makes its passage through the hole in 
 the tube-plate often very difficult. The scale may be cracked off by inserting a red-hot 
 iron bar into the tube. Tubes which have been removed can often be used again in 
 shorter boilers after cutting off their battered ends, or they may be lengthened by braz- 
 ing new ends to them. 
 
 If a tube-plate is bulged from any cause, tie it to the opposite tube-plate by means 
 of rod-braces which take the place of some of the tubes, and are held at the ends by 
 nuts and stout washers, or substitute stay-tubes secured by nuts for a number of tubes 
 expanded in the ordinary manner. 
 
 Leaky seams of old boilers which cannot be calked properly may be made tight by 
 driving an iron cement into them, forming a rust- joint. The following composition for 
 such a cement is given by Roper : cast-iron borings or turnings, 19 Ibs. ; pulverized 
 sal-ammoniac, 1 Ib. ; flour of sulphur, Ib. ; should be thoroughly mixed and passed 
 
406 STEAM BOILERS. CHAP. XVII. 
 
 through a tolerably fine sieve. Sufficient water should be added to wet the mixture 
 through. It should be prepared some hours before being used. A small quantity of 
 sludge from the trough of a grindstone will improve its quality. Instead of sal-am- 
 moniac, urine may be used. 
 
 When a boiler is old and much worn leaks in seams, around rivet and stay-bolt 
 heads, and at the ends of tubes frequently become very general. When time is want- 
 ing to subject the boiler to thorough repairs the leaks may frequently be stopped tem- 
 porarily by introducing into the boiler oatmeal, or some other substance which is 
 changed into a paste by boiling, and, after steam is formed, finds its way into every 
 crevice or point of least resistance. The presence of such substances in the boiler pro- 
 duces, however, generally violent foaming, and they often obstruct the passages of 
 water-gauges and clog the valve-seats. 
 
 When the water-legs and bottoms of old boilers are worn out generally, so that they 
 can no longer be made tight by calking and patching, they may be filled with cement. 
 Level the boiler by trimming ship till a little water in the boiler stands at the same depth 
 everywhere. Close the mudhole-doors. Prepare a mixture of three parts of Portland 
 cement and one of sand, or equal parts of Roman cement and sand, with water, thin 
 enough to ran, in a trough placed in the fire-room at a height slightly above the man- 
 holes between the crown-sheets. An inclined shoot leads from this trough to each man- 
 hole. The cement must be mixed in the trough in sufficient quantity at once and with 
 some quickness, as it sets rapidly under water. The level of the cement in the boiler 
 must be kept some inches below the grate-bars, and the extremities of the feed and 
 blow pipes must be kept clear. 
 
 14. Preservation of Boilers. A proper observance of the rules for the manage- 
 ment of boilers given in the present chapter is essential for their preservation. The 
 action of the various causes tending to deteriorate steam boilers, and the means by 
 which their influences may be neutralized, will be described in detail in the following 
 chapter. 
 
 The most efficient method of protecting a boiler against internal corrosion consists 
 in covering its surfaces with a thin, adhesive layer of ordinary boiler-scale. This scale 
 must be formed as soon as possible after steam is raised, before oxidation has com- 
 menced on the iron surfaces : and it must be very thin, otherwise it will soon crack with 
 the expansions and contractions of the plates, and the moisture entering and retained 
 between the iron and the scale will aggravate the evil of corrosion. It is recommended 
 to keep the water in the boiler at about tJtree times the density of ordinary sea-water 
 for a short time after steam is raised, in order to produce this protective layer of scale. 
 
SBC. 14. MANAGEMENT OF BOILERS. 407 
 
 In England a thin coating of Portland cement has sometimes been substituted for 
 scale on the interior surfaces of boilers. It is recommended that all boilers under con- 
 struction should receive such a coating, which should be renewed after the boiler has 
 been tried under steam. The cement is ground very fine, and two parts of cement are 
 mixed with one part of sand. Before applying it the iron should be scraped quite bare, 
 otherwise the cement will not adhere to it. It is then applied with a common white- 
 wash-brash like paint ; as soon as one coat is dry another is laid on, two or three coats 
 being applied in this manner. The cement becomes quite hard, almost like iron ; the 
 thinner it is applied the better, otherwise the expansion and contraction of the plates 
 will throw it off. In some cases it has been found to stick upon the sides and tops of 
 furnaces for two or three voyages. It is recommended to fill with it the holes caused by 
 pitting, after scraping the respective parts quite clean. It is sometimes used as a foun- 
 dation upon which a firmly-adhering layer of ordinary scale is allowed to accumulate. 
 When it is properly applied it is not washed off by the feed, but on the Sultan and 
 Olatton its use was discontinued because particles of it were carried over into the 
 engines ; this is ascribed to the faulty manner in which it was applied. Sometimes a 
 layer of cement about 1 inch thick is applied to the bottom of cylindrical boilers in- 
 side to protect them against corrosion. 
 
 To prevent the corrosion of boilers when not in use they must be kept dry, and the 
 iron must be covered inside and outside with a protective coating. Every part of the 
 boiler should receive, when new, one or two coats of metallic paint. Boilers that are to 
 be laid up for some length of time receive often inside a coat of oil, which is applied 
 with a syringe at parts which cannot be reached with a brush. Fish-oil was formerly 
 frequently used for United States naval boilers, but its use is now interdicted and 
 hydrocarbon oils are to be used instead, because their decomposition is not accompanied 
 by the formation of fatty acids. Boilers built for the English naval service have been 
 completely filled with linseed or mineral oil. A pressure of 15 or 20 pounds was pro- 
 duced by means of a force-pump, provision being made for the escape of air from the 
 upper corners of the boiler ; the oil was then run out, and the process was renewed every 
 six months. In other cases the boilers have been heated gently to dry them, and two 
 coats of boiled linseed-oil have been applied. In all such cases the surfaces must be 
 carefully cleaned of rust and loose scale before the oil is applied to them. 
 
 In the English navy two different systems are used for the preservation of boilers not 
 in use. The one is known as the wet and the other as the dry system. The former con- 
 sists in keeping the boilers filled completely with sea-water mixed with carbonate of 
 soda ; 25 pounds if soda-ash, or 50 pounds if ordinary crystal-soda, be used for every 
 
408 STEAM BOILERS. CHAP. XVII. 
 
 100 cubic feet of sea- water in the boiler. This is to be dissolved and placed in the bot- 
 tom before running up the boiler. The sufficient saturation of the water with soda 
 should be tested by placing a piece of clean, new iron with some of the mixture in a 
 bottle for a night ; if the iron rusts more soda must be added. Special care must be 
 taken that every part of the boiler is filled, the air being allowed to escape from the 
 highest part of the boiler. 
 
 Instead of soda, slaked lime is also used, in the proportion of 8 pounds of lime to 
 every 1,000 gallons of sea- water. This lime must be carefully cleaned out before getting 
 up steam, or there may be heavy priming. 
 
 The dry process consists iu removing all water from the boilers to dryness, by means 
 of stoves if necessary ; after which shallow iron pans containing altogether two or three 
 hundred-weight of quicklime are placed in the bottoms, over the furnaces, and above 
 the tubes, the quantity of lime depending on the size of the boiler. In addition a sheet- 
 iron tray of burning coal is to be placed in the ashpits or furnaces till the coal is coked ; 
 then it is introduced into the boiler and the latter immediately closed air-tight. The burn- 
 ing coke consumes much of the oxygen of the air in the boiler, and increases the effi- 
 ciency of the dry lime. At least every six months the boiler is to be opened for inspec- 
 tion, and, if the lime is found to be much slaked, the pans at the bottom, which can be 
 removed without considerably changing the air, are to be taken out and refilled with 
 fresh lime. In addition to this, when the atmospheric dampness is extreme, light fires 
 in the ashpits are sometimes found to be necessary. 
 
 Weston found that the amount of oxygen lost by the air in boilers where the dry 
 process had been used varied from nothing (in two cases) to 90 per cent. The dry -lime 
 process frequently fails from the extreme difficulty of absolutely excluding moisture. 
 By introducing the burning coke into the boiler 60 per cent, of the oxygen in the 
 boiler may be consumed at once. After opening a boiler which has been treated 
 by the dry-lime process, air must be allowed to circulate through it before any one 
 enters it. 
 
 15. Extract from "Instructions for the Care and Preservation of the 
 Steam-machinery of United States Naval Vessels (1879)." 
 
 "16. A thin deposit of scale will be useful in protecting the interior surfaces of boil- 
 ers from corrosive action, and the use of zinc in the boilers will, it is believed, divert 
 corrosion from the iron. In order that the best results may be obtained it is deemed 
 advisable to ensure metallic continuity between the zinc and the boilers. 
 
 "17. No tallow or oil of vegetable or animal origin is to be put into the boilers for 
 any purpose whatever, but Cranes mineral oil, or its equivalent, will be used. This 
 
SEC. 15. MANAGEMENT OF BOILERS. 409 
 
 prohibition applies to all boilers in nse in the navy under cognizance of this Bureau, of 
 whatever type or service. 
 
 "18. The dry-pipes and drains of the steam-drums are to be examined frequently 
 to ascertain if the holes in them are clear. 
 
 "19. The boilers, when empty, are to be kept dry by such means as are at the dis- 
 posal of the officer in charge. The water-bottoms and lower parts of the fronts are to be 
 kept free from scale, rust, and ashes, and well painted. 
 
 "20. The boilers are not to be used as water- tanks for fresh water, nor for trimming 
 ship. 
 
 "21. The exteriors are to be kept as dry as possible, and nothing wet or com- 
 bustible is to be stowed over or around them. The bilges in the fire-room are to be 
 kept dry and well whitewashed. 
 
 "22. Sudden changes of temperature in the boilers are to be avoided, and, when 
 time will permit, at least three hours should be occupied in raising steam from cold 
 water. 
 
 "23. When not under steam a cover must be fitted to prevent water from going 
 down the smoke-pipe. 
 
 " 24. The uptakes are to be kept free from dirt and well painted. 
 
 " 25. The number of hours each boiler has had fires within it since the ship was com- 
 missioned is to be stated in each quarterly report. 
 
 "26. After the fires are hauled, and before the water is blown out of the boilers, the 
 furnaces and ashpits should be closed. 
 
 "27. The mineral oils which are to be used for interior lubrication float on the sur- 
 face of the water in the boilers without being decomposed, and the surface-blows are to 
 be used as rarely as possible, in order that the oil may not be blown overboard." 
 
CHAPTER XYIII. 
 
 CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 
 
 1. General Causes of the Deterioration of Boilers. The deterioration of 
 boilers consists in the accumulation of calcareous scale and sediment within them, and 
 in the diminution of their strength and the starting of leaks in consequence of corrosion 
 and fracture and of the burning and distortion of the plates. 
 
 The deterioration of boilers is due to a variety of known causes, many of which are 
 in a great measure avoidable, while for some of them no reliable practicable method of 
 prevention has yet been discovered. There are some destructive agencies at work, short- 
 ening the life of boilers, the action of which is not fully understood ; and instances of 
 rapid deterioration of boilers occur from time to time for which no definite cause can be 
 assigned. In general the untimely deterioration of boilers may be traced to some of the 
 following causes viz. : 
 
 Inferior quality of or defects in the materials used in the construction of the boilers, 
 especially the want of strength and ductility, and the presence of cracks, laminations, 
 and surface-defects in plates. 
 
 Bad workmanship, causing injuries to the materials by punching, drifting, and burn- 
 ing, indenting the plates in calking, defective welding and riveting, and want of tight- 
 ness in seams. 
 
 Deficiency of structural strength of the boiler and improper methods of connecting 
 the various parts, causing severe local strains to be produced by the steam-pressure 
 or by the unequal expansion and contraction of certain parts in consequence of varia- 
 tions of temperature. 
 
 Faulty design of the boiler, causing inaccessibility for cleaning and repairs, and 
 defective circulation of the water preventing the free escape of steam from heating- 
 surfaces. 
 
 Mismanagement of the boiler : subjecting it to great differences and sudden variations 
 of temperature ; overheating the parts exposed to the direct action of the fire or hot 
 gases ; letting the steam-pressure exceed the safe limit ; allowing the formation of thick 
 deposits of scale, and neglecting to clean and repair the boiler in time, to keep it dry, to 
 protect the surfaces by paint, etc. 
 
 410 
 
SEC. 1. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 411 
 
 The use of sulphurous fuel and of impure feed- water containing corrosive substances 
 and producing deposits of solid matter. 
 
 Galvanic action in consequence of the presence of heterogeneous metals in the boiler, 
 which either enter in its construction or have been introduced with the feed- water. 
 
 A gradual deterioration of boilers after they have been put in use appears unavoid- 
 able ; but while stationary boilers frequently last twenty years, the life of marine boilers 
 ranges, under favorable conditions, from nine to twelve years, and in naval vessels is 
 often limited to six years of use. 
 
 The greater durability of stationary boilers compared with that of marine boilers is 
 mainly owing to two circumstances. In the first place, stationary boilers can generally 
 be made of such a size and capacity that there is no need of urging the fires, and of such 
 a form that they are accessible in every part, that the circulation of the water and the 
 escape of the steam from their heating-surfaces is unobstructed, and that the alterations 
 of form due to variations of pressure and temperature do not produce severe local 
 strains ; while in marine boilers these conditions have frequently to be sacrificed on ac- 
 count of the restrictions with regard to weight and space imposed upon the designer. 
 In the second place, the feed- water of stationary boilers is generally purer than that of 
 marine boilers, since the water used for the purpose is either originally more free from 
 injurious salts and acids, or can be passed through purifiers of ample capacity which 
 would not be permissible on board of vessels on account of their bulk and weight. 
 
 A perceptible diminution of the endurance of marine boilers has taken place since 
 the introduction of surface-condensers and of high steam-pressures. 
 
 The shorter life of boilers in naval vessels compared with that of boilers in merchant- 
 vessels is to be ascribed principally to the irregular manner in which the former have 
 often to be worked. The boilers of merchant- vessels are generally worked under uni- 
 form conditions for regular periods, with certain intervals of rest during which the boilers 
 may be cleaned and repaired. On the other hand, in boilers of naval vessels steam has 
 frequently to be raised very quickly. They are sometimes kept under steam for long 
 periods, either working with full power or lying under banked fires ; and at other times 
 steam is raised and lowered, and the fires are started and hauled, after short intervals. 
 Under these conditions the boilers of naval vessels are subjected to frequent changes and 
 inequalities of temperature, and to frequent exposure of their imperfectly-dried surfaces 
 to the action of the atmospheric air ; their cleaning and repairs cannot take place at 
 regular intervals, and the facilities for effecting thorough repairs are frequently wanting 
 during long cruises on distant stations. 
 
 The formation of thick deposits of scale on the heating-surfaces of boilers not only 
 
412 STEAM BOILERS. CHAP. XVIII 
 
 diminishes greatly their evaporative efficiency, bnt leads frequently to injuries affecting 
 their strength and durability by causing the plates to become overheated. The condi- 
 tions under which scale is formed, its character, and the method of preventing its forma- 
 tion by blowing-off have been discussed in sections 5-10, chapter xvii. Various other 
 means of preventing its formation will be discussed in sections 10 and 11 of the present 
 chapter. 
 
 The external corrosion of the shell of marine boilers caused by leakage from seams, 
 rivets, man and hand holes, water-gauges, etc., the leakage of water through the deck, 
 and the action of the bilge-water and of wet ashes on the bottom and front of boilers is 
 easily prevented by stopping promptly all leaks, keeping the surfaces of the boiler 
 clean, and protecting it with a coat of paint. (See section 12, chapter xvii.) The sup- 
 ports on which the boiler rests must be arranged in such a manner that the bottom of 
 the boiler is easily accessible. External corrosion will also take place when boilers 
 rest directly on oaken keelsons, or when copper bolts are allowed to come in contact 
 with the iron of the shell of boilers. (See section 1, chapter xiv.) 
 
 The deterioration of the plates forming the heating-surfaces of boilers, produced by 
 overheating, variations and differences of temperature, the corrosive action of the gases 
 of combustion and of sulphuric acid distilled from soot, and the various causes of the 
 internal corrosion of boilers, will be discussed separately in the following sections of the 
 present chapter. Many of the causes of the deterioration of boilers may be prevented 
 by observing the directions for the management of boilers given in chapter xvii. The 
 internal corrosion of boilers is mainly due to the exposure of their damp surfaces to the 
 atmospheric air, the presence of chloride of magnesium and of fatty acids in the feed- 
 water, and the galvanic action of heterogeneous metals and especially of particles of 
 copper, introduced with the feed-water, in contact with iron. 
 
 In some parts of the boiler the iron plates may be raised to a sufficiently high tem- 
 perature to decompose the steam in contact with them, the oxygen combining with the 
 iron and forming, according to circumstances, some one of the oxides of iron. There 
 seems to be no doubt that this action frequently takes place in the uptake and in 
 superheaters, in which places the corrosion of the iron plates generally presents a very 
 different appearance from what it does at other parts of the boiler. Professor Hoff- 
 man concludes from direct experiments that the temperature of the iron has to ex- 
 ceed 356 Fahr. before decomposition of pure water takes place ; and Professor Barff 
 states that at a temperature of 650 Fahr. the black oxide of iron will be formed when 
 iron is exposed to the action of superheated steam. 
 
 The question whether wrought-iron or steel is more liable to corrosion has to be con- 
 
OF THE 
 
 UNIVERSITY 
 
 SBC. 1. CAUSES AND PREVENTION OF THE DETERIORATION OP BOILERS. 413 
 
 sidered as remaining undecided, the testimony in reference to this point being very con- 
 flicting. Steel plates having very nearly the same chemical composition, and exposed 
 in the same boiler to apparently identical influences, have shown in some cases very dif- 
 ferent results as regards corrosion. It is, however, generally believed at the present 
 day that the softer and purer kinds of steel and wrought-iron are more liable to cor- 
 rosion than the harder and less pure ones. 
 
 Corrosion does not attack the surfaces of iron and steel plates in a uniform manner, 
 and this fact is easily explained by the presence of structural differences in the plates. 
 Wrought-iron and steel cannot have a perfectly homogeneous structure from the very 
 nature of the processes of manufacture. The former is an aggregation of fibres welded 
 together by squeezing, hammering, and rolling, or separated by thin layers of impuri- 
 ties ; steel ingots are known to be traversed by innumerable air and gas cells, the walls 
 of which are more or less perfectly welded together by hammering and rolling. A 
 highly-polished plate of iron or steel exposed to atmospheric influences will show de- 
 tached spots of rust at first, which gradually enlarge and spread over the whole surface. 
 Powerful acids attack different parts of plates in a different degree, and thus develop 
 their irregular structure. 
 
 In steam boilers, however, several other circumstances combine to make corrosive 
 action very unequal. At places where the protective coating of the black oxide of iron 
 or of adhesive scale has been detached and the clean iron is left exposed, and where the 
 fibres of the metal have been loosened by bending, welding, etc., corrosion will com- 
 mence and make the most rapid progress. Corrosive substances may be deposited on 
 the plates in detached masses, and thus exert a powerful action. 
 
 Frequently the surfaces of plates exhibit detached cavities of small extent, but vary- 
 ing as to depth from a shallow depression to an actual perforation of the plate ; some- 
 times only a few of these cavities lying far apart occur in a plate ; at other times they 
 lie so close together that the plate presents a honeycombed appearance. This phe- 
 nomenon is called pitting. In the milder cases this form of corrosion may have been 
 caused by surface-defects due to cinder-spots, etc., but it is probably more frequently 
 produced by the intense local action of particles of corrosive substances, especially of 
 fatty acids, or by the galvanic action of particles of heterogeneous metals. 
 
 When the plates exhibit deep furrows following well-defined directions the phe- 
 nomenon is called grooving. These furrows are frequently found to run at a short dis- 
 tance from and parallel to seams, and their formation is explained by assuming that 
 the repeated bending of the plates at these places in alternate directions has gradually 
 detached the protective coating of scale and opened or broken the fibres of the metal, 
 
414 STEAM BOILERS. CHAP. XVIII. 
 
 and has thus facilitated corrosive action. Not unfrequently such a line of weakness is 
 produced by cutting into the surface of the plate in chipping the edge of the lap for the 
 purpose of calking it. 
 
 2. Deterioration caused by Overheating and by the corrosive Action of 
 the Gases of Combustion. Under the influence of the intense heat existing in the 
 furnace and combustion-chamber the metal is liable to rapid oxidation, especially in 
 laps and rivet-heads, where, in consequence of the greater thickness of metal, the diffe- 
 rence of temperatures at the water and fire sides is relatively great. The metal is also 
 liable to be destroyed by corrosive gases given off by the burning fuel. When sulphu- 
 rous fuel is burned the plates exposed to a high temperature are acted upon with great 
 rapidity, successive thin coats of bisulphuret of iron being formed on their surface. 
 
 In boilers worked with a strong artificial draught the cinders and fine particles of 
 coal carried along by the blast gradually wear away the metallic surfaces against which 
 they strike. This action has been observed especially in locomotives, where the front 
 tube-sheet has been found to be much injured by it, and where the use of copper tubes 
 had to be abandoned on that account. 
 
 The tubes and plates of superheaters located in the uptakes of boilers are liable to 
 become overheated, especially in starting the fires before steam has formed ; this may 
 be prevented by filling them with water, which is to be drained off after steam has 
 formed in the boiler. The upper part of the front of boilers in the uptake are exposed 
 to the same injury. The rules of the Board of Trade (English) require that the flat 
 ends of all boilers, as far as the steam-space extends, and the ends of superheaters 
 should be fitted with shield or baffle plates where exposed to the hot gases in the 
 uptake. 
 
 When parts of the boiler exposed to an intense heat as the furnace-crown, the top 
 of the back-connection and horizontal tubes are left bared of water the metal soon at- 
 tains a red heat. Iron tubes and plates collapse or bulge out and warp when they are 
 overheated, and their strength is frequently permanently impaired ; good, tough iron 
 becomes brittle and weak when it is burnt. Brass loses its strength and disintegrates 
 at a much lower temperature than that at which iron is injured. Horizontal fire-tubes 
 of brass are quickly burnt when they are left bare of water, but in vertical tubes the 
 water may be carried far below the upper tube-plate without injuring them. 
 
 The overheating of furnace-crowns and back-connections is frequently caused by the 
 accumulation of scale or the formation of a saponaceous sediment. (See section 4 of the 
 present chapter.) 
 
 Laminations existing in plates prevent the ready transmission of heat through them. 
 
SEC. 2. 
 
 CAUSES AND PREVENTION OF THE DETERIORATION OP BOILERS. 
 
 415 
 
 When such defects exist in plates forming the furnace or the combustion-chamber the 
 layer in the plate nearest to the fire becomes greatly overheated, and, expanding, bulges 
 out. When the extension of the metal has exceeded the limit of elasticity the metal 
 does not resume its original shape after the heat has been discontinued, and a blister is 
 formed, which increases in size with repeated applications of heat, and is sooner or later 
 fractured, either at the apex when its thickness is uniform, or near the edge when it is 
 thinnest there. 
 
 When the temperature of the plate exceeds a certain limit the phenomenon of the 
 spheroidal condition of water is produced. According to Boutigny, this may take 
 place when the temperature of an iron plate is as low as 350 Fahr. When the shape 
 or position of the heating-surface prevents the ready escape of the steam from it, as in 
 horizontal water-tubes, or when the fires are urged much in a boiler with defective cir- 
 culation of the water, or when the thick scale which has accumulated on a heating- 
 surface cracks and becomes detached by the expansion of the overheated plate, the tem- 
 perature of the latter may have become high enough to produce the spheroidal condi- 
 tion of the water ; and although the repulsion between the water and the plate may con- 
 tinue only for a few seconds, this time may be sufficient, with an intense heat, to soften 
 the iron so that it is forced outward by the steam-pressure. In the depression thus 
 formed sediment is very apt to lodge and accumulate, favoring a repetition of the pro- 
 cess, and by alternate heating and cooling this part of the plate will be either cracked 
 or burnt out. 
 
 
 Relative tenacity. 
 
 Temperature. 
 
 Experiments made by 
 committee of Franklin 
 Institute, 1833-33. 
 
 Experiments made by Kollmann, 1877-78. 
 
 Degrees Fahr. 
 
 Wro ught-iron. 
 
 Fibrous 
 wrought-iron. 
 
 Fine-grained 
 wrought-iron. 
 
 Bessemer steeL 
 
 32 
 
 100 
 
 IOO 
 
 IOO 
 
 IOO 
 
 212 
 
 .... 
 
 IOO 
 
 IOO 
 
 IOO 
 
 392 
 
 93-8 
 
 95 
 
 IOO 
 
 IOO 
 
 572 
 
 91.7 
 
 90 
 
 97 
 
 94 
 
 932 
 
 66.8 
 
 38 
 
 44 
 
 34 
 
 1292 
 
 30.0 
 
 16 
 
 23 
 
 18 
 
 1652 
 
 
 
 6 
 
 12 
 
 9 
 
 I8 3 2 
 
 .... 
 
 4 
 
 7 
 
 7 
 
 Experiments on the tenacity of wrought-iron and steel at different temperatures, 
 made by Kollmann in 1877-78, give much lower results for the tenacity of these metals 
 at temperatures exceeding 570 Fahr. than the experiments made by a committee of 
 
416 STEAM BOILERS. CHAP. XVIII. 
 
 the Franklin Institute in 1832-33. The foregoing table, showing the relative tenacities 
 of wrought-iron and steel at different temperatures, gives a brief summary of the re- 
 sults of these experiments. 
 
 3. Strains produced by sudden Variations and great Differences of Tem- 
 perature. If the ends of a plate are rigidly fixed so that it is incapable of altering 
 its length, while at the same time it cannot bend sideways, an increase in temperature 
 of 1 Fahr. will subject it to a compressive stress of about 150 Ibs. per square inch, and 
 a decrease of 1 Fahr. will produce under the same conditions a tensile stress of equal 
 intensity ; and these stresses are totally independent of the sectional area of the plate. 
 
 When steel boiler-plates were first introduced great difficulty was experienced from 
 the unequal expansion of plates which had left the rolls at different temperatures. The 
 coldest-rolled plates expanded most on being reheated the first time ; and, consequently, 
 when plates that had left the rolls at different temperatiares were riveted together there 
 was a constant strain on the joint, which often resulted in the fracture of the plate. 
 This difficulty was overcome by annealing all the plates after they left the rolls. 
 
 The cracks which are frequently found in the lap-joints of furnaces, extending either 
 from the rivet-holes to the edge of the plate or from hole to hole in the seam, are pro- 
 bably in most cases caused by the unequal heating and sudden cooling of the plates. 
 The double thickness of plates at the lap in contact with the fire causes their tempera- 
 ture to be greater than where only a single thickness exists, and a corresponding ten- 
 dency to expand. When a current of cold air rushes in thrcragh the open furnace-door 
 and impinges against the heated and expanded joint, it cools off the outer plate sud- 
 denly, producing a tendency to contract. But the inner plate of the lap still retains its 
 original temperature and resists contraction, thus throwing a sudden tensile strain on 
 the outside plate, which, if sufficiently severe or often repeated, will produce fracture, 
 especially as the iron around the rivet-holes is frequently already injured by punching 
 or drifting. 
 
 Long furnace-flues should be allowed to accommodate themselves to the expansion 
 and contraction due to the varying temperatures either by applying the Bowling hoop 
 to them (see section 6, chapter ix.) or by turning the flanges which secure them with a 
 large radius. If this is not done they are apt to crack through the bend of the flanges, 
 especially when they are secured by angle-irons, or the end plates of the boiler to which 
 they are attached have to buckle with the longitudinal expansion of the flues, and the 
 constant repetition of this movement inevitably results in the destruction of the plate ; 
 the more rigidly the plates are stayed the more severe is this strain on them. 
 
 A great difference of temperature exists often in the upper and lower half of furnace- 
 
SEC. 3. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 417 
 
 flues, especially when the feed- water is cold. The resulting unequal expansion throws 
 severe strains on the longitudinal and transverse joints, besides weakening flues by dis- 
 torting their circular cross-section. 
 
 These strains are very severe in the case of the shell of cylindrical boilers of large 
 diameter, the bottom of which has the temperature of the feed-water while the upper 
 portions have the temperature of the steam. The difference of these temperatures may 
 amount to more than 200 Fahr. The effect of these strains is that the circumferential 
 seams at the bottom are frequently leaky, while the longitudinal seams are generally 
 tight. Many boilermakers double-rivet the circumferential joints on this account, 
 although single-riveting would be sufficient to resist the strains produced by cold-water 
 pressure. In several double-end boilers these strains have caused the fracture of the 
 plate between the rivet-holes in the circumferential seams. On account of these strains 
 the shell of such boilers should be made of a soft, ductile iron. The difference of tem- 
 perature should be reduced as much as possible by facilitating the circulation of the 
 water and by increasing the temperature of the feed- water. 
 
 The general rules and regulations prescribed by the Board of Supervising Inspectors 
 of Steam -vessels (1879) provide that "the feed- water shall not be admitted into any 
 boiler, on board of any steam- vessel subject to the jurisdiction of this Board, at a less 
 temperature than one hundred (100) degrees Fahrenheit for low-pressure boilers, and 
 one hundred and eighty (180) degrees Fahrenheit for high-pressure boilers ; nor shall 
 cold water be admitted into any such boiler while the water is at a less temperature than, 
 the sumranding atmosphere." Boilers carrying a steam-pressure exceeding 60 pounds 
 to the square inch are to be considered as high-pressure boilers. 
 
 Plate XXXVI. represents some specimens of rivets taken from the bottom of a cir- 
 cumferential seam of an externally-fired, cylindrical flue-boiler, abont 4 feet in diameter. 
 The seam was situated in the furnace near the bridge-wall, and had repeatedly given 
 trouble by leaking. The expansion and contraction of the lap had evidently caused the 
 fracture of the rivets, gradually detaching the conical heads from the shanks, so that 
 they were held in many cases only by a few fibres. The corrosion of the extremity 
 of the detached shanks indicates that this action had extended over a considerable 
 length of tune. The unequal strains thrown on the rivets had been intensified by the 
 leverage of the rivet-heads, and the fracture of the fibres, commencing at the circum- 
 ference of the shank, had taken place, as it gradually extended toward the centre, in 
 nearly every case at the same distance from the outer surface of the conical head, so 
 that the detached heads were concave at the side where fracture had taken place. This 
 uniformity in the appearance of the fractures seems to indicate that the iron of the rivets 
 
418 STEAM BOILERS. CHAP. XVIII. 
 
 was made brittle by hammering after the rivets got cold, the injury extending to a 
 nearly uniform depth from the surface. 
 
 In locomotive boilers the difference of the temperatures of the flat sides of the shell 
 and of the fire-box, which are tied rigidly together by closely-spaced stays, produces 
 often very destructive strains. Since the top and side plates of the fire-box are not at 
 liberty to expand freely, they pucker at the ends, causing the joints to leak and often 
 fracturing the plates or the stay-bolts. The tube-sheet is likewise distorted, the outer 
 rows of the tube-holes become oval, and the tube-ends either crack or leak. These evil 
 effects may be greatly lessened by turning the flanges with a large radius, which allows 
 the plates to accommodate themselves to the varying movement of expansion and 
 contraction. It has also been proposed to use flexible stays instead of screw-stays near 
 the extremities of the side and front plates of the furnace. 
 
 4. Formation of certain Saponaceous Deposits in Land Boilers. In an 
 article by M. Maurice Jourdain, in the first report of the Parisian Association of the 
 Owners of Steam-apparatus, an account is given of the formation of a peculiar deposit 
 in land boilers under certain conditions. When boilers are fed with water containing 
 greasy matter certain particular circumstances, still imperfectly defined, cause a light 
 grayish, pulverable deposit to cover the iron directly exposed to the fire. This powder, 
 which, according to chemical analysis, is composed principally of lime-salts and 
 magnesia and greasy matter, possesses the peculiar property of being impervious to 
 water. 
 
 " It is very easy," says M. Jourdain, "to understand the effect produced by this 
 lack of permeation. The water, running along the plate without wetting it, is main- 
 tained in a spheroidal state. Under these conditions the iron plate can be highly 
 heated without communicating any sensible amount of heat to the water, which covers 
 without touching it. This state continues until the temperature of the iron is suffi- 
 ciently high to burn out the lime-soap which overlays it. At that moment the iron 
 plate, suddenly uncovered, is brought into contact with the water, which causes a 
 partial explosion and a sudden cooling of the metal. This suffices to deteriorate the 
 boiler. 
 
 "In fact, the deteriorations observed in the cases where that gray powder existed 
 have always been very serious." 
 
 In 1864 six boilers erected in the iron-works of Mr. Borsig, in Silesia, showed, 
 during the first days of use, leaks from the joints of the iron plates exposed to the 
 fire, which daily increased. When the boilers were stopped a great many cracks were 
 discovered extending over a large surface, besides loose rivets, blisters, etc. Three 
 
SBC. 5. CAUSES AND PREVENTION OP THE DETERIORATION OP BOILERS. 419 
 
 
 
 other boilers put in operation to replace the former ones did not give any better result. 
 After forty-eight hours of use the effect of irregular heating or overheating of the iron 
 was indicated by the escape of water and other phenomena. Once, while one of these 
 boilers was in operation, a rumbling noise was heard, followed by a detonation accom- 
 panied by a violent jerking of the boiler, as if it were at the point of exploding. These 
 phenomena ceased after precautions had been taken to prevent the presence of greasy 
 matter in the feed-water. 
 
 Several other cases are described where equally destructive effects were produced by 
 the formation of the powdery substance, which always contained carbonate of lime and 
 magnesia mixed or combined with greasy matter. 
 
 In discussing these observations M. Delaunay states that he found that the soap 
 formed by the combination of carbonate of lime with fatty matter does not adhere to 
 the plates of a boiler, and is permeable ; and he concludes, therefore, that the presence 
 of magnesia may be essential to give to this substance the peculiar character which pro- 
 duces such destructive effects on boilers. He continues then: "A final peculiarity 
 which we can select from the detailed reports made on various experiments where the 
 said phenomena have occurred is that the feed-water which caused them produced 
 always but little incrustation. 
 
 "It is comprehensible, in fact, that, since the greasy matter contained in the con- 
 densed water is always present in small quantities, its isolating action cannot become 
 manifest when it is surrounded, so to speak, by a mass of chalky matter, the precipi- 
 tation of which operates iinceasingly, adding at each instant new layers of deposit to the 
 old ones. The conditions which seem to determine the production of the phenomena of 
 isolation and of the spheroidal state, the consequences of which we have described, are, 
 therefore : 1st, the relative purity of the water ; 2d, the presence of magnesia in the 
 water. Let us, however, remark that these conclusions are only probable, and cannot 
 be considered as scientifically demonstrated." (L. Delaunay, ' Etudes sur les Geriera- 
 teurs d Vapeur d Haute Pression.') 
 
 5. Corrosion of Steam Boilers by Sulphuric Acid present In the Soot. 
 Several years back the attention of French engineers was directed to the rapid corrosion 
 of land boilers on the exterior surfaces of plates where soot was allowed to accumulate. 
 Several such cases are described in an article in the ' Annales des Mines et des Fonts et 
 Chaussees,' 1876. 
 
 In one case the iron was reduced in thickness from 0.47 inch to 0.067 inch in five 
 years. The corrosion, which was wholly on the outside, was attributed by Mr. Dou- 
 ville. Mining Engineer, to the action of the oxygen and sulphurous acid in the gases of 
 
420 STEAM BOILERS. CHAP. XVIII. 
 
 combustion in the presence of water. He took large scales of the oxide of iron from 
 the corroded parts, and he found therein sulphur, but was not able to determine its 
 state of combination. 
 
 In another similar case " two specimens of the deposits left by the smoke on the in- 
 jured iron have been analyzed ; they gave between 52 and 53 per centum of sulphate of 
 iron. One gave 1.42 per centum of free sulphuric acid, the other gave 12 per centum 
 nearly. The deposits formed on the rest of the boiler also contained sulphuric acid, but 
 in notably less quantity, and no sensible deterioration of the metal had resulted 
 from it." 
 
 Some examples of exterior corrosion in consequence of the condensation of the 
 aqueous vapor in the smoke on the cold parts of boilers have been pointed out by Meu- 
 nier-Dollfus, Director of the Alsatian Association of Steam Boilers. Respecting a case 
 described by him it is stated that the corrosion was principally on the cold or but little 
 warmed portions of the heaters, and had for first cause the sulphurous acid dissolved in 
 the water of condensation deposited from the smoke. It was ascertained that in the 
 presence of air and of this watery acid there was first oxidation of the iron and then 
 formation of the sulphate of the oxide of iron. 
 
 The following conclusions are drawn from the investigation of the cases considered : 
 "When the smoke-deposits on boiler-surfaces distant from the furnace are rendered 
 moist by any accidental cause, the sulphurous acid in the gases of combustion deter- 
 mines the attack upon the metal by the formation of the sulphate of the oxide of 
 iron. 
 
 " The attack can take place, while the boiler is in use, on such of its metallic surfaces 
 as may be wetted by leakage from the boiler itself, or by water infiltrated through, the 
 masonry or derived from the condensation of the aqueous vapor in the gases of com- 
 bustion by contact with surfaces relatively cold. It can also be produced, while the 
 boiler is out of use, by means of the humidity of the air in the flues." (See Journal of 
 the Franklin Institute, November, 1877.) 
 
 The following is an extract from an essay on " The Acid Products of the Combustion 
 of Coal," by M. Vincotte, translated in the Journal of the Franklin Institute, March, 
 1880: 
 
 "The gases of combustion deposit on the heating-surfaces of boilers different sub- 
 stances of great importance as regards the durability of the metal composing those sur- 
 faces and its power of transmitting heat. These substances are principally soot, tarry 
 matter, sulphuric acid, and ammoniacal salts. 
 
 " The quantity of soot in the gases depends on the kind of coal consumed and on 
 
SEC. 5. CAUSES AND 1'REVENTION OF THE DETERIORATION OP BOILERS. 421 
 
 tlie intensity of the chimney-draught ; but whether great or small, a portion is always- 
 deposited on the heating-surfaces. 
 
 " The soot in immediate contact with the surfaces is cooled by them below the tem- 
 perature of combustion ; but when the deposit attains a certain thickness the portion 
 most distant from the surfaces burns whenever the hot gases of combustion passing 
 over it contain sufficient free oxygen. The thickness of the soot-deposit which escapes 
 combustion depends on the temperature of these gases at any particular point con- 
 sidered, and should, therefore, after the boiler has been some time in use, be found to 
 increase from the furnace to the chimney. 
 
 " On all the heating-surfaces where the gases of combustion have a sufficiently ele- 
 vated temperature to burn the outer portion of the soot-deposit- the unburnt inner por- 
 tion is found covered with a white layer of ash from the burnt portion, and this ash 
 has a thickness limited only by its cohesion ; generally it increases until the surfaces 
 are swept. 
 
 " On the heating-surfaces where the temperature of the gases of combustion is too 
 low to ignite the soot the latter remains black, and its thickness continually increases 
 until it falls off by its own weight, which does not happen soon." 
 
 After giving the analysis of several specimens of soot-deposit taken from different 
 boilers and heaters, which contained in nearly every case various quantities of ferric 
 sulphate, ferrous sulphate, and free sulphuric acid, he continues : 
 
 " There does not appear a satisfactory theory of the formation of the sulphuric acid. 
 It may, indeed, be said that all the coals burned in boilers contain sulphur, whose com- 
 bustion would naturally produce sulphurous acid, which encounters amid the gases of 
 combustion free oxygen, aqueous vapor, and other substances necessary to its trans- 
 formation into sulphuric acid ; but what those substances are or how they react is un- 
 known. Be that as it may be, it is found on all the heating-surfaces not coated with 
 pitch, and forms there, in immediate contact with the iron beneath the soot, a very thin 
 layer of ferric sulphates, with which it remains mixed. 
 
 "The quantity of acid thus found on the heating-surfaces is so much the greater as 
 their temperature is lower. It increases from the furnace to the chimney, and when 
 the boilers are fitted with feed-water heaters the acid is most abundant on them, where 
 it is found not only in contact with the metal, but throughout the whole layer of soot 
 there, which is impregnated with it. ... 
 
 ' To sum up, when a boiler is heated with semi-fat coal, and examined after the fire 
 is withdrawn, the heating-surfaces are found covered with the following products of the 
 combustion : 
 
422 STEAM BOILERS. CHAP. XVIII. 
 
 "1. Above the fire is a thin layer of very dry pitch mixed with soot and ordinarily a 
 little acid and astringent ; if, however, portions of the surface have been overheated the 
 pitch will have disappeared from them. Upon the pitch is some clinker, derived from 
 the ash mechanically carried up, which, being intercepted by the rivet-heads and other 
 projections, is partially fused there. 
 
 "2. Farther on are found three very distinct layers. The first, in immediate contact 
 with the iron, is thin and white ; it is composed of from 80 to 90 per centum of ferrous 
 and ferric sulphates, 2 to 3 per centum of free sulphuric acid, a small quantity of 
 ammoniacal salts, sulphate of lime, etc. Under this layer the iron is white and clean ; 
 sometimes it is brilliant. 
 
 "The next layer is black soot mixed with a little acid, and with salts of iron and 
 ferric oxide, derived probably from anterior decompositions. This layer of black soot 
 increases in thickness as we proceed towards the chimney. 
 
 "Lastly, the outer layer is composed of a white or pinkish substance, very soft, 
 very adhesive, and very dry, composed of alumina, silica, and sulphate of lime, and, 
 when of great thickness, its external portion sometimes melts and forms small, greenish, 
 vitreous grains. Ordinarily the layer is compact, but sometimes when very fat coal is 
 used it is flaky. 
 
 "3. Still farther towards the chimney the outer layer diminishes in thickness and 
 disappears after having become grayish. There is then found only the white and acid 
 substance in contact with the iron, and the black soot, which sometimes becomes from f 
 to of an inch thick when not removed by sweeping. 
 
 "With fat coal the difference is that the layer of pitch extends farther at the ex- 
 pense of the surface covered by the acid. 
 
 "As soon as the boiler is put out of use the different products just enumerated 
 become rapidly modified, and an examination finds them under very varied forms. 
 
 "The sulphuric acid quickly attracts the surrounding humidity, becomes diluted, 
 and penetrates by imbibition into the soot which absorbs it. 
 
 "The mixture of ferrous and ferric salts acts in an analogous way, depending, how- 
 ever, in a very notable manner on its composition, whether ferrous or ferric. When 
 there is only a little ferrous sulphate it attracts enough humidity to furnish its water of 
 crystallization, and crystallizes into a very dry substance, which is only transformed 
 slowly, and has not, under ordinary circumstances, further action on the metal. The 
 ferric salts, on the contrary, when in certain quantity, are deliquescent or not, accord- 
 ing to the hygrometric state of the flues. I have had specimens which in one corner of 
 my laboratory were deliquescent and in another warmer corner were not. 
 
SBC. 6. CAUSES AND PREVENTION OP THE DETERIORATION OF BOILEB& 423 
 
 "The soot, when sufficiently moistened, diminishes greatly in volume. It often be- 
 comes loosened, twists like a peeling, and hangs from the heating-surfaces in tatters, of 
 which a large portion falls at the first accession of heat. 
 
 "The acid, once slightly diluted, attacks the iron. The ferric salts act in the same 
 manner, and after a time, depending upon the humidity of the flues, the white layer in 
 contact with the iron grows yellowish, and then, little by little, changes into a thick 
 layer of rust impregnated with sulphate of iron. It is by the appearance of this layer, 
 after the boiler has been some days out of use, that we ascertain whether the flues are 
 humid. If it grows yellowish after two or three weeks we can conclude that the thick- 
 ness of the metal is rapidly undergoing a general diminution, and this corrosion is due 
 to humidity, whose cause should be discovered and removed. 
 
 " As regards the preservation of boilers, all the preceding is summed in some leading 
 facts : 
 
 "First. Nearly all the heating-surfaces of a boiler are covered with sulphuric acid, 
 but this acid attacks them only in an insensible manner. Boilers forty years old are 
 met with whose heating-surfaces have always been covered with acid without diminish- 
 ing their thickness ^ of an inch. 
 
 "Second. As soon as this acid in any way acquires humidity it becomes very corro- 
 sive and rapidly pits the iron. 
 
 " TJiird. An examination of the flues shows at a glance, simply by the appearance 
 of the products, if humidity be present. If it be, the cause must at once be discovered 
 and the effects ascertained.'' 
 
 6. Corrosion due to the Presence of Oxygen and Carbonic Acid in 
 Water. The most common cause of the corrosion of boilers is the exposure of unpro- 
 tected iron surfaces to the combined action of the atmospheric air and of water. The air 
 present in the feed- water and the natural moisture of the atmosphere are sufficient to 
 produce rusting / but this action is more intense when, after the fires are hauled, the 
 boilers are only partly emptied and are left standing for some time in a damp condition, 
 or when leaks keep the outside of boilers wet. 
 
 It has been asserted by some engineers that distilled sea-water has a peculiarly 
 destructive effect on iron. The corrosive action of such distilled water is, however, to 
 be ascribed to the presence of some corrosive substance carried over during the process 
 of distillation, or of atmospheric air, oxygen, or carbonic acid ; because perfectly pure 
 water does not produce corrosion in iron at ordinary temperatures. The carbonic acid 
 may either be originally present in the free state or may be produced by the decompo- 
 sition of some carbonate. 
 
STEAM BOILERS. CHAP. XM 11. 
 
 The following aocotant is given by M. Cornut, Chief Engineer of the '"Association du 
 Nord," of experiments made by Scheurer-Kestner and Meunier-Bollfus to test the ac- 
 tion of different waters on iron. 
 
 They took three flasks, eaeh holding ten litres, and tilled them with water. 
 
 The//-.*/ flask contained water free of lime-salts but highly aerated. 
 
 The aftvnd flask contained water holding lime-salts in solution, and likewise air and 
 carbonic acid. 
 
 The M//W flask contained distilled water from which the free oxygen had been re- 
 moved by boiling. 
 
 In each flask a well-cleaned and polished bar of iron was placed. 
 
 All ntHvssjirypiwautionswe.lv taken to ensure exact ness in these experiments, which 
 continued several weeks. 
 
 Tho.tf/\s7 flask was the first to show signs of oxidation. Yellow streaks soon made 
 their appearance- in the water and the bar became gradually covered with pustules of 
 rust. After all the o\\gen of the air had been consumed the phenomena of oxidation 
 ceased ; and the bar, having Ixvn taken out of the flask and cleaned Jind then put back. 
 remained as bright as if it had been varnished. 
 
 The ,m>//</ flask showed the same phenomena of oxidation, but they appeared more 
 slowly, and the yellow streaks were intermixed with white streaks formed by lime-salts 
 which were precipitated. After the lapse of some time the iron bar was taken from the 
 flask, cleaned and put back, and again Ixrjune covered with a layer of oxide. Conse- 
 quent ly the lime-salts deposited on the metal interfered with and retarded its oxidation. 
 
 The Mini flask showed no sign of oxidation ; the Ivir remained bright. 
 
 Kxperiments on the oxidation of iron by Professor K. rrau-Calvert, of Manchester 
 v - Memoirs of the Philosophical Society of Manchester.' ."nh vol., Tnh series", proved that 
 no oxidation takes place in iron immersed in dry oxygen or in pure and dry carbonic 
 acid. Damp oxygen acts feebly, damp carbonic acid not at all. A mixture of oxygen 
 and carKmie acid in a dry state does not oxidi/e the iron : a mixture of the two gases in 
 a damp state produces a very rapid oxidation of the iron. First peroxide of iron is 
 formed, then carlxwate, and finally a mixture of oxide and hydrate of the sesquioxide. 
 An iron i\xl plunged into a bottle filled with the ordinary water of the city of Manches- 
 ter, containing in solution oxygen and carbonic acid, was covered with rust at once. 
 An iron plate, immersed in a bottle filled with the same water which had been boiled so 
 as to remove the oxygen and carbonic acid, showed no trace of oxidation after several 
 
 The k Third Report of the Admiralty Committee on Boilers' contains the following 
 
Sc. 7. CAUSES AND PREVENTION OP THE DETEIIIOHATION OP BOILERS. 425 
 
 to tin- ilc.strurt jvi- action o|' water in combination with oxyir'-n ari'l car- 
 bonic acid on boilers : 
 
 " The water with which the boilers are filled, and likewise that which is supplied to 
 make up the quantity changed by blowing-off and the unavoidable waste of steam, con- 
 tains air and carbonic acid in solution ; the air dissolved by water contains a la'rger pro- 
 portion of oxygen than ordinary atmospheric air, and it is this oxygen which contri- 
 butes so greatly to the corrosion of iron." 
 
 From a series of experiments, carried out by the committee to illustrate this action of 
 oxygen upon iron immersed in water under different circumstances at the ordinary tem- 
 peratures, it appears that pure distilled water perfectly free from solid matter allows of 
 more corrosion than sea-water. 
 
 " Mr. Lant Carpenter gives as the mean of thirty analyses the following relative pro- 
 portions of oxygen, nitrogen, and carbonic acid in 100 volumes of the gases dissolved in 
 .surface sea -\\-aicr: Oxygen, 25.1; nitrogen, 54.2; carbonic acid, 20.7; the average pro- 
 portion of these mixed gases amounts to 2.8 volumes in 100 volumes of water ; and in 
 order to avoid the corrosion due to the presence of these substances, and also the varia- 
 ble conditions of density and scale, the discontinuance of blowing-off, together with the 
 substitution of distilled sea-water (where fresh cannot be carried) for sea-feed to make 
 up waste, would be advisable. 
 
 " Perfectly dry air has no action upon compact iron at the ordinary temperature. 
 On the other hand, water perfectly free from air is also without action upon iron at the 
 ordinary temperature. Water in the state of steam, when passed over iron heated to a 
 sufficiently high temperature, does oxidize it. ... 
 
 " Much misapprehension appears to exist with reference to the action of 'pure water ' 
 upon iron, as, after what has been stated, it will be seen that the corrosion or oxidation 
 which has teen described by many of the witnesses and others to the action of pure 
 waterier se should properly be attributed to the oxygen contained in the air dissolved 
 by such water, the water acting as a means of transfer for the oxygen to the iron 
 a property which it would possess to an unlimited extent." 
 
 7. Corrosive Action of the Chloride of Magnesium. The affinity of magne- 
 sium for chlorine is so weak that even at the temperature of boiling water the chlorine 
 decomposes the water, forming hydrochloric acid and magnesia according to the fol- 
 lowing equation : 
 
 MgCl, + H,O = MgO + 2HC1. 
 The committee appointed by the British Admiralty to investigate the deterioration of 
 
426 STEAM BOILERS. CHAP. XVIH. 
 
 boilers made experiments from which they inferred that this decomposition takes place 
 much more slowly than has frequently been believed to be the case. 
 
 " Some chloride of magnesium was dissolved in water and evaporated to dryness ; 
 the process was repeated (say 20 times) in order to see if the chloride would be de- 
 composed completely. At the end of the experiment the residue (after heating 
 strongly) was digested with water, filtered, and a solution of nitrate of silver added : 
 an abundant precipitate of chloride of silver showed that the decomposition was very 
 incomplete. 
 
 "It was ascertained that so long as the contents of the dish were fluid, or even 
 after solidification commenced, the water evaporated showed no signs of acidity, but 
 that, when the last portions of the water were passing off, very marked acid reaction 
 took place, and the pungent vapor of hydrochloric acid was distinctly perceptible. 
 
 "From these experiments it may be inferred that no decomposition of chloride of 
 magnesium occurs during the simple ebullition and concentration of sea-water (as had 
 been suggested), and that the water distilled from sea- water, when carefully done, 
 contains nothing more than that distilled from fresh or land water. When the distilla- 
 tion is conducted on a large scale and rapidly, varying quantities of the saline contents 
 may be carried over mechanically ; but the evil effects attributed to the use of distilled 
 sea- water are doubtless due to other causes. 
 
 " In the presence of iron, as will be shown further on, the stability of chloride of 
 magnesium is affected, with the production of proto-chloride of iron (ferrous chloride) ; 
 but the constant though feeble alkalinity of sea-water precludes the action, or even 
 the existence, of hydrochloric acid in the free state. 
 
 "The instability of this salt [chloride of magnesium] in the presence of iron, even at 
 the boiling-point of water, and in the absence of air, as pointed out by Professor Wag- 
 ner (Dingier' s PolytechniscTies Journal, October, 1875), would be considerably increased 
 at the higher temperatures at present realized in marine boilers. . . . 
 
 "In one of the experimental tubes at Sheerness (A No. 21) there were dissolved 
 3,000 grains of chloride of magnesium in rain-water ; after six months' work the rod in 
 the tube was found to be corroded somewhat irregularly from below upwards, which is 
 contrary to the usual direction. . . . 
 
 "The water in this tube was examined, and found to contain proto-chloride of iron 
 in solution, while at the bottom of the tube there was a deposit of red oxide of iron 
 which contained small quantities of magnesia ; now, proto-chloride of iron and hydrate 
 of magnesia react upon each other, with the reproduction of chloride of mag- 
 nesium and oxide of iron, which is gradually converted into the red oxide by the 
 
SEC. 8. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 427 
 
 
 
 action of air ; the proportion of iron in solution was equal to 6.37 grains in one gallon. 
 A continuous reaction such as this goes far to explain why marine boilers raising steam 
 from sea-water suffer so much more from corrosion than boilers on shore fed with fresh 
 water. 
 
 "Another source of corrosion connected with the presence of chloride of mag- 
 nesium is due to the hygroscopic nature of this salt. When chloride of magnesium 
 is present in the deposits formed in the boiler it absorbs moisture eagerly from the 
 atmosphere, and thus offers the means of transferring oxygen to unprotected iron 
 surfaces. 
 
 " The phenomenon known familiarly to engineers as sweating proceeds from the same 
 cause, and may always be seen in the form of globules on the surfaces of old boilers in 
 damp weather. These globules are acid to test-paper, and contain iron in solution in 
 the state of protoxide ; on exposure to the air they become covered with a thin film of 
 peroxide of iron ; and if in that condition the drops are gradually dried up the film 
 (hollow) remains, and if the surface of the iron be kept quite dry decay ceases, but 
 recommences when the temperature of the iron falls to that of the atmosphere and 
 allows of fresh moisture being deposited." (Admiralty Committee on Boilers, Third 
 Report.} 
 
 8. Corrosive Action of Fatty Acids. "The following is an explanation of the 
 changes which a fatty body undergoes in presence of certain calcium and magnesium 
 salts, especially the carbonates. When fatty matters are heated to a temperature of 
 about 140 or 160 Fahr. with calcium carbonate, they form a kind of insoluble lime- 
 soap, which adheres to the sides of the containing vessel. As the temperature rises, 
 however, this lime-soap is decomposed into free fatty acid Tisually oleic acid and a 
 readily-decomposable variety of basic lime-soap. These two substances at once seize 
 hold of the iron and dissolve it. The presence of the grease is not less destructive even 
 if the proportion of calcium and magnesium salts is very low, as, under great pressure, 
 a trilling amount of lime-salts will suffice to determine the re-solution of a neutral fat 
 into free acid and glycerine." (G. ft. Tweedie, in "Iron," Sept. 21, 1878.) 
 
 Such a decomposition is, in fact, effected by water alone when it is made to act upon 
 fat at high temperatures and pressures, and this process, called water-saponification, has 
 been made use of to separate stearic acid from tallow for the purpose of manufacturing 
 hard candles. The temperature at which this water- saponification takes place is, accord- 
 ing to some authorities, about 400 Fahr., while others assert that even a temperature 
 of about 270 Fahr. would be sufficient for the process. In several instances it has been 
 found that the iron boilers used in this process of water-saponification for the manufac- 
 
428 STEAM BOILERS. CHAP. XVIII. 
 
 % 
 
 ture of candles corroded rapidly and in a very irregular manner, the plates being deeply 
 pitted, especially in the neighborhood of the water-level. 
 
 The corrosive influence of water and fatty acids upon iron at ordinary boiler tempe- 
 ratures was tested in 1864 by Professor A. W. Hofmann, of the Royal College of 
 Chemistry, by direct experiments which are described by him in a letter to Messrs. 
 Humphrys & Co., appended to the 'Eeport of the Admiralty Committee on Boilers.' 
 He says : 
 
 " Iron tubes containing rods of different varieties of iron, and fitted with hermeti- 
 cally-closing caps, were charged with water and stearic acid, the latter having been 
 separated from tallow by the ordinary process of lime-saponification. These tubes were 
 exposed for three weeks to a temperature of from 264 to 285 Fahr., corresponding to a 
 pressure of from 2f to 3 atmospheres. On opening them the inner surface of the tubes, 
 as well as the iron rods, proved to be powerfully corroded ; a large proportion of oxide 
 of iron was found in conjunction, and apparently in combination, with the fatty acid 
 floating in the liquid. Under the circumstances described this eifect can have taken 
 place only by an actual decomposition of water. Indeed, when the experiment was 
 repeated with iron rods enclosed with water and fatty acids in a glass tube the ends of 
 which had been drawn out and sealed before the blow-pipe, it was found that after six 
 hours' exposure to a temperature of 356 Fahr., corresponding to a pressure of ten 
 atmospheres, very appreciable quantities of hydrogen had been evolved. When the 
 drawn-out point of the tube was softened in the blow-pipe flame the compressed gas 
 forced its way through the glass and exploded with -a loud sound. To ascertain the 
 dependency of this effect, which was observed in several consecutive experiments, on the 
 fatty acid itself, and not on traces of mineral acids possibly contained and masked 
 therein, the experiment was repeated at the reduced temperature of 212 Fahr., when 
 no hydrogen was found to be evolved. Had a trace of free mineral acid been present 
 hydrogen would have inevitably been disengaged, even at the temperature of boiling 
 water. It was, moreover, established by careful experiments that iron, when heated 
 with water alone to 356 Fahr., yielded no trace of hydrogen. 
 
 " The rapidity of the corrosive action of fatty substances, in the presence of water, 
 upon iron considerably increases with the augmentation of temperature, but even at 
 ordinary atmospheric temperatures it takes place in a lesser degree." 
 
 The iron-soap formed by the action of fatty acids on iron is decomposed again by 
 heat into oxide of iron and free fatty acid. A series of observations and experiments 
 made on this subject by Professor V. Wartha, of Buda-Pesth, are recorded in Dingier 's 
 Polytechnisches Journal. Wartha was led to this study by the analysis of deposits 
 
SEC. 8. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 429 
 
 found on the damaged parts of the heater of a steam-engine. The analysis of that de- 
 posit gave him an oleate of the oxide of iron and free oleic acid. 
 
 "I made then by synthesis," says Prof. Wartha, "a comparative trial with free 
 oleic acid, which, after having remained a long time in contact with the air, turned a 
 yellowish brown. I then took a few cubic centimetres of that fatty acid, added water 
 and mixed it with iron filings, then heated it, when the mass began to swell and boil 
 and disengaged great quantities of hydrogen, and a glutinous mass of oleate of iron 
 remained, brown in color and soluble in ether. This mass contained 11 per cent, of oxide 
 of iron, and was in every respect like the substance found in the heater. 
 
 "The explanation of the reactions which I have just described is very simple. In 
 the factory spoken of the exhaust-steam is used to heat the feed- water. The fatty acid 
 that is to say, the oleic acid formed by contact with steam and under the influence 
 of heat, is carried off with the condensed steam, and arrives thus in the heater. In that 
 apparatus the drops of condensed oil stick to the sides as a pasty mass, and under the 
 influence of heat the iron is then attacked at the point of contact. 
 
 "Under the influence of pressure, and with the aid of hot water, oxide of iron and 
 free fatty acid are continually formed, and the latter combines with a fresh portion of 
 iron to form oleate of iron ; so that the drop eats away the iron, and burrows, so to 
 speak, in the metal. After a very short time the metal plates were thus perforated 
 and the apparatus commenced to leak. 
 
 "By the preceding explanation it is easy to account for the manner in which a com- 
 paratively small quantity of oleic acid can, in a very short time, perforate an iron plate 
 7 millimetres (0.276 inch) thick." 
 
 Regarding the action of fatty substances upon copper Professor Hofmann speaks as 
 follows in the letter referred to above : 
 
 "Under like influences the behavior of metallic copper greatly differs from that 
 of iron. At ordinary atmospheric temperatures fatty acids have no action upon 
 copper, provided atmospheric air be excluded, and, indeed, even at high temperatures 
 metallic copper resists the action of fatty acids in a remarkable manner. The immense 
 copper condensers used at Messrs. Price & Co.'s Candle- works for condensing the pro- 
 ducts obtained by superheated steam saponification are week by week exposed to the 
 simultaneous action of water and fatty acids at temperatures varying from 520 to 550 
 Fahr. ; they remain perfectly intact. It is only when the process is interrupted from 
 Saturday night till Monday morning that air gets into the tubes and that a slight corro- 
 sion takes place, which is indicated by the appearance of a bluish film on the inner sur- 
 face of the condenser-tubes. The experience collected at Price's Candle- works, for 
 
430 STEAM BOILERS. CHAP. XVIII. 
 
 which I am indebted to the kindness of Mr. Wilson, entirely agrees with the results of 
 experiments which I have made myself upon the subject. In these experiments rods of 
 copper were enclosed together with water and fatty acids in iron tubes and submitted to 
 a temperature of from 264 to 285 Fahr., corresponding to a pressure of from 2J to 3 
 atmospheres ; under these conditions the copper rods were found to be scarcely altered. 
 Similar results were observed when copper rods were heated with water and fatty acids 
 in glass tubes as high as 356 Fahr. ; the fatty acid remained colorless, and not a trace 
 of hydrogen was evolved. On the other hand, copper plates boiled with water and 
 fatty acids in an open flask, provided with a glass tube for the purpose of condensing 
 the steam and fat, were powerfully attacked in the course of a few hours." 
 
 The water in boilers generally contains a sufficient quantity of lime-salts to deter- 
 mine the decomposition of fatty bodies, and, when the pressure of the steam exceeds 30 
 pounds per square inch above the atmosphere, its temperature is probably sufficient to 
 cause the decomposition of tallow and of the animal and vegetable oils used for lubri- 
 cating the steam-cylinders and valves. The conditions necessary for the decomposition 
 of these fats and the combination of their acids with copper exist after the engines are 
 stopped and atmospheric air enters the valve-chests, condensers, pumps, and pipes. 
 When the engines are started again the copper salts thus formed are carried along with 
 the feed-water into the boilers, and as soon as they are brought there in contact with 
 metallic iron the fatty acids leave the copper and combine with an equivalent quantity 
 of iron, the metallic copper being deposited in the form of minute particles. The quan- 
 tity of iron primarily oxidized in this manner is relatively small, 32 parts of copper 
 combined with fatty acids being sufficient to oxidize 28 parts of iron. But the second- 
 ary action described by Professor Wartha, and consisting in the decomposition of the 
 oleate of iron and the recombination of the free fatty acid thus formed with metallic 
 iron, is a very dangerous source of corrosion. 
 
 "The insolubility of the fatty acids in water necessitates their immediate contact 
 with the metallic iron surfaces in order that they may effect corrosion ; this contact 
 would be greatest at or near the water-line, until the fat had become so far mixed me- 
 chanically with the particles of solid matter in the boiler as to give it a greater specific 
 gravity, when it would, as it does in practice, sink to the bottom of the boiler, where it 
 sometimes assumes the form of globular masses of varying size produced by the rolling 
 motion of the ship." (See ' Third Report of the Admiralty Committee on Boilers?} 
 
 An instructive case of the deterioration of iron, apparently due to the action of fatty 
 acids, was described by John A. Tobin, Passed Assistant Engineer, U.S.N., in a lecture 
 delivered before the Massachusetts Institute of Technology. 
 
SBC. 8. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 431 
 
 The bottom sheets of the horizontal steam-drums of the TJ. S. S. Swatara, having a 
 thickness of f inch, were found to be badly corroded, after two and a half years' use, as 
 high as the water resulting from condensation or carried into the drums by foaming 
 had risen, and particularly along the bottom, which was covered with a dark, greasy 
 sludge mixed with a noticeable quantity of oxide of iron. The corrosion of the plates 
 had taken place in the form of pitting and confluent honeycombing, extending from the 
 merest impressions to a depth equal to the thickness of the plates. Professor Nichols, 
 of the Massachusetts Institute of Technology, analyzed a sample of the greasy deposit, 
 and found therein copper in combination with various fatty acids, such as butyric, oleic, 
 stearic, palmitic acid, etc., and he found also particles of metallic copper present in the 
 scale on the plate. After new bottoms were put in the drums, and wrought-iron steam- 
 pipes connecting the drums with the boilers were substituted for the copper ones, corro- 
 sion was almost completely prevented by thoroughly draining and cleaning the drums 
 once a month. 
 
 To prevent the action of the fatty acids on condenser-tubes and on steam and feed 
 pipes they are tinned when they are made of copper or brass. 
 
 The iron of the boilers may be protected against the action of the fatty acids by a 
 coating of paint or cement or by a thin layer of scale deposited on the heating-surfaces. 
 It is necessary to form the scale before the action of the fatty acids on the iron has 
 commenced, because afterwards it is very difficult to make the scale adhere to the iron, 
 although the water may be maintained in the boiler at a high degree of saturation. 
 
 It is also proposed to neutralize the fatty acids by means of carbonate of soda or 
 caustic lime, forming a soap. (See section 11 of the present chapter.) 
 
 Filters are used to remove grease as we! 1 as other foreign matter from the feed- water 
 before it enters the boiler. (See Selderfs Filter, section 14, chapter xv.) 
 
 Perkins avoids entirely the use of lubricants in the steam-cylinders and valves of his 
 high-pressure engines by employing a peculiar alloy for the wearing-surfaces of his 
 valves and pistons. 
 
 The formation of fatty acids is completely avoided by using mineral oils instead of 
 animal or vegetable oils or fats for lubricating the steam-cylinders and valves. (See 
 section 15, chapter xvii.) 
 
 To illustrate the difference in the action of tallow or vegetable oils and of mineral 
 oils upon copper, the Admiralty Committee on Boilers placed coils of sheet-brass in 
 common tallow and in mineral oil, and heated them day by day for four months, air 
 having free access to the surfaces. "1029.30 grains of sheet-brass, 12 inches long and 4 
 inches wide, lost 14.10 grains in the tallow, which was colored green ; a similar piece of 
 
432 STEAM BOILERS. CHAP. XVIII. 
 
 i 
 
 the same brass, weighing 1101.40 grains, 11.90 inches long and 4.10 inches wide, im- 
 mersed in the mineral oil, lost 0.20 grain, the oil becoming darker in color." 
 
 9. Corrosion of Boilers by Galvanic Action. When electro-heterogeneous 
 metals are brought in contact in the presence of acids or saline solutions galvanic action 
 takes place that is to say, the solution is decomposed, oxygen being disengaged at the 
 electro-positive pole of the galvanic couple. 
 
 Copper has a strongly electro-negative character relatively to metallic iron. 
 
 The contact of these two metals in the presence of water containing even a minute 
 quantity of saline constituents produces galvanic action ; the disengaged oxygen of the 
 water combines at once with the iron, while the hydrogen escapes at the pole formed by 
 the copper, leaving the latter unaltered. In this manner a minute quantity of copper 
 may cause the oxidation or corrosion of a large quantity of iron, if metallic contact is 
 maintained between the two metals. 
 
 On account of this action the use of copper tubes in marine boilers is inadmissible. 
 Copper is frequently carried into the boiler by the feed-water ; minute particles may be 
 abraded from the copper and composition pipes by the steam and water flowing through 
 them ; and the salts formed by the combination of copper with fatty acids are decom- 
 posed when they come in contact with the iron of the boiler, and metallic copper is de- 
 posited. (See section 8 of tlie present chapter.') 
 
 Particles of metallic copper have frequently been found on pitted and honeycombed 
 boiler-plates, and this peculiar form of corrosion has been ascribed to the continued 
 galvanic action of these detached particles. It is, however, difficult to decide how 
 much of the corrosion was due to the action of fatty acids and how much of it was 
 caused by the galvanic action of the copper. The Admiralty Committee on the Deterio- 
 ration of Boilers was inclined to ascribe but a small share of the corrosion of boilers to 
 the latter cause, because the conditions which usually prevail in marine boilers prevent 
 immediate contact between the small particles of copper and the clean metallic surfaces 
 of the iron plates. 
 
 Lead is likewise an electro-negative metal relatively to iron. The corrosion of the 
 plates in the vicinity of manholes and mudholes is ascribed to galvanic action caused 
 by the presence of lead, derived from the white and red lead used in making the 
 joints. 
 
 In order to prevent galvanic action it has been recommended to obtain electro-homo- 
 geneity in the boiler by using only one kind of iron, and iron tubes instead of brass tubes, 
 and, when steel is used, to employ it for every part of the boiler and not in combina- 
 tion with iron. It is, however, impracticable to get large quantities of iron or steel of a 
 
SEC. 9. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 433 
 
 perfectly uniform character, which would be necessary in order to ensure perfect elec- 
 tro-homogeneity. 
 
 There is also no direct evidence that brass tubes hasten the corrosion of boilers ; on 
 the contrary, the inner surface of iron tube-plates with brass tubes appears generally 
 to be little attacked by corrosion. 
 
 The introduction of copper into boilers is prevented by tinning copper and composi- 
 tion steam-pipes, feed-pipes, and condenser-tubes, by using mineral oils as lubricants 
 for the steam-cylinders and valves, and by filtering the feed- water. 
 
 Galvanic action may be effectually prevented by keeping the surfaces of the boilers 
 covered with a coating of firm scale or paint. 
 
 Feldbacher's plan of lining the interior of the boiler with sheet-copper has been 
 tried on land boilers, and, it is claimed, with good results, corrosion being prevented 
 and incrustation greatly lessened. Another patentee coats the interior of the boiler 
 with a mere film of copper deposited from a solution of cyanide of copper. Other 
 patents have been secured for coating the interior surfaces of boilers with tin, also for 
 enamelling and electro-coppering iron tubes. But none of these plans have come into 
 extensive use in marine boilers. 
 
 The film of black magnetic oxide of iron which covers a boiler-plate when it leaves 
 the rolls in a finished state affords to the iron complete protection against corrosion as 
 long as it adheres firmly and remains unbroken. When this film is thick it is easily de- 
 tached from the plate, and its thickness depends upon the temperature of the iron at the 
 time of rolling. Much of this film is detached by the rough usage to which the plates 
 are subjected in transportation and in the boiler-shop, and by the various processes of 
 boilermaking. 
 
 Professor Barff produces a firmly-adhering coating of this black oxide artificially 
 by heating articles of cast or wrought-iron, the surfaces of which have been carefully 
 cleaned, to a high temperature in a closed chamber from which the atmospheric air 
 has been removed, and subjecting them to the action of superheated steam. He 
 states that the temperature required for this process ranges from 650 to 1,500 Fahr. ; 
 the lower temperature and longer treatment for wrought-iron, the more rapid treatment 
 and higher temperature for cast-iron. The application of this process to finished boilers 
 has been suggested, but has not been practically tested. 
 
 A film of this black oxide protects iron against the attack of powerful acids. From 
 experiments made by Schonbein and others it appears that this oxide of iron has an 
 electro-negative character far greater than copper and nearly equal to that of platinum. 
 This fact may explain the rapid corrosion of iron at places where this film has been 
 
434 STEAM BOILERS. CHAP. XVIII. 
 
 broken, since a powerful galvanic action might be produced under favorable circum- 
 stances by the contact of this oxide and of the pure iron, the former forming the nega- 
 tive and the latter the positive pole of a galvanic battery. On this account some manu- 
 facturers of boilers remove carefully the whole film of oxide from boiler-plates, since it 
 cannot be maintained unbroken in a boiler and would rather hasten than prevent its 
 corrosion. 
 
 The corrosive effect of galvanic action on iron boilers may be prevented by rendering 
 them the negative pole of an electric battery. This is generally and most successfully 
 accomplished by placing within the boiler and connecting with the shell a metal which 
 has a stronger electro-positive character than iron. The metal generally used for this 
 purpose is zinc. 
 
 1O. The Use of Zinc for the Prevention of Corrosion and Incrustation 
 of Boilers. Very conflicting accounts are given by engineers about the efficacy of zinc 
 in preventing the corrosion of boilers. From the varied testimony offered on this 
 point, and from the results of a series of experiments, the Admiralty Committee on 
 Boilers drew the following conclusions, which are given in their 'Third Report '- 
 viz. : 
 
 "Apart from any consideration as to the existence of galvanic action in boilers, the 
 protective value of zinc may be stated as follows : If a boiler is worked in the ordinary 
 manner with sea- water its exposed surfaces will be vulnerable to the action of all the 
 corrosive influences which may be present capable of affecting iron ; but if zinc be in- 
 troduced and applied in the manner which has already been pointed out i.e., perfect 
 metallic continuity ensured between it and the iron galvanic action is set up between 
 the two metals, and the latter is compelled, by the presence of one of a more electro- 
 positive nature, to assume a negative condition towards corrosion or oxidation. 
 
 " Such being the case, the metallic condition of the iron is preserved at the expense 
 of the zinc, which loses in course of time its metallic nature by oxidation, in which 
 latter condition it ceases to afford protection and must therefore be renewed at intervals. 
 In cases where this metallic contimiity has not been effected the zinc would share with 
 the iron surfaces of the boiler any corrosive action that might be present, in proportion 
 to the surface exposed, which, in any case, would be relatively small. There would be 
 no electro-chemical relation between the metals, and the different results observed by 
 marine engineers may have depended upon the fortuitous circumstance that in some 
 cases metallic continuity has been unintentionally effected in suspending the zinc from 
 the stays of the boiler ; this seems to be a very probable explanation of the discordant 
 opinions held by many as to the protective value of zinc," 
 
SEC. 10. CAUSES AND PREVENTION OF THE DETERIORATION OF BOILERS. 435 
 
 Zinc has also been found to prevent the formation of adhesive scale in land boilers 
 fed with calcareous water. The mode of action of zinc in this respect is described by 
 Brossard de Corbigny, in an article in the Annales des Mines of 1877, which has been 
 translated for the Journal of the Franklin Institute of January, 1878, and from which 
 the following is extracted : 
 
 " When the water is but little calcareous the deposits, instead of forming solid and 
 adherent scale, remain in the state of fluid mud, easily removable by simple washing. 
 The iron being clean and not rusted, no picking or scraping is needed, which effects a 
 great economy of time, hard labor, and oversight. 
 
 " When, however, the water is strongly calcareous the deposits are as coherent and 
 stony as though the zinc had not been employed ; but, what is extremely important, this 
 scale, while acquiring its thickness and hardness, does not adhere to the iron. It can 
 be pulled off by hand, or, at worst, detached without much effort, leaving the iron 
 clean. A simple washing removes it from the boiler, and in this case, as in the previ- 
 ous, picking and scraping are avoided." 
 
 The following hypothesis is advanced to explain this action : 
 
 " The two metals, iron and zinc, surrounded by water at a high temperature, form a 
 voltaic pile with a single liquid, which slowly decomposes the water. The liberated 
 oxygen combines with the most oxidizable metal the zinc and its hydrogen equivalent 
 is disengaged at the surface of the iron. There is thus generated, over the whole ex- 
 tent of the iron influenced, a very feeble but continuous current of hydrogen, and the 
 bubbles of this gas isolate at each instant the metallic surface from the scale-forming 
 substance. If there is but little of the latter it is penetrated by these bubbles and re- 
 duced to mud ; if there is more, coherent scale is produced, which, being kept off by 
 the intervening stratum of hydrogen, takes the form of the iron surface without adher- 
 ing to it. 
 
 "Zinc introduced into a boiler whose surfaces had been imperfectly freed of scale 
 has the property of detaching the remainder. This effect is well explained by the action 
 of a feeble disengagement of gas beneath the scale, raising it little by little and separat- 
 ing it from the iron." 
 
 With the selenious water of the slate-works of Angers the addition of zinc gave no 
 useful result ; the scale adhered strongly to the iron ; but the writer does not venture to 
 say whether this result was to be attributed to the more coherent nature of pure sul- 
 phate of lime, of which the scale almost entirely consisted, or to an insufficient quantity 
 of zinc employed. 
 
 The action of zinc with sea- water and acid water was not investigated by the writer. 
 
436 
 
 STEAM BOILERS. 
 
 CHAP. XVIII. 
 
 Fig.147a. 
 
 With water containing free acid, however, no good results would be obtained, as the 
 zinc would be corroded very rapidly and too large a quantity would be required. 
 
 De Corbigny recommends that in boilers fed with fresh water from half a pound to 
 one and a half pounds of zinc should be used for each five square feet of water-heating 
 surface during several months, and that the zinc should be used in the form of pigs or 
 cubical masses, and not in the form of chippings or sheets, since in the latter case the 
 electro-chemical action, being exerted over too large a surface, becomes too rapidly 
 exhausted. 
 
 Other writers recommend 20 inches of area for each horse-power, and for marine 
 boilers at least a quarter of a pound for each square foot of grate-surface. 
 
 The zinc slabs should be distributed over different parts within the water-space of 
 the boiler ; but they should not be suspended directly over the furnaces or back-connec- 
 tions, where the oxide of zinc, as it drops down, may cause blistering and burning. 
 When zinc is placed in steam drums and pipes the oxide is sometimes carried by the 
 steam-currents into the valves and causes injury to them. 
 
 Figure 147a shows an arrange- 
 ment adopted in some boilers for 
 ] securing the zinc. Zinc slabs about 
 15" X 12" X 1" are bolted to iron 
 clamps and secured by them firmly 
 to the iron stay-rods or tubes. It is 
 important that, so far as it is practi- 
 cable, no water should get between 
 the zinc and the iron. To ensure 
 perfect metallic continuity between 
 the two metals the zinc is some- 
 times soldered to the iron or zinc studs are screwed into the shell of the boiler. 
 
 The ' Steam Manual,' issued by the English Admiralty (1879), contains the following 
 instructions regarding the use of zinc in boilers : " The zinc slabs appear to deteriorate 
 either by gradually wasting away or by gradual change of substance. In this latter 
 case, which is the most common, the zinc slab gradually becomes black in color, brittle, 
 and friable ; and when it has been long exposed the mere pressure of the hand is suffi- 
 cient to reduce it to powder. Experience alone will enable the engineer officer to deter- 
 mine how far the zinc may be allowed to decay before renewal ; but, until, that experi- 
 ence is gained, it is recommended that the zinc should be renewed as soon as the decay 
 has penetrated a quarter of an inch below the surface. Under ordinary conditions of 
 
SEC. 11. CAUSES AND PREVENTION OP THE DETERIORATION OP* BOILERS. 
 
 working the zinc slab may be expected to last, when the boilers are new and the zinc 
 good, not less than from two to four months under water. 
 
 " Good zinc is to be used for these slabs. Zinc ' bottoms ' must not be used, and the 
 remains of slabs which have been in use must not be recast on board ship for further 
 use. If upon examination they are not sufficiently decayed to be rejected altogether 
 the decayed parts may be chipped off and the slab refitted and kept in use for some 
 time longer. When the slabs are finally removed from the boilers they are to be broken 
 up, the decayed portions thrown aside as useless, and such of the remains as appear to 
 consist of good zinc preserved and returned into store." 
 
 11. Action of various Substances upon the Incrustative and Corrosive 
 Ingredients of Feed-waters. The substances used to prevent the formation of hard, 
 adhesive scale and the internal corrosion of boilers are either introduced directly into 
 the boiler, or they are mixed with the feed- water and made to act upon the salts and 
 acids contained in the water before it enters the boiler. When the latter method is used 
 tanks of large capacity have generally to be provided in order to allow time for the 
 completion of the process of purification. On board of vessels there is generally no 
 room for such tanks, and the action of neutralizing the incrustative and corrosive pro- 
 perties of the feed- water has to take place, in part at least, within the boiler. 
 
 . Many of the substances recommended for the prevention of incrustation act by a 
 mechanical process, while others effect a chemical reaction, either changing the carbon- 
 ates and sulphates of lime into soluble salts of lime or producing insoluble salts of lime 
 and magnesia, which are either precipitated as an incoherent mass or remain suspended 
 in the water without agglutinating. 
 
 The anti-corrosive remedies are intended especially to change the chloride of magne- 
 sium and the fatty acids into substances which do not corrode the metal of the boiler, 
 are not readily redecomposed under the conditions obtaining in a steam boiler, and 
 either pass off with the steam in the form of gas or can be removed from the boiler by 
 the ordinary process of blowing-off. 
 
 The selection of any of these remedies should always be preceded by a careful ana- 
 lysis of the water used for feeding the boiler. The presence of certain substances may 
 not only neutralize the effect of a chemical agent, but cause reactions which produce 
 even more harmful substances. In several instances remedies have been proposed which 
 increased corrosion while they prevented incrustation, and others produce practical dif- 
 ficulties which render their use not only inconvenient but dangerous. 
 
 The chemical treatment of the water of marine boilers is more difficult than that of 
 the water of most land boilers, on account of the greater number of ingredients con- 
 
438 STEAM BOILERS. CHAP. XVIII. 
 
 tained in the former, and because the method of purification in tanks is not admissible 
 on board of vessels. 
 
 "The present large capacity of marine boilers using sea- water alone renders any 
 complete chemical treatment of the water, with the object of preventing corrosion, ex- 
 tremely difficult, as, besides sulphate of lime, sea-water contains salts of magnesium, 
 which, in any attempt towards a condition of alkalinity by lime or carbonate of soda, 
 must be decomposed with production of a bulky precipitate of magnesia ; and this, 
 added to the sulphate of lime, may induce priming, and would increase the difficulties 
 resulting from the accumulation of solid matter on the evaporating surfaces, unless the 
 quantity of scale be limited by substituting distilled sea-water for the feed from the sea 
 itself." (See ' TJiird Report of the Admiralty Committee on Boilers.') 
 
 The many anti-corrosive and anti-incrustative remedies offered for sale contain, in 
 some form or other, one or several of the following ingredients combined in a more or 
 less judicious manner : 
 
 Oil-cakes, potatoes, and other starchy matter are sometimes put into the boiler to 
 prevent the formation of hard scale by enveloping the particles of lime so that they are 
 deposited in the form of mud ; but they produce frothing, and often cause the boiler to 
 foam badly. 
 
 Glue, offal of hoofs and horns, tobacco-juice, Irish moss, peat, tow, hemp, etc., act 
 in a similar manner. 
 
 Clay likewise causes the lime-salts to settle down as a soft mud ; but a grave objec- 
 tion to its use are the gritty particles of sand contained in it, which, earned over by the 
 steam, injure the valve-faces. 
 
 Various varnishes or lacquers are proposed to prevent incrustation, but the pre- 
 sence of fatty matter in them often increases corrosion, and they afford insufficient pro- 
 tection against incrustation. 
 
 The action of petroleum or paraffine-oil as a protection from incrustation is involved 
 in obscurity, but its effects are reported on favorably. It is said to penetrate and rot 
 the scale, making it porous and easily removed, the quantity used being about one 
 quart per week for a twenty-five horse-power boiler. The most efficient qiiality is the 
 heavy, unrefined oil, the other being soon expelled by the heat. The oil may also be ap- 
 plied to the iron as a coating. (See section 14, chapter xvii.) 
 
 Glycerine has been used in Europe in steel boilers to prevent the adhesion of scale. 
 
 Milk of lime or hydrate of lime added to water takes up the excess of carbonic acid, 
 which enables the water to hold the carbonate of lime in solution, and the latter is thus 
 precipitated. This process, known as Clark's, recommends itself through its cheapness 
 
SKC. 11. CAUSES AND PREVENTION OP THE DETERIORATION OF BOILERS. 439 
 
 and simplicity where tanks can be provided for carrying it on. It is used to purify the 
 water of locomotive and stationary boilers. 
 
 A solution of caustic lime being added to sea- water has the effect of decomposing the 
 chloride of magnesium and sulphate of magnesia, with the formation of the correspond- 
 ing salts of lime, magnesia being liberated. 
 
 Hetet (Pharmacien-en-chef de la Marine) has made extensive experiments at Brest 
 with a process of neutralizing the fatty acids contained in the condenser-feed by means 
 of lime-water. The quantity of lime used by him is equal to about one-tenth of the 
 weight of oil used as a lubricant in the cylinders of the engines. The lime forms with 
 the grease an insoluble soap, which is found in the bottom of the boilers or of the puri- 
 fying-tank. This process should always be carried on in tanks, and the water should 
 be filtered, so that the lime-soap does not enter the boiler, where, under the influence of 
 heat and in contact with the iron, it would probably be decomposed. (See section 8 
 of the present chapter.} 
 
 Carbonate of soda, caustic soda, and. potash cause the carbonates of lime and mag- 
 nesia to be precipitated by combining with the free carbonic acid, forming soluble car- 
 bonates or bicarbonates of soda and potash. 
 
 Carbonate of soda decomposes the sulphate of lime, forming sulphate of soda, which 
 is held in solution, and carbonate of lime, which is precipitated. 
 
 By combining with the free carbonic acid in the feed- water these alkalies remove a 
 very harmful corrosive element. (See section 6 of the present chapter.) They also effec- 
 tually neutralize the fatty acids by forming a soluble soap which is not readily decom- 
 posed. 
 
 The ' Admiralty Instructions ' relating to the machinery of English naval vessels pro- 
 vide that, with a view to determine the condition of the water in the boilers as regards 
 its acidity, neutrality, or alkalinity, the water of each boiler is to be tested with test- 
 paper at least once per day when under steam. Should the water in the boilers be 
 found to be in an acid condition a small quantity of carbonate of soda is to be sup- 
 plied to the boiler to neutralize the acid in the water. The soda should be put into 
 the condenser or hot- well, from which it will be pumped into the boilers with the feed- 
 water. 
 
 It is recommended to use carbonate of soda in connection with lime or other in- 
 gredients for the prevention of scale and corrosion. First, by adding a solution of 
 caustic lime to the feed- water, the chloride of magnesium and sulphate of magnesia are 
 to be decomposed, with formation of the corresponding salts of lime, magnesia being 
 liberated ; then carbonate of soda is to be supplied, which will decompose those two 
 
440 STEAM BOILERS. CHAP. xvm. 
 
 lime-salts, forming carbonate of lime, which, is precipitated, and chloride of sodium and 
 sulphate of soda, which remain in solution. 
 
 "The use of carbonate of soda in limited quantity, with the idea that it neutralizes 
 the effect of fatty acids in marine boilers raising steam from sea- water, is questionable, 
 because so long as chloride or sulphate of magnesia remains in the water the carbonate of 
 soda is immediately decomposed with the result as stated above. It is only when those 
 two salts have been wholly decomposed, and the sulphate of lime in solution has been 
 converted into carbonate, that the carbonate of soda can be potential as such. So far as 
 the committee are aware this condition has never been realized in marine boilers, and it 
 is for this reason that the benefit which would have resulted from the use of a limited 
 quantity of soda (in fresh water) has not existed. . . . 
 
 " The quantity of soda required for the complete decomposition of the magnesium 
 and calcium salts is easily ascertained. To replace magnesium in any of its compounds 
 by sodium it is necessary to make use of 23 parts of the latter for 12 parts of the 
 former, and similarly to replace the calcium (the equivalent of which is 20) also requires 
 23 parts ; or, taking the oxides of these metals, lime, magnesia, and soda, their respec- 
 tive equivalent values are 28, 20, and 31. There are in a ton of ordinary sea- water 
 19,837 grains of magnesium and 7,233 grains of calcium, and to replace these two by 
 sodium there are required 46,339 grains of that metal. The weight of crystallized car- 
 bonate of soda which contains 23 parts or one equivalent of sodium is 143, so that 
 46,339 x 143 -f- 23 = 288,107 grains, or 41 pounds, 2 ounces, 232 grains, are the quantity 
 of crystals of carbonate of soda required for one ton of sea- water." (Report of Admi- 
 ralty Committee on Boilers.} 
 
 Hyposulphite of soda keeps the salts of lime in solution ; oxalate of soda causes their 
 precipitation as a loose deposit ; but both are expensive. 
 
 Proto-chloride of tin, silicate, phosphate, and arseniate of soda have been tried, 
 but are too expensive, and are not found to answer their purpose sufficiently 
 well. 
 
 Chloride of ammonium, or sal-ammoniac, acts on the carbonate of lime and magne- 
 sia, converting them into chlorides, while their carbonic acid goes to the ammonium, 
 forming carbonate of ammonia, which escapes with the steam. Sulphate of lime is not 
 affected by sal-ammoniac. 
 
 Tannic acid forms insoluble tannates of lime and magnesia by the decomposition 
 of the carbonates. These tannates have a low specific gravity and float about the water 
 instead of agglutinating into a crust. The sulphates and chlorides are not acted upon 
 by tannic acid, and will incrustate notwithstanding its presence. Free tannic acid at- 
 
SEC. 11. CAUSES AND PREVENTION OP THE DETERIORATION OF BOILERS. 441 
 
 tacks the iron, and its presence in a boiler is a source of great danger. Its effect may be 
 neutralized by carbonate of soda. 
 
 Tannin-bearing bodies, like oak and hemlock bark, logwood, catechu, leather, or ex- 
 tracts made of them, are frequently introduced into boilers for the prevention of scale. 
 When the solid bodies are used there is danger of pieces of bark, etc., being blown over 
 by the steam into the valves and pipes. 
 
 Tannate of soda, being introduced into a boiler where the carbonates of lime and 
 magnesia are present, gives up to them its tannic acid to form the insoluble tannates, 
 and takes up their carbonic acid to form carbonate of soda, which in turn reacts on the 
 sulphate of lime, converting it into carbonate of lime and bringing it thus under the 
 inihience of a fresh portion of tannate of soda. The constant presence of the soda pro- 
 tects the iron from the corrosive action of tannic acid as well as of carbonic acid. The 
 same reaction takes place between tannate of soda and the already existing scale, but 
 more slowly than when the salts are held in solution. 
 
 Acetic acid converts the carbonates into soluble acetates ; but the iron is equally ex- 
 posed to its attack. Sulphate of lime is not altered by it. 
 
 Molasses, fruit, cider, vinegar, etc., containing more or less acetic acid, have been 
 used in boilers for the prevention of scale. 
 
 Tannic acid as well as acetic acid may be advantageously used to purify the water in 
 tanks, the excess of acid being neutralized by carbonate of soda before the water is fed 
 into the boiler. 
 
 It has also been proposed, for the purification of water in tanks, to convert the earthy 
 carbonates into soluble chlorides by hydrochloric acid, and to neutralize the excess of 
 acid by filtration through carbonate of baryta (witherite) in the form of coarse powder. 
 The soluble chloride of barium thus formed decomposes the siilphate of lime, forming 
 chloride of calcium which is kept in solution, and insoluble sulphate of baryta which is 
 deposited. (See " Steam Boiler Waters and Incrustations" by Dr. J. G. Rogers, Jour- 
 nal of the Franklin Institute, 1872.) 
 
 When boilers have been fed with water containing a large quantity of organic mat- 
 ter, sewage, bilge-water, etc., it has been found that they remained remarkably free 
 from corrosion, and that the formation of scale with such water was even impossible. 
 The latter effect is probably due to the mechanical action of the organic matter which 
 envelops the solid scale-producing matter as soon as it is formed. The anti-corrosive 
 effect of the organic matter may be ascribed to its absorption of the free oxygen con- 
 tained in the water. 
 
CHAPTER XIX. 
 
 BOILER-EXPLOSIONS. 
 
 1. Causes of Boiler-explosions. Kupture takes place in a boiler either when 
 any part is not sufficiently strong to resist with a reasonable margin of safety the stress 
 produced by the working pressure under ordinary conditions, or when the stress ex- 
 ceeds the ultimate strength of any part in consequence of an excessive rise of pressure 
 in the boiler, or in consequence of great variations and differences of temperature, sud- 
 den shocks, weakening of plates by overheating, or similar abnormal conditions. 
 
 The weakness of a boiler may be due to the faulty design of its bracing, to the use 
 of inferior material or to injuries received by it in the processes of boilermaking, to 
 bad workmanship, and to gradual deterioration produced by the various causes de- 
 scribed in chapter xviii. By exercising vigilance in supervising the building of a boiler, 
 and by conducting with intelligent care the periodical inspections and tests, the actual 
 condition of a boiler may be known, and the steam-pressure which it can bear with 
 safety may be determined accordingly. The original cause of any leak or sign of weak- 
 ness should always be determined promptly and exactly, and removed in making re- 
 pairs, otherwise deterioration will not only continue, but may even be aggravated by 
 the means adopted for stopping the leak or strengthening the weak place, especially 
 when the defect is produced by varying bending strains near the attachment of braces 
 or the laps of joints. 
 
 Various theories have been proposed to account for the instantaneous generation of 
 excessive pressure within steam boilers by the conjuncture of unusual conditions ; seve- 
 ral of these will be discussed farther on. Such theories have been brought forward 
 especially in order to explain the frequent occurrence of explosions at the moment 
 preparations are commenced for starting the engines after a boiler has been lying with 
 banked fires for some time. 
 
 When the fires are burning actively, and the steam is allowed to accumulate in the 
 boiler by keeping the stop and safety valves closed, the pressure will increase with 
 great activity. In a large marine boiler of ordinary form experimentally exploded at 
 Sandy Hook, N. J., in 1871, the pressure rose in thirteen minutes from 29f pounds to 
 
 443 
 
SBC. 1. BOILER-EXPLOSIONS 443 
 
 53J pounds per square inch ; the furnaces contained a wood-fire, and the water in 
 the boiler stood 15 inches above the upper row of tubes. (See section 5 of the present 
 chapter.) 
 
 A gradual excessive accumulation of steam-pressure is always due to mismanage- 
 ment and carelessness ; it is, however, a frequent cause of boiler-explosions. It cannot 
 take place when the boiler has a safety-valve of sufficient size, properly located and in 
 working order. But when the safety-valve cannot relieve automatically the boiler of 
 an excess of pressure, the latter may accumulate to a dangerous degree rapidly after 
 steam has been raised and the engines are not started or are suddenly stopped, and 
 more gradually when the engines are worked slowly or when the boilers are kept under 
 steam with banked fires. The danger of getting an overpressure of steam in the boiler 
 is greatly increased when the steam-gauge is inoperative or gives wrong indications. 
 
 The following circumstances frequently make safety-valves inoperative in practice : 
 When the safety-valve is not placed directly on the boiler, but on the steam-pipe be- 
 yond the stop-valve, the latter may be closed and shut off the steam from the safety- 
 valve. In repairing or cleaning the boiler the workmen sometimes plug up the aperture 
 of the valve-chamber from inside the boiler, or put a blank flange between the safety- 
 valve and the escape-pipe to prevent the drip of hot water from leaky valves ; several 
 cases of explosion have happened in consequence of the neglect of removing this ob- 
 struction before the boiler was closed and steam was raised. In cold weather the 
 escape-pipe may be obstructed by the freezing of the water which has accumulated 
 therein. Safety-valves are sometimes purposely overloaded and tied or wedged down 
 by reckless attendants, or the rise of the valve may be prevented by some obstruction 
 accidentally placed over the valve-chamber. The valve may stick fast in consequence 
 of corrosion or of the accumulation of grease and dirt at the articulations of the lever 
 and in the guide-sleeve of the stem ; or it may be prevented from rising by the bending 
 of the lever, stem, or guide-spindle ; or the too closely fitted guide-feathers may be 
 jammed in the valve-chamber in consequence of distortion or unequal expansion caused 
 by irregular strains or differences of temperature. To the latter cause the explosion of 
 the boiler of the British armored vessel Thunderer was attributed. 
 
 To place safety-valves beyond the control of reckless ignorance and carelessness 
 boilers are provided with lock-up safety-valves. For the design, construction, and 
 arrangement of safety-valves see section 11, chapter xv. Safety-valves should be 
 opened by hand soon after steam has commenced to form, and from time to time after- 
 wards while steam is on the boilers, to make sure that they are in working order. 
 
 The overheating of plates may be caused by a deficiency of water in the boiler, or 
 
444 STEAM BOILERS. CHAP. XIX. 
 
 by the formation of deposits of solid matter on the heating-surfaces, or by the accumu- 
 lation of steam on the heating-surfaces. The overheated parts may be sufficiently 
 weakened to give way at once at the ordinary working pressure either by fracturing or 
 by bulging ; or they may be left in a deteriorated condition after cooling down ; or 
 their excessive and unequal expansion may produce injurious strains in the parts with 
 which they are connected. When highly-overheated plates are suddenly partially 
 cooled by coming in contact with water their local contraction may produce sufficiently 
 severe strains to rupture them. The ways in which rupture or severe strains are pro- 
 duced in boilers by sudden variations of temperature have been described in section 3, 
 chapter xviii. 
 
 Violent shocks are liable to cause rupture in boilers which are in a strained condition 
 in consequence of the steam-pressure within them. Such shocks may be produced by 
 heavy weights falling on the boiler or by the collision or grounding of the vessel. The 
 impact of the water carried upward by the rush of steam when the safety-valve or a 
 stop- valve is opened suddenly may sometimes be sufficient to produce rupture. 
 
 Boilers have been severely jarred by the detonation of inflammable gases which had 
 collected in their flues. When a boiler is kept with covered banked fires and tightly- 
 closed damper the gases distilled from bituminous coal, not being able to escape from 
 the flues, may form in them a detonating mixture with the atmospheric air, and, under 
 such conditions, an explosion may occur when, by uncovering the fires, the gases are 
 ignited. More or less severe accidents of this kind have occurred with stationary boil- 
 ers, but they cannot take place where dampers are not employed, as is usually the case 
 in marine boilers. 
 
 Successive explosions of several boilers working side by side have been repeatedly 
 observed, and may be explained by assuming that the explosion of one of the boilers 
 produced severe jars or concussions in the adjoining ones, or that the latter were per- 
 forated by detached pieces from the exploded boiler. The rupture of a pipe or drum 
 connected with several boilers may be the cause of their explosion. In a war-steamer 
 the perforation of boilers by shot or splinters in action is a dangerous source of explo- 
 sion. Locomotive boilers have exploded because the shell was pierced by a broken 
 connecting-rod, or because the steam-dome was carried away by coming in contact with 
 a tunnel or an overhead bridge. 
 
 The rupture which takes place in a boiler may be only of a local character and not 
 affect the strength of the boiler as a whole ; this is often the case when rivets or small 
 portions of plate are blown out or when small tubes burst or collapse. Or the damage 
 done to any part may cause the opening of seams in such a manner as to allow a gradual 
 
SBC. 8. BOILER-EXPLOSIONS. 445 
 
 discliarge of the steam, thereby relieving the boiler of an overpressure. In these cases 
 the principal danger consists in the liability to scalding by hot water and steam. 
 
 The rupture of a boiler is called an explosion when it causes the sudden liberation 
 of large masses of the steam and water confined in the boiler, by which means a dy- 
 namic force is developed which aids in the demolition of the whole structure and hurls 
 detached portions with great violence. This will take place when the rupture deprives 
 other parts of the boiler of an essential support, and thus throws additional excessive 
 strains on them, as in the case of broken stays or braces, and rents in circular shells ; or 
 when the original weakness extends over a large portion of the structure, which gives 
 way at nearly the same instant, and thereby sets free suddenly an enormous amount of 
 energy stored up in the steam and heated water. 
 
 2. Various Theories concerning Boiler-explosions. The extreme violence of 
 many boiler-explosions has caused the very general impression that some unusual, 
 instantaneously -generated forces are required to produce them. Although it has been 
 demonstrated that the energy stored up in water heated to a temperature corresponding 
 to the ordinary working pressures of steam boilers is more than sufficient to produce 
 the effects of the most violent boiler-explosions, and although it has been repeatedly 
 shown by direct experiment that a moderate and gradually-accumulated steam-pressure 
 in a weak boiler does produce all the phenomena of a violent explosion, there are still 
 many advocates of special theories which explain these phenomena by the sudden gene- 
 ration of abnormal forces within the boiler in consequence of the concurrence of extra- 
 ordinary circumstances. It is of the gravest importance that the true causes of every 
 boiler-explosion should be clearly understood, in order that their recurrence may be 
 guarded against. The tendency to ascribe explosions to obscure causes rather than to 
 regard them as the natural results of conditions which can be prevented by intelligent 
 care exercised in the design, construction, and management of boilers, has, no doubt, its 
 origin in the desire to escape responsibility for the results of culpable neglect, and has 
 been productive of much mischief by engendering carelessness on the part of owners 
 and attendants of steam boilers. Before accepting any theory as a correct explanation 
 of the causes of an explosion it is not sufficient to prove that the circumstances iipon 
 which the theory is based could exist in the case in question, and that they were capa- 
 ble of producing the phenomena observed, but it must be shown by direct evidence that 
 they did exist, or that no other known causes could or were likely to produce the 
 same results. 
 
 Little need be said about the theory which ascribes boiler-explosions to an electric 
 discharge ; no intelligent explanation has eVer been given of the supposed generation of 
 
446 STEAM BOILERS. CHAP. XIX. 
 
 electricity of sufficiently high tension in a steam boiler to produce the effects of a violent 
 explosion. This idea probably had its origin in an imperfect knowledge of the fact 
 that a current of steam sometimes exhibits electrical phenomena, the development of 
 electricity being solely due, as Faraday has shown, to the friction of the suspended 
 watery particles against the sides of the orifice through which the steam issues. 
 
 Another theory assumes the possibility of the formation and detonation of a mix- 
 ture of hydrogen and oxygen gases in a steam boiler. The hydrogen set free in conse- 
 quence of the corrosion taking place within a boiler, and the free oxygen contained in 
 the feed- water, are relatively small quantities and so diffused in the mass of steam that 
 they cannot form an explosive mixture. 
 
 At a meeting of the American Academy of Sciences (1877) it was shown that steam 
 could be decomposed by simple heat into the constituent gases of water viz., hydro- 
 gen and oxygen. The apparatus consisted of a flask in which water was heated, a tube 
 conveying the steam to a closed platinum crucible, where it was again heated by a spirit- 
 lamp, and a tube which carried thence the superheated steam and the liberated gases to 
 an ordinary pneumatic trough, where the mixed gases were collected in a test-tube 
 while the steam was absorbed. The gases thus collected were exploded by a lighted 
 match. The heat employed was a little over ordinary redness, but did not reach 
 whiteness. 
 
 In order that a detonating mixture of oxygen and hydrogen may be formed by a 
 similar process in a steam boiler, a considerable portion of the plates enclosing the 
 steam-space must be raised to a bright-red heat ; then the steam must be condensed by 
 the injection of cold water, which must not come in contact with the red-hot plates ; 
 the heat of the latter may ignite then the explosive mixture of gases. It is, however, 
 highly improbable that all these essential conditions will be fulfilled in a steam boiler, 
 and it is more likely that rupture will take place in consequence of the weakened con- 
 dition of the red-hot plates or of their sudden cooling by the injection of cold water. 
 
 Several theories have been proposed to explain how large masses of water might be 
 instantaneously converted into steam and thus prodiice an overpressure in a boiler. 
 
 It has been assumed that, especially when the feed- water is very greasy, the phe- 
 nomenon of the spheroidal condition of water might be produced on a large scale in a 
 boiler, and that when a large mass of water is instantaneously vaporized under these 
 conditions the resulting increase of pressure and the impact of the water thrown up 
 against the shell of the boiler by the suddenly-formed steam would be sufficient to rup- 
 ture the boiler. While there is reason to believe that this phenomenon occurs fre- 
 quently on a small scale on the plates forming* the furnace-crowns and the combustion- 
 
Sic. 3. BOILER-EXPLOSIONS. 447 
 
 chamber, and increases their liability to deterioration, there is no evidence that large 
 masses of water have ever been affected similarly in a boiler in such a manner as to 
 generate sufficiently great forces to produce an explosion. 
 
 AYhen water is deprived completely of air and is kept perfectly motionless its tem- 
 perature may be raised many degrees (according to Tyndall, 100 Fahr.) above the boil- 
 ing-point ; but with the slightest agitation the surplus of heat in the water is expended 
 in the sudden formation of an equivalent mass of steam. It is claimed that this pheno- 
 menon may occur in a steam boiler and cause its destruction by an instantaneous in- 
 crease of pressure and by the impact of a large mass of water thrown violently against 
 the walls of the boiler by the steam formed in an explosive manner. To produce this 
 superheated condition of the water the boiler must be supposed to have been standing 
 for a long time undisturbed, with closed valves and low fires, no circulation of the 
 water taking place within it. The sudden opening of the safety-valve or steam stop- 
 valve, or the starting of the feed-pump, would then be sufficient to produce the agitation 
 of the water which causes the sudden generation of steam. In the case of a marine 
 boiler the vessel likewise must lie perfectly motionless during the period of superheat- 
 ing. But we have no direct evidence that these conditions have ever been the cause of 
 the explosion of a boiler, and it is very doubtful whether a large mass of water could 
 be highly superheated in a boiler without circulation. 
 
 Another theory is based on the supposition that the steam in the boiler has become 
 highly-superheated in consequence of the overheating of the surfaces in contact with 
 it ; that then, by some means, a large mass of water is carried up in the form of spray, 
 and, mingling with the steam, is at once vaporized by the surplus of heat in the latter. 
 The thorough mingling of a large mass of water with the steam is, however, not easily 
 effected in practice, and, even in extreme cases, the whole surplus of heat in superheated 
 steam would not be sufficient to produce a considerable increase of pressure by the 
 vaporization of water. 
 
 In many cases the vaporization of large masses of water coming in contact with 
 highly-overheated plates is supposed to produce the sudden increase of pressure which 
 is regarded as the cause of the explosion. A large amount of plate raised to a red heat 
 might, no doubt, contain a sufficient quantity of heat to produce this effect ; but gene- 
 ral experience as well as direct experiment have demonstrated the fact that water thrown 
 on a red-hot plate does not absorb heat with sufficient rapidity to generate suddenly a 
 large quantity of steam. Several experiments were made by the Manchester Steam- 
 users' Association to test this theory. Some small empty boilers were made red hot 
 and water was forced into them, but in every case the boiler failed to explode. In most 
 
4A8 STEAM BOILERS. CHAP. XIX. 
 
 boilers the feed-entrance is near the bottom, and when the feed is turned on the water 
 will rise gradually up to the overheated plates and will not be scattered over them. 
 Severe overheating of plates in a boiler frequently opens the riveted seams, and thus 
 offers a means of escape to the steam and prevents an excessive accumulation of 
 pressure. 
 
 In nearly every case in which severe overheating of portions of the boiler has taken 
 place previous to an explosion it is reasonable to suppose that the loss of strength in 
 the overheated plates, or their strained condition when suddenly cooled off, woiild be 
 sufficient to cause rupture even with the ordinary working pressure ; and while an in- 
 crease of pressure produced by the sudden vaporization of a certain quantity of water 
 and a violent projection of water may have occurred simultaneously and to a certain de- 
 gree intensified the disruptive force, it is the weakened condition of the boiler which 
 must be regarded as the primary cause of the explosion. 
 
 Assuming that a boiler explodes either in consequence of a sudden reduction of its 
 strength or of a gradually -accumulated overpressure, rupture commences at the weakest 
 part and continues, following the lines of least resistance, in such parts as have been re- 
 duced in strength or are left insufficiently supported after the primary fracture. De- 
 tached portions of the boiler are projected with more or less violence, impelled by the 
 unbalanced force of steam pressing on their surfaces, and a force of corresponding mag- 
 nitude reacts on the opposite walls of the boiler. The steam expands to atmospheric 
 pressure, and a portion of the heated water vaporizes as the superincumbent pressure 
 diminishes. The steam suddenly generated in the body of heated water carries along a 
 mass of water, the impact of which assists in the work of destruction. . 
 
 The rapidity with which these consecutive effects are produced depends on the 
 nature, location, and extent of the fractures. The weight and temperature of the water 
 and steam determine the amount of energy stored up in a boiler ; but the violence of 
 an explosion, or the work done in exploding a boiler, depends greatly on the rapidity 
 with which the energy stored up in a boiler is liberated. 
 
 In discussing one of the experimental steam-boiler explosions at Sandy Hook, N. J., 
 in 1871, Professor R. H. Thurston presented the following calculations of the energy 
 stored up in the boiler and of the work done by the liberated forces : 
 
 " The steam boiler referred to weighed 40,000 Ibs., and contained about 30,000 Ibs. of 
 water and 150 Ibs. of steam, all of which had a temperature of 301 Fahr., when, at the 
 moment before explosion, the steam-pressure was 53 pounds above that of the atmos- 
 phere. 
 
 "When the explosion took place the whole mass at once liberated its heat, until 
 
SEC. 2. BOILER-EXPLOSIONS. 449 
 
 it had cooled down to the temperature of vapor under the pressure of the atmos- 
 phere. 
 
 "In this act the water gave off 30,000 x 89 = 2,670,000 British thermal units, and 
 the steam lost the difference between its total heat at 301 and that of 212 Fahr., or 
 150 x 27. 2 = 4, 080 thermal units. The sum 2, 670, 000 + 4, 080 = 2, 674, 080 thermal units 
 has an equivalent in mechanical energy of 2,674,080 X 772 = 2,064,389,760 foot-pounds, 
 and this was sufficient to have raised the whole boiler and contents, weighing 70, 000 Ibs., 
 to a height of 29,491.282 feet more than five miles. This represents the maximum 
 possible effect. 
 
 " The least effect would have been produced had the liberation of heat and the pro- 
 duction of additional quantities of steam, within the mass of water and at its surface, 
 been so sluggish as to have given no assistance in propelling the fragments of the rup- 
 tured boiler, the whole destructive work being done by the simple expansion of the 
 steam which filled the steam-spaces. 
 
 " The total amount of mechanical energy set free from the steam alone was 4,080 X 
 772 = 3, 149,760 foot-pounds, or sufficient to raise the whole boiler through a space of 78.74 
 feet and, water included, 44.99 feet. Owing to the greater inertia of the lower part of 
 the boiler, and particularly of its inelastic burden of water, the principal part of this 
 work was undoubtedly performed upon the upper portion and steam-chimney of the 
 boiler, weighing probably 6,000 Ibs. ; and, if entirely expended in this direction, the 
 work thus done was equivalent to raising this 6,000 Ibs. to a height of 525 feet. 
 
 "This latter case is capable of treatment in quite a different way from the above. 
 As the boiler was completely torn in pieces, the steam must have expanded pretty 
 equally in all directions, except where checked in its downward movement, and may 
 probably be treated as if forming a rapidly-expanding hemisphere of vapor, its centre 
 being in the steam-space of the boiler. 
 
 "The expansion of this hemisphere would have continued until the tension of the 
 steam was rediiced to that of the surrounding atmosphere, and would have continued 
 through a mean distance, as given by an approximate estimate, of 4.5 feet. The mean 
 pressure would be 25 pounds above the atmosphere nearly. 
 
 " The area of cross-section of the steam-drum was 4,071 square inches, and 4,071 X 
 25 X 4.5 = 457,987.5 foot-pounds, the amount of work done in its projection. 
 
 " The weight of the steam-drum, which was J inch thick, 6 feet diameter, and 8 feet 
 8 inches high, was, with its braces, 2,500 Ibs., and 457,987.5 -f- 2,500 = 183.2, the height 
 in feet to which the drum might have been thrown by the simple expansion of the con- 
 fined steam. In fact, the steam-drum had attached to it, when found after the explo- 
 
450 STEAM BOILERS. CHAP. XIX. 
 
 sion, a considerable part of the boiler-top, which, being comparatively light and being 
 acted upon by similar pressures, must have considerably accelerated rather than 
 retarded its ascent. 
 
 "The actual height of ascent of this piece was variously estimated by the 
 spectators at from 200 to 400 feet." (Journal of the Franklin Institute, March, 
 1872.) 
 
 3. Phenomena of Boiler-explosions. "When a boiler gives way from over- 
 pressure or sudden contraction a rent may be formed or a piece of plate blown out. 
 The former is the most usual manner of yielding ; but in both cases it will depend upon 
 the strength, nature, and arrangement of the material bounding the initial fracture, as 
 well as its position, and also upon the pressure, temperature, and amount of water 
 and steam in the boiler, whether the contents will gradually escape through the open- 
 ing already made, or whether in their violent rush they will increase the extent of 
 opening, and make it easy for the steam behind to tear the boiler into several pieces 
 and cause a violent explosion. 
 
 "Now, to make this more clear, we shall first consider the influence of the position 
 of fracture. Many cases have occurred of manhole-lids on the crowns of horizontal 
 boilers being blown aloft either from defect of fastening down or defect of material. 
 When the manhole is properly fortified with a mouth-piece or ring the cover is pro- 
 jected aloft, the contents gradually escape through the hole, and the boiler is left on its 
 seat (if this be sufficiently strong to withstand the recoil), and probably no further dam- 
 age is done, except to the boiler-house roof. Should, however, the same accident 
 happen to a manhole-cover underneath the boiler, placed near the ground, the effect will 
 be very different, and it will depend upon the weight of the boiler and water contained, 
 size of manhole, pressure of steam, and distance of aperture from the ground whether 
 the boiler and its contents will be merely raised a little from its seat, or whether it 
 will be shot aloft like a rocket by the unbalanced pressure on the discharge of steam. 
 If the manhole were in the side of a vertical boiler, and near the top, the blowing-off 
 of the lid into an open space in front would probably topple over the boiler if it 
 were not well supported. 
 
 "Again, if the manhole in our first case were without any provision for strength- 
 ening the plate surrounding it, and if the edges of the plate were reduced in strength 
 by fractures or by corrosion and wear, the rush of steam and water on the lid blow- 
 ing off would probably start a rent in the shell, which a high pressure within the boiler 
 would continue along the lines of least resistance, and the result would be a violent 
 explosion, the severed plates being carried in different directions. 
 
SBC. 3. BOILER-EXPLOSIONS. 451 
 
 "The remarks respecting the blowing-away of the manhole-cover apply also to 
 the case of a piece of plate being blown out." (Wilson.) 
 
 When a single stay gives way in a boiler there is a strong probability that the 
 adjoining stays will also give way in rapid succession on account of the greatly-in- 
 creased load thrown on them. The unsupported plate bulges out and finally tears, 
 usually through the line of rivet-holes in the seams. Unless braces are much reduced 
 in sectional area by corrosion, their weakest part is generally at the weld or in their 
 fastenings. The angle-irons to which braces are attached tear frequently through the 
 rivet or bolt holes. When the bulging of plates exceeds a certain limit the stay-bolts 
 are drawn through them, especially when they are simply screwed in or secured by 
 riveted heads. 
 
 When large flues collapse without fracturing to a great extent the steam or water 
 issuing through the cracks or opened seams may reduce the pressure, or, by putting 
 out the fires, check the increase of pressure sufficiently to avert an explosion. If the 
 fracture is of larger extent there is great danger of scalding by hot water and steam ; 
 the furnace-doors will be blown open and the fire scattered over the fire-room floor. 
 If a flue is ruptured to such an extent that it no longer acts as an efficient stay for 
 the plates to which it is attached, a violent explosion will be the probable result, 
 the ends of the boiler being blown away in opposite directions. 
 
 Similar results will be produced when a number of tubes are either fractured or 
 pulled out of the tube-plates. 
 
 The reaction of the steam and water issuing from the collapsed furnace-crown of 
 a locomotive boiler or of a vertical fire-tube boiler frequently sends the boiler into the 
 air to a great height like a rocket. 
 
 An extensive rupture in a cylindrical boiler generally results in a violent explosion 
 and the total destruction of the boiler, because its various parts are connected in such 
 a manner as to form essential supports for each other. On the contrary, large portions 
 of the flat stayed surfaces of a rectangular boiler may give way without seriously 
 weakening the rest of the boiler. 
 
 The violent explosion of the large rectangular boiler of the British armored vessel 
 Thunderer, in 1876, commenced with the giving- way of some stay-bolts which tied the 
 boiler-front to the uptake. The whole front above the smoke-connections was torn off 
 the shell through the lines of rivet-holes in the seams and thrown down, and the 
 portion of the uptake to which the front was stayed was torn and bent out of shape ; 
 but the rest of the boiler was uninjured. 
 
 The reports and papers published by the Hartford Steam-boiler Inspection and In- 
 
452 STEAM BOILERS. CHAP. XIX. 
 
 surance Company contain descriptions and illustrations of several instructive cases of 
 explosion of locomotive and marine boilers, in which the semi-cylindrical top gave way, 
 rupture commencing at and following a line of grooving near the horizontal seams. In 
 one case the whole semi-cylindrical top was blown off, fracture taking place almost 
 simultaneously at both sides through lines of grooving near the horizontal seams which 
 joined the top to the lower rectangular part of the shell, and extending through the 
 transverse seams connecting the top to the front plate and to the cylindrical barrel of 
 the boiler. 
 
 In other cases a piece, extending nearly the whole length of the semi-cylindrical top, 
 was torn off on one side of the boiler. The fracture followed likewise a line of hori- 
 zontal grooving near a seam, and extended in a transverse direction to lines where the 
 shell was strengthened by stays or by the flanges of steam-domes, the detached pieces 
 bending over on these lines as on hinges. In one case the piece thus torn off the side 
 of a locomotive boiler was about 5 feet long and 1 feet wide, and had an area of 
 nearly 7 square feet. The pressure on this area was but a little less than 70 tons at 
 the pressure of 130 pounds per square inch. The reaction of the force set free by the 
 rupture overturned the locomotive. 
 
 When a cylindrical shell gives way at a longitudinal seam, or by tearing through a 
 longitudinal line of weakness produced by corrosion or grooving, the fracture may be 
 confined to a single plate, and, by continuing through the circumferential seams, detach 
 the belt of which it forms part from the shell, or the longitudinal fracture may con- 
 tinue through several belts. The latter is generally the case when the longitudinal 
 seams of adjoining plates are in the same line. When they break joint the rent ex- 
 tends sometimes in a diagonal direction through adjoining plates till it strikes again a 
 longitudinal seam, along which it continues. As the fracture extends in a transverse 
 direction after the longitudinal rupture has taken place, the cylindrical plates are flat- 
 tened out, and in this manner the rivet-heads in circumferential seams are frequently 
 torn off. When the longitudinal rupture takes place near the bottom of the boiler the 
 detached plates will probably be thrown some distance ; but when this line of fracture 
 is near the top the plates may be found in nearly their original location. There is a 
 strong probability that when a circumferential belt of plates has been torn off a cylin- 
 drical shell, the two ends of the boiler will separate and will be thrown a greater or less 
 distance in opposite directions. But in some cases, when the two ends are tied together 
 strongly by their braces, stays, and flues, or when the destructive forces spend them- 
 selves promptly through the vent made by the original fracture, the boiler remains 
 otherwise uninjured and is not moved from its seat. 
 
SEC. 4 BOILER-EXPLOSIONS. 453 
 
 In the case of the boiler-explosion on the steamer Westfield, in the harbor of New 
 York, in 1871, rupture commenced at one side of the cylindrical shell of the boiler 
 along a horizontal line of grooving. The fracture extended through the width of one 
 plate and continued through the transverse seams, detaching one belt from the rest of 
 the shell and flattening it out. This portion of the shell was found lying directly op- 
 posite the original position in the boiler before the explosion. The front and the back 
 of the boiler had separated and were thrown some distance in opposite directions. 
 
 4. Investigation of Boiler-explosions. " In investigating the cause of a com- 
 plicated explosion the relative weights, positions, shapes of the scattered pieces, and 
 the direction taken by them must first of all be carefully noted, and their original posi- 
 tions in the boiler be assigned to them, along with the positions of the different mount- 
 ings, manner of staying, and absence or presence of means for strengthening domeholes, 
 manholes, tubes, combustion-chambers, etc. The original shape of the shell and large 
 flue-tubes should be ascertained as accurately as possible. The primary rent is then to 
 be sought for. In many cases the direction taken by the heavier pieces is a guide to 
 this, as the fractured plates, if free to move, will shoot off, the light pieces along with 
 and in the direction of the first rush of steam, and the heavier pieces in an opposite 
 direction. 
 
 "That this, however, is not always the case is obvious ; as, for instance, when the 
 boiler turns over before separating, or where the direction a piece of the shell would 
 take, if free to move, is changed by part of it clinging for a time to the larger mass to 
 which it may be attached. 
 
 "All the edges of the plates and angle-irons along the lines of fracture should be 
 carefully examined in search of weak places, such as thinness caused by grooving and 
 corrosion, external and internal, wasting of rivet-heads, defective rivet-holes, insuffi- 
 cient lap, old flaws and fractures, patching and other signs of repair, indications of 
 softening or deterioration by overheating, condition of low-water indicating apparatus, 
 safety-valves, and pressure-gauges. 
 
 "A close examination of the shape of the rivet-heads and of the shapes and sizes of 
 the plates and arrangement of seams throughout the boiler will usually lead to detec- 
 tion of repairs when these are not obvious at first sight. The color and nature of the 
 fractures, and whether they be short or jagged, are the only guides to the length of 
 time they have existed and how they are produced. 
 
 "Overheating from shortness of water usually declares itself by the bulging and 
 buckling of the plates, by breaking off the incrustation on one side, and by producing 
 a burnt appearance, along with removal of soot, etc., on the other side, by the starting 
 
454 STEAM BOILERS. CHAP. XIX. 
 
 of joints and melting of fusible plugs, and in furnace- tubes also by forming corruga- 
 tions parallel with the ring-seams. These corrugations are produced by the excessive 
 expansion of the plates at the part where they occur. . . . 
 
 "One or more of the defects above indicated will in most cases be found to be the 
 cause of explosion, which may have occurred at the ordinary working pressure. But if 
 no such defects can be found, and the calculated strength of the boiler be sufficient for 
 the alleged working or blowing-off pressure, the condition of the safety-valves, levers, 
 weights, springs, double-eyes, pipes, or branches must be still more closely enquired 1 
 into and the strength of the plates at fractures carefully tested. The alleged blowing- 
 off pressure must be carefully checked by calculating the weight upon the valve, and 
 the accuracy of the pressure-gauge as well as its condition should be ascertained, and 
 anything else suggested by the nature of the case that may throw light upon the man- 
 ner in which the overpressure has been brought about." ( Wilson.) 
 
 5. Experimental Steam-boiler Explosions. A series of instructive experi- 
 ments on the explosion of steam boilers was made by the United Railroad companies 
 of New Jersey, under the direction of Francis B. Stevens, at Sandy Hook, N. J., in 
 November, 1871, in the presence of a number of prominent engineers. 
 
 The first experiment was made with a boiler of the type represented in figure 2, 
 Plate XXI. It was 28 feet long, and the cylindrical portion of the shell was 6 feet 6 
 inches in diameter and of iron a full quarter-inch thick. The boiler had been thirteen 
 years in use, and the last inspector's certificate had allowed 40 pounds of steam to be 
 carried in it. Before the final trial it had been repaired and tested by hydrostatic 
 pressure to 82 pounds per square inch without fracture, and then had been subjected 
 to a steam-pressure of 60 pounds per square inch without fracture. 
 
 On its final trial a heavy wood-fire was built in the furnaces, the water standing 12 
 inches deep over the flues. The pressure rose rapidly until it reached 90 pounds, when 
 leaks appeared in all parts of the boiler, and at 93 pounds a rent at the rear of the 
 steam-drum, where it joined the shell, became so great that the steam passed off more 
 rapidly than it was formed. After the fires were extinguished it was found that at the 
 point where the steam-drum joins the shell the latter had been drawn downwards, and 
 each crown-sheet, originally flat and stayed to the roof of the shell, had been forced 
 down and bulged, to an extent of about 2 inches, between two rows of the stays referred 
 to, pulling the outside shell with it away from the lower sheet of the vertical steam- 
 drum, thus opening a seam, venting the boiler, and preventing an explosion. 
 
 The second experiment was made on a flat rectangular box, 6 feet long, 4 feet high, 
 and 4 inches wide over all, made to represent the water-leg of a boiler. The two side- 
 
SEC 5. 
 
 BOILER-EXPLOSIONS. 455 
 
 plates were of the best flange fire-box iron, ^ inch thick. They were held together by 
 a single row of rivets at their edges, passing through a frame made of wrought-iron 
 bars 3f inches wide and 2 inches deep, mitred at their ends. The side-plates were 
 braced together every 8f inches one way and 9^ inches the other way of their surfaces 
 by screw stay-bolts of 1 inches diameter with their ends slightly riveted over. This 
 box had been subjected to a hydrostatic pressure of 138 pounds per square inch without 
 fracture, and to a steam-pressure of 102 pounds per square inch. 
 
 This box was set on one edge between walls of brick masonry, and it was filled with 
 water up to about five-sixths of its height ; a strip at the top of the box about 15 inches 
 wide projected beyond the masonry. The enclosed portion of the box was heated by 
 two small furnaces in which wood-fires were built. The pressure rose in 33 minutes 
 from to 165 pounds per square inch. 
 
 " When the pressure reached 165 pounds to the square inch the box exploded with a 
 loud report, completely demolishing the brickwork by which it was enclosed. The two 
 sides were hurled in exactly opposite directions, and to about equal distances, at right 
 angles to their surfaces. The fracture had occurred in one plate only, and was along 
 the whole riveted seam joining it to the frame. For a large part of the length of the 
 seam this plate was torn out between the rivets, and for the remaining part the rivets 
 were sheared. The other plate was not fractured nor were the bars of the frame broken ; 
 the plate and the frame remained riveted together, but not uninjured, all the bars of 
 the latter being bent considerably inwards, forming an irregular curve of from four to 
 six inches versed-sine. Both plates were bulged out irregularly, so as to be about nine 
 inches dishing, and the bulging took place near the bars. Not one of the bolts was 
 broken, and neither the threads upon their ends nor the threads in the plate were 
 stripped or injured, but the slight riveting-over of the ends of the bolts was broken off 
 in all of them." . . . "Between the bolts there was a small permanent stretching of 
 the plates, giving each space between the bolts a slightly dishing or bulged form in 
 addition to the general bulging of the plates, thus forming a system of secondary bulges, 
 as it were, and around every bolt both plates were strongly marked by a congeries of 
 circular crispations." (See ' Report of United States Naval Engineers ' and Journal 
 of the Franklin Institute, 1872, Nos. 1 and 2.) 
 
 The following account of the third experiment is taken from the report of the board 
 of United States naval engineers, consisting of B. F. Isherwood, E. S. De Luce, and 
 Sidney Albert, Chief Engineers United States Navy, detailed by the Secretary of the 
 Navy to witness these experiments : 
 
 " The boiler that was exploded during this experiment was built by T. F. Secor in 
 
456 STEAM BOILERS. CHAP. XIX. 
 
 1845, and taken out of the steamboat Bordentown in August last, after having been 25 
 years in use. When taken out the inspector's certificate allowed it to be worked with 
 a pressure of 30 pounds per square inch. It was a horizontal fire-tube boiler, with the 
 tubes returned immediately above the furnace and combustion-chamber. 
 
 " It had but one furnace, and that was 11 feet 5 inches in width, with grate-bars 7 
 feet in length. The top of the furnace and the top of the combustion-chamber were 
 flat, and braced to the flat top of the shell above them by rectangular braces 2 inches 
 by J inch in cross-section, placed 17 inches apart crosswise the boiler and 12 inches apart 
 lengthwise the boiler, each brace holding a flat surface of 204 square inches, to which it 
 was attached by crow-feet so arranged that the flat surface between the sustaining rivets 
 was 12 inches square. The flat water-spaces were braced, at intervals of 8 inches in one 
 direction and 12 inches in the other, by 1-inch diameter screw-bolts, each of which held 
 a flat surface of 96 square inches. The iron plates of the boiler were a large J inch 
 thick. . . . 
 
 " The shell of the boiler was rectangular, with the exception that the vertical sides 
 were joined to the flat top by quadrantal arcs of 37 inches radius. All the seams were 
 single-riveted. Upon the centre of the top of the boiler was a cylindrical steam-drum 
 of 6 feet diameter and 8 feet 8 inches height. The flat water-space at the front of the 
 furnace was 4J inches wide, and that at the back end of the boiler was 5 inches wide, 
 including thicknesses of metal. The width of the boiler was 12 feet 2 inches, its length 
 15 feet 5 inches, and its height, exclusive of the steam-drum, was 8 feet 6 inches. 
 
 " The shell was braced very unequally. Each upper brace, 1 inches large in diame- 
 ter, sustained the pressure upon a surface 28 by 12 inches, or 336 square inches ; and 
 each rectangular vertical brace adjacent to the sides, 2 inches by J inch cross-section, 
 sustained the pressure upon a surface 19 by 12 inches, or 228 square inches ; these were 
 the weakest places. 
 
 " The following were the grate and the water-heating surfaces of the boiler : 
 
 Grate-surface 79| sq. feet. 
 
 Heating-surface in furnace 180 
 
 " "in combustion-chamber and back-connec- 
 tion 103 
 
 " in tubes 2,171 
 
 " " in uptake 64 " 
 
 Total heating-surface 2,518 
 
SEC. 5. 
 
 BOILER-EXPLOSIONS. 
 
 457 
 
 " On the 3d of September last this boiler was subjected to a hydrostatic pressure of 
 60 pounds per square inch, when twelve crow-feet gave way. After being repaired it 
 was again subjected on the 4th of November last, when erected at Sandy Hook, to a 
 hydrostatic pressure of 69 pounds per square inch, which it bore without fracture ; and 
 on the 16th of November last it was subjected to a steam-pressure of 45 pounds per 
 square inch, which it also sustained without fracture. 
 
 " The fuel used in the experiment was wood, and the water-level in the boiler was 
 15 inches above the highest point of the tubes. When the fire had been brought to 
 steady action the pressure of the steam gradually increased at the following rate, com- 
 mencing with the pressure of 29i- pounds per square inch : 
 
 Time P.M. 
 Hours. Minutes. 
 
 Steam-pressure in pounds per 
 square inch above atmosphere. 
 
 Time P.M. 
 Hours. Minutes. 
 
 Steam- pressure in pounds per 
 square inch above atmosphere. 
 
 12 
 
 21 
 
 29* 
 
 12 
 
 3<> 
 
 46* 
 
 12 
 
 23 
 
 33^ 
 
 12 
 
 31 
 
 4 8i 
 
 12 
 
 2 5 
 
 37i 
 
 12 
 
 3 2 
 
 So 
 
 12 
 
 27 
 
 4i 
 
 12 
 
 33 
 
 5 2 
 
 12 
 
 29 
 
 44i 
 
 12 
 
 34 
 
 53i 
 
 "At the pressure of 50 pounds per square inch some of the braces in the boiler gave 
 way with a loud report, and when the pressure of 53 pounds was reached the boiler 
 exploded with terrific violence. The steam-drum and a portion of the shell attached to 
 it, forming a mass of about three tons weight, were hurled to a great height in the air 
 and fell to the earth at about 450 feet from the original position of the boiler, crushing 
 several trees in their fall. Two other large fragments fell at less distances, while smaller 
 ones were thrown much farther. Almost the whole of the boiler was literally torn into 
 shreds, which were scattered far and wide, the only portion remaining where the boiler 
 had been being the tubes. These, though considerably distorted, were otherwise unin- 
 jured. Both tube-plates had been blown from the tubes in opposite directions and at 
 the same moment, for nearly all the tubes were found lying in a heap on the ground 
 immediately beneath the place they had occupied in the boiler, the riveting of their 
 ends over the plates having been simultaneously stripped. The top of the furnace and 
 the top of the combustion-chamber, which in the boiler were immediately beneath the 
 tubes, had entirely disappeared into debris, as had also the sides and ends of the shell. 
 The boiler seems to have first yielded by the fracture of the upper row of horizontal 
 braces. The loud report heard when the pressure attained 50 pounds per square inch 
 was probably caused by their breaking. The larger masses were all thrown in one direc- 
 
458 STEAM BOILERS. CHAP. XIX. 
 
 tion at right angles to the side of the boiler, but the smaller fragments were projected 
 radially in all directions as from a centre. Two heavy bomb-proofs, constructed of large 
 timbers and sand for the protection of the other boilers, were dislodged, and a part of 
 the fence of the enclosure was destroyed by the impact of the flying fragments. The 
 crow-feet in most cases remained firmly attached to the shell, and the braces had parted, 
 probably in the welds, leaving the ends still secured to the crow-feet. The screw-bolts 
 which braced the flat water-spaces had slipped from their fastenings in the plate with- 
 out injury to the screw-threads either upon them or in the plate. The latter was per- 
 manently bulged or dished between the bolts, and this stretching of the metal had, by 
 its enlargement of the holes, allowed the screw-ends of the bolts to draw out without 
 injury to the threads either on the bolts or in the plates. 
 
 " The ground beneath, and for a considerable distance around where the boiler stood, 
 was saturated with the water of the boiler in fact, made into mud and the adjacent 
 grass and small shrubbery were so drenched that an ordinary boot was wet through by 
 walking among them. At seven minutes before the explosion took place the water- 
 gauge on the boiler was examined and found to indicate the water-level 15 inches above 
 the top of the tubes. 
 
 " The conclusions to be drawn from this experiment are the following : 
 
 " First. An old boiler, containing a large mass of water above the highest point of its 
 heating-surface, can be exploded with such complete destruction as to reduce it into 
 mere debris, and hurl the fragments in all directions with a force that no ordinary con- 
 struction of building or vessel could withstand. 
 
 "Second. That the pressure required for so devastating an explosion is the very 
 moderate one of 53f pounds per square inch. 
 
 " Third. That with only a wood-fire, generating a far less quantity of heat in equal 
 time than a coal-fire, there were required only 13 minutes to raise the pressure from the 
 inspector's working allowance of 30 pounds per square inch to the exploding pressure of 
 53f pounds per square inch, showing that a few minutes' absence or neglect of the en- 
 gineer, coupled with an overloaded or inoperative safety-valve, are all that are needed to 
 produce the most destructive steam-boiler explosion, even with an old and unequally- 
 braced boiler, in which it might be supposed a rupture of the weakest part would pre- 
 cede other fracture, and allow the escape of the pressure without doing further injury. 
 
 " Fourth. That in accounting for either the fact of an explosion or for its destructive 
 effects there is no necessity for hypotheses of low water, enormous pressures, instan- 
 taneous generations of immense quantities of steam, superheated steam, the formation 
 of hypothetical gases, development of electricity, etc., etc. The most frightful catas- 
 
SK. 5. BOILER-EXPLOSIONS. 459 
 
 trophe can be produced by simply gradually accumulating pressure of saturated steam 
 to a strain at which the strength of the boiler yields, nor need that pressure be much 
 above what is ordinarily employed with boilers of this type. 
 
 " Fifth. That there is no flashing of boiler- water into steam at the moment of an 
 explosion. On the contrary, with the exception of the small portion of this water 
 vaporized (after the reduction of the pressure owing to the rupture of the boiler) by the 
 contained heat in it between that due to the temperature of the steams of the explod- 
 ing pressure and of the atmospheric pressure, it remains unchanged, and is thrown 
 around, drenching the objects near it and scalding whomever it falls upon. 
 
 "Sixth. The weakest portion of the boiler-braces was in their welds. 
 
 " Seventh. The equal stretching in all directions of the boiler-plates between the 
 screw-bolts, due to their bulging under the pressure, was sufficient to permit the slipping 
 out of the bolts without injury to the screw-threads either upon them or in the plates. 
 
 "Eighth. That this experiment has conclusively disposed of several theories of steam- 
 boiler explosion, replacing vague conjecture and crude hypothesis with exact experi- 
 mental facts, and, by thus narrowing the field for the search of truth, has made the dis- 
 covery more probable." 
 
INDEX. 
 
 Acetic acid, Preventing incrustation by 441 
 
 Acid products of the combustion of coal 420 
 
 Acidity of water in boilers. Testing the 439 
 
 Adamson joint for furnace-flues 227 
 
 Admiralty Committee on Boilers on chemical treat- 
 ment of the feed-water of marine boilers, 438, 440 
 
 on corrosion due to fatty acids 430,431 
 
 on corrosion due to oxygen and carbonic acid 
 
 in feed-waters 425 
 
 on corrosion of iron by chloride of magnesium. 426 
 on decomposition of chloride of magnesium. . . 426 
 on use of carbonate of soda in marine boilers. . 440 
 
 on use of zinc in boilers 434 
 
 Air, Chemical and physical properties of 34 
 
 Flow of, to grate of furnace 43, 46 
 
 Moist, causing corrosion 423 
 
 Weight of, at different temperatures 45 
 
 Weight of, required for combustion of fuel 40 
 
 Weight of, required for combustion of various 
 
 substances 37 
 
 Air-admission in excess of quantity theoretically re- 
 quired 40, 49 
 
 through bridge-walls 149, 324 
 
 through furnace-doors 327 
 
 Algoma, TJ. S. S., Chimney for boilers of 295 
 
 Steam-jet for boilers of 300 
 
 American Tube-works, Seamless drawn brass and 
 
 copper tubes 270 
 
 Ammonia 34 
 
 Ammonium, Chloride of 440 
 
 Angle-iron rings, Welding 210 
 
 Testsof 150 
 
 Weight of, per foot 87 
 
 Ash, Proportion of, in coals 52, 53, 54 
 
 Specific heat of 41 
 
 Ashpans 315,328 
 
 PA6B 
 
 Ashpit, Dimensions of 136 
 
 doors, Forms of 327 
 
 doors, Begulating the draught by 48 
 
 Ashcroft furnace-door 326 
 
 grate 323 
 
 spring safety-valve 345 
 
 Back-connection see Connection, back. 
 
 Back-draught 50, 377 
 
 Barff s method of protecting iron 433 
 
 Bar-iron, flat, Weight of 85 
 
 round and square, Weight of 84 
 
 Strength of 83,99,100 
 
 Testsof 103 
 
 Baryta, Carbonate of 441 
 
 Beading the ends of tubes 273 
 
 Bearing-bars 322 
 
 Belleville water-tube boiler 283 
 
 Bending plates 159, 164 
 
 Bending-rolls 163 
 
 Berryman heater, The 354 
 
 Bilge-water, Corrosive action of 399 
 
 preventing incrustation and corrosion 441 
 
 Birmingham wire-gauge 86 
 
 Bisulphuret of carbon 34 
 
 Blast-pipe, Head produced by 47 
 
 of locomotives 299 
 
 Bleeding- valve 350 
 
 Blisters, Cutting out 404 
 
 Formation of 415 
 
 Blowing down boilers 385 
 
 Blowing-machines, Work to be done by 47 
 
 Blowing-off causing increased formation of scale 396 
 
 Loss of heat and water in consequence of .... 397 
 
 Blow-off cocks 3*5 
 
 Blow-pipes 336 
 
 461 
 
462 
 
 INDEX. 
 
 PAGE 
 
 Blow-valves 335 
 
 Boiler of steamer Estelle 130, 132, 284 
 
 of steamer Lookout 130, 156 
 
 of steamer Lord of the Isles 130, 229, 275 
 
 of U. S. S. Daylight .127, 130, 132 
 
 of U. S. S. Kansas 130, 132 
 
 of U. S. S. Lackawawna 130, 142, 264 
 
 of U. S. S. Mahaska 130, 132 
 
 of U. S. S. Miantonomoh and class 130, 144 
 
 of U. S. S. Morse 130, 132 
 
 of U. S. S. Nipsic 130,154, 289 
 
 of U.S. S.Plymouth 130 
 
 of U. S. S. Shockokon 130, 132 
 
 Belleville water-tube 283 
 
 Davey-Paxman 286 
 
 Dickerson's marine 129 
 
 Double-end cylindrical 127, 130 
 
 Drop-flue 130, 132, 260 
 
 Dry-bottom 226, 314 
 
 Emery's connected-arc marine 124, 214 
 
 Flue 130, 132, 260 
 
 Herreshofl coil 130, 132, 284 
 
 Howard marine water-tube 282 
 
 Lamb and Sumner 261 
 
 Locomotive see Locomotive boiler. 
 
 Martin's vertical water-tube 125, 130, 263 
 
 Perkins's water-tube 281 
 
 Return-flue 130, 132, 260 
 
 Stimer's differential tubular 265 
 
 Superheating 311 
 
 Vertical fire-tube 129 
 
 Water-tube 280 
 
 Boiler-building, Systems of 231 
 
 Boiler-experiments 369 
 
 Boiler-explosions see Explosions of boilers. 
 
 Boiler-iron, Brands of 78 
 
 Strength of 78,83 
 
 Boiler-keelsons 313 
 
 Boiler-mountings 319 
 
 Boiler-plates, Dimensions of 79 
 
 Examination of 105 
 
 Forge-tests of 102, 149 
 
 tests of, English Admiralty 104, 149 
 
 tests of, French Government 104 
 
 tests of, United States laws and regulations 
 
 regarding 92 
 
 with thickened edges 199 
 
 Boiler-power 122 
 
 Boiling-point of water at different pressures 71 
 
 of brine 395, 396 
 
 PAGE 
 
 Bolts for fastening braces, Strength of 242 
 
 wrought-iron, Dimensions and weights of 89 
 
 wrought-iron, Shearing strength of 252 
 
 Bowling-hoop for furnace-flues 227 
 
 Braces see Stays. 
 
 Brass boiler-tubes 74, 266 
 
 Composition and properties of 74, 83 
 
 Tenacity and ductility of, at different tempe- 
 ratures 75 
 
 Thermal conductivity of 58 
 
 tubes, Dimensions and weights of 270 
 
 Brick, Thermal resistance of 56 
 
 Bridge- walls, Air-admission through 149, 324 
 
 Calorimeter over 136 
 
 Fqrms of 230, 324 
 
 Brine, Boiling-point of 395, 396 
 
 Density of 389, 396 
 
 Brine-pumps 335 
 
 Bronze, Composition and properties of 74, 83 
 
 Tenacity and ductility of, at different tempe- 
 ratures 75 
 
 Buckled plates, Stiffness and strength of 117 
 
 Butt-joints see Joints. 
 
 Calcareous deposits, Composition of 390 
 
 Formation of 391 
 
 Thermal resistance of 56 
 
 Calcium, Chloride of 390 
 
 Other salts of see Lime. 
 
 Calking 204 
 
 Calking-tools 205 
 
 Calorific intensity of fuels 40 
 
 power of coals 38, 39, 53, 54 
 
 power of fuels, Dulong's law 38 
 
 power of various substances 37 
 
 Calorimeter for measuring actual quantity of heat in 
 
 steam 372 
 
 of chimney 137, 292 
 
 of tubes 137, 263, 265 
 
 over bridge-wall 136 
 
 Carbon, Chemical equivalent of 34 
 
 Combustion of 35 
 
 Heat developed by combustion of 37 
 
 Proportion of, in different coals 52, 53, 54 
 
 Proportion of, in iron affecting welding 207 
 
 Proportion of, in steel 79, 206 
 
 Radiation of heat from incandescent solid. ... 61 
 Temperatures of combustion of 41 
 
 Carbonic acid, Chemical and physical properties of . . 34 
 produced by combustion of coal 35, 37 
 
INDEX. 
 
 463 
 
 Carbonic acid producing corrosion of iron 423 
 
 Volume and weight of, present in sea-water. . . 389 
 Carbonic oxide, Chemical and physical properties of . 34 
 
 Combustion of 35 
 
 Heat developed by combustion of 37 
 
 produced by combustion of carbon 35, 51 
 
 Temperature of combustion of 41 
 
 Cast-iron, Physical and mechanical properties of 83 
 
 Tenacity and ductility of, at different tempe- 
 ratures 76 
 
 Thermal conductivity of 58 
 
 Use of, in boiler-construction 77 
 
 Cement, Pilling water-bottoms of boilers with 406 
 
 Iron, for leaky seams 405 
 
 Portland, a substitute for scale 407 
 
 used for boiler-beds 313 
 
 used for boiler-covering 352 
 
 Charcoal, Calorific power of 38 
 
 Combustion of 35 
 
 Charcoal-iron, American 78 
 
 Check-valves 333 
 
 Chemical equivalents of various substances 34- 
 
 Chimneys, Calorimeter of 137, 292 
 
 Fixed 294 
 
 Forms and dimensions of 291 
 
 Height of 46, 47 
 
 Hoisting 295 
 
 Radiation of heat from 46, 293 
 
 Temperature of 46, 47 
 
 Chimney-draught 45, 64 
 
 Chimney-gas, Volume of, per pound of fuel 42 
 
 Weight of a cubic foot of 45 
 
 Chimney-stays 294 
 
 Circular arcs, Stress in 119 
 
 Circular flues, Resistance to collapse of 109, 229 
 
 Circulation of water 59, 223 
 
 Defective 386 
 
 Methods of improving the 387 
 
 Cleaning boilers 398 
 
 fires, Loss of heat by 50 
 
 fires, Process of 377 
 
 Clinker 36 
 
 Coal, Calorific power of 38 
 
 Character and efficiency of various kinds of . . . 
 
 52, 53, 54 
 
 Energy developed by combustion of 39 
 
 Incombustible matter in 48 
 
 Methods of firing with different kinds of 378 
 
 Moisture absorbed by .... 49 
 
 Rates of combustion of, in boilers 43, 45 
 
 Coal, Vaporific power of, in a boiler 39, 66 
 
 Weight of air required for combustion of 40 
 
 Coke, Combustion of 35, 43, 54 
 
 Moisture absorbed by 49 
 
 Combustion, Conditions necessary for perfect 40 
 
 of coal, Acid products of the. . . 420 
 
 of constituents of fuels 35 
 
 of various substances 37 
 
 Rates of 42, 45, 134 
 
 Temperatures of 35, 40 
 
 Volume of products of 42 
 
 Combustion-chambers in different types of boiler 229 
 
 Steam generated on plates of 61 
 
 Composition see Brass and Bronze. 
 
 Conduction of heat, Laws of 55 
 
 Conductivity, Thermal, of wrought-iron determined 
 
 by Forbes 56 
 
 of several metals determined by B. F. Isher- 
 wood 58 
 
 Connections, Back, Arrangement and Construction of 230 
 
 Evaporation from surfaces of 61 
 
 Proportions of 136 
 
 Front 287 
 
 Connection-doors 291, 327 
 
 Convection of heat 55 
 
 Copper, Alloys of 74,76,83 
 
 Corrosion of 429 
 
 Galvanic action of 432 
 
 Lining and coating boiler-shells and tubes 
 
 with 433 
 
 Physical and mechanical properties of 73, 83 
 
 Thermal conductivity and resistance of 56, 58 
 
 tubes. Dimensions and weights of 270 
 
 Use of, in boilermaking 31, 73, 221 
 
 Corrosion by galvanic action 432 
 
 by oxygen and carbonic acid in water 423 
 
 by sulphuric acid in soot 419 
 
 External, of boilers 314, 413 
 
 Irregular action of, on iron plates 413 
 
 of boilers when not in use, Preventing the 407 
 
 of steam-drums of U. S. S. Swatara 431 
 
 of superheating-surfaces 310 
 
 Corrosive action of chloride of magnesium 425 
 
 of fatty acids 427 
 
 of gases of combustion 414 
 
 Corrosive ingredients of feed-waters, Neutralizing. . . 437 
 
 Corrugated flues and plates 221 
 
 | Couste on the formation of deposits in boilers 
 
 389, 391, 396 
 
 Covering for boilers 351 
 
464 
 
 INDEX. 
 
 Cylindrical shells. Resistance to internal fluid pres- 
 sure 108 
 
 Resistance to external fluid pressure 109 
 
 Experiments on collapsing strength of 113 
 
 Damper 48,293 
 
 Davey-Paxman boiler 286 
 
 Daylight, U. S. S., boiler of, Description of 127 
 
 Dimensions and weight of 130 
 
 Economic evaporation of 132 
 
 Dead-plate 319 
 
 De la Beche and Playfair, Experiments on efficiency 
 
 of coals 53 
 
 Diagonal lap-joints 198 
 
 Dickerson's marine boiler 129 
 
 Dissociation 41 
 
 Drain cocks and pipes 332, 351 
 
 Draught, Artificial 47, 298 
 
 Efficiency of boilers with 65 
 
 Experiments with 302 
 
 Natural 45, 64 
 
 of a furnace 43 
 
 Drifting 171 
 
 Drilling boiler-plates 168 
 
 Dry-pipes 331 
 
 Dudgeon's tube-expander 272, 278 
 
 Emery, C. E., Connected-arc marine boiler. .119, 124, 214 
 
 English Board of Trade rules for blow-off cocks 336 
 
 circular flues 229 
 
 cylindrical boiler-shells 218 
 
 girder-stays 242 
 
 safety-valves 343, 346 
 
 stays and stayed plates 239 
 
 tests and inspection of boilers 363, 367 
 
 English boiler-iron, Brands of 79, 83 
 
 English Navy, Preservation of boilers in the 407 
 
 Specification of boilers for the 148 
 
 Testing boilers in the 363 
 
 Tests of boiler-plates in the 102, 149 
 
 Tests of steel for ships and boilers of the 104 
 
 Use of zinc in boilers of the 43C 
 
 Use of carbonate of soda in boilers of the 439 
 
 Equivalent of heat, Mechanical 39 
 
 Equivalents, Chemical 34, 35 
 
 Erection of boilers 816 
 
 Escape-pipes 347 
 
 Estette, boiler of steamer, Description of 284 
 
 Dimensions and weights of 130 
 
 Performance of 132 
 
 Eutaw. U. S. S., superheater of, Description of 311 
 
 Efflciencyof 68 
 
 Evaporation from furnace and back-connection 60 
 
 from tube-surfaces 61, 262, 264 
 
 Influence of felting boilers on their economic. . 63 
 
 Influence of rate of combustion on the 66 
 
 in boilers of various types 123, 132 
 
 Evaporative efficiency of vertical water-tube and 
 
 horizontal fire-tube boilers 125 
 
 Evaporative efficiency of boilers, Conditions affecting 
 
 the, and measure of 64, 122 
 
 Evaporative power of coals 52, 53, 54 
 
 Expanding tube-ends 271 
 
 Expanding-tools 272 
 
 Expansion of metals by heat 83 
 
 Stress produced by 140 
 
 Explosion of boilers. Causes of 442 
 
 Experimental 454 
 
 Investigation of 453 
 
 Phenomena of 450 
 
 Theories regarding 445 
 
 Work done by 448 
 
 of locomotive boilers 452 
 
 on the Thunderer 451 
 
 on the Westfield 453 
 
 Eye-bars, Experiments on, by Charles Pox 243 
 
 by Board of U. S. Naval Engineers 253 
 
 Factors of safety in boiler-construction 139, 219 
 
 Fairbairn, Collapsing strength of cylindrical flues. . . . 109 
 
 Proportions of single-riveted lap-joints 195 
 
 Proportions of double-riveted lap-joints 197 
 
 Strength of riveted lap-joints 190 
 
 Fan-blowers, Experiments with 304 
 
 Use of, for marine boilers 298, 300 
 
 Work to be done by 47 
 
 Fatty acids, Corrosive action of, in boilers 430 
 
 Corrosive action of, on copper 429, 431 
 
 Corrosive action of, on iron 428 
 
 corrosive action of, Neutralizing the. ..431, 439, 440 
 
 Decomposition of 427 
 
 Favre and Silbermann, Heat of combustion of various 
 
 substances 36, 38 
 
 Paying-surfaces 167 
 
 Feed-pipes 335 
 
 Feed-pumps 354 
 
 Peed-valves 333 
 
 Feed-water, Chemical treatment of 437 
 
 Grease in 418, 430 
 
 Introduction of, into boilers 60, 334 
 
INDEX. 
 
 465 
 
 Feed- water. Purification of 431, 437 
 
 Temperature of 123, 417 
 
 Felt, Cowhair, covering for steam boilers 352 
 
 Loss of heat through different thicknesses of. . C2 
 Felting, Influence of, on economic evaporation of 
 
 boilers 63 
 
 Ferruling tubes to reduce their calorimeter 265 
 
 Increase of holding power by ._ 278, 279 
 
 Filter, Selden's 354 
 
 Fires, Banking 380 
 
 Cleaning 377 
 
 Starting 375 
 
 Fire-bos-iron 78, 221 
 
 Fire-room, Arrangement of 122 
 
 Air-tight 301 
 
 Fire-tube boiler, horizontal, Evaporative efficiency 
 
 of 66,125,262 
 
 Vertical 129, 262, 309 
 
 Weight and space required for 133 
 
 Firing, Rules to be observed in 376, 380 
 
 Waste of fuel from careless 49 
 
 Flame 35, 36 
 
 Flange-iron 78 
 
 Flanging 164 
 
 Flat plates, strength and stiffness of unstayed 115 
 
 Flat surfaces in boilers 138, 238 
 
 Floats 339 
 
 Flow of air to grate, Velocity of 43, 46 
 
 gases in flues, Resistance to 44, 46 
 
 solids 98 
 
 Flues, Collapse of 451 
 
 collapsing strength of. Experiments on. 109, 113, 222 i 
 
 Corrugated. 231 j 
 
 Expansion and contraction of 227, 416 
 
 Factor of safety for 140 : 
 
 Flow of gases in 44, 46 
 
 Formulae for strength of 110, 111, 229 
 
 Laying-off plates for 159, 160 
 
 Welding 208 
 
 Flue-boilers 260 ! 
 
 Foaming 381. 385 ( 
 
 Forbes, P., on the thermal conductivity of wrought- 
 
 iron 56 
 
 Fox, Charles, Experiments on fastening of braces... . 243 
 
 Fox, S., Corrugated flues and plates by 221 : 
 
 French naval boilers, Capacity of steam-room in 306 
 
 Grate bars of 320 
 
 Hoisting chimneys of 295 
 
 Removable tubes in 276 
 
 Staying of 237, 245, 248 i 
 
 French naval boilers. Tests of 302 
 
 French practice in proportioning single-riveted joints 196 
 tests for steel boiler-plates 104 
 
 Friction in riveted joints 173 
 
 of gases in flues 44, 43 
 
 Fuel, Calorific power and intensity of 38, 40 
 
 Combustion of 35, 40 
 
 Elementary constituents of 34 
 
 Incombustible matter and moisture in 48 
 
 Thickness of bed of, on grate 48, 50, 376 
 
 Waste of, in solid and gaseous state 49, 51 
 
 Furnaces, construction of, General considerations re- 
 garding 221 
 
 Corrugated flues and plates for 221 
 
 cylindrical, Construction of 226 
 
 cylindrical, Securing, in boiler 228 
 
 cylindrical, Strengthening-hoops for 227 
 
 cylindrical, strength of, Rules for 229 
 
 Draught of 43 
 
 Efficiency of 47 
 
 Evaporation from plates of 61 
 
 in rectangular boilers. Construction of 223 
 
 in rectangular boilers, Double tier of 126 
 
 in cylindrical boilers. Arrangement of 126 
 
 Proportions of 136 
 
 Temperature of 41, 60 
 
 Furnace-crowns, Collapse and repairs of 404 
 
 Staying applied to 286, 248, 251 
 
 Furnace-doors, Air-admission through 327 
 
 Construction and forms of 325 
 
 Martin's or Ashcroft's 326 
 
 Prideaux's 3>6 
 
 Fusible plugs 339 
 
 Galloway tubes 260 
 
 Galvanic action in boilers 432 
 
 Gases, Volume of, &t different temperatures 42 
 
 of combustion, Resistance to flow of, in flues. 43, 46 
 
 of combustion, Weight of, per cubic foot 45 
 
 Gauge-glasses 337 
 
 Gauges, water, Arrangement and forms of 337 
 
 water, Indications and control of 381 
 
 Steam 340 
 
 Georgeanna, steamer, Superheating apparatus of. ... 68 
 
 Girder-stays 241, 250 
 
 Glance, IT. S. tug, Furnace of boiler for 225 
 
 Grashof, Formula for collapsing strength of flues. . . . 112 
 
 Grate, Dimensions of 136 
 
 Form of 319 
 
 Indicated horse-powers per square foot of. 123 
 
464 
 
 INDEX. 
 
 Cylindrical shells, Eesistance to internal fluid pres- 
 sure 108 
 
 Resistance to external fluid pressure 109 
 
 Experiments on collapsing strength of 113 
 
 Damper 48,293 
 
 Davey-Paxman boiler 286 
 
 Daylight, U. S. S., boiler of. Description of 127 
 
 Dimensions and weight of 130 
 
 Economic evaporation of 132 
 
 Dead-plate 319 
 
 De la Beche and Playfair, Experiments on efficiency 
 
 of coals 53 
 
 Diagonal lap-joints 198 
 
 Dickerson's marine boiler 129 
 
 Dissociation 41 
 
 Drain cocks and pipes 332, 351 
 
 Draught, Artificial 47, 298 
 
 Efficiency of boilers with 65 
 
 Experiments with 302 
 
 Natural 45, 64 
 
 of a furnace 43 
 
 Drifting 171 
 
 Drilling boiler-plates 168 
 
 Dry-pipes 331 
 
 Dudgeon's tube-expander 272, 278 
 
 Emery, C. E., Connected-arc marine boiler. .119, 124, 214 
 
 English Board of Trade rules for blow-off cocks 336 
 
 circular flues 229 
 
 cylindrical boiler-shells 218 
 
 girder-stays 242 
 
 safety-valves 343, 346 
 
 stays and stayed plates 239 
 
 tests and inspection of boilers 363, 367 
 
 English boiler-iron, Brands of 79, 83 
 
 English Navy, Preservation of boilers in the 407 
 
 Specification of boilers for the 148 
 
 Testing boilers in the 363 
 
 Tests of boiler-plates in the 102, 149 
 
 Tests of steel for ships and boilers of the 104 
 
 Use of zinc in boilers of the 436 
 
 Use of carbonate of soda in boilers of the 439 
 
 Equivalent of heat, Mechanical 39 
 
 Equivalents, Chemical 34, 35 
 
 Erection of boilers 316 
 
 Escape-pipes 347 
 
 Estelle, boiler of steamer, Description of 284 
 
 Dimensions and weights of 130 
 
 Performance of , 132 
 
 Eutaw, U. S. S., superheater of, Description of 311 
 
 Efficiency of 68 
 
 Evaporation from furnace and back-connection 60 
 
 from tube-surfaces 61,202,264 
 
 Influence of felting boilers on their economic. . 63 
 
 Influence of rate of combustion on the 66 
 
 in boilers of various types 123, 132 
 
 Evaporative efficiency of vertical water-tube and 
 
 horizontal fire-tube boilers 125 
 
 Evaporative efficiency of boilers, Conditions affecting 
 
 the, and measure of 64, 122 
 
 Evaporative power of coals 52, 53, 54 
 
 Expanding tube-ends 271 
 
 Expanding-tools 272 
 
 Expansion of metals by heat 83 
 
 Stress produced by 140 
 
 Explosion of boilers. Causes of 442 
 
 Experimental 454 
 
 Investigation of 453 
 
 Phenomena of 450 
 
 Theories regarding 445 
 
 Work done by 448 
 
 of locomotive boilers 452 
 
 on the Thunderer 451 
 
 on the Westfield 453 
 
 Eye-bars, Experiments on, by Charles Fox 243 
 
 by Board of U. S. Naval Engineers 253 
 
 Factors of safety in boiler-construction 139, 219 
 
 Fairbairn, Collapsing strength of cylindrical flues. . . . 109 
 
 Proportions of single-riveted lap-joints 195 
 
 Proportions of double-riveted lap-joints 197 
 
 Strength of riveted lap-joints 190 
 
 Fan-blowers, Experiments with 304 
 
 Use of, for marine boilers 298, 300 
 
 Work to be done by 47 
 
 Fatty acids, Corrosive action of, in boilers 430 
 
 Corrosive action of, on copper 429, 431 
 
 Corrosive action of. on iron 428 
 
 corrosive action of, Neutralizing the. ..431, 439, 440 
 
 Decomposition of 427 
 
 Favre and Silbermann, Heat of combustion of various 
 
 substances 36, 38 
 
 Paying-surfaces 167 
 
 Feed-pipes 335 
 
 Feed-pumps 354 
 
 Feed-valves 333 
 
 Feed-water, Chemical treatment of 437 
 
 Grease in 418, 430 
 
 Introduction of, into boilers 60. 334 
 
INDEX. 
 
 465 
 
 Feed-water, Purification of 431,437 
 
 Temperature of 123. 417 
 
 Felt, Cowhair, covering for steam boilers 352 
 
 Loss of heat through different thicknesses of. . 62 
 Felting, Influence of, on economic evaporation of 
 
 boilers 63 
 
 Ferruling tubes to reduce their calorimeter 265 
 
 Increase of holding power by 278, 279 
 
 Filter, Selden's 354 
 
 Fires, Banking 380 
 
 Cleaning 377 
 
 Starting 375 
 
 Fire-box-iron 78, 221 
 
 Fire-room, Arrangement of 122 
 
 Air-tight 301 
 
 Fire-tube boiler, horizontal, Evaporative efficiency 
 
 of 66,125,262 
 
 Vertical 129, 262, 309 
 
 Weight and space required for 133 
 
 Firing, Rules to be observed in 376, 380 
 
 Waste of fuel from careless 49 
 
 Flame 35,36 
 
 Flange-iron 78 
 
 Flanging 164 
 
 Flat plates, strength and stiffness of unstayed 115 
 
 Flat surfaces in boilers 138, 238 
 
 Floats 339 
 
 Flow of air to grate. Velocity of 43, 46 
 
 gases in flues, Resistance to 44, 46 
 
 solids 98 
 
 Flues, Collapse of 451 
 
 collapsing strength of, Experiments on. 109, 113, 232 
 
 Corrugated 231 
 
 Expansion and contraction of 227, 416 
 
 Factor of safety for 140 
 
 Flow of gases in 44. 46 
 
 Formula; for strength of 110, 111, 229 
 
 Laying-off plates for 159, 160 
 
 Welding 208 
 
 Flue-boilers 260 
 
 Foaming 381. 385 
 
 Forbes, P., on the thermal conductivity of wrought- 
 
 iron 56 
 
 Fox, Charles, Experiments on fastening of braces 243 
 
 Fox, S., Corrugated flues and plates by 221 
 
 French naval boilers, Capacity of steam-room in 306 
 
 Grate bars of 320 
 
 Hoisting chimneys of 295 
 
 Removable tubes in 276 
 
 Staying of 237, 245, 248 
 
 French naval boilers, Tests of 362 
 
 French practice in proportioning single-riveted joints 196 
 tests for steel boiler-plates 104 
 
 Friction in riveted joints 178 
 
 of gases in flues 44, 46 
 
 Fuel, Calorific power and intensity of 38, 40 
 
 Combustion of 35, 40 
 
 Elementary constituents of 34 
 
 Incombustible matter and moisture in 48 
 
 Thickness of bed of, on grate 48, 50, 376 
 
 Waste of, in solid and gaseous state 49, 51 
 
 Furnaces, construction of, General considerations re- 
 garding 221 
 
 Corrugated flues and plates for 221 
 
 cylindrical, Construction of 226 
 
 cylindrical, Securing, in boiler 228 
 
 cylindrical, Strengthening-hoops for 227 
 
 cylindrical, strength of, Rules for 229 
 
 Draught of 43 
 
 Efficiency of 47 
 
 Evaporation from plates of 61 
 
 in rectangular boilers. Construction of 223 
 
 in rectangular boilers, Double tier of 126 
 
 in cylindrical boilers. Arrangement of 126 
 
 Proportions of 136 
 
 Temperature of 41, 60 
 
 Furnace-crowns, Collapse and repairs of 404 
 
 Staying applied to 236, 248, 251 
 
 Furnace-doors, Air-admission through 327 
 
 Construction and forms of 325 
 
 Martin's or Ashcroft's 326 
 
 Prideaux's 826 
 
 Fusible plugs 339 
 
 Galloway tubes 260 
 
 Galvanic action in boilers 432 
 
 Gases, Volume of, it different temperatures 42 
 
 of combustion, Resistance to flow of, in flues. 43, 46 
 
 of combustion, Weight of, per cubic foot 45 
 
 Gauge-glasses 337 
 
 Gauges, water, Arrangement and forms of 337 
 
 water, Indications and control of 381 
 
 Steam 340 
 
 Georgeanna, steamer, Superheating apparatus of 68 
 
 Girder-stays 241, 250 
 
 Glance, IT. S. tug, Furnace of boiler for 225 
 
 Grashof, Formula for collapsing strength of flues 112 
 
 Grate, Dimensions of 136 
 
 Form of 319 
 
 Indicated horse-powers per square loot of. .... 123 
 
468 
 
 INDEX. 
 
 Martin furnace-door 326 
 
 grate 833 
 
 Miantonomoh, U. S. S., Dimensions and weights of 
 
 boilers of 130 
 
 Specifications for boilers of 144 
 
 Stays for boilers of 247 
 
 Mineral wool. 352 
 
 Moisture in fuels 48 
 
 Deterioration of boilers due to 407, 423, 427 
 
 Morris, Tasker & Co., Lap-welded iron boiler-tubes. . 269 
 
 Morse, U. S. S., Dimensions and weights of boilers. . 130 
 Economic evaporation of boilers 132 
 
 Murphy shaking-grate 322 
 
 National Tube-works Co., Lap-welded iron boiler- 
 tubes 268 
 
 Nipsic, U. S. S., Blow- valves and pipes for 336 
 
 Chimney for 296 
 
 Dimensions and weights of boilers for 130 
 
 Peed-valves and pipes for 334, 335 
 
 Front-head and tube-plate of boilers for 159 
 
 List of materials for boilers of 154 
 
 Smoke-connections and uptakes for 289 
 
 Stay-tubes for boilers of 275 
 
 Stop-valves and steam-pipes for boilers of .331, 332 
 Water-gauges for boilers of 338 
 
 Nitrogen, Chemical and physical properties of 34, 36 
 
 Proportion of, in coals 53, 54 
 
 Oatmeal used for stopping leaks in boilers 406 
 
 Oil, Coating interior of boilers with 407 
 
 mineral, Preventing corrosion and incrusta- 
 tion by 431, 438 
 
 vegetable and animal, Decomposition of 427 
 
 Oil-cakes used to prevent incrustation 438 
 
 Olefiant gas, Chemical and physical properties of. ... 34 
 
 Combustion of 35, 37 
 
 Temperature of combustion of 41 
 
 Organic matter in feed-water 441 
 
 Overheating of plates, Deterioration caused by 414 
 
 Explosions caused by 444, 447 
 
 Oxidation of iron by superheated steam 412 
 
 Admiralty Committee on Boilers on 425, 426 
 
 Experiments by P. Crau-Calvert on 424 
 
 Experiments by Scheurer-Kestner on 424 
 
 Oxide of iron, Black 433 
 
 Oxygen, Chemical and physical properties of 34 
 
 Corrosion of iron by 423 
 
 present in feed-water causes corrosion 425 
 
 present in fuel, Effect of 36,37 
 
 Oxygen, present in various coals 53, 54 
 
 Volume of, dissolved in sea-water 425 
 
 Weight of, required for combustion of various 
 substances 37 
 
 Painting boilers 400 
 
 Patching boilers 403 
 
 Pauksch's boiler-tubes 276 
 
 P6clet on draught of boilers 43 
 
 on transmission of heat 58 
 
 Percussion, Eifects of, on strength of iron and steel. 172 
 
 Percussion water-gauge 389 
 
 Perkins's tubulous boiler 281 
 
 Petroleum, Preventing corrosion and incrustation by 438 
 
 Phosphor in iron and steel 78, 206 
 
 Phosphor-bronze 74, 76 
 
 Pitch of rivets in single-riveted lap joints. .. 188, 195, 196 
 
 in double-riveted lap-joints 189, 197 
 
 in double- welt butt-joints 201, 202 
 
 Pitting 413 
 
 Planing boiler-plates 163 
 
 Plate-iron, Brands of, used in boilermaking 78 
 
 Weight of 84, 86 
 
 Plymouth, U. S. S., Chimney of 295 
 
 Dimensions and weights of boilers of 130 
 
 Superheaters of 809 
 
 Potash used for purifying feed- water 439 
 
 Pressure, External fluid 109 
 
 Internal fluid 107, 108, 119 
 
 Prideaux's furnace-door 326 
 
 Prosser's expanding-tool 272, 278 
 
 Providence Steam-engine Co.'s steam-riveting ma- 
 chine 172 
 
 Pumps, Brine ' 335 
 
 Feed 354 
 
 Relative efficiency of injectors and 359 
 
 Punch, Forms of 166 
 
 Sizes of die and 166 
 
 Punched and drilled rivet-work, Strength of 203 
 
 Punched holes, Shape of 167 
 
 Strained zone around 167, 182 
 
 Punching, Loss of tenacity due to 167, 184 
 
 Power required for 167 
 
 Process of 165 
 
 Quinnebaug, U. S. S., Chimney for 298 
 
 Radiation of heat from felt-covered boilers and pipes. 62 
 
 from boilers and pipes, Means of preventing.. . 351 
 
 from incandescent carbon 61 
 
 Laws of 55 
 
INDEX. 
 
 469 
 
 PAGK 1 
 
 Rankine, W. J. M., Formula for efficiency of boilers. 65 I 
 
 Formula for efficiency of heating-surfaces 59 
 
 Rule for area of safety-valves 343 
 
 Repairing boilers 403 
 
 Reverse- valve 351 
 
 Rivets, corroded, Specimens of 417 
 
 Crushing action between plates and 185, 186 
 
 Dimensions of 187, 195, 203 
 
 Formsof 176 
 
 Leaky 402 
 
 Oval 199 
 
 Pitch of 188, 189, 301 
 
 Shearing strength of 184, 192 
 
 Steel 81, 176 
 
 Testing 103 
 
 Weights of 91 
 
 Rivet-holes, Countersunk 177 
 
 Half-blind 171 
 
 Laying off 158 
 
 Punching and drilling 167, 168 
 
 Rivet-iron 83 
 
 Rivet-points, Forms of 176 
 
 Length of shank required for 177 
 
 Riveted joints see Joints, riveted. 
 
 Riveting, Chain and zigzag 178, 189 
 
 Cold 175 
 
 countersunk, Strength of 191 
 
 Hand and machine 171, 173, 175 
 
 over tube-ends 273, 278, 279 
 
 Riveting-machines, Steam and hydraulic 172, 173 
 
 Rodman testing-machine 93 
 
 Saddles for cylindrical boilers 315 
 
 for dry-bottom boilers 314 
 
 Safety, Factors of 139, 219 
 
 Safety-valves, Area of 342, 343 
 
 Arrangement and forms of 341, 343 
 
 Construction of, in accordance with regulations 
 of supervising inspectors of steam-vessels. . . . 344 
 
 Derangement of 443 
 
 Effective opening of 342 
 
 lever, Formulae relating to 346 
 
 lever, Practical method of loading 347 
 
 spring, Rules of Board of Trade (English) for. 346 
 
 Sal-ammoniac 440 
 
 Saline matter of sea-water. Constituents of 388 
 
 Salinometer, Description of the 393 
 
 Graduating the 394 
 
 pots 340 
 
 Salt, common, Weight of, contained in sea-water. . . . 389 
 
 Saponaceous deposits in land boilers 418 
 
 Saturation of water in boilers, Testing the 393 
 
 with regard to sulphate of lime 395 
 
 Scale in boilers, Composition of 390 
 
 Formation of 391 
 
 Preventing the formation of 393, 435, 437 
 
 Protection against corrosion by 406 
 
 Scaling boilers 400 
 
 Screw-stays 243 
 
 Experiments on 255 
 
 Sea-water, Density of 388 
 
 Distilled 423, 435 
 
 Salts contained in 389 
 
 Specific heat of 395 
 
 Volumes of gases dissolved in 425 
 
 Sectional boilers 280 
 
 Selkirk's tube-beader 273 
 
 Sellers & Co., Wm., Portable riveters by 174 
 
 Self-adjusting injectors by 359 
 
 Size of die in punching-machines of 166 
 
 Setting boilers 313 
 
 Shaw's spiral nozzles 350 
 
 Shearing boiler-plates 163 
 
 Shearing strength of bolts and pins in braces 242 
 
 of rivets 184, 193 
 
 of wrought and cast iron 83 
 
 of wrought-iron bolts, Experiments on 252 
 
 Shell of boilers, cylindrical, Arrangement of tubes 
 
 and furnaces in 126 
 
 cylindrical, Building and constructing. 170, 216 231 
 
 cylindrical, Laying off plates for 159, 160, 161 
 
 cylindrical, Rules for strength of 217 
 
 cylindrical, Rupture at longitudinal seams. . . . 453 
 cylindrical, Strains caused by differences of 
 
 temperature 417 
 
 cylindrical, Weakened by steam-domes. 308 
 
 cylindrical, Weakened by manholes 329 
 
 cylindrical, Welding the seams of 209 
 
 Formsof 123,214 
 
 Quality of iron for 78, 215 
 
 Rectangular 123, 126, 215 
 
 Shell-iron 78, 215 
 
 Shock, W. H., Experiments on holding power of 
 
 boiler-tubes 277 
 
 Experiments on influence of hammering on 
 
 tenacity and ductility of wrought-iron bars. 100 
 Experiments on shearing wrought-iron bolts. . 252 
 Shoekokon, IT. S. S., Dimensions and weights of 
 
 boilers of 130 
 
 Economic evaporation of boilers of 132 
 
470 
 
 INDEX. 
 
 Silicon in iron 78, 206 
 
 Smoke, Formation of 36 
 
 Losses due to formation of 51 
 
 Smoke-connections see Connections, front. 
 
 Smoke-pipe see Chimney. 
 
 Socket-bolts 243 
 
 Leaky 403 
 
 Soda, Arseniate, hyposulphite, oxalate, and phos- 
 phate of 440 
 
 Sulphate of 439 
 
 Tannate of 441 
 
 Use of, for preventing corrosion of boilers. . . . 407 
 Use of, for purifying feed-water 439 
 
 Sodium, Chloride of, in sea-water 389, 890, 391 
 
 Soot, Formation and character of 36, 420 
 
 Sulphuric acid in 419 
 
 Specific gravity of coals 52, 53, 54 
 
 Specific heat see Heat, specific. 
 
 Specifications of boilers 141 
 
 for boilers of the English navy 148 
 
 for boilers of U. S. S. Lackawanna 142 
 
 for boilers of U. S. S. Miantonomoh 144 
 
 Spherical forms in boiler-construction 138 
 
 shell, Resistance of, to internal fluid pres- 
 sure 107 
 
 Spheroidal condition of water 415, 446 
 
 Stay-bolts, Experiments on screw 255 
 
 Forms and dimensions of 243, 259 
 
 Leaky 403 
 
 Stay-domes 252 
 
 Stayed plates, Experiments on strength of 255, 258 
 
 Formulas and rules for strength of 238, 239, 258 
 
 Strains in, caused by differences of tempera- 
 ture 418 
 
 Staying, Methods of 234 
 
 Stay-plates , 251 
 
 Stays, Corrosion of 238 
 
 Experiments to determine proportions of pins, 
 
 eyes, and shanks of 253 
 
 Factor of safety for 141, 238 
 
 Fastenings of 242, 247, 249 
 
 Fitting and adjusting 249 
 
 for chimneys 294 
 
 Girder 241, 250 
 
 Gusset 241, S51 
 
 Proportioning 237 
 
 Rupture of 451 
 
 Strains on 117,119,237 
 
 Various forms of 245 
 
 Stay-tubes 274 
 
 Steam decomposed by heat 446 
 
 Dry 67 
 
 Properties of 71 
 
 superheated, Boiler explosions ascribed to 447 
 
 superheated, Density of 67 
 
 superheated, Dynamic efficiency of 67 
 
 superheated, Isherwood's experiments with.. . . 68 
 
 superheated, Oxidation of iron by 412 
 
 Weight of, discharged from an orifice per 
 
 second 342 
 
 Steam-drums, Arrangement and forms of 307 
 
 Corrosion of 431 
 
 Laying off plates for 162 
 
 Weakening of cylindrical boiler-shells by 308 
 
 Steam-jet, Efficiency of 300 
 
 Experiments with 302 
 
 Forms and arrangement of 300 
 
 Koerting's 301 
 
 Steam-pipes 332 
 
 Steam-room, Capacity of 306 
 
 Height of 307 
 
 in different boilers 130 
 
 Insufficient 306 
 
 Steam stop-valves 331 
 
 Steel bars, Appearance of fracture of 101 
 
 boilerplates, Experiments en strength of riv- 
 eted joints of 193, 204 
 
 boiler-plates. Tenacity of 80, 82, 83 
 
 Corrosion of 80, 413 
 
 Physical and mechanical properties of 83 
 
 plates, Annealing 81, 167, 416 
 
 plates, Power required to punch 167 
 
 plates, Weightof 86 
 
 plates, Welding 206 
 
 rivets , 81,83, 176 
 
 Tenacity and ductility of, at various tempera- 
 tures 76, 415 
 
 Tests of 82, 104 
 
 tubes 267 
 
 Use of, in boiler-construction 79, 80, 81 
 
 Stimer's differential tubular boiler 265 
 
 Stop-valves 331 
 
 Strains due to variations and differences of tempera- 
 ture 416 
 
 in riveted joints 180, 191 
 
 Strength, Apparent, of riveted plates 180, 192 
 
 Apparent, of test-specimens 96 
 
 of iron and steel, Effect of percussion on 172 
 
 of materials in riveted joints 183 
 
 of various metals 83 
 
INDEX. 
 
 471 
 
 PAOE | 
 
 Strength of wrought-iron bars, Effect of rolling and 
 
 hammering on , 98 
 
 Shearing, of wrought-iron bolts 253 
 
 Strengthening-hoops for furnace-flues 227 
 
 Strengthening plates and rings around manholes 329 
 
 Stress, Effects produced by 97 
 
 Stub's wire-gauge 271 
 
 Sulphates of lime and magnesia see Lime and 
 Magnesia. 
 
 Sulphur, Chemical and physical properties of 34 
 
 in fuel 36, 53, 54 
 
 in iron 77 
 
 Sulphuretted hydrogen, Properties of 34 
 
 Sulphurous acid, Chemical and physical properties of. 34 
 
 Sulphuric acid in soot 419 
 
 Superheated steam see Steam, superheated. 
 
 Superheated water in boilers. ... 447 
 
 Superheaters, Arrangement and forms of 68, 309 
 
 Corrosion of 310 
 
 Efficiency of 69, 310 
 
 Lloyd's rule for cylindrical shell of 218 
 
 Superheating, Methods of 69 
 
 Superheating boilers of United States naval vessels. . 311 
 
 Superheating steam-pipes 333 
 
 Supervising inspectors of steam-vessels, Rules of, con- 
 cerning fusible plugs 339 
 
 concerning hydraulic tests of boilers 364 
 
 concerning safety-valves 343, 344 
 
 concerning temperature of feed- water 417 
 
 concerning tests of boiler-plates 92 
 
 concerning water-gauges 337 
 
 Swaging tubes 265 
 
 Swafara, U. S. S., Corrosion of steam-drums 431 
 
 Sweating of boilers 427 
 
 Tallow, Decomposition of 428, 429, 431 
 
 Use of, in boilers 386, 408 
 
 Tannate of soda 441 
 
 Tannic acid 440 
 
 Temperature, Differences of, in a steam boiler 60, 417 
 
 differences and variations of, Strains produced 
 
 by 140, 416 
 
 Influence of, on tenacity of metals 75, 415 
 
 Influence of. on limit of saturation 395 
 
 of chimney of marine boilers 47, 66 
 
 of combustion of various substances 40 
 
 of feed-water 123, 417 
 
 of furnace in marine boilers. . . 41, 60 
 
 of gases in tubes of boilers, Differences of. . . 62 
 of gases in uptake, Method of determining 372 
 
 Temperature of ignition 35 
 
 of steam at different pressures 71 
 
 Testing-machine, Rodman's 93 
 
 Tests of boiler-plates for English naval vessels 149 
 
 of boiler-plates, Forge 102 
 
 of boiler-plates, U. S. laws and regulations 
 
 about 92 
 
 of boilers by expansion of water 366 
 
 of boilers, by steam 366 
 
 of boilers, Hammer 369 
 
 of boilers, Hydraulic 363 
 
 of boilers, Laws and regulations regarding. . . . 362 
 
 of steel for boilefs 82, 104 
 
 Test-specimens, Form and dimensions of 96 
 
 Thermal conductivity, Formulae for 56 
 
 of wrought-iron, Forbes on 56 
 
 of various metals, Isherwcod on 58 
 
 Thermal resistance of various metals 66 
 
 of plates and tubes in a boiler 59 
 
 Thermal unit, British 37 
 
 Thermometers 341 
 
 Thurston, R. H., Formula for area of safety-valves. . 343 
 
 Work done in exploding a boiler 448 
 
 Tin, Alloys of copper and 74, 83 
 
 Proto-chloride of 440 
 
 T-iron rings for furnace-flues 227 
 
 Tests of 151 
 
 used for staying and stiffening boiler-plates. 236, 251 
 
 Weight of 88 
 
 Trenton, U. S. S. , Smoke-connections and uptakes of 290 
 
 Tube-beader, Selkirk's 273 
 
 Tube-brushes 398 
 
 Tube-expanders 272 
 
 Tube-scrapers 398 
 
 Tube-sheets, Bulged 405 
 
 Drilling holes in 271 
 
 Laying off, for boilers of U. S. S. Nipsic 159 
 
 Raymond's recessed 272 
 
 Stay-rods for 274, 276 
 
 Tubes, Arrangement of, in marine boilers 125, 262 
 
 Brass 74, 266 
 
 Calorimeter of 137, 263 
 
 Collapse of 451 
 
 conical, Laying off plates for 160 
 
 Copper 266 
 
 Dimensions and spacing of 138, 262 
 
 Drawn seamless 267 
 
 Evaporation from 61, 263, 264 
 
 Expanding 271 
 
 Ferruling and swaging 265, 278 
 
472 
 
 INDEX. 
 
 Tubes, Fire 264 
 
 fitting, Bad workmanship in 368 
 
 Galloway 260 
 
 Hanging 286 
 
 Lap- welded iron 266, 268 
 
 Leaky 383, 405 
 
 Pauksch 276 
 
 Plugging leaky 383 
 
 Kemovable 276 
 
 Scaling 401 
 
 Secured by various methods, Holding power of 277 
 
 Stay 274 
 
 Steel .- 267, 268 
 
 Sweeping 398 
 
 Water 263, 266 
 
 Tweddell's hydraulic machine-tools 173 
 
 hydraulic tube-expander. 273 
 
 Uptakes, Arrangement of 287 
 
 must be air-tight 291 
 
 of cylindrical boilers 289 
 
 of rectangular boilers 288 
 
 Securing chimney to 294 
 
 U. S. laws regarding factor of safety in marine boilers 140 
 
 regarding inspection of marine boilers 366 
 
 regarding tests of boiler-plates 92 
 
 regarding tests of marine boilers 362 
 
 U. S. naval boilers, Ashpit-doors of 328 
 
 Blow-valves of 335 
 
 Chimneys of 295, 296, 298 
 
 Connection-doors of 327 
 
 Furnace-doors of 325 
 
 Grate-bars of 320 
 
 Hydrometer used for 393 
 
 Manhole-covers for 330 
 
 Rectangular 215, 223, 225 
 
 Regulations for care and preservation of 408 
 
 Removable tubes used in 277 
 
 Saddles for 314, 315 
 
 Salinometer-pots of 340 
 
 Specifications for 142, 144 
 
 Superheaters of 309, 311 
 
 Tests of 362 
 
 Water-gauges for 338 
 
 Vacuum-valve 351 
 
 Vapor, aqueous, Chemical and physical properties of 34 
 
 Vaporific power of coals 39, 52, 53, 54 
 
 of dry pine-wood 52 
 
 of various substances 37 
 
 Ventilators, Fire-room, of U. S. S. Miantonamoh 148 
 
 Volume of air 34, 42 
 
 of gases at different temperatures 42 
 
 of steam at different pressures 71 
 
 of various gases 34 
 
 Washers, Standard sizes and weight of 90 
 
 Water, Circulation of, in boilers 59, 223, 386 
 
 decomposed by heat alone 446 
 
 Distilled sea 423, 425 
 
 of different seas, Analysis and density of 388 
 
 present in fuels 36 
 
 Chemical and physical properties of 34, 71 
 
 Spheroidal condition of 446 
 
 Superheated 447 
 
 Water-bottom of boilers 225 
 
 Filling, with cement 406 
 
 Water-bridge 230 
 
 Water-gauges 337, 381 
 
 Water-gauge glasses 337 
 
 Water-legs 226 
 
 Water-level 138 
 
 Water-room in boilers 306 
 
 Water-saponification 427 
 
 Water-spaces in boilers 137 
 
 Water-tubes, Horizontal 280 
 
 Vertical 125, 263, 266 
 
 Weight of atmospheric air per cubic foot 34, 45 
 
 of atmospheric air and oxygen required for com- 
 bustion 37, 40 
 
 of boilers of a given power 133 
 
 of boilers of various types 130 
 
 of chimney-gas per cubic foot 45 
 
 of drawn brass and copper tubes 270 
 
 of flat bar-iron per foot 85 
 
 of lap- welded iron boiler- tubes 268, 269 
 
 of rivets 91 
 
 of sheet and plate iron 86 
 
 of steam 71 
 
 of steel plates 86 
 
 of various gases per cubic foot 34 
 
 of various metals per cubic foot 83 
 
 of wrought angle-iron 87 
 
 of wrought-iron bolts 89 
 
 of wrought-iron plates and bars 84 
 
 of wrought T-iron 88 
 
 Welded joints, Strength of 211 
 
 Styles of. 208 
 
 Welding, Theories and instructions regarding 206, 210 
 
 angle-iron rings 210 
 
INDEX. 
 
 473 
 
 Welding boiler-plates 207 
 
 cylindrical boiler-shells 809 
 
 front plates of boilers 210 
 
 furnace-flues 208 
 
 Weslfield, steamer, Boiler-explosion on 453 
 
 Wire-gauge, Birmingham 86 
 
 Stub's 271 
 
 Wood, Calorific power of 37, 52 
 
 Moisture in 49 
 
 Workmanship, Examples of bad 367 
 
 Rules of Board of Trade (English) regarding. . 218 
 
 Wrought-iron, Appearance of fractures of 101 
 
 Corrosion of 31 
 
 Ductility of 83, 100 
 
 Ductility and tenacity affected by hammering 
 and rolling 98 
 
 Wrought-iron, Ductility and tenacity at different 
 
 temperatures 76, 415 
 
 Physical and mechanical properties of 83 
 
 Thermal conductivity of 56, 58 
 
 used in boiler-construction 31, 77 
 
 Wyoming, U. S. S., Experiments with boilers of 266 
 
 Tcmtie, U. S. S., Supports for boilers of 315 
 
 Zinc, Alloys of copper and 74, 83 
 
 Corrosion and incrustation prevented by 434 
 
 Instructions regarding its use in English naval 
 
 boilers 436 
 
 Thermal resistance of 55 
 
 OF THE 
 
 UNIVERSITY 
 
 OF 
 
Plate L 
 
 1 
 
 =n 
 
 -- 2 
 
 -i-^n-r--!! "-^i^-. 
 
 r> 
 
 "i 
 
 i . n i 
 
 - 1 t ^Vn V 
 
 n ij. : 
 
 -^n 
 
 :& 
 
 Hfi 
 ^g 
 
 ^pO;La33l-T^>l 
 
 -^ ' oa ^ 
 
 n 
 <> 
 
 1 ^J HH 
 
 JLX-l 
 
OF THE 
 
 UNIVERSITY 
 
 OF 
 
Experiments on Tens 
 Conducted at tfu 
 
 CMefEngit 
 
 
 Bl. 1 B2. B3. Bl. 1 // 
 
 < 
 
 11 
 
 ! 
 
 
 <i 
 
 e4 
 
 4 
 
 Jl 
 
 
 
 
 tl 
 
 <! 
 
 rf 
 
 ! 
 
 r 
 
 
 1 
 
 -; 
 <! 
 
 * 
 
 
 
 
 
 ( 
 N 
 
 5 
 
 - 
 
 5 
 4 
 
 
 i 
 
 i 
 
 5 
 < 
 
 ;c 
 
 ^ 
 
 
 
 1 
 
 <i 
 < 
 
 fill I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Bl. B2. B3. Bt. 
 
 69 
 
 C3. 
 
 C3. 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.625 
 
 0.648 
 
 0.629 
 
 0.620 
 
 0.634 
 
 0.778 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.307 
 
 0.330 
 
 0.311 
 
 0.302 
 
 0.316 
 
 0.475 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 4.285 
 
 4.284 
 
 4.275 
 
 4.302 
 
 4.250 
 
 4.190 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 0.410 
 
 0.409 
 
 0.400 
 
 0.427 
 
 0.375 
 
 0.315 
 
 
 
 
 
 
 
 
 27100 
 
 27000 
 
 26500 
 
 26600 
 
 26450 
 
 28050 
 
 10000 
 
 20450 
 
 14000 
 
 18000 
 
 10250 
 
 16850 
 
 13175 
 
 54200 
 
 540OO 
 
 53000 
 
 53200 
 
 52900 
 
 56100 
 
 20000 
 
 40900 
 
 28000 
 
 36000 
 
 20500 
 
 33700 
 
 26350 
 
 
 ^ rk-ft>^>)ta' ^/ 
 
 V. ro -f ty ff 
 
 Jot/OO 
 
 ai%ff5 
 
 rfU-L'J i .O 
 
hngth of Wrought Iron, 
 mngton Navy Yard, 
 
 : 
 
 **'. Shock U. S. JV. 
 
 Plate II. 
 
 ' <.'-. C3. C4. 
 
 S 
 
 * 
 
 T> 
 
 -f 
 
 pi 
 
 rt 
 
 
 
 SKETCH 
 OF 
 SPECIMENS 
 BEFORE 
 TESTING. 
 
 1 "l 
 
 ( } 1 
 
 1 IVi 
 
 rfi ir 
 
 y. 
 
 i rf 
 
 RlRfl 
 
 f-S 
 
 V ' 
 
 
 S S 
 
 
 
 
 5 
 
 Q ( 
 
 ! l 
 
 
 2 
 
 
 il! 
 
 il i 
 
 n J 
 
 n 
 
 > 
 
 r~ \ ^~ 
 
 ' -. 
 
 \ } ' 
 C3. C4. 
 
 1 
 
 
 "-N .^- --S 
 
 U-j 
 
 ri 
 
 H 
 
 
 
 3 
 
 
 rf 
 
 J-T 
 
 J^ 
 
 - v 
 
 
 
 
 iQI 
 
 B 
 
 i M 
 
 
 
 
 
 
 
 so 
 
 H 
 
 ___ 
 
 w 
 
 
 " 
 I 
 
 n 
 
 
 H 
 
 __1 
 X 
 
 
 
 
 o o 1 
 
 Ji^ 
 
 Q 
 
 5 
 
 Is h 
 
 
 N 
 
 
 TV -i 
 
 LJ^r 
 11 
 
 U 
 _J~T 
 
 | 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 74 
 D2. 
 
 71 
 
 D3, 
 
 \ ( 
 
 ii 
 
 63 
 
 63* 63"' 63 
 
 ft 
 
 63 
 E6. 
 
 1 
 
 E6. 
 
 -^^ ^^ 
 
 6S 
 E7. 
 
 E7. 
 
 6S 
 
 SKETCH 
 OF 
 SPECIMENS 
 AFTER 
 TESTING. 
 
 J 1 D1 - 
 
 E3. 
 
 E4. E5. 
 
 1 E8. 
 
 E3. 
 
 K 
 
 I 
 
 i 
 
 E.*. 
 
 II 
 
 H W 
 
 1 
 
 -i | Dl. 
 
 m 
 
 D3. 
 
 
 r 
 
 El. 
 
 E2. 
 
 E4. 
 
 
 
 
 0.798 
 
 0.798 
 
 0.798 
 
 > 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 0.798 
 
 O.798 
 
 DIAMETER 
 BEFORE TESTING. 
 
 0.775 
 
 0.7 5 O 
 
 0.770 
 
 0.745 
 
 0.650 
 
 0.640 
 
 0.635 
 
 0.616 
 
 0.643 
 
 0.625 
 
 0.627 
 
 0.625 
 
 DIAMETER 
 AFTER TESTING. 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.000 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0.500 
 
 0300 
 
 0.000 
 
 0300 
 
 AREA OF SECTION 
 BEFORE TESTING. 
 
 0.472 
 
 0.442 
 
 0.466 
 
 0.436 
 
 0.332 
 
 0.322 
 
 0.317 
 
 0.298 
 
 0.325 
 
 0.307 
 
 0.309 
 
 0.307 
 
 i AREA OF SECTION 
 AFTER TESTING. 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3 
 
 .875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 3.875 
 
 LENGTH 
 BEFORE TESTING. 
 
 4.175 
 
 4.265 
 
 4.205 
 
 4.265 
 
 4 
 
 .200 
 
 4.29O 
 
 4.290 
 
 4.290 
 
 4.300 
 
 4.290 
 
 4.265 
 
 4.349 
 
 LENGTH 
 AFTER TESTING. 
 
 0.300 
 
 0.390 
 
 0.330 
 
 ' 0.39O 
 
 0.325 
 
 0.415 
 
 0.415 
 
 0.415 
 
 0.425 
 
 O.415 
 
 0.390 
 
 0.474 
 
 ELONGATION. 
 
 2Z75t 
 
 25300 
 
 25300 25700 
 
 262OO 
 
 26200 
 
 26000 25900 
 
 2650O 25700 
 
 25950 26000 
 
 ABSOLUTE 
 BREAKING STRAIN 
 
 S150i 
 
 ) 50600 
 
 0601 
 
 ) 5^00 
 
 524OO 
 
 52400 
 
 52000 51800 53000 51400 
 
 51900 5200C 
 
 > BREAKING STRAIN 
 PER SQUARE INCH. 
 
 
 !~rt - 
 
 12.5 
 
 
 
 
 
 / AVER. BREAK, STRAIN 
 PER SQUARE INCH. 
 
 &JAJ4& - 
 
 
 O<i J- 
 
 
 
 
 
 MJttCLL A arRUTHERS, EHG'S, M.Y. 
 
Kg. 1. 
 
 Plate III. 
 
 BOILER FOR U. S. S." DAYLIGHT." 
 
 9 - 
 
 /OOOOOOOOOO ; OOOOOiOO OOO O0OOOOOOO' 
 
 poooooooooJpooootaoooo Sboooooooi 
 
 OOOOOOOOOO |OOOOOJOO OOO OOOOOOOOOOi 
 
 oooooooooo ! oooooooooo oooooooooo 
 loop oo oo ooo iooooo'oo ooo ooo oo oo ooo] 
 
 o 
 
 TO 
 
 15- 
 
 
 
 
 ,15" 
 
 i -8 
 
 3" 
 
 * 
 
 
 
 
 
 
 
 
 
 IjL \f >. 
 
 \ 
 
 
 T 
 i 
 
 i 
 
 i 
 
 \ 
 
 | 
 
 
 
 i 
 i 
 J 
 
 
 I] 
 
 i 
 
 
 
 | 
 
 
 
 ' 
 . 
 
 
 00 
 
 
 -2-0 
 
 ,1 I 
 
 
 
 
 i 
 
 'I 1 
 
 i 
 
 
 f^T 
 
 CD 
 
 
 6'-3- - - -19- ^ 
 
 -OD 
 '^ ** tj' " ' " 
 
 f 
 
 
 i 
 
 
 !T 
 
 
 -6-8 
 
 
 e-< 
 
 d 
 
 i 
 
 i 
 
 
 "s 
 
 i 
 
 
 ~;r- 
 
 ^ 9'll" '' 
 
 Kg. 2. 
 
 TWO BOILERS FOR U. S. S. " KANSAS.' 
 
 !^ 7 -'37UF-" : 
 
 ; O : : O '.' '.': O O O O O O O O :h '.'; O O O O O O O O O O O G 
 i O Oii C-O-Q.O O O O O '.'i 
 
 8 - ii- " l; 
 
 ^/^/^ /-w\ /^i? 
 
 oooo;oooc 
 
 OOOQ-OOOC 
 
 oooqpoo 
 
 OGOCTOOCK 
 OOOOIOOOC 
 -..-. OOOO.OO OC 
 
 jjooooooivo \Sp_oop bop/ o 
 
 'o 
 
 ;'.VOOOOOOOO 
 
 coooooooo 
 
 '- 
 
 -3'- 4"- 
 
 ? 
 
 -6^-10'-, 
 
 ig. 3. 
 
 TWO BOILERS FOR U. S. S. " MAHASKA." 
 
 , 610- IO-38? 
 
Plates IY. & V. 
 
 1 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 1 
 
 < 
 
 
 to. 
 
 
 
 s 
 
 ** 
 
 
 
 
 
 o 
 o 
 
 z 
 
 H 
 O 
 
 z 
 
 n 
 3D 
 
 m 
 
 O 
 
 Z 
 -I 
 
 O 
 O 
 
 o 
 
 2 
 o 
 
 < 
 
 m 
 
 -o 
 
 > 
 
 (0 
 
 CO 
 
 O 
 
 O 
 O 
 O 
 
 2 
 O 
 
 < 
 m 
 
Plate VI. 
 
 BOILER FOR U. S. S. "LACKAWAh 
 
 CI1 120 1 2 3 4 5 6 
 
 Scale: N^H- 1- ! I \ h= t i 1 
 
 11-9- ^ 
 
 -7-10- 
 
 rf 
 
 
 
r 
 
Of THE 
 
 UNIVERSITY 
 
 Of 
 
o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 BOILER FOR U. S. S. "LACKAWAN 
 
 \ 
 
 Scale: 
 
 12 6 1 2 3 
 
 Section A,- B. 
 
Plate YH. 
 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 6 OOOOO O 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 OOOOOOO 
 
 oooooo o 
 
 OOOOOOO 
 
 8 ooooo o 
 oooooo 
 
 OOOOOOO 
 
 00000 o 
 oooooo 
 ooooo o 
 
 OOOOOOO 
 OOOOOOO 
 OOOOOO O 
 O OOOOO O 
 
 oooooo 
 oooogo. 
 
 00000000 o 
 OCOQ O OOOO 
 ooooooooo 
 oooooooooo 
 
 8 OOOOOOOOO 
 ooooooooo 
 oooooooooo 
 oooooooooo 
 ooooooooco 
 
 OOOOOOOGOO 
 
 oooooooooo 
 
 OOOOOOOOOO 
 
 oooooooooo 
 oooooooooo 
 oooooooooo 
 
 " OOOOOOOO 
 
 oooooooo 
 oooooooo 
 
 "OOOOOO 
 
 oooooo 
 oooooo 
 
 ~ OOOO 
 
 oo 
 o 
 
 ; 
 
 oo 
 
 8 
 
 oOcoo" - 
 ooooo _ _ 
 
 OOOOOQO 
 
 OOOQ8 8888 
 8888 
 
 gQQQ 
 
 _ooo5oo66 , 
 
 OOOOOQOOOO 
 
 ooooooooo ~ 
 
 00_ 
 
 ..88888 
 
 ooooooooo 
 
 OQ OOOOOOO 
 
 oooooocoo 
 
 uOOOOOOOOO 
 
 o ooooooooo 
 oooooooooo 
 
 8 OOOOOOOO 
 oooooooo 
 
 . 
 
 6 
 
 
 o 
 
 OOOO 
 
 ooooo_ 
 ooooooOoo 
 
 OQOOOOOOQ- 
 
 oooooooooo 
 
 OOOOOOOOOO 
 
 ooooooooc 
 oooooooooo 
 9QQ5QOQQQQ 
 
 
 
 ooooo 
 
 80 o o o 
 o oo o 
 00 ooo 
 ooooo 
 
 ooooo 
 ooooo 
 
 oo5o 
 
 , oo 
 
 OOOOOOO 
 
 - OOOOOOO 
 
 oooooooo 
 
 _ OOOOOOO O 
 OOOOQOOOOO 
 
 oooooooooo 
 oo 
 
 ooo 
 
 oo 
 
 OOOOOOOOO 
 
 ooooooooo 
 oooooooooo 
 ooo oc ooooo 
 
 OOOOOOOOOO 
 rS ^O O O QD O O 
 
 
 ooooc 
 o oooc 
 ooooc 
 
 8 ooooc 
 ooooc , 
 ooooooc 
 
 8 oooc 
 oooc 
 ooooc 
 
 ooooc 
 
 oooooooc 
 
 OOOOOOC ' 
 
 ooooooc 
 
 oooooooooc 
 ooooooooc 
 ooooooc - 
 ooooooc . 
 o o go o o Q.C 
 
 Section E. - F. 
 
 
 1 
 
 \ 
 
 1 \ 
 
 :| i 1 1. i \ 
 
 i 
 
 Section C.-D. 
 
 -: _ 
 
SIX BOILERS FOR U. S. S. 
 
 BUREAU OP STEAM ENGINEE 
 
 Scale: 
 
 12 
 
 
 
 >> 
 
 f 6600660001 
 ooooooooo 1 
 
 f OOOOOOOOO! 
 fOOOOOOOOOO; 
 
 M|oboooooooo| 
 
 lopqgopogogi 
 
 ^TOQQ^OT^ 
 
 I^^H"' ~-i*L~: 
 
 111' 
 
 OOOO 
 OOOO 
 OOOO 
 OOOO 
 
 OOOO 
 
 o 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 
 o 
 
 o 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 o 
 o 
 o 
 o 
 o 
 
 o 
 o 
 o 
 o 
 
 o 
 o 
 
 \ 
 
 o 
 o 
 
 o 
 
 o 
 
 o 
 o 
 
 ) O O OOOOO OOOOO OOOOO OOOOO OO 6% 
 )OO OOOOO OOOOO OOOOO OOOOO OOOOL 
 )OO OOOOO OOOOO OOOOO OOOOO OOOO'; 
 )OO OOOOO OOOOO OOOOO OOOOO OOOO 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 OOOOOOOOOOOOOOOOO OOOOO O'O O C 
 
 ' 
 
 /Q: 
 
 vQ\ 
 
 /Jtl 
 
 -7J'-- 
 
 (o o g o o o o ' 
 \o/ 
 
 OO 
 
 o 
 
 o 
 
 ooo< 
 o 
 
 
 
 o o 
 ID o o 
 
 \o\ 
 
 
 o// 
 
 
 of 
 
 _QO< 
 
 8 
 
 --10- 
 
 QO o> 
 
 fo/ 
 
 \O; 
 
IANTONOMOH & CLASS. 
 
 EC&, U. S. NAVY DEPABTMEJTT. 
 
 =Feet. 
 
 Plate 
 
 
BRACING OF BOILERS 
 
 FOR U. S. S. MIANTONOMOH & CLASS. 
 
 Plate IX. 
 
 Scale 
 
 a 6 9 12 15 18 21 
 
 U.S.S."AMPHITRITE:' 
 
 U.S.S, MONADNOCKV 
 
 <TT)FOK FRONT A BACK HEADS. 
 
 FOR BACK CONNECTION. 
 
 I 
 
 U.S.S."TERROR." 
 
 =i \ 
 
 FOR FRONT & BACK HEADS 
 
 U.S.S."MIANTONOMOH 
 
Plate X. 
 
 GO 
 n 
 O 
 37 
 0) 
 O 
 
 r- 
 m 
 73 
 
 H 
 m 
 
 O 
 O 
 ?; 
 O 
 
 c 
 
 H 
 
 H 
 m 
 
 (O 
 
 n 
 O 
 
 73 
 00 
 O 
 
 37 
 CO 
 
 CO 
 
 
 
 CO 
 
 o 
 
 CO 
 
If \ 
 
 ' 
 
 JA^JTO 
 
 OOOOOOOO 1 
 
 ooooooooo^ 
 ooooooooo ^ 
 oooooooooo 
 oooooooo 
 oooooo 
 
 OOOOQ 
 
 BOILER FOR STEAMER 
 
 BUREAU OF STEAM-ENGINEERING, 
 
 Grate-surface, 
 
 Heating-surface, ..... 
 
 Calorimeter of tubes, . . . . 
 
 Ratio of calorimeter to grate-surface, . 
 Ratio of grate to heating-surface, 
 Capacity of steam-room, . 
 
 Weight of boiler, 
 
 Weight of water (to \\ inches above tubes), 
 
 TUBES. 
 
 Number, 
 
 Outside diameter, 
 Thickness, . 
 Total length, 
 
.<:OKOUT." 
 
 Plate XL 
 
 1878. 
 
 23.00 sq. ft. 
 . 518.97 sq. ft. 
 . 3.14 sq. ft. 
 i to 7.32 
 i to 22.56 
 . 94.5 4 cub. ft. 
 20,910 Ibs. 
 8,740 Ibs. 
 
 . 114 
 
 2-j- inches. 
 . No. t2\V. G. 
 . 6 feet 2| ins. 
 

SIX BOILERS FOR U. S. S. "NIPSIC." 
 
 BUREAU OF STEAM-ENGINEERING, JUNE, 1877. 
 
 The boilers are to be constructed of the best American charcoal flange-iron. All seams not in contact with th 
 fire to be double-riveted. All the plates to be planed on the edges, the seams to be butt-jointed (planed) and covere 
 with double longitudinal straps on the shell, inch thick ; the straps on the heads to be single, inch thick ; all t 
 be double-riveted and calked perfectly tight. Gussets and stay-plates to be made of the best flange-iron, eac 
 riveted to the shell and heads by a flange 2^ inches wide, and an angle-iron 2^ X 2-J- inches ; the grain of the iron to b 
 placed in the direction of the strain on the stay-rods. Stay-domes a, b, c to be made of the best flange-iron, \ inc 
 thick. All plates used in the construction of these boilers to stand a test not less than 55,000 pounds per square inch 
 
 The sections of the furnaces are to be riveted together so as to bring all the welded seams in line with eac 
 other. Special care must be taken to have the welded seams of the furnaces come below the grate-bars, about on th 
 lines marked d d. 
 
 The inner row of rivets around manhole to be countersunk, flush on the inside. Tubes to be of drawn bras! 
 and to be expanded by the Prosser tool. 
 
 Scale: 
 
 12 
 
 JL 
 
 Feet. 
 
 o o o Q-o o o o o o o o 
 ob oe CK>O o o o o o o 
 booooooooooo 
 
 O O O Q,> OOOOOOOO 
 O O O O D O O O O O O O O 
 ":00O OOOOOOOO 
 
 ^oooooooooo 
 
THICKNESS OF PLATES. 
 
 Shell and circular butt-straps, 
 Heads and back-connections, 
 
 Tube-sheets, 
 
 Furnaces, . . . , . . 
 Gussets and angle-irons, 
 
 RIVETS. 
 
 Diam. of rivets for shell f inch, pitch 2$ ins. 
 Do. do. heads f inch, pitch 2} ins. 
 
 Do. do. connections inch, pitch if ins. 
 
 WEIGHTS. 
 
 Plate XIL 
 
 i 
 
 TV 
 TV 
 i 
 
 inch. Wrought-iron, 
 Cast-iron, . 
 Tubes, brass, 
 Total weight of boiler, 
 
 22,213 
 2,469 Ibs. 
 2,665 
 
 . 27,347 Ibs. 
 Weight of fresli water, 6 inches above tubes, 13,060 Ibs. 
 
 TUBES. 
 
 Number, . 
 Outside diameter, 
 Thickness, 
 Length, . 
 
 Grate- surf ace, one boiler, 
 
 Heating-surface, one boiler, 
 Calorimeter, one boiler, 
 
 32.00 sq. ft. Ratio of grate to heating-surface, 
 821.80 sq. ft. calorimeter to grate-surface , 
 
 4.52 sq. ft. 
 
 . 1 66. 
 
 2-J- inches. 
 . No. 13 W. G. 
 6 feet 3 inches. 
 
 i to 25.6 
 i to 7.1 
 

 
 

 
Plate XIII. 
 
 o 
 m 
 
 H 
 
 > 
 
 r 
 
 CO 
 
 o 
 
 n 
 
 m 
 H 
 
 z 
 
 
 
 O 
 TI 
 
 09 
 O 
 
 r 
 m 
 
 c 
 
 
 
 CO 
 CO 
 
 CO 
 
 q 
 

 


Plate XIV. 
 

TWO BOILERS FOR S. S. "LORD OF THE IS 
 
 DESIGNED BY ADAM MILLER, LONDON, 1879. 
 
 Area of grate-surface, 
 Number of tubes, 
 Diameter of safety-valve, 
 Working pressure, . 
 
 171 sq. ft. 
 780. 
 
 Si ins. 
 
 75 H>s. 
 
 Longitudinal seams of boiler-shell fitted with triple-riveted butt-straps, each \ 
 i inch diameter, 4-^ inches pitch ; holes drilled after plates are bent. Circumfere 
 riveted, lap-jointed ; rivets -J-f inch diameter, 3^ inches pitch ; holes in inner stra 
 bending, and those in outer strakes drilled when in place. 
 
 Longitudinal seams of superheater fitted with triple- riveted butt-straps, each 
 inch diameter, 4^ inches pitch ; holes drilled after plates are bent. Circumfere 
 riveted, lap-jointed ; rivets i-j^ inches diameter, 3^ inches pitch ; holes punched 
 boiler-shell. 
 
 Stay-tubes 4 inches diameter, f inch thick, with thread cut into the body of stay, 
 diameter of 3 inches. 
 
 Jj_ 
 
 oooooooo 
 
 OOOOOOO 
 
 ooooooo 
 ooooooo 
 
 ooooooooo 
 
 OQOQQOQQQ 
 
 oodooOooTd 
 
 oooO 
 
 
 Heating-surface, tubes, 
 plates, 
 superheatei 
 Total heating-surface, 
 
 Scale: 
 
.IS." 
 
 Plate XV 
 
 [,786.88 sq. ft. 
 1,176.00 sq. ft. 
 616.00 sq. ft. 
 
 ;c7888sd ft = 
 
 
 BUTT JOIST FOE 
 1 BOILER SHELL. 
 
 J f> rs 1 rr~ 
 
 
 thick ; rivets 
 seams double- 
 anched before 
 
 thick ; rivets 
 seams double- 
 rilled same as 
 
 ngan effective 
 i 
 
 
 ,< 
 
 
 / o o oto o o o 
 o o o o o o o 
 o lo o o .LO o o o 
 
 o 
 
 
 
 i 
 
 O O O O *T^D O O o 
 
 o o o o o o o 
 \ o o o o o o o o 
 
 
 
 
 <J 
 
 
 
 " n 
 
 r 
 
 
 
 
 
 
 C 
 
 BUTT JOINT FOR 
 SUPERHEATER. 
 
 - 
 
 1 
 
 
 < 
 
 \ 
 
 1 
 
 1 
 
 
 
 . * 
 
 i 
 
 
 
 
 
 ^"b ooo oTo o oo 
 
 O] OOoOoOO 
 
 olo o o cio o o o 
 
 cj o o o o*f=o " 
 -^5-4 o o o o o o c 
 .^~ i o o o oLo o o p 
 
 ft 
 
 
 
 1 
 
 1 
 
 \ 
 
 l M 
 
 
 
 ; 
 
 
 
 
 r 
 
 
 
 
 ^S 
 
 Feet. 
 

 
Grate-surface - 4.75 sq. ft. 
 Heating --119 " 
 Calorimeter .916 
 IVeigTit of toiler 3100 Jbs. 
 water 900 
 
 BOILER FOR 8* X 8' EN 
 
 
5NE, U. S. S. CUTTERS. 
 
 Plate XVI. 
 
THE 
 
 R 
 
 OF 
 
-- 
 
 
 . n 
 
 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 
 oo ooo 
 ooooo 
 ooooo 
 ooo oo 
 
 ooooo: 
 
 OOOOO' 
 
 o o o o b j 
 
 ooooo 
 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 ooooo ooooo 
 
 00000:00000 
 ooooojo oo oo 
 
 OOOOOiOOOOO 
 
 OOOOOlpOOOO 
 -38|->4P^B- 
 
 QOOOOjOOOOO 
 
 00000:00000 
 ooooolooooo 
 ooooo! ooooo 
 
 o oooo 
 ooo oo 
 o oooo 
 ooooo 
 
 ooooo 
 ooooo 
 ooooo 
 ooooo 
 
 ooooo 
 ooooo 
 ooooo 
 ppoqq 
 
 ooooo 
 ooooo 
 ooooo 
 ooooo 
 
 ooooo o 
 
 .'OOOOO O 
 
 oooooio 
 _oqoqo_q 
 
 o o o o o o 
 _ ooooo o 
 
 ! OOOOOiO 
 
 oojDqqiq 
 
Plate XVII. 
 
 TOILERS FOR 
 -"PLYMOUTH. 
 
 Feet. 
 
 OOOOOOi 
 
 oooooooo 
 oooooooo 
 
 OOOOOOOOi 
 
 oooooooo 
 
 OOOOOOOOv 
 
 oooooooo/ 
 ooooooo / 
 OOOOOQ/J 
 
 \ 
 
 \ 
 
 Plate XVII. 
 
 Section through. Centre of Boiler. 
 
 13 - 6" 
 
 4 
 
 1 
 
 I 
 
 / 
 
OF THE 
 
 UNIVERSITY 
 
CO 
 37 
 
 > 
 
 o 
 
 z 
 o 
 
 n 
 O 
 7) 
 
 CO 
 O 
 
 r 
 m 
 
 30 
 
 c/> 
 
 -n 
 
 C 
 0> 
 
 
 
 o> 
 
 
 
 X 
 
 v% 
 
 T3 
 
 r 
 -< 
 
 TH 
 
 
 
 . 
 ( 
 
 
 
 \ 
 
 
Plate XVIII. 
 
 : 
 


--fflfc J e- -e- 
 
 r~f 
 
 
 
 
 
 
 
 
 
 
 'Mfa W7 
 
 (D 
 
 _ 
 
 
 
 T 
 l 
 
 
 
 
 m 
 
 '" 
 
 f 
 | 
 
 
 V 
 i 
 
 
 1 
 
 
 
 
 r h H h fj hy-i w h ^5; 
 l^ 1 
 
 , 
 
 
 
 O 
 
 m 
 
 > 
 
 r 
 <s> 
 
 n 
 O 
 n 
 
 oo 
 O 
 
 r 
 m 
 
 c 
 
 
 
 0) 
 
 
 
 0) 
 
 
 *s 
 
 \ 
 
 o 
 
 r 
 
 o 
 
 c 
 
 H 
 
 I 
 
Plate XIX. 
 
 "- ~* 
 
 
 . . . - 
 
 "" i""~ 
 
 
 
 
 i 
 
 1 
 
 . 
 
 m 
 
 
 j 
 
 L- . . 
 
 2-7" \ 
 
 i 
 
 1 
 
 V 
 
 2-10" 
 
Single Shear 
 
 Mean of 44149 Hw.per" 
 
 Sheared at 9000 
 
 No.l 
 
 Sheared at 8900 
 
 No.2 
 
 Sheared at WOO 
 
 Sheared at 9*00 
 
 .62 
 
 N0.4HIHH 
 
 Jfcon o/ 39263 16.j)er 
 
 Sfcjred at 12900 
 
 Sheared at 12900 
 
 Sheared at 13300 
 
 / 
 
 \ 
 
 _ 
 
 Sheared at 127S 
 
 la 
 
 .50 
 
 Noj'l 
 lOll 
 
 11 
 
 1 
 
 
 Jfean of 39553 H>s.pr ' 
 
 Sheared at 18800 
 
 No. 
 13 
 
 Sheared at 19500 
 
 No.14 
 
 Sheared at 19650 
 
 No. 15 
 
 Sheared at ISSOi 
 
 No. 1C 
 
 O/41SD8 pr ' 
 
 Slieared at 3C5M 
 
 .90 
 
 Jfean<!/-W708B>.per " 
 
 Grand Af. 0/41033 Uis.per 
 
 Double Sliear 
 
 Sheared at 16175 
 
 No.l 
 
 Sheared at 16800 
 
 Sheared at 10060 
 
 Sheared at 10400 
 
 No.4 
 
 1 I 
 
 aeon o/ 77318 (6s..per 
 Double Shear 
 
 Sheared at 25650 
 
 No.7 
 
 Sheared at 23700 
 
 No.8|[ 
 
 Sheared at 23600 
 
 f 
 
 Sheai 
 
 Jfean of 79536 i&s.per ' 
 Double Shear 
 
 ' 
 
 \ 
 
 Sh 
 
 z 
 
 sored at 36650 
 
 No.13 
 
 II 
 
 V 
 
 
 Sheared at 39iOO 
 
 .77 
 
 DPOD1 
 
 Mean of 76789 lbs.per "" 
 Double Shear 
 
 Sheared at 51600 
 
 No.19 
 
 Could not be screwed up tightly, therefore weak. 
 Sheared at i7650 
 
 Kot screwed up tightly. 
 Sheared at 4C300 
 
 Mean of 75293 Ibs.per " 
 Double Shear 
 
 Grand Mean 0/78030 Ws.per * 
 Double Shear 
 
 / 
 
 \ 
 
 2 
 
 Sheared at & 
 
 No.25 
 
 
 
 D 
 
 Sheared at 61700 
 
 No.26 
 
 / 
 
 \ 
 
 L 
 
 B 
 
 1 
 
 ?Aearcd at C 
 
 3650 
 
 
 111 
 
 
 No.S7 
 
 / 
 
 \ 
 
 
 Shearing Attachment 
 
 
 i i i i ! I i 
 
 
 i 
 
 : i | 
 
 
 
 
 
Xo.5 
 
 Sheared at MM 
 
 n | Xo.G 
 
 Plate XX. 
 
 X 
 
 Xo.ll 
 
 Sheared at 15750 
 
 ." 
 
 S/tcar 
 
 Ko.l7|l 
 
 It 
 
 OanmfaiUMO 
 
 Experiments on Shearing Wrought Iron Bolts, 
 
 Conducted at the Washington Xary Yard, 
 
 CUtf Engineer Wm. H. Shock, U. 8. N. 
 
 i ; I'^xo 
 
 0.22 
 
 i'.f.l. ?,i.'L'l.) 
 
 No.28 
 
 1.M 
 
 I 
 
 Xo.29 
 
 Sheared at XTVO 
 
 Xo.24 
 
 Sheared at Stito 
 
 Tl 
 
 Xo.30 
 
 D 
 
 
 V 
 
 X. ! t 
 
 1 i 
 
 
 
 
 
 
 
 
 a 
 
 Shearing Attachment 
 
 m 
 
 Sheared at 17575 
 
 Xo.5 
 
 Sheared at 17GOO 
 
 Sheared at 24600 
 
 f JXo.ll 
 
 .S4 
 
 
 
 No.12 
 
 andatlMM 
 
 No.lC 
 
 Sheared at 'MM 
 
 No.17 
 
 .79 
 
 No. 18 
 
 Sheared at 51300 
 
 No.22 
 
 << at OM 
 
 Xo.23 
 
 Sheared at 51375 
 
 Xo.24 
 
 No.28 
 
 Sfenred a/ 63535 
 
 Xo.20 
 
 Sheared at C3500 
 
 m 
 
 L U 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 
 
 
 
 1 1 
 
Fig.l. 
 MARINE FLUE BOILER. 
 
 
 
 o 
 
 Plate XXI. 
 
 \ 
 
 Fig- 2. 
 BOILER FOR U. S. S. "SHOCKOKON." 
 
 11-6- 
 
 f 
 
 5 J 
 
 e 
 
 ^ 
 
 ^ 
 
 
 32 
 
 Fig. 3. 
 BOILER FOR U. S. S. "MORSE." 
 
 

 
SMOKE BOX EM> 
 
 METHODS OF SECURING BOILER TUBES. 
 
 Fig- I- Fig. 2. 
 
 Plate XXII. 
 
 Fig. 6. 
 TWEDDELL'S HYDRAULIC TUBE EXPANDER. 
 
 Fig. 7. 
 SELKIRK'S TUBE BEADER. 
 
E xpe rime fit i 
 Tubes before 
 
 No. 5 
 
 2 1150 Ibs. 
 
 No.C 
 
 12000 Ibs. 
 
 Tvbes after exper/'tcr' 
 
 Scale: 
 
 1 2 
 
 N0.25&86 
 
 8225 Ibs. 
 
 14 100 Ibs. 
 
 So. 6,13 d- 14,29 dt 30,41 it 42, 
 are in accordance with the i'.S.A'aval Practice, 
 
 Tubes before experiment with brass femUes. 
 
 No.31&32 
 
 19450 Ibs. 
 
 Tubes after experiment with brass ferrule*. 
 
Brass Tubes, 
 with iron ferrules. 
 
 Composition nuts. 
 
 Plate XXIII. 
 
 No. 23 24 
 
 28310 Ibs. 
 -Mean 
 
 f h iron ferrules, 
 
 i Inches. 
 
 Prosser's process. 
 
 >"o.45&-4( 
 
 U 
 
 Dudgeons process. 
 
 Experiments conducted at the Washington Navy Yard, Jan, 1877, 
 Chief Engineer W>n.H.*J,<,,-k. J'.S.X. 
 
 V; 
 
 - 
 
 : ' 
 
 
 -UJi 
 
 jfetnod 
 of 
 fastening 
 
 JEW 
 
 '1 
 
 -. 
 
 Strain 
 in 
 
 pounds 
 
 1 
 
 2.5 .!)" '2.4* ", 
 
 Proaser 
 
 /ran 
 
 75 
 
 2,:.0 
 
 2 
 
 
 
 2.41 - 
 
 
 
 " 
 
 " 
 
 SO200 
 
 3 
 
 
 
 
 
 2.15 
 
 
 
 Dudgeon 
 
 " 
 
 n 
 
 12750 
 
 4 
 
 " 
 
 M 
 
 UM 
 
 
 
 " 
 
 H 
 
 ,. 
 
 1GOUO 
 
 6 
 
 
 
 " 
 
 LM 
 
 . 
 
 
 
 .Vane 
 
 70 
 
 21150 
 
 c 
 
 " 
 
 ,. 
 
 LM 
 
 
 
 Prosser 
 
 
 
 71 
 
 12UUU 
 
 ; 
 
 .< 
 
 u 
 
 
 
 .. 
 
 .< 
 
 Iron 
 
 u 
 
 27500 
 
 8 
 
 
 
 -.-.: 
 
 . 
 
 Dudoeon 
 
 .. 
 
 ti 
 
 4GOOO 
 
 a 
 
 
 " 
 
 _'.::! 
 
 i 
 
 
 
 
 
 72 
 
 30300 
 
 10 
 
 
 
 
 J.ll 
 
 .. 
 
 
 
 M 
 
 
 
 3GOOO 
 
 11 
 
 .< 
 
 
 LM 
 
 
 
 Prosaer 
 
 
 
 
 
 25300 
 
 12 
 
 .. 
 
 .. 
 
 
 
 .. 
 
 .. 
 
 .. 
 
 .. 
 
 2G400 
 
 11 
 
 
 
 " 
 
 
 
 . 
 
 A'./tj'U/ci 
 
 .Vune 
 
 75 
 
 10450 
 
 11 
 
 M 
 
 .. 
 
 tt 
 
 ,i 
 
 
 
 " 
 
 
 
 27000 ; 
 
 15 
 
 
 
 
 
 tt 
 
 
 
 
 
 iron 
 
 .. 
 
 40150 
 
 1C 
 
 .. 
 
 
 
 
 
 
 
 
 
 " 
 
 .. 
 
 JS600 
 
 17 
 
 " .. 
 
 ., , 
 
 , 
 
 
 
 .. 
 
 
 *ooo 
 
 
 
 U 
 
 a 
 
 a 
 
 a 
 
 a 
 
 - 
 
 21400 
 
 19 
 
 t 
 
 _ 
 
 - 
 
 M 
 
 
 
 Iron 
 
 - 
 
 3U30U 
 
 20 
 
 a 
 
 - 
 
 _> ID 
 
 U 
 
 
 
 ft 
 
 - 
 
 41050 , 
 
 21 
 
 S.( 
 
 .9 
 
 2.0" 
 
 
 
 Dudgeon 
 
 Acme 
 
 c;.-, 
 
 7050 
 
 22 
 
 - 
 
 ft 
 
 - 
 
 - 
 
 u 
 
 
 
 _ 
 
 KIM 
 
 23 
 
 - 
 
 - 
 
 - 
 
 - 
 
 
 
 /ron 
 
 04V 14400 
 
 -! 
 
 
 
 - 
 
 
 
 " 
 
 " 
 
 13SOO 
 
 4 r 
 
 > 
 
 V 
 
 V 
 2.5 
 
 1=::^ 
 
 Method 
 of 
 
 '; -'< I END 
 
 A 
 
 
 Strain 
 pound* 
 
 .'/ 
 
 M 
 
 
 Dudgeon 
 
 .Vtwie 
 
 - 
 
 8100 
 
 26 
 
 
 
 - 
 
 .. 
 
 M 
 
 
 
 
 
 - 
 
 8150 
 
 27 
 
 H 
 
 - 
 
 " 
 
 u 
 
 " 
 
 /ron 
 
 - 
 
 14250 
 
 25 
 
 
 
 .- 
 
 .. 
 
 M 
 
 
 
 " 
 
 - 
 
 14550 
 
 90 
 
 
 
 
 
 < 
 
 
 
 Proaser 
 
 Xune 
 
 7* 
 
 14450 
 
 10 
 
 
 
 - 
 
 .. 
 
 .. 
 
 
 
 
 
 M 1 150OO 
 
 11 
 
 - 
 
 
 
 
 
 M 
 
 D 7 .' 
 
 u 
 
 1 1707* 
 
 12 
 
 
 
 " 
 
 
 
 
 
 " 
 
 
 
 21S2S 
 
 U 
 
 
 
 .. 
 
 .. 
 
 M 
 
 
 
 Brass 
 
 32250 
 
 34 
 
 
 
 M 
 
 
 
 
 
 M 
 
 " 
 
 " 1 11400 
 
 30 .. 
 
 
 
 M 
 
 
 
 JVoer 
 
 " 
 
 227511 
 
 36 
 
 
 
 .. 
 
 
 
 M 
 
 
 
 
 
 22:iiu 
 
 18 
 
 
 
 B 
 
 .1 
 
 . 
 
 " 
 
 " 
 
 70 ! 17150 
 .. 17100 
 
 1* 
 
 N 
 
 
 
 " 
 
 
 
 Dudgeon 
 
 
 
 24SOO 
 
 40 
 
 
 
 
 
 .. 
 
 
 
 .. 
 
 .. 
 
 <. 
 
 2JOOO 
 
 41 
 
 .. 
 
 . 
 
 
 
 
 PWr 
 
 .Vane 
 
 75 
 
 1000 
 
 42 
 
 _ 
 
 
 
 - 
 
 - 
 
 
 
 
 
 _ 
 
 111-50 
 
 41 
 
 
 
 
 
 _ 
 
 - 
 
 /' '.. 
 
 - 
 
 - 
 
 IMM 
 
 11 
 
 w 
 
 - 
 
 - 
 
 a 
 
 
 
 
 
 a 
 
 22250 
 
 45 
 
 a 
 
 - 
 
 ~ 
 
 _ 
 
 - 
 
 Brass 
 
 - 2U5!*) 
 
 46 
 
 
 
 - 
 
 - 
 
 - 
 
 a 
 
 a 
 
 27550 
 
 47 
 
 
 
 - 
 
 _ 
 
 - 
 
 Pro5er 
 
 
 
 . 1 15250 
 
 45 
 
 
 
 
 ' 
 
 
 
 
 JIC--50 
 

 
Tubes bej 
 
 No. 1 
 
 29050 Ibs. 
 
 No. 2 
 19950 Ibs. 
 
 No. 3 
 2C900 Ibs. 
 
 -^^r^-^f 
 
 No. 4 
 25525 Ibs. 
 
 No. 5 
 20250 Ibs. 
 Tubes < 
 
 No. 11 
 
 29TOO Ibs. 
 
 * N C5 
 
 a: 
 
 No. 12 
 
 29650 Ibs. 
 
 No. 13 
 11300 Ibs. 
 
 i ^_- 
 
 n 
 
 f " i 
 
 No. 14 
 
 14800 Ibs. 
 
 r "i 7 i" 
 
 JZD 
 
 No. 15 
 
 8850 Ibs. 
 
 Tubes aj 
 
 > 
 
 _J 
 
 *> r 
 
 cr 
 
 RUSSELL A STRUTHERS, ENG'S, N,Y. 
 
1-illlUlt. 
 
 Plate XXI V. 
 
 Experiment with Iron Tubes, with Copper, Iron & Steel tube plates. 
 
 Conducted at the Washington Navy Tar<l, Jan. 1877. 
 
 by 
 
 Chief Engineer Wm. H. Shod, U. S. JV. 
 
 No. 6 
 
 17300 Ibs. 
 
 experiment. 
 
 No. 7 
 20450 Ibs. 
 
 No. 8 
 
 24650 Ibs. 
 
 _J 
 
 No. 9 
 
 22700 Ibs. 
 
 No. 10 
 21600 Ibs. 
 
 No. 16 
 5950 Ibs. 
 
 Tubes expanded by 
 Dudgeons process. 
 
 No. 17 
 
 221 00 Ibs. 
 
 Ins. Ins. Sq.in Ins. In. 
 
 ( Copper ring in Ion 
 
 i 3 'i 2'; .981 2'i Iron 7-16 Steel. X 1 tube-plalc. Ends 
 
 f tube nveted over. 
 
 ring in lower i 
 'off 
 
 ., 
 
 " 
 
 " 
 
 " 
 
 
 " Riveted over. 
 
 :: :: 
 
 
 
 
 
 Copp- ?* 
 
 
 
 
 29-16 
 
 
 .4 
 
 Steel. H Partly riveted. 
 
 " " Tron femilcs. 
 
 :: .. 
 ., . 
 
 - 
 
 '* 
 
 K 
 
 
 '* Simply expanded, 
 i Tube-plate boles ta- / 
 
 ., .. 
 
 
 29-16 
 
 K 
 
 
 
 i Tube-plate holes ta- ) 
 " - per 2*h ' -213-16' 
 ; ( and 25i - 29-16 
 
 39050 
 
 IW50 1 
 20900 
 
 25525 
 
 302^0 
 
 17330 
 
 8850 
 
 " . 595 
 
 71 22100 
 
 

Plate XXY. 
 
Scum Pipe 
 
 33 
 
 Z 
 
 m 
 
 oo 
 o 
 
 m 
 
 X) 
 
 00 
 
 < 
 
 H 
 
 I 
 O 
 
 > 
 
 J) 
 
 
 td 
 
Plate XXVI. 
 
 rr ? 
 
 m 
 
 u 
 m 
 
 H 3 
 
 C <? 
 
 oo r 
 
 C 
 
 CO 
 
 O 
 
 r 
 m 
 
Fig.l. 
 THE HERRESHOFF COIL BOILER. 
 
 Plate XXVII. 
 
 i 
 
 IJJPTJPJCT l T 
 
 N ii* i I 
 
 ijiiiiij if i 1 ' 
 
 'i >l j i' '.! i ! i i! ! 
 
 I i i 
 
 HNL JF 
 
 Fig. 2. 
 THE BELLEVILLE BOILER. 
 

Pig. i. Plate XXVIII. 
 
 BOILER FOR 8X8 ENGINE, U. S. S. CUTTERS. 
 
 Grate-surface-5 .33 sq.ft . 
 Steam-Room-4.10 cub.ft. 
 
 Heating-surface-150 sq.ft. 
 Weight of Boller-2110 Ibs. 
 
 33=0: 
 
 Fig. 2. 
 THE DAVEY-PAXM AN BOILER. 
 
42 
 
Plate XXIX. 
 
 CD 
 
 1 m 
 
 30 
 
 V) 8 
 
 O "H 
 TI C 
 
 <=! 
 
 co 
 
 ' a 
 
 z 
 
 o 
 
 T3 30 
 GO 00 
 
 o -n 
 
 f_ v 
 
OF THE 
 
 UNIVERSITY 
 
 cr 
 
^^_ 
 
 
Plate XXX. 
 
 U. S. S."NIPSIC". 
 
 12 1 2 3 4 5 
 
 Scale: litiiilmui I I =1= E3 Teet, 
 
 J.A.H. 
 
15" 
 
 Plate XXXI. 
 
 ^ 
 
 
 1 \ 
 -if 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 _ 
 
 _- 
 
 Rnr 
 
 
 Tig. 3. 
 
 Fig. 4 
 
Plate XXXII. 
 
 STEAM STOP VALVES & FEED 
 
 VALVE % FOR BOILERS OF 
 
 U. S. S. "NIPSIC." 
 
 rig. 2. i 
 
 c _*_. ; _ _ :! j 
 
/ .^--.^ 
 
 ' /f ^\v\ 
 
 , ^Ky>hv 
 
 ^f^7 \ 
 
 ^ KX 
 
 UJ- 1 
 
Plate XXXIII. 
 
 
 H 
 m 
 
 73 
 
 O 
 
 > 
 
 c 
 
 O 
 
 m 
 
 n 
 O 
 so 
 
 CD 
 O 
 
 r 
 m 
 
 O 
 -n 
 
 C 
 
 O 
 

Plate XXXIV. 
 
 
 Fig.l. 
 
 SAFETY VALVE FOR BOILERS, 
 U. S. S. "NIPSIC. 
 
 Fig. 2. 
 
 SAFETY VALVE APPROVED BY THE 
 BOARD OF SUPERVISING INSPECTORS OF STEAM-VESSELS. 
 
 Fig.3. 
 ASHCROFT'S SAFETY VALVE. 
 
Kg.l. 
 
 KOERTING'S JET APPARATUS. 
 
 Plate XXXY. 
 
 SELLERS' SELF-ADJUSTING INJECTOR. 
 
 ^ 
 
 rig. 3. 
 
 KOERTING'S UNIVERSAL 
 LIFTING INJECTOR. 
 
 gu'iiiu. ' *-\ 
 
 Water 
 

 Plate XXXYI. 
 
 SPECIMENS OF RIVETS AND RIVET-HEADS 
 
 FROM BOILERS OF COPPER-ROLLING MILL, 
 
 NAVY YARD, WASHIKGTOIf, D. C. 1879. 
 
 

 - 
 
 
 YE. 01 143 
 
 
 

 
 . - 
 
 /. . *%&. A 
 
 ^ 
 
 _< 
 
 
 
 
 
 
 . . 
 
 .* 
 
 ' 
 
 
 
 , 
 
 f 
 
 ; 
 
 
 
 *. 
 
 -^ 
 
 ' < 
 
 '** 
 
 1 , 
 
 \ , 
 
 ' , : 
 
 ^-> . I 
 
 ^ 
 
 '>* - 
 
 ^ ^