m PRACTICAL TREATISE ON HIGH PRESSURE STEAM BOILERS INCLUDING RESULTS OF EECENT EXPERIMENTAL TESTS OF BOILER MATERIALS, TOGETHER WITH A DESCEIPTIOW OF APPROVED SAFETY APPARATUS, STEAM PUMPS, INJECTORS AND ECONOMIZERS IN ACTUAL USE. BY / WILLIAM M. BARR '^^v..../?.6A^.../tr 33: Of u; *..■-•>:.> INDIANAPOLIS, I^^OF v//^- YOHN BROTHER^ 1880. V b '^n COPYRIGHT. WILLIAM M BAEE, 1879. -^J &'J- INDIANAPOLIS: BAKER & RANDOLPH, PRINTERS. PREFACE. This book is not put forth so much as a specimen of book making as it is a record of notes, memoranda, experi- ments, practice, and experience gathered, during several years in which the writer has been connected in one way or another, with the design and manufacture of high pressure steam boilers. 'No person at all acquainted with boilers would expect to receive at the hands of anybody an entirely original treatise on this subject. This book contains a considerable amount of practical information never before published, gleaned from the author's own experience, as well as val- uable contributions from several of his friends, who have had large experience in boiler making. The chapters on the strength of iron and steel have been carefully compiled from tests made with samples sheared from plates actually delivered for boilers under contract, and were not selected samples taken from the mill with a view to getting high results. The tables show that this country possesses materials for boiler construc- tion having qualities which are not surpassed by any iron or steel in any market. There is little doubt that the boiler of the future will be of steel, and considerable space has been given this material: many tests have been made, and records of the results appear, for the first time, in these pages. This book comes far short of being what the writer would like to have it, many subjects of interest to boiler VI PREFACE. makers and steam users have been omitted. Marine boilers have not been included, as that is a class of work whic.h is under a more intelligent direction than generally found in the smaller shops, and have less need of such data as here furnished ; to have included it, would have made a larger and more expensive book than the writer felt justified in undertaking. The same is also true in regard to locomo- tive boilers. Most of the smaller boiler shops are in charge of men who were once journeymen boiler makers, and " set up business on their own account." These per- sons are, as a class, good boiler makers, but having had little experience in estimating and designing work, other than the particular kind on which they have had a long experience, are often at a loss how to proceed ; it is pos- sible that this book may prove of service to many such. The object has not been to make this a book specially for boiler makers, but a hand book for engine builders, architects, and steam users, as well. The writer wishes to express his deep sense of obliga- tion to his many friends who have contributed and assisted in the preparation of much of the data which appears in these pages, and especially to Mr. David Greig, Leeds, England, Mr. George H. Atkinson, Pittsburg, Penn., Mr. J. M. Allen, Hartford, Conn., Mr. Coleman Sellers, Philadelphia, Penn., and many others. Free use has been made of recent papers read before engineering societies, and bearing on this subject; from which extracts have been made, and the proper credit given in the body of the work. Indianapolis, Ind., December, 1879. CONTENTS, PAGE I. Introduction 1 II. Cast iron as a material for steam boilers 8 III. Wrought iron as a material for steam boilers 15 IV. Steel as a material for steam boilers 29 V. Testing wrought iron or steel for boilers 66 VI. Riveted joints '. 87 VII. Weldingj flanging and influence of temperature..- 131 VIII. Strength of boilers 148 IX. Heating surface and boiler power 188 X. Externally fired boilers 218 XI. Internally fired boilers 257 XII. Boiler setting 304 XIII. Feed apparatus 337 XIV. Heaters and economizers 374 XV. Safety apparatus 393 XVI. Incrustation and corrosion 422 XVII. Sectional boilers 437 ERRATA. Page 131. Eleventh line from the botton, read nearly instead of clearly. Page 168. See foot note, giving correct titles to tables L and LI. Page 171. Tenth line from top, read flues instead of tubes. Page 257. Last line, read originating instead of originally. Page 305. Engraving. The foundation at E is shown under the dimension line Y. It should have been placed under the center line of the front. The error is in the engraving only; the figures are correct. Page 309. Thirteenth line from bottom, read cleaning instead of clearing. CHAPTER I INTRODUCTION. Conditions Demanded in Boiler Construction — Materials of Construc- tion — Impurities Present in Crude Iron. How to generate steam economically is the great prob- lem of the steam engine of to-day. For many years past engineers have been devoting more time to the perfection of the mechanism of the engine than to the design and construction of the boiler and fur- nace. As a result, v^e have engines of superior excellence, supplied with boilers and furnaces faulty in design, coupled with an extravagant waste of fuel in service. In order to obtain anything like a proper economy in the use of fuel in generating steam, the phenomena of combustion must be understood to properly design and construct the furnace; the strength and properties of mate- rials to properly design the boiler. It is scarcely half a century since boiler pressures rarely exceeded ten pounds per square inch above the atmosphere. Gradually, however, as the superior economy and advan- tages of high pressure steam became more generally known, and the properties of the different materials of construction became better understood, there came also a demand for better boilers. This, in turn, required at the hands of the manufacturer greater discrimination in the selection of a material not only, but for improved designs and better workmanship. This has, in part, been accomplished, in-as- much as boilers are in very common use carrying a regular steam pressure of one hundred pounds per square inch, and (2) 2 A TREATISE ON STEAM BOILERS. occasionally as high as one hundred and fifty pounds. These high pressures are the result partly of experimental inquiry, but are mainly due to a better understanding of the principles of thermodynamics, for it is in accordance with its teachings that these higher pressures have been steadily adopted, so that now there is a very general tend- ency, arguing from pure theory, toward extreme pressures, under the belief that the highest economy is to be based upon, and is in fact, simply a question of pressure. This leads us to the conditions demanded in the construction of a steam generator, which are. Safety while working under high pressures. Simple in construction. Thorough circulation of water in all parts of the boiler. Economical in the use of fuel. Durability in service. Facility of examination, cleaning and repairs. The first requisite in a boiler would seem to be that of safety, for without this all the other conditions are of little value. By safety is meant that pressure of steam which a boiler can generate and hold without danger of rupture. The safety of a boiler depends upon its form; the materials of which it is made; and the details of its con- struction. The conditions of safety and durability depend largely upon the selection of a suitable material — one which shall have considerable hardness, and at the same time, a high tensile strength combined w^ith a reasonable degree of toughness. Practically, the only available materials for the con- struction of steam generators are — cast iron, wrought iron and steel. Each of these have properties which are of value in this connection, though most of the steam boilers now in use are made of wrought iron. Formerly, copper had been used in boiler construction, especially for fire boxes for locomotives, and the internal heating surfaces in IMPURITIES IN IRON. marine boilers. This material has been almost entirely abandcned in boiler construction, notwithstanding its supe- rior conducting power over iron or steel ; the causes which have led to its abandonment are high first cost, inferiority in hardness and tensile strength as compared with the other materials named. The materials of construction hold such an important place in boiler design that some space should be given in a Avork of this kind to the consideration of the crude and finished iron employed in boiler making. Before any steam generator can be properly designed there must be a knowledge of the properties of the materials which are to enter into its construction; and as castings, wrought iron and steel, have one common starting point, it will not be thought out of place to give in briet outline the foreign elements which are contained in and which give character to iron and steel. The crude cast iron as it comes from the blast furnace, no matter what impurities it may contain, is known under the name of jpig iron. These are usually classed as either white or gray irons ; and it is probable that the difierence in the qualities, or properties, leading to this classification are due more to the infiuence of the contained carbon than to any other cause. The white iron approaches more nearly the character of an alloy than the gray iron, which partakes more of the nature of a mechanical mix- ture. The latter iron is employed for foundry use, the former for the manufacture of wrought iron. The prin- cipal impurities in pig iron are sulphur, silicon, phosphorus, manganese, carbon. Sulphur is almost always found in pig iron and has a remarkable infiuence on its quality. White irons appear to contain more of it than the gray varieties. At low heats it will in a measure prevent fiuidity in cast iron, causing it to assume a mushy appearance, which may be entirely overcome by the application of a higher heat, when the A TREATISE ON STEAM BOILERS. mushy appearance changes to that of a more perfect fluid. Sulphur diminishes the strength of castings in a very high degree by causing them to be cold-short, brittle and hard. Silicon is always present in crude as well as in refined irons, being next after carbon the commonest impurity met with in iron. When in combination with crude iron, the proportion is usually found to be greater in the gray than in the white varieties. Quantities as small as one-half of one per cent causes crude iron to be brittle, and its pres- ence in castings is regarded as injurious to quality, the best castings being those which contain it in the least amount. Phosphorus has the effect to harden cast iron, and to increase its fusibility. It enters into chemical union w^ith iron and is present in quantities rarely exceeding one to one and a quarter per cent in ordinary white or gray irons. In combination with iron it renders it close and compact, and has the tendency to make it cold-short when reduced to low temperatures or near the freezing point. Pig iron containing phosphorus melts easily, becomes very fluid, is easily managed in refining, and when contained in wrought iron up to a limit of one-fourth of one per cent has no perceptible effect on its welding power, except that it requires it to be done at a low heat. Manganese, in its chemical properties, is in many respects like iron, and will form similar compounds. These two metals have an aflinity for each other, and during the oper- ation of reducing the iron from the ore they enter into such an intimate relation as to form an alloy. Manganese causes iron to be more fluid when melted, and to be hard and brittle when cold. White iron contains more man- ganese than gray irons. Some ores yield an iron contain- ipg as much as ten and twelve per cent of manganese, which has the property of containing in its composition as much as one-twenty-fifth of its weight of carbon in a state IMPURITIES IN IRON. of chemical combination ; this iron is extensively used in steel making by the Bessemer process, and is generally known by its German name, speigel-eisen. Carbon is always present in pig iron. The quantity so found is variable, and its effect on the character of the iron is by no means certain. Sometimes it appears to be a chemical union, when it then partakes somewhat of the nature of an alloy. At other times it seems to have no affinity for the iron, and its presence in a free state in the pores, or rather between the crystals of the iron, is regarded as little else than a mechanical mixture. These differences are not altogether due to the quantity of carbon present, but rather to its state, or particular form of com- bination. Iron, when pure, is soft and malleable. Carbon adds to its hardness, renders it more brittle and lowers the degree of fusibility when a sufficient quantity of carbon is present to make what is known as cast iron. In this state it can neither be forged nor welded. The quantity of carbon in pig iron commonly ranges from one and one-half to three and one-half per cent. How carbon enters into combination with iron can hardly be said to have been satisfactorily explained. It is certain, however, that a small quantity of carbon has a very marked effect upon a large mass of iron. The state in which carbon is present in cast iron may be determined, 1. By the appearance of the fracture. 2. By dissolving a portion of the iron in either diluted sulphuric or muriatic acids. The former of the two is generally employed in deter- mining the approximate quality of pig iron and castings. By the latter method there is more certainty in arriving at the actual condition of the contained carbon. The acid acting upon the iron, but not upon the carbon, the latter 6 A TKEATISB ON STEAM BOILERS. is left in a free state in the solution, if it so exists in the sample tested. After dissolving gray irons, the solution will contain numerous black particles, which have all the properties of ordinary graphite. A white iron known to have more carbon in its composition than the gray, when similarly dissolved will present fewer particles of carbon in the free state, because the contained carbon is in a chem- ical combination with the iron and is present in the solu- tion as a carbide ; thus presenting fewer particles of graphite or carbon than the former solution. White cast iron — So far as observed, this seems to have all the qualities of a perfect alloy ; that is, the mixture of carbon and iron produces a metal having the different properties from the materials which enter into its compos- ition.* It is harder than gray iron, or any iron in which the carbon is present in a state of mechanical mixture only. The fracture of white iron presents a surface of silvery whiteness, with but little luster, showing little or no free carbon. It is extremely hard, will resist the action of the file or chisel ; it is very brittle, and is unfit for any of the uses to which ordinary castings are applied. It naelts at a lower heat than gray iron, but does not become liquid at this low temperature, assuming, rather, a pasty condition, which may be overcome by the application of a still higher temperature. Mottled cast iron contains about the same quantity of carbon as the white and gray varieties. The condition in which it is held is about half combined and half in a free, state. It is very hard, brittle and not so elastic as gray iron. Gray cast iron is made from the pig irons in which there is the least carbon chemically combined, and the greater * Carbon being a non-metalic substance, it will be understood that the use of the word alloy in this sentence is unconventional, its use being applied to mixtures of metals only. CHILLED IRON, portion in the free state, as graphite. A general average would show about one per cent of carbon chemically combined and about two and one-half per cent of free graphite. Foundry pig is usually sold in the market as either Xos. one, two, three, ^o. 1 being the softest and ^o. 3 the hardest. No. 1 is usually of a dark gray color, having large granular crystals, between which particles of carbon are seen and may be easily detached. In melting it becomes quite fluid and accurately fills the mold in which it is poured. It is not so hard or so strong as either of the other two irons named, and to make the best castings, portions of the "heat" should be made up of the lower grades, ^o. 2 pig-iron is harder and has a finer grain than ]^o. 1, owing to more of the carbon being in a combined and less in the free state. When properly mixed with Xo. l.it makes the best castings for machinery or for any use in which strength and durability are required. 1^0. 3 iron is not in general suitable for castings which require working in the machine shop. It is hard, brittle and is saitable only for mixing with higher grades of iron in the production of heavy castings. Chilled iron is produced by a change in the condition of the carbon present in the cast iron from a state of mechan- ical mixture to that of chemical combination, brought about chiefiy by a sudden cooling while in a molten state. When pig iron is melted, or in a fiuid state, it is not improbable that the carbon which was present in a free state in the pores of the iron is dissolved and unites with the iron. In ordinary cooling the carbon would separate from the iron as in the pig iron before melting, but when the fluid iron is suddenly cooled the carbon is condensed by the con- traction of the iron and forced to remain in chemical union with it, instead of disengaging itself and collecting in the pores of the casting. CHAPTER II. CAST IRON AS A MATERIAL FOR STEAM BOILERS. Arguments in Favor of It — Objections to its Use — Effect of the Impurities found in Cast Iron — Tensile Strength of Cast Iron — Elastic Limit of Cast Iron — Defects in Castings — Behavior of Cast Iron in the Fire — On Designing Cast Iron Boilers. Cast iron as a material for boilers — The arguments in favor of cast iron as a material for steam generators and which must of necessity be confined to sectional boilers, are, 1. That the transmission of heat thrcyagh plates of an equal thickness of cast or wrought iron is in favor of the former. 2. That in point' of durability it excels wrought iron, for the following reasons : a. It will resist corrosion better than wrought iron. h. It is unaffected by the chemical impurities of feed water or the acids found in the pro- ducts of combustion. c. On account of its granular structure it is not possible for it to blister when subjected to a high heat in the furnace. d. It is not liable to be strained by inequality of temperature. 3. As the parts must of necessity be small, they are capable of resisting very high pressures and are not dependent on any system of stays or braces for strength. 4. Its low first cost, together with the certainty that any number of parts can be made exact duplicates of each OBJECTIONS TO CAST IRON. 9 other, and the facility with which these parts can be fitted not only for original use but to replace defective or worn out sections. 5. A defective section replaced by a new one renders a cast iron boiler as good as new, which is claimed as a very great advantage over wrought iron boilers, as a patch can never equal in strength the original plate. Objections — The objections urged against cast iron as a material for steam boilers are, 1. That it is an unsuitable material, in consequence of its treacherous nature when subjected to high or unequal temperatures^ * 2. That the cooling strains in the manufacture often produce flaws or other defects in the castings which are hidden to the eye and do not become apparent by merely testing them by hydraulic pressure, and which may, with- out a moment's w*arning, lead to a sudden and disastrous fracture. 3. Cast iron seldom gives warning by any of the indi- cations of weakness which characterize or precede the failure of wrought iron. 4. Cast iron being a crude product, there is no cer- tainty that castings can be made uniform in strength or in other qualities. 5. Cast iron boilers are objected to on account of defi- cient circulation due to their construction, and especially to the fact that they must be made in small pieces; that there is difficulty in getting the steam generated into the steam room of the boiler, without priming; but this is a question of design rather than one of material — yet, as the material can only be used in certain sizes and forms, the objection is entitled to consideration. The arguments here presented both for and against cast iron as a material for steam boilers are, in the main, those 10 A TREATISE ON STEAM BOILERS. offered by engineers and boiler makers when addressing actual or prospective purchasers. The reasons given for the rejection of cast iron as a material for steam boilers vs^ill doubtless be observed to be miscellaneous rather than specific. An inquiry into the relative properties of cast iron and wrought iron ought but does not show why the latter is preferred to the former material. Scarcely any two mines furnish iron ore of the same composition and it is doubtful whether any two fur- naces make an iron having precisely the same qualities. From this it will readily be understood why there may be some difficulty in making castings possessing certain qual- ities in the same degree. This latter, it may be said, is not absolutely essential to safety so long as the castings are sound and strong. The impurities usually found in cast iron and which give it character are, sulphur, which renders iron hot-shorty silicon or phosphorus, rendering it cold-short, and carbon, which gives to it its fusibility. The degree of hardness or softness of castings depends somewhat, but not entirely, upon the quantity of carbon contained in its composition. The carbon present in hard and brittle castings is often chemically combined with the iron ; while soft and tough castings contain perhaps the same percentage of carbon mechanically combined. The selection of iron for the foun- dry is a very important one, but can not be entered into here. The selection, however, must be such that castings shall possess moderate hardness, closeness of grain, strength and toughness. A comparison of the properties of cast and wrought iron will show that ordinary castings have sufficient strength for boilers, so that it is not on this account, but because of its unsatisfactory behavior in the furnace at a high temperature that has had most to do with its rejection. STRENGTH OF CAST IRON. 11 From a mean of many experiments it may be said that ordinary castings have a tensile strength of about fifteen thousand pounds per square inch, or 6.69 gross tons. When special care has been exercised in the selection and mixture of pig iron, castings may be made of a higher tensile strength, and tests show that a strength of twelve to fifteen tons per square inch may be obtained. This, however, should be regarded as a maximum attainment and does not refer to ordinary castings, nor especially to thin cored work required in the sections for cast iron boilers. The elastic limit of cast iron varies somewhat, but is not far from one-third of its breaking strain. This would give ^ve thousand pounds per square inch as the utmost limit of safety in common castings. Allowing a factor of safety for cast iron boilers of ten, a working pressure would then be allowed of .223 ton or ^yq hundred pounds per square inch of section. This would give for sections three-eighths inch thick a safe working pressure of one hundred and eighty-eight pounds per square inch. If it were a question of strength merely, this would be quite sufficient to meet every case in ordinary practice. But every experienced foundryman knows that castings can not be relied upon with any degree of certainty. Frac- tures in cooling are likely to occur at any point where two surfaces join each other at right angles. If they differ in thickness, or if the two pieces are of any considerable size, this is almost sure to be the case. Blow holes are so frequentl}^ found in castings that their presence is gener- ally admitted in all ordinary work; as they are mostly below the surface there is no determining where they are located, to what extent they exist, or in what direction they lead. In addition to this, the process of cooling in the mold after the casting is made, introduces a class of abnormal strains which are brought about by the cooling or fixing of one portion of a casting over that of another. 12 A TREATISE ON STEAM BOILERS. These strains are of a very complex character and fre- quently of themselves will distort if not fracture the piece containing them: Annealing is frequently resorted to in order to counteract or neutralize these strains. In large castings slow cooling is practised as much as possible, the effect of which is to develop a coarse, uneven grain, being finest near the surface and growing coarser and more irreg- ular toward the center. Where pieces join each other, cavities are likely to occur by reason of the irregular grouping of the crystals, which is one of the principal causes of these irregular strains. After a casting has been poured and consolidation begun, then, the more rapidly it can be safely cooled, the finer and more even will be the grain, and for any given metal the greater will be its strength. The cooling, in order to obtain the best results, should be uniform through- out the mass. To attain this, it may be necessary to uncover some of the thicker portions of the casting. It the cooling be unequal and at the same time quite rapid, injurious strains are brought into action which may have the effect, as already stated, to fracture the casting at a weaker point. The quality of cast iron may be judged somewhat by the appearance of the surface of fracture while still fresh. Soft and tough castings are coarser grained and have a less silvery luster than very hard castings. The judging of the quality of castings at sight can only be acquired by experience. Cast iron in the fire — The effects of intense heat on cast- ings is to melt off" all sharp projections and those parts necessary to the bolting of the pieces together. The m'etal almost invariably changes from the bright, granular appear- ance characteristic of good castings, to very coarse, uneven grains, having scarcely any metallic luster. It becomes THE HARRISON BOILER. 13 extremely brittle and is so unlike its former state that it is utterly unfit for further use in the foundry in the pro- duction of castings requiring strength. The continued heating and reheating of any metal would in time destroy it, but cast iron seems to be less able to withstand the efi:ect8 of severe heat and repeated cooling than wrought iron. So far, the behavior of cast iron in the fire has been anything but satisfactory and at present it meets with but little favor among engineers as a material for steam boilers. That form has much to do with its durability and safety ought to be admitted. The only cast iron boiler which has had an extended sale in this country is that designed by the late Joseph Harrison, Jr., and for many years manufactured by him at his works in Philadelphia. Mr. Harrison was an accom- plished and successful engineer, who gave many years of valuable time in improving the details of this boiler and conducting experiments on a large scale, which, fortun- ately, his abundant means enabled him to do. It is prob- able that any suggestion calculated to make this boiler a success had at least an intelligent and impartial trial. Not- withstanding all this, the boiler can not be said to have ever become popular. During the autumn of 1878, a gentleman contracted, through the writer, for a wrought iron boiler, to replace a Harrison boiler, which had for thirteen years previously furnished the steam for driving the machinery of his mill. He stated that during this time the boiler worked to his satisfaction. There are other examples which go to show that when suitable irons are employed, and the boiler properly designed and cared for when in use, cast iron may be used in the construction of steam boilers. The essential requis- ites seem to be, that pieces be small and free from angular projections, or changes of direction if these by any means 14 A TREATISE ON STEAM BOILERS. necessitate an increase of thickness at the line of juncture. The castings should be of uniform thickness throughout and contain no external bolting flanges, or other projec- tions, in the fire. Where boilers are made wholly of cast iron and subject to internal or bursting strains, the sections should be, pre- ferably, as nearly spherical as possible, and should in no case have flat surfaces of any considerable extent forming either the outside or inside of a boiler. Every section in a cast iron boiler must be strong enough to withstand the pressure of steam without any system of bracing, or stays of any kind, except those necessary to the bolting of the parts together to make a complete boiler. In the construction of boilers partly of wrought and partly of cast iron, the strains upon the latter should be. those of compression rather than those of extension. CHAPTER III WRuUGriT IRON AS A MATERIAL FOR STEAM BOILERS. Tenacity and Ductilit}^ of Iron — Properties of Iron, as Modified by Working — Welding — Texture of Wrought Iron — Effect of Cinder in Iron — Elasticity — Elastic Limit — Malleability — Flexure — Defects in Boiler Plates — Varieties of Plate Iron — Tests of Boiler Plate — Homogeneous Iron. Wrought iron is prepared, usually, from the harder vari- eties of pig iron, by a succession of processes such as refin- ing, boiling or puddling, squeezing, hammering, rolling, etc.; the primary object being to rid the iron of all the foreign substances contained in it which are calculated to reduce its strength and malleability, and, secondly, to pre- pare it in convenient size and shape for manufacturers' use. Wrought iron has for many years past been the princi- pal material employed in the construction of steam gener- ators of whatever kind. It has many qualities which make it a very desirable material for the purpose. That quality of boiler plate is judged to be the best which has the greater tensile strength, combined with ductility and malleability. These properties are affected in some measure by the impurities existing in the pig iron from which the plate iron is made, as well as the subsequent working the iron receives before being rolled out into plates. It is impossible to eliminate all the impurities in cast iron during its conversion into wrought iron. The follow- ing table gives the chemical analysis of a sample of boiler plate having a tensile strength of fifty-five thousand pounds per square inch : 16 A TREATISE ON STEAM BOILERS. Iron 99.20 Carbon 04 Manganese. 17 Silicon 15 Sulphur, 03 Phosphorus 21 Oxygen 20 100.00 The above iron contained, and is included in the above analysis. 0.80 per cent of cinder. The noticeable thing in any analysis of wrought iron is the small percentage of contained carbon. In order to show how nearly the impur- ities in pig iron are removed during the process of con- version, the chemical analysis of an average sample of white iron is given below, by which a comparison is easily instituted: Iron , 89.14 ,-. , fCombined 2 45 Carbon-^ _ -,_ (Free 87 Manganese 2.71 Silicon 1.11 Sulphur 2.51 Phosphorus .91 100.00 Wrought irons should possess in a good degree the fol- lowing properties : Tenacity, Welding power. Ductility, Each of these properties are influenced in some measure by the impurities in the iron, which may produce the fol- lowing defects : 'O Cold-short iron is very brittle when cold, cracking badly, or breaking if bent at a sharp angle or doubled; but may be forged and welded at a high heat. This defect TENACITY OF IRON PLATES. lY occurs in irons which have an excess of phosphorus. Red- shorty or hot-short iron, may be tenacious when cold, but easily broken when hot; it welds with great diffi- culty, though tough and reliable when taken directly from the bar and used cold. Red shortness occurs in iron con- taining an excess of sulphur. Tenacity is that property in a material by which it resists a force which tends to separate or tear it asunder. This is a very important property in irons intended for steam boilers. The tensile strength of American boiler plate will range from forty thousand to sixty thousand pounds per square inch. Unless portions of the plates have been actually tested or the plates are known to have been made from blooms of the very best quality, it is not safe to assume a greater tensile strength than forty- tive thousand pounds per square inch of section. This applies to such irons only as are stamped by reputable makers as C. H. I^o. 1, and higher grades; these latter are usually designated by some private brand or trade mark. Some of these special irons are stamped and guaranteed sixty thousand pounds. Boiler plates may possess high tensile strength at the expense of other qualities, such as homogeneousness and toughness. There are manufacturers of boiler plate who express doubts as to whether an iron suitable for steam ' boilers can be made having all the necessary qualities and at the same time possess a tensile strength greater than lifty-live thousand pounds per square inch. They assert that the iron becomes harder and more brittle as the ten- sile strength increases, and that the properties of hardness and brittleness introduced into the sheets by far outweigh any advantages which may be gained by the increased ten- sile strength. (8) 18 A TREATISE ON STEAM BOILERS. Toughness is an invaluable property in boiler plate., and means a combination of qualities, such as hardness, tenacity and ductility, by which the material is better ena- bled to withstand the effects of irregular strains, and frac- tures induced by concussion or bulging. Ductility is that property which a material possesses — like iron, for example — of being drawn out without break- ing. This elongation is produced by subjecting the iron to a tensile stress higher than the elastic limit when a perma- nent change of form takes place. It is found that tenacity has more influence upon the ductility of metals than mal- leability. We are thus led to expect that there will be something in common between the tensile strength and ductility of wrought iron. This will be affected somewhat by the quality of the original bar and the treatment it receives by subsequent working. The following table* shows the effect produced by dif- ferent modes of working, changes of temperature, etc. The conclusions given are founded upon a large number of experiments by Mr. Kirkaldy and others : TABLE I. ON THE PROPERTIES OF IRON, AS MODIFIED BY WORKING. TENSILK STRKNGTH. DUCTILITY. Reducing diameter by roll- ing Increased Reduced. Turning or removing the skin No alteration No alteration. Reducing diameter by forg- ing Annealing Increased Reduced. Reduced Increased. Welding ,. ( Reduced from between^ (^ 4.1 and 43.8 per cent, j ■ Reduced. =■' From "Notes on Building Construction," Rivington's London, 1879. TEXTURE OF WROUGHT IRON. 19 TABLE I — Continued. TENSILE STRENGTH. DUCTILITY. Stress sadiiJetilj' applied Additional Hammering Hardening in water or oil... Cold Toiling— plates Cold rolling — bars Reduced 18.5 per cent Increased Reduced in nearly all cases. Reduced. Increased Reduced. Doubled Destroyed. Reduced 60 per cent. Increased 50 per cent No difference Galvanizing Effect of frost, 23° F Reduced 2.3 per cent Reduced 3.6 per cent Reduced 8 per cent. Effect of frost, stress sud- denlY applied Reduced between and 30 per cent. Texture of wrought iron — Irons are usually said to be in texture either fibrous or granular. When wrought iron has been forged under a hammer directly from a bloom the forging presents a granular or jagged grain; this grain is not uniform in size in large forgings, being coarser the center and finest near the surface. If the in process of hammering be continued, it will become, when reduced to smaller bars, uniformly fine grained. If, however, instead of this continued hammering, the original forged billet be run through a train of rolls the texture will have changed from granular to fibrous. M. Janoyer in a paper on the texture of iron* main- tains that iron presents but a single texture, and that is the granular one ; all others are only metamorphoses of this, due to defective temperature at the moment of finish- ing, which does not permit Complete welding of the entire mass. He suggests classifying wrought iron into welded, non-welded and imperfectly welded irons, instead of fibrous and granular. When iron is pure and homogeneous its tex- ture is granular. The operation of puddling consists in stirring a mass of spongy iron in the midst of a bath of -'Journal Franklin Institute, vol. 68. 20 A TREATISE ON STEAM BOILERS. cinder, which prevents the intimate approximation of its particles. This opposes a thorough welding of the mass and favors the production of a fibrous texture ; since dur- ing the subsequent working, the molecules can slide over each other, thus giving to the iron its fibrous appearance. The temperature at which iron is rolled has much to do with determining its texture ; for example, if two or more bars of crude granular iron be laid one above the other to form a fagot, and this fagot be raised to a welding heat and passed through a set of rolls, the result will be granular iron, if the welding temperature be maintained ; if, however, the temperature falls below the welding point the texture will then be fibrous instead of granular, because of the unequal temperature of the bar, which permits the naolecules or particles of the iron to slide over each other during the process of rolling. Cinder — All wrought irons contain more or less cinder in their composition, and the fibrous texture of iron may almost always be traced to its presence, especially when worked in the rolls at too low a temperature. The pres- ence of cinder always prevents perfect welding. Squeezing' the blooms as they come from the puddling furnace will remove a considerable portion of the cinder, but all blooms intended for boiler plates should be worked under a heavy steam hammer until all the cinder is worked out of it, if such a thing is possible. The presence of the cinder, oxide of iron, or any other substance between the surfaces of the two plates of iron will prevent their welding ; these foreign 'substances between plates are the cause of blisters in boiler plates, by preventing perfect welding. Malleability is that property by which bodies may be drawn out by forging or hammering. Soft and fibrous are more malleable than hard or granular irons. DEFECTS IN IRON PLATES. 21 Boiler plates seldom require reducing in thickness, or otherwise wrought, except at joints in which three or more plates intersect. Any iron at all suitable for boilers will possess this property in a sufficient degree. Flexure — A very important property in iron for boiler plates is that of flexure, or bending. In every act of bending or flanging boiler plate there are two forces to be overcome : 1. The extension of the metal on the outside of the curve. 2. The compression of the metal on the inside. As might be expected, thoroughly welded or granular irons bend easier than flbrous. Boiler plates which will stand flanging or bending to a right angle both with and against the grain when heated to a cherry red, and with- out cracking or breaking in the curve,, will be found suit- able for any ordinary boiler work. The lower grades of iron will scarcely stand such a test, except at a high heat and for narrow widths. The defects in iron boiler plates are principally imperfect welding, brittleness and low ductility, all of which may be largely overcome by a proper selection of materials in the earlier stages of its manufacture and by a careful manipu- lation during the successive operations of reheating, weld- ing, and especially by a thorough working under a heavy steam hammer. Ordinarily, the selection of particular brands of iron for the manufacture of boiler plate is entirely beyond the control of the persons who are to use the iron. Hence, irons of this class are usually guaranteed by the makers to be of a certain tensile strength. This is usually satisfactory to pur- chasers, on the general belief that if the specimen tested has an average tensile strength of^ say, fifty thousand pounds per square inch of section, it possesses the other qualities 22 A TREATISE ON STEAM BOILERS. needed for a good boiler plate not requiring flanging, and in this manner for plates required for any service. Varieties of plate iron — The wrought iron plates now regularly offered in the market are known as either C. — C. 'No. 1 or C. H.— C. H. No. 1— and C. H. No. 1 flange. C. IRON, or charcoal iron, is the common boiled or pud- dled iron, rolled into bars or plates. This grade of iron is porous, and will become very brittle with repeated heating and cooling. It will not stretch much before breaking and will break suddenly. Its tensile strength ranges usually from thirty to forty thousand pounds per square inch. It is only suited for tank work, and ought never to enter into any portion of boiler construction. C. No. 1 IRON, or C. H. iron (charcoal hammered, as it is oftener known), is the same iron as the above, except that it is subjected to more careful working and is ham- mered into suitable blooms before rolling. This iron very much resembles the common iron in its general qualities, having but little elasticity and breaking with a sudden jerk. Like the above, it becomes very brittle by repeated heating and cooling, though somewhat stronger than C. iron ; its tensile strength ranging from thirty-five to forty- five thousand pounds per square inch. It is not a suitable iron for boiler construction. C. H. ]^o. 1 SHELL IRON is made from C. H. blooms, with the addition of selected scrap, the whole being thoroughly welded under a heavy steam hammer and afterward rolled into plates. This iron, like the two others just described, is injuriously affected by repeated heating and cooling, which has the effect to render it brittle. This is the quality of plate generally used in the construction of land boilers using pressures of steam below eighty or ninety pounds TENSILE STRENGTH OF IRON PLATES. 23 per square inch. It rarely enters into the construction of boilers for river or ocean service; its principal defect being a lack of homogeneity and imperfect welding. Its tensile strength is from forty to fifty thousand pounds per square inch. Shell irons are often made of a much better quality and higher tensile strength than the above, when ordered for any particular purpose. The following table gives the mechanical tests to which ten samples were subjected, and which were taken from boiler plates rolled for river steam- boat service ; five samples from Phillips, Nimick & Co., and five from Lloyd, Son & Co., both firms manufacturing at Pittsburg, Pa. The tests were made by Mr. G-eorge H. Atkinson, inspector of steam boilers at that point, the testing machine used being the design of Rehlie Brothers and of the kind furnished the United States government. TABLE II. TENSILE STRENGTH OF C. H. No. 1 BOILER PLATE. SAMPLE. BREAK- ING WEIGHT. TENSILE STRENGTH PER SQUARE INCH. ELONGA- TION IN PARTS OF AN INCH. TIME CONSUMED IN TEST, IN MINUTES AND SECONDS. WEIGHT ON MA- CHINE AT WHICH ELONGA- TION COM- THICK- NESS. WIDTH. REMARKS. MENCED. .25 .96 18,000 75,000 .125 MIN. SEC. 4.30 14,500 Stamped Phil- .26 1. 00 17,600 67,692 .1875 4.00 14,500 lips, Nimick & Co., C. H. .25 1.00 17,000 68,000 .1875 4.00 14,000 No. 1, 57.000. .26 1.00 18,600 71,538 .1875 5.00 15,000 Short speci- men. .26 .90 16,800 71,794 .1875 3.30 14,000 .24 1.00 14.900 62,083 .1875 4.00 13,600 Stamped Lloyd, .24 1.00 14,700 61,250 .125 4.00 13,500 Son & Co., Pittsburg, .24 1.00 13,800 57,500 .1875 3.30 12,400 57,000. Short •2* 1.00 13,9j0 57,916 .T875 4.00 13,000 .specimen. .24 1.00 14,200 59,166 .1875 3.30 12,500 24 A TREATISE ON STEAM BOILERS. C. H. 'No. 1 FLANGE IRON is Similar to the above, the diflerence being that only the very best scrap iron and charcoal hammered blooms are used. The greatest care is exercised in the selection of materials, and the working in the forge is such as to insure thorough welding. In tex- ture it is less fibrous and more granular than any of the irons preceding it. On account of its nearer approach to a homogeneous structure, it is less liable to blister or crack in the fire. It will stand repeated heating and cooling, and should have good flanging qualities. The tensile strength should never fall below fifty thousand pounds per square inch, and does not often exceed sixty-five thousand pounds. The elastic limit will vary from eighteen thous- and to twenty-five thousand pounds per square inch, and will stretch from twenty-five to thirty per cent in ordinary two inch specimens. This is the highest grade of iron regularly offered in the market and is quite extensively used in the construction of marine boilers and for the heads and other flange plates of land boilers. Plates of this quality of iron are usually branded with the name of the maker and the guaranteed tensile strength ; thus: SMITH, JONES & CO. C. H. No. 1 FLANGE, 57,000 This method of stamping was introduced in order to meet the requirements of the government regulations with reference to the quality of plates entering into steam boil- ers intended for use on board steam vessels in the United States. The pressure of steam allowed to be carried is determined upon the shape pf the boiler and the tensile strength of the material ; hence the figures stamped upon the plates ought always be below the actual tensile strength of the plates bearing them. A sample sheared from several plates bearing the stamp of the makers and intended for SLIGO FLANGE IRON. 25 steamboat boiler service were taken to the Custom House, Pittsburg, Pa. and tested by Mr. Atkinson, with results as given below : TABLE III. TENSILE STRENGTH OF PHIFJJPS, NIMICK A- CO., C. H. No. 1 FLANGE IRON". o7,ono. THICK- NESS. .26 .26 .25 .25 PLK. BliKAK- ING WEIGHT. WIDTH. 1.00 20,600 1.00 16,700 1.00 19,:mo 1.00 19,900 TENSILE ELONG.\- STRENGTH TION IN PARTS OF SQUARE AN INCH. INCH. 79,230 .1875 64,230 .1875 77,200 .1875 79,600 .1875 TIME CONSUMED IN TKST, IN MINUTES AND SECONDS. MIN. SEC. 5.00 3.30 4.00 4.30 WEIGHT ON THE MACHINE AT WHICH ELONGA- TION COM- MENCED. 16,500 14,500 16,000 16,500 RKMARKS. Short specimen. U. 8. regulation The writer was shown at the works at Phillips, I^imick & Co., Pittsburg, Pa., another grade of flange iron named by them sligo c. h. no. i flange, which was guaranteed sixty thousand pounds tensile strength at its lowest limit- Specimens of both hot and cold flanging shown at their works attest the superior quality of this brand of iron. Its ten- sile strength, as given by them, w^as from sixty thousand to sixty-five thousand pounds per square inch, w^ith an elastic limit of from twenty thousand to twenty-two thousand pounds and a stretch of twenty-eight to thirty per cent. Samples of this iron, taken from plates rolled for a boiler intended for a western steamboat and tested by Mr. Atkinson, gave results as follows: 26 A TREATISE ON STEAM BOILERS. TABLE IV. TENSILE STRENGTH OF PLATES STAMPED, PHILLIPS, NIMICK A CO. H. No. 1, SLIGO, 60,000. SAMPLE. BREAK- ING WEIGHT. TENSILE STRENGTH PER SQUARE INCH. ELONGA- TION IN PARTS OP AN INCH. TIME CONSUMED IN TEST, IN- MINUTES AND SECONDS. WEIGHT ON THE MACHINE AT WHICH ELONGA- TION COM- MENCED. THICK- N ESS. WIDTH. 1.00 1.00 1.00 .80 1.00 .86 REMARKS. .23 .23 .23 .31 .24 .28 14,800 16,200 15,600 16,100 15,700 16,600 64,347 70,434 67,826 64,919 65,416 68,936 .1875 .1875 .1875 .1875 .25 .1875 MIN. SEC. 3.00 3.30 3.30 4.00 3.30 4.00 13,000 14,000 14,000 15,400 13,500 14,500 Short specimeti U. S. regula- tion. This firm make another and higher grade of iron which they call sligo special. It is a high grade of flange iron, specially adapted for the construction of all kinds of steam hoiler work. This iron will stand working into any shape in which it is possible to work iron, and the makers claim that its qualities are improved by repeated heating and cooling, an assertion borne out by exhibiting specimens which had many times been reheated and cooled and then doubling the plate cold. The guaranteed tensile strength was given at from sixty-two thousand to sixty-eight thou- sand pounds per square inch, with an elastic limit of twenty-five thousand to twenty-eight thousand pounds and a stretch of from thirty to thirty-three per cent. Another grade of iron is manufactured by them called SLiGO FIRE BOX IRON, having qualities the same as the pre- ceding iron, and a tensile strength from sixty-four thousand to seventy thousand pounds per square inch and elastic limit from twenty-eight thousand to thirty thousand pounds, with a stretch from thirty to thirty-three per cent. HOMOGENEOUS IRON. 27 These irons are free from anything like brittleness, are tough and have a homogeneous granular texture, with occasional fibers of silky luster in bending fracture. The writer regrets that he w^as not able to obtain at this time samples from plates rolled to order, that special tests might be made. The figures given above are those resulting from tests made by the company in their laboratory upon a Rehlie Brothers' testing machine, for their own guidance in its manufacture. Homogeneous iron or (as it is oftener called) mild steel, is a somewhat recent term, used to designate a wrought iron of uniform granular texture throughout its mass ; it is not necessarily to be considered as purer than other wrought irons and may contain in some degree most, if not all, the elements usually considered as impurities in pig iron. The term homogeneous in this connection simply implies that the iron is of the same kind or of the same nature throughout the pjate. It should be entirely free from cinder, as it would be impossible to make a homogen- eous iron with cinder in its composition, for the reason that it has no affinity for and being of an entirely different nature from iron, will not combine with it; the presence of cinder in any iron prevents contact or perfect welding, by keeping the molecules of iron asunder and is one of the reasons for the fibrous character of ordinary wrought bar and plate irons. Homogeneous iron can best be made by a suitable pre- paration of the iron by either the Bessemer or Siemens- Martin process, or by melting wrought iron in crucibles and then casting into a solid ingot, from which the plates or bars may afterwards be made. This material is known in the market under the names of homogeneous iron, mild steel and hom.ogeneous steel. The name of ingot iron has also been proposed. The latter is, 28 A TREATISE ON STEAM BOILERS. perhaps, more nearly correct than the three former and it is probable will come into general use in time. As iron does not sensibly harden unless it contains at least 0.30 per cent of carbon, it would appear that the use of the terrp. steel is scarcely allowable. At present, however, the ques- tion among boiler makers is, broadly, Iron vs. Steel, and in order to keep the two separate the term steel will be used in this book to designate the particular material just described, though homogeneous or ingot iron is, as already said, more nearly correct. CHAPTER IV. STEEL AS A MATERIAL FOR STEAM BOILERS. Faults of the Earlier Steel Plates — Qualities in Steel which Recom- mend it as a Material for Boilers — Its Nature must be Studied — — The Defects of Steel — Homogeneous Plates — Impurities which Affect the Quality of Plates — Tensile Strength of Steel Plates — Crucible Steel Plates — Bessemer Steel Plates — Open Hearth Steel Plates. Steel is usually spoken of as an intermediate metal between wrought and cast iron, its position being deter- mined by the quantity of carbon contained in its compos- ition. For the higher grades of steel this may be true, but as the quantity of carbon in steel boiler plates is often less than is found in samples of wrought iron, this definition, then, is defective. The difference between steel and wrought iron does not consist entirely in the quantity of carbon contained in the former over the latter, but rather that steel has been melted and cast into a malleable ingot, which is an entirely different thing from puddling and one in which the quantity of carbon contained in it has noth- ing to do, especially when present in very small quantities, as in mild or very soft steels. When there is carbon enough in steel to cause hardening when suddenly cooled, it then plays an important part in its quality and imparts to it properties which a,re not wanted in boiler plates, but which are valuable in steel intended for tools and other purposes. Steel is characterized by a fine granular texture, and when the contained carbon amounts to 0.40 to 0.50 per cent, it has the property of hardening and taking a tern- 30 A TREATISE ON STEAM BOILERS. per. There are several varieties of steel, diflering in strength, hardness and ductility. The particular quality of steel best suited for boiler plate contains from 0.12 to 0.20 per cent of carbon, or so little carbon as to permit a red heat and sudden quenching, without destroying the property of flexure. Ahigherpercentage of carbon increases the tensile strength, at a loss of ductility. The advantages of steel as a material for boilers were recognized many years ago and was so employed to a moderate extent. A leading article in Engineering, 1878, says : " It was not till about fifteen years ago, when plates of Bessemer steel were offered to the makers in quantities, that the use of steel for boiler making can be said to have become fairly established. Even up to the present day its use to any considerable extent for stationary boilers has been confined, with few and unimportant exceptions, to some half-dozen boiler works in the Manchester district, but these are of the very highest standing. Only two of these, however, have used steel extensively for shells, the rest having con- tented themselves by using it chiefly for the furnace tubes. Bessemer steel plates have been used for boilers of various kinds by upwards of fifty other makers in different parts of the kingdom, but as a rule against the advice of these makers, and (shall we say consequently?) often with unsat- isfactory results." "For marine boilers, steel plates have been used only to a very limited extent. Of forty of the best known firms of marine engine builders, including those who make for the Admiralty, up to a very recent date only about half a dozen had used steel plates and half of these would not have used them if they had had their own way." The steel furnished in this country, as well as that made abroad during most of this time, not possessing the proper- ties required to make its employment a success, fell into dis- WORKING STEEL PLATES. 31 favor and has been for along time under a cloud; many man- ufacturing establishments well known to the writer declin- ing to have anything to do with it — others employed it because contracts called for it, but with the understanding there should be no recourse for damage in case of failure. No doubt this has had much to do with the little attention given to the production of a reliable steel boiler plate. At this time, however, a marked change is observed in manu- facturers and users alike. There is a growing demand for steel boilers, not only in this country, but in Europe. From the present outlook it seems almost certain that the boiler of the future will be of steel. Steel as a material for steam boilers recommends itself on account of its homogeneity, tensile strength, malleabil- ity, ductility, freedom from laminations and blisters. It requires greater care in working than is usually given to iron. It is a higher material and requires a higher intel- ligence to properly work it. This intelligence means a knowledge of the properties and peculiarities of the mate- rial. Steel differs so much from wrought iron that in order to work it properly its nature must be studied and understood. To demand that it shall conform to all the ordinary practice of working w^rought iron is absurd. If it can do so, well ; if not, then the method of working must conform to the nature of the mate- rial. Steel is not a nriaterial of definite quality and its properties vary w^ith each change of quality. It can be made almost as hard as a diamond, certainly hard enough to cut glass. There is no substance known which equals in elasticity a good steel watch spring. It is possessed of a toughness which is unapproached by any other kind of metal ; it has strength in all directions and before it breaks it will yield even to fifty per cent. It njay be hardened, tempered or annealed at will. During these processes the material is studied and is worked in such manner as is best 82 A TREATISE ON STEAM BOILERS. suited to its quality. IS^o one thinks of subjecting bar iron and bar steel to the same Ireatment in the forge or work- shop. It is not unlikely that many failures in steel boiler plates have arisen from the want of this very precaution at the outset. A steel plate was used just as an iron plate^ and because it failed under such treatment the material was condemned as untrustworthy and dangerous — a sweep- ing verdict, which can come only of impatience, careless- ness or ignorance. There is no doubt that many of the earlier faults in steel were due to imperfections in manufacture or impru- dent handling and cooling after rolling. But now that plates are carefully made and annea^led after shearing to dimensions, the burden of the responsibility rests largely upon the boiler maker. Quoting again from Engineering the writer sums up his review of steel for boilers as follows: " That of some eighty boiler makers who have fairly tried steel plates, only some eight or nine can be said to have persevered with its use and used it extensively; that where the use of steel plates has been persevered in against the advice and feeling of the boiler maker, the result has generally been unsatisfactory ; that it may be taken for granted that the prejudice on the part of boiler makers against the use of steel is, as a rule, inversely proportion- ate to the extent of their acquaintance with it. It would appear that those makers who have not been alive to the difference required in the working and treatment of iron and steel, or who have gone timidly to work and let the workman find out for themselves the best way to treat steel, have usually had trouble and have only been too glad to receive a confirmation of their adverse opinion." The defects in boiler 'plates — For steel, the principle defects are brittleness, low ductility, and flaws induced by DUCTILITY IN STEEL PLATES. 33 the presence of cavities formed by bubbles of air or gas in the original ingot. The two former may partially be overcome by a still further removal of the foreign substances which aflect the softness of steel and by reducing it to a more nearly pure iron. The latter is not so easily overcome ; it is doubtful whether a cavity once formed by a bubble of air or gas in the body of an ingot can ever be welded by subsequent hammering or working of any sort, owing to the interior surface of the cavity being lined with a film of oxide which may be brought into close surface contact, but not welded. Such a cavity, flattened down during the process of hammering and rolling into a mere surface con- tact, must be regarded as an incipient fracture, which may at any time spread to almost any extent and in any direc- tion, when the conditions are such as to induce it. The harder the steel the greater the certainty of such exten- sion of fracture; this tendency is diminished as softness and ductility are increased. In steel plates ductility is a property of very great importance, for without it plates are liable to give way without any of the usual indications of failure or even a moment's warning. Other things being equal, ductility increases in this material as its tensile strength is dimin- ished. It is only in homogeneous irons or mild steels, as they are usually called, which' possess this property in the highest degree, and these are not usually made having a tensile strength higher than about seventy thousand pounds per square inch; a reduction to sixty thousand or even fifty-five thousand pounds will be found to be still more ductile. Some experiments by Mr. Charles Huston on American steels exhibited the following results: (4) 34 A TREATISE ON STEAM BOILERS. TABLE V. TENSILE STRENGTH. CONTRACTIOK OF AREA, PER CENT. Crucible steel (not quite hard enough to temper) Crucible steel (ordinarily soft) 78,3G6 64,0(10 54,600 26.66 36.83 Siemens-Martin steel (exceptionally soft) 47. The increase in ductility, in proportion to the decrease in tensile strength, is quite marked. There is a limit to the amount of ductility which can be given homogeneous plates, arising from the practical difficulty in the manufacture of solid ingots. This diffi- culty is not entirely confined to mild steels, though the ingots are apt to be more spongy in a soft and ductile metal than in the harder varieties. This fact has engaged the attention of steel makers for some years and plans for compressing the fluid steel have been suggested by several prominent manufacturers, among whom are Sir Henr^^ Bessemer and Sir Joseph Whitworth. The latter subjects the molten steel to a pressure of some six tons per square inch, by which all cavities are closed up, the gases contained in them driven out, the metal being compressed to about seven-eighths of its original bulk, its density and strength being greatly increased. Owing to the groat cost of compressing steel by either of the above methods, it can not be at present adopted in the commer- cial production of boiler plate. The writer saw at tbe Edgar Thompson steel works, w4iat is now their regular practice, the compression of steel ingots by steam. After pouring the ingot a cap is placed over the top of the mould and securely fastened by a key, making a steam tight joint. A flexible tube leads from this cap to a conveniently arranged steam pipe. A pressure of about seventy-five pounds of steam is used in compress- HOMOGENEOUS STEEL PLATES. 35 ing the fluid ingot, and has given very satisfactory results. The absence of anything calculated to impair the quality of the ingot is a valuable feature in the process. Homogeneous steel plates are expected to possess in a good degree, tenacity and ductility, and to be more nearly equal in these properties when tested both lengthwise and across the grain than is usual in fibrous wrought iron plates. In Mr. Kirkaldy's tests of Krupp's and Yorkshire iron plates, the differences in tensile strength were found to be as follows: LENGTHWISE OF THE GRAIN. Krupp — Stress per square inch of fractured area 85,144 lbs. Yorkshire — Stress per square inch of fractured area 61,140 lbs. ACROSS THB GRAIN. Krupp — Stresss per square inch of fractured area 65,359 lbs. Yorkshire — Stress per square inch of fractured area 54,110 lbs. These specimens were nnannealed. The figures show an average of nine specimens of Krupp's iron and an average of eighteen of the Yorkshire iron. It will be observed that in Krupp's iron the difference in tensile strength, when taken in the two directions, stands 85,144 to 65,359, or the iron is 30.3 per cent stronger in the direction of the fiber than across it. And similarly the Yorkshire iron has an increased strength of thirteen per cent in the direction of the fiber over that taken from across the plate. Mr. Kirkaldy made some tests of the Landore-Siemens steel for the English Admiralty in 1875, in which it was shown to be a remarkably homogeneous metal, with results as follows : Unannealed plates, 0.37 inch thick, 10 inches between supports, ultimate strength per square inch length- 36 A TREATISE ON STEAM BOILERS. — , • wise of grain, 72,878 pounds; ultimate strength per square inch across the grain, 72,670 pounds; or a difference of only .00286 per cent, showing it to be much superior in this particular property than either of the two former irons. A homogeneous steel plate will be a doubtful gain if secured at the expense of even a partial loss of ductility over the very best iron plates now manufactured. One of the principal faults of a homogeneous plate is its liability to fracture from very slight surface or edge imper- fections when under high tension — imperfections which would scarcely, if ever, affect a fibrous iron plate. In such a case the stronger steel plate would obviously be inferior to an iron plate, not in strength, but in trust- worthiness. A fracture once begun in a homogeneous plate will extend from the edge into the body of the plate if that be the direction of least resistance. In this respect it is almost the very opposite of iron, which usually con- fines its fractures to the line of rivet holes, or if in the body of the plate the fracture usually follows the direction of the fiber slowly and does not extend in the rapid man- ner in which it is apt to do in a steel plate. So far as correcting mere fractures in a plate are con- cerned much can be said in favor of a fibrous over a homo- geneous material. It is not an uncommon practice where fractures are discovered in iron plates to stop its extension by simply drilling a hole at the end of the fracture and inserting a rivet — the fractures often being repaired without removing the plate. There is little doubt that a homogeneous plate will resist strains which induce fracture much longer than iron plates of the same thickness, but once the fracture is started a homogeneous plate will allow its extension in a shorter time and to a greater extent than a tough fibrous plate would. Still, with all its drawbacks, a good tough IMPURITIES IN STEEL. 37 homogeDeous metal of reasonable tensile strength and high ductility is, all things considered, the best material that can be selected for boiler construction. The ordinary impurities which affect the quality of steel are phosphorus, sulphur and silicon. Phosphorus renders steel cold-short, and as boiler plates are usually worked cold, the less there is of it in the plates the better. The highest allowable limit in good steel boiler plate is 0.08 per cent and should not exceed 0.05 or 0.06 per cent if possible; it having no perceptible effect on plates at that percentage. Sulphur renders steel hot-short and thus affects the working in the steel works rather than in the boiler shop, except in flange plates. Sulphur should not exceed 0.05 per cent in steel boiler plates and even at this percentage the plates should contain at least 0.25 of manganese in order to counteract the hot- short effects of the sulphur. Silicon in steel boiler plate, even in small quantities, renders it hard and decreases its ductility. It ought not to exceed 0.05 per cent in any steel intended for steam boilers. Copper is sometimes found in steel and when present in any appreciable quantity renders steel hot-short and has a marked effect upon its welding properties when present in quantities exceeding 0.03 to 0.05 per cent. The effect of carbon in steel is to increase its hardness and to decrease its fusibility and welding power. The following table shows the effect of different quan- tities of carbon in steel and iron : 38 A TREATISE ON STEAM BOILERS. TABLE Vr. SHOWING THE CHARACTERISTICS OF IRON AND STEEL FOR DIFFERENT PROPORTIONS OF CONTAINED CARBON. {Bauermann's Metallurgy). NAME. PERCENTAGE OF CARBON. PROPERTIES. 1. 2. 3 Malleable iron-" Steely iron Steel 0.25 0.35 0.50 1.00 to \M 1.75 1.80 1.90 2.00 6.00 Is not sensibly hardened by sudden cooling. Can be slightly hardened by quenching. Gives sparks with a flint, when hardened. Limits of steel of maximum hardness and teuacitv 4 Steel 5 Steel Superior limit of welding steel. Very hard cast steel, forging with great difficulty. Not malleable hot. fi Steel 7 Steel 8 Cast iron Lower limits of cast iron, can not be hammered. 9 Cast iron Highest carbureted compound obtainable. The percentage of carbon in the above table is greatly in excess of that used in the manufacture of boiler plate ; the quantities in actual use for this grade of metal may be found in the analyses of the different samples of steel as given in this chapter. Tensile strength of steel boiler steel — This is a subject on which opinions have, in the past, widely differed. There is little doubt that the earlier steel boiler plates were made of too high tensile strength and too little ductility. At present most English engineers require that the plates shall in no case exceed twenty-nine tons (64,960 lbs.) per square inch. It is found, however, that steel with a strength of twenty-six tons (58,240 lbs.) per square inch will weld better and with more certainty than steel of a higher strength. A mild steel is more eas- ily worked and less likely to be injured by careless hand- ling than steel of high grade, and if it can be kept as low '"Wrought iron, not malleable cast iron. REQUIREMENTS OF LLOYD's REGISTER. 39 as sixty thousand pounds tensile strength per square inch, preserving ductility and toughness, it will be amply strong and will meet every ordinary requirement in boiler con- struction. Requirements of steel plates entering into the construc- tion of steam boilers made under the supervision of Lloyd's Register of British and Foreign Shipping : 1. "The material to have an ultimate tensile strength of not less than twenty-six tons (58,240 lbs.) and not more than thirty tons (67,200 Hbs.) per square inch of section, 2. "A strip cut from every plate used in the construc- tion of the furnaces and combustion chambers and strips cut from other plates taken indiscriminately, heated uni- formly to a low cherry red heat and quenched in water of 82° Fahrenheit, must stand bending to a curve of which the inner radius is not greater than one and a half times the thickness of the plates tested. 3. "All the holes to be drilled, or if they be punched the plates to be afterwards annealed. 4. "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. 5. "The boilers upon completion to be tested in the presence of one of the society's engineer surveyors, to not less than twice the intended working pressure." The three competing steels now in this market are cru- cible steel, Bessemer steel and the Siemens-Martin steel. The latter is oftener known as Open Hearth steel. These are to be regarded as distinguishing processes rather than three different kinds of steel, as they do not necessarily produce a material having chemical or mechan- ical properties widely differing from one another. Crucible steel boiler plate — The practice of Park, Brother CO -* w -! a a EH ■< 1-3 P 511 12 32 7 38 INCH. 1.00x0.310 1.00x0.311 1.00x0.323 1.00x0.325 SQ.IN. 0.310 0.311 0.323 0.325 SQ.IN. 0.151 0.132 0.154 0.159 INCHES in's. lbs. LBS. 19,800 20,330 21,660 22,715 LBS. LBS. 63,871 65,370 66,935 69,277 P.CT_ 52.0 ' 58.0 52.4 51.0 P.CT.. 513 515 517 The numbers obtained by dividing the breaking strain by the area of least section after fracture, sometimes called ''the tensile strength per square inch of fractured area," give a valuable measure of the toughness of the material. These numbers are as follows, for the several specimens : No. 506 159000. No. 507 160000. No. 508 140000. No. 509 138000. No. 510 161000. No. 511 131000. No. 512 .153000. No. 513 155000. No. 514 155000. No. 515 140000. No. 517 143000. No. 518 166000. The strains were applied gradually in all cases. With specimen Fo. 516, the breaking strain was not observed; but ]^o. 518, of the same material, was after- ward broken and the result recorded. 60 A TREATISE ON STEAM BOILERS. REMARKS BY OTIS IRON AND STEEL COMPANY. The above samples contained the following percentages of combined carbon : No. 28, .0014 per cent ; No. 12, .0013>^ per cent; No. 32, .0014 per cent; No. 7, .0012 percent; No. 38, .0014 per cent. Samples marked ^^^^^' were taken from across the sheet. Sample marked 28 H was heated red hot and cooled in water before being broken, and although the strength is increased, there is no perceptible increase in the hardness when tried with the file. Samples No. 28, No. 12, No. 32, No. 7, were uniform in mixture and quality of stock used. Several very severe mechanical tests, to show the qual- ities of open hearth homogeneous boiler plate, were made with cold plates at the works of this company at Cleve- land, Ohio, in the presence of the writer, to whom the samples were also given. The pieces tested were sheared off the ends of plates, which were then being cut to dimen- sions in their ordinary business routine. Several unannealed samples from plates five-sixteenths of an inch thick were folded down in the usual manner by blows given by a heavy hammer. Samples were also sub- jected to a shearing test, in which a piece of steel three inches in width was sheared up to within an eighth of an inch of the edge without exhibiting any signs of fracture, the "shearing" being depressed more than half an inch on the opposite edge of the plate. A number of these " shear- ings" were made at a distance of about five-eighths of an inch apart. A sample about eight inches square, taken oft another plate selected at random on the floor of the mill, was folded over fiat and afterward folded again at right angles to the first, the whole being then hammered flat, making a specimen about four inches square and one and a quarter inches thick, without exhibiting any signs of fracture. A selection of five tests was taken from their labora- tory record to show the range in tensile strength allowable in this system of manufacture. The tests were made with the ordinary two-inch specimens, with results as follows: OPEN HEARTH STEEL. 61 TABLE XVI. SHOWING THE RANGE IN TENSILE STRENGTH OF OPEN HEARTH STEEL PLATES MADE BY THE OTIS IRON AND STEEL COMPANY. i% ^ testing machine and figure 4. weights added until it breaks. This breaking weight, mul- tiplied by the original fractional area of the specimen, is taken as the tensile strength per square inch of the mate- rial tested. The lowest grade of wrought iron, entering into the €onstruction of the shells of steam boilers, should stand a tensile strain of at least 45,000 pounds per square inch. The best irons range from 60,000 to 75,000 pounds and occasionally higher. 68 A TREATISE ON STEAM BOILERS. For steel boiler plates the tensile strength is to he kept as low as possible and insure sound and homogeneous ingots or plates. In crucible steel it is difficult to get it below sixty thousand pounds and should not exceed seventy-five thousand. In Bessemer and Siemens-Martin steels it may vary from fifty-five thousand to seventy thousand pounds. TABLE XX. STRENGTH OF AMERICAN IRON BOILER PLATE. TESTS MADE AT THE U. S. TREASURY DEPARTMENT, WASHINGTON. THICKNESS TENSILE STRAIN REDUCED NO. IN IN POUNDS PER AREA PER HOW TESTED. INCHES. SQUARE INCH. CENT. 128 i 61,538 36 With the grain. 129 i 59,125 18 Across the grain. 136 k 58,373 38 With the grain. 135 i 53,333 9 Across the grain. 126 5 62,871 38 With the grain. 127 T^ 58,765 20 Across the grain. 134 5 TS 62,195 43 With the grain. 133 5 T6 60,202 10 Across the grain. 124 f 61,481 30 With the grain. 125 1 58,653 22 Across the grain. 132 3 8 60,408 47 With the grain. 131 3 8 57,377 15 Across the grain. 148 3 8 56,270 25 With the grain. 149 1 54,461 17 Across the grain. 146 ^ 61,918 33 With the grain. 147 i 63,469 6 Across the grain. Nos. 128, 129 were one-fourth inch iron reduced to the square of its thickness. Nos. 135, 136 were of the same iron and were nearly one inch wide. Nos. 126, 127 were of small, and 133, 134 were of larger area. Nos. 124, 125, 148, 149 were cut exactly the square of the thickness, and Nos. 131, 132 were of the same iron whose area approximated one-fourth of one square inch. Nos. 146 and 147 were samples of one-half inch, cut the square of its thickness. It may be of interest to compare the relative tensile strengths of American with English boiler plates. The ENGLISH BOILER PLATES. 69 figures in table XXI it will be observed, average consider- ably lower in the table immediately preceding. One thing in favor of the figures given of any tests made in England is, that test pieces are as a rule eight or ten inches long, while in this country they are usually the "short," though sometimes two inches long, and rarely six or eight inches, a|difl:erence which will be explained further along in this chapter. TABLE XXI. TENSILE STRENGTH AND DUCTILITY OF ENGLISH BOILER PLATE, AS DETERMINED BY MR. KIRKALDY'S EXPERIMENTS. DISTRICT IN WHICH THE IRON IS MADE. NAMES OF MAKERS OR WORKS AND BRANDS. DIRECTION OF THE GRAIN AND THICKNESS IN INCHES. TEARING WEIGHT PER SQ. INCH OF ORIGINAL SECTION IN LBS. CONTRAC- TION OF AREA FRAC- TURED. PERCENT ULTI- MATE ELONG- ATION OR TEN- SILE SET AF- TER FRAC- TURE. PER CT. Yorkshire Lo wni oor T 5 L. T6 c. ^ L. f c. f L. f c. 1 L. i c. i L. T6" L. f to 1^ C. f to ^ L. 1 C. 1 L. i to f c. i to f 5] ,990 50,512 56,000 46,211 52,237 • 46,435 .55,821 50,445 54,835 45,584 44,957 44,021 .51,2.51 46,704 53,402 41,776 19.7 12.1 17.8 13.2 15.3 6.9 17.2 9.0 12.5 4.6 8.7 69 13.1 10.2 10.6 3.7 13.2 Yorkshire •. Lowmoor 9.3 Yorkshire Farnley 14.1 Yorkshire Farnley 7.6 Yorkshire fiowling. 11.6 Yorkshire Bowling 5.9 Staffordshire Bradley % Crown S. C Bradley % Crown S. C Thorney croft, Best Best Thorney croft, Best Best Lloyds, Foster, %, Best.... Lloyds, Foster, %, Best.... Consett, Best Best. 12.5 Staffordshire 5.5 Staffordshire 11.2 Staffordshire 4.6 Staffordshire 5.3 Staffordshire 4.6 North of England... 8.9 North of England... Consett, Best Best 6.4 Scotland Glasgow, Best Best 9.0 Scotland Glasgow, Best Best 2.6 L. signifies lengthwise, or in the direction of the grain. C. signifies crosswise, or across the grain. 70 A TREATISE ON STEAM BOILERS. This subject has received a great deal of attention in England, and I quote from a paper by Mr. Marlett, chief examiner, Lloyd's Register, 1878, as follows: "Another point of our investigations which has received our most anxious attention, is as to the limits within which the tensile strength should be confined. " In the committee's circular, the limits are from twenty- six to thirty tons (58,240 to 67,200 lbs), and this agrees with the Admiralty requirements, but the weight of evidence we have been able to collect since the issue of that circular is in favor of somewhat higher limits. Mr. Sharp, of the Bolton Co., Mr. Webb, of Crewe, and Mr. Ellis, of Messrs. J. Brown & Co.'s works, urge that the upper limit might be raised to thirty-two tons (71,680 pounds) per square inch, without the slightest fear of obtaining brittle plates, so long as the temper and other tests are enforced. The Dutch government stipulate for a tensile strength of from twenty-seven to thirty-one tons (60,480 to 69,440 pounds) in their present contracts with the Bolton and Landore steel companies, and in contracts for boiler plates and other uses, the limits are fixed as high as thirty-three tons 73,920 pounds), although the steel has still to be mild and ductile. It is said by some that when steel gets down to about twenty-six tons (58,240 pounds) in tensile strength, it begins to be more spongy and is less capable of being welded than steel of twenty-eight tons (62,720 pounds) per square inch, and it is urged that steel between thirty and thirty-two tons (67,200 to 71,680 pounds) strength, if it fulfills all the other conditions of ductility, is abetter material than the weaker, and sometimes less ductile, material having a tensile strength of twenty-six tons (58,240 pounds). Indications seem to show that these lower limits are more easily reached by the Siemens steel than by the Bessemer and an advantage is claimed for the lat- ter at the higher limits. After giving the matter our most SIZE AND SHAPE OF SAMPLES. 71 carefal consideration, we are of the opinion that it would on the whole be preferable to fix the limits at twenty-seven to thirty-one tons (60,480 to 69,440 pounds) per square inch, rather than twenty-six to thirty tons (58,249 to 67,200 pounds)." The size and shape of samples for testing — It was but a few years since that the only experimental test to which boiler plate was subjected was to determine its tensile strength. It was then the custom to make test pieces short, or, rather, of no particular length, and with little or no uniformity of cross section, except at the point where rupture was to occur. The pieces were, however, usually of one of the outlines, as given in the accompanying sketches, in which figure 5 is designated as a long and fig- ure 6 as a short specimen : X. J- Figure 5. Figure 6. In testing a plate to determine its tensile strength merely, this is perhaps well enough; but as nearly all tests are now required to show both tensile strength and ductil- ity, it is recommended that test pieces be of the same length and sectional area, in order that results of different tests may be tabulated, and thus form an intelligent and ready means of comparison, preventing much needless con- 72 A TREATISE ON STEAM BOILERS. fusion, which must necessarily arise where specimens are of different lengths. Experiments made to determine how the different lengths of specimens tested affects the percentage of elong- ation, show most clearly that if a fixed percentage of elong- ation is required, the specimen should be of fixed length. If, on the other hand, latitude is permitted in the length of the specimens to be tested, so as to suit the different test- ing machines, there should be a sliding scale for percentage of elongation. The different percentages of elongation for different lengths of specimens were found by experiment to be as follows : * STRETCH 6f homogeneous STEEL PLATES, Eight inch specimen 20 percent. Six inch specimen 25 percent. Four inch specimen 32 percent. Two inch specimen 37^ percent. The reason for these differences of percentages in elong- ation is obvious, and arises from the fact that near the point of fracture the elongation is much greater than at other parts of the specimen. With material, therefore, of equal quality, the shorter the specimen tested the higher will be the percentage of elongation; or, on the other hand, comparatively hard and brittle steel might easily be made to show a required twenty per cent of elongation by making the specimen sufficiently short for the purpose. In testing samples of steel boiler plate it is of the utmost importance that the pull be exactly in the line of the sample, so that when fracture sets in, it shall be a break and not a tear. A material may be torn asunder at a pressure very much below^ that required to break it or pull it apart. Care should also be taken that no imperfections or "nicks" exist in the test portion of the sample. '•■ Committee Report, Lloyd's Begister. UNITED STATES GOVERNMENT TESTS. 73 The eight inch specimen was first adopted by the French Admiralty and afterward by the English Admir- alty, then by Lloyd's Register of English and Foreign Shipping, and thus a standard length forced itself upon the attention of manufacturers, so that it is now in very general use in Europe. It is much to be regretted that our own government still uses the short test specimens, particularly as little or no modifications would be required in the present testing machines in order to use the eight inch specimen. This would enable a direct comparison of American and foreign tests, which are always of great interest and value. TJ. S. Government Tests — The instructions to local inspectors of steam boilers, so far as relates to the ten- sile strength of boiler plate, and contained in the rules and regulations prescribed by the Board of Supervising Inspectors of steam vessels, are as follows: " Rule 3 — Every iron or steel plate intended for the con- struction 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 center of the plate, 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. "When a sheet of boiler iron is found by the inspector with one or more stamps upon the same, the inspectors shall in every such case be governed and rate the tensile 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 74 A TREATISE ON STEAM BOILERS. boilers, by the United States inspectors, shall be as follows, viz: " The inspector shall visit places where marine boilers are bein^ 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 of the plates, the inspector shall cause two pieces to be taken from each sheet to be tested, the area of one of which shall equal one-quarter of one square inch, the area of the other shall equal the square of its thickness, and the force at which these pieces can be parted in the direction of the liber or grain, represented in pounds avoirdupois — the former multiplied by four, the latter in proportion to the ratio of its area — that piece showing the greater tensile strain shall be held to be the tensile strength of the plate from which the test pieces were 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 marine boilers; provided, always, that the said plates possess the other qualities required by law, viz, homogeneousness, toughness and ability to withstand the eifect of repeated heating and cooling; but should these tests prove the marks on 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 noth- ing herein shall be so construed as to prevent the manu- facturers from restamping such iron at the lowest tensile strain indicated by the samples, provided such restamping is done previous to the use of the plates in the manufacture of marine boilers. " In the following table will be found the widths — expressed in hundredths of an inch — that will equal one- quarter of one square inch of section of the various thick- nesses of boiler plates. The signs + (plus) and — (minus) UNITED STATES GOVERNMENT TESTS. 75 indicate that the numbers against which these signs are placed are a trifle more or less, but will not in any instance exceed one-thousandth of an inch. " The gauge to be employed by inspectors'and others to determine the thickness of boiler plates and the widths in the table will be the Darling, Brown & Sharp's gauge, of Providence, Rhode Island, and will be furnished by the Treasury Department. This gauge has been approved by the Board of Supervising Inspectors : 3^^ = 133- 21 =119 — 23 = 109 + Y' = 100 .26 =96— .35 =71 .29 =86— f^ = 67 3^- = 80 ^^^ = 57 .33 =76+ y^ = 50 1 inch. v^ • • r\ 1 inch. Figure 7. "All samples intended to be tested ontheRiehle testing machine must be prepared in form, according to the above diagram, viz, eight inches in length, two inches in width, cut out at their centers in the manner indicated. Two small center punch marks must be made on samples, one inch each side of their center, for the purpose of ascertain- ing their elongation or ductility. "In commencing a test, the person conducting the same must first apply weights to within four thousand pounds of one-quarter of the tensile strength marked upon the sample, and, after pumping the machine to equilibrium, apply the remaining weights at intervals of about fifteen seconds, until the sample is parted. " The smaller w^eights must be applied last, and should a sample part immediately on the application of such a weight, the weight last applied must be rejected. 76 A TREATISE ON STEAM BOILERS. "The machine must be kept at equilibrium during the application of the weights, and, after the first application is made, the point where elongation commences must be ascertained by applying a pair of dividers to the center punch marks, at every additional weight, until the test is completed. ''All tests made of boiler material must be recorded upon a table showing the following : " Date when tests were made. " From whom samples were obtained and by whom tested. " Material, iron or steel. " Stamp or label on samples, which must be the same as stamps on the material from which they are taken. " Thickness of samples, expressed in hundredths of an inch. " Width of samples, expressed in hundredths of an inch. •' Strain at which each sample parted. "Strain per square inch of section. '^Elongation of samples, expressed in hundredths of an inch. " Time consumed in tests, expressed in minutes and sec- onds. " Weight on machine at which elongation commenced. Elongation — In order to get anything like satisfactory results from any experiments made to determine the per- centage of elongation that any given sample of either iron or steel plates is capable of yielding, the samples ought to be at least six inches long, or better still, eight inches — the standard employed by the English and French governments. The samples should be rough polished on one side and ruled with lines at any convenient distance, say one-fourth inch apart. The elongation may easily be measured before or after breaking and the flow of metal observed at differ- TESTING BOILER PLATES. 77 ent portions of the piece tested. The English Admiralty and Lloyd's Register both require that steel plates enter- ing into the construction of ships and boilers shall stand an elongation of twenty per cent in an eight inch speci- men. This would require an elongation of thirty-seven and a half per cent on a tAVO inch specimen, as deduced from the Lloyd's experiments See page 72. Iron boiler plate varies considerably, but will elongate from six to twenty per cent in samples of the same length as the above. Reduction of area — When wrought iron and steel sam- ples are broken to ascertain this tensile strength, the orig- inal area of cross section is always reduced and this reduction of area is a good index in determining the suitability of the material for boilers. When iron or steel plates are under high tension in a testing machine, the reduction of area will depend largely upon the inherent hardness or softness of the samples; thus, a soft fibrous iron will stretch and soon show a reduced area in which the fracture will occur. The amount of elongation and reduction of area will be found to be greater than if the iron had been of higher tensile strength. Hard specimens, either of iron or steel, stretch very little and in breaking do so suddenly with very little reduction of area. Soft steel elongates more than iron plates and will suffer a contraction amounting to from thirty to sixty per cent of original area. The writer was shown by Mr. Atkinson several sam- ples of "Sligo" iron, by Phillips, Nimick & Co., stamped fifty-seven thousand pounds and having an actual tensile strength of seventy-one thousand pounds, which showed after breaking an average contraction of area of thirty- five per cent. 78 A TREATISE ON STEAM BOILERS. Tenacity and ductility are so closely associated that separation is almost impossible, and the tendency of experts now is to require irons of a certain tensile strength to suffer a certain reduction of area before breaking, or when the elastic limit is reached. At a meeting of boiler plate manufacturers, held at Philadelphia, November, 1878, after ^. very intelligent discussion of this subject, it was '^Resolved, That in the judgment of this meeting, plates should not' be used in a steamboat boiler that showed a contraction of area less than twelve per cent. We therefore recommend that all boiler plate, stamped with a tensile strain of under forty-five thousand pounds, should show contraction area of twelve per cent; forty-five thousand and under fifty thousand, should show fifteen percent ; fifty thousand and under fifty-five thousand, should show twenty-five per cent; fifty-five thousand and over should show 35 per cent." In some tests made by Mr. Kirkaldy, in 1876, upon Essen and Yorkshire plates, one hundred and twenty-eight specimens were tested, with results as follows — specimens ten inches long in central portion, by two inches in width : TABLE XXII. ESSEN. YORKSHIRE. Elastic stress 25,144 pounds. 48,028 pounds. 74,542 pounds. 83.8 per cent. 1.94 per cent. 7.76 per cent. 22.70 per cent. 27,477 pounds. 45,515 pounds. 56,875 pounds. 18.6 per cent. 0.85 per cent. 6.41 per cent. 14.80 per cent. Ultimate stress Stress for fractured area Contraction of area Extension at 30,000 pounds Extension at 40 000 pounds Extension ultimate TESTING BOILER PLATES. • 79 Elasticity is that property which all bodies have in a greater or less degree and by which they retain their form when acted upon by any force which tends to distort their original figure. Elasticity is said to be perfect when a body acted upon by a force which distorts it will immediately and com- pletely recover its original form, when the force is removed. Elasticity is said to be imperfect when such a force per- manently alters or changes the shape of the figure either wholly or in part. The difiterent kinds of elasticity are known by names corresponding to the dififerent kinds of strains to which bodies can be subjected and are known under the several names — tension, compression, flexure and torsion. The elastic limit of any material represents the load which it is capable of receiving before it becomes perma- nently fixed or set, and from which it will not recover when the load is removed. Thus, there are limits to ten- sion, compression, flexure and torsion, beyond which the addition of a further application of weight or force will sooner or later lead to rupture. Reference is made, not to sudden changes in stress and shocks, but to gradually increasing strains. This defini- tion is theoretically worthless, for a limit so definite is not probable and much less is it proven.* Such investigators as Hodgkinson and Clark have observed that there are permanent changes of form under very small loads. At present we must be content with defining this limit with Fairbairn, as that stress below which the. changes in form are approximately proportioned to the forces, while above this they increase much more rapidly. All experiments, up to the present time, have shown that when the elastic limit is passed, the tensile resistance is considerably increased, while ductility and tenacity dimin- *Weyrauch. 80 A TREATISE ON STEAM BOILERS. ish ; the metal becoming brittle and having little power of resistance to shock. In experiments at the Woolwich Arsenal, an iron rod, four times raptured by pull, gave the successive values of 3,520, 3,803, 3,978, 4,186. " It is found by experiment that, up to the limit of elas- ticity, the displacements suffered by the molecules of the body are sensibly proportional to the stress which causes them, so that a double displacement is caused by a double straining force; a triple displacement by a triple straining force; and so on."* The elastic limit is usually determined by weighing the force required to produce a perceptible and permanent change of form in the sample tested; this weight, divided by the area of the sample, gives the approximate elastic limit. The elastic limit of wrought iron is generally taken at one-half its tensile strength. Experiments made at Wash- ington on bars from five-eighths to two inches diameter show that for the particular grade of iron required for chain cables and for ship-building purposes generally, the elastic limit does not vary much from fifty-seven per cent of its tensile strength. Tests made of rivet steel from the Edgar Thompson steel works show, on three-quarter inch bars turned down to one-half inch^diameter, an elastic limit'of 41,000 pounds, the sample having 64,000 pounds tensile strength, with twenty-nine per cent elongation in a three inch specimen and fifty per cent reduction of area at point of fracture. The carbon in this steel was 0.11 per cent. For steel boiler plates from the same company having an area of .2282 (.62 X .36), the elastic limit was 37,634 pounds, the tensile strength being 58,690 pounds. ^Anderson. Strength of Materials, page 4. TESTING BOILER PLATES. 81 Percussion tests are seldom resorted to, for the reason that very few irons will stand such a test. It is sometimes employed in testing plates for ship building, and in all such tests the superiority of steel over iron plates is clearly shown. Percussion tests have been made by allow- ing a ball weighing nearly thirty-four hundred pounds to fall on the unsupported middle of steel and iron boiler plates from distances varying from five feet six inches to twelve feet high. The first blow from a height of five feet six inches cracked the iron plate, and these cracks were much extended when the plate was turned up and struck on the other side from a height of eight feet. With steel boiler plates the first blow was from ^ve feet six inches high; this produced no flaw. The plate was then turned over ^nd struck from a height of eight feet six inches; it was then turned over again and struck from a height of ten feet; and was again turned over and struck from a height of twelve feet, and still no crack or flaw found in it. • « Bulging tests — These are seldom made, and so far as the writer is aware, are never required in any specifications for boiler plate. Experiments were made by Mr. Kirkaldy in 1875 to ascertain the resistance of plates to and the effects under bulging stress, and are tabulated in his report on Essen and Yorkshire wrought iron plates. Fifty-four speci- mens, each twelve inches diameter, were pressed into an aperture ten inches in diameter, the "bulger" being five inches in diameter and having a rounded end turned to a radius of five inches. The stress was gradually increased until the specimen was pushed through the aperture or until the specimen gave way either by cracking or burst- ing. These experiments were made on plates having a nominal thickness of three-eighths, one-half and ^ve~ (7) 82 A TREATISE ON STEAM BOILERS. eighths of an inch. The three-eighths inch plates stood the test better than the latter, and are given in the follow- ing table taken from the report: TABLE XXTII. RESULTS OF EXPERIMENTS BY MR. KIRKALDY TO ASCERTAIN THE RESISTANCE TO BULGING STRESS OF WROUGHT IRON PLATES. NOMINAL THICKNESS THREE-EIGHTHS INCH. PLATES UNANNEALED. BKAND. Kriipp Krupp Krupp Mea Farnley Lowmoor .... Bowling Monkbridge Taylor's Cooper & Co Mea^ INCH. .44 ,44 .44 .44 .42 .38 .40 .37 .39 .38 .39 STRESS IN POUNDS BULGED, INCHES. 0.81 0.82 0.82 0,82 0.77 0.92 0.74 0.86 0.80 0.85 0.83 1.34 1.35 1.36 1.35 1.39 1.54 1.35 1.47 1.42 1.47 1.44 1.75 1.79 1.80 1.78 1.85 2.06 1.78 1.97 2.12 2.15 2.16 2.14 2.32 2.71 2.46 2.51 2.58 2.64 2.67 2.63 ULTIMATE. O t^ K INCHBS. 3.28 3.28 3.26 3.27 3.24 3.20 3.22 2.75 1.84 1.65 2.65 POUNDS. 139,940 139,780 137,560 139,093 116,810 102,780 114,4i0 110,880 54,720 51,220 '91,805 Uncracked. Uncracked. Uncracked. Uncracked. Uncracked. Cracked. Burst. Burst. Burst. The English A^dmiralty tests for irons entering into the construction of steam boilers are as follows : TESTING BOILER PLATES. 83 TABLE XXIV. TENSILE STRENGTH REQUIRED OF WROUGHT IRON SUPPLIED THE ENG- LISH GOVERNMENT. HOW TKSTED. TENSILE STRAIN. CLASS OF IRON. IN TONS (2,240 POUNDS) PER SQUARE INCH. IN POUNDS PER SQUARE INCH. BB or 1st class plate ii-on and sheet iron, 34 inch thick and above J With the grain 1 Against the grain... j With the grain 1 Against the grain... (with the grain 1 Against the grain... - With the grain X AVith the grain 22 18 21 18 20 17 22 22 49,280 40,320 BB or 1st class boiler plate iron ^ inch thick and above 47,040 40,320 B or 2d class plat^ or sheet iron.... Angle, Bulb, T, , , or other iron of ordinarv form 44,800 38,080 49,280 Best merchant iron, BB bar iron, y^, round, segmental. Fire bar iron 49,280 Forge tests are made by bending tbe samples of iron over a corner of a cast iron slab, of which the edge is slightly rounded. The plates to be tested may be either hot or cold, and are tested both with and across the grain. The test consists in determining the angle through which the plate will bend without showing signs of fracture. This, of course, depends upon both the quality of the iron and the thickness of the plate. The next table contains the requirements of the Eng- lish government in forge tests, and will be found to be well adapted for testing American, irons. The best Amer- ican irons will stand a severer test than that required of the BB 21 ton T. S. English iron. 84 A TREATISE ON STEAM BOILERS. TABLE XXV. SHOWING THE FORGE TESTS, BOTH HOT AND COLD, REQUIRED BY THE ENGLISH GOVERNMENT FOR PLATE AND SHEET IRONS. POSITION OF THE TxRAlN IN THE TEST. PLATE IRON. SHEET IRON KIND OF IRON — - COLD. HOT. COLD TESTED. THICKNESS. ALL THICK- NESSES UP TO 1 INCH. HOT, I IN. i IN. 1 IN. 1 IN. 15° 5° 10° T. S. Best Best, 21 tons. Best Best, 18 tons. Best, 20 tons Best, 17 tons Lengthwise.. Crosswise Lengthwise.. Crosswise 70° 30° 55° 20° 35° 15° 30° 10° 25° 10° 20° 5° 125° 90° 90° 60° 90° 40° 75° 30° 125^* 90^ 90° 60^ The angles given in the above table is that through v^hich the plate is bent, commencing at the horizontal, and is not the angle between the sides of the plate after it i^ bent. Mr. Kirkaldy^s experiments — The investigations of Mr, Kirkaldy, founded upon an elaborate series of experiments made by him on iron of every description and quality, led him to the following conclusions, among many others : " I. The breaking strain does not indicate the quality, as hitherto assumed. *'2. A high breaking strain may be due to the iron being of superior quality, dense, fine and moderately soft, or simply to its being very hard and unyielding. "3. Alow breaking strain maybe due to looseness and coarseness in the texture or to extreme softness, though very close and fine in quality. MR. kirkaldy's experiments 85 "4. The contraction of area at fracture, previously overlooked, forms an essential element in estimating the quality of specimens. "5. The respective merits of various specimens can be correctly ascertained by comparing the breaking strain jointly with the contraction of area. " 6. Inferior qualities show a much greater variation in the breaking strain than superior. " 7. Greater differences exist between small and large bars in coarse than in fine varieties. •"8. The prevailing opinion of a rough bar being stronger than a turned one is erroneous. "9. Rolled bars are slightly hardened by being forged down. ''10. The breaking strain and contraction of area of iron plates are greater in the direction in which they are rolled than in a transverse direction. (The experiments show the difference to be about ten per cent)." CHAPTER VI RIVETED JOINTS. Effects in Punching Plates — Experiments on Drilled and Punched Holes — Experiments on Ordinary and Spiral Punching — Strength of Riveted Joints — Single Riveted, Hand, Steam and Hydraulic Riveting — Double Riveted Lap Joints — Single and Double Riveted Butt Joints — Experiments on Thick Steel Plates by Punching and Drilling — Loss Due to Punching — Experiments on Chain and Zig- Zag Riveting — Testing Rivets — Testing Stay Bolts — Shearing Tests of Rivet Iron and Steel — Steel Rivets — Proportions for Single Riveted Lap Joints — Double Riveting — Calking. The only practical method of joining plates in the con- struction of boilers is by riveting. This is at best a very expensive and unsatisfactory way of making a joint, and the difficulties begin at the very outset by the loss of strength occasioned in punching the plates, and occurs by reason of, 1. A reduction of area through tlie line of rivet holes, and 2. By the disturbing influence of the punch on the remaining metal, still farther reducing its tensile strength. The bad effects of punching are, in general, more appar- ent in steel than in iron plates. It has been observed that when ordinary mild steel plates, having a tensile strength of upwards of seventy thousand pounds, have been tested after punching and before annealing, there is a loss of strength variously estimated from five to forty per cent of the original plate, depending somewhat on the hardness and the thickness of the plate. DRILLED AND PUNCHED HOLES, 87 The observed changes in the material in the line of punched holes are, increased hardness, alteration of struc- ture and loss of ductility. From, specimens tested, which had been cut from dif- ferent portions of the same plate and in the same line of punched holes, there does not appear to be a uniform dis- tribution of strain over the entire surface of the plate, but the disturbance of material is confined to within a very- short distance around the hole, extending from one to three-sixteenths of an inch. This inference is drawn from the fact that by drilling and reaming out punched holes and then testing the plate, making proper allowance for the reduced area, no perceptible decrease of strength is noted. TABLE XXVI. SHOWING RESULTS OF EXPERIMENTS MADE TO ASCERTAIN THE EFFECTS PRODUCED BY DRILLED HOLES AND BY PUNCHED HOLES UNDER PULLING STRESS, KRUPP'S WROUGHT- IRON. TESTS BY MR. KIRKALDY. Size of specimen — holes not de- ducted Size of specimen — gross area, square inches Ultimate stress per square inch... Ultimate stress — total Difference, or loss per square inch Difference or loss — per cent Elongation of holes — fractured... Elongation of holes — unfrac- tured Elongation of holes — total — inch. LENGTHWAY. CKOSSWAY. DRILLED. PUNCHED. DRILLED. PUNCHED. 8 X .44 8" X .44" 8" X .44" 8 X .44" 3.52 3.52 3.52 3.52 33,005 lbs. 28,006 lbs. 30,053 lbs. 24,329 lbs. 116,180 lbs. 98,580 lbs. 105,790 lbs. 85,640 Iba. 19,590 lbs. 26,534 lbs. 20,142 lbs. 26,101 lbs. 37.2 4S.6 40.1 51.7 .34 inch. .17 inch. .27 inch. .15 inch. .18 inch. .05 inch. .13 inch. .03 inch. .52 .22 .40 .18 88 A TREATISE ON STEAM BOILERS. TABLE XXVI— Continued. LENGTHWAY. CROSSWAY. DRILLED. PUNCHED. DRILLED. PUNCHED. Elongation of holes — total — per cent 30.6 Fibrous. 52,595 K)s. 13.0 Fibrous. 54,540 lbs. •23.5 Fibrous. 50,195 lbs. 10.6 Appearance of fracture Fibrous. Solid plate, ultimate sti-ess per square inch 50,430 lbs. \ y • • • • • • • • r \ The drilled holes were made exactly the same size as those punched : Diameter 0.85 inch X 4 holes = 3.40 inches, or 42.5 per cent of the width of the specimen. All the speci- mens were uuannealed. The engraving, fig- ure 8, represents thie shape of the specimens tested, being 8 inches Figure 8. in width of Central por- tion, with the rows of rivet holes two and a half inches apart between their centers; the pitch of the four holes across the plate being two inches, the one row being to exhibit the elongation of the holes after the plate was pulled asunder, the other to show the shape of the holes without being fractured. The punched holes were conical, as usual, being larger on the exit than on the entrance side of the plate. Those drilled were all made exactly to the smaller size, and thus suitable for the same sized rivet. It will be observed that in the first line of the table the space occupied by the rivet holes is not deducted as cus- tomary in making calculations on riveted joints, and that the gross and not the net area is stated. Mr. Kirkaldy's reason for doing so is, that it is better to give the total stress borne by the specimens of gross sectional area in pounds per square inch instead of the reduced area, so that any one can divide it by the net area, instead of the gross, area, should they prefer to do so. STRENGTH OF PUNCHED PLATES. 89 The strength of the solid plate, or that without the holes, was taken from other tests of the sanne material, and is given in the last line of the table, in order to facilitate <3omparisons. The difference in strength between that of the solid plate and that with the holes represents the loss due to the latter. As already shown in the foot note to the table, 42.5 per cent of the plate was removed in forming the four holes. The actual loss appears as follows: In the plate with drilled holes 37.2 per cent loss wheu tested lengthway of the plate, and 40.1 per cent when tested ^rossway; or a mean loss of 38.65 per cent for the two directions. In the plates with punched holes the loss, when tested lengthway of the plate, was 48.6 per cent and 51.7 per cent when tested crossway of the plate, showing a mean, loss of 50.15 per cent for the tw^o directions. To summarize we have, then. Loss due to punching, mean 50.15 Loss due to drilling, mean 40.01 Showing a mean loss of 10.14 per cent, due to punching over drilling. The ultimate stress borne by a specimen is greatly affected by the hardness or softness of the material and by the shape of the specimen. The softer the material the more rapidly does its sectional area become reduced by the specimen stretching and consequently in the amount of stress sustained. When the breadth of a specimen is reduced to a minimum atone point, a greater resistance is offered to its stretching than when formed parallel for some distance; and as the stretching is checked so will also the contraction of area, and with it will be an increase in the ultimate stress.* In all punched holes in boiler plates which the writer has measured, there Has been the same conical taper resulting * Kirkaldv. 90 A TREATISE ON STEAM BOILERS. from the use of a die larger than the punch. This is the common method of fitting punches and dies for boiler work, originating, doubtless, in a necessity for a larger die because of a lateral motion of the punch, due to the imper- fect fitting of the slide to which the punch is secured. Then, afterwards, as punching machines were better built and had none of that lateral motion, the same practice of fitting punch and die continued, under the belief that it was necessary to good punching. The fact that some of the best examples of punching now on record was done in a machine in which the punch and die accurately fitted each other, shows that this matter of enlargement of the die may easily be overdone. The ordinary clearance for five-eighths and three- quarters inch dies is nearly ^-^ of an inch; the punch being made on size and the clearance allowed in the die. Cold punched nuts, as for example, those made by Hoopes & Townsend, Philadelphia, when taken as exam- ples of "commercial" punching as distinguished from experimental merely, are of considerable interest in this connection, owing to the entire absence of the conical holes spoken of in the preceding paragraph. It has already been shown in the table collated from Mr. Kirkal- dy's experiments, that there is a loss in punching iron plates over drilling, approximating ten per cent, Hoopes & Townsend have long been of the opinion that if properly performed, punching does not weaken good iron farther than by the simple reduction of area. In order to deter- mine the truth or falsity of this opinion they prepared ^= \ ^ ^ a number of test ^^- ^""^ J ^ ^i: : ^- ^ ^^^^^F^M T that represented in figure 9. These were Figure 9. made of bar iron IJ X f inch, and one of each pair had a hole f|- inch in STRENGTH OF PUNCHED IRON. 91 diameter drilled, and the other specimen the same sized hole punched in it. The specimens were then planed down next the hole, as represented in the engraving, so as to leave a thickness of three-eighths inch on each side of the hole. The other pairs had one-quarter, three-six- teenths and one-eighth inch respectively. These speci- mens were then broken by subjecting them to a tensile strain in one of Richie Brothers' testing machines, with the following results :* TABLE XVII. , STRENGTH OF PUNCHED AND DRILLED IRON BARS. HOOPES & TOWNSEND. THICKNESS OF BAR. f inch. f inch. I inch. f inch. I inch. f inch. f inch. 4 inch. THICKNESS OUT- SIDE OF HOLE. _3_ 16 inch, inch, inch, inch, inch, inch, inch, inch. PUNCHED BAR BROKE AT 31,710 31,380 18,820 18,750 14.590 15,420 10,670 11,730 pounds, pounds, pounds, pounds, pounds, pounds, pounds, pounds. DRILLED BAR BROKE AT 28,000 26,950 18,000 17,590 13,230 13,750 9,320 9,580 pounds, pounds, pounds. pounds, pounds, pounds, pounds, pounds. From the engraving it will be seen that it was the por- tion of the iron immediately next to the hole, and which is usually supposed to be most affected by the action of punch or drill, which had to resist the strain. It will be seen that, in any case, the punched bars had the greatest strength, indicating that the punching had the effect of strengthening instead of weakening the iron. These experiments have given results just the reverse of similar experiments made on specimens of boiler plates; but * These tests were undertaken at the suggestion of and were first pnblished in the Railroad Gazette. 92 A TREATISE ON STEAM BOILERS. Messrs. Iloopes & Townsend argue that it is due, first, to the kind of material used, which is a tough and ductile iron, and second, to the method of punching. If a brittle and granular iron was used, the effect of the punching would be to crumble or disintegrate the iron in the imme- diate vicinity of the action of the punch; or if the punches and dies employed were so proportioned as to have a ten- dency to split open the bar, the metal around the hole would also be strained injuriously. But in manufacturing nuts they use a punch which fits accurately into the die, and the machines employed are heavy enough and made to work with sufficient accuracy so that the iron being punched is subjected to direct vertical pressure alone, with- out exerting any lateral or bursting strains in the iron. The efiect is, that the metal is compressed and thus made more dense and stronger. That some such action takes place seems probable from the appearance of the holes in the nut, which are straight and almost as smooth as though they were drilled. Kennedy^ s patent spiral shearing punch — A eleven-sixteenths inch punch, the size used in ordinary five-eighths inch riveting, is shown full size in figure 10. This punch derives its name from the fact that in its operation it performs its work in a circle, in the same manner that a shear does — in a straight line. Thus, to shear a hole two inches in diameter in a given plate of iron, is about the same as to shear ofi* a bar of iron of the same thickness, a little more than six inches in width. It is well known that to cut off a given plate of metal with the blades of the outter parallel, requires an amount of power and conse- quent strain upon the machine far beyond what it would Figure 10. SPIRAL AND FLAT PUNCHING. 9^ if the blades were only a few degrees angular to each other. This is just the difference between the flat and the "spiral shearing punch." It would seem a matter of some surprise, then, that with this knowledge, and in view of the enormous and growing extent to which iron and steel are used, that so little change, to say nothing of improvement, has been made in the punching of holes. Some few attempts have been made in that direction, which are too familiar to need special notice ; still, nothing has remained but the hard and costly and damaging method of the common flat punch. It is hard and expensive, because it not only requires a punching machine to be at least one-third heavier and stronger to meet the strain, but also requires at least fifty per cent more power to do the same work that can be done with the spiral punch. But the economy in power, the cost, and strain, and wear of machinery, which was the flrst object in the mind of the inventor, proves to be but a small part of the real value of the invention. So serious and well known is the injury and weakening of the surrounding parts after punching thick steel plates and sometimes iron plates, has led to a prejudice against punching at all where the strength of the material is of importance, and, therefore, resort has generally been had to the tedious and costly process of drilling. Experiments were made in steel plates cut to sample, as shown in figure 11, suitable for testing in a machine, , v^-i ^ ^ 0.43 inch fhicK. with results as follows : figure u. Two holes were punched in each specimen, as seen in the cut, one with a flat punch and the other with a spiral punch. When tested, all the specimens broke through the hole punched with the flat punch. ^ ! i J f i — ^ .7S| r A 94 A TREATISE ON STEAM BOILERS. In tests made to determine the relative amount of power required to operate the two kinds of punches, it was found that a seven-eighth inch "spiral punch" penetrated a five-eighth inch plate, at a pressure of twenty-two to twenty-five tons, while a seven-eighth "flat punch" in the same plate required thirty-three to thirty-five tons, thus showing a dead loss of ten tons of pressure on each hole, beside the additional strain and wear of machinery. The following table supplies some interesting data in regard to punched plates, as well as a comparison of the two punches: TABLE XXVril. RESULTS OF EXPERIMENTS MADE AT CREWE ON THE TENSILE STRENGTH OF SAMPLES OF THE SAME PLATE PUNCHED WITH KENNEDY'S SPIRAL AND ORDINARY PUNCHES RESPECTIVELY, BY MR. F. W. WEBB. BREAKING WEIGHT OF PLATE. ELONGATION. AREA OF PLATE UNDER TENSION. DIAME- TER OF HOLE. ACTUAL. PER SQUARE INCH. ON TWO INCHES OF LENGTH PER CENT. REMARKS. ACROSS HOLES. POUNDS. POUNDS. .885 45,350 63,752 .11 5.5 .7114 .885 45,000 60,318 .23 11.5 .7461 .895 42,400 57,495 .]4 7.0 .7375 1 .89 .89 37,050 42,800 51,287 60,692 .03 .06 1.5 3.0 .7224 ,7052 Punched with "ordinary" punch. .90 45,150 61,047 .07 3.5 .7396 1 .895 39,400 55,465 .09 4.5 .7032 Mean 42,393 58,579 .104 5.2 .7236 STRENGTH OF RIVETED JOINT?. 95 TABLE XXVIII— CoNTiKUEij. BIAME- TER OF HOLE BREAKING WEIGHT OF PLATE. ACTUAL. POUNDS .885 45,850 .88 48,000 .88 46,200 .88 44,250 .88 45,500 .895 47,600 .885 45,600 Mean 46,143 .885 40,350 .89 41,800 .895 44,350 .885 45,400 .885 42,100 .89 45,450 .89 34,300 Mean 41,9r)4 ELONGATION. PER SQUARE INCH. POUNDS. 63,285 67,672 63,584 61,254 64,148 66,084 61.476 63,929 55,693 59,274 63,073 62,664 58,109 62,915 47,480 58,458 ON TWO INCHES OF PER LENGTH CENT. ACROSS HOLES. .27 13.5 .25 12 5 23 11.5 .12 6.0 .26 13.0 .27 13.5 .09 4.5 .21 10.6 .21 10 5 .08 4.0 .24 12.0 .24 12.0 .24 12.0 .23 11.5 .07 3.5 .19 9.3 AREA OF PLATE UNDER TENSION. .7245 .7093 ,7266 .7224 ,7093 .7203 .7418 .7220 .7245 .7052 .7032 .7245 .7245 .7224 .7224 .7181 REMARKS. Punched with "Kennedy's spiral" punch. Punched with both punches. Fracture oc- curred in ev- ery case thro' the "ordina- ry" punch hole. Strength of riveted joints — The first reliable data on the strength of riveted joints was given by Sir William Fair- bairn in 1838, which was deduced from tests made with single and double riveted joints in plates of w^rought iron one-quarter inch thick. The relative values given by him were. 96 A TREATISE ON STEAM BOILERS. Tensile strength of the solid plate 100 Tensile strength double riveted lap joint 70 Tensile strength single riveted lap joint 56 Since that time the percentages as given above have been in almost constant use by engineers and are still gen- erally accepted. As differences in material, kind and number of rivets, as well as varieties of arrangement and spacing of rivet holes became more common, tests were also made from time to time with more or less varying results, some of which are presented in this chapter. Mr. W. Bertram's experiments, as given by Mr. D. K. Clark in his Manual of Rules and Data, shows that for three thicknesses of specimens tested, viz, three-eighths, seven-sixteenths and one-half inch wrought iron plates, having a tensile strength of twenty tons (44,800 pounds) per square inch, that the averages of all the lap joints show that the three-eighths inch joint is the strongest, that the seven-sixteenth inch is nearly as strong, and that they are about one-quarter stronger than one-half inch lap joints, relatively to the thickness of the plate, thus: TABLE XXIX. COMPARATIVE STRENGTH OF RIVETED JOINTS. THICKNESS OF PLATKS. Strength of plate, per cent Strength of single riveted joint by hand.. Strength of double riveted joint by hand.. 1 INCH. ^« INCH. 100 100 40 50 59 70 4 INCH. 100 60 72 The test specimens were each four inches wide and contained two rivets each, three-quarters of an inch diam- eter, placed two inches apart, center to center. Three tests of each were made and averages taken for the numerical value of percentages as given. RIVETED JOINTS. 97 The figures in the above table show that for single or double riveted joints, thin plates are to be preferred to thick ones. Experiments by David Greig and Max Eyth^ Leeds ^ England — In some tests of riveted joints made by these gentlemen (1879), in which each construction of joint was represented by four specimens of exactly the same dimensions, two being of steel and two of iron, one specimen of each material was of Brown's and one of CammelFs make. This distinction was made, not for the purpose of testing two different materials, but to get a fair average result. Tests of these two materials, made on samples two inches wide by three-eighths inch thick, gave an average breaking strain per square inch of solid plate as follows : TOXS, POUNDS. Cammell's iron 21.9 49,056 Cammell's steel 24.0 53,760 Brown's iron 22.6 50,624 Brown's steel 27.6 61,824 This gives. Average of iron 22.25 49,840 Average of steel 25.80 57,792 The iron specimens were invariably riveted together with iron rivets, the steel with steel rivets. The thickness of all plates was nominally three-eighths inch ; the rivets, except in four cases, were nominally five-eighths inch, the holes being drilled eleven-sixteenths inch in diameter. The slight difference in the thickness of the plates was reduced by calculation in working out the experiments to a uniform thickness of three-eighths inch. Test pieces were prepared, as shown in table XXX, to determine the relative value of punching and drilling. All the specimens were alike in their dimensions, two (8) 98 A TREATISE ON STEAM BOILERS. pieces, six and a half inches wide, forming a single riveted lap joint, with four, five-eighths inch rivets, one and five- eighths inch pitch. The punched and drilled holes were of the same diameter, viz, ^ inch. The die for the punch was 1^ inch diameter, or the usual -^ inch clearance. The conical hole produced hy punching measured at the top 0.708 inch and at the bottom 0.790 inch, there being no measurable difference between iron and steel in this respect. Four of these specimens were of iron, two being drilled and two. punched, all steam riveted. In the punched specimens the conical holes were placed with their smaller ends in contact. Of the two drilled speci- mens, one broke through the plate, the other sheared in rivets. The average strength of the same proved to be 50.4 per cent of the strength of the solid plate. TABLE XXX. STRENGTH OF LAP JOINTS, SINGLE RIVETED. v_ J ; • ; • ; • : • /""""" P'IGURE ^ 12. PLATE SIX AND ONE-HALF INCHES WIDE, THREB-EIGHTHS INCHES THICK, FOUR FIVE-EIGHTHS RIVETS, ONE AND FIVE-EIGHTHS INCH PITCH. Greig and- Myth. DESCRIPTION OF SPECIMEN. AVERAGE BREAKING STRAIN OF SPECIMEN IN POUNDS. BREAKING STRAIN OF SOLID PLATE PER INCH OF AVIDTH, IN POUNDS. BREAKING STRAIN OF SPECIMEN PER INCH OF WIDTH, IN POUNDS. STRENGTH OF SEAM PER CENT OF SOLID PLATE. > . « Q 05 H 03 SHEARING RESISTANCE OF RIVET IRON PER RIVET, IN POUNDS. STRENGTH OF SPECI- MEN, PER CENT OF NORMAL STRENGTH OF MATERIAL. Iron plate, drilled holes Iron plate, punched holes.. 61,350 49,400 18,700 18,700 9,438 7,600 50.4 40.6 16,325 15,810 PLATES 84.1 75.5 KIVBTS. 103 RIVETED JOINTS. 99 TABLE XXX— Continued. DESCRIPTION OF SPECIMEN. AVERAGE BREAKING STRAIN OP SPECIMEN IN POUNDS. BREAKING STRAIN OF SOLID PLATE PER INCH OF WIDTH, IN POUNDS. Eh W O S C ^^ "^ 5 Z P sag < u -> aw^ p; 03 E^ w o 12,300 12,950 13,060 STRENGTH OF SEAM PER CENT OF SOLID PLATE. a > . aP < '^ « SHEARING RESISTANCE OF RIVET IRON PER RIVET, IN POUNDS. STRENGTH OP SPECI- MEN, PER CENT OF NORMAL STRENGTH OF MATERIAL. Steel plate, drilled holes 8teel plate, punched holes and annealed • 79,800 , 84,700 84,900 21,700 21,700 21,700 50.6 59.6 60.2 19,950 21,175 21,225 18,440 18^440 18,440 PLATES RIVETS. 108 115 ^Steel plate, punched holes and unannealed 115 Second Series of Tests for Different Modes of Riveting. DESCRIPTION OF SPECIMEN. Iron, hand riveted lrf)n, steam riveted Iron, hydraulic riveted. Steel, hand riveted Steel, steam riveted Steel, hydraulic riveted IN OF PER H, IN IN OF INCH )UNDS % a > . TANCE PER UDS. BKEA SPEC UNDS X«CW H H ►- Q STRA PER IN PC o ^ a o o P ^ESIS IRON PON fa o z'^feo gaa* £ f- 5 ^ 2 S^ ^2 S '- < 2 2 Pi S - 5 o p^ S 1 H w ^ 2 (J a 2z ^>^H- M IH H p^ p:; a a <: ^ o u w £d 5 £ a; » < r > > « S to 2 H a H a a M 56,550 61,850 n s-§ ^ ago a: w K 18,700 21,700 18,700 21,700 PM U4 "> . a M !^ ^ Q H t, izi 2o5 ss2 o w " W S W ■< t. !=^ M « M »*'(-. r^ 1 tt M a -^ "^ o S a ® Oh B^ . "» !^ 9 j ..^ a w 3 So H^ O - CO pj i Hi a O « w 15,810 18,440 S a o a PLA TES RIVKT8 113 104 114 99.4 102. Iron plates have iron rivets ; steel plates steel rivets. These tests show a reduction in strength of seam by increasing the number of rivets in a seven and a half inch plate, from six to eight, and also their diameters from five- eighths to three-quarters inch. The plate containing the six rivets, five-eighths inch diameter, had two and a half inch centers of rivets lengthwise of the seam, the second row of rivets being in a line one inch distant, making the diagonal centers of rivets one and five-eighths inches. 102 A TREATISE ON STEAM BOILERS. One of the iron specimens broke through the plates, the fracture following a zigzag line through the rivet holes ; the other sheared the rivets. The strength of the seam was 64.6 per cent the full strength of the plate. In the first case (zigzag fracture), the section of the plate along the line of fracture broke with 113 per cent of its normal tearing strain. The steel samples broke through the rivets, the latter showing a shearing resistance only two per cent above the normal. This reduction of strength, as compared with former test pieces, is caused by tlie absence of bending of the joint, owing to the double row of rivets keeping the plates more rigidly in line. The strength of the seam was the highest obtained, being 70 per cent of the solid plate. In the second series of tests in table XXXI, the test specimens were of the same dimensions as the first, viz, seven and a half inches wide, three-eighths inch thick, the number of rivets being increased to eight and their diam- eters to three-quarters of an inch. The pitch of these rivets was one and seven-eighths inch lengthwise of the seam, and one and seven-eighths inch centers of rivets to the second row, or equidistant in any direction. Although the joint is very rigid, it is rather weak in the plate against direct tensile strain and was sure to break in a straight line across the rivets. Its strength proved to be 59.2 of the solid plate for iron and 62.9 per cent for the steel specimens, showing again the great advantage gained by the effect of a double row of rivets in preventing the bending of the joints under stress. RIVETED JOINTS. 103 TABLE XXXII. STRENGTH OF BUTT JOINTS, SINGLE AND DOUBLE RIVETED, SINGLE AND TWO COVERS, IRON AND STEEL PLATES, THREE-EIGHTHS INCH THICK, RIVETS FIVE-EIGHTHS OF AN INCH IN DIAMETER. "V r Figure 14. X ■\. DKSCRIPTION OF SPECIAIEN. Iron plate, GJ^ inches wide, single cover, single riv- eted, four % inch riv- ets in each plate. Fig- ure 14.. Steel plate, 63^ inches wide, single cover, single riv- eted, four % inch riv- ets in each plate. Fig- ure 14 Iron plate, 7)^ inches wide, single cover, double riveted, six % rivets in each plate. Figure 15... Steel plate, IY2, inches wide, single cover, double riveted, six y^ rivets in each plate. Figure 15... ^r j^ __ A. Figure 15. Greig and Eyth. o 2; ^ 'A w e . < U tn a a Q ■< !5 !zi 03 5 '-I 58,000 78,800 78,000 110,750 O 03 S !^ ? - « P. a 05 W S w H r; y Q M &. >-j i-i a 05 03 r ;1 fe ^ 18,440 18,440 Si S a H Oh u. ^ 03 H •a ^ i^ a &j an 05 0: H a W oi « H H a S CS Ph ■< »5 ►, ^ H !2i 05 ftj a H g !z; w PLATES BIVBTS 82.3 103 107 104 104 A TREATISE ON STEAM BOILERS. TABLE XXXII— Continued. DIMENSIONS OF SPECIMKN. Iron plate, 6)^ inches wide, two covers, single riv- eted, four Ys inch riv- ets in each plate, Fig- ure 14 Steel plate, 63^ inches wide, two covers, single riv- eted, four % inch riv- ets in each plate. Fig- ure 14 Iron plate, 1%, inches wide, two covers, double riv- eted, six % inch rivets in each plate. Figure 15 Steel i>late, 73^ inches wide, two covers, double riv- eted, six 5^ inch rivets in each plate. Figure 15 c5>^ ^« R s w s . < H M w « ft ., fe o Bo^ <;z;i2i M >i '-' H < 76,850 95,150 89,150 110,500 O K H "" hi !<5 S ft M O O >1 ft ^§ fe o O fc 18,700 21,700 18,70n 21,700 fe fc 11,746 13,100 n,8S6 14,740 W B < H g 1-! O g fe H W CO PL, 62.7 60.4 6.3.5 68.0 ;2; « S ^ II, ft ^ ^ 5 in !z; s H £ 6- « 3 « W fe 3 K o K <^ H S ►J 1^ H M pj o u ;; H ffi « "* ^ 55 Ph ■< o o 02 SIZE. 1-3 < O 1 H 03 . H 02 02 ? H ! »» V y • • • • • • 1* • • • ;• • ( X Figure 20. NET ULTIMATE STRENGTH OF JOINT. DESCRIPTION OF JOINT. \ INCH PLATES. /^ INCH PLATES. 1 INCH. PLATES. AVERAGE FOR THE THREE PLATES. Entire plate PER CENT. 100 40 40 59 PER CENT. 100 50 54 70 PER CENT. 100 60 •52 72 PER CENT, 100 Single riveted, hand, fig- ure 19 50 Single riveted, machine, figure 19 49 Double riveted (chain), Figure 20 67 We thus have an average gain by chain riveting over single riveting in SHEARING TESTS OF IRON. 113 One-half inch plates of ...>: 19 per cent. Seven-sixteenths inch plates of. 18 per cent. Three-eighths inch plates of. 16 per cent. These joints were all made with three-quarter inch rivets, arranged in the specimens at two-inch centers. Testing rivets — The strength of a riveted joint depends so much on the strength of the rivets which enter into it, that it is of the utmost importance to know the quality of the materials of which they are made, before putting any work on them. One of the simplest and at the same time a severe test, is to upset a rivet on an anvil under a heavy hammer, say to one-half its original length and without splitting it. The writer has employed this test in deter- mining the quality of supplies, and finds that specimens selected at random, and which will stand this test, have usually all the- other qualities of a good rivet; but as the strain brought upon rivets is that of shearing, tests should be made to determine the resistance to the separation of the rivet at the line of plates composing the joints, which may be either single or double shear, the former being the ordinary practice in this coun- try, the latter the exception. ^^^^^^ Single shearing is clearly ^^^^Iflil^^^^ shown in figure 21, which repre- \ J lJ' Ji " ' '^M \\,\\7 il'/m/ sents a rivet in a single riveted ■ 'Si^ . . , 1 . . Figure 21. joint, undergoing separation. Mr. Wm. H. Shock, chief engineer U. S. ]^., made some shearing tests of iron for stay bolts with results as given below. This quality of iron is not as high as that usually employed in the manufacture of rivets, but still of good quality. The iron was made up into bolts with nuts, instead of being riveted into the testing plates, as is the usual practice. There were sixty of these bolts in all ; twelve 114 A TREATISE ON STEAM BOILERS. each of the following sizes, viz : one-half, five-eighths, three-quarters, seven-eighths and one inch diameter. These were forged in the usual manner, without any reference whatever to the experimental tests to be made. The fol- lowing table gives the results of the tests : TABLE XXXVI. RESULTS OF EXPERIMENTS ON SHEARING STRAINS OF IRON BOLTS, HY AVILLIAM H. SHOCK, CHIEF ENGINEER, UNITED STATES NAVY. SINGLE SHEAR. K ■ 1 to 6. 7 to 12. i:{ to 18. 19 to 24. DOUBLE SHEAR. S f- w h3 -^ " .515 .646G .7833 .9033 25 to .30 11.036 I Mean of the above.! H Q 9183.3 12808.3 19025. 26562.5 34358.3 <; !z; i^ 3 -I g ■^ i^ o O H 1^ ■ *l IB CC ?. li< "^ < o g a o fe O (1, z a - << o .20831 .32837 .48192 .64089 .84395 W CO M 44,085 39,00C 39,477 41,446 40,071 40,817 a . S5 31 to 36 37 to 42 43 to 48 49 to 54 55 to 60 Mean H ^ W t-H h; fe '^ s. f in 3 inches. 29 .35. inch. .096 inch. 49 51 7,500 fl)s. 38,206 lbs. 11,750 11>s. 59,850 His. 11^ in 3 inches. 28 .33 .086 43.6 56.4 inch, inch. Single riveted lap joints — This is the simplest form of a riveted joint and is used almost exclusivel}^ in riveted seams in boiler shells when of forty inches or less m diameter. Some manufacturers begin double riveting at thirty- six: inches, but this is the exception rather than the rule. The single rivited joint, though easily made, is at best but about one-half the tensile strength of the solid plate. Among the defects of construction may be mentioned the liability of tearing the plate through the line of rivet figure 22. holes, as shown in tigure 22. This is liable to occur in any case where the rivet holes are punched too close together, thus reducing the strength of the plate below the shearing strength of the rivets. This is a fault which one may be easily led into and is perhaps the commonest defect in boiler construction. 122 A TREATISE ON STEAM BOILERS. Figure 23. Another cause of failure, though not nearly so common as the former, is that of punching the holes too near the edge of the plate. When the distance from the edge of the rivet is too near the edge of the plate the latter is likely to give way in front of the rivet, as in figure 23. This defect is easily remedied by simply allowing a wider margin; and in consequence may be easily overdone, for if the edge of the plate be too far from the joint it makes calking the joint steam tight a much more difficult matter, owing to the spring or elasticity of the plate. On the other hand, if the rivet holes have their centers too far apart and the distance from the edge of the hole to the edge of the plate be such that the plate can not yield, as in figures 22 and 23, then there is a possibility of .,.._-.^SII shearing off' the rivets, as in figure '^ 24. This is likely to occur when the rivets are too small in diara- figurr 24. eterfor the thickness of the plate. The ultimate strength of a plate depends upon its area of cross section ; and the loss of area caused by punching the holes for the insertion of the rivet, reduces the strength of the plate simply in that amount. With rivets, however, the case in quite dissimilar, for the strength of the rivet increases as the square of its diameter. In the former case the strength of the plate consists merelj^ in that of the net area through the line of rivet holes; in the latter the resis- tance to shearing increases with the increased area of the rivet. Other things being equal, that is the best joint in which the strength of the plate and the resistance of the rivet to shearing are equal to each other. Hundreds of tests have been made to determine by direct experiment the best proportions for single riveted joints. PROPORTIONS FOR RIVETED JOINTS. 123 The differences in quality of boiler plate and rivets, together with the great uncertainty as to the exact effect of punching iron plates, have, so far, prevented anything like the determining either by calculation or experiment of what might be accepted as the exact, or better, perhaps, the best proportions for riveted joints. The writer has examined many formulas and finds that in most cases they are suited only to the one or two thicknesses of plates for which they were evidently intended, being usually three- eighths and seven-sixteenths inch, and have the appear- ance of having been worked out for the seams in Cornish or other large diameter internally fired boilers. The thinner plates, one-fourth inch, for example, in English tables of proportions for riveted joints, give one- half inch as the proper diameter of rivets. This is not in accordance with American practice; five-eighths inch rivets in one-fourth inch plates being almost universal. The spacing of rivets is also greater in this country than in England. The following tables were compiled by the writer for his own use, partly from theoretical deductions, partly from tests made on riveted joints, and also by a comparison of these with the practice of successful and intelligent manu- facturers. It will be observed that the dimensions are empirical, yet they have served a good purpose and on the whole are quite reliable. This matter of spacing rivets is, at best, approximate only, and may be changed within narrow limits ; it often happens that spacing will not come out even, and in such cases, whether the centers shall be increased or decreased, rests entirely upon the judgment of the designer. In general, it has been the practice of the writer to use the proportions in the table, and when spaces occur at the end to put in the extra rivet instead of throw- ing it out. 124 A TREATISE ON STEAM BOILERS. TABLE XLII. SHOWING DIAMETER AND SPACING OF RIVETS IN SINGLE RIVETED LAP JOINTS. •THICKNESS DIAMETER LENGTH CENTER OF RIVET TO EDGE OF PLATE. CENTER TO OF PLATE. OF RIVET. OF RIVET. CENTER OF RIVETS. ^ inch. ^ 1 13 T6 u ^ inch. 1 n ] If ^ inch. 5 8 n 1 n f inch. f If lA 2 ^^ inch. 3 4 2 1t^ 91 ^8- ^ inch. 7 8 2i 1 3 93 -^ inch. "8 2^ 1 3 ^ 8 01 -■2" f inch. ] 2f 1 3 2f ^^ inch. 1 3 ll^ 2| f inch. H 3} If 3 Double riveting — Boilers ought to be double riveted, if for no other reason, simply as a matter of economy, for the strength of the single riveted joint, which is only about one-half the strength of the solid plate, is increased by about twenty per cent by double riveting, without any such corresponding increase in cost. The strength of a joint depends largely upon the strength of the rivets, and these must be so disposed in the joint as to utilize the strength of a larger number than can be used in single riveting, and at the same time increase the net sectional area of the plate in the line of punched holes in the joint. The waiter does not think it a good plan to change the diameters of rivets in fixing upon single or double riveted joints. The two tables therefore contain the same diameters of rivets for the same thickness of plates. DOUBLE RIVETED LAP JOINTS. 125 TABLE XLIII. SHOWING DIAMETER AND SPACING OF BIVETS IN DOUBLE RIVETED LAP JOINTS. Figure 25. THICKNESS RIVETS. CENTER TO CENTER CENTER CENTER OF EDGE OF TO TO PLATE. DIAMETER LENGTH. PLATE, CENTER. CENTER. CENTER. A B b C D E F i 8 ^ 2 n 1 9 ^T6 5 ITT f n 2i 2 1 21 ^3T 3 8 1 If 1 3 2| 9 1 ^ 8" 123 ^3^ 7 3 2 1 3 2| 9 1 If 1 7 '8 2i 1 •i J 8 • 3 9 7_ '■le 129 ^32 9 - TIT 7 "8 ^ 1 '^ ' 8 3i 2A 2 5 8 ] 93 1 ■> 3* 2f 2i 11 TIT 1 3 1 •> 3f 2f 2A 1 H ^4 1 ^ -^ 4 4 3 2i Tbe distance F is approximate only, column E being the exact distance. Calking — In boiler making calking is a process of upsetting the overlapping edges of plates by means of a tool called a calking chisel. A full size representation of the calking end is given in figure 25 and marked '^old 126 A TREATISE ON STEAM BOILERS. style calking." When two rough boiler plates are riveted together they are not .steam tight; the object of calking is to "upset" the edge of the overlapping plate and drive it firmly down upon the one underneath. This operation does not, of course, alter the character of the joint; it sim- ply forces the edge of the outside plate down firmly upon the lower one and thus a joint at first approximately tight is rendered altogether so by this simple operation. The edges of plates ought always to be planed or sheared to the proper bevel for calking before riveting together. The angle of plates best suited for calking is about 20° less than a right angle. The practice of chip- ping seams after riveting is altogether wrong, as it endangers the strength of the plate underneath by the frequent and inevitable markings caused by the slipping of the chisel in the hands of the chipper. The markings are ruinous to the plates containing them, and are, no doubt, a frequent cause of disaster. Aside from the injury done the plates by careless chipping, the operation of calking by means of a sharp edge, even though it approximate a right angle, is also destructive to the lower plate, by form- ing a slight indentation the v^hole length of the seam so operated upon. An improved form of calking, patented by Mr. James W. Connery, Philadelphia, Pa., is shown in figure 26, and named by him "concave" calking, after the appearance of the finished joint. The object of concave calking is to bring together the seams of a boiler after riveting, in such a manner that they shall be perfectly steam tight and at the same time not in any manner injure the under sheet. This is eiFectually accomplished by the use of a tool with a semi-cylindrical end, producing a concave depression in the bevelled edge of the lap ; slightly dividing the plate calked and driving the divided part towards the rivets, forming a bearing from one-half to three-quarters of an inch, thereby forming a concavp: calking. 127 proper junction of the two surfaces and increasing the strength of the joint, without in any manner injuring the surface of the under plate. The accompanying cut illustrates the difference between the old and the new systems. Figure 2g. Full Size. The old plan is to chip or plane the edge of the over- lapping sheet, leaving a solid angle to it of about 80°, and then to drive up, by means of a hammer upon the tool shown at the right, the under face of the tool resting upon the under sheet, until the angle of the metal of the upper sheet has assumed something near the form there shown. With a tool of this form it is impossible to thorough!}^ upset or calk the metal of the upper plate without a more or less injury or scoring of the Under plate; and this will not be the only injury done the under sheet, for it is weU known that in all processes of hammering, rolling and otherwise compressing iron, it becomes harder and more dense, and as there is nothing in the process of calking with this tool which makes any change in the material of 128 A TREATISE ON STEAM BOILERS. the under plate, it follows that after indentations and chan- nel] ngs are made in the under plate by the calking tool, the extreme edge of the upper plate, while being hardened and compressed, will be imbedded in the under plate, thus aggravating the injury done with the tool. These effects are plainly shown in plates which are cut apart after the most careful calking, and is well illustrated in the figure as giving to the plate that starting point of fracture with which all mechanics in metal are familiar. With Connerv's improvement a concave depression is produced in the bev- elled edge of the lap, the crown of the tool being entered in the edge of the plate at such a distance from the under plate as will leave, when finished, a considerable thickness of metal between the concave groove and the lower plate ; the surface of the compressed and hardened metal, driven down upon the lower plate, will be too large to cause any appreciable disturbance of the surface of the under plate, while the tool can, under no circumstance, injure or mar the lower plate in any way. It will be readily seen, too, that this form of tool, com- mencing as it does on a small surface for indentation of the edge, must result in carrying the compression or condensa- tion of the iron of the lap to a much greater depth than is possible with the old method, thus tending to bring about a permanent strain upon the iron through the line of rivets in a much less degree. This is indicated on the left of the cut by the deep wedge of dark shading running nearly into the rivets. Many of our first class establishments have adopted this method of calking, among which is the Baldwin Loco- motive Works of Philadelphia, who have been using this improvement exclusively for several years upon their hun- dreds of locomotives and indorse it in the highest degree. The writer is so fully impressed with the value of thi& invention that he does not hesitate to recommend it in all CONCAVE CALKINa. 129 cases as being superior to any other method of calking of which he has any knowledge; such a thing as grooving or injuring the lower plate by calking being practically impossible, and w^hich gives this invention its chief value. The following tests were made at the Washington E^avy Yard: "Five plates of different thicknesses were riveted together and the four seams on one side were calked by the Connery process, and those on the opposite side (by differ- ent boiler makers employed in the yard) by the ordinary process. Upon cutting the sheets apart, in every case, it was found that the bearing surfaces of the sheets were about double that of the seams calked by the ordinary method, and that there was no injury done to the under sheets, whereas the under sheets of the seams calked in the ordinary way were slightly indented in some cases, and in others were channeled or grooved about one-thirty-second of an inch in depth, depending upon the skill and care of the workmen employed. Several seams were also calked by the above process on an experimental cylinder which was subsequently tested to its collapsing pressure of one hundred and thirty-four pounds per square inch, without the slightest leak, whereas a number of leaks made their appearance in seams calked by the ordinary process." Locomotive boilers tested by hydraulic pressure to more than three hundred pounds per square inch, and afterwards used with a steam pressure of one hundred and fifty pounds, showed no leakage at seams with this calking. 'No joints in a boiler are more difficult to get tight than those which are single riveted. This is due to a par- tial distortion of the joint, caused by the shell of the boiler assuming a more perfect cylindrical form when pressure is applied than that given it in the boiler shop during the process of manufacture. Figure 27 shows, by means of the dotted lines, the curved surface of the cylindric part of the shell, and the (10) 130 A TREATISE ON STEAM BOILERS. full lines the actaal positioa of the joint; and it is just here that the mischievous effects of the grooving caused Figure 27. Full Size. by chipping and sharp calking become fully apparent, and which is shown in an exaggerated degree in figure 26, by the breaking of the plate. CHAPTER VII WELDING, FLANGING AND INFLUENCE OF TEMPERATURE. Welding Boiler Plates — Advantages Claimed — Objections to Welding Externally Fired Boiler Shells — Practical Difficulties in Welding Long Seams — Strength of Welded Joints — Welded Rings for Boilers — Flanging — Influence of Temperature on Boiler Plates — Mr. Isherwood on the Franklin Institute Experiments of 1837. Welding boiler plate joints — -.Very little attention has been given in this country to the production of welded boiler plate joints. The few experiments that have been made have been so crudely done that no intelligent opinion can be formed as to the relative costs of welded and riveted work. In England welded joints have been in use for sev- eral years and for some purposes is steadily growing in favor, though it is not practiced in boiler construction to any great extent as yet. The advantages claimed for this form of joint over the ordinary riveted joint are, 1. That the welding approximates more clearly the original strength of the plates than the best forms of riv- eted joints can possibl}^ do, besides being entirely free from the bad eflects of punching and loss of strength occasioned by drifting, as well as the injury done the plates by cold hammering. 2. The welded joint needs no calking, and thus, next to drifting, one of the greatest evils through bad workman- ship in boiler construction is rid of entirely. 3. By welding the rings in the shell of a boiler they may be re-rolled after the work is done on them, and thus 132 A TREATISE ON STEAM BOILERS. a perfectly cylindrical shell can be produced, a thing^ impossible in the ordinary lap joints. If an entire shell could be welded, it would remove at once the objectionable two thicknesses of plate in the fire and the trouble incident to the accumulation of deposit which is likely to form around the joints iind rivet heads; and, further, there being no jointed seam, entirely precludes such a thing as corrosion caused by leakage at the lap of the plates or around loose or imperfectly fitted rivets. Whether a welded joint is to be preferred over a riv- eted one, will depend upon circumstances. In an inter- nally fired boiler it is important that the main flue should be truly c^^lindrical, as the resistance to collapse depends largely upon this. The best makers usually employ in its construction a butt riveted joint with the seam underneath. The objections to this are, that it is impossible to perfectly calk such a seam when once in place; and then the seam of rivets along the bottom of the flue will prevent the ready removal of ashes and dust which accumulates along its whole length. Should there be a leaky joint, a thing we may almost certainly count on, there will in time be quite an accumulation of hard baked ashes and cinders the whole leno^th of the flue. In case the fuel used con- tained sulphur, there would be more or less of sulphurous oxide mixed up with the ashes or deposited along the sides or bottom of the boiler flue, and which, if once wet or dampened, will attack the flue by external corrosion and seriously impair its strength. In such a case a welded and perfectly tight flue would possess a marked advantage over the other, to say nothing of that gained by tbe truly cylin- drical form; this advantage, it should be understood, refers to the facility and certainty in cleaning and freedom from leaks, and not that the corrosive action would be less under the same conditions. WELDING BOILER PLATES. 133 In such a flue, as just described, the pressure tends to collapse and thus to tighten the weld. An imperfect weld might, in such a flue, escape detection for a long time, but which would soon make itself apparent in any case where internal pressures were employed. In externally fired boilers, the main advantage of welded seams over riveted ones appear to be the getting rid of all the double thicknesses of plates in the fire. This is at all times a desirable thing to do. In boilers of this type, the straii>s are from within, outward, and the safety of the shell depends entirely upon the tensile strength and ductility of the plates and the soundness of the weld. The strength of a riveted joint is known to within a very small percent- age of the weight required to tear it apart. For welded joints, unfortunately, no such exact data exists, and from the nature of the joint it is exceedingly difl3.cult to arrive at anything even approximating its actual strength ; not that experimental tests are wanting, nor that suflicient time has been denied the subject in order to make the fullest investigation in regard to the effect upon the plates joined by welding, the weld itself, or the proper mechanical manipulation of the plates in the fire. We have all this, and it only goes to show what the possibilities are, but gives no data as to probabilities in actual practice, on a large scale, with even good facilities and skilled workmen. The welding of two plates in a well made open fire is attended with greater- risks than the welding of two bars of iron. The reasons for this are quite obvious. In the case of the bars, their ends are in the center of the fire and entirely shut off" from the injurious eff'ects of free oxygen, if the fire is properly made. When a thick fire is built upon the tweer, the air passing up through it gives up its oxygen to the highly heated carbon, and carbonic acid gas is formed as the result of this union, and in passing up still further through this bed of burning coal, the carbon in the 134 A TREATISE ON STEAM BOILERS. upper portion of the lire may take up a portion of the oxygen in the carbonic acid gas, and carbonic oxide gas. is formed. J^either of these gases has an injurious effect upon iron so far as welding is concerned, and in the case of the two bars referred to above, they are in this highly heated chamber of gases formed by the sides and cover of the fire, and may be readily brought to a welding heat without any fear of oxidation, for there is no excess of oxygen in the fire to come in contact with the iron. In the case of the plates it is somewhat different, for the fire being hottest in the center and of lower tempera- ture toward the edges, it is not possible to confine the plates to a chamber of heated gases from which oxygen is excluded, for no such chamber exists, and can not, from the nature of the case. Further, every movement of the plates brings the more or less highly heated portions in contact with the air, when oxidation instantly occurs, form- ing an oxide of iron or hard cinder which prevents welding. There is at the same time a partial loss of iron, but this is not of so much account as the bad effects resulting from the presence of the cinder in the weld. It is not practicable to heat any considerable length of plate in an ordinary forge or flanging fire at one time, and as the oxidation referred to is sure to occur in a greater or less degree, the surfaces must be protected from oxidation by means of a flux. The one generally used is sand; this is composed of silicon and oxygen. ' The action of the flux may be said to be two-fold : first, in forming a vitreous coating over the iron, and second, in reducing the temper- ature of the parts to which it is applied, and arises from the circumstance that iron is usually "scarfed" at the place where it is to be welded ; we thus have a thick and a thinner portion of the same plate exposed to the action of the heat, the thinner portion of the plate being nearest the center of the fire, and arrives at a welding heat long before WELDING BOILER PLATES. 135 the thicker p.ortioa of the plate attains a similar heat. If the action of the heat was not checked, this thinner edge would be burned away long before the plate was brought to the welding point. In order to prevent this the sand or other flux is used, and in coming in contact with the highly heated iron it is melted and absorbs so much heat from the iron that it gives the latter a vitreous coating. This sili- cate combines with the iron and covers that portion of it which is of sufficiently high temperature to melt the sand. This silicate being of a very refractory nature, will last some time in the fire before it burns oiF the iron, and in this manner serves to protect the thinner parts of the iron, while the thicker portion is absorbing the heat and arriv- ing at a welding condition. In using sand as a flux, care must be exercised that it be kept, or afterward cleaned off, the inside of the joint where the two scarfed edges of the plate are to be welded, as its presence in the weld prevents perfect contact, and thus weakens the joint. For small work, borax is the flux generally employed in the forge for welding. It prevents oxidation in the same manner as already described for sand. There are many ways of making a welded joint and they will vary anywhere from good to bad in strength and soundness. Scarf welding is on the whole to be preferred to simply lapping the plates and then welding. In scarf- ing, the edges of the plates should be upset and then thinned down, not exactly to a sharp edge, but say one- sixteenth of an inch, or perhaps less. The exact thickness is of course no material part of the process of making a good joint ; neither is the thinning of so much importance as the upsetting of the edge of the plate to a thickness greater than that of the plate itself, the object being that when the weld is made it may then be finished down with suitable "flatters" to the regular thickness. 136 A TREATISE ON STEAM BOILERS. Ill the manufacture of welded boilers as a business, it would be necessary to construct a special heating appa- ratus, which would probably consist of an external and an internal gas furnace, operating on the principle of the blow pipe, in which the flame of the burning gas would be directed against such portions of the joint as needed the greater heat. Such an apparatus could be made in which no free oxygen could reach the heated plates, and thus welds could be made without the use of a flux of any kind. The plates could be heated the whole length at one time and when brought to the proper heat could be welded by pressure instead of by hammering. What the future may bring forth, it is impossible even to conjecture, but at present welded boiler plate joints, especially when intended for externally fired boilers, are untrustworthy and are almost sure to contain imperfec- tions in the weld which the usual hydraulic test fails to indicate, but which will reveal themselves sooner or later in the expansion and contraction incident to heating and cooling in actual service. Fjgure 28. Figure 29. Figure 30. In Mr. Bertram's experiments on welded joints the lap welded test pieces, figure 30, were inferior in strength to those scarf welded, figure 29. WELDING STEEL PLATES. 137 The specimens tested were four inches wide by three- eighths, seven-sixteenths and one-half inch in thickness. The lap of the joint was one and a quarter inches. The results were as follows : » TABLE XLIV. Strength of entire plate, per cent Strength of scarf welded joint, fig. 29, Strength of lap welded joint, fig. £0, f INCH PLATE. 100 Faulty. 50 /e INCH PLATE. 100 106 69 % INCH PLATE. 100 102 62 From the above data it appears that the strength of joints united by lap welding are scarcely better than single riveting, or about forty per cent weaker than the plates which compose the joint. Scarf welding, on the contrary, equaled the strength of the plate. 'No doubt the shape of the joint under severe stress had much to do with the low- ering of its strength in consequence of the indirect pull. In regard to the welding of steel boiler plates Mr. Daniel Adamson says: "After many trials and many failures in attempt- ing to weld steel boiler plates, the writer found it necessary to ascertain in all cases the composition of the metal before putting any labor on it, and from a large experience it is now considered desirable that the carbon should not exceed one-eighth of a per cent, while the sulphur and phosphorus should, if possible, be kept as low as .04 per cent, silicon being admissable up to the extent of -^^ of 1 per cent. Farther experience is yet required to ascertain what exact composition gives the most satisfactory results by welding. At present some preference may be said to be given to the Martin-Siemens 138 A TREATISE ON STEAM BOILERS. class as compared with Bessemer metal, when both are of about the same chemical compositi-on." Weldle^s rings for boilers — Several years ago (1865?) Mr. Ramsbottom designed a machine to work annular ingots of Bessemer steel, or other metal, into cylinders of such length and thinness that they may be put between rollers and rolled round and round and reduced to the thinness required to make boilers. The machine consisted of a mandrel, on which the hoop or annular ingot was placed- On each side of tbe mandrel was placed a roller, the sur- faces of the roller and mandrel being grooved diagonally in opposite directions, thus leaving diamond shaped projec- tions on them. The rollers were intended to be driven by steam or other power, and were pressed against the hoop or ingot, which is enlarged in diameter and expanded lengthwise by the pressure and the lateral action of the projections on the rollers and mandrel. When the hoop or ingot has been thus partially expanded it was then to be put on another mandrel of larger diameter and again acted upon by the rollers; or it might be put on a revolv- ing mandrel and thus expanded both in length and diameter by a roller which is traversed to and fro in the direction of the axis of the hoop. In this arrangement the hoop and traversing roller must be pressed together. It has also been proposed and in a measure carried out, to forge the annular ingots by means of a steam or power hammer into the cylindrical sections of a boiler ready to rivet into a continuous shell. From the i)resent outlook it does not appear that either of these methods are likely to supersede the making up of Hat sheets into shells and large flues, because of the increased cost of manufacture of the cylindrical weldless hoops over that of flat iron rolled, and then riveted or welded. FLANGING. 13^ Flanging — The exterior flanging of a boiler head and small flue holes is about as severe a test as plate iron gen- erally receives in the process of boiler construction. By far the greater number of heads used in this country are hand flanged; there are very few boiler making establish- ments having enough flanging 'to do to warrant the erection of suitable furnaces and machines. The few machines which are in use, however, attest the superiority of machine over hand flanging. In the first place the heads are perfectly round, an important matter- of detail in boiler construction ; in the next j)lace the flange is turned perfectly true and at right angles to the face of the head. In hand flanging it is almost impossible to secure either or both of these in the same plate. The heating is done in the ordinary forge fire and liable to all the objec- tions of overheating one portion of the plate while other portions are not of sufficiently high temperature to insure the best working. In flanging an iron or steel plate, it should be done with wooden mauls, bending the plate over a cast iron former. The blows should be light and distributed over as large a surface as possible, avoiding anything like short bends in turning the flange. The heating, when done in an ordin- ary flange fire, must of necessity be local, and, hence, will require the greater care in working. As the flanging approaches completion by successive stages of heating and hammering, care must be exercised that the plate, if of steel, is not ruined by cracking or splitting, which may be induced by internal strains. To avoid this in subsequent working or handling, it should be immediately annealed by heating the whole plate gradually and evenly, until brought to a low red heat, and then allowing it to cool slowly, not disturbing it until entirely cold. The writer has used a considerable number of heads,, machine flanged, by Phillips, Nimick & Co., and aside from 140 A TKEATISE ON STEAM BOILERS. the superior quality of "Sligo" iron, these heads seem to possess an advantage over the hest hand flanged work by the strengthening of the plate in the curve, as shovv^ii in the annexed engravings. Figure 31 represents the thinning of the curve occa- sioned by the stretching of the plate over the cast iron former in hand flanging, the dotted line representing the normal curve and the middle line the actual thickness of metal. Fiorure 32 is a representation of the thickening of the curve, taken from machine flanged heads, made on the machine used by the above named firm. The normal curve, it will be noticed, falls considerably within the actual line of the metal. The advantages gained by the strengthening of the head at that particular point are quite obvious, and are not likely to be underestimated. Figure 33 is an engraving made from a photograph taken from a boiler head, and is, all things considered, one of the best specimens of machine flanging the writer has yet seen. The dimensions of the head are as follows Diameter outside, 72 inches. Diameter large flue, 40 inches. Diameter 2 holes, 12 inches. Diameter 2 holes, 8 inches. Diameter 6 holes, 6 inches. Diameter 1 hole, 5 inches. The influence of temperature on boiler plates — The influ- ence of diflerent temperatures on the strength of iron or steel is a question of great importance in engineering con- struction, and probably nowhere more so than in steam boilers. It is supposed that the efl*ect of repeated changes in the temperature of iron plates brings about certain molecular changes, which destroys the cohesion of the iron in the same manner that it would be destroyed by the con- tinued vibrations of a plate caused by any external force. INFLUENCE OF TEMPERATURE ON PLATES. 141 It is well known that the continued reheating and cool- ing of iron will shortly render it entirely worthless, if it approach a red heat. Plates cut out of that portion of old hollers suhjected to the action of the fire are almost invariably hard and brittle, and will seldom show one- fourth of its original ductility, together with a marked decrease in tensile strength. This can be referred to na FlGUKK 33. other cause than that produced by molecular changes brought about by the long continued action of the fire. Wrought iron is apt to blister or crack in overheating and changes its structure from fibrous to coarse granular,, losing in tensile strength and ductility and becomes^more 142 A TREATISE ON STEAM BOILERS. brittle. The loss in tensile strength in overheated iron plates has doubtless been a cause of many boiler explo- sions and can only be explained by the possibility that the continued variation and diiferonces of temperature of the outer and inner surfaces of the plate have diminished the cohesion of the fibers or laminse composing it. If this is true of fibrous iron, what would be the effect of the temperature on granular iron? Sir William Fair- bairn ascertained, in his experiments, that on the whole cast iron of average quality loses strength when heated beyond a mean temperature of one hundred and twenty degrees, and that it becomes insecure at the freezing point or under thirty- two degrees Fahrenheit. Cast iron yields to the fire sooner than wrought iron ; it loses a consider- -able percentage of strength at about two hundred degrees, and when red hot will scarcely sustain its own weight. The effect of variations of temperature on mild steel has not been so closely observed as that on wrought iron, and while we have any reason to believe that molecular changes are undergone in the material in consequence of repeated heating and cooling, it does not appear that the strength is diminished in any considerable amount when the plates nre used in steam boilers. In case the boiler should become short of water, or in event of plates being overheated in such portions of the boiler as are not protected by the water, then the effect of excessive and continued over- heating is to render the plates coarse granular, losing in tensile strength and toughness. Wrought iron will the better withstand the effects ot repeated heating and cooling in proportion as it is free from cinder; hence, a tine granular and homogeneous iron will resist the bad effects of reheating better and longer than a course fibrous iron, because the latter is made fibrous by containing in its composition more or less cinder,, which, owing to the lower temperature of the finishing ANNEALING. 143 • heats, was prevented from escaping. As a result, it can never equal in strength and ductility the granular iron, and is apt, with even a moderate degree of overheating, to become extremely brittle ; and it is for this reason that it should never be used for the fire sheets of boilers. Heating to redness and slow cooling — or in other words, annealing — has an eflect on steel in which it is rendered more ductile than before, but at a reduction of its tensile strength ; such a plate once annealed does not apparently change by reheati'ng, unless the temperature is higher than that to which it was first brought. By this it is not to be understood that no molecular changes are going on in the metal because of the lower temperature, but rather that the destructive changes in steel plates are less in degree than in iron plates if there is no sudden cooling. Steel plates, in order to anneal them properly, should be brought up to a temperature higher than that at which the final work was done on them. A cherry red will be found to be somewhat higher than the final working heat, and in boilers higher than any subsequent heat. Steel plates must not be^ annealed at too high a temperature ; that is, the temperature must not be near the melting point, because it will change its texture and crystallize by slow cooling, thereby losing in tenacity, ductility and elasticity, rendering the plate worthless. . The operation of heating and sudden cooling produces efiects directly the opposite of the above, and its subse- quent behavior is not unlike that of similar pieces which have been brought to a strain exceeding the elastic limit; that is, the- tensile strength is increased, as is also its brit- tleness. Any metal, therefore, which will harden in cooling is not fit to enter in boiler construction because of this very property of becoming brittle and thus reducing its power to endure sudden variations in load or resistance to shocks. 144 A TREATISE ON STEAM BOILERS. In regard to the accidental overheating of steel plates^ caused by low water, there have been many instances of collapse or of bulging, but such a thing as fracture in con- nection with overheating is almost unknown. Mr. Adamson observes that few or no malleable metalSy such as wrought iron or mild steels, can be found in the open market that possess a range of endurance at all vary- ing temperatures, say, from cold up to red heat, but nearly all ordinary bar or boiler iron and mild steels will endure considerable percussive force when cold and up to 450°" Fahrenheit, 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 poorer class of metals at this temperature, which may be called a color heat, varying from a light straw" to a purple and dark blue, are simply rotten. Some of these peculiar properties are illustrated by a series of tests of various qualities of metal; for example: ordinary merchant iron shows that it may be bent cold, or it may be bent red hot without signs of breakage or much distress. Examples are not wanting to show that irons will endure this bending test when cold or when red hot, but such a heat as can be induced by placing the metal into a bath of boiling tallow, registering a temperature of about 610° Fahrenheit, these metals break through by being bent, lose most of their malleability and snap off short under the action of the hammer. The same unfortunate element is exhibited by the mild class of Bessemer and Martin-Siemens steel, with this dif- ference, that they bent better cold and more pleasantly when hot, but both break up by percussive action at the medium temperature before named, the Martin-Siemens enduring somewhat better than the Bessemer class under these tests. During several years of observation Mr. Adamson has come to the conclusion that no metal containing much INFLUENCE OF TEMPERATURE ON IRON. 145 above a trace of sulphur can endure bending at this color heat, while at the same time the phosphorus must be low; in fkct, such endurance can only be obtained by a compar- atively pure iron, unalloyed by any other ingredients. Experiments made by Mr. Adamson, with bars of iron one inch in diameter and ten inches long between supports, when under tensile strain gave the following mechanical data : Permanent set induced per square inch. ...36, 287 pounds. Maximum strain per square inch 53,476 pounds. Elongation under maximum strain 18 percent. Final breaking strain on original area per square inch 50,929 pounds. Percent of elongation 20.5 A piece of this same iron subjected to chemical analysis yielded, Iron 99.44 Carbon trace Manganese 0.10 Silicon 0.16 Sulphur 0.01 Phosphorus 0.29 100.00 This iron was tested with a view to examine its power of endurance at a low heat, and at temperatures varying from 500° to 600° Fahrenheit it was found very difficult to get a bent piece ; and by referring to the composition of this iron, it will be found to contain a large measure of phosphorus, which in some degree may explain its lack of power to resist percussive force at the heats just named ; nevertheless, this cheap ordinary iron is much more valuable for many practical purposes than pure and com- paratively expensive wrought irons.. Franklin Institute experiments — The influence of temper- ature on the strength of wrought iron boiler plates was (11) 146 A TREATISE ON STEAM BOILERS. investigated in 1837 by a commitee of the Franklin Insti- tute and their conclusions were, that, the tenacity of boiler plates increased with the temperature up to five hundred and fifty degrees Fahrenheit, at which point the tenacity began to diminish. The tensile strength per square inch of section At 32° Fahrenheit was 56,000 pounds. At 570° Fahrenheit was 66,500 pounds. At 720° Fahrenheit was 55,000 pounds. At 1050° Fahrenheit was 32,000 pounds. At 1240° Fahrenheit was 22,000 pounds. At 1317° Fahrenheit was 9,000 pounds. Mr. Isherwood, in a contribution to the Franklin Insti- tute Journal in 1874, shows, in a very convincing manner, that the committee erred in judgment in continuing the experiments with the same specimens successively after rupture. He says : " In the experiments the same piece of iron was suc- cessively ruptured and gave as a general result just what might have been expected, namely, increased tenacity at each rupture under ordinary atmospheric temperatures; but the committee failed to detect the reason and left the naked fact standing in their tables without explanation. The experiments made by the committee under high tem- perature were vitiated by the same cause, as they were made on the same piece of iron after it had been broken — often several times — under low temperatures. The com- mittee did not perceive that the greater tenacity of the iron observed under the high temperature might be due to the fact that the iron was then necessarily fractured at a stronger point than under the preceding low temperatures; but they compared, in all cases, the tensile strength obtained from the first trial under low temperatures with the tensile strength under the high temperature often after several fractures had been made under the low temperature and the weakest points thereby eliminated. INFLUENCE OF TEMPERATURE ON IRON. 147 '•' The tenacity thus found under the high temperature, was of course, as much too great, comparatively, as the tenacity under the low temperature, for the number of fractures made, exceeded the tenacity at the first fracture under the low temperature. Yet, obvious as is this deduc- tion, the committee ignored it and attributed the entire increase of tenacity shown under the high temperature to the influence of that temperature alone, while, in fact, this increase was mainly, if not wholly, due to the elimination of the weakest points by the several previous fractures of the iron made under low temperatures. "As far as I am aware, this fact of the necessarily increasing tenacity of the iron at successive fractures, as a consequence of the continued elimination of weaker points by each preceding fracture, is now pointed out by me for the first time. The failure to perceive it caused Professor Walter R. Johnson to attribute an actual increase of strength conferred on the iron by the simple process of stretching, whereas this result was solely due to the removal of weak points. "Combining this error with that of the increase of strength assumed to be due to high temperatures, but really due to the same cause, led him to propose what he termed * thermo-tension ' treatment of iron as a means of increasing its tenacity. The whole principle of his process, however, being based on fallacious assumptions, its practical appli- cation proved worthless. "From a careful comparison of all the experiments I have been able to collect concerning the influence of tem- perature on the tenacity of wrought iron, the results show that between the temperatures of zero and 550° Fahren- heit, this influence is exactly nil, developing the important fact that between these limits no provision need be made by the engineer for effect of difference of temperature." CHAPTER VIII. STRENGTH OF BOILERS. Ultimate Strength — Factor of Safety— Safe Working Load — Strength of Riveted Shells — Collapsing Pressures — Strength of Welded Tubes — Stay Bolts and Braces — Steam Domes — Man Holes. The strength of a boiler will depend upon the material of which it is made ; the form and dimensions of its exte- rior and interior portions; the strength of intersecting joints, such as steam domes, nozzles, etc.; the strength of the riveted joints, and that of the system of stays which bind the portions of the whole together. The ultimate strength of a boiler is seldom or never called in question, except in connection with its safe work- ing pressure ; the former being necessary, however, to the determining of the latter. The strength of iron and steel plates, both single and double riveted, have already been given, but it yet remains to ^x upon the strength of riveted shells and flues in their actual form and dimensions before the w^orking pressure can be set with safety. The ultimate strength of a boiler is the greatest pressure which it. is capable of withstanding without danger of rupture. It is not necessary that the failure occur at the moment of over pressure, but whether it is likely to occur at all by a continued application. Experiments of this kind are both difficult and costly, and are therefore rarely made. Knowing the longitudinal and transverse strength of iron or steel plates, the strength of riveted joints, and in part, the many destructive influences which are at work and daily lessening the strength of the boiler, a certain frac- FACTOR OF SAFETY. 149 tion of the ultimate strength, called a factor of safety, is selected as a basis of calculation at which boilers are con- sidered safe, after taking into account all the contingencies incident to boiler making and subsequent use (and, shall I say abuse?) in regular service. A factor of safety, in steam boilers, is a unit employed to show in what proportion a given pressure is less than the ultimate strength of the boiler. If a boiler is capable of withstanding an ultimate pressure of nine hundred pounds per square inch, and is used at a pressure of one hundred and fifty pounds, there is said to be a factor of safety of six with reference to the lower pressure, as com- pared with the ultimate strength. The numerical value given a factor of safety is the relation which it bears to the ultimate strength, and not that of the elastic limit; just what that figure should be for boilers has never been agreed upon, but has been narrowed down to either six or eight ; so that in ordinary boiler construction for land use, no very great discrepancies are likely to occur by the use of either in the regular course of business. In this country six is the ordinary factor of safety employed in all kinds of boiler work; in England it varies between six and eight. It is the practice among the best class of boiler makers in this country to make no boilers less than one-quarter inch thick, no matter if the factor of safety should reach ten or even twenty. This practice results mainly from the difficulty in calking the seams in the plates. 150 A TREATISE ON STEAM BOILEES. TABLE XLV. SHOWING THE TENSILE STRENGTH OF IRON AND STEEL PLATES, WITH SINGLE AND DOUBLE RIVETED JOINTS, AND THE SAFE WORKING STRENGTH PER SQUARE INCH OF SECTION, ALLOWING AS A FACTOR OF SAFETY ONE-SIXTH OF THE ULTIMATE STRENGTH. ULTIMATE STRENGTH OF SAFE WORKING LOAD OF RIVETED STRENGTH OF RIVETEE JOINTS. JOINTS PER SQUARE INCH. SOLID PLATE IN POUNDS PER SQUARE INCH. SINGLE RIVETED AT 66 PER CENT. DOUBLE RIVETED AT 70 PER CENT. SINGLE RIVETED. DOUBLE RIVETED. SOLID PLATE. 45,000 25,200 31,500 4,200 5,250 7,500 50,000 28,000 35,000 4,667 5,833 8,333 55,000 30,800 38,500 5,133 6,417 9,167 60,000 33,600 42,000 5,600 7,000 10,000 65,000 36,400 45,500 6,067 7,583 10,833 70,000 39,200 49,000 6,533 8,167 11,667 75.000 42,000 52,500 7,000 8,750 12,500 The elastic limit of wrought iron is not far from one- half its tensile strength ; assuming it to be one-half, then the safe working load of solid plate as given in the above table has a factor of safety of only three as compared with the elastic limit. It is much easier to make tests for ultimate strength than for the limit of elasticity and the results are more definite ; it is for this reason, probably, more than any other, that the factor of safety is made referable to the ultimate, rather than the elastic strength of the material. If we had nothing to deal with other than the pressure necessary to tear the boiler shell asunder in the line of rivet holes, or in the line of its least strength, the problem of strength in design would be a very simple one. Unfor- tunately this is not the case. Every one at all convers- ant with the details of boiler construction knows that STRENGTH OF BOILERS. 151 too many boilers are sent out with internal strains result- ing from bad workmanship, which no doubt in some cases will equal the intended working pressure. The efiect of these strains is to reduce the ultimate strength of the boiler and should always be taken into account. As there is no practical way of doing so we can only assume that a part of the. factor of safety is already expended. But in what proportion ? Perhaps no better answer can be given to this question than the data furnished in the following circular, issued by the English Board of Trade. The strength of boilers — The following circular, issued by the English Board of Trade, is for the information of engine and boiler makers, to enable them to know under what instructions the inspectors of the board of trade act in recommending the pressure of steam to be carried in boilers within their jurisdiction: *' When boilers are made of the best material, with all the rivet holes drilled in place and all the seams fitted with double butt straps of at least five-eighths the thickness of the plates they cover, and all the seams at least double-riv- eted with rivets having an allowance of not more than fifty per cent over the single shear, and provided that the boilers have been open to inspection during the whole period of construction, 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 sat- isfaction of the board's surveyors. But when the above conditions are not complied with, the conditions in the following scale must be added to the factor six, according to the circumstances of each case : 152 A TREATISE ON STEAM BOILERS. A .15 B .3 C .3 D .5 E* .75 F .1 G .15 H .15 I .2 J- .2 K .2 L .1 M .3 N .15 O 1. P .1 Q .2 R .1 S .1 T .2 V .25 X'' 1.65 To be added when all the holes are fair and good 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 doable 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 circumferential seams are not fitted with double butt straps and double riveted ; when the boiler is of such a length as to fire from both ends, or is of unusual length, such as flue boilers. 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 sat- isfied 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. Where marked * the allowances 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 : STRENGTH OF BOILERS. 153 (Pitch — diameter of rivets) X 100 f Percentage of strength of plate at — = < joint as compared with the solid Pitch. [ plate. (Area of rivets X No. of rows of rivets) X 100 f Percentage of strength of = i rivets, as compared with Pitch X thickness of plate. ( the solid plate.* " Then take iron as equal to twenty-three 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 scale given in this circular : (51,520 X percentage of strength of joint) X twice the thickness of the plate in inches. Inside diameter of the boiler in inches X factor of safety. Pressure to be allowed I per square inch on the safety valves. "Plates that are. drilled in place must be taken apart and the burr taken ofi', and the holes slightly countersunk from the outsides. 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 fiber. The rivet holes may be punched or drilled out of place, but when drilled in place must be taken apart and the burr taken off and slightly counter- sunk 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^TnoMAS Gray." Strength of riveted shells — The bursting pressure of a cylinder of either wrought iron or steel may be estimated as follows: Multiply together the tensile strength of the * If the rivets are exposed to double shear, multiply the percentage as found by 1.5 154 A TREATISE ON STEAM BOILERS. material in pounds per square inch and its thickness; divide this by the radius of the shell in inches, which will give the bursting pressure in pounds per square inch. By this rule a cylinder forty-eight inches diameter, one-quarter inch thick, of iron having a tensile strength of forty-five thousand pounds per square inch would yield, at a pressure of 468.75 pounds, as follows : 45.o_o^...2_5 ^ 468,75. This is true only of a continuous shell without any joint; as it is not practicable to construct such a boiler with our present appliances a deduction must be made for the seam of rivets. If single riveted the strength would be reduced to, say, fifty-six per cent of the above, which would lower the pressure to 262.5 pounds ; or if double riveted, to seventy per cent, which would ^x the bursting pressure at 328. pounds. The safe working pressure, allowing a factor of safety of six, would be ^6.|^j^ = 78.12 pounds per square inch. It will be observed that the factor of safety does not take into account whether the shell is single or double riveted. As there is approximately twenty per cent difler- ence between the net results of the two percentages it is too large to be overlooked. It is customary to double rivet all longitudinal seams in boilers over forty -four inches in diameter, but not the circumferential seams. The stress on the end of a boiler is the area of the head multiplied into the pressure, and for the same shell as above would be, 48 X 48 X .7854 = 1809.6 square inches area. The sec- tional area of the metal in the shell = 48 X 3.1416 X .25 = 37.7 square inches; this at 45,000 pounds per square inch would be capable of sustaining a load of 1,696,500 pounds before rupture ; or dividing this by the area thus, - floi y = 937.5 pounds necessary to produce transverse rup- ture, or twice that of the longitudinal seams. If the same reduction be made for riveted joints as in the preceding example then 937.5 X .56 = 525 pounds as the ultimate STRENGTH OF BOILERS. 155 strength to resist rupture, as against 262.5 pounds for a single riveted, or 328 pounds for double riveted longitu- dinal seams, which goes to show that nothing is to be gained by double riveting circumferential seams in ordi- nary cylindrical shells. If tubes or flues are inserted in the heads, their com- bined areas are to be deducted from the area of the head; this has the eflfect in many cases to reduce the pressure on the boiler heads more than one-half. The following tables, XLYI and XLYII, show the safe working pressure per square inch for iron or steel boilers, either single or double riveted : TABLE XLVI. SHOWING THE SAFE WORKING PRESSURE FOR SINGLE RIVETED IRON CYLINDER BOILERS, FROM TWENTY-FOUR TO SEVENTY-TWO INCHES IN DIAMETER, EMPLOYING A FACTOR OF SAFETY OF SIX. Single Riveted Iron Shells. DIAMETER OF THICKNESS OF SHELL. TENSILE STRENGTH PER SQUARE INCH. BOILER. 40,000 45,000 50,000 55,000 24 Y% inch. PRESSURE. 104 PRESSURE. 117 PRESSURE. 130 PRESSURE, 143 \ inch. 139 156 174 191 -^ inch. 174 195 217 239 26 ^ inch. 96 108 120 132 \ inch. 128 144 160 176 j\ inch. 160 180 200 220 28 -^ inch. 89 100 112 123 \ inch. 119 134 149 164 ^ inch. 149 167 186 205 30 ^ inch. 83 94 104 . 115 \ inch. 111 125 139 153 ^ inch. 139 156 174 191 156 A TREATISE ON STEAM BOILERS. TABLE XL VI— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARi : INCH. OF OF SHELL. BOILER. 40,000 45,0C0 50,000 55,000 PRESSURE. PRESSURE. PRESSURE. PRESSURE. 32 ■^ inch. 78 88 98 107 ^ inch. 91 102 114 125 ■^ inch. 130 146 163 179 34 ■^ inch. 74 83 92 101 ^ inch. 98 110 123 . 135 ^ inch. 123 138 153 169 36 ■^ inch. 69 78 87 96 ^ inch. 92 104 116 127 ^ inch. 116 130 145 159 38 ^ inch. 66 74 82 90 1^ inch. 88 99 110 121 ^ inch. 110 123 137 151 40 T^ inch. 63 70 78 86 ^ inch. 83 94 104 115 3^ inch. 104 117 130 143 42 3^ inch. 60 67 74 82 ^ inch. 79 89 99 109 ^ inch. 99 112 124 136 44 ■^ inch. 57 64 71 78 ^ inch. 76 85 95 104 ^ inch. 95 107 118 130 46 ^ inch. 54 61 68 75 ^ inch. 72 82 91 100 TS inch. 91 102 113 125 48 y\ inch. 52 59 65 72 ^ inch. 70 78 87 96 STRENGTH OF BOILERS. 157 TABLE XL VI— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER 40,000 45,000 50,000 55,000 PRESSURE. PRESSURE. PRESSURE. PRESSURE. 48 3^ inch. 87 98 109 120 50 I inch. 67 75 83 92 ■^ inch. 83 94 104 115 f inch. 100 112 125 138 52 I i-nch. 64 72 80 88 ■^ inch. 80 90 100 110 f inch. 96 108 120 132 54 *i inch. 62 69 77 85 ^ inch. 77 87 96 106 f inch. 93 104 116 127 56 ^ inch. 60 67 75 82 • ^ inch. 75 84 93 102 f inch. 89 100 112 123 58 ^ inch. 57 65 72 79 ^ inch. 72 81 90 99 f inch. 86 97 108 119 60 ^ inch. 56 63 70 77 ^ inch. 70 78 87 95 f inch. 83 94 104 115 66 ^ inch. 51 57 63 69 ^ inch. 63 71 79 87 f inch. 76 85 95 104 72 1^ inch. 46 .52 . 58 64 ■^ inch. 58 65 72 80 f inch. 69 78 87 96 158 A TREATISE ON STEAM BOILERS. The pressure given in the above and in the next tables for plates of 45,000 and 50,000 pounds tensile strength, agree closely with the best practice in this country for diameters ranging from thirty-six to forty-eight inches. Although single riveted seams may be strong enough for any pressure that may be required in any particular case, yet double riveting is to be recommended always, because the strength is increased thereby some twenty per cent; even then, it is thirty per cent below the strength of the solid plate. TABLE XLVII. SHOWfNG THE SAFE WORKING PRESSUEE FOR DOUBLE RIVETED IRON CYLINDER BOILERS, FROM TWENTY-FOUR TO SEVENTY-TWO INCHES DIAMETER, EMPLOYING A FACTOR OF SAFETY OF SIX AND ADVANC- ING THE PRODUCT TWENTY PER CENT FOR DOUBLE RIVETING. Double Riveted Iron Shells. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER. 40,000 45,000 50,000 55,000 PRESSURE. PRESSURE. PRESSURE. PRESSURE. 24 3^ inch. 125 140 156 172 \ inch. 167 187 209 229 -^ inch. 209 234 260 287 26 3^ inch. 115 130 144 158 \ inch. 155 173 192 211 3^ inch. 192 216 240 264 28 -^ inch. 107 120 134 148 \ inch. 143 161 179 197 1% inch. 179 200 223 246 30 ^ inch. 100 113 125 138 \ inch. 133 150 167 184 Y% inch. 157 187 209 229 32 Y% inch. 94 106 118 128 STRENGTH OF BOILERS. 159 TABLE XLVII— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF 1 SHELL. BOILER. 40,000 45,000 50,000 55,000 PRESSURE, PRESSURE. PRESSURE. PRESSURE. 32 I inch. 109 122 137 150 ^g- inch. 156 175 196 215 34 -^Q inch. 89 100 no 121 I inch. 118 132 148 162 ^ inch. 148 166 184 203 36 j^ inch. 83 94 104 115 \ inch. 110 125 139 152 fQ inch. 139 156 174 191 38 ^ inch. 79 89 98 108 I inch. 106 119 132 145 ^ inch. 132 148 164 181 40 3^ inch. 76 84 94 103 I inch. 100 113 125 138 ^\ inch. 125 140 156 172 42 ^ inch. 72 80 89 98 J inch. 95 107 119 131 3^ inch. 119 134 149 163 44 ^ inch. 68 77 85 94 i inch. 91 102 114 125 3^ inch. 114 128 142 156 46 3^ inch. 65 73 82 90 J inch. 86 98 109 120 -^ inch. 109 122 136 150 48 ^ inch. 62 71 78 86 ^ inch. 84 94 104 115 ^ inch. 104 118 131 144 160 A TREATISE ON STEAM BOILERS. TABLE XLVII— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER. 40,000 45,000 50,000 55,000 50 ^ inch. PRESSURE. 80 . PRESSURE. 90 PRESSURE. 100 PRESSURE. 110 ^ inch. 100 .113 .125 138 1 inch. . 120 134 150 166 52 ^ inch. 77 86 96 106 ^ inch. 96 108 120 132 f inch. 115 130 144 158 54 ^ inch. 74 83 92 102 ^ inch. 92 104 115 127 f inch. 112 125 • 139 152 56 I inch. 72 80 90 98 ^ inch. 90 101 112 122 1 inch. 107 120 134 148 58 I inch. 68 78 86 95 j^ inch. 86 97 108 119 f inch. 103 116 130 143 60 J inch. 67 76 84 -92 ^ inch. 84 94 104 114 f inch. 100 113 125 138 66 I inch. 61 68 76 83 ^ inch. 76 85. 95 104 f inch. 91 102 114 125 72 ^ inch. 55 62 70 77 3^ inch. 70 78 86 96 f inch. 83 94 104 115 STRENGTH OF BOILERS. 161 In the tables following, the writer makes a distinction between ordinary iron boiler plate and what is called in the tables "high grade iron.'*' By this is meant flange iron, and what is sometimes called fire box iron; or in other words, the very highest grades of wrought iron plates, by whatever name they may be called. For ordinary shells made of C. H. No. 1 iron, 45,000 to 50,000 pounds is as high a tensile strength as it is safe to assume without test- ing; the 40,000 pounds iron is not recommended for any service in which high pressures are to be used. It is not probable that manufacturers will have frequent calls for iron boilers from 60,000 to 70,000 pounds tensile strength. Such irons are made, however, and could be furnished if ordered. There are western river steamboats which have boilers made of iron averaging not far from 65,000 pounds tensile strength. If it is necessary to order this grade of iron for a boiler, samples should be cut from each sheet at the rolling mill, numbered or marked for test- ing before doing any work on the plate. If the samples (or coupons as they are generally called) do not come up to the required test the sheet is to be rejected. In the case of steel plates the tensile strengths should be chosen from 60,000 to 65,000 pounds tensile strength, and ought not to exceed 70,000, and in no case more than 75,000 pounds. The ordinary temper and bending tests will suffice for steel of the three first grades; for the fourth or last, in addition to these, it should be tested for elongation and contraction of area. (12) 162 A TREATISE ON STEAM BOILERS. TABLE XLVIII. SHOWING THE SAFE WORKING PRESSURE FOR SINGLE RIVETED STEEL OR HIGH GRADE WROUGHT IRON CLYLINDER BOILERS, FROM TWENTY- FOUR TO SEVENTY-TWO INCHES IN DIAMETER, EMPLOYING A FAC- TOR OF SAFETY OF SIX. Single Riveted, Steel or Wrought Iron Shells. DIAMETER L OF THICKNESS. OF TENSILE STRENGTH PER SQUARE ; INCH. BOILER. SHELL. 60,000 65,000 70,000 75,000 ""'' 24 3 1 inch. PRESSURE. 156 PRESSURE. 169 PRESSURE. 182 PRESSURE. 195 \ 1 inch. 208 226 243 260 5 inch. 260 282 304 325 26 3 1 inch. 144 156 168 180 i ^ inch. 192 208 224 240 A^ inch. 240 260 280 300 28 T6" ^ inch. 134 145 156 167 i inch. 179 193 208 223 5 inch. 223 242 260 ■ 279 30 1^^ inch. 125 135 146 156 i inch. 167 181 194 ,208 T^ inch, 208 226 243 260 32 3 TS" J inch. 117 127 137 147 i ^ inch. 156 163 182 195 5 1 T6 J inch. 195 212 228 244 34 3 1 inch. 110 119 129 138 i 1 inch. 147 159 172 184 5 T5^ inch. 184 199 214 230 36 ^^ inch. 104 113 122 130 i inch. 139 150 162 174 3^g^ inch. 174 188 203 217 STRENGTH OF BOILERS. 163 TABLE XLVIII— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER. 60,000 65,000 70,000 75,000 PRESSURE. PRESSURE. PRESSURE. PRESSURE. 38 ^^ inch. 99 107 115. 123 I inch. 132 143 154 164 -^ inch. 164 178 192 206 40. ^ inch. 94 101 109 117 1 inch. 125 135 145 156 -^^ inch. 156 169 182 195 42 y^6 inch. 89 97 104 112 I inch. 119 129 139 149 3^ inch. 149 161 174 186 44 ^ inch. 85 92 99 107 I inch. 114 123 133 142 ^ inch. 142 154 166 178 46 A inch. 82 88 95 102 ^ inch. 109 118 127 136 ^ inch. 136 147 159 170 48 Y^g inch. 78 84 91 97 ^ inch. 104 113 121 130 ^g^ inch. 130 141 152 162 50 \ inch. 100 108 116 124 ^ inch. 124 J 35 145 156 f inch. 150 162 175 188 52 I inch. 96 104 112 120 ■]^ inch. 120 130 140 150 1 inch. 144 156 168 180 164 A TREATISE ON STEAM BOILERS. TABLE XLVIII— Continued. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER. 60,000 65,000 70,000 75,000 54 i ^ inch. PRESSURE 93 PRESSURE. 100 PRESSURE. 108 PRESSURE. 116 T^6 inch. 136 3 25 135 145 1 inch. 139 150 162 174 56 i i nch. 89 96 104 111 5 Tl" inch. 111 121 130 140 3 8 inch. 134 145 156 167 58 i ^ inch. 86 93 100 108 ^ inch. 108 117 126 135 1 inch. 129 140 151 162 60 1 1 4- J inch. 83 90 97 104 5 16 inch. 104 113 121 130 1 inch. 125 135 146 156 66 4 J nch. 76 82 88 95 5 inch. 95 103 110 1J8 1 inch. J 14 123 133 142 72 i- 1 inch. 69 75 81 87 5 T6 inch. 87 94 101 108 f inch. 104 113 122 130 A word of caution may not be out of place just here in regard to using thinner plates of steel because of its higher tensile strength; for example: the substituting of a 42 X 3^ X 70,000 lbs. shell made of steel, instead of a 42 X i X 50,000 lbs. shell made of iron would not be recommended by any boiler maker who cared anything for his reputation, and the reasons are quite obvious, the principal one being STRENGTH OF BOILERS. 165 that the tightness of a riveted joint is not increased because of the increased tensile strength of the plates, and it would be a very difficult matter to keep such a boiler tight, espe- cially if of considerable length. The writer does not favor the use of plates less than one-quarter inch thick for boil- ers when exceeding thirty inches in diameter, whether of steel or iron ; neither does he recommend single riveting for boilers of any diameter when constructed of steel or of iron having these high tensile strengths. TABLE XLIX. SHOWING THE SAFE WORKING PBESSURE FOR DOUBLE RIVETED STEEL OR HIGH GRADE IRON CYLINDER BOILERS, FROM TWENTY-FOUR TO SEVENTY-TWO INCHES DIAMETER. EMPLOYING A FACTOR OF SAFETY OF SIX, AND ADVANCING THE PRODUCT TWENTY PER CENT FOR DOUBLE RIVETING. Double Riveted Iron or Steel Shells. DIAMETER THICKNESS TENSILE STRENGTH PER SQUARE INCH. OF OF SHELL. BOILER, 60,000 65,000 70,000 75,000 PRESSURE. PRESSURE. PRESSURE. PRESSURE. 24 ^ inch. 187 203 218 234 ^ inch. 250 271 292 312 ^ inch. 312 338 365 390 26 ^ inch. 173 187 202 216 I inch. 230 250 269 288 ^^ inch. 288 312 336 360 28 j\ inch. 161 174 187 200 ]- inch. 215 232 250 268 Y^e inch. 268 290 312 335 30 ^ inch. 150 162 175 187 \ inch. 200 217 233 250 3^ inch. 250 271 292 312 166 A TREATISE ON STEAM BOILERS. TABLE XLIX— Continued. DIAMETER OF THICKNESS OF SHELL. TENSILE STRENGTH PER SQUARE INCH. BOILER. 60,000 65,000 70,000 75,000 32 x% inch. PRESSURE. 140 PRESSURE. 152 PRESSURE. 164 PRESSURE. 176 i inch. 187 196 218 234 ^ inch. 234 254 274 293 34 ^ inch. 132 143 155 166 i inch. 176 191 206 221 ^^ inch. 221 239 257 276 36 ^ inch. 125 136 146 156 ^ inch. 167 180 194 209 ^ inch. 209 226 244 260 38 x% inch. 119 128 138 148 i inch. 158 172 185 197 y5^ inch. 197 214 230 247 40 T%inch 113 ]21 131 140 i inch. 150 162 174 187 ^ inch. 187 203 218 234 . 42 3^ inch. 107 116 125 134 i inch. 143 155 167 179 ^ inch. 179 193 209 223 44 T^ inch. 102 110 119 128 I inch. 137 148 160 170 ^ inch. 170 185 199 214 46 ^^ inch. 98 106 114 122 J inch. 131 142 152 163 T^ inch. 163 176 191 204 48 3^ inch. 94 101 109 116 STKENGTH OF BOILERS. 167 TABLE XLIX— Continued. DIAMETER OF THICKNESS OF SHELL. TENSILE STRENGTH PER SQUARE INCH. BOILER. 60,000 65,000 70,000 75,000 48 I inch. PRESSURE. 125 PRESSURE, 136 PRESSURE. 145 PRESSURE. 156 -fQ inch. 156 169 182 194 50 ^ inch. 120 130 139 149 ^ inch. - 149 162 174 187 1 inch. 180 194 210 226 52 I inch. 115 125 134 144 -^ inch. 144 156 168 180 f inch. 173 187 202 216 54 I inch. 112 120 130 139 -^ inch. 139 150 162 174 1 inch. 167 180 194 209 56 I inch. 107 115 125 133 -^ inch. 133 145 156 168 f inch. 161 174 187 200 58 I inch. 103 112 120 130 -^ inch. 130 140 151 162 1 inch. 155 168 181 194 60 I inch. 100 108 116 125 ^ inch. 125 136 145 156 f inch. 150 162 175 187 66 I inch. 91 98 106 114 ^ inch. 114 124 132 142 1 inch. 137 148 160 170 72 1^ inch. 83 90 97 104 ^ inch. 104 113 121 130 f inch. 125 136 146 156 168 A TREATISE ON STEAM BOILERS. Collapsing pressures — The best experimental data rela- ting to the collapsing pressures for flues or tubes are those of Sir William Fairbairn. The pressure necessary to col- lapse a flue was found to vary nearly according to the fol- lowing laws : Inversely as the length. Inversely as the diameter. Inversely as a function of the thickness, which is nearly the power whose index is 2.19 ; but which for ordinary practical purposes may be treated as sensibly equal to the square of the thickness. By these formulas the 2.19 power of the thickness multiplied by 806,300, and divided by the product of diame- ter in inches by the length in feet, is undoubtedly correct for thin flues of certain lengths. The 2 power of the thickness is also correct for another class of thicker flues. In the following tables both of these formulas are used — the 2.19 in the right hand triangle, and the 2 in the left, i^either of these formulas appear to apply to heavy flues of great lengths. This, to a certain extent, is on account of the laps acting upon the principle of Fairbairn's bands. In the tables of internal pressure, one-fifth of the value of ordinary boiler iron (say 50,000 pounds to the inch of section) is taken to be safe; while in the external one-third is taken; this is on account of the great variation in the tensile strength of iron. Note — Headings to Tables L and LI sliould read, ''Showing Safe Working Pressures against Co//ap,sg,". according to Fairbairn'.s .for-niula, etc. COLLAPSING PRESSURES. 169 ^ . ©cOL';cceowc«c^icoco«:i-toc-*c1lCC5CCi ^ ©' ci C5 00 i--^ r-^ ?o >c ■* eo c4 cn'*rfcCCCeOC0CO!N'M(NC;C^iMC-1!N->-ir-.T-i^»-iT-i,-l,-i,-iTH^ s o 00 lo CC 1ft C-. --c c »-< c; X CO X CO CO X CO © ic X X lo X CC lo c: CC -* (M (^ mxxi— cocooic;coc:oOTt o ^ bq t^ cc i6 -* 00 to 1--^ d id d 1-4 •>#' — ' 00 d -#' oi d d i^ d ^ co' -0u':iC'*-*->t-f'+C0C0C0C0C0C0C0C1C'lC^i-*©C5i--co t^-*TrcCT-ii^ccccxi— -j'lcot^r^xc^ C4 lO (>1 co' t^ o4 o oc oc o6 o c-i co' " d d CO d 00 id CO -4 d o6 d CO -^ 00 d -r CO 1 ''"' .r-OOCO-^CO — OOC-. OCOO COCC'CCcClOi.OlClC'-f-^'^'T-J'COCOCOeO 1 P3 O C^r-r-^^ — T-,-.^ 1 fa cc Si o cc T^ c i~- — -T C-. IC Lc cc CO O^t T— ■irOiCOT© — T-r-CO (-^ 00 f-;-*C0'-;OJ-^cC^r^-:t;— * -: X id co' d x'. d CO d r4 -*' o^' c 0-1 w P3 »o ^ en CO ira CO (M i-H o o C-. X cc 00 t^r-CCCOCOCCi.OIClClO'f't-*-* t-1 < w < P3 O5cccoc;: lo -^ — ' o; CO CO o c-.-co co ifXN -^ 00 r^o -r — I— X i- ;5! (MC50oot~oocor~ooic-TO-^ccoo>iic:d d -* oc CO cc -1^ c CO 1-4 -H d -M< d d CO o CO ^-4 CO r-1 1- -f i-H ci 00 CO ic ^ CO (M M — c o c; c; C-. 00 ooxi^r-t^cococo o O O CO c 'c^c^oj--r- T-r- — ^T-< — ^^.^.,-1 fa! oq it.i t- o-i Clio -< 02 r-i CO 00 CO — 0^ .o ic CO cc cc 00 = t^ cc -t CO e ="'H ooooco c; c^ CO C; ::; C-. c; oi CO -i; 1-; M c: C2 CO c; -^ -H C; -* -* c- Ifclcc ) d r4 CO -^ d d ^ d c^ d oc cc d d 1' t4 ^' id d — ' d id CO -4 -f — ^ £■ CO o -MCOt-^r-r^ti— OCClOCOCMSOlXI^t-^ClCIC^CO. CO--OJ'— .-i=C-. 05X 1 CClC-t-*CCCOCOC-)CN0>lC-ieqO^'-'r-,,-^,-l^r-l.-.r-l,-.^. < ^r-"- S o H }2; < C •—1 X H tc ^ o '^cot-xc-. o — oico-ti.occt^xc-. o.-HOieo^uocoi-xcire^-^ccxooj s < 2 ,-^^^,-r-^.-.,-l^rH(NOJC-l4'M0flC0C0C0C0COt-t z O a; Z '"' ^ 5 5 1 ^ 170 A TREATISE ON STEAM BOILERS. O I o lOOCJSiOOOOl^t- T}^c^IOpo^~«5lCr^cClc^l. t-T-«3 r^icioooiMOrfot— lOQOoocjr- oo c^ •— lO i-i J-l05i-lCOCO'-'CCSO-*00535 sr' 00 c^i c5 o -^ lo tjj— f< o CO CO Q r^ id c4 o i >OG(JOjC0050'MOOC-'r-OOtO:0 o to o o CO oj o CO t-^ "si r-^ cj o t-- -* 1-i ci t--^ ic CO OD- lO CO Cvl O 03 0> 00 t~ [t-.'jS o to.in.in.o -^ -f -*< ■* -^OJ'CcM^OfM-^-* Ot-iiCaoooto-*co ot^iococOCO— (CO ioc<>asi:^io-*(M — i-ioaiooooot^t^t^oo^oiooiooiO'Ti (N(Mr-lT-li-ft— rHT-(T-(r-l TjiO5CDCr>C0 — t^iO'^1000T-i'#OTj^ 0.130 14.126 3^ 0.119 8.357 5 0.140 17.497 3% 0.119 9.687 It will be understood that this table is to be used in connection with and is to be regarded as supplementary to table LXVI. This will be found quite useful in the re-arranging of grate surface. 248 A TREATISE ON STEAM BOILERS. \ TABLE LXX. SHOWING THE INTERNAL AREAS OF TUBES FOR THE DIAMETERS AS GIVEN BELOW, AND FOR THE NUMBER OF TUBES GIVEN IN TABLE LXVI, ON PAGE 240. THE AREAS ARE GIVEN IN SQUARE FEET. DIAME- TER DIAMETERS OF TUBES. BOILER 3 3i ^ 3f 4 41 5 36 1.10 1.14 1.10 1.35 1.22 1.18 1.22 38 1.35 1.14 1.22 1.35 1.37 1.18 1.70 40 1.44 1.68 145 1.55 1.53 1.37 1.70 42 1.90 1.88 1.86 1.75 1.91 1.96 2.19 44 2.03 1.78 186 2.02 1.91 1.96 1.94 46 1.77 1.88 1.97 1.88 1.76 2 06 1.94 48 . 2.J1 1.88 2.09 2.02 1.98 2.06 2.19 50 2.32 , 2.08 2 21 2 29 2.29 2.26 2.43 52 2.41 . 2.47 2.79 2.56 2.44 2.55 2.55 54 2.79 2.72 2.79 2.56 2.75 2.75 2.55 56 3.04 2.82 3.19 3.23 3.13 3.14 2.79 58 3.13 3.27 3.19 3.23 3.44 3.14 3.40 60 3.38 3.36 3.60 3.70 3.51 3.53 3.65 Should it be found necessary to re-arrange the grate surface from that given in the preceding tables, it is recom- mended that the lengths of grates be kept the same and diminished in width, rather than shortened. The next table shows the relation of tube to grate area. The increase shown in the line opposite forty-six inches diameter of boiler is due to the less number of tubes in the boilers, brought about by the insertion of an 8X12 man hole instead of a 6X8 hand hole, which was used in the smaller boilers. This table is to be regarded as supplementary to tables LXYIII and LXX. RELATION OF GRATE TO TUBE AREA. 249 TABLE LXXI. SHOWING THE RELATION OF GRATE AREA, AS GIVEN IN TABLE LXVIII, TO THE TUBE AREA AS GIVEN IN TABLE LXX. THIS TABLE EXPRESSES THE RATIO IN FRACTIONS OF THE GRATE SURFACE. THE VALUES WERE OBTAINED BY DIVIDING THE GRATE BY THE TUBE AREA. DIAME- TER DIAMETER OF TUBES. OF BOILER 3 13.2 H 3f 4 4^ 5 36 13.6 13.6 11.1 12.3 12.7 12.3 38 11.6 13.8 12.8 11.5 11.5 13.3 9.2 40 11.3 9.7 11.2 10.5 10.7 11.9 9.6 42 9.7 9.8 9.9 10.5 9.6 9.4 8.4 44 9.4 10.7 10.3 9.5 10.0 9.7 9.8 46 18.6 10.6 10.1 10.6 11.3 9.7 10.3 48 9.8 11.0 9.9 102 10.4 10.0 9.4 50 10.6 11.8 11.1 10.7 10.7 10.9 10.1 52 10.5 10.3 9.1 9.9 10.4 10.0 10.0 54 9.4 9.7 9.4 10.3 9.6 9.6 10.3 56 10.7 11.5 10.2 10.1 10.4 10.4 11.6 58 10.7 10.2 10.5 10.4 9.7 10.7 9.9 60 10.2 10.3 9.6 9.3 9.8 9.8 9.5 There are more tabular boilers fitted with three-inch tubes than perhaps any other size, but in some sections of the country three and a half and four-inch tubes are more common. The next two tables give the same particulars as those given in the three-inch tables. The shells, heads, man holes, etc., are in no respect different from those con- tained in table LXYII. Boilers having diameters from thirty-six to forty inches inclusive are not given, as three and a half inch tubes are not often put in boilers of such small sizes. 250 A TREATISE ON STEAM BOILERS. TABLE LXXII. SHOWING PROPORTIONS, HEATING SURFACE, WEIGHT AND HORSE POWER OF TUBULAR BOILERS FITTED WITH 3>^ INCH TUBES. SHELL. NUMBER OF. TUBES. HEATING SURFACE, f SHELL AND WHOLE QF TUBES. WEIGHT. HORSE DIAME- TER. LENCxTH. SHELL. TUBES. TOTAL. POWER AT 15 FEET, INCHES FEET. SQ. FEET. POUNDS. POUNDS. POUNDS. 42 12 32 440 2,242 1,640 3,882 29 3 14 32 514 2,506 1,914 4,420 34.3 16 32 586 2,770 2,187 4,957 39.1 .18 32 660 3,034 2,460 5,494 44.0 20 32 733 3.298 2,734 6,032 . 48.9 44 12 32 444 2,341 1,640 3,981 29.6 14 32 519 2,617 1,914 4,531 34.6 16 ,32 "592 2,893 2,187 5,080 39 5 18 32 666 3,169 2,460 5,629 44.4 20 32 740 3,445 2,734 6,179 49.3 46 12 34 470 2,465 1,743 4,208 31.S 14 34 548 2,756 2,034 4,790 37.2 16 34 626 8,047 2,324 5,371 41.7 18 34 706 3,338 , 2,615 5,953 47.1 20 34 784 3,629 2,905 6,534 52.3 48 12 36 497 2,569 1,846 4,415 33.1 14 36 579 2,871 2,153 5,024 38.6 16 36 662 3,173 2,461 ■ 5,634 44.1 18 36 745 3,476 2,768 6,244 49.7 20 36 828 3,778 3,076 6,854 55.2 50 12 38 523 3,410 1,948 5,358 34.9 14 38 609 3,811 2,273 6,084 40.6 16 38 697 4,212 2,597 6,809 46.5 18 38 784 4,613 , 2,922 7,535 52.3 20 36 871 5,014 3,247 8,261 58.1 THREE AND A HALF-INCH TUBULAR BOILERS. 251 TABLE LXXII— Continued. SHELL. HEATING WEIGHT. SURFACE, HORSE NUMBER 2 ■r> /^ ITT 17 T> -g- SHELL POWER OF AND AT DIAME- TUBES. 15 FEET. LENGTH. WHOLE OF SHELL. TUBES. TOTAL. TER. TUBES. INCHES FEET. SQ. FEET. POUNDS. POUNDS. POUNDS. 52 12 48 637 3,752 2,461 6,213 42.5 14 48 743 3.988 2,871 •6,859 49.5 16 48 849 4,404 3,281 7,685 56.6 18 48 955 4,820 3,691 8,511 63.7 20 48 1,062 5,236 4,101 9,337 70.8 54 12 48 640 3,844 2,461 ■ 6,305 42.7 14 48 747 4,276 2,871 7,147 49.8 16 48 855 4,708 3,281 7,989 57.0 18 48 • 962 5,140 3,691 8,831 64.1 20 48 1,068 5,573 4,101 9,674 71.2 56 12 55 722 4,017 2,820 6.837 48.1 14 55 843 4,464 3,289 7,753 • 56.2 16 55 962 4,911 3,759 8,670 64.1 18 55 1,083 5,358 4,229 9,587 72.2 20 55 1,203 5,805 "4,699 10,504 80.2 58 12 55 726 4,300 2,820 7,120 48.4 14 55 848 4,762 3,289 8,051 56.5 16 55 968 5,224 3,759 9,183 64.5 18 55 1,089 5,686 4,229 9,915 72.6 20 55 1,210 6,148 4,699 10,847 80.7 60 12 62 786 4,601 3,178 7,779 • 52.4 14 62 917 5,077 3,708 8,785 61.1 16 62 1,048 5,553 4,238 9,791 69.9 18 62 1,178 6,029 4,768 10,797 78.5 20 62 1,309 6,505 5,297 11,802 87.3 252 A TREATISE ON STEAM BOILERS. It is not a common thing to see tubular boilers fitted with four-inch tubes. There are some sections of the coun- try, however, where they are much in favor. Most of them are made of large diameters — that is, in the neighborhood of ^ve feet. Those who have used them speak of them in the highest terms. No doubt much of their popularity is due to the verv eflacient circulation of water in the boiler, thereby preventing priming, and all the annoyance and trouble incident to it. TABLE LXXIII. SHOWING PROPORTIONS, HEATING SURFACE, WEIGHT AND HORSE POWER OF TUBULAR BOILERS FITTED WITH FOUR-INCH TUBES. SHELL. HEATING WEIGHT. NUMBER OF TUBES. SURFACE, f SHELL AND WHOLE OF TUBES. HORSE POWER AT 15 FEET. DIAME- TER. LENGTH. SHELL. TUBES. TOTAL. INCHES FEET. SQ. FEET. POUNDS. POUNDS. POUNDS. 48 12 26 428 2,569 1,660 4,229 28.5 14 26 498 2,871 1,937 4,808 33.2 16 26 570 3,173 2,213 5,386 38.0 18 26 641 3,476 2.490 5,966 42.7 20 26 713 3,778 2,766 6,544 47.5 50 12 30 482 3,410 1.915 5 325 32.1 14 30 562 3,811 2,234 6,045 37.5 16 30 643 4,212 2,553 6,765 42.9 18 30 722 4 613 2,873 7,486 48.1 20 30 803 5,014 3,192 8,206 53.5 52 12 32 511 3,752 2,043 5,795 34.1 14 32 596 3,988 2 383 6,371 39.7 16 32 681 4,404 2,734 7,138 45.4 18 32 766 4,820 3,064 7,884 51.1 20 32 852 5,236 3,405 8,641 56.8 FOUE-INCH TUBULAR BOILERS. 25a TABLE LXXIII— Continued. SHELL. NUMBER OF TUBES. HEATING SURFACE, f SHELL AND WHOLE OF TUBES. WEIGHT. HORSE POWER AT 15 FEET. DIAME- TER. LENGTH. SHELL. TUBES. TOTAL. INCHES FEET. SQ. FEET. POUNDS. POUNDS. POUNDS. 54 12 36 565 . 3,844 2,298 6,142 37.7 14 36 660 4,276 2,681 6,957 44.0 16 36 754 4,708 3,064 7,772 50.3 18 36 849 5,140 3.447 8,587 56.6 20 36 942 5,573 3,830 9,403 62.8 56 12 41 632 4,017 2,617 6,634 42.1 14 41 738 4,464 3,054 7,518 49.2 16 41 843 4,911 3,490 8,401 56.2 18 41 949 5,358 3,92,6 9,284 63.3 20 41 1,054 5,805 4,362 10,167 70.3 58 12 45 686 4,300 2,873 7,173 45.7 14 45 802 4,762 3,352 8,114 53.5 16 45 916 5,224 3,830 9,054 61.1 18 45 1,030 5.686 4.309 9,995 68.7 20 45 1,144 6 148 4,788 . 10,936 76.3 60 12 46 704 4,601 2,936 7,537 46.9 14 46 821 5,077 3,426 8,503 54.7 16 46 939 5 553 3,916 9,469 62.6 18 46 1,055 6,029 4,405 10,434 70.3 20 46 1,172 6,505 4,894 11,399 78.1 The following evaporative test of a tubular boiler, in a flouring mill at Bellevue, Ohio, is by Mr. Holmes. The dimensions of the boiler are as follows : Diameter of boiler 60 inches. Length of boiler 15 feet. Number of four-inch tubes 51 Grate surface 26 square feet. 254 A TREATISE ON STKAM BOILERS. The coal used is an Ohio variety, known as Massillon lump. The duration of trial was ten hours. OBSERVATIONS. Total amount water weighed to boiler ...28,700 lbs. Total amount coal weighed to furnace 4,050 lbs. Total amount ash and clinker weighed dry 220 lbs. Total amount combustible 3,830 lbs. Average temperature feed water in tank 99° Average temperature gases in uptake 479° Average temperature air in fire room 93° Average pressure steam in boiler 72 lbs. Average percentage water in steam None. PERFORMANCE. Coal per hour 405 lbs. Combustible per hour 383 lbs. Water per hour 2,870 lbs. RESULTS. Pounds of water evaporated at 72 lbs. pressure and temperature of 99° per pound of coal 7 08 lbs. Equivalent evaporation from pressure of atmos- phere and temperature of 212° per pound of coal 8.15 lbs. Equivalent evaporation from pressure of atmos- phere and tempeiature of 212° per pound of combustible 8.61 lbs. This boiler was afterwards reset, and an equivalent evaporation at atmospheric pressure from and at 212° was obtained of 11.7 pounds of water, per pound of net com- bustible, or a gain of over 36 per cent. A compound tubular steam boiler, built by Mr. E. H. Ash- croft, Boston, Mass., is shown in figure — . The engraving was made from a photograph of a boiler having the follow- ing dimensions : COMPOUND TUBULAR STEAM BOILER. 255 256 A TREATISE ON STEAM BOILERS. Length of boiler 12 feet. Diameter 54 inches. 118 tubes, each 3 inches diameter 12 feet long. Diameter of steam dome 32 inches. Length of steam dome 12 feet. Heating surface.. 1,281 square feet. Horse power 85 This boiler is being well received, and results show it to be economical in fuel. Tests show an equivalent evap- oration from and at 212° of over ten pounds of water per pound of coal. The writer expected to be able to give a detailed account of the tests made to determine the capa- city and economy of this boiler, but they were not received at the time of making up this form for the press. CHAPTER XL INTERNALLY FIRED BOILERS. The Cornish Boiler — The Lancashire Boiler — The Fairbairn Boiler — The Galloway Boiler — Vertical Flue Boilers — The Shapley Boiler — The Baxter Boiler — Vertical Tubular Boilers — Snyder's Vertical Boiler — Flynn's Vertical Boiler — Suiter's Boiler — Portable Boilers — Semi-Portable Boilers — Locomotive Boilers. The internally fired boilers, in common use in this coun- try, are either vertical flue or tubular boilers, or of the locomotive type. In Europe, horizontal boilers, fitted with internal flues, are very common and of a type rarely seen in this country. In England, the internally fired boilers are usually of the Cornish or the Lancashire varieties ; there has been a growing dislike to externally fired boilers in that country for many years, during which time the above named boilers have been growing in favor, and are now so thoroughly intrenched behind public opinion, that it would require a remarkably good showing in economy and durability in a rival to gain similar popularity. The Cornish boiler owes its name to the circumstance of its having been first introduced in Cornwall, England. The original inhabitants of Cornwall were Celts, speaking the Cornish language, which, though now extinct, or no longer spoken by the people, the name of Cornish still lives, and is applied to many technical names — for example, Cornish mining, Cornish agriculture, etc. ; hence, Cornish boilers, meaning thereby a particular kind of boiler orig- inally in or peculiar to Cornwall. (18) 258 A TREATISE ON STEAM BOILERS. This boiler was introduced early in the present centuryj by Richard Trevithick, an English engineer, born in Corni wall, and one whose name is inseparably connected with' the modern steam engine. This boiler consists of a horizontal cylindric shell, with flat ends and fitted wdth one large flue passing through from front to back of the boiler and securely fastened to ■ the two ends by riveted joints. This large flue contains the grate on which the fuel is burned, the products of com- bustion passing through the flue to the back end of the boiler, returning, by a suitable arrangement of the brick work along the sides of the boiler to near the front end, thence downward and along the bottom of the boiler to the rear end, and from thence to the chimney. This ar- rangement of exterior flues is shown in figure 62. A Cornish boiler is to be distinguished from a single flue boiler in its having the furnace arranged in the flue, and thus being an internally fired boiler. The usual course of heated gases in any arrangement of boiler and furnace is from below upwards. It was first shown by Peclet and is now generally recognized, that a great advantage in point of thorough convection of heat and consequently in economy of fuel, is gained by causing the course of the hot gas to be on the whole from above downioards ; because then, the hottest strata of the furnace gas, being uppermost, spread themselves out above the denser and colder strata which are below, and so diff'use themselves more uniformly throughout all the passages than they do when made to ascend from below.* Figure 62. "• Rankine. CORNISH BOILERS. 259 It would naturally be inferred from an inspection of the engraving that the heating surface is of the best possible arrangement to insure economy of fuel. The feed water enters the boiler near the bottom, where the water is cool- est ; as it rises becomes more highly heated until the sur- face is reached, where the steam is given off. There is ample facility for circulation, and the conditions are favor- able to rapid evaporation. The large water surface lessens the tendency to priming and thus practically insures dry steam. Under the most favorable conditions, as regards con- struction, fuel used and rate of combustion, about eight pounds of water may be evaporated per pound of coal. In this respect the Cornish boiler is about on an equality with our ordinary cylinder boiler. The rate of combus- tion in Cornish boilers is not far from ten pounds of coal per square foot of grate, in good ordinary firing, and from this down to ^ve or six pounds in slow firing. These boilers must, of necessity, be of large diameter when required to be of any considerable power. This also necessitates a large flue, in order to afford the necessary grate area and heating surface. Increasing the diameter of the fine decreases its power to resist collapse, whieh may occur either by overpressure or overheating. These flues are often strengthened by means of heavy wrought iron rings, which are secured to the ends w^liere the sec- tions are to be joined to form a continuous flue. Figure 63 represents such a joint. This ring of angle iron gives the flue great stiff- ness, and increases its power to resist collapse by shortening the length of the figure es. span between supports. A much better device is that of Mr. Adamson, shown in figure 64. The flue is made in sections, with welded seams and flanged ends, which are 260 A TREATISE ON STEAM BOILERS. secured end to end by riveted joints, as shown, render- ing collapse almost impossible; it is, also, a very superior expansion joint, thus preventing the shell being strained, as is often the case with plain flues, and presents a further Figure 64. • i i i i advantage m that both the rivets and edges of the plates are kept entirely free from the action of the fire. When a fire is started in any internally fired boiler, having a large flue, such as the Cornish or Lancashire, the flue will be heated first, and will expand in length a con- siderable distance before the external plates, or the shell of the boiler, has received any considerable degree of heat. Unless some provision is made for this unequal expansion, it is likely to lead to a great deal of annoyance by leaky joints, if not to something more serious in the way of rup- ture. The following table gives the principal dimensions of Cornish boilers, as taken from the catalogue of Abbott & Co., IN'ewark-upon-Trent, England: The Lancashire boiler is an internally fired boiler, and differs from the Cornish in having two flues and furnaces instead of one. It was introduced, in 1844, by Fairbairn and Hetherington in Manchester, and is, therefore, called a Lancashire boiler. The insertion of two smaller flues in the shell of a boiler, instead of one large one, was to strengthen the boiler against collapse. This is, perhaps, the most popular for large boilers of any in England to-day. Figure 65 represents a longitudinal section, and figure 66 a cross section, of a Lancashire boiler. CORNISH BOILERS. 261 TABLE LXXIV. SHOWING THE PRINCIPAL DIMENSIONS OF "NEWARK" STANDARD SIZE CORNISH BOILERS. SHELL. HORSE POWKR. 1 DIAME- TER. LENGTH FT. IN. FT. IN. 2 2 9 6 4 3 9 7 6 6 4 3 10 8 4 4 13 lO 4 6 14 12 4 8 15 14 4 9 16 16 4 9 17 6 " 18 5 18 DIAME- TER OF FLUE.- FT. IN. 1 3 2 2 2 2 2 2 2 2 4 2 4 2 6 2 9 DOME. DIAME- TER. FT. IN, 1 1 3 1 6 1 6 1 9 1 9 1 9 1 9 2 FT. IN. 1 1 6 1 9 1 9 2 2 2 2 6 2 6 THICKNESS AND QU iLITY OF PLATES. APPROX- WEIGHT SHELL ENDS. FLUE. DOME. INCH. INCH. INCH. INCH. POUNDS 5 T? 16^ ^ % BB % B 5 BB 1,900 % B 3^ BB % B % BB 3,248 ■% B >^ BB Ys B % BB 4,480 %B ^ BB Ys B % BB 5,824 %B y^ BB % B % BB 6,720 % B K BB % B % BB 7,280 % B % BB % B % BB 7,840 % B ^ BB % B % BB 8,512 % B 3^ BB T^ B % BB 10,080 For properties of B and BB iron i)lates, see pages S3-4. Figure 65. These engravings represent the arrangement and set- ting of the boiler at the trials at Mulhouse, and do not contain the welded and flanged flues, as shown in detail at -In every case, Low Moor, or equal quality plates, are put over the fire in the flues. —Abbott & Co. 262 A TREATISE ON STEAM BOILERS. figure 64. The principal dimensions of the boiler used in the trial are as follows : * Shell, 6 feet 6f inches diameter by 25 feet 9 inches long; two flues, each 2 'feet S^-q inches diameter. The shell plates were 0.63 inch thick, the end plates 0.748 inch thick, and the flue plates 0.51 inch. The combined width of the fire grates is 4 feet 6f inches, and their length 5 feet 1 inch ; this length includes 6-^ inches formed by the ^^^» parts of the bars resting on iron Figure 66. supports. Taking the effective length of the grates, therefore, at 4 feet 6^ inches, we get a fire grate area of 20.5 square feet. TABLE LXXV. GIVING AN ABSTRACT OF RESULTS OBTAINED IN EVAPORATIVE TESTS WITH THE LANCASHIRE AND FAIRBAIRN BOILERS, AND USING SAARBRUCK COAL.f [MULHOUSE EXPERIMENTS]. Coal consumed per day of eleven hours Net fuel consumed per day of eleven hours Water evaporated per day of eleven hours Equivalent evaporation from and at 212° Actual evaporation per pound of coal Actual evaporation per pound of net fuel Equivalent evaporation from and at 212° per pound of coal Equivalent evaporation from and at 212° per pound of net fuel- Equivalent evaporation from and at 212° per square foot of heating surface per hour Mean temperature of escaping gases Weight of air supplied per pound of coal burnt., LANCASHIRE BOILER. 3,628 lbs. 3,261 lbs. 24,178 lbs. 28,247 lbs. 6.66 lbs. 7.41 lbs. 7.79 lbs. 8.66 lbs. 4.19 lbs. 555° 13.99 lbs. FAIRBAIRN BOILER. 3,648 lbs. 3,263 lbs. 25,852 lb 30,187 lbs. 7.09 lbs. 7.92 lbs. 8.27 lbs. 9.25 lbs. 2.70 lbs. 322° 14.98 lbs. * Engineering. t For analysis and calorific value of Saarbruck coal, see Combustion of Coal, page 187. LANCASHIRE BOILERS. 263 The total heating surface of the boilers is 612.5 square feet, divided as follows : SQUARE FEET. Surface of outer shell exposed in side and bottom flues 271.03 Surface of internal flues, deducting parts below the grates , 333.25 Surface at back end of boiler 8.22 Total 612.50. The next table gives the sizes of Lancashire boilers, as taken from the catalogue of Abbott & Co,, for the first three boilers in the table; the two following are from Tangje Brothers & Holman, London, England : TABLE LXXVI. LANCASHIKE BOILERS. SHELL. DIAME- TER OF FLUE.* DOME. THICKNESS AND QUALITY OF PLATES. APPROX- HORSE IMATE POWER. . WEIGHT. DIAM. LENGTH DIAM. HEIGHT SHFLL ENDS. FLUE. DOME. FT. IN. ft: IN. FT. IN. FT. IN. FT. IN INCH. INCH. INCH. INCH. POUNDS. 20 5 9 19 2 2 3 2 6 l^B ^ BB Ys B % BB 13,440 25 5 6 25 2 2 3 2 6 l^B y^ BB %B % BB 17,024 30 6 26 2 .4 2 6 3 ^B y^ BB %B T^BB 20,720 35 fi q R1 2 ly^ 2 9 R 4 1 32,480 36,400 40 7 34 3 1 4 3 The following test of a Lancashire boiler shows the evaporative power, rate of combustion, and much other data of interest : Boiler trials at the South Metropolitan Gas Works, Lon- don, England, December 19-21 and January 2-4, 1877-8 : 264 A TKEATISE ON STEAM BOILERS. Diameter of boiler 6 feet 6 inches. Length of boiler 25 feet Cinches. Diameter of each furnace (2) 2 feet 3 inches. Grrate surface (whole) 27.75 square feet. Grate surface (as bricked up) : 16 square feet. Total heating surface 679 square feet. Heating surface, deducting lower half of furnaces 504 square feet. Proportion of air space through bars to total grate surface 0.2 to 1. This boiler was set in the usual manner. The products of combustion, after passing to the end of the boiler, returned to near the front end, thence downward to the lower flue and along the bottom to the rear end of the boiler, and thence to the chimney. The furnaces were fitted with rocking bars, the bars being zigzag instead of straight. The rocking arrange- ment could not be 'used on the days in which the grates were bricked up. EVAPORATIVE TESTS. ( GRATE. WHOLE. PARTIAL. Duration of trial, hours 8.50 11.75 Temperature of feed water , 50.3° 49.1° WATER FED INTO THE BOILER. Per hour, in pounds 1,228 1,288 Per square foot of total heating surface, pounds 1.81 1.89 Per pound of coal, pounds (total) 6.88 9.98 Equivalent evaporation from and at 212° pounds 8.0 11.4 Equivalent evaporation from and at 212°, per pound of combustible 8.5 11.7 FUEL USED. Cefn (South Welsh) coal in both tests ; percentage of non-combustible and of water in the fuel 6.0 3.0 Total fuel per hour, pounds 178.7 129.0 Total fuel per square foot of grate sur- face, pounds 6.4 8.1 THK FAIRBAFRN EOILER. 265 GRATK. WHOLK. PARTIAIi. Total fuel per indicated horse power, pounds 3.64 --, T^. „ ^ THeavy and black, Condition ot fire -{ about 8 inches (, thick. Cost of evaporating one gallon of water, cents 0.31 Factor of evaporation 1.16 Rate of transmission of heat (thermal units per square foot of total heat- ing surface per minute), in heat units. 33.7 Steam pressure in boiler, above atmos- phere, pounds 52 5 Indicated horsepower 49.1 Barometer 29.91 Temperature of the air in boiler house.. 47.0° Grate surface, square feet 27.75 Total heating surface, square feet 679 2.79 Lifi^ht and brig:ht,. about 5 inches thick. 0.21 1.14 34.7 55.9 46.2 30.37 44.5° 16.0 679 The Fairhairn boiler, shown in figures 67 and 68, is a modification of the Lancashire boiler, and might be said to ^^»^^^^^^'.M^fc^.#/M^^M^:^S;M€?//f^/^e^^ Figure 67. be an internally fired elephant boiler. It consists of three cylindric shells, two of these — each traversed concentrically by an internal flue — being placed side by side at a short 266 A TREATISE ON STEAM BOILERS. FlGUKK 68. distance apart, while the third is placed above and between them, being joined to them by suitable connecting tubes. The boiler shown in the above engraving is a representation of the one used in the experi- ments at Mulhouse, and while possessing all the salient points of the Fairbairn boiler, is slightly modified in design, the two flues in the lower shells being placed eccentrically as shown. * These two lower shells are each 4 feet 1^ inch diameter by 25 feet 9 inches long, and the flues they contain are 2 feet 3^^ inches diam- eter. The lower cylinders are each connected by three tubes or mouth pieces with the upper cylinder, which is 3 feet 8^ inches in diameter by 22 feet llf inc?ies long. The upper cylinder is made of plates J inch, and the two lower of plates 0.53 inch thick, while the internal flues are made of 0.51 inch and the ends of 0.71 inch plates. The grates, which are contained in the internal flues of the lower cylinder, are precisely identical with those of the Lancashire boiler already described. This boiler being one of the three experimented with by the Societe Alsaci- enne de Constructions Mechaniques, Mulhouse. These were the French or Elephant boiler, described on page 223, the Lancashire boiler, described in this chapter, and the Fair- bairn boiler. The heating surface of the latter is 1017.48 square feet, divided as follows : * Engineering. THE GALLOWAY BOILER. 267 SQUARE FEET. Surface exposed by the upper cylinder 144.88 Surface exposed by the two lower cylinders to the second "run" of gases 314.49 Surface exposed by the two lower cylinders to the third "run" of the gases 182.96 Surface exposed by six connecting tubes 34.04 Surface exposed by two internal flues, deducting surface below grates 333.27 Surface exposed at front of upper cylinder 3.84 Total 1.017.48 It will be seen, on reference to the engraving of tliis boiler, that the setting is so arranged that, on leaving the internal flues of the lower shells, the gases return to the front end along the sides and bottoms of the two lower cylinders, and thence pass to the chimney between these cylinders and the third one above, a mid-feather wall dividing the flues so that the products of combustion from the two furnaces do not unite until just before entering the chimney. An abstract of results obtained in the tests at Mul- house may be found in table LXXV, in comparison with that of the Lancashire boiler. The Galloway boiler, shown in figure 69, is a modification of the Lancashire boiler, in which the two furnaces at the front end unite in one back flue of an irreg- ular oval form. This flue consti- tutes the chief feature in the "Gal- loway boiler," in which are placed conical water tubes, fixed in an up- right position, in such a way as to support the flue and to intercept and break up the flame and heated gases, when passing from the fire grate or furnaces to the chimney. Along the sides of the flues there are also placed several wrought FlGUKE G9. 268 A TREATISE ON STEAM BOILERS. iron stops or bafflers, which deflect the currents of heated air and cause them to impinge against the tubes, so as to absorb all the available heat possible. The conical water pipes, or " Galloway tubes," as they are now generally called, present a direct heating surface to the action of the flame or heated gases, and thus eft'ects a great saving in fael ; they also promote rapid circulation of water and thereby maintain that uniform temperature which is so essential to the durability and safety of all steam boilers. Unequal expansion or contraction is avoided and its attendant evils, undue strains and event- ual rupture. Messrs. W. & J. Galloway exhibited three of their boil- ers, in the British section, at the Centennial exhibition, in which an important improvement over their former designs was shown. This improvement consists in the arching ot the bottom part of the oval back flue, by means of which greater facilities arc fornished for cleaning and examining the lower part of the boiler when required. A further advantage is also obtained by having the conical tubes all radiating from one center, as shown in the engraving; they are consequently one uniform length and are inter- changeable. These boilers were each seven feet diameter by twenty- eight feet long. The shell was of Bessemer steel plates, three-eighths inch thick, with double riveted longitudinal seams. The two furnaces were each two feet nine and a half inches diameter by seven feet six inches long, made of steel plates, in three rings, flanged and riveted together, as already described on page 260. The main flue contained in it thirty-three conical water tubes, each ten and a half inches diameter at the top, or large end, and Rve and a half inches diameter at the lower end. These tubes are welded and flanged from one plate, and thus present no joints VERTICAL BOILERS. 269 other than the flange joints by which they are attached to the flue. The following data shows the evaporating power of this boiler, as determined at the Centennial exhibition; one trial using anthracite, and the other trial using bitu- minous coal. Two regular trials were made with each kind of coal — one for economy, the other for capacity : ANTHRACITE. BITUMINOUS. Pressure of steam above atmosphere.. 70.06 70.12 Temperature of steam, average 310° 310° Temperature of uptake, average 303° 324.6° Temperature of feed water, average... 56° 55° , Coal consumed per square foot of grate per hour 8.87 lbs. 7.27 lbs. Water evaporated per pound of coal.. 8.51 lbs. 9.18 lbs. Water evaporated per pound of com- bustible 9.58 lbs. 10.07 lbs. Water evaporated per hour 2,946 lbs. 2,603 lbs. Water evaporated per square foot of heating surface per hour 3.03 lbs. 2.67 lbs. Percentage of moisture in the steam... 0.22 0.57 Number of pounds of saturated steam evaporated at 70 lbs. from 212°, equivalent to total heat units de- rived from the fuel — Per pound of coal 9.94 10.69 Per pound of combustible 11.19 11.72 Per square foot of heating surface 3.53 3.11 Horse power, at 121^ square feet 77.88 77.88 Horse power, on the basis of 30 lbs. of water actually evaporated per hour, per horse power.. 98.19 86.77 Vertical boilers — There has been a very great demand in this country, within a few years past, for small internally fired vertical boilers. These are used for furnishing steam 270 A TREATISE ON STEAM BOILERS. for small engines, pumps, heating, etc. The simplest form of these boilers is shown in fig- ure 70. When used for heating, where pressures are only five to ten pounds, it is not usually the practice to put in stay bolts or braces; but when used for fur- nishing steam for small engines or pumps, or for any purpose where the pressures may be anywhere from fifty to seventy-five pounds, the stay bolts and braces should always be put in. Sometimes a ring is put in between the outer shell and the fire box at the bot- tom, and at other times the fire box is flanged, as shown in the engraving. The writer has made iiU them both ways, but prefers the Figure 70. latter. Each boiler should have one or naore hand holes just above the crown sheet, and at least three in the bottom, as shown in the engraving. These band holes are quite essential to inspection and cleaning and should never be omitted. The writer once saw a boiler of this description, in which the space between the fire box and the shell had completely filled with scale^ and if the boiler had been used for any purpose in which it would have been necessary to use even a moderate pres- sure, disastrous results must have certainly followed. Whether the blame could attach to the owner or not, it cer- tainly could to the boiler maker, who was guilty of little less than criminal negligence in not putting them in. The wri- ter has seen what might be called a clever trick in evading this known duty in boiler construction in order to save a few dollars — that is, by the insertion of two or three one VERTICAL FLUE BOILERS. 271 inch or one inch and a quarter pipe plugs. This is not sufficient, and nothing less than a 2X3 hand hole should ever be used, even in the smallest boilers, and as much larger as the circumstances will permit. The ring around the fire door opening should be pre- ferably of wrought iron, though cast iron is often used. If a ring is to be used at the bottom of the boiler instead of fianging the fire box, as shown in the engraving, it should be of wrought iron alioays. The following table gives the principal dimensions of vertical fiue boilers. The shell being \ inch thick in all cases, the fire box -^ inch thick, the outside heads f inch for all sizes up to 44 inches, inclusive, and -^ inch thick for larger diameters. The inside heads are ^ inch thick up to 40 inches, and f inch for the other sizes. TABLE LXXVII. PUOPORTIONS AND WEIGHTS OF VERTICAL FLUE BOILERS AS SHOWN IN FIGURE 70. SHELL. FIRE BOX. FLUE. GRATE AREA. HEATING SURFACE. HORSE POWER AT 9 FEET. WEIGH T. DIAM. HEIGHT DIAM. HEIGHT DIAM. AREA. INCHES INCHES INCHES INCHES INCHES SQ. FT. SQ. FEET. SQ. FEET. POUNDS. 30 60 25 83 9 .44 3.41 24.0 2.7 1,087 32 06 27 36 9 .44 3.98 28.9 3.2 1,236 3i 72 29 39 10 .55 4.59 34.0 3.8 1,409 36 78 31 42 11 .66 5.24 39.6 4.4 1,585 38 84 32 45 11 .66 5.59 43.6 4.8 1,777 40 90 34 48 12 .79 6.31 49.9 5.5 2,012 42 96 36 51 12 .79 7.07 55.8 6.2 2,245 44 102 38 5i 13 .92 7.88 62.9 7.0 2,473 46 108 39 57 14 1.07 8.30 68.8 7.6 2,786 48 114 41 «0 15 1.23 9.17 76.6 8.5 3,036 These weights do not include either the grate or hase, but do include stays and braces 272 A TREATISE ON STEAM BOILERS. The diameters of the outer shell and for the fire hox, a» given in the above table, are inside measure. The height of the boiler is that from the bottom of the lower joint to the top of the upper head. In regard to this height, if it is found to be inconveniently high, it may be lowered 12 to 18 inches for the lower half of the table without inter- fering with the heating surface or decreasing the boiler power. The height of the fire box is from the bottom of the boiler to the lower side of the head. The water space for boilers 30 to 36 inches, inclusive, is 2^ inches ; from 38 to 44 inches it is 2^1 inches, and from 46 to 48 inches it is 3^^ inches. The diameter of the flue is inside measure. In boilers of this class the fire box is the main thins^ as a matter of course, and ought to be large and roomy. If it were thought advisable, the heights given in the table might be reasonably extended. In the examples given above, the fire boxes have parallel sides ; if they were inclined, as shown in figure 71, it would improve the cir- culation, and add but very little, if anything, to the cost. The Shapley boiler, as made by the Knowles Steam Pump Works, is shown in sectional elevation in figure 71. This boiler is made in two sections, the lower section containing the greater part of fire box and the vertical tubes ; the latter are situated between the fire box and out- side shell, having their lower terminus in two base flues, extending from either side of ash pit entrance to smoke stack at the rear of the boiler. The upper section is prin- cipally a reservoir for steam. The fire box extends a short distance into the upper section, and the products of com- bustion are conveyed through cross tubes, to the vertical tubes as indicated by the arrows, thence downward to the base flues and so to the chimney. The tubes and crown sheet are removed as far from the intense heat of the fire as the size of the boiler will permit ; this also insures a large com- THE SJIAPLEY BOILEK. 273 bustion chamber, a thing which is alwa3^s to be secured^in internally fired boilers whenever possible. The tubes are (19) 274 A TREATISE ON STEAM BOILERS. well protected from the action of the fire, and are quite accessible in case they need repairs. The Baxter boiler — The boiler furnished with the well known Baxter engine, as built by the Colt's Patent Fire Arms Manufacturing Company, is shown in figure 72, which is a representation of the boilers regularly furnished with their engines, with the single exception of their small- est size, or two horse power, which is illustrated in figure 74. Referring to figure 72, it will be seen that all the Zi-xrAAAiJi^ Figure 72. heating surfaces are below the water line. The combus- tion chamber is large, and of a form to insure economy of fuel. The fire box is provided with descending flues. THE BAXTER BOILER. 275 passing through the water space and communicating with a jacket surrounding the water space and extending up to the water line of the boiler, so as to leave the dome uncovered, and to which the engine is at- tached, as shown. Figure 73 is a horizontal section showing the arrangement of the. descending Hues, the furnace . door, the grates, and the jacket surrounding the boiler containing the figure 73. heated products of combustion on their way to the chim- ney. The design of this boiler is such as to insure a pro- per circulation of water, hence there is little or iio danger of priming. The smallest size, or two horse power, is shown in sec- tional elevation in figure 74, and differs from the one al- ready described in having no descending flues, as shown in the other engravings, having instead an internal chamber, or fire box, with an annular heating chamber between it and the inside of the boiler. Vertical tubular boilers-;— The commonest form of a vertical tubular boiler is shown in fig- ure 75. It does not differ from the vertical fine boiler, already Figure 74. described, except in having tubes instead of fines above 276 A TREATISE ON STEAM BOILERS. the furnace. This is the form of boiler usually supplied with the numerous small vertical engines now offered in the market. When "properly made, it is an economical boiler, and with proper management will be found to be quite durable in service. The tubes in vertical boilers, especially if short ones are employed, should not be of large diameter. The diam- eters usually employed are two, two and a half and three inches. The number of tubes may be such that their aggregate area shall equal one-eighth of the grate area. The following table (LXXYIII) gives the principal proportions of vertical tubular boilers having the same size of hre box as given in table LXXVII, for vertical flue boilers. This is higher than vertical fire boxes are usually made for the diameters given. The writer" attaches so much more importance to fire box heating surface than to tube surface that he recommends high fire boxes rather than long tubes, especially as the heating surface proper is that only to the water line and not to the upper limit of the tubes. The height of water carried above the crown sheet in vertical boilers is scarcely ever more than twelve inches; the value of the tube surface may be easily over estim- ated by not taking into account the comparatively small portion of the whole surface actually utilized. I limiiiiMim. „ FiGUllK VERTICAL TUBULAR BOILERS. 277 TABLE LXXVIII. PKOPORTIONS AND WEIGHTS OF VERTICAL TUBULAR BOILERS, AS SHOWN IN FIGURE 75. SHELL. FIRE BOX. TUBES. HORSE GRATE AREA. HEATING SUKFACE POWER AT WEIGHT. DIAM. HEIGHT DIAM. HEIGHT NO. DIAM. 12 FT." INCHES INCHES INCHES INCHES INCHES SQ. FEET. SQ. FEET. POUNDS. 30 60 25 33 36 2 3.41 61.2 5.1 1,163 32 66 27 36 42 2 3.98 77.7 6.5 1,367 34 72 29 39 48 2 4.59 95.5 8.0 1,583 36 78 31 42 55 2 5.24 116.7 9.7 1,825 38 84 32 45 36 2)^ 5.59 109.9 9.2 1,995 40 90 34 48 42 2% 6.31 134.5 11.2 2,301 42 96 . 36 51 58 2>^ 7.07 1G1.4 13.4 2,604 44 102 38 54 53 2>^ 7.88 187.7 15.6 2,941 46 108 39 57 54 2)^ 8.30 . 202.7 16.9 3,195 48 114 41 60 60 23^ 9.17 235.4 19.6 3,611 50 120 43 60 66 2>^ 10.08 277.7 23.1 4,000 52 120 45 60 72 2% 11.04 290.0 24.2 4,225 54 120 46 60 51 3 11.54 266.6 22.2 4,291 56 120 48 60 55 3 12.57 286.0 23 8 4,518 58 120 50 60 60 3 13.64 309.3 25.8 4,815 60 120 52 60 66 3 14.75 335.6 28.0 5,083 These weights do not include the grates, base or fit- tings of any kind, but do include hand hole plates, stays, braces, etc. ■ In horizontal tubular boilers the grate area may be made of any size best suited to the fuel to be used and the quantity to be burned. In vertical tubular boilers the grate area is fixed by the diameter of the fire box, and the fuel must be selected with reference to the most economi- cal consumption. Anthracite nut coal or crushed coke ■■' If the ordinary number of tubes are put in the head, then use 15 as a divisor. 278 A TREATISE ON STEAM BOILERS. will, in general, be found to give the best results when burned in vertical boilers than if bituminous coal is used, unless the latter is verj slowly burned and sparingly fired. It should be broken up into small pieces not larger than a hickory nut, or about the size of anthracite nut. One of the inconsistencies in rating boiler power by total heating surface is shown by a comparison of the thirty-six and thirty- eight inch boilers in the above table. The fire box in the latter boiler is one inch larger in diam- eter and three inches higher; the tube area in proportion to the grate is practically the same, yet the larger boiler rates nearly a half horse power less than the smaller one. In comparing the above table with almost any manu- facturer's published list of vertical boilers, the first noticea- ble thing which will attract the reader's attention will be, doubtless, the small number of tubes for the diameters given. The writer has before him three lists of this kind — all of them, as manufactures, are of very high standing — two of them American and one English. The number of tubes called for in the above table, for a 48 inch boiler =: 60, 2J inches in diameter; one of the American lists for the same diameter has 97, 2J inch; the other 88, 2J inch tubes. The English list has 30, 2J inch tubes. There is probably no heating surface of so little value as the tubes in a vertical boiler ; from half to two-thirds of their length is in the steam space and thus performs no useful service in evaporating water. The value of the remaining half or one-third, as the case may be, is in con- tact with the water, but, on account of their position with reference to the furnace, and thus presenting no surfaces against which the heated gases can impinge, it is to be regarded as heating surface of the very lowest order. The most effective heating surface in boilers of this class is that of the fire box; and the tube area should not greatly exceed that necessary for draft, merely. It is bet- VERTICAL TUBULAR BOILERS. 279 terto have a large fire box and few tubes, than a small fire box and many tubes. In the table given above, the tube area is fully fifty per cent greater than that necessary for draft, so that, the number of tubes given in the table ought not to be exceeded ; deducting one-third from the tabular numbers will give the smallest number of tubes admissible, Avhich is, one-eighth of the grate area. Between these two limits may be considered good if not common practice. In England, the tube area for vertical boilers is fixed by the grate area ; in this country, no account is taken of the grate area, but as many tubes are placed in the head as it will contain. By the latter method it is easy to figure large powers, but it would be a gain to manufacturers and users alike to leave out the surplus tubes and employ a smaller divisor for the rating. It is efliciency and not extent of heating surface that is needed. The upper end of the vertical tubes, as shown in figure 75, are liable to waste away by being continually heated, and in time will often prove very troublesome. Much of this is due, perhaps, to the too rapid firing before the steam is on the boiler. Many cases of this kind have come within the observation of the wri- ter, and some very curious phe- nomena, in connection with the wasting of the upper end of the tubes, have been disclosed. To obviate any difficulty of this kind the tubes may be shortened and the products of combustion pass up into a receiving chamber, from Fl&LKE 76 280 A TREATISE ON STEAM BOILERS. which they may then pass into the chimney, as shown in the sectional elevation in figure 76. Bj this arrangement in the design of a boiler the tubes are wholly protected by the water, and will outlast those, as shown in the boiler on page 276. The length of the tubes should be such that at least six inches of water should be above them in ordi- nary steaming. The upper chamber must have depth enough to be able to use an expander in setting the upper ends of the tubes. This upper chamber contracts the steam space and largely reduces the water surface. The engrav- ing does not show the best proportions for a boiler of this kind for rapid steaming; the tubes being too numerous and too long, the upper chamber too large in diameter in proportion to that of the boiler; still it conveys the idea. In cases where these boilers have been used for heating, they have given satisfaction and are well liked. A modification of the above design is given in figure 77, in which it will be observed that the upper chamber is conical instead of parallel, as in the boiler just described. This design is that of the boilers furnished by the Niles Tool Works, Hamilton, Ohio, with their small engines, from two to twelve horse power. The following table is compiled from their practice. This is an excellent form of boiler and is capable of yielding good evaporative results. Figure 77. 282 A TREATISE ON STEAM BOILERS. TABLE LXXIX. PROPOriTIONS OF VERTICAL TUBULAR BOILERS, BY THE NILES TOOL WORKS. SHELL. TUBES. HORSE POWER. DIAMETER. HEIGHT. NUMBER. DIAMETER. 2 24 52 18 2i 4 28 62 27 2J 6 30 66 37 2J 8 33 71 42 2^ 10 36 77 55 2J 12 42 80 69 2* Each horse power in this table is based on lo square feet of lieating surface. ■ The commonest and at the same time the worst fault of small vertical tubular boilers is that of priming. There is no doubt that much of this trouble is due to having too many tubes in the boiler, which may and often does have the effect to retard the circulation immediately over the crown sheet. Priming may be induced through other causes, such as bad feed water, sudden reductions in pressure, etc. "Whatever may be the cause it is a troublesome and dan- gerous occurrence, and one which needs to be overcome at any cost. The i^ew York Safety Steam Power Compa'ny introduce in their vertical tubular boilers a baffle plate through which all the tubes pass at about the water level. A vertical section of tfieir- boiler illustrating this detail of construction is shown in ffgure 78. A large tube hangs from the center of this plate nearly to the crown of the furnace and an annular space is left around the outside of the baffle and between it and the circulator sufficient for the easy escape of the steam and SNYDER S VERTICAL BOILER. 28B water. . The effect of this arrangement is to stop the current of steam and water tending to shoot up between the tubes, and compel it to flow outward and escape be- tween the baffle and cir- culator, at which point the steam and water separates, most of the water flowing over the circulator, as before de- scribed, while the re- mainder of the water falls on the top of the baf3.e plate and flows through the tube in its center, thus keeping up a constant current over the center of the crown sheet and among the tubes. It will be ob- served ■ that the steam is taken off from the very center of the boiler, and as the steam is delivered at the outer edge of the baffle it must flow inward between and around the tubes on its way to the engine and become dried and slightly super heated. This improved arrangement not only secures thorough circulation and dry steam, but by its use the operator is enabled to keep as much of the fire surface wetted as he may wish, by simply locating the baffle at the desired point. \.v.Roi3Eins .sc.w.yr Figure 78. Snyder's Vertical Boiler— A novel design for small boil- ers is shown in figure 79. It is not an internally fired boiler, and does not properly belong to this chapter. *It i 28i A TREATISE ON STEAM BOILERS. is manufactured by Mr. Ward B. Snyder, IS^ew York city, to supply a popular demand for a small and low priced steam motor. Figure 79 is a sectional view of the boiler. The letters in the cut indicate spaces as follows : A, dome top or smoke bonnet; B, steam space; C, water space; D, furnace or fire box ; E, ash pit. This boiler consists of a-j-^g-- inch wrought iron lap welded I" cylinder, with the heads fitted, as shown in the engraving. A tubular stay rod, which also acts as a ^ue, is secured to the two heads. The engraving shows but one tubular stay ; others may be added if thought neces- sary. For steam yacht boilers the makers recommend from five to ten of these tubes, ac- cording to size of main boiler, which serves to keep the main body of water steady, in case of the rolling of the boat. A number of side tubes are fitted to the shell of this generator, as shown in elevation in figure 80, through which there is a free circulation, throwing continuously a stream of mixed water and steam upon the surface of the water in main boiler, the steam ascending and the water descending, as indicated by the arrow points, while outside and around these tubes there is a free circulation of the heat and abundant room for the combustion of gases. These side tubes, instead of being fastened in by the use of an expander, are held by bushings threaded inside Figure 79. Snyder's vertical boiler. 285 and out to fit the taper threads on the outside of the small tubes and the holes in. the central shell B which receive them. The stay 2, figure 79, is fastened in the same way to the top and bottom heads. One fact is worthy of notice in reference to putting in boil- er tubes in this way, viz, that the tubes and stays must be brought to an absolute fit be- fore the thread of the bushings can be entered or started, con- sequently they impose no strain upon the boiler of themselves^ as is too often the case of or- dinary riveted stays. The whole of the boiler pro- per is secured to the upper plate, marked 20 in figure 79, and is suspended in the inner casing marked 8, around which is still another casing, marked 7. The air supply for the fur- nace may be made to enter at the top of the boiler, at 20, and figukk so. pass down into the ash pit between the casings 7 and 8. This will supply the fire with heated air, thus adding to the economy of fuel and preventing loss of heat by radia- tion. The following table gives the principal dimensions of these boilers : 286 A TREATISE ON STEAM BOILERS. TABLE LXXX. SNYDER'S VEETICAL BOILERS. HOKSE CENTRAL SHELL. SIDE TUBES. HEAT- ING SUK- FACE. BOILER. WEIGHT POWER. DIAM. LENGTH THICK. NO. DIAM. LENGTH DIAM. HEIGHT PLETE. INCH. INCHES. INCHES. INCH. INCHES. SQ. FT. INCH. INCHES. POUNDS. 1 10 24 1^ 32 % 16 14 20 50 380 2 12 30 5 16 38 % 24 24 25 61 700 3 15 39 T^^ 44 % 27 35 25 68 ^ 850 4 15 39 ^ 88 % 50&36 48 30 70 1,150 5 15 42 fo 88 % 50 & 42 58 30 73 1,250 6 15 46 1% 95 % 60&42 69 35 77 1,300 7 16 50 1% 105 % 60&50 83 35 81 1,425 Flynn's Vertical toiler — A vertical boiler, having many of the good features already recommended, are contained in the design by Mr. Daniel Flynn, Fall River, Massachu- setts, and shown in elevation in figures 81 and 82. Its chief peculiarity lies in an enlargement or belt around the waist or middle portion, which is enclosed with and forms a part of the boiler shell, and which, in combination with the provision for returning gases, contributes greatly to the efiiciency of the invention. Figure 81 is a side elevation, showing, on the right hand, the outside of the casing, and on the left, the same broken away, presenting a perpendicular section of the interior arrangements. Figure 82 is a horizontal section of the boiler through XY. In Figure 81, A is the grate, B the fire chamber, and C and C the surrounding interior and exterior shells. The products of combustion follow- ing the direction of the arrows in the engraving, arising from B, first pass through the fire tubes, aaaa, into the mixing chamber, E. From this receptacle, the gases have their exit through the large openings, FFF, and after FLYNK S VERTICAL BOILER. 287 Figure 81. 288 A 'JREATISE ON STEAM BOILER; having imparted a portion of their heat in the ordinary manner, are retained by the conical casing, P, wliich incloses the space, Gr. They are consequently compelled to descend through the fire tubes, 6, into an annular chamber, H, which is inclosed in a conical casing, Q. Thence the gases rise through the exterior circle of fire tubes, c, and pass Figure 82. into the large space, I, and finally are discharged through the chimney at the apex, the object of forcing them through this circuitous course being to gain the full ben- efit of every particle of heat. The particular enlargement above referred to consists of the space between the annu- lar tube sheets. Mi M^, and the outer casing, R, in which are the circles of tubes, b and c. A special point of advantage, to which attention is directed, is the arrangement of the water spaces. A cen- tral chamber, W, will be noticed, extending above the crown sheets as far as the mixing chamber, E. At this flynn's vertical boiler. 289 point, it is reduced in size to a tube, W, which terminates at the bottom of the steam drum, S, its open upper end being surrounded by a perforated cover, V, which prevents a too violent upward motion of the current generated in the lower chamber. In connection with the other water spaces which lie between the systems of tubes, surrounding the fire chamber and occupying the interior of the surround- ing casings of the mixing chamber, and finally cover the lower portion of the steam drum; this central chamber adds greatly to the already large separating surface, so that steam may be rapidly disengaged without carrying up water into the steam pipes. For easy access to all parts of this boiler, for repairs^ ample provision has been made. By removing the cover- ing at Q the tubes, h and c, may be readily cleaned, the refuse falling out at H, by its own weight. The opening of the door at L permits entrance to the space, I, after which, the door, P, being displaced, access may be had to the chamber, Gr. Through the opening, O, the interior of the steam drum may be reached. At U is the steam pipe, its inner end, T, opening upwards in order to prevent its becoming obstructed through priming of the boiler. To the left of the illustration is the appliance for the test cocks and glass water gauge, which, it is claimed, prevents these appendages from being choked or otherwise rendered inoperative. Its form is plainly shown and needs no spe- cial explanation. The efficiency of this boiler has been amply tested and with successful results. Attention is called to the liberal size of the grate, which, it will be noticed, is of much larger area than could be aflbrded if the lower portion of the boiler were made on a cylindrical instead of on a con- ical form. As regards economy, its consumption of fuel is claimed not to exceed two and a half pounds of coal per hour per horse power. In a recent letter to the writer, (20) 290 A TREATISE ON STEAM BOILERS. Mr. Fiynn says that he has obtained an evaporation of llj pounds of water per pound of Cumberland coal. Ample steam space is afforded, which maybe increased by making the steam drum of any required height. The outside covering forms a jacket which confines the heated gases around the interior steam generator, so that every available portion of heat contained in the escaping gases is utilized. Suiter's patent steam boiler — This boiler is of the fire box, fire and water tube variety, and consists of furnace, fire throat, combustion chamber and horizontal return tubular l3oiler — the whole united to operate together. About the filre box, throat and combustion chamber are water spaces similar to the water legs in ordinary fire box boilers; and circulating pipes are provided from the bottom of combus- tion chamber to the bottom of the fire box, and from the sides of the horizontal boiler to the side spaces in the fire box. A steam pipe from the steam space in the top of the combustion chamber to the steam space in the horizontal boiler is also provided. The grate bar is somewhat novel, and consists of the ordinary straight single bar depressed to form a fire basket in the center and provided with spaces at the ends to admit air over the fire. It will be observed that the combustion chamber is so designed that the pressure on the upper and lower sheets tends to tighten the joints of the vertical tubes and thus require no special staying, the circulating pipes acting as supports. Messrs. Slusser & Suiter, Cincinnati, Ohio, have in use at their works a small boiler, as shown in Figure 83; it is of the following dimensions: Horizontal shell four feet ten inches long by twenty-six inches diameter, with twelve three-inch tubes whole length of shell. Fire box twenty-six inches diameter inside by eighteen inches ver- SULTER S BOILER. 291 292 A TREATISE ON STEAM BOILERS. tical depth from bottom of boiler to grate, with two inches water space around the fire chamber and ash pit. Ash pit same diameter as fire box and 12-^ inches deep, with a water bottom two inches deep, connected with the water spaces around the fire chamber. The fire throat, which takes the place of the ordinary bridge wall, has a constant width of twenty-six inches and a vertical depth ranging from ten inches at the front end to 'G.ve inches at the back end; the length of throat parallel with axis of boiler is eighteen inches. The bottom of the throat is an arc of a circle of thirteen inches radius and has a two-inch water space connected with the water spaces of fire box and combustion chamber. The combustion chamber has a diameter inside of twenty-six inches and a vertical height in center of twenty-seven inches. The bottom of the com- bustion chamber is provided with an entrance hole for examinatidn of the interior and removal of soot and ashes as these may collect in the use of the boiler; this hole is surrounded by an annular water space two inches deep. The water space above the crown plate of combustion chamber and the annular space around the entrance hole are connected by vertical water tubes to secure circulation. The chimney is of sheet iron, twelve inches diameter and about thirty-two feet high from surface of fire grate, con- nected to front of horizontal section of boiler by the usual breeching. An evaporative test was made in April, 1878, underthe direction of Mr. John W. Hill, consulting engineer in that city. The coal fired during the trial was Pittsburg, taken from the pile in the boiler room ; this was weighed and dumped in charges of 25 pounds. The water was meas- ured to the boiler in charges of 300 pounds, by duplicate tanks connected with suction of feed pump. TEST OF SULTER's BOILER. 293 Calorimeter tests of the quality of steam produced exhibited a slight super heat ; hence, all the water pumped into the boiler was evaporated. DIMENSIONS OF FURNACE AND BOILER. Length of horizontal shell 4 feet 10 inches. Diameter of horizontal shell 26 inches. Diameter of fire box inside..- 26 inches. Diameter of combustion chamber inside 26 inches. Horizontal tubes 12, 3 inches. Vertical tubes 6, 3 inches. Heating surface 100 square feet. Grate surface 1 .983 square feet. Cross section of tubes 84.82 square inches. Heating to grate surface 50.43 Grate surface to cross section of flues.... 3.36 Cross section flues to chimney 75 DATA FROM THE TRIAL. Duration of trial 9 hours. Temperature of atmosphere '. 73.8° Temperature of water to boiler 146.5° Pressure by steam gauge (corrected) 93.93 Water delivered to boiler 4965. lbs. Water entrained •. None. Coal fired 594.5 lbs. Ash and clinker returned 37.5 lbs. RESULTS OF TRIAL. Steam per pound of coal (from feed).... 8.35 lbs. Steam per pound of coal from and at 212° 9.25 lbs. Steam per pound of . combustible 8.906 lbs. Steam per square foot of heating surface per hour, 5.52 lbs. Coal fired per square foot of grate surface per hour 33.31 lbs. Percentage of non-combustible, in coal 6.3 The boiler is entirely unprotected from loss of heat by radiation, and, according to Mr. J. C. Hoadley's deduc- tions, eleven per cent of the total heat developed was wasted in this direction ; whilst, had the boiler been well protected by brick side and end walls and overhead arch? 294 A TREATISE ON STEAM BOILERS. with an air space between the brick-work and surfaces of boiler, the loss by surface radiation would have been reduced to about three per cent, and with other conditions the same, the trial would have developed an evaporation per pound of coal from and at 212° Fahrenheit of ^-^.f^^^ =10.07 pounds. Several years ago Mr. Hill made a series of evaporation trials on five small locomotive fire box boilers, the heating surfaces in which were, for the First boiler.. 95 sup. feet. Second boiler 85 sup. feet. Third boiler 102 sup. feet. Fourth boiler 84 sup. feet. Fifth boiler 85 sup. feet. And the evaporation per pound of Pittsburg coal, from and at 212° Fahrenheit, was for the First boiler 6.00 lbs. Second boiler 5.07 lbs. Third boiler 5.54 lbs. . Fourth boiler 6.12 lbs. Fifth boiler 6.44 lbs. Taking the average evaporation of these five boilers at 5.83 pounds, then by this data the Suiter's boiler is capable of doing (f.ff = 1.585) nearly sixty per cent more work> or furnishing sixty per cent more steam with the same expenditure of coal. Taken together, the heating surface was less and the grate surface more in the five boilers mentioned than in the Suiter's, and by the ordinary methods of estimating boiler capacity, would be reckoned equal in power to the Suiter's ; hence the comparison of economic effects is fair, and exhibits the relative value of the latter boiler in a striking manner. PORTABLE ENGINE BOILERS. 295 During the trial the handling of the coal was not the best possible, and the boiler was set in the building in such a manner that the air currents freely circulated around, facilitating the absorption of heat by the atmos- phere from the naked surfaces of the arrangement. Mr. Hill says : " Considering the size of the boiler, I regard the economy obtained as excellent ; and am of the opinion that there is merit sufficient in it to justify the construction and trial of similar boilers of larger dimen- sions." The question of durability and facility of repair can only be determined by continuous use for a reasonable length of time. Portable boilers — The demand for a portable engine for agricultural purposes has been increasing for many years Figure 84. past, and is now a very large and important branch of industry. The boilers supplied with this class of engines are often of some one of the vertical tubular varieties, but more generally a modification of the locomotive type. 296 A TREATISE ON STEAM BOILERS. Figure 84 is a sectional elevation of a boiler designed by the writer for the Atlas Engine Works, Indianapolis, Ind. There are several hundred of them nov^ in use, and, with proper care, form a very good and serviceable kind of boiler for the purpose intended. The writer does not wholly approve of the water bot- tom to the fire box. This was made so originally in order to meet the requirements of Southern planters for use in or around their cotton gins; they were apprehensive that open bottoms, fitted with the ordinary ash pans, might endan- ger their premises, and required something which seemed to offer a better security against fire. This boiler is fitted with a fire box, having an arched top and strengthened by means of stay bolts, as shown in the engraving. The reversing of the head at the fire box was done to secure better riveting_and calking. This boiler steams rapidly and is very economical in the use of fuel. The following are the principal dimensions, as sup- plied with portable engines. The latter size is not usually supplied with wheels, but mounted on skids instead. It can be mounted, however, if desired : TABLE LXXXI. PRINCIPAL DIMENTIONS OF PORTABLE BOILERS. HORSE POWER 8 10 15 Diameter of boiler 26 in. • 8 ft. 30 in. 33 in. 21 in. 19 4J ft. 28 in. 8 ft. 7 in. 30 in. 34 in. 23 in. 27 5 ft. 30 in. Length of boiler 10 ft. Length of fire box 40 in. Height of fire box 40 in. Width of fire box 26 in. No. of 2^-inch tubes 28 Length of tubes 6 ft. SEMI-PORTABLE BOILERS. 297 In some " practical " tests made at the works with very inferior coal as fuel, and feed water at a temperature of 65° Fahrenheit, the eight horse power boiler evaporated twelve cubic feet of water per hour, the ten horse power evapo- rating fifteen cubic feet in the same time. The evapora- tion was under a pressure of eighty pounds per square inch. The boilers were in the condition usually delivered to the trade, and the test was as near as possible the same as the firing would have been in the hands of the purchaser, except that it was conducted with a view to ascertain the actual evaporative capacity of the boiler instead of an economy trial. TABLE LXXXII. SEMI-PORTABLE BOILERS BY ATLAS ENGINE WORKS. ENGINE. HORSE POWER. 15 20 25 30 40 Diameter of cylinder .... Length of stroke 8 in. 12 in. 160 30 in. 10 ft. 40 in. 40 in. 26 in. 28 6 ft. 16 in. 18 in. 12 in. 20 ft. 9 in. 14 in. 150 32 in. 12 ft. 54 in. 45 in. 28 in. 28 7 ft. 18 in. 20 in. 12 in. 20 ft. 10 in. 16 in. 140 36 in. 12^ ft. 54 in. 48 iTi. 32 in. 38 7* ft. 20 in. 22 in. 14 in. 25 ft. 10 in. 20 in. 120 40 in. 13 ft. 54 in. 49 in. 36 in. 49 8 ft. 24 in. 24 in. 16 in. 30 ft. 12 in. 20 in. Revolution s 120 Diameter of boiler Length of boiler 42 in. 15 ft 4 in. Length of fire box Height of fire box Width of fire box No. of 2J inch tubes Length of tubes 54 in. 50 in. 37 in. 58 9 ft. Diameter of dome Height of dome Diameter of stack ,. Length of stack • 24 in. 24 in. 20 in. 30 ft. 298 A TREATISE ON STEAM BOILERS. The above table gives the proportions for boilers of the same style, but of larger sizes. Locomotive boilers — It is not within the scope of the present work to enter into the details of locomotive con- FiGURE 85. Class A. struction, in which the boiler figures so largely. Much that has already been said in regard to boilers in general applies to locomotives. There are also many details of contruction, which are peculiar to different builders and to certain roads. These could not be entered into without departing from the original purpose of the writer. It will suffice, perhaps, to give in brief outline the princi- pal dimensions of the locomotive boilers in use on the Figure Class B. Pennsylvania railroad as a guide merely for the propor- tioning of this kind of boilers for stationary uses. A short description is appended to each engraving to show LOCOMOTIVE BOILERS. 299 the particular service for which each class of engine is intended. In using the proportions in the table for stationary purposes the size of the boiler is about right for single cylinder engines of the sizes given in the table, if used with natural draft. The combustion of fuel is not as econ- FiGURE 87. Class C. Anthracite. omical "on the road" as when the rate is lower; which would be the case when used with ordinary chimney or force draft. Figure 85 is a representation of the boiler used with engines in class A, 17X24 cylinders, which are the one& employed for passenger trains on the main line, except in Figure Class C. Bituminous. the mountain districts. The principal dimensions are given in table LXXXIII. The shell and fire box of this class of boilers, and, indeed, all the boilers on this road, are 300 A TREATISE ON STEAM BOILERS. wholly of steel ; the tubes are, in all cases, of wrought iron, lap welded. The boilers in class B are somewhat larger than in class A, and supply 18X24 cylinders. These engines are used mainly in the mountainous districts for passenger service. Figure 89. Class D. The driving wheels are sixty-two inches diameter as against sixty-eight in class A. The tubes are increased in number and decreased in length over class A; the total heating surface being nearly the same. There are two styles of boilers for the engines coming within class C. The one represented in figure 87 is for Figure 90. Class E. burning anthracite coal, and the one in figure 88 is for bituminous coal. The cylinders for both styles of boilers are 17X24. This engine is used for passenger, local and fast freight trains. The number and size of the tubes vary LOCOMOTIVE BOILERS. 301 between these two boilers, a smaller tube being used for the anthracite than for the bituminous coal. The rate of combustion being slower for the anthracite coal, the grate is of larger area, and the heating surface in the fire box increased nearly forty per cent. The tube area is some- FiGURE 91. Class F. what less than for bituminous coal. The total heating surface divided by the fire grate area is 60.5 for the bitu- minous, and 39.86 for the anthracite burning boiler. The driving wheels are sixty-two inches diameter. The engines in class D are intended for ordinary freight service. The tubes are larger in diameter than for any of Figure 92. Class G. the boilers preceding it and of greater length. The fire- grate area is also less, the grate bars being but sixty inches in length ; the total heating surface in the fire box being ninety-six feet as against one hundred and fifteen in class C, bituminous. The cylinders in engines of this class are^ 302 A TREATISE ON STEAM BOILERS. 18X22, with six driving wheels, fifty-six inches in diameter. The engines in class E are intended for freight service in the mountain districts. The cylinders are the same as for class D, viz, 18X22. The drivers are six inches less in diameter. The tubes are longer than in class D, as is also the length of the fire box. The total heating surface in this boiler is greater than any preceding it, except class C, anthracite. The boilers in class F differ from those already referred to, in the absence of the "camel back." This does not reduce the width of the fire box, but does reduce the height. This engine is used for making up trains and for Figure 93, Class H. general yard service. The cylinders are 15X18 inches, the driving wheels are 44 inches diameter. The tank is placed over the boiler, as shown. The engines in class G are used for passenger service on branch lines. The cylinders are 15X22 inches, and have 56-inch driving wheels. This is a good form of boiler to use for stationary engines. The writer prefers it to what is known as the camel back, as shown in figures 85 to 90. As the rate of combustion is less when used in a building from what it would be " on the road," the fire box might be lengthened if thought necessary. EnOINE' Thickness of boiler platei Thickness of boiler plate Thickness of boiler plate; Maximum internal diamt Maximum internal diamt Height from top of rail t( Number of tubes Inside diameter of tubes. Outside diameter of tubei Length of tubes between Number of internal diam Length of fire box at bot1 Width of fire box at bott Height of crown sheet ab Thickness of inside fire b Thickness of inside fire b Thickness of tube sheets, {Diameter of cy Length of strot H External heating surface Internal heating surface c Fire area through tubes.. Firegrate area Heating surface of fire bo Total heating surface witl Total heating surface witl External tube surface div Total heating surface divi Fire grate area divided bj Diameter of smoke stack., Least sectional area of ch Fire Grate area divided Pressure of steam per squ Effective pressure per squ Capacity of tank Capacity of coal tank... CLASS o, PASSENGER ENGINE. CLASS H, SHIFTING ENGINE. 5 T6 Vs % 44% 70 130 1% 2 115 65.7 64% 35 52% 5 3^ 15 22 652.31 sq. ft, 574.0 sq. ft, 2.17 sq. ft, 13.3 sq. ft. 69.04 sq. ft. 721.35 sq. ft. 640.04 sq. ft. 9.44 .')4.23 6.13 17 in. 1.58 sq. ft. 8.50 125 lb."*. 100 lbs. 1,600 gals. 6,500 lbs. 5 T6" Vs % 47% U% 63>^ 91 2% 2% 156% 69.5 54.5 35 46>^ 5 TS" 15 22 776.13 sq. ft. 699.79 sq. ft. 2.51 sq. ft. 13.2 sq. ft. 79.11 sq. ft. 855.24 sq. ft. 778.90 sq.ft. 9.80 64.79 5.25 15 in. 1.23 sq. ft. 10.8 125 lbs. 100 lbs. 2,200 gals. 5,000 lbs. CLASS I, FREIGHT ENGINE. 5 % 7 .55% 53% 77 138 2>i 2^ 153 68.1 96 34% 43>i •^M 5 TS 20 24 1,158.65 sq. ft. 1,043.28 sq. ft. 3.75 sq. ft. 23.0 sq.ft. 100.91 sq. ft, 1,259.56 sq. ft. 1,144.19 sq. ft. 11.48 54.76 6.16 20 in. 2.18 sq. ft, 10.54 125 lbs. 100 lbs. 3,000 gals. 8,000 lbs. \. TABLE LXXXIII. LOCOMOTIVE BOILERS, PENNSYLVANIA RAILROAD. Thickness of boiler plat«s, barrel and dome Thickness of boiler plates, outside fire box slope ThiekQessof boiler plates, waist and smoke box Maximum Internal diameter of boiler Maximum internal diameter of wagon top Heightfrom top of rail to center of boiler Number of tubes Inside diameter of tubes Outside diameter of tubes Length of tubes between sheets Number of internal diameters in length of tube Length of fire box at bottom (inside) Width of fire box at bottom (inside) Height of crown sheet above top of grate Thickness of inside fire box sheets, sides Thickness of inside fire box sheets, front, back and crown Thickness of tube sheets {Diameter of cylinder Length of stroke HEATING SURFACE. External heating surface of tubes Internal heating surface of tubes Fire area through tubes Firegrate area Heating surface of fire box Total heating surface with external area of tubes Total heating surface with Internal area of tubes External tube surface divided by fire box area Total heating surface divided by fire gi^te area Fire grate area divided by tube area Diameter of smokestack Least sectional area of chimney Fire Grate area divided by least sectional area of chimney Pressure of steam per square inch Effective pressure per square inch estimated at g boiler pressure- Capacity of tank Capacity of coal tank 66% 35 60% 920.52 ,sq. ft. 815.4 sq. ft. 3.16 sq. ft. 16.1 sq. ft. 131.72 sq. ft. ,052 24 sq. ft. 947.12 sq. ft. 6.98 65.36 5.09 18 in. 1.77 sq. ft. 9.11 125 lbs. 100 lbs. 2,400 gals. 8,000 lbs. % % 51-% 2Ji 63.5 601^ }4 941 .87 sq. ft. 858.70 sq. ft. 3.37 sq. ft. 17.6 sq. ft. 115.11 sq. ft. 1,066.98 sq. ft. 973 51 sq. ft. 8.18 00.5 5.22 18 in. 1.77 sq. ft. 9.96 125 lbs. 100 lbs. 2,400 gals. 8,000 lbs. BITUMINOUS PASSENGER ENGINE. % 51% 48iJ4 911.87 sq. ft, 858.70 sq. ft, 3..37 sq. ft, 17.11 ,sq. ft. 115.11 sq. ft. 1,056.98 sq. ft. 73.51 sq. ft. 8.18 60.5 5.22 18 in. 77 sq. ft. 9.96 126 lbs. 100 lbs. i,400 gals. 5,000 lbs. ANTHRACITE PASSENGER ENGINE. % 125fk 71.7 119% 34^ 44 1,002.60 sq. ft 878.45 sq. ft, 3.06 sq. ft 29.13 .sq. ft, 1.68.56 .sq. ft. 1,161.15 sq. ft. 1,036.95 sq. ft. 18 in. 1.77 sq. ft. 16.48 125 lbs. 1011 lbs. 2,40(1 gals. 8,000 lbs. % 50% 2X 67.5 69J^ 57% 997.25 sq. ft, 886..65 sq. ft, 3.28 sq. ft, 14.6 sq. ft, 95.97 sq. ft, 1,093.22 sq. ft. 982.52 sq. ft. 10.39 75.38 4.42 18 in. 1.77 sq. ft. 8.20 125 lbs. 100 lbs. 2,400 gals. 8,000 lbs. % % 51K 48% 2J4 2% 146tt 64.9 6714 35 57^ X 6 18 22 984.83 sq 885.6 sq 3.39 sq 16.34 sq. 111.27 .sq. ,096.10 sq. ft. 996.87 sq. ft. 8.85 67.18 4.85 18 in. 1.77 sq. ft 9.24 125 lbs. 100 lbs. 2,400 gals 8,000 lbs Vs % 43% 42% 63K 74.5 44Ji 651.22 sq. ftl 678.5 sq. ft 1.93 sq. ftl 10.7 sq. ft; 61.29 sq. ft, 712.51 sq. ft. 639.79 sq. ft; 10.62 66.59 5..54 15 in. 1.23 sq. ft. 8.72 125 lbs. 101) lbs. 820 gals. 1,500 lbs. % 44% m 66.7 64% 52% 652.31 sq. ft 574.0 sq. ft 2.17.si|. ft 13.3 sq. ft, 69.04 sq. ft, 721.35 sq. ft. 640.04 sq. ft. 9.44 54.23 6.13 17 in. 1.58 sq. ft. 8.,60 125 lbs. 100 lbs. 1,600 gals. 6,500 lbs. % % 47^ 44% 633^ 91 2% 2K 46K 776.13 sq. ft, 699.79 sq. ft. 2.61 sq. ft. 13.2 sq. ft. 79.11 sq. ft. 855.24 sq. ft. 778.90 sq. ft. 9.80 64.79 5.25 15 in. 1.23 sq. ft. 10.8 125 lbs. 100 lbs. 2,200 gals. 5,000 lbs. 65% ,53% 2Ji 2>^ 43^ 1,168.65 sq. ft. 1,043.28 sq. ft. 3.75 sq. ft. 23.0 sq. ft. 100.91 sq. ft. 1,259.56 sq. ft. 1,144.19 sq. ft. 11.48 54.76 6.16 20 in. 2.18 sq. ft. 10.54 125 lbs. 100 lbs. 3,000 gals. 8,000 lbs. LOCOMOTIVE BOILERS. 303 The boilers in class H are very similar to those in class F. The boilers are larger in diameter, as are also the diameter of the tubes. The firebox is about ten inches longer. The cylinders are 15X22 inches. The driving wheels are forty-four inches in diameter. Figure 94. Class I. The engines belonging to class I are represented in figure 94. This is the largest and heaviest class of engines in use on this road. The boilers are peculiar in their construction, as seen from the above engraving. These engines are for heavy freight service in the mountain dis- tricts. The cylinders are 20X24 inches. The driving wheels are 50 inches in diameter. For the constructive details of this engine and boiler, and for the others already given, the reader is referred to ''Engineering," volume 24 or to '^ The Pennsylvania Eailroad," which is a reprint of the articles contained from week to week in "Engineer- iDg." The principal dimensions of these boilers are collated from the tables published in Engineering, and given in table LXXXIII, which may be of very great service to those interested in boilers of this type. \. CHAPTER XII. BOILER SETTING. Ordinary Boiler Settings — Grates — Force Draft — Resistance of Air in Passing through Pipes — Sizes of Pipes required for Grate Areas — The Jarvis Furnace — The Butman Furnace — The Pierce Furnace. After deciding which kind of a boiler is best adapted for any particular service, the question of boiler setting should then be carefully con- sidered. There are all sorts of ideas as to how a boiler should be set; many of them good and quite as many more at variance with the actual re- quirements. It matters little what the particular design of the furnace may be, if it affords complete combustion. The requirements in furnace con- struction have already been presented in Combustion of Coal, to which the reader is referred, particularly to Chapter V, on combustion. Chapter YI, on air required for furnace combustion. Chapter YII, on the furnace. Chapter YIII, on the products of combustion. A very common form of boiler setting is shown in figure 96, where but little money is to be expended in its erection. It is by no means the ideal furnace ; yet it fur- nishes good evaporative results when properly fired. Figure 95. BOILER SETTING. 805 The following table gives the ordinary proportions ; the letters in the table are those corresponding to dimension lines as given in the engraving bearing the same letters : k^£_> Figure W. TABLE LXXXIV. PROPORTIONS FOR FLUE AND TUBULAR BOILER SETTINGS. A II dimensions are in inches. A. B. "■[ D. E. F. G. H. I. J. K. L. M. N. O. p. Q- K. s. T. U. V. w. X. Y. z. 36 144 ■25 170 18 48 42 18 44 18 16 18 51 9 13 16 81 13 24 45 13 oil 71 74 24 38 144 ■25 170 18 48 42 18 44 18 16 20 51 9 13 16 82 13 24 47 13 51 73 75 24 40 144 25 1 170 18 48 42 18 44 18 16 20 51 9 13 16 84 13 24 49 13 51 75 76 24 42 144 21 170 18 52 46 18 44 18 16 20 51 9 13 16 85 13 24 51 13 55 77 77 24 44 144 21 1 170 18 52 46 18 44 18 16 20 51 9 13 16 86 13 24 53 13 55 79 78 24 46 144 21 170 18 52 46 18 44 18 If 16 20 51 9 13 16 88 13 24 55 13 55 81 79 24. 4S 144 21 170 18 52 46 18 44 18 « 16 20 51 9 13 16 89 13 30 57 13 55 83 80 24 50 144 13 170 18 60 54 18 44 18 2 16 20 51 9 13 16 90 ■13 30 59 13 63 85 81 24 52 144 13 177 18 60 54 18 44 18 02^ 16 20 51 9 13 18 ?2 18 30 61 18 63 97 82 24 54 144 13 177 18 60 54 18 44 18 16 20 51 9 13 18 93 18 30 63 18 63 99 83 24 56 144 1 177 18 72 66 18 44 18 16 20 51 9 13 18 94 18 30 65 18 75 101 84 24 58 144 1 177 18 72 66 18 44 16 16 20 51 9 13 18 96 18 30 67 18 75 103 85 24 60 144 1 i 177 18 72 66 18 44 18 16 20 51 9 13 18 97 18 30 69 18 75 105 86 24 (21) 306 A TREATISE ON STEAM BOILERS. The distance given in column D is from the center of the cast iron front to the back end of the rear wall. The distance E is intended to be that of two bricks. There are sections of the country where bricks will not lay to nine-inch centers. In any such case the thickness Figure 97. This will may be varied to suit the size of the bricks, apply to columns H, J, P, S and V. The distance K will vary with the diameter of the mud drum, if one is used. The drum may, in general, be one- third the diameter of the boiler and may extend from out- side to outside of the walls and should always be fitted with a man hole. In building the wall, proper allowance must be made for expansion. Figure 97 shows a mud Figure 98. drum suitable for a' single setting and figure 98 for a dou- ble setting. Sometimes the mud drum is placed in the direction of the boiler instead of across it ; in that case the nozzle is fitted at one end of the drum, the other end passing out through the rear wall. BOILER SETTING. 307 The distance given in column L is that, when the noz- zle is riveted to the second sheet from the rear end, where the sheets have twenty-four inch centers. The thickness, S and V, are for single walls, but double walls are recommended instead, as being more economical in fuel. The distance W is that from the face of the bridge wall to the inside of the fire front. The grate bars have a "rest" of about one inch at each end on the bridge wall plate and on the bearing bar attached to the. fire front. The fronts are intended for separate breechings to be attached to the boiler. When a smoke box is formed by a continuation of the shell, a deeper lining around the fire will be required if the front is brought out "fiush" with end of the boiler. This will also change distances D andG as well as W. The distance, Z, is to the under side of the grate. The engraving shows the grate bars slightly depressed at the back end. This a very common practice in setting grates. The writer does not attach any special importance to it. The distance from the under side of an externally fired boiler to the top of the grates will vary with the kind of fuel to be burned. For bituminous coal, so far as the writer's observation goes, thirty inches appears to be the best distance. There are several "batteries"' of boilers known to the writer in which the distance is thirty-six inches, and in one instance forty-eight inches. It is not apparent that anything is gained by this extra distance over thirty inches. For semi-bituminous coal the distance may be twenty to twenty-four inches, and for hard anthracite about eighteen inches. These distances may be varied somewhat to suit local conditions. The furnace walls, as shown in figure 95, are brought in to the boiler sides on the line of the diameter. It is recommended that they be carried up to the water line or just below it, and thus increase the heating surface. 308 A TREATISE ON STEAM BOILERS. It will be observed that the longitudinal dimensions are for boilers twelve feet long in all cases. In whatever amount the boiler is longer than that distance, columns B, C and D only are changed, unless the rear support to the boiler is to be brought nearer the furnace, in which case the distance, L, will be increased and C decreased. This setting has a cast iron plate at the back end of the boiler, instead of being arclied, as is sometimes the case, and shown in figure 105. The writer prefers a plate, especially for tubular boil- ers, as it enables the end of the tubes to be quickly got at for repairs or examination and affords a good light at the same time. No objection exists, however, to the brick arch, and it may be commended for the facility which it affords the return of the gases through the tubes or flues. The walls are carried up, as shown in the engraving, and filled in over the top of the boiler with some good non-conducting material. There are a number of good non-conductors in the market, which may be used, or a brick arch may be carried over the top of the boiler. This arch may rest upon a thin wooden lining laid over the top of the boiler. AYherever brick work is to come> in contact with the sbell of the boiler, the joints should be made with fire clay instead of lime mortar; fire clay should also be used with fire brick in the furnace. Brick should be used throughout in the boiler setting, and is to be preferred to stone for the foundations. There are localities in which boilers are seldom or never set in the manner shown in figure 96, but have roll- ers underneath, as shown in figure 105; or rest upon the side walls, the boilers having cast iron lugs riveted to the shell, as shown in figure 61. The writer has fitted them in each of the styles indicated, but prefers the mud drum, either as shown, or passing out through the rear wall. RYDEll't^ GRATE BAR. 309 The grate surface is composed alternately of solid metal and free spaces in about equal areas, or, perhaps, the solid metal slightly exceeds the spaces. The grates should have depth instead of width for the needed strength. It is cus- tomary to make the grates of cast iron, though bars of inch or inch and a quarter square iron are often used. For burning anthracite coal a grate made of wrought iron tubes, through which the feed water is made to pass, has been used with good results. A modilication of this grate is frequently applied to locomotives. The ordinary grate bar is too well known to need any description. Figure !H) is an engraving of a grate bar made by Albright Sc Stroh, Mauch Chunk, Pa. This grate FiGUKR ilO. bar may be used with coal, wood or sawdust, and presents more free space than is usual for grates of this class; it is of liofht w^eio;ht, and after a trial it has been found to with- stand warping in heavy tiring. The clearing of tires by means of slice bars and hooks is a very difiicult and exhausting kind of labor, especially if the grates are of any considerable length. The radiant heat from an open furnace door during the hot summer months is very trying, and none but experienced tiremen, as a rule, can stand it. Ryder's reciprocal gratehars — Figure 100 is an engrav- ing of a grate bar manufactured by Mr. J. F. Montgomery, Taunton, Mass., and is known as Kyder's patent reciprocal grate bar, which was designed to overcome this labor- ious and ditiicult operation. These bars are simple in construction, and consist of a series of movable and sta- tionary bars. The movable bars (ev^ry other one being 310 A TREATISE ON STEAM BOILERS. stationary) are moved backward and forward several inches by a lever in front of the boiler, through the ash pit door. The movable bars resting on friction rolls are raised above the stationary bars a little, and have a cor- rugated surface for friction, which thoroughly disturbs the coal, destroys the clinkers and removes all the ashes, thus opening up a thorough and uniform draft over the entire Figure l^U. tire surface. By the uniformity of draft and ventilation thus gained, a more perfect combustion of all the gases is obtained. Steamship engineers and (iremen of long experience sailing out of New York and Boston, and ihose who are running stationary engines and boilers with these bars, report excellent I'csults in the economy of coal, and that an easv ireneration and unitbrm pressure ot steam is obtained with lii^ht lires — tiring often with little coal at a time. RYDER S GRATE BAR. 311 The bars should be worked as often as appears necessary to keep the fire clear from ashes and clinkers. The danap- ers in the flues should be kept well closed, in order to intensity the heat and retain it in contact with the heating surface of the boiler and tubes, instead of rapidly forcing it tlirough them by strong draft, and out of the smoke stack or chimney, with the gases of the coal half consumed, thus wasting nearly one-half the value of the coal. Keeping the damper closed as much as possible is a good practice at all times. The writer has referred to it once or twice previously in this book. It is claimed that the reciprocal grate bars develop a new and successful method for the ventilation of fires, as they produce a level sur- face of coal over the entire grate, ensuring, by the re- ciprocal action, a uniform or equal consumption of coal ; that the bars will not warp or crook: that with three vibrations of the lever, more execution can be performed in the ven- tilation and cleaning of the fire, than by the use of the poker or slice bar in one hour's time; that abetter fire can be obtained, with less draft back of the bridge wall, less fuel and less labor than with any other bars in use, and no loss (but a gain) in steam, figure loi. while cleaning fires, as the fire doors are never open wiiile cleaning them. 312 A TREATISE ON STEAM BOILERS. Force draft — In case the draft should be deficient or sluggish, a steam jet or blast nozzle may be used with advantage, something like the one shown in figure 101, and placed in the interior and at the base of the chimney, as shown. This is not equal in efiiciency to the forcing of the air under the grates. This usually requires some special adaptation of the apparatus to the ash pit or fur- nace and is not so easily or cheaply applied, but in the "long run" there is economy by so attaching it, if it is to be a permanent fixture. Figure 102 represents one form of apparatus for a force draft, as designed by Schutte & FORCE DRAFT. 313 Goehri ng, Philadelphia, Pa. Figures 101 and 103 are also bj the same firm. AVhen the blower is attached below the grates, as shown in figures 102 and 103, tbe ash pit must be fitted with a close fitting door, that the pressure of the air be compelled ^^^^^/^^ i«%% ;?^%««S%S« %5?«S4i«««% «S«%S5^^ %%%%?J2i9 ■''/////'■////yK '^/y//yy////A %y ■J 375 H 10 1.250 9| 4 500 6.] 12 1,500 10 5 625 t 15 1,875 11 C) 750 '>h 20 2 500 12i 7 875 8 25 3,125 13* Mr. Sturtevant has given this subject a great deal of attention for many years, and has made many and costly experiments to determine the frictional resistance of air in passing through long tubes or pipes. The following table was calculated for the use of those putting up blast pipes, who, knowing little or nothing of the frictional resistance of the air, are apt to think that because the combined area of four 6-inch pipes is the same as one 12-inch pipe, that the four pipes will convey the same quantity of air, with the same ease and freedom, that the 12-inch will ; whereas it actually does take 5.7 — almost six 6-inch pipes. Again, sixteen 3-inch pipes have the combined area of one 12-inch pipe, but in actual practice it takes just thirty-two 3-inch pipes to do the work of one 12-inch. This is due to the excess of friction for every cubic foot of air in the small pipes over that in the large: 316 A TREATISE ON STEAM BOILERS. U5 rH rt< CM rH '^ CO !>1 Tfl iH '"' ■^ ?a Ol IC 30 iH '^ '"' '"' 02 rH oi uO CO CM Oh iH T-^ r-^ ^ CM O O 00 CO ". CO c t r-. oi CO CO CO t^ t^ PQ «i 3 CO l.o T— ' CO to ira I^ o CO o> ^ c-i o^' CO ■^ lO to oi ^ W P5 O Pm lo •■C CO -M CO t~ s o _ CO :0 -^ cm' ■CO '^ i6 t-^ o ^ '^ "-• H '* CO CO 1— ( t^ '•£> c; ■M o C-. r/* 1^ ^ '"' (>i -* ■n" r-^ ri CM ca 00 (N CO c^ CO C^J ■jS o Ol o r~ CO lO CO OJ Ft CO 03 ^ lO (N t- (^ CO o CM Ol o X t^ c^ CM c The action of the blast, through these side jets, is to prevent the localization of an intense heat underneath the H boiler at the furnace, and to carry it along the sides, and distributing it over a greater area, and promoting a current of hot gases along the sides of the boiler instead of confin- ing it along the bottom, as is generally the case in the ordin- ary settings. The forcing of the air into the cast iron box in the furnace is accomplished by means of an injector, shown in figure 110, and is from the designs of Messrs. Schutte & Goehring. In this jet blower, the in- ducing current is a jet of steam, the quantity of w^hich is controlled by the spindle H, in the steam nozzle; the in- Lduced current of the air is formed in a series of nozzles of increasing area. The pur- pose of these nozzles is to regu- late the admission and proper mixture of the inducing and induced currents in such pro- portions as not to lose power by sudden shocks. The quan- tity of air to be admitted is regulated by a steam valve, H? and is always under the control of the fireman, and may DISCHAEGB. Figure 110. be adjusted as circumstances require. THE BUTMAN FURNACE. 327 Figure 111. Figure 111 is a section at C, showing the compartment for heated gas over the top of the boiler. The line of the pendants, already described, is shown by the two lines, one on either side, and just below the water line of the boiler. Figure 112 shows a section at the rear end of the boiler, and marked D, in figure 107. The arch is built down upon the boiler at this point, to prevent any such thing as a current of heated gases over the top of the boiler; they are thus carried down below the water line, and below the bottom line of pen- dants already referred to. o The area of openings at the sides and above the rear bridge wall are pro- portioned to the area of o]3enings in the flues or tubes ; practically, these side openings are made about one-third greater than the combined areas of the tubes or flues. Figure 113 shows the method of utilizing the waste heat escaping up the chimney. The pipe supplying air above the Figure 112. 328 A TREATISE ON STEAM BOILERS. fire starts at the base of the chimney and ascends to near the top, thence downward and from the bottom to the jet blower, where it is forced into the cast iron box between the grates, as shown in figure 109. The furnace doors, as usually fitted to this furnace, are shown in elevation in figure 114. Unlike the ordinary fire door, it is hinged at the top instead of the side. The door is slightly more than counterbalanced by means of two side weights, shown in the engrav- ing. This weight has a segment of a gear cast in- side, as shown in figure 115, into which a similar toothed segment, fitted to the door, is geared, and thus the movements of the door and weight are controlled by each other. On the same central shaft, around which the weight oscillates, is also secured a deflecting plate, which is shown in both figures 115 and 116. When the door is closed, as shown in figure 115, the de- flecting plate is entirely within the housing. A butterfly register is at- tached to the door, through which a greater or less quantity of air may be admitted, as it may appear to be needed. When the deflect- ing plate is down, as shown in figure 115, the air passes under- FlGURE 113. neath its lower edge, and is thus brought, into close sur- face contact with the bed of burning fuel. . When the furnace door is opened in order to supply fresh fuel, the deflecting plate is thrown out horizontally, as shown in figure 116 ; the object of this defiecting plate being to prevent the cold air from impinging directly against the bottom of the boiler, but to so direct its course that it shall mingle with the heated gases immediately THE BUTMAN FURNACE. 329 over the fire, passing otf with them without any local cool- ing of the lower part of the boiler shell. The grate ased in this boiler setting is quite narrow, and, as will be seen in figure 117, one is placed on each side of the central wall in the furnace. It belongs to that class known as oscillating or rocking grates, and is provided on I Figure 114. its upper surface with oblique cutting edges and interlock- ing fingers, which also form parallel cutting edges. The cross bar is corrugated in the direction of its length, and tapers from the top to its bottom ; the fingers men- tioned in the preceding paragraph are attached to it. These are '' staggered," as shown in the engraving, so as to present a series of irregular orifices, which shall allow a 830 A TREATISE ON STEAM BOILERS. free and full supply of air to the fuel. These fingers are made semi-circular, and when the grate is being " rocked " or " shaken/' the same distance is preserved as when in its normal condition, thus prevent- ing the loss of any small fuel on the grate from falling through in- to the ash pit. Each grate bar has an arm projecting downward, as shown ; these pass through suitable openings in a connecting bar, by which they may all be moved at once, which can be done without opening the furnace door. When the bars are shaken, the whole surface of the fuel is broken up, allowing at the same time complete access of air to the burning fuel. Rocking grates are to be re- commended where the character of the fuel will permit, because they always prevent in great measure, and oftentimes wholly, Figure 115. the cloggiug Up of the frcc spaccs between the bars. In burning any fuel likely to clinker, the spaces are liable to become filled up, and thereby inter- fere with the draft. By means of a rocking or oscillating grate, the clinkers are prevented from forming, by grind- ing or breaking them up as they form. The following test of a boiler set with the Butman furnace at the flouring mill of G. W. Cunningham, Tifl5.n, Ohio, is by Isaac Y. Holmes, M.E., who reports as follows: This boiler was set with the Butman furnace and started on the first of August, 1876. It has therefore been in constant use over a year. On the third of August, 1876, I made an examination and test of THE BUTMAN FURNACE. 331 the same, a fall report of which was forwarded you at that time. The result showed an evaporation at pressure of atmosphere and tempera- FlfJURE 116. ture of 212° of 11.33 pounds of water per one pound of Masillon lump coal. The object you had in view in making this second trial, as stated Figure 117. to me, was to ascertain if the Butman furnace would retain its econ- omic efficiency after being subjected to a year's work, and, also, how 332 A TREATISE ON STEAM BOILERS. the various portions of the structure had withstood the wear and burning of the sarne. Herewith is submitted a summary of results of the tests made on the twentieth instant (August, 1877): DU.TY. Driving a four-run grist mill Clifton mill COAL. The kind used on this trial was Masillon nut DURATION. Continuous firing 7 hours OBSERVATIONS. Total amount water weighed to boiler ...^ 15,376 lbs. Total amount coal weighed to furnace 1,233 lbs. Total amount ash weighed dry 87 lbs. Total amount combustible 1,136 lbs. Average temperature feed water in tank 184° Average temperature gases in uptake 340° Average temperature air in pipe 132° Average temperature air in fire room 97° Average pressure steam in boiler 75° PERFORMANCE. Coal, per hour 176 lbs. Combustible, per hour 162 lbs. Water, per hour 2,196 lbs RESULTS. Pounds of water evaporated at 75 pounds pressure and temperature of 184° per one pound of coal 12.47 Pounds of water evaporated from pressure of atmosphere and temperature of 212° per one pound of coal 13.23 Equivalent evaporation from pressure of at- mosphere and temperature of 212° per one pound of combustible 14.36 The evaporation of 13.23 pounds of water at pressure of atmos- phere and temperature of 212° shows very conclusively that instead of losing its efficiency, it gives a higher rate of evaporation than at first,* although this is no doubt due to the familiarity of the fireman with the furnace. THE PIERCE FURNACE. 333 As regards the deterioration of the brick work or other portions of the structure will say, there was no perceptible wear to any portion of it, and I should not think that even the line of brick at the surface of the burning fuel would not require renewing for several years to come. The boiler was 60 inches X 18 feet, with fifty-six 3J- iiich tubes; was fitted with two grates, as shown in figure 117, each 15 J inches X 4 feet. The mud drum was 20 inches X 8 feet. Two things are noticeable in the above m ////////////////////M//.^ '''' '■■""'"■'■'■■:■ w//////////////////////////////////////////////////////M///m^^^^ FlUUKE 118. test: the moderately high temperature of the feed water and the low temperature of the escaping gases. This is the highest rate of evaporation known to the writer, and if it were not for the ability and unusual care with which Mr. Holmes conducts his tests, the results might be discredited. The Pierce furna ce — The accompanying engravings show an improved design for boilers' furnaces, by Henry M. Pierce, LL.D., of Grand Rapids, Mich. 334 A TREATISE ON STEAM BOILERS. Figure 118 is a vertical longitudinal section, showing the boiler in position in the furnace. In this design the fire grate is wholly removed from the boiler and arched over, as may be more clearly seen in figure 119. The radiant heat of the fuel is not given out in this case so as to be absorbed by the boiler shell, but by the arch of fire brick overhead and along the fiue. The advantages which such a form of construction offers toward effecting the complete combustion of hydrocarbon gases, when pro- perly supplied with air, are too apparent to need any special ex- planation. The furnace is fired in the usual manner, the products pass- ing over a bridge wall, at a high temperature, into a roomy com- FiGURE 119. bustion chamber of equal if not still higher temperature. In this chamber, jets of heated air is admitted through the back of the bridge wall, through a perforated plate, as shown in figure 120. This supply of air in a chamber, at a temperature of over 1,000° Fahrenheit, has the eft'ect to convert the volume of carbonic oxide into carbonic acid gas. The chamber being large, af- fords ample time for complete combus- tion ; the passage of the gases through it is slow, because of its large area. The combustion being complete, the gases pass underneath the boiler, and from thence through the tubes to the chimney. A furnace, constructed after Dr. Pierce's designs, was on exhibition during the Exposition for the year 1879, at Figure 120. THE PIERCE lURNACB. 335 Pittsburo^, Pa., and is preferred to the one just described, when viewed from the wri- ter's standpoint. Figure 121 is a vertical cross sec- tion, showing the boiler, furnace and combustion cham- ber. Figure 122 is a plan of the setting, showing the grate ' and the combustion chamber back and along side of the grates. The fuel is burned upon the grate, and, as in the heated chamber, say in its details of construction, to be Figure 121. setting first described, in a highly 2,500° Fahrenheit. Jets of air are Figure 122. admitted along the sides of the s^rate, as shown in the 336 A TREATISE ON STEAM BOILERS. engraving. The walls are hollow, as shown by the black lines in the plan and elevations, and which serve to heat the air before its admission to the combustion chamber. This furnace attracted a great deal of attention during the exposition. It was fired with Pittsburg coal, which was completely burned, and without the escape of any smoke. . Suitable openings were made in the combustion chamber, which showed that the visible products were y///r///t //W/A •////////. '//////J f//M '/,///v(ii^r:!!^^^^^SS:!^iii^.i^0y,;t m/y/yy wj////, r//m/, »//////. '///f/y/ r,y/y/y/. /,/yMy/ r,/y//,. -myy/yj ,yy/,//j yy/yM ryyyyy.vy wyyyA vyyyy/, '/,i?S^^ SSSSSSi SSSSSSSi iiSSSS^^Sii^/^y/yyyy, vyyyyyyy vyyyyyy. vyyyyyy, ryyyyyyy vyyyyyj vy/y/yy/y 'yyyyyy 'yyyyyyy. vyyyyyy. vyyyyyy yyyyyyy ry.vyyyy, yyyyyyy. // f^s&isssss&issssssi?sssssss&'sssss&si/iy//y/////y///yyyyyy'/^^^ Sji«N» -js^N-v N^vNc ^NNN^ jNiSiS; ryyyyyy 'yyyyyyyi yyyyyyyy /yyyyyyy, yyyyyyy, yyyyyyy vyyyyyyy vyyyyyj yyyyyyy^ vyyyyyy ryyyyyyjryyyyyy.y.vyyyyyfyyyyy^^^^ yy^ yyyyyyyy ryyyyyyy. 'yyyyyyy, yyyyyyyy 'yyyyyyj /yyyyyyy. vyyyyyy, yyyyyyy /yyyyyyy 'yyyyyy/ /yyyyyy/ /yyyyy/y* yyyyyyy. 'yyyyy/^ SS S»»i ■S»t»* 'J^S^» ;N»»N S5NKS S»K« »"»»: •N^^^^s: •{« SNNi tNSNiSi CS8NKi ««i«N> SiNIN >^jiW 5^»^» ^^^J^« ^N^^^^ ^^c^-jN^ x^ /yyyyyy /yyyyyyy,yyyyyyyy /yyyyyyy, /yy 'yyyyyyy. 'yyyyyyyy /yyyyyyy /yyyyyyyy yyyyyyy. yyyyyyyj yyyyyyyy 'yyyyyy, yyyyyyy/ /yyyyyy, yyyyyyy, 'yyyyy, vy 'yyyyyyj /yyyyy Y/////y//yyyyyy////////jf//////y.////////.y/////y.'////////,y/////y'^^^^ v/y:///////yY///////////////jV///////,'//////y////////,y////y/y.///////.^^^ WMyyy /yyyyyyy /yyyyyyy, yyyyyyy/ ryyyyyy/ /yyyyyy, yyyyyyyy, yyyyyyy 'yyyyyyy, 'y/yy/yA ■///////, 'yyyyyyy, z 'yyyyyy /yyyyyyj vy. 'yy,'yy/y/My//////y.////////.W///M.///////jy//////J.'///////jy/////yV//^^^^ /yyyyyy, 'yyyyyyy, yyyyyyyy 'yyyyyyy, /yyyyyyj /yyyyyy, yyyyyyyy 'yyyyyyj /yyyyyyy /yyyyyyy. /yyyyyyy ryyyyyyy 'jyyyyyy. 'yyyyyyy, V, yyy/y/yy//y.//y///y/jy///////y///////jy/////y.y//////yy///////,y/////jy//////y,////^^^^ yyyyyyy /yyyyyyy /yyyyyyyy yyyyyyyy. 'yyyyyyy. 'yyyyyyy /yyyyyyyy /yyyyyy, yyyyyyy, yyyyyyy, yyyyyyy, /yyyyyyy yyyyyyy 'yyyyyyy /y y/yyy/////yyy//////,y//////jy///////,y/////yy///////,y//////y:y//////.//////^^^^ 'yyyyyy, yyyyyyy/ /yyyyyyy 'yyyyyyyy /yyyyyyy 'yyyyyyj yyyyyyyyj yyyyyyy ryyyyyyy 'yyyyyyy. yyyyyyy 'yyyyyyy ', yyyyyy, vyyyy" ' ///j y///y/y, y//////y ///////y 'yyyyyyyy /yyyyyy. yyyyyyyy /yyyyyyy 'yyyyyyj yyyyyyyy /yyyyyy, yyyyyyy, y /yyyyy yyyyyy,/, /... ,.,„,„„.rj,j„j,tj,j,jr,'jjjjjjj.yjfjjj,rjjjyjjy.jyjyyyy.'jiyyyyy.yyyyyyy.'yyy y/y 'yy/yyyyyyy. 'yyyyyyy/ /y/yy/yy 'yyyyyyyy. ; /yyyyyy.'yyyyyyyj/ 'yyy, yyyyyyyy /yyyyyyy. 'yyyyyy, ryyyyyyy 'yyyyyy, ryyyyyy. yyyyy. yy, yyyyyyj ryyyyy, ■yyyyyyyryyyyyyyjyyyyyyy.ryyyyyyyyyyyyyyy/y/yyyyyjyyyyyyyyryyyyyy.ryyyyyyj/j yyy, ryyyyyyy. yyyyyyy 'yyyyyyy /yyyyyyy 'yyyyyy, yyyyyyy, yyyyyyyy, yyy//yy 'yy///, '. yyyyyyy, yyyyyyy 'yyyyyyyy 'yyyyyy, yyyyyyy, yyyyyyyy 'yyyyyyy ryyyyyy, /, yyyyyy, 'yyyyyyy. ' Figure 123. destroyed immediately after they left the grate. Figure 123 shows a vertical longitudinal section, with the arched opening between the two combustion chambers. CHAPTER XIII. FEED APPARATUS. Power Pumps — Strainers — Removing Sand from Feed Water — Water Chargers — Steam Jet — Steam Pumps — Dayton Cam Pump — The Cope & Maxwell Pump — Dean Brothers' Pump — The Knowles Pump — Auxiliary Pumps — Seller's Injector — Hancock's Inspirator — Schutte & Goehring's Injector — Pratt's Automatic Boiler Feeder — Snowden's Feed Pipe — Moore's Boiler Feeder. Boilers are usually supplied with water by means of a pump or injector. Pumps may be divided into two classes ; power pumps, or those driven by a belt, and steam pumps, or that class in which there is combined a steam cylinder and a pump. So long as the machinery is in motion, a power pump may be operated at a lower cost than a steam pump ; the latter, however, may be operated at any time when steam is on, and is, on the whole, to be preferred even at its greater first cost and subsequent outlay for oper- ating. Whatever device may be selected for feeding boilers, it should be arranged in matters of detail so as to permit a constant feed into the boiler which shall exactly equal the evaporation; but this should not be the limit to its capacity. In selecting a pump, allow one cubic foot of water per hour for each horse power of the boiler; the smallest size for the pump should be not less than twice that capac- ity when running at ordinary speed, and four times the (23) 338 A TREATISE ON STEAM BOILERS. capacity may often be found to be a useful reserve in case of ieaky valves, pipes, etc., which are not only liable to occur, but at a time when it may not be convenient to take the pump apart for repairs. When a power pump is used, a combined lift and force pump is recom- mended ; and, in pumping from a well, the delivery should be into a tank of sufficient size to supply the boiler for at least half a day. This will, in all ordinary cases, allow ample time for any small repairs that may be needed to the lifting or well pump. The force pump may draw the water from the tank and supply the boilers continuously. This Figure 124. arrangement of pumps and tank is not always practicable, especially for very large powers, but whenever it can it ought to be done. The engraving, figure 124, shows a very neat and compact single acting lift or well pump, by Chandler & Taylor, Indianapolis, Ind. It is fitted with leather valves, which are conveniently accessible. The piston is packed with hemp or jute packing. The counter shaft is fitted with tight and loose pulleys, and is thus self contained. POAYER PUMPS. 339 A boiler feed pump by the same firm is shown in figure 125. It is similar in design to the well pump, dififering only in the plunger and valves. The connecting rod pin is near the middle of the plunger, and be- ing always within the bearings of the pump barrel and gland, the plunger needs no other guides. The valves are of metal and suitable for pumping hot water. These may both be mounted on the same base and driven by the same belt. Pumps should be fitted with strainers and foot valves; one similar to figure 126 will be found quite reliable. It is in- tended to screw on the lower end of the pipe and should be placed near the bottom of the well. If a driven well is used there is always likely to be more or less trou- ble with sand in the water for some time after, in which case a device similar to figure 127, by W. and B. Douglas, Middletown, Conn., may be attached to the feed Figure 125. Figure 126. 340 A TREATISE ON STEAM BOILERS. pipe and collect a large portion of the sand in the lower end of the chamber, from which it may be withdrawn by the removal of the plug FlUURE 127. shown in the engraving. Steam jets are often used for supplying a tank with water either from a well or stream, within a reasonable distance. A very simple and convenient device for raising water is shown in figure 128, made by Moore & Kerrick, Indianapolis, Ind. If the well is not deep, it may be placed near or above the ground, but may be carried any distance down the well to keep it within the atmos- It is so simple as to need no special explanation. The quantity of water delivered is regulated by the steam valve, which may be located at any convenient place in the engine or fire room. Steam pumps — Notwithstand- ing the lower cost at which power pumps may be operated, it is still, all things considered, in the inter- est of true economy to use a steam pump instead. The requirements of a steam boiler feeder are. That it shall have no dead cen- ters. That it shall be tdmple in con- struction and durable in service. figure ris. pheric lift of the water. DAYTON CAM PUMP. 341 That the working parts shall be readily accessible. That it will not stop while there is sufficient steam to drive it, and if at rest it shall start at any portion of the stroke. The water valve chambers must be so constructed Figure 129. that by simply removing a cover the valves may be quickly got at for cleaning or repairs. That it shall pump hot or cold water equally well. Dayton cam pump — This pump, shown in elevation in figure 129, and in section in figure 130, is by Smith, Vaile & Co., Dayton, Ohio. Figure 130. By reference to the horizontal sectional view, it will be seen that it is a direct and double acting steam piston pump, having a plain slide valve, similar to the ordinary D valve of an engine. This valve is moved by two levers, A a, on a shaft B, being placed at right angles and form- 342 A TREATISE ON STEAM BOILERS. ing a bell lever Motion is imparted to these levers by a cam C, bolted to the piston rod and working with it, a pin on the lever A working in a groove of the cam C ; also, a fixed support or pocket, S, which holds a sliding V shaped plunger, P, and a spiral steel spring. The operation of the valve movement is as follows : The cam, C, near the ter- mination of the stroke of the pump piston, brings the Y shaped or pointed lever, A, in contact with the Y shaped FlGUBK 131. plunger, P, forcing it back in the pocket, S, and compress- ing the spiral spring contained in the pocket. The move- ment of the piston continues until the points have passed, when the forcible reaction of the spiral spring, and the pressure of the inclined faces of the Y shaped points serve to move the lever, A, and, through it and the small lever, a, to throw the steam valve, Y, sufficiently to partially open the steam port for the return stroke. The same operation is performed on the return stroke at its termination, only the lever. A, is thrown in the opposite direction. The arrangement of the water valves is extremely simple.- The pump, being double acting, there are two suction and two DAYTON CAM PUMP. 343 discharge valves, all contained in one water box on the side of the water cylinder. Bv reference to the sectional view of water box, a clear idea is obtained of the arrangement FlGUUK 132. VALVE MOVING SHAFT. AUXILIARY VALVK. AUXILIARY CYLINDER. AUXILIARY VALVE SE AT. MAIN STEAM VALVE. of the valves, valve seats, stems, springs and plugs, and the manner of putting them in. This pump possesses, what all pumps should, and that is, large steam and water passages. By an inspection of the 344 A TREATISE ON STEAM BOILERS. details of its moving mechanism, lb will be seen that It can never make a short stroke ; each stroke must be completed before the steam valve will open, to cause it to make a stroke in the opposite direction. Figure 131 is a pump fitted with two cylinders, one for hot and the other for cold water. This is a step in the right direction, and can not fail of appreciation. STEAM CHEST AND VALVE SHAFT. The Cope ^ Maxwell pump made at Hamilton, Ohio, is shown in elevation in figure 132, and is, in this illustration, shown as a combined lift and force pump. The details of the valve movement will be understood by reference to the following description : The steam chest, G, is cast in one piece with the steam cylinder head, and has neither bolt, nut, screw or joint of any kind, inside of it or about it, except its cap or cover. On removing the cap the entire valve movement can be lifted out, or any part of it removed and another substi- tuted, ready for instant operation, without requiring to be fitted and without breaking or making a single joint, con- nection or attachment. There are no ports passing through gasket joints; no long crooked or small ports or holes to COPE & MAXWELL PUMP. 345 get stopped up with dirt; no pockets to retain water and incur risk of damage from frost, and no point about it requiring adjustment. . The main steam valve, A, is a flat slide valve, and is cast in one piece with the auxiliary pistons, B B, and the seat, C, of the auxiliary steam valve, E, as shown in engraving. It is moved continuously through the first half of its travels by the power of the main piston, and through the last half by the power of the auxiliary pistons, B B. The auxiliary steam valve, B, is also a flat slide valve, and moved continuously by the power of the main piston. The auxiliary steam cylinders, H H', are composed of two plain hoods or caps, each locking in to a central con- necting piece, D, which holds them in position. They are put together or taken apart without the use of tools. It simply requires to be set down in its place in the steam chest, without further attention, to keep it in place. Being located in the steam chest, it is constantly surrounded by live steam, securing the best possible steam jacket without special provision. The valve moving shaft, F, is a plain shaft extending through the back wall of the steam chest, connected on the outside to the main piston rod by means of lever and connecting rod, and terminating on the inside in a collar provided with two lugs, the lower one locking into the cen- tral or connecting piece of the auxiliary steam cylinders, H H', so they shall move together, and the upper, locking into a recess in the back of auxiliary steam valve so as to give the desired motion to it. Operation — The reciprocating motion of the main piston communicates a rocking motion to the valve moving shaft, F, which, by means of its connection with the auxiliary steam cylinder, moves it back and forth. The main steam Figure 133. DEAN BROTHERS PUMP. 347 valve, A, being cast in one piece with the pistons of the auxiliary cylinder, is moved with the cylinder, and it is so arranged that the main piston in making its full stroke causes the main steam valve. A, to move from its end to its mid position, cutting off both steam supply and exhaust in time to arrest the motion of the piston at the desired point, thus giving almost absolute uniformity to the length of strokes and furnishing it a very superior cushion. Figure 134. Just before the main steam valve reaches its mid posi- tion, the auxiliary steam valve, E, moved by means of its connection with the valve moving shaft, F, reaches the proper position for giving steam to the auxiliary cylinder and motion to its piston. The main steam valve being one piece with this piston is thus carried from its mid position to the end of its travel, reversing the steam and its exhaust ports and the motion of the main piston. Figure 133 shows the pump, as arranged for deep well pumping. The lower or lifting pump may be at any depth below the surface of the ground. The connection between 348 A TREATISE ON STEAM BOILERS. the lower chamber and the base of the pump is ordinary iron pipe, the lifting rod passing up through the center, as shown. Dean Brothers' pump-^This pump is shown in elevation in figure 134, in which a vertical longitudinal section of the water cylinder is also shown, with the valves in place. Figure 135 is a plan of the pump, showing the steam valve and ports, and the opening at the side of the water cylinder for the water supply. Figure 135. It is a simple slide valve engine, combined, in a novel and compact manner, with the best form of a double act- ing pump. Care has been taken that the parts can be readily examined or removed, and all parts subject to wear have means of adjustment. It has but one steam valve. This is a fiat slide valve, which embodies the most favorable conditions for tightness even after the wear consequent upon a long use. It is provided with a fiy wheel, which causes it to run without concussion or jar, turns the centers softly and allows the water valves to seat quietly. The stroke is always the same. This wears the cylinders per- fectly true, and discharges the full amount of water against the heaviest pressure. The Knowles 'pumjp—Thi^ pump is shown in vertical longitudinal section in figure 136. Mr. L. J. Knowles, THE KNOWLES PUMP. 34^ 350 A TREATISE ON STEAM BOILERS. Warren, Mass., was probably the first to introduce the direct acting, positive motion steam pump in this country. This pump has had a very large sale, and has proven a great success. The steam valve is a common flat slide valve, actuated by means of the valve driving piston, both of which are shown in detail in the engraving. The steam valve of the pump being an ordinary flat slide valve, does not have a rotary motion, but simply a horizontal motion, the same as any slide valve. A flat valve embodies the most favorable conditions for securing tightness in the pro- cess of manufacture and re- taining it in constant service. The slight rotary motion im- parted to the valve driving piston, by the rocker arm, simply puts it in a position to be driven horizontally by the steam, in which motion it carries the slide valve with it, both being directly con- nected together. The driving piston is entirely independent of the exhaust- steam for cushioning, thereby working with the same certainty and exactness when exhausting into vacuum (working condensing) as when exhausting into the atmos- phere. It will always start at any point of the stroke, and recent improvements applied to the pump insure entire freedom from jar or pounding under varying conditions of pressure. Auxiliary pumps — In every large manufactory, or in any case where the quantity of water required is too great to Figure 137. AUXILIARY PUMPS, 351 Figure 138, 352 A TREATISE ON STEAM BOILERS. have storage in suitable tanks, there should be an independ- ent source of water supply for use at such times as the steam may not be on the main boilers. This will be found useful in washing out the main boilers during the process of cleaning. It may also serve a useful purpose as a fire engine, by starting a fire in the boilers about quitting time and place it in the care of the watchman during the night, and thus be ready for service at any moment. Figure 138 shows such a boiler and pump as made by the Knowles Steam Pump Works. Should there be a difficulty in starting the pump on account of any defect in the suction pipe, a "charger," as shown in figure 137, may be employed. This should be of sufficient size to charge the water cylinder and fill the ports at least twice, after which there will be no trouble. Injectors — This very ingenious and useful device for sup- plying boilers with water was invented by M. Giffard, of France. It was an innovation on old methods, and at- tracted the attention of engineers everywhere. Among the first to appreciate the value of this invention was the firm of William Sellers & Co., Philadelphia, Pa., who in- vestigated its action, satisfied themselves as to its entire practicability, and at once received the right to manu- facture. It was a fortunate circumstance, indeed, that this instrument, almost unknown and wholly untried in this country, had as its sponsor a firm whose reputation was already well established and who not only gave it their fullest endorsement, but labored diligently to improve it. The earlier instruments were in many respects faulty in their practical workings, but by successive changes and improvements, added year by year, the old injector has almost entirely lost its identity in the new. The principle, however, remains the same. SELLER S INJECTOR. 353 The Seller's injector is shown in elevation in figure 139, and represents the latest improved form. It is known as the ^' Self- Adjusting 1876 Injector/' and will be best explained by reference to figure 140, which is a sectional view of the Figure 139. same instrument. It will be observed that it is self con- tained; that is, there is contained within the instrument itself the necessary steam and check valves required in its ordinary service. The injector is operated by a single movement of the lever H, and its action may be traced through the instru- ment as follows : The steam, water, and boiler connections are indicated in the sectional view, and need no further description. By the movement of the lever H, the cross- head I slides on the guide-rod J, and thus communicates motion to the rod B, which passes through the stufi&ngbox into the interior. A valve W, is secured to the rod B, and (24) i 354 A TREATISE ON STEAM BOILERS. has its seat on the upper side of another valve, X. The receiving tube A, contains both of these valves, and the passage of steam through this tube is prevented or con- trolled by the valve X. By a close examination of the FiGUKE 140. engraving, there v^ill be seen a hollow spindle beginning at W and terminating at C. This spindle passes through the valve X, and may be moved independently of it for a short distance, but by a further movement of the lever H, the valve X is raised fi'om its seat by means of a stop attached to the hollow spindle, formed by an enlargement of the spindle a short distance back of the valve X. Thus the first movement of the lever is to admit steam to the center of the spindle, by the unseating of the valve Wand without disturbing the valve X. It will be understood that what has just been described belongs to the steam side of the injector. SELLER S INJECTOR. 355 The water enters, as indicated by the arrow, into the chamber surrounded by the cylinder marked M M. Inside of this cylinder is a piston ]^ I^, which terminates in a gradually contracting nozzle at a point just beyond C 0. This piston is fitted to slide in the cylinder M M. By a slight movement of the handle H, a jet of steam will issue from the central hole in the spindle and a partial vacuum will be formed in JN" 1^ ; the water will be drawn into this tube, and forced through the delivery tube D. When sufficient water has passed through the instrument to flow "solid" out of the waste-orifice P, then the lever H may be drawn out to its full extent. By this single movement, the valve E, is closed by means of the rod L, and the valve X opened, which will result in a continuous flow of water past the check valve into the boiler. The rod L, shown in connection with the valve R and the lever H, is fitted with two stops, shown at T and Q. When the lever is thrown forward, as shown in the engrav- ing, the valve R is raised from its seat by the screw shown at its lower extremity. When the lever is pulled back, so as to fully open the valve X, the valve R will be closed by the action of the stop T on the rod L. The lever H may now be moved at any point between the stops T and Q without affecting the waste valve ; and b^ the movement of this lever the amount of water to be delivered is regu- lated. The guide rod J is fitted with a number of teeth, as shown. A small click shown at V is hinged to the lever H, and is free to drop into the notches between the teeth in J. When the proper adjustment has been made by the lever H, for the water delivery, the click engages the space below and the injector will continue to deliver a quantity of water corresponding to the area of opening and pressure of steam. The steam supply must be adjusted by the operator; the water supply is self regulating. If too much water is 356 A TREATISE ON STEAM BOILERS. delivered, some of it will escape through O into C, and, pressing on the piston JST 'N, will move the combining tube away from the delivery tube, thus throttling the water supply; and if sufficient water is not admitted, a partial vacuum will be formed in C, and the unbalanced pressure on the upper side of the piston l!^ E" will move the com- bining tube toward the delivery tube, thus enlarging the orifice for the admission of water. From this it is evident that the injector, once started, will continue to work with- out further adjustment, delivering all its water to the boiler, the waste valve being kept shut. By placing the hand on the starting lever it is easy to tell whether or not the injector is working; and if desired, the waste- valve can be opened momentarily by pushing the rod L, a knob on the end being provided for the purpose. These injectors are made in several sizes and numbered 2, 3, 4, etc. These are not arbitrary numbers but repre- sent the diameter of the smallest part of the delivery-tube expressed in millimeters. Thus a N^o. 5 injector means that the tube through which the water is driven in passing thr„ough the delivery tube, is ^ve millimeters in diameter. A 1^0. 8 injector has a tube eight millimeters in diameter; and so for each of the sizes. The non- adjustable injector with fixed nozzle^ non-lifting^ is shown in elevation in figure 141, and in section in figure 142. The latter figure, it will be observed, is reversed in the engraving, but will be none the less easily understood. This injector differs from the one already described in being non-adjustable, and having no valve attached to it. The interior arrangement will be comprehended at a glance, after reading the description of the self adjusting instru- ment. This injector is best suited to localities where it may be operated under practically constant conditions — that is, where the steam pressure is nearly constant at all times. seller's injector. 357 H 7^ a < ^ Ph !^ «1 \^ S3 o tc !ri t? &< o • o a (i* w M ,, &< o z ^ 63 -^ P S O N eu * a ii fe o 5^ ^; i-H rt (M IM T-H — T-( rt ■» T-H O o ■* iM n lO OO CO *"* IM !M CO •*! CO CO o CO 00 00 '^ -* CO -N CO CO CO ■M lO 00 CO 00 (N CD CO r^ CO CO GO CO o ^ lO c^ •* eo r^ t^ CO CO ^^ -# •^ o 1-t c^ "* c~ o '^ 3C CO C5 (M \-~ 00 r~ (N (M -* LO 1^ ^ ^ ■* C>1 ; 00 "^ GC o> c: c^ CO CO o C5 CO (M o c5 CO lO CO a -* CO o5 00 crj o CO CO CO 00 LO 00 o ^ <>; lugs are often rendered in- operative because they become cover- ed over with scale, which may be of suf- ficient thickness to Figure 188. withstand the pres- sure of steam after the metal has been melted below it. SAFETY PLUGS. 421 To avoid the possibility of the safety plug becoming inoperative through the accumulation of scale over it, a device for its protection is shown in figure 79, p. 284. The plug is contained in the pipe 3 at a point shown by an enlargement of the pipe about half way up from the bottom of tlie boiler, and above it a continuation of the same pipe extends upwards beyond the water level, so that as the water never comes in contact with the metal there is no chance of scale ever forming over the metal of the plug. The security against the water falling dangerously low, afforded by the use of the safety plug,^has been fully appreciated, but when improperly made or set they may become a source of danger instead of safety. CHAPTER XVL INCRUSTATION AND CORROSION. Soft and Hard Water — Carbonate of Lime — Sulphates of Lime and Mag- nesia — Incrustation — Prevention and Removal of Scale — Kemp's Boiler Cleaner — Use of Tannic Acid — Starchy C'ompounds — Car- bonate of Soda — Crude Petroleum — Tannate of Soda — External Corrosion — Internal Corrosion — Pitting of Plates — Grooving, Feed water maybe divided into two distinct classes: Fresh and salt; as we have nothing to do with marine boilers in this book, the latter will not be considered. Fresh water may be either soft or hard. By soft water, is meant pure water; that is, water which, upon evaporation, leaves no mineral residue which had been held in solution. It may contain impurities which are held mechanically, such as sand, mud, etc., which may also be removed mechanically by filtering, or percipitation in tanks where the water is allowed to be- come quiescent. The localities in which pure or soft water abounds are not numerous. It is the best water for boil- ers whenever it is practicable to obtain it. Rain water may and often is collected in cisterns for small powers; this is quite practicable and is recommended. Hard water is to be distingjiished from soft in its con- taining in solution salts of lime, magnesia, iron, etc.; the two former being by far the most common. There aro localities in which the sulphates of lime and magnesia pre- dominate, though the carbonates are oftener met with. INCRUSTATION. 423 There are other substances, such as silica, alumina, salt, etc., which are often found in feed water. Carbonate of lime is well known to us under the names of limestone, marble or chalk. Its presence in feed water may be accounted for in this way: rain water falling upon the surface of the ground is taken up by the soil, and in its passage through it absorbs more or less carbonic acid, there present as a result of organic decay. Cold water dissolves about its own volume of carbonic acid, whatever be the density of the gas with which it is in contact; this property decreases as the temperature of the water in- creases, so that at the boiling point it is scarcely percep- tible. When water so saturated comes in contact with lime- stone or marble it dissolves it and w^e liave wells or springs which yield hard water. When this water is heated to the boiling point, the carbonic acid is given off and the car- bonate of lime remains in the boiler. For a time it is held mechanically in suspension, but gradually attaches itself to the boiler, forming a soft scale by allowing it to dry to the shell after blowing out, the furnace walls being still 'hot. Mr. J. M. Allen says, in his annual report* for 1873: "It has generally been supposed that a deposit in a soft state caused little or no injury to a boiler; but our experience has proved conclusively that the contrary is true. The impalpable powder found 'in a boiler, when empty and dry, is mainly carbonate- of lime, and on account of its lightness it is long held in suspension. When the water, from constant evaporation and little or no blowing, becomes sat- .urated with this material, it is rendered unfit for generating steam on account of the resistance offered to the escape of the steam bubbles, and to the free convection of heat. A deposit of slush or sludge col- lects on the bottom, around the seams, and in fire box boilers around the furnace sheets and in the water legs. ' Its presence is detected by leakage at the seams, fractures at the edge of the plates, and in the *lliirtford Steam Boiler -Inspection -and Insurance Co., J. M. Allen, President, Hart- ford, Conn. 424 A TREATISE ON STEAM BOILERS. line of rivets, and by over heating, and consequent depressions of por- tions of the plates where it rests. " This action may be better understood by those who have watched the process of making what is known as "hasty pudding." As the corn meal and water begin to boil the diflficulty which the steam or vapor generated at the bottom has in escaping is manifested by the sputtering manner in which the surface of the mush is thrown about. If vigorous stirring is not kept up, it burns on the bottom, and acts very much as the slush or sludge from lime does in steam boilers. " This difficulty is greatly aggravated if grease finds its way into the boilers; the grease appears to combine mechanically with the car- bonate of lime and sinks on the plates when the boilers are at rest. It becomes a loose, spongy mass, which is not carried off by the circula- tion, but, by its contact with the plates, keeps the water from them, and, by offering resistance to the free transmission of heat, causes over heating and burning of plates. Before we had fully investigated this subject, our opinion was, in many instances, where boilers were leaking badly and showed indications of having been burned, that it was caused by the carelessness of the engineer in starting his fire, with no water in the bailer." What has been said in resrard to carbonate of lime is also in the main true of magnesia. Sulphate of lime is known to nearly every one under its common name, plaster of paris. It is soluble in nearly four hundred parts of water, at a temperature of 95° and almost if not completely insoluble at a temperature of 290°, or slightly more than that corresponding to forty pounds steam pressure. When once found in the boiler, there is no such thing as re-dissolving it by a mere reduc- tion of pressure, as it forms much more rapidly during the day by evaporation than the water will re-dissolve during the night, should the temperature ever get so low as 95°. The formation of scale in a boiler in which sulphate of lime predominates is quite irregular; being more than twice as heavy as water, it can not long remain in suspen- sion, and is found in deposits of varying thickness, the INCRUSTATION. 425 hardness varying according to the substances in combina- tion with it, and the heat to which the whole may have been subjected; and forms, perhaps, the most troublesome scale that the steam user has to contend with. Carbonate of magnesia will be found in feed water in localities in which magnesian limestone abounds. It is not usually found in any great quantity, as compared with the other salts named. In its behavior in the boiler, it is not unlike carbonate of lime at similar temperatures. There are a number of other substances which act in a manner analogous to those already referred to, but as they are in general so small in quantity, when compared with the others, they need not here be particularly described. The impurities in feed water, when consisting of salts of lime and magnesia, produce incrustation; when an excess of acid is contained in the water, corrosion takes place. Incrustation, when allowed to form in any boiler, has the effect to reduce its steaming capacity and also induces over heating of the plates, by reason of its being a non- conductor of heat Its presence also prevents a satisfactory internial examination of a boiler, as it covers the joints and other portions which should be laid bare to make the inspection thorough. Many attempts have been made to calculate the loss of heat by the accumulation of scale. The results show great loss, but exactly how much is not entirely known; it is placed by different observers as follows : ^ inch thick requires an increase of 15 per cent in fuel. i inch thick requires an increase of 30 to 60 per cent in fuel. J inch thick requires an. increase of 60 to 150 per cent in fuel. This last line is given for what it is worth. 426 A TREATISE ON STEAM BOILERS. The prevention and removal of incrustation is a subject which interests every one having a boiler fed with hard water. The usual means of prevention and removal are, by blowing off; the use of chemical agencies which render the impurities more soluble ; the use of some mechanical device, fitted to the interior of the boiler, which will col- lect the deposit and which may then be removed, cleaned and replaced. Blowing off is the easiest and readiest method of get- ting rid of surface impurities which will prevent the free escape of steam at the surface of the water. There are many devices for this purpose; the one described below is said to yield excellent results by those who have them in use. Kemp's boiler cleaner^ as manufactured by James F. Hotchkiss, Bay City, Mich., is shown in figure 189, as attached to the boiler when in use. A is a box or reservoir located above or upon the arch wall of the boiler. lu marine boilers the reservoir may be suspended from the deck frame above; from this reservoir three pipes extend; the first pipe, B, enters the rear part of the top sheet of the boiler or generator, and is con- nected with a horizontal pipe, which is adjusted a little below the water line. At either end of this horizontal pipe is an enlarged mouth, C, partly submerged, but extending a little above the surface of the water — the mouths being of a diameter to allow several inches varia- tion in the water line. The second pipe, D, leading from the reservoir A, enters the other end of the boiler, in sim- ilar manner, terminating below the water surface. When the boiler is heated, a constant current of water is immediately established through the bell mouth C, and pipe B, filling the reservoir A, and, cooling to a certain extent, it returns to the boiler by the pipe D. It will be KEMP S BOILER CLEANER. 427 observed that the up flow pipe is placed about midway between the fire bridge and the back end of the boiler, at a point where the water is presumably hottest. On the other hand, the down flow pipe enters the front or cooler portion of the water, and, while the water may rise and fall in the boiler to any moderate extent, the enlarged mouths, C, will constantly maintain a current free from steam, from the surface. As the sediment and impurities are chiefly sep- arated from the water by figure 189. ebullition in that part of the boiler where the horizontal pipe C is located, they are immediately drawn in by the current and carried into the reservoir A; here, the current, weakened by expansion, can support the impurities no longer, and they settle in the reservoir, and are retained, until blown off* through the third pipe E, as seen in the engraving. The reservoir may be located at any desired point above the level of the water line, as most conven- ient, and occupies no appreciable room. It usually holds about three gallons of water. When the boiler is in use the stop cocks should always be left open. To wash the reservoir out, open cock E for about half a minute once each day, or as often as necessary. This is simply all the attention required in a general way. By placing a vessel under the blow off* pipe, the amount of deposits can be easily ascertained. In severe cold weather, where the boiler is exposed and allowed to stand unused a day or more at a time, shut the 428 A TREATISE ON STEAM BOILERS. stop cocks D B, and open cock E, leaving the reservoir empty; but so long as the water is warm in the boiler, it will circulate through the reservoir and keep it from freezing. The use of chemical agents seems to be a favorite one for the prevention or removal of scale. There are hundreds of "boiler compounds," some of them of surpassing excel- lence, others perfectly useless, if not positively hurtful. Nearly all compounds for this purpose have either tannic acid or soda as the active agent. Yegetabie matter does not, as a general thing, act injuriously on the plates and if it contains any considerable quantity of tannic acid, it may prove of great value in arresting or preventing incrusta- tion ; and for this reason caoutchouc (crude india rubber) nutgalls, logwood^ hemlock, mahogany, etc., are often employed as scale preventives. The chemical action of the tannic acid is to decompose the carbonates in the water, and thus to form a tannate of lime, instead of a car- bonate, and in the same manner for carbonate of magnesia. Tannates of lime and magnesia are not soluble, and being of light specific gravity, are held mechanically in suspen- sion by the circulation of the water. These particles of tannate of lime or magnesia floating in the water do not have a tendency to unite to form masses by their adhesion, and may easily be blown out of the boiler with the water. This chemical reaction does not occur in water contain- ing sulphate of lime, and for this reason an analysis of the water should be had before using any compounds having tannic acid as the principal ingredient. Molasses, potatoes, vinegar, etc., etc., have found their way into boilers as preventives. The latter containing acetic acid, decomposes the carbonates of lime and mag- nesia, forming acetates, which, being soluble, are kept in solution and do not form scale. There is danger, however, INCRUSTATION. 429 that an excess of free acid will act injuriously on the plates. In regard to potatoes or any starchy compounds, they may and often do prove serviceable. Their action seems to be mechanical, and to envelope the solid particles of lime and prevent adhesion. Scale has not only been prevented, but actually removed, by simply using potatoes in the boiler. It should be borne in mind that starchy ingredients, of whatever kind, have a tendency to produce frothing, and thus to deceive the fireman as to the correct water level. These have little or no action on sulphate of lime, but are to be confined to water having carbonates only. Carbonate of soda, caustic soda, potash and other fixed alkalies have been used with varying results. These will decompose sulphate of lime and form sulphates of soda or potash, which will be retained in solution and the carbon- ate of lime precipitated. Alkalies do not injuriously attack the plates of the boiler, and may often prove beneficial in neutralizing the effect of free acids present in the water. Petroleum has been used with marked results. Mr. Allen says, in the same report already quoted from, "We have a specimen of scale in this office nearly one and a half inches thick, that was removed from a boiler in the west by crude petroleum, or what is known as unrefined, black, earth oil. 1 am aware that there is great prejudice against using anything of the kind in steam boilers, but earth oils are very different from animal oils. They are very volatile, and in an experience of several years where hundreds of boilers have been treated with it, we have found no injury to plates or tubes, and the boilers have been kept free from scale. Petroleum works better where sulphate of lime predominates, than in waters impregnated with carbonate of lime. We would not advise it in connection with the latter. I desire to impress upon all persons the importance of careful attention to their boilers when sol- vents of scale or purgers are used. It often happens that scale is thrown off and allowed to accumulate on the bottom of the boiler, and from want of attention, not being removed, the boiler becomes 480 A TREATISE ON STEAM BOILERS, burned and nearly or quite ruined. If a purger is used, the boiler should be often opened and as often thoroughly cleaned." Tannate of soda is a compound which, it is claimed, will hold both the carbonate and sulphate of lime in solu- tion. This compound was finally decided upon after a laborious chemical research, by Jos. G. Rogers, M.D., Madison, Ind. Its action may be described thus: Tannate of soda decomposes the carbonates of lime and magnesia as they enter, tannates being precipitated in a light, floc- culent, amorphous form, so that they do not subside at all in the boiler, but are retained in suspension by the boiling currents until they find their way into the mud receiver, where they settle into a loose, mushy mass, 'which may be easily blown out from time to time. The carbonate of soda, formed in the reaction, is retained in solution, becom- ing a bicarbonate by appropriation of the free carbonic acid in the water. This decomposes the sulphate of lime, the resulting sulphate of soda being retained in solution, and the carbonate of lime being acted upon by fresh por- tions of the tannate of soda as above. The constant pres- ence of the alkali protects the iron from all action, either of the carbonic or tannic acids. The same reaction takes place between the tannate of soda and the already existing scale, with like results, but more slowly, some weeks being generally required, in practice, in removing the deposit, if it exists in any considerable quantity. Zinc has been used with some success with water con- taining bicarbonate of lime; it has no eftect on sulphate of lime. The disappearance of zinc in the boiler is no indi- cation that it is preventing scale, as it may be reduced by galvanic action. This portion of the chapter might be almost indefin- itely extended, and be of no practical use when concluded. CORROSION. 431 The writer has given in brief outline the action of the commonest and perhaps the best anti-incrustators. One thing must be done before any one of the several substances named can be recommended; that is, an analysis of the water, showing whether it contains carbonates, sulphates or acids, which of each, and in what quantities. Then and only then, can an intelligent recommendation be given. The writer's advice to boiler owners is that, no preparation be bought or used until after such analysis by a competent expert or chemist. External corrosion is frequently caused by the exposure of the shell of the boiler to the weather. It often occurs that boilers have no other protection than simply a loose board roof which, even in ordinary rain storms, leak at every joint. If the boilers were always under steam thei bad consequences would be comparatively light, but the greater mischief occurs when the boilers are cold. When- ever rust appears on the surface of a boiler it means loss of iron, loss of strength, and consequently is less able to withstand high pressure. The danger is increased if the action of the rust be confined to certain portions of the boiler and continuous deterioration be going on ; this is likely to occur along the line of brick work, in externally fired boilers, near the water line. In exposed situations this rate of corrosion may amount to one-sixteenth inch in a single year. This is perhaps exceptional; but a boiler would soon be rendered worthless if only half that waste of iron was going on and which is not at all exceptional. Another source of corrosion is that which proceeds from leaky joints, either from the riveted seams, man or hand holes, or from imperfectly fitted attachments. When the leak occurs around a rivet a new one should be put in ; if in the seam, it should be re-calked inside and out. The gaskets used between the shell and hand, or man hole 432 A TREATISE ON STEAM BOILERS. plates, are often so imperfectly fitted that it is the excep- tion to find them perfectly tight. A gasket of vulcanized rubber, say three-sixteenths to one-fourth inch thick, is recommended rather than a plaited one made of hemp, when used with high pressures. A gasket recently brought to the attention of the writer is shown in figure 190, and is the invention of Mr. C. S. Stoy, Butler, Indiana. It consists of a thin copper shell, filled with pack- ing, as shown. There is little doubt that it will make an excellent joint and may be used over and over again. But whatever packing may be used the joint must be tight, and a leak, however trivial around a boiler, FiiiUKK lyo. should be immediately repaired. Internal corrosion — This wasting of the plates is doubt- less caused, in the main, by the action of acids in the water. It has also been attributed to galvanic action. A noticeable thing in connection with internal corrosion is its want of uniformity; its appearance is not unlike ordi- nary rusting and is not difficult to detect. The following extract is from Mr. Allen's report for 1875, and agrees with the writer's own observations. I am also indebted to Mr. Allen for the accompanying engravings: "The work of corrosion is insidious, whether it is external or internal. A boiler that is set in brick work may leak at the seams, and corrode the plates adjoining, and yet there may be no indication of danger. So, by the use of impure water, a very dangerous process may be going on inside the boiler. 1 n boilers covered more or less with scale its presence is often detected by red streaks where the scale is cracked. It attacks the edges of plates at the joints, and around the rivet heads. Sometimes it will attack two boilers working side by side. One will be corroded in the front part, and the other I CORROSION. 433 in the back part. Sometiraes different sheets in the same boiler will be corroded, while others remain intact. Again, boilers will be found in what is known as a pitted condition. This is manifested by small spots in close contact, being eaten into the sheet. It looks like a^ pock marked face and is sometimes confluent; and what is strange about this is, that often certain sheets in the boiler will be attacked while others will remain clean and smooth, and the iron will bear the -^s^sNvNNv .- --^^^^^^ \.. "":"2i Figure 191. Section on A B. same brand on each plate. It is well known that iron ore, even from the same mine, is not always chemically the same; certain impurities will be found in some places which do not exist in others. And in the manufacture of boiler iron there is no doubt but that the sheets are chemically slightly different; hence, when the boiler is constructed the presence of acids in water may excite galvanic action. This would account for the different manner in which boilers are affected. The following figure will illustrate the effects of corrosion: "This was discovered by inspection. The outer and inner side of the sheet is shown in the drawing ; also a cross section. The hole in the center of the sheet was made by the inspector's chisel. The iron was little thicker than paper ; the piece of plate can be seen in this office. (29) 434 A TREATISE ON STEAM BOILERS. Figure 192 represents a portion of the inner plate of a water leg of a locomotive boiler. Corrosion attacked the plate around the stay bolts, as represented by the radiating dark lines. The other end of the stay bolts' was eaten nearly off. We judge from appearances that in tapping out the holes for the stay bolts, a strain was brought to bear which disturbed, the fiber of the iron, or perhaps I should say the skin of the iron ; imperceptibly, however, to the casual observer. Figure 192. "The difficulty was further aggravated by the unequal expansion and contraction of the two sides of the water leg. The inner sheet forming the side of the fire box was subjected to greater heat, and this continual, though imperceptible action assisted in increasing the difficulty. " Impure water found this disturbed point the most open to attack, and the result is as we find it here. The furrows are eaten in quite deep, and it looks like the work of a tool." Pitting — When internal corrosion occurs in isolated spots it is called pitting. This generally occurs near the joints of the plates, but not unfrequently directly on their faces. It is not uncommon to find pitting up in the steam dome, and from the fact that it is as likely to occur away from as in contact with the water, it is now generally believed to be due to galvanic action. GROOVING. 435 Figure 193 is loaned by Mr. Allen to show the peculiar corrosive effects found in a boiler fed by swamp ivater, and is engraved from a sample now in the office of the Hart- ford Steam Boiler Insurance and Inspection Company. Gi'ooving— There is not a well defined and satisfactory explanation which will wholly account for this destructive Figure 193. action in boilers. It is believed, however, that it is caused by the constant changes of form which take place in a boiler by the alterations of pressure, and thus induce a hinging or buckling action of the plates, particularly along the lines of riveted joints. In the ordinary method of making boilers, it is impossible to make a shell perfectly round, and when such a boiler is subjected to steam pres- sure the tendency is to make it a true cylinder, and this is the cause of the buckling or hinging above referred to. If the plates are made of fibrous iron, they are loosened every time this occurs, and it is greatly aggravated by the 436 A TREATISE ON STEAM BOILERS. continued changes of temperature to which the whole is exposed. The iron being less firm at this point than else- where, corrosion becomes all the more easy and certain, and is further assisted by imperfect or too rigid bracing at certain points, and too slack at others. It is not now, as was formerly, believed to be due to galvanic action. CHAPTER XVI I. SECTIONAL BOILERS. The Babcock & Wilcox Boiler— The Boot Boiler— Kelley's Boiler— The Firmenich Boiler. Sectional boilers are not as yet common property, and the writer can only refer in a general way to the details of construction as practiced by the several makers, whose idesigns may be described. No attempt can be made in a single chapter to trace the development of sectional boilers, nor to describe all those now in the market ; three or four examples will be given to illustrate the present practice of the leading manufacturers. The Babcock Sf Wilcox boiler is shown in front elevation in figure 194, in longitudinal sectional elevation in figure 195, and in cross sectional elevation in figure 196 ; these three engravings, for convenience of reference, are printed on the same sheet. This boiler is composed principally of lap welded w^rought iron tubes four inches in diameter, arranged in sections, having seven or eight tubes in each. These sec- jblons are inclined at an angle of about 15°, as shown in the sectional elevation. These sections are connected with each other, and with a horizontal mud drum at the bottom, and rear of the boiler, and also by vertical passages at each end, with the two horizontal steam and water drums shown in the cross sectional elevation. The water fills all the tubes and extends half way up and into the steam and water drums. THE BABCOCK & WILCOX BOILER. FlQURE 196. 438 TREATISE ON STEAM BOILERS. The end connections are in one piece for each vertical row of tubes ; these tubes do not lie one above the other in a vertical line, but are staggered so as to receive heat either by radiation or direct impact by the flow of heated gases upwards through the spaces between them. These tubes are not threaded, but secured to the end connections by the use of an expander in a manner similar to that in which tubes are fixed in an ordinary tubular boiler. The connections between the mud drum and the steam and water drum are made in the same manner, and thus dis- pensing entirely with bolted joints and packing. By this arrangei;nent are secured freedom from strains induced by unequal expansion, and a means of rapid and thorough circulation of water. The water inside of the tubes when heated has a tendency to rise toward the higher end, and flows upward and into the steam and water drum. The steam is here given ofi*, if the water is of suflSciently high temperature. The back connection secures a downward current, and thus a continuous circulation is established, preventing the evils arising from the destructive strains consequent upon unequal temperatures. This rapid circu- lation also prevents, in a measure, the formation of scale upon the heating surfaces, sweeping the particles away and depositing them in the mud drum, from which they may be blown out. The provision for cleaning the boiler, both internally and externally, is quite complete. Hand holes opposite each end of each tube, man holes in the drums, and a bon- net to the mud drum, permit access to all parts of the inte- rior, while side doors admit of the removal of accumulated dust and ashes from the exterior of the heating surfaces, either by blowing, brushing, or any other well known means. The proportions of this boiler were adopted after numerous experiments with boilers of varying capacity; BABCOCK & WILCOX BOILER. 439 and experience has established that it can be driven to the utmost, and still be free from the objections always attach- ing to boilers of small capacity — carrying a steady water level and steam pressure, and always furnishing dry steam. The cubical capacity of this boiler, per horse power, is equal to that of the best practice in tubular boilers of the ordinary construction. The fire surface being of the most eft'ective character, these boilers will, with good fuel and a reasonably economical engine, greatly exceed their nominal power, though it is seldom economy to work a boiler above its nominal power. The space occupied by this boiler and setting is equal to about two-thirds that of the same power in tubular boilers. The following is an abstract of a test of a Babcock & Wilcox boiler, by Charles E. Emery, C.E., l^ew York city. This test was made in February, 1879, at the Raritan Woolen Mills, Earitan, E". J. There were two boilers in use, containing 4,080 square feet of heating surface, and 103 square feet of grate surface, the capacity of the two boilers being rated jointly, by the makers, at 360 H.P. The experiment commenced at 6.01 a. m., and closed at 6.38 p. M. In starting, steam was raised by spreading the banked fires left from the previous day. When the pres- sure reached 80 pounds the fire was hauled, all refuse removed, and fires started anew with wood, which in the cal- culation has been considered equal in calorific value to -^ its weight of coal. The fires were maintained with coal during the day, finally hauled, allowed to cool, the com- bustible portion deducted from the coal charged, and the refuse weighed separately. The experiment was closed when the boilers stopped making steam at 80 pounds pressure, with water in the glass gauges at same height as at starting. During the trial, all the coal consumed was weighed in an iron wheelbarrow, balanced when empty by a fixed 440 A TREATISE ON STEAM BOILERS. weight, and each barrow load was adjusted at the scale to weigh 200 pounds net. All the water evaporated was measured in a tank provided with a heavy float connected through a fine chain to an index showing the water level on an exterior scale, divided decimally. By weighing water out of the tank, its capacity was found to be 5,172 pounds of water between the limits employed. A complete record was kept of the coal, water, steam pressure and various temperatures, and the quality of the steam was tested with a calorimeter at frequent intervals. The proprietors of the mill took the proper business pre- caution of stationing observers at each point, who kept entirely independent records, agreeing with those taken by the assistants. The coal used was clean nut coal from the Lackawanna region. It had been exposed to the weather during the winter, and when first taken from the pile was wet, but a sufiacient quantity for the trial was brought under shelter a few days in advance, so that the coal actually used was bright and appeared dry. The results of the trial are as follows : Average steam pressure 71.63 Average temperature of fire room 44.00 Average temperature of water in feed tank 90.47 Average temperature of water entering boiler after passing through a heater in flue 110.59 Average temperature of up-take boiler No. 1 by pyrometer (evidently wrong) 381.87 Average temperature of flue beyond feed water heater 453.23 Wood used in starting fires, 750 pounds, equivalent of coal (730x .4) 292 pounds. Coal put in furnaces during experiment 19,827 pounds. Total of above 20,119 pounds. Combustible in refuse at close of experiment 820 pounds. Total coal consumed, including equivalent of wood 19,299 pounds. Refuse from coal removed during experiment 749 pounds. Refuse from coal at close of experiment 2,134 pounds. Total 2,883 pounds. Actual percentage of refuse (2.883 -^ 19,299 x 100= ) 14.94 per cent. Combustible consumed (19,299—2,883=) 16,416 pounds. Coal Avith 12 per cent refuse agreed upon equivalent to that actually consumed [16,416-^ (100—12) = ] 18,654.5 pounds." THE ROOT BOILER. 441 Total weight of water actually evaporated at i)ressure of 71. C3 pounds from temperature 110.59° 101,573.28 pounds. Equivalent evaporation at pressure of 70 pounds from temperature of 180^, as agreed upon 172,592.58 pounds. Evaporation per pound of coal, with 12 per cent of refuse, at presureof 70 pounds, from temperature of 180^ 9.252 pounds. Evaporation per pound of combustible, atmospheric pres- sure from temperature of 212^^ 11,221 pounds. On the basis that ain^ good engine, under fair condi- tions, will require hut 30 pounds of water per hour for horse power, these boilers developed 464 H. P., or 104 horse power in excess of that required bj the contract. 7^he Root boiler manufactured by the Abendroth & Root Manufacturing Company, Brooklyn, New York, is shown in sectional elevation in figure 197, and in front elevation in figure 198. By reference to the engravings the construction of this boiler will readily be understood. The tubes are four inches in diameter and inclined as shown in the sectional elevation. These tubes are screwed into cast iron caps as shown in figure 199, in which A is the tube, B the B^^ cap and E a stud by which a [ B J ^--' — x c,^ x x.~xx. <^^ ^ triangular elbow is held in nitl^^Tl r__^^^ - - -- ^ place; this elbow is shown in ^ ^J|^^S^:^^^^^ elevation in fie^ure 201, and in w^mmzSa'^ — ^^--- --..v,,^^ partial section in figure 200. figure 199. It will be observed that each tube screws into a square end cap in which are three openings. These caps are exactly alike, except those in the lower row, which differ in size and require a modified elbow connection, as is clearly shown in the sectional eleva- tion. The manner of securing the trian- gular elbow J) is shown in the enlarged FIGURE 200. engraving, figure 201. The object of this square end cap and triangular elbow is to permit a free circulation of water through and exit of steam from the 442 A TREATISE ON iSTEAM BOILERS. tubes. The steam is collected from the tubes into the steam drum S, which is placed over the top of the boiler, as shown. This form of construction afiords ready facility for cleaning or renewal of tubes. The water should not be carried too high if there is a severe drain on the boiler, as it is likely to induce priming. When steam is generated in the tubes, it rises to the steam drum through the triangular elbows (return bends would perhaps be nearer correct), seen at the front end of the boiler. In pass- ing; throu2:h the bends the current is broken, and if any water is min- 0^1 ed with the steam it is thrown back into each tube. This breaking and reversal of the current to prevent priming is now gen- erally recognized as the correct way to do it, and no doubt is the explanation of the remarkable freedom which this boiler has from this troublesome and dangerous occurrence. If the demand upon the boiler is constant, a steam drum need not be supplied, but when the requirements are irregular, it is then perhaps best to have it. The grates may be of any of the common or oscillat- ing varieties now in the market. The course of the products of combustion are clearly shown by the' direction of the arrows. The tubes have what the makers call '' bridge wall blocks," which are shown at A. These are built up to any required height, FiGUKK 201. ,*f THE ROOT BOILER. Sectional Elevation. Figure 197. THE ROOT BOILER. Front Elevation. Figure 198. 11 kelley's sectional boiler. 44B to insure every part of the tubes being in contact with the heated gases. The following test was made at the Centennial Exhi- bition, Philadelphia, Pa. : Heating surface, square feet 1,590. Grate surface, square feet 42. Coal used (anthracite), pounds 3,053.9 Ashes, pounds 320.2 Steam pressure, pounds 69.94 Temperature of feed water, degrees 64. 59 Water evaporated by calorimiter tests, pounds 27,146.69 Water evaporated per one pound of coal from tem- perature of feed, in pounds 8.89 Water evaporated per one pound of coal from and at 212°, pounds 10.35 Water evaporated per one pound of combustible from temperature of feed, pounds 9.93 Water evaported per one poured of combustible from and at 2 1 2°, pounds 11. 565 Water evaporated per square feet of heating surface from temperature of feed, pounds 2.22 Water evaporated per square feet of heating surface from and at 212°, pounds 2.48 Kelley's sectional boiler, by William E. Kelley, New Brunswick, E". J., is shown in sectional elevation in figure 203, and in detail showing the manner of securing the tubes and the circulation of the water through the boiler, in figure 202. The tubes in this boiler are 3 inches in diameter, and screwed into the vertical chamber, as shown in the detailed engraving. These tubes, with the exception of those in the upper row, are inclined at an angle of about one to eight, and connected at one end only, and that at the vertical chamber. The tubes are therefore left free to expand separately without afiecting others, and in like manner may be removed for examination or repair. A cap is placed at the back 444 A TREATISE ON STEAM BOILERS. end to close each tube. Inside of these tubes are placed partition plates, as also shown. The inclined tubes are always full of water, the water line of the boiler being at W. L. The heat from the fuel Figure 202. on the grate in rising first comes in contact with the lower half of the inclined tube, the upper half of the tube being in a measure shielded or protected from the direct heat or flame by the lower half; consequently tjie greatest amount of steam will be generated from the surface of the lower half of the tube. The steam thus made would rise to the upper side of the tube, were it not intercepted by the par- « THE KELLEY SECTIONAL BOILER. FIGURE 203. kellby's sectional boiler. 446 tition plate; this causes the steam to move along the under side of the partition plate, and along the outside of the pocket in the front chamber, and thence into the dome, first passing through the horizontal pipe, as will be here- after explained. The upper half of the tube, as before stated, will be less exposed to the direct action of the fire; hence the water will flow down the coolest part of the tube, and through an opening, D, in the rear end of the partition plate, and thence up through the lower half of the tube, as before stated, the circulation being accelerated by the volume of steam that is seeking an exit from the lower half of the tube into the front chamber. The pockets in the front chamber tend to keep the downward and upward currents separate. The arrows on the cut indi- cate the direction of the circulation. The free exit of the steam from the front chamber into the dome is obstructed by a partition running entirely across the chamber, near the top and above the water line, W. L., and the steam is compelled to pass along under the partition in the horizontal pipe, and through the opening in said partition, and then along over it, through the upper half of the tube into the chamber and dome ; as the tubes are in the heat and above the water line, the steam is made very dry, and all moisture and water that would otherwise be carried out of the boiler is converted into steam in passing through the horizontal tubes. The following test was made at Philadelphia, during the Centennial Exhibition : Heating surface, square feet 662. Grate surface, square feet 27.5 Coal used, anthracite, pounds 2,380.95 A shes , pounds 204. 5 Steam pressure, pounds 69.95 Temperature of feed water, degrees 66.95 Water evaporated by corrected calorimiter tests, pounds 18,710.53 446 A TREATISE ON STEAM BOILEKS. Water evaporated per one pound of coal from temperature of feed, pounds 7.858 Water evaporated per one pound of coal from and at 212°, pounds. 9.139 Water evaporated per one pound of combustible from temjDerature of feed, pounds 8.636 Water evaporated per one pound of combustible from and at 212°, pounds 10.94 Water evaporated per square foot of heating sur- face from temperature of feed, pounds 3.52 Water evaporated per square foot of heating sur- face from and at 212°, pounds 4.13 The Firmenich boiler^ by J. Gr. & F. Firmenicli, Buf- falo, l!^. Y., is shown in sectional elevation in figure 204. This boiler consists of two partially cylindrical wrought iron shells at the bottom, separated sufficiently to admit the requisite width of grate. From the upper and flat side of these lower cylinders or mud drums, pipes extend upward and are connected to similar drums at the top. Surmounting these two upper drums is still another, having suitable connections with the two lower ones, and acts as the steam drum or reservoir for the boiler. The lower drums vary in size from twelve to twenty- four inches, and the upper steam and water drums, from twenty to thirty-six inches diameter; the lengths varying according to the size of the boiler. The vertically inclined heating tubes are from two to three inches in diameter, the latter size being used in all boilers over twenty-five horse power. This boiler offers good facilities for exter- nal and internal examination. Figure 204 is a cross sec- tional elevation, showing the arrangement of drums and tubes. The following figures are taken from the report of the economy trials at Philadelphia during the Centennial Exhibition: THE FIRMENICH POILEH, 447 Heating surface, square feet 1,078-88 Grate surface, square feet 15.84 Coal used (anthracite), pounds 1,482.35 Ashes, pound? 153.25 Steam pressure, pounds 70.06 Temperature of feed water, degrees.. GS.94 Water evaporated, calorimiter tests, pounds 13,23;^. 6 Water evaporated per one pound of coal from temperature of feed, in pounds 8.93 Water evaporated per one pound of coal from and at 212°, pounds 10.34 Water evaporated per one pound of combustible from temperature of feed, pounds 9 95 Water evaporated per one pound of combustible from and at 212°, pounds 11.53 Water evaporated per square feet of heating surface from temperature of feed, pounds 1.33 Water evaporated per square feet of heating surface from and at 212°, pounds 1.775 The tubes in this boiler being nearly vertical, pre- vents the accumulation of soot or ashes on them. The combustion chamber is of unusual dimensions, and bj properly arranging for the admission of air the combus- tion should be complete and a very high temperature main- tained at all times. Those who have this boiler fjgurk 204. in use, speak well of it and claim a saving in fuel over same evaporation with ordinary boilers. the INDEX, PAGE Abbott & Co 260 Abendroth & Root 441 Absorption of heat 187 Acetic acid in boilers 428 Acids in feed water 432 Action of fire on plates 141 Adatnson, D., on Bessemer steel 49 Adam son, D., on Siemens-Martin steel.. 64 Adamson's joint for flues 260 Addition to factor of safety 152 Advantages of steel for boilers 30 Air forcing apparatus 326 Air required for grate surface 315 Air, resistance of through pipes 315 Air-space boiler covering 391 Albright & Stroh 309 Allen, William & Sons 383 Alumina in water 423 American boiler plate, strength of 17, 68 American Linen Co 320 American Steam Gauge Co 407 Analysis of Bessemer steel 50 Analysis of crucible steel 41 Analysis of iron plate 16 Analysis of open-hearth steel 57 Analysis of Workington iron 57 Annealing 143 Annealing after punching 106, 109 Annealing east iron 12 Annealing flanged heads 139 Annealing thick steel plates 107 Anti-incrustators * 431 Arching back connections 308 Area, reduction of by tests 77 Areas of lap-welded tubes 247 Area of tubes for vertical boilers 276 Asbestos covering BSO Ashcroft, E. H 2o4 Ashcroft safety valve 399 Atkinson, George H 23 Atlas Engine Works 296 Automatic boiler feeder 309 Auxiliary pumps 350 Babcock & Wilcox boiler.... 437 Babcock & Wilcox economizer 388 Back plates 308 Baffle plates 282 Baxter boiler 274 Bead on the ends of tubes 177 Bending of joints under stresa 102 Bending tests 66 Bending wrought iron 21 Bertram's experiments, riveted joints.... 96 PAGE Bessemer pig analysis 45 Bessemer steel 43 Bessemer steel analysis 49 Bessemer steel blooms, tests of 50 Bessemer steel boiler plates 46, 51 Bessemer steel, elastic limit of 80 Bessemer steel, limit to T. S 68 Bessemer steel rivets .' 120 Best iron, strength of 67 Bituminous coal, rate of combustion 247 Blast gates 314 Blast nozzle 312 Blast pipes, sizes for 315 Blisters in wrought iron 141 Blow holes in castings 11 Board of Trade, English 151 Boiler feeder, requirements of 340 Boiler rests 308 Boiler setting 304 Borden Cumberland coal 320 Borntraeger, H. W 51 Bourdon's pressure gauge.... 406 Bowling iron 69 Boyd, William 106 Bracing cast iron boilers 14 Bradley's crown iron 69 Breaking and tearing samples 72 Breaking strain vs. quality 84 Brittle boilerplates 17 Brittle iron, punching of 92 Brown, Aug, P 419 Brown's iron and steel (.Eng.) 97 Bulging tests 81 Bulging tests of steel 55 Burning steel 119 Butman, T. R ^^20 Butt joints, strength of 103 Cadman, A. W. & Co 419 Calking 125 Calking chisel 125 Calking, Connery's 126 Calking flues in place 132 Calking, grooving caused by 130 Calking tool for tubes 179 Cambria Iron Co 221 Camel back boilers 302 Cammel's iron and steel 97 Caoutchouc for scale 428 Carbon in boiler steel 30 Carbon in castings 10 Carbon in iron 5 ("arbon in steel 38, 53 Carbonate of magnesia 425 INDEX. 449 PAGE Carbonate of soda for scale 429 Carbonic acid gas and iron 134 Carbonic oxide gas andiron 134 Castings not uniform 9 Castings, qualitj' of 12 Cast iron, corrosion of.. 8 Cast iron, elastic limit of 11 Cast iron, factor of safety in 11 Cast iron, flaws in 9 Cast iron for boilers 8 Cast iron, how affected by heat 142 Cast iron, strength of 11 Caustic soda for scale 429 Chain and zigzag riveting 109 Chain joints, strength of Ill Chalmeis-Spence Co 391 Channeling by calking 128 Chandler & Taylor 238 Charcoal plate irons 22 Chemical removal of scale 428 . Chilled iron 7 Chimney draft 312 Chipping seams 126 Chipping the ends of tubes 177 C. H. No. 1 iron 22 C— iron 22 Cinder in boilerplates 142 Cinder in homogeneous iron 27 Cinder in wrought iron 20 Cinder prevents welding 134 Circulating generator. Steads' 384 Circulation and convection 213 Circulation and locating tubes 241 Circulation in boilers 370 Circulation in economizers 388 Circulation in vertical boilers 283 Circulation of water 212 Classitication of wrought iron 19 Cleaning fires 309 Clearance in punch and die 90 Concave calking 126 Coal required per hour 246 Cohesion of iron affected by heat 140 Coil heater 375 Cold feed water, injury by 370 Cold forge tests of steel 55 Cold punched nuts 90 Coldshort iron 10, 16 Collapse of boiler flues 132 Collapsing pressure 168 Color heat 144 Colts' Patent Fire Arms Manufacturing Co 274 Combined safety and stop valve 397 Compound tubular boiler 254 Compressing steel ingots ,.,, 34 (80) PAGE Compression gauge cocks 418 Conducting power of metals I89 Conduction 189 Conduction of heat by liquids 190 Conduction, resistance to 196 Connery, J. W 126 Consett, Best Best iron 69 Consolidated Safety Valve Co 399 Construction of boilers 2 Contraction of area in samples 85 Contraction of area recommended 78 Contraction, strains produced by 371 Convection 186 Convection and circulation 213 Convection of heat.., 190 Cooling strains in cast iron 9, 11 Cope & Maxwell 344 Copper 2 Copper boilers 196 Copper in steel 37 Copper, transmission of heat through.... 196 Cornish boilers 257 Cornish boilers, H. P. of 211 Corrosion 132 Corrosion, external 431 Corrosion, how detected 432 Corrosion induced by strains 443 Corrosion, internal 432 Corrosion, rate of 431 Covering boilers and pipes 389 Cracking of plates 141 Crosby's safety valve 400 Crucible steel, limit to T. S 68 Crucible steel plates 39 Crown bars 174 Crown sheet 174, 191 Cunningham, G. W 330 Cylinder boilers 219 Cylinder boilers, circulation in 214 ( lylinder boilers, H P. of 211 Cylinder boilers, setting 220 Damper 220 Damper, partial closing of 311 Dangerous connections 396 Darling, Brown & Sharpe 75 Dayton cam pump 341 Dean Brothers 348 DeBruner, H. G 41 Decay of tubes 279 Deep well pump 346 Defective joints 122 Defects in iron plates 21 Defects in large tubes 132 Defects in .steel plates 32 Diameter of stay bolts 175 Diaphragm pressure gauges 408 450 INDEX. PAGE Diffusion of heat in water 213 Direct transfer of heat 195 Domes for boilers 180 Domes, proportions for 182 Double riveted lap joints 96 Double riveting recommended 158 Double riveting, table of 126 Double walls for furnaces 307 Douglas & Sons 222 Douglas, W. B 339 Draft, forced 312 Draught circulation 215 Drifting tests 49, 67 Drilled and punched holes 87, 105, 1('7 Dry steam 241 Ductility and tenacity 18, 34, 78 Ductility in steel plates 33 Dudgeon's tube expander 179 Dutch government, limit to T. S 70 Economizers 386 Economizers and boilers 387 Economizer, Babcock & Wilcox 388 Economizer, circulation in 388 Economizer, functions of 387 Economizer tubes, cast iron 389 Economy in double riveting 124 Edgar Thompson Steel Works. 49 Bdson, M.B 413 Effect of heat on cast iron 142 Elasticity 79 Elasticity and elongation 109 Elasticity in steel plates 106 Elastic limit 79 Elastic limit of Bessemer steel 80 Elastic limit of cast iron.. 11 Elastic limit of wrought iron 80 Elephant boiler ..: 223 Elongation, percentage of 72 Elongation tests 76 Emery, Charles E 439 Emission of radiantheat 187 English boiler plate 96 English government tests for iron 83 Equalizing diameter of pipes 316 Equivalent evaporation 204 Essen and Yorkshire plates 78 Evaporation 203 Evaporation, factors of 207 Evaporation in Cornish boilers 259 Evaporation in flue boilers 237 Evaporation modified by heating surface 195 Evaporation per horse power.. 210 Evaporation trials, fire box boilers.. 294 Evaporative capacity, portable boilers..., 297 Evaporative efficiency 197 Expander, Dudgeon's 179 PAGE Expander, Prosser's 177 Expansion of large flues 260 External corrosion 431 External heating surface 191 Externally fired boilers 218 Extent of heating surface 198 Eyth, Max 97 Factor of safety 149 Factor of safety, additions to 152 Factor of safety, cast iron 11 Factors of evaporation 207 Fairbairn & Hetheriugton 260 Fairbairn's boiler 265 Fairbairn on riveted joints ; 95 Fan blowers 314 Farnley iron 69 Faults in steel plates 36 Faults of steel rivets 119 Feed apparatus 337 Feeding water into steam room 322 Feed water and cast iron 8 Feed water, heating 375 Feed water, point of admission 370 Fernald, F. L 55 Fibrous and granular iron 20 Fibrous iron changed to granular 141 Fire, action of on plates 141 Fire box, copper 196 Firebox heating surface 278 Fire box, high 276 Firebox iron 26, 161 Fire box in vertical boilers 272 Fire box, large or small 279 Fire clay to be used 308 Fire door rings 271 Fire, for heating steel 119 Firmenich, J. G. & F 446 Fitting domes to boilers 181 Five-flue boilers 229 Flanged work to be annealed 139 Flange iron 24, 161 Flanging 139 Flanging heads 232 Flanging tests, steel 56 Flanging wrought iron 21 Flat surfaces, staying of 175 Flexure ••. 21 Flue boilers 226 Flue boilers, proportions for vertical 271 Flue boilers, setting 305 Flue boilers, tests of 236 Flue boilers, 6 inch 232 Flue boilers, vertical ^ 270 Flue heating surface 192 Flue in Cornish boilers 259 Flues acting as stays.. 176 INDEX. 451 PAGE Flues, diameter of 226, 230 Flues of large diameter 259 Flues, securing to heads 176 Flues, strength of 229 Flues, strengthening 171 Flues, thicknessof 171, 229 Flues, riveting to heads 233 Flynn, Daniel 286 Fly wheel pump 348 Force draft 312 Forging boiler shells 138 Forge tests 60, 83 Formulas for riveted joints 123 Foundations to be brick 308 Foot valves 339 Four inch tubular boilers 252 Fractured area and T. S 59 Fractures by punching steel 109 Fractures in cast iron 11 Fractures in steel plates 33, 36 Franklin Institute experiments 145 French admiralty test specimens 73 French boiler 223 Frictional resistance riveted joints 119 Fuel for vertical boilers 277 Fuel saved by heating feed 375 Furnace design 304 Furnace door, Butman's 329 Fusible plugs 420 Gain by heating feed water 374 Galloway boiler 267 Galloway boiler, circulation in 215 Galloway tubes 268 Galloway, W. & J 185 Gases, action of heated 194 Gases, flow of in boilers 192 Gas furnace 136 Gaskets 432 Gauge cocks 418 Gauge, pressure 405 Gauge, water 419 Generator, Stead's 384 Giflfard, M 352 Glasgow Best Best iron 69 Grade of iron for boilers 67 Granular and fibrous iron 20 Graphite 7 Grate area 202 Grate area and safety valves 394 Grate area for tubular boilers 246 Grate area in vertical boilers 277 Grate and tube areas 239, 249 Grate bars 309 Grate bars, Butman's 331 Grate bars in Suiter's boiler 290 Grates, distance from boiler 307 PAGE Grates, length of 248 Grate and heating surface 211 Gray cast iron 6 Green's feed water heater 380 Greig, David 97 Grooving 435 Grooving of plates by calking 130 Hancock, John T 361 Hancock's inspirator 361 Hand and machine flanging 140 Hand holes in boilers 270 Hand hole plates 241 Hand riveting, tests of 117 Hard iron or steel, strength of 77 Hardening of plates 143 Hard water 422 Harrison boiler 13 Heads, flanging of 232 Heads for portable boilers 296 Heads, staying of 176 Heads, thickness of 242 Heat, conduction of 190 Heat, effects of on cast iron 12, 142 Heat, rate of transmission 208 Heat, reclaiming from exhaust 874 Heat, transferof 186, 195 Heat, transmission of 196 Heater, coil 375 Heater and boiler feeder 369 Heater and economizers 374 Heater, Green's 380 Heater, Stilwell's 376 Heater, Victor 383 Heating and cooling plates 74 Heating and grate surface 211 Heating feed water, gain by 374 Heating steel rivets 119 Heating surface 191 Heating surface and evaporation 195 Heating surface, extent of 198 Heating surface in fire box... 278 Heating surface in flue boilers 227, 231 Heating surface in flues 192, 201 Heating surface, position of 192 Heating surface in shell 199 Heating surface in tubes 200 Heating surface in vertical boilers 276 Height of fireboxes 272 Hemlock for scale 428 Herrick, J. A 57 High fireboxes 276 High grade iron 161 Hill, John W 292 Hoadley, J. C 293 Holes, conical when punched 89 Holes leading into domes 181 452 INBEX. PAGE Holley, A. L 221 Holmes, Isaac V 235 Holt, John P : 409 Homogeneous iron 27 Homogeneous plates, properties of 35 Homogeneous plates, stretch of 72 Hoopes & Townsend 90 Horse power of boilers 209, 238, 277, 278 Hotchkiss, James F 426 Hot-short iron 10, 17 Hot tests of steel 56 Hussey, Howe & Co 43 Huston, Charles, oa steel 33 Hydraulic riveting 100, 117 Impure iron 433 Impurities in castings 10 Impurities in steel 37 Incrustation and corrosion 422 Indirect transfer of heat 195 Ingot iron 27 Injector, GifFard's 352 Injector, Seller's 352 Injector, Schutte & Goehring's 365 Injury by cold feed water 370 Inspirator, Hancock's 361 Inspecting plates 73 Internal corrosion 432 Internally fired boilers 257 Internal heating surface 191 Iron, corrosion of 433 Iron CJad Manufacturing Co 384 Iron, elastic limit of 150 Iron from Rodger's bed ore 57 Iron modified by working 18 Iron plate, analysis of 16 Iron, transmission of heat through 196 Iron, treacherous at low heats 144 Iron, tensile tests for rivets 115 Isherwood, B. F 146 Jarvis' furnace 317 Jarvis, K. M 317 Johns, H. W 389 Johnson, R. W 147 J oints f or large flues 259 Joints, riveted 86 Joints under stress 102 Kelley's sectional boiler 443 Kelley, Wm. E 443 Kemp's boiler cleaner 426 Kennedy's spiral punch 92 Kent, R 41 Kirkaldy, D., on properties of iron 18 Kirkaldy on testing iron 84 Knowles, L. J 348 Knowles Steam Pump Works 272 Krupp's iron.,... 35 PAGE Kunkle's safety valve 403 Lap joints, riveted 121, 124 Lap welded joints 136 Lancashire boiler 260 Lancashire boiler, H. P. of 211 Landore-Siemens steel 35 Lane's pressure gauge 407 Large fire boxes 279 Law in regard to boiler plates 74 Leaky joints 431 Leaky tubes 179 Leaking through rivet holes 371 Length of specimens 72 Lifting pumps 388 Lignite, combustion of 318 Lime, carbonate of 422 Lime extractor, Stilwell's 376 Lime, sulphate of 422 Limit of elasticity 79 Limit to tensile strength 63, 70 Liquids, conduction of heat by 190 Lloyd's Register 39, 70, 73 Lloyd's, Foster, best iron. 69 Lloyd, Son & Co 23 Locomotive boilers 298 Loss of heat by scale 425 Long and short specimens 58 Loss by punching 89 Lower grade of iron for boilers 67 Lowmoor iron 69 Low temperature, effects on iron 144 Low T. S. of steel 61 Lugs for boilers 308 Lunkenheimer's safety valve 403 Machine flanging 139 Mahogany for scale 428 Malleability 20 Manganese in iron 4 Manganese in steel 46, 53 Man holes 184 Martell, Mr 70 Metals, conducting power of 189 Mild steel 27 Mississippi gauge cock 418 Molasses for scale 428 Molecular changes in iron 140 Montgomery, J. F 309 Moore, George W 372 Moore & Kerrick 340 Moore's boiler feeder 372 Mud drums 306 Mud, removing from feed 378 Napier, James R 196 Nashua Iron and Steel Co 55 Newark-Cornish boiler 261 New Jersey Zinc Co 45 INDEX. 453 PAGE New York Safety Steam Power Co 282 Nicks in samples 72 NilesTool Works 280 Non-adjustable injector 356 Northcote, Henry 208 North-of-England iron 69 Nozzles for boilers 183 Nuts, cold punched , 90 Nut galls for scale 428 Open hearth steel 52 Otis Iron and Steel Co 58 Overestimating tube efficiency 276 Overheating steel plates., 142 Overstamped plates vs. law 74 Oxidation in welding 134 Oxygen, free 136 Oxygen must be kept from steel 119 Park, Brother & Co 39 Patching ; 9 Peat, combustion of 318 Peclet's experiments on heat 196 Pennsylvania R. R. boilers 298 Percussion tests 81 Petroleum for scale 429 Phillips, Nimick & Co 23 Phosphorus in iron 145 Phosphorus in steel 37 Phosphorus not removed by the Besse- mer process 46 Pierce, Henry M 333 Pierce's furnace 334 Pig iron 7 Pipes, passage of air through 315 Pitting 434 Planing edges of plates 126 plates for boilers 143 Plates injured by calking 127 Plates overstamped vs. law 74 Plates to be stamped by law 73 Plates, welding of 131 Pook, Samuel H 55 Portable boilers 295 Position of heating surface .'.. 192 Post & Co 410 Potash for scale 429 Potatoes for scale 428 Power pumps 338 Pressure gauges 405 Pressure gauge, Bourdon's 406 Pressure gauge, Edson's recording 413 Pressure gauge, Holt's 409 Pressure gauge. Lane's 407 Pressure gauge. Post & Co 410 Pressure on boiler heads 155 Pressure on rivet heads 118 Prevention of scale 426 PAGE Priming 282 Properties of iron, modified 18 Properties of steam 205 Proportions for riveted joints.. 122 Proportions for stay bolts 175 Proportions for steam drums 183 Proportions for vertical boilers 277 Prosser's tube expander 177 Puddling 19 Pumps 337 Pumps, auxiliary 3.')0 Pumps, capacity of 337 Pumps for deep wells 346 Pumps, power 338 Pumps, steam 340 Punch, action of on plates 109 Punch and dies 90 Punch, Kennedy's spiral 92 Punched and drilled holes 87, 1('5 Punched holes, conical 89 Punching and annealing 48 Punching, bad effects of 86 Punching brittle iron 92 Punching, experiments on 93 Punching good iron 90 Punching, loss of strength in 89 Punching steel plates 47 Punching thick plates 106 Quality of boiler plate 15 Qualities required by law 7* Radiant heat 187 Radiant heat from wood 189 Radiation 186, 188 Ramsbottom's welding machine 138 Rate of evaporation 204 Rating boilers by heating surface 278 Raritan Woolen Mills 439 Reaming out punched holes 87 Recording gauge, Edson's 413 Records of U. S. tests, how kept 76 Red short iron 17 Reduction of area in tests 77 Refuse fuel 313 Reheating and cooling 141 Removal of scale '. 426 Requirements of iron 83 Requirements of a steam pump 340 Resistance to collapse 171 Resistance to conduction 196 Resistance to shearing 56 Return steam trap 369 Richards, C. B 58 Richardson's safety valve 398 Rings for man holes 184 Rivet heads, pressure on 118 Rivet-iron tests 115 454 INDEX. PAGE Riveted joints 86 Riveted joints, friction al resistance in.. 119 Riveted joints, strength of 98 Riveted joints, ultimate strength of 15') Riveted shells, strength of 148 Rivets, spacing of 123 Rivets, steel — faults of 119 Rivets, testing 113 Rivets of steel 56 Riveting, double 124 Riveting, single... 121 Riveting, influence of pressure on 118 Riveting in flues 233 Rocking grates, recommended 330 Rodger's bed ore, analysis of iron 57 Rogers, Joseph G 430 Root's boiler 441 Rough iron not the strongest 85 Rubber gaskets 432 Rusting of boilers 431 Ryder's grate bar 309 Safe load on stay bolts 173 Safety, factor of 149 Safety and stop valve 397 Safety apparatus 393 Safety plugs... 420 Safety valve 393 Safety valve and dangerous connections.. 395 Safety valve and grate area 394 Safety valve, Ashcroft's 399 Safety valve, Crosby's 400 Safety valve, diameters of 394 Safety valve, Kunkle's 403 Safety valve, Lunkenheimer's 403 Safety valve, Richardson's 398 Safety valve, how connected 394 Safety valves, table of 404 Salt in feed water 423 Samples, nicks in 72 Samples, size and shape of 71 Samples, for elongation tests 76 Sand, removing from feed water 378 Scale and location of tubes 241 Scale, chemical agents for 428 Scale, formation of 423 Scale in boilers 373 Scale, injury to boilers by 423 Scale, loss of heat by 425 Scale, preventives to be used 431 Scale, prevention and removal 426 Scarfing and welding.... 134 Scarf-welded joints... 136 Schutte& Goehring 313 Schutte & Goehring injector 365 Scrap steel 56 Seams not to be chipped ,. 126 PAGE Sectional boilers 437 Sellers, William & Co 352 Semi-portable boilers 297 Setting boilers 304 Setting cylinder boilers 220 Setting grate bars 307 Setting internally fired boilers 258 Shapley's boiler 272 Shearing and tensile strains 11* Shearing rivets in joints 122 Shearing steel rivets 56 Shearing tests of stay bolts 113 Shearing tests of steel 60 Shells for boilers, thickness of 242 Shell iron 22 Shock, W. H 113 Short and long specimens 59 Siemens-Martin steel 52 Siemens-Martin steel, limit to T. S 68 Silica in water 423 Silicon in Bessemer pig 44 Silicon in iron ^ 4 Silicon in steel. 37 Singer, Nimick & Co 62 Single riveted joints, calking 129 Single riveted joints, strength of 66, 98 Single riveted lap joints 121 Single riveting, table of 124 Six inch flue boilers 232 Size and shape of samples 71 Sligo Iron 25, 77 Slusser & Suiter 290 Small fire boxes 279 Smith, Vaile& Co 341 Snowden, Thomas, feed pipe 371 Snyder's vertical boiler 283 Snyder, Ward B 284 Societe Alsacienne, etc 266 Soldiers' Home, Ohio, tests at 236 Solid drawn tubes 172 South Metropolitan Gas Works, tests at.. 263 Space between tubes 240 Spacing rivets 123 Specimens, length of 72 Spiegel-eisen 44 Spiral and flat punching 93 Staffordshire iron 69 Stamping boiler plates 73 Stay bolts 172, 270 Stay bolt tests 113 Stay bolts with nuts 174 Staying boiler heads 176 Staying flat surfaces 175 Stay rods 176 Steads' circulating generator 384. Steam Boiler Applianoe Co 367 INDEX. 455 PAGE Steam drum 183 Steam domes 180 Steam, dry 241 Steam generators and economizers 387 Steam jets for lifting wating 340 Steam jet blowers, sizes of 314 Steam jet for draft 312 Steam pipes, covering for 392 Steam, properties of 205 Steam pumps 340 Steam room 183 Steam room, feeding water into 372 Steam riveting, tests of 117 Steam trap, return 369 Steel, advantages of for boilers 30 Steel bars, strength of 56 Steel boiler plate, carbon in 30 Steel, definition of 29 Steel for boilers 29 Steel, incipient fractures in 33 Steel injured by punching 108 Steel, nature of must be studied 31 Steel not injured by drilling 108 SI eel plates, annealing 143 Steel plates, defects in 32 Steel plates, ductility of 33 Steel plates, experiments on thick 106 Steel plates, faults of 36 Steel plates, limit to T. S 68 Steel plates, punching 47 Steel plates, strength of 68 Steel plates, stretch of 72 Steel plates, welding of 137 Steel plates, why a failure 32 Steel rivets, HG, 119 Steel rivets, burning of 119 Steel rivet tests 116 Steel, scrap 56 Steel, shearing tests 60 Steel, tensile strength of 38 Steel, texture of 29 Steel, treacherous at low heats 144 Strength of American plate 69 Strength of boilers 148 Strength of butt joints 103 Strength of riveted joints 95 Strength of plates in flanging 140 Stilwell & Bierce 376 Stilwell's lime extractor 376 Stone for boiler foundations 308 Stop and safety valve 397 Stoy, C. S 432 Strainers 339 Strains in boilers 371 Strains in cooling 9 Strength of punched and drilled holes..., 91 | PAGE Strength of riveted sheUs 148, 153 Strength of stay bolts ;.... 173 Strength of steel plates 106 Strength of welded joints 133 Strengthening flues 171 Stretch of homogeneous plates 72 Stretching Iron 77 Sturtevant, B. F 314 Sulphate of lime 424 Suiter's boiler 290 Tangye Brother & Holman 263 Tanks for water reserve 338 Taonate of soda for scale 430 Tannic acid for scale 428 Tearing and breaking samples 72 Tenacity and ductility 18, 34, 78 Tensile and shearing strains 114 Tensile strength of boiler iron 23 Tensile strength, limit to 17 Tensile strength of steel .38 Tensile strength of Bessemer steel 51 Tensile strength of crucible steel 42 Tensile strength of open hearth steel 01 Tensile strength of rivets, steel 120 Tensile strength, specimensfor 07 Test, bending 66 Test, bulging..; 81 Test, drifting 67 Test, forge 83 Test, percussion 81 Test, temper 66 Tests, how U. S. to be made 73 Tests, how U. S. records to be kept 76 Tests of iron, Kirkaldy's conclusions 84 Testing steel plates 40, 161 Tests of rivet iron 11,3-5 Temperature, effect of low on iron 144-6 Temperature, escaping gases 387 Texture of flange iron 27 Texture of steel 29 Texture of wrought iron 19 Thickness of flues 171 Thickness of plates for boilers 165 Thick steel plates 106 Thin fires must not be used 119 Thornycroft best best iron 69 Thurston, R. H., on H. P. of boilers 212 Three inch tubular boilers 243 Three and a half inch tubular boilers.... 250 Tidal circulation 216 Tightness of joints 165 Torsional tests of rivet steel 120 Torsional tests of steel for boilers 50 Toughness in boilerplates 18 Toughness in plates, U. S. law 74 Transfer of heat 186 456 INDEX. PAGE Transmission of heat 196 Transmission of radiant heat 189 Trevithick, Richard 258 Tube areas, table of 248 Tube and grate areas 239, 249 Tube area in vertical boilers 278 Tube expander 177 Tubes and circulation 238 Tubes as braces 177 Tubes as heating surface 192 Tubes, cutting to length in place 178 Tabes, defects in large 132 Tubes, distance between 241 Tubes for vertical boilers 276 Tubes, Galloway's 268 Tubes, height'of 240 Tubes injured by firing 279 Tubes, interfering with circulation 214 Tubes, length of 239 Tubes, location of 241 Tubes, number in. a boiler 239 Tubes, space between 240 Tubes, strength of 172 Tubes, tool for calking 179 Tubes, wasting of 279 Tubular boilers 238 Tubular boilers, 3 inch 243 Tubular boilers, 3>^ inch 25i) Tubular boilers, 4 inch 252 Tubular boilers, 6 inch 234 Tubular boilers, compound 254 Tubular boiler, test of 253 Tubular boilers, length of 239, 245 Tubular boilers, proportions 238 Tubular boiler setting 305 Tubular boilers, vertical 276 Turned iron, not weakened 85 Two-flue boilers 227 Ultimate strength of boilers 148 Ultimate strength of riveted joints 150 U. S. tests, how made 75 U. S. Treasury, tests of iron 68 Value of tube surface 193, 278 Varieties of plate iron 22 Vertical boilers 269 PAGE Vertical boilers, fire boxes for 272 Vertical cylinder boilers 222 Vertical flue boiler 270 Vertical tubular boiler 275 Vertical tubular boiler, faults of 282 Vertical tubular boiler, proportions 277 Vibrations of a plate 140 Victor heater 383 Vinegar for scale ». 428 Water bottoms 296 Water charger 350 Water, feeding into steam room 372 Water gauges 419 Water, hard or soft 422 Water not a good conductor 213 Water, required for boilers 337 Water, required per H. P 210 Watts' rule for H. P 211 Water space, vertical boilers 272 Weakening of shell by domes 180 Webb, F. W., on Bessemer steel 48 Webb, F. W., on punching 94 Welded boilers 136 Welded joints, strength of 133 Welding blooms 20 Welding boilers 131, 136 Welding, oxidation in 134 Welding prevented by cinder 134 Welding, scarf 135 Welding steel plate^.;.. 137 Weldless rings for Wl6rs 138 White iron I 6, 16 Williams, C. Wye...} 193 Wilson, Kobert f 215 Workington iron, analysis of 57 Wrought iron, classification of 19 Wrought iron, elastic limit 80 Wrought iron for qoilers 15, 67 Wrought iron, proi)erties required 16 Wrought iron, texiture of 19 Yorkshire iron....j 35, 69 Yorkshire and Essen plates 78 Zigzag and chain jriveting 109 Zinc for scale ; 430