r kk 1LLBR METQflLP#PflRKIH. :- >»- ,j THE TREATMENT OF STEEL. A SERIES OF CIRCULARS ON Heating, Annealing, Forging andTempering, ; ISSUED BY THE f CRESCENT STEEL WORKS. I ,J.%i.i.h PITTSBURGH : MILLER, METCALF & PARKIN. 1881. Copyright, 1881, By MILLER, METCALF & PARKIN. i KKIGHT & LEONARD . 1 ^ijOiu PREFACE. FROM time to time our friends ask of us questions con- cerning steel, its correct treatment and management, and we have therefore thought it proper to combine in one form circulars issued by us in previous years. At the same time we add some new and, we hope, interesting matter. We trust that it may be accepted as an effort on our part to assist in overcoming difficulties connected with the working of steel. Many will find in it nothing new; others may be able to select something of interest and profit. M. M. & P. Pittsburgh, July i, 1881. THE TREATMENT OF STEEL. AN English steel manufact- urer remarks: "We do not analyze for carbon because we find by long experience that the eye can judge of the percentage of carbon in an ingot of cast steel of the highest tempers from the ap- pearance of the fracture more accurately than the chemist can ascertain by any method of analysis hitherto discov- ered." We do not indorse this as a whole, but to show the re- liability of " inspection by the eye " we give the result of tests made in 1874 and 1875. In April, 1874, we selected eight ingots, by our numbers, from the ingots as they were piled up for use. The figures in the second column show the amounts of carbon in these eight ingots as determined by analysis. In March, 1875, twelve ingots were selected in the same way, one being below the series of eight and the others being within the limits of the eight, Nos. 5, 7 and 10 being inter- polations. It will be seen that the carbon runs up with the numbers 7 8 THE TREATMENT OF STEEL. without a break, thus proving by careful analysis that our methods were careful and reliable. The mean difference in carbon is only seven-hundredths of one per cent. As there is no possibility of a skillful and careful person mistaking any one of these numbers for another, our patrons can judge for themselves of the proba- bility of receiving steel of uniform temper and properly adapted to the work to be done. The structure of the ingot is invariably the same for the same amount of carbon, and the observance of this rule is the steel maker's guide and the steel user's safety. Analyses made by Prof. John W. Langley, University of Michigan, Ann Arbor, Michigan : No. I 2 3 4 5 6 7 ■■ 8 9 io ii 12 Mean differences Carbon. March, 1875, .302 .490 •529 .649 .So i .S41 .S67 .87 1 •955 1.005 1.05S 1.079 .071 Carbon. April, 1874. .404 •599 •7S9 '.856 .867 •939 1036 1. 116 .102 Carbon. Differ- ences. Mean of the two preced- ing columns. .302 •447 •564 .719 .So 1 .S 4 S .S67 .S69 •947 1.005 1.047 1.097 •1*45 .117 •155 .0S2 .047 .019 .002 .07S •0 5 S .042 .050 .072 Iron by Differ- ence. 99.614 99-454 99363 99.270 99.119 99.0S5 99.044 99.040 9S.900 9S.860 9S.752 93-753 The figures in this table are tractions of one per cent. ON ANNEALING, |VVING to the fact that the opera- tions of rolling or hammering steel make it very hard, it is frequently necessary that the steel should be annealed before it can be conveniently cut into required shapes for tools. Annealing or softening is ac- complished by heating steel to a red heat and then cooling it very slowly, to prevent it from getting hard again. The higher the degree of heat, the more the steel will be softened, until the limit of soft- ness is reached when the steel is melted. It does not follow that the higher a piece of steel is heated, the softer it will be when cooled, no matter how slowly it may be cooled ; this is proved by the fact that an ingot is always harder than a rolled or hammered bar made from it. Therefore, there is nothing gained by heating a piece of steel hotter than a good bright cherry red ; on the contrary, a higher heat has several disadvantages : First — If carried too far, it may leave the steel actually harder than a good red heat would leave it. IO THE TREATMENT OF STEEL. Second — If a scale is raised on the steel, this scale will be harsh, granular oxide of iron, and will spoil the tools used to cut it. It often occurs that steel is scaled in this way, and then because it does not cut well, it is customary to heat it again, and hotter still, to overcome the trouble ; while the fact is, that the more this operation is repeated, the harder the steel will work, because of the hard scale and the harsh grain underneath. Third — A high scaling heat, continued for a little time, changes the structure of the steel, destroys its crystalline property, makes it brittle, liable to crack in hardening, and impossible to refine. Again, it is a common practice to put steel into a hot fur- nace at the close of a day's work, and leave it there all night. This method always gets the steel too hot, always raises a scale on it, and worse than either, it leaves it soak- ing in the fire too long; and this is more injurious to steel than any other operation to which it can be subjected. A good illustration of the destruction of crystalline struct- ure by long continued heating may be had by operating on chilled cast iron. If a chill be heated red hot and removed from the fire as soon as it is hot, it will, when cold, retain its peculiar crys- talline structure ; if now it be heated red hot, and left at a moderate red for several hours ; in short, if it be treated as steel often is, and be left in a furnace over night, it will be found, when cold, to have a perfect amorphous structure, every trace of chill crystals will be gone, and the whole piece will be non-crystalline gray cast iron. If this is the effect upon coarse cast iron, what better is to be expected from fine cast steel ? ON ANNEALING. II A piece of fine tap steel after having been in a furnace over night will act as follows : It will be harsh in the lathe and spoil the cutting tools. When hardened it will almost certainly crack; if it does not crack it will have been a remarkably good steel to begin with. When the temper is drawn to the proper color and the tap is put into use, the teeth will either crumble off or crush down like so much lead. Upon breaking the tap the grain will be coarse and the steel brittle. To anneal any piece of steel, heat it red hot ; heat it uni- formly and heat it through, taking care not to let the ends and corners get too hot. As soon as it is hot take it out of the fire, the sooner the better, and cool it as slowly as possible. A good rule for heating is to heat it at so low a red that when the piece is cold it will still show the blue gloss of the oxide that was put there by the hammer or the rolls. Steel annealed in this way will cut very soft; it will harden very hard, without cracking, and when tempered it will be very strong, nicely refined and will hold a keen, strong edge. ON HEATING STEEL. WING to varying instructions on a great many different labels, we find at times a good deal of misapprehension as to the best way to heat steel ; in some cases this causes too much work for the smith, and in other instances disasters follow the act of hard- ening. There are three distinct stages, or times of heating : First, for forging. Second, for hardening. Third, for tempering. The first requisite for a good heat for forging is a clean fire and plenty of fuel, so that jets of hot air will not strike the corners of the piece ; next, the fire should be regular, and give a good uniform heat to the whole part to be forged. It should be keen enough to heat the piece as rapidly as may be, and allow it to be thoroughly heated through, without being so fierce as to overheat the corners. The trouble in the forge fire is usually uneven heat, and not too high heat. Suppose the piece to be forged has been ON HEATING STEEL. 1 3 put into a very hot fire, and forced as quickly as possible to a high yellow heat, so that it is almost up to the scintillating point. If this be done, in a few minutes the outside will be quite soft and in nice condition for forging, while the mid- dle parts will be not more than red hot. The highly heated soft outside will have very little tenacity ; that is to say, this part will be so far advanced toward fusion that the particles will slide easily over one another, while the less highly heated inside parts will be hard, possessed of high tenacity, and the particles will not slide so easily over each other. Now let the piece be placed under the hammer and forged, and the result will be as shown in Fig. i. Fig. i. Fig. 2. The soft outside will yield so much more readily than the hard inside, that the outer particles will be torn asunder, while the inside will remain sound, and the piece will be pitched out and branded "burned." Suppose the case to be reversed and the inside to be much hotter than the outside ; that is, that the inside shall be in a state of semi-fusion, while the outside is hard and firm. Now let the piece be forged and we shall have the case as shown in Fig. 2. The outside will be all sound and the whole piece will appear perfectly good until it is cropped, and then it is found to be hollow inside, and it is pitched out and branded "burst." In either case, if the piece had been heated soft all 14 THE TREATMENT OF STEEL. through, or if it had been only red hot all through, it could have been forged perfectly sound and good. If it be asked, why then is there ever any necessity for smiths to use a low heat in forging, when a uniform high heat will do as well, we answer : In some cases a high heat is more desirable to save heavy labor, but in every case where a fine steel is to be used for cutting purposes, it must be borne in mind that very heavy forging refines the bars as they slowly cool; and if the smith heats such refined bars until they are soft, he raises the grain, makes them coarse, and he cannot get them fine again unless he has a very heavy steam hammer at command and knows how to use it well. In following the above hints there is a still greater danger to be avoided : that is incurred by letting the steel lie in the fire after it is properly heated. When the steel is hot through it should be taken from the fire immediately and forged as quickly as possible. " Soaking " in the fire causes steel to become " dry " and brittle, and does it more injury than any bad practice known to the most experienced. By observing these precautions a piece of steel may al- ways be heated safely, up to even a bright yellow heat, when there is much forging to be done on it ; and at this heat it will weld well. The best and most economical of welding fluxes is clean, crude borax, which should be first thoroughly melted and then ground to fine powder. Borax prepared in this way will not froth on the steel, and one half of the usual quantity will do the work as well as the whole quantity unmelted. After the steel is properly heated it should be forged to shape as quickly as possible, and just as the red heat is leav- ON HEATING STEEL. 1 5 ing the parts intended for cutting edges, these parts should be refined by rapid light blows, continued until the red dis- appears. For the second stage of heating for hardening great care should be used, first, to protect the cutting edges and work- ing parts from heating more rapidly than the body of the piece; next, that the whole part to be hardened be heated uniformly through, without any part becoming visibly hotter than the other. A uniform heat, as low as will give the required hardness, is the best for hardening. BEAR IN MIND, that for every variation of heat which is great enough to be seen there will result a varia- tion in grain, which may be seen by breaking the piece, and for every such variation in temperature there is a very good chance for a crack to be seen. Many a costly tool is ruined by inattention to this point. The effect of too high heat is to open the grain, to make the steel coarse. The effect of an irregular heat is to cause irregular grain, irregular strains, and cracks. As soon as the piece is properly heated for hardening it should be promptly and thoroughly quenched in plenty of the cooling medium, water, brine or oil, as the case may be. An abundance of the cooling bath, to do the work quickly and uniformly all over, is very necessary to good and safe work. To harden a large piece safely a running stream should be used. Much uneven hardening is caused by the use of too small baths. For the third stage of heating; to temper, the first impor- tant requisite is again uniformity. The next is time; the i6 THE TREATMENT OF STEEL. more slowly a piece is brought down to its temper, the better and safer is the operation. When expensive tools, such as taps, rose cutters, etc., are to be made, it is a wise precaution, and one easily taken, to try small pieces of the steel at different temperatures, so as to find out how low a heat will give the necessary hardness. The lowest heat is the best for any steel, the test costs noth- ing, takes very little time, and very often saves considerable losses. FURNACES. |E present in this connection sketches ^Jl of a cheap and handy furnace for [J use in a blacksmith shop, adapted jl especially for heating steel, and more [J particularly for heating steel for hard- ening. The furnace is so simple that the sketches need no explanation; for binders, ten pieces of old rail about six feet long with one end set in the ground and the tops tied by ^ inch rods are all that is necessary, with a piece of iron about 3x5^ inches running around near the top, and set in flush with the bricks. The distinctive features of this furnace are the fire bed and a good damper on the stack. In an experience of many years we have found nothing better than the Tupper grate- bar with half-inch openings. These bars set in as shown make a level, permanent bed, and give an evenly distributed supply of air to the fuel. In such a furnace as this one set of bars will last for years and remain level. While on the subject of grate-bars we may as well say that the satisfactory and safe working of this furnace would be entirely defeated by any attempt to use either square wrought-iron bars or ordinary straight cast-iron bars. Such bars always warp, get pushed out of place, and allow a rush of air through at one place and no air at another. This causes hot and cold places in the furnace and produces 17 i8 THE TREATMENT OF STEEL. uneven heating, which is the chief source of cracking in hardening, and also the air rushing through the large holes will burn the steel. A bar must be used which will remain level and in its place, and the smaller and more numerous the openings are the better will be the result. Clean hard coke is the only proper fuel for such a furnace and for such work. The furnace should be filled full up to the fore plate, or better, a little higher, — with coke in pieces SEC Tl OH OF STACK ^fa_ -9'" 1 13'™ SG > (4/ IN) * CO CO no larger than an ordinary man's fist, — but the smaller the better. When it is used for heating for forging purposes the damper may be left high enough to run the furnace as hot as may be required, — if necessary, a welding heat can be obtained. When used for hardening the furnace should be got as hot as is needed before the steel is put into it, then when the steel is put in the damper should be dropped down tight. The door, which is 12 inches high and 24 inches wide, should be nicely balanced by a lever and weight, with a rod FURNACES. J 9 in a handy place so that the operator can pull it up easily and turn over his pieces from time to time, so as to get his heat perfectly uniform. In the clear gas of a coke fire the whole interior of a fur- nace can be seen easily, and every piece can be watched as 1 m -l3lN~ i <4& EL & •- 60 IN.. ii — ir -|| ii n ii — ir SECT/DN AT AB D it ought to be. Time, care, watchfulness and absolute uni- formity of heat are the essentials necessary for success in hardening steel. Every large shop should have such a furnace, and should have one man trained to its use, to do the hardening and tempering for the whole shop. Such a furnace in the hands of a careful man in any railroad shop in the country would pay for itself every year and save the man's wages besides. 20 THE TREATMENT OF STEEL. The furnace will consume very little coke at any time, and when not in use, with the damper down it will stay hot a long time and waste the coke but a trifle. There is no more absurd nor wasteful system than that of requiring a smith at his anvil to harden and temper his work. His fire is not fit to heat in, to begin with, and he never has time to do his work properly if it were. From such a furnace as is here described we harden all A a i CO £5 SECTION AT CD B sorts of tools, taps, small dies, large rolls, rotary shear knives, and shear knives as large as five feet long, which is the whole length of the furnace. The steel which is tempered best is that which is the finest in the grain and the strongest. The best way to test both grain and strength is to hammer out a piece to about 1 X A x H inch, a foot or so in length, and temper to a high blue, or pigeon-wing, and when cold to break it off in little pieces with a hand hammer. FURNACES. 2 1 A little practice will soon enable a man to determine, first, whether he heated his piece to just the right point. The file and the appearance of the grain will determine this point. Next, when a little experience as to heat has been gained, he will know by the strength and grain whether his steel is really good, or whether it is "dry" and poor. In every shop there are plenty of worn-out tools of all sorts in the scrap-heap ; the temperer should be allowed to spend all of his leisure time in hammering these out and testing them as above. The steel will cost nothing, and the knowledge gained will pay for the time over and over again. We say to our friends again that if you will heat even the finest steel in dirty slack fires you will get dirty coats of sul- phurous oxides on it, and no good results. EFFECTS OF HEAT UPON STEEL. WE now present an illustration of the effect of heat upon steel, which is a direct reproduction upon paper of the grain of the steel by means of the heliotype process. Description. — Take a bar of steel of ordinary size, say about an inch by a half, and heat six or eight inches of one end to a low red heat, and nick the heated part all around the bar at intervals of half to three- quarters of an inch, until eight or nine. notches are cut. This nicking is done at red heat, to determine the fractures at the nicks. Next place the end of the bar in a very hot fire and heat it white hot until it scintillates at the extreme end, leaving the other parts enough out of the fire to heat them only by conduction. Let the end remain in the fire until the last piece nicked is not quite red red-hot, and the next to the last barely red hot. Now, if the pieces be numbered from one to eight, com- mencing at the outer end, No. i will be white or scintillating hot, No. 2 will be white hot, No. 3 will be high yellow hot, No. 4 will be yellow or orange hot, No. 5 will be high red Fold-out Placeholder This fold-out is being digitized, and will be inserted at future date. o t( e; it ei th q* m he N. EFFECTS OF HEAT UPON STEEL. 27, hot, No. 6 will be red hot, No. 7 will be low red hot, No. 8 will be black hot. As soon as heated, let the bar be quenched in cold water and kept there until quite cold. After cooling, the bar should be carefully wiped dry, especially in the notches. An examination by the file will reveal the following, if high steel has been used : No. 1 will scratch glass, Nos. 2, 3 and 4 excessively hard, Nos. 5 and 6 well hardened, No. 7 about hard enough for tap steel, No. 8 not hardened. In breaking off the pieces over the corner of the anvil they should be caught in a clean keg or box, to keep the fractures clean and bright. No. 1 will be as brittle as glass, No. 2 will be nearly as brittle as glass, Nos. 3, 4 and 5 will break off easily, each a little stronger than the other, Nos. 6 and 7 will be very strong, and much stronger than No. 8, or the bar unhardened. Place the pieces in the order of their numbers fitting the fractures, then upend each one, beginning with No. 1 and following with each in the order in which they lie, and the result will be fractures as shown so beautifully in our illus- tration, each differing from the other. No. 1 will be coarse, yellowish cast, and very lustrous; No. 2 will be coarse and not quite so yellow as No. 1 ; No. 3 will be finer than 1 or 2, and coarser than No. 8, and will have fiery luster; No. 4, like No. 3, not quite so coarse, yet coarser than No. 8; No. 5 will be about the same size grain as No. 8, but will have fiery luster; No. 6 will be much finer than No. 8, will have no fiery luster, will be hard through and very strong. This is what is called refining by hard- ening. No. 7 will be refined and hard on the corners and edges, and rather coarser, and not quite so hard in the 24 THE TREATMENT OF STEEL. middle. This is about the right heat for hardening taps, milling tools, etc., the teeth of which will be amply hard, while there will be no danger of cracking the tool. No. 8 illustrates the original grain of the bar. In nine cases out of ten the bar will crack along the middle to the refined piece. In the illustration the crack shows very plainly in No. 4, but we have never known this crack to extend into the refined piece, although we have repeated the experiment many times. We learn from this experiment the following: First, "0" Any difference in temperature sufficiently great to be seen by the color will cause a corresponding difference in the grain, "b" This variation in grain will produce inter- nal strains and cracks. Second, Any temperature so high as to open the grain so that the hardened piece will be coarser than the original bar will cause the hardened piece to be brittle, liable to crack, and to crumble on the edges in use. Third, A temperature high enough to cause a piece to harden through, but not high enough to open the grain, will cause the piece to refine, to be stronger than the untem- pered bar, and to carry a tough, keen, cutting edge. Fourth, A temperature which will harden and refine the corners and edges of a bar, but which will not harden the bar through, is just the right heat at which to harden taps, rose-bitts and complicated cutters of any shape, as it will harden the teeth sufficiently without risk of cracking, and will leave the mass of the tool soft and tough, so that it can yield a little to pressure and prevent the teeth tearing out. These four rules are general, and apply equally well to any quality of steel or to any temper of steel. Steel which is so mild that it will not harden in the ordi- EFFECTS OF HEAT UPON STEEL. 25 nary acceptance of the term will show differences of grain corresponding to variations in temperature. To restore any of the first seven pieces shown to the original structure as shown in No. 8, it is only necessary to heat it through to a good red heat, not to a high red, allow it to stay at this temperature for ten minutes to thirty minutes, according to the size of the piece, and then to cool slowly. If upon the first trial the restoration should be found incom- plete, and the piece upon being fractured should still show some fiery grains, a second heating continued a little longer than the first would cause a restoration of fracture. This property of restoration is not peculiar to any steel, and its performance requires no mysterious agencies beyond those given above. It should be distinctly borne in mind that a piece restored from overheating is never quite as good as it would have re- mained if it had never been abused, and we strongly advise that no occasion should ever be given for the use of this process of restoration except as an interesting experiment. The original and proper strength of fine steel can never be fully restored after it has once been destroyed by over- heating. ON TEMPER OF STEEL. [HE word temper, as used by the steel maker, indicates the amount of carbon in steel ; | thus, steel of high temper is steel containing much carbon ; steel of low temper is steel containing little carbon ; steel of medium temper is steel con- taining carbon between these limits, etc. Each number of our carbon circular represents a temper, and besides these numbers we use intermediate ones, amounting to some twenty in all. As the temper of steel can only be observed in the ingot, it is not necessary to the needs of the trade to attempt any description of the mode of observation, especially as this is purely a matter of education of the eye, only to be obtained by years of experience. The act of tempering steel is the act of giving to a piece of steel, after it has been shaped, the hardness necessary for the work it has to do. This is done by first hardening the piece, generally a good deal harder than is necessary, and then toughening it by slow heating and gradual softening until it is just right for work. 26 ON TEMPER OF STEEL. 2J A piece of steel properly tempered should always be finer in grain than the bar from which it was made. If it is neces- sary, in order to make the piece as hard as is required, to heat it so hot that after being hardened it will be as coarse or coarser in grain than the bar, then the steel itself is of too low temper for the desired work. In a case of this kind, the steel maker should at once be notified of the fact, who should immediately correct the trouble by furnishing higher steel. If a great degree of hardness is not desired, as in the case of taps, and most tools of complicated form, and it is found that at a moderate heat the tools are too hard and are liable to crack, the smith should first use a lower heat in order to save the tools already made, and then notify the steel maker that his steel was too high, so as to prevent a recurrence of the trouble. In all cases where steel is used in large quan- tities for the same purpose, as in the making of axes, springs, forks, etc., there is very little difficulty about temper, be- cause, after one or two trials, the steel maker learns what his customer requires, and can always furnish it to him. In large, general works, however, such as rolling-mill and nail factory, or large machine works, or large railroad shops, both the maker and worker of the steel labor under great disadvantages from want of a mutual understanding. The steel maker receives his order and fills the sizes of tempers best adapted to general work, and the smith usually tries to harden all tools at about the same heat. The steel maker is right, because he is afraid to make the steel too high or too low, for fear it will not suit, and so he gives an average adapted to the size of the bar. The smith is right, because he is generally the most hur- ried and crowded man about the establishment. He must forge a tap for this man, a cold nail knife for that one, and a 28 THE TREATMENT OF STEEL. lathe cutter for another, and so on ; and each man is in a hurry. Under these circumstances he cannot be expected to stop and test every piece of steel he uses, and find out exactly at what heat it will harden best, and refine properly. He needs steel that will all harden properly at the same heat, and this he usually gets from the general practice among steel makers, of making each bar of a certain temper, according to its size. But if it should happen that he were caught with only one bar of say inch and a quarter octagon, and three men should come in a hurry, one for a tap, another for a punch, and an- other for a chilled roll plug, he would find it very difficult to make one bar of steel answer for all of these purposes, even if it were of the very best quality, and the chances are that he would make one good tool and two bad tools. There is a perfectly easy and simple way to avoid all of this trouble ; and that is, to write after each size the pur- pose for which it is wanted, as for instance : Track tools, smith tools, lathe tools, taps, dies, cold nail knives, cold nail dies, hot nails, hot or cold punches, shear knives, etc. This gives very little trouble in making the order, and it is the greatest relief to the steel maker. It is his delight to get hold of such an order, for he knows that when it is filled he will hardly ever hear a complaint. Every steel maker worthy of the name knows exactly what temper to provide for any tool, or if it is a new case, one or two trials are enough to inform him, and as he always has all of his twenty odd tempers on hand, it is just as easy — and far more satisfactory to both parties — to have it made right as to have it made wrong. The Sheffield manufacturer, previously referred to, calls ON TEMPER OF STEEL. 29 attention to this same experience, and very truthfully re- marks : " For many purposes, indeed, temper is of more impor- tance than quality. Nothing is more common than for steel to be rejected as bad in quality, because it has been used for a purpose for which the temper was unsuitable. We may divide consumers of steel into three classes : First, those who use their own judgment of what percentage of carbon they require, and instruct the manufacturer to send them steel of a specified temper; second, those who leave the se- lection of the temper to the judgment of the manufacturer, and instruct him to send them steel for a specified purpose ; and third, those who simply order steel of a specified size, leaving the manufacturer to guess for what purpose it is required. It cannot too often be reiterated of how much importance it is, when ordering steel, to state the purpose for which it is going to be used." And again : " You may depend upon it there is nothing so dear as cheap steel. It must be more economical to put five shil- lings' worth of labor upon steel that costs a shilling, to pro- duce a tool that lasts a day, than to put the same value of labor upon steel that costs only ninepence, to produce a tool that lasts only half a day. I am sure that the system adopted by some large consumers of buying tool steel by tender is one which in too many cases defeats the object for which it was instituted, and, by lessening the price, and con- sequently deteriorating the quality, causes the steel bill to be lessened at the cost of the labor bill, so that extravagance instead of economy is the result. In fact, it is an illustra- tion of the proverb about being penny wise and pound foolish." ON GAUGES. IN consequence of the absurdities and anomalies existing in our present system 1 of gauges, we recommend the use of the i inch as a unit of measurement. There are in use at the present time three standard gauges, as follows : Nos. 1 . 2 . 3- 4- 5 • 6. 7 ■ 8 . 9- io ii 12 i3 i5 16 17 18 19 20 21 22 London. Decimals of one inch. •oS 3 .072 .065 .05S •CM9 .040 •035 •0315 .0295 Stubs'. Decimals of one inch. .300 .2S4 •259 .2 3 S .220 203 .1S0 •165 .148 •!34 .120 .109 •095 •0S3 .072 .065 .058 .049 .042 •035 .032 .028 Brown & Sharpe's. Decimals of one inch. 2S930 25763 22942 20431 1S194 16202 H42S 12S49 JI443 101S9 09074 oScSi 07196 0640S 05706 05082 04525 04030 03589 03196 02S46 025347 ON GAUGES. 31 Nos 23 24 25 26 27 28 29 SO 31 32 33 34 35 36 37 3S 39 40 London. Decimals of one inch. .027 .025 .023 .0205 •O1S75 .0165 •0155 •OI37S .OI225 .OII25 .OIO25 .OO95 .009 .OO75 .0065 •00575 .005 •OO45 Stubs 1 . Decimals of one inch. 025 022 020 01 8 016 014 013 012 010 009 008 007 o°5 004 Brown & Sharpe's. Decimals of one inch. 022571 020I OI79 OI594 014195 OI264I OII257 OIOO25 O1S928 00795 OO70S 0063 OO561 OO5 OO445 OO3965 OO3531 OO3144 In some cases the difference between two numbers falls as low as two thousandths of an inch, in others it is only one thousandth, etc. It may be possible to make one gauge to any of these standards which shall be so accurate as to defy the detection of an error, and with the same care it may be possible to make a thousand such gauges; but every mechanic, and every person accustomed to making accurate measurements of the best work, knows that it is simply impossible to obtain abso- lute accuracy in such pieces of work when produced in large quantities, and it is impossible commercially, on account of the cost. Every one knows of the wonderful accuracy of the Whit- worth gauges, and also their enormous price, which makes them almost unsalable. In regard to ordinary wire gauges, they are notoriously 32 THE TREATMENT OF STEEL. inaccurate, because they cannot be made accurate and be at all salable. In a recent case a sample under discussion measured on one gauge tight twenty-three, and on the other light twenty- four, and our customer said it was neither by his gauge. A new gauge in our possession has its No. 23 so much larger than its No. 22 that the difference can be easily de- tected by the naked eye, yet No. 23 ought to be two to four thousandths smaller than No. 22. If we were to roll No. 23 by that gauge how "would our customer get what he wanted, unless his gauge accidentally contained the same blunder? Another trouble is with the wearing of the gauges, for which there is no remedy, and we imagine that no man ever throws away a gauge because it is worn out; on the contrary, it represents an outlay of several dollars, he is used to it, he measures everything by it, and he is mad when anything does not measure to suit it. A still more serious difficulty arises from a very common mode of ordering. We frequently have orders for such a gauge "light " or "tight," "full" or "scant," "heavy "or "easy," or such a number and one-half, for instance, 15^2. The latter is terribly confusing to a roller; he almost al- ways takes it to mean that it is to be thicker than the whole number, and is pretty certain to make it 14^-2 for 15^ if he is not warned beforehand. Then in regard to the terms "light," "easy," etc., we have, for instance, the differences between Nos. 27 and 28 in the three systems, as follows: .00225 .002 .001554 or, two hundred and twenty-five hundred thousandths, two thousandths, and fifteen hundred and fifty-four millionths. ON GAUGES. 33 How is it possible for a roller to know just how many millionths of an inch another man, whom he never saw, means when he says No. 28 "full," or No. 27 "easy"? and how is he to guess how many thousandths of an inch the other man's gauge is wrong in its make, or how many hun- dredths it has worn in years of steady use? This is no fancy sketch ; the above are every-day difficulties in this age, when every man knows just what he wants and will have nothing else, and yet has no better way of telling his wants than to say I want such a gauge " tight," when probably his gauge differs from every other gauge that was ever made. There is a very easy and simple way out of this whole snarl, and that is to abandon fixed gauges and numbers altogether, and use the micrometer sheet metal gauges, which meas- ure thousandths of an inch very accurately, and even a quarter of a thousandth may be neatly measured. They are very simple, so that any boy of ordinary intelli- gence can be taught to use one in a very few minutes. They have very easy arrangements for readjustment when worn, and even when worn considerably they can be used accu- rately, without adjustment, by making allowance for the error in reading at the zero line. We find that mechanics like to work to them, and that there is very little trouble to get sheet rolling done to within a thousandth of an inch on fine sizes. COMPRESSED RODS. IN 1878 we concluded to undertake the manufacture of compressed drill rods, an article hitherto never successfully made in this country ; and we are now credited with the ability to furnish a product which, for quality and finish, cannot be surpassed. We are prepared to furnish the following sizes : COMPARE GAUGE WITH EXACT SIZES GIVEN IN THOUSANDTHS OF AN INCH. Sizes in Sizes in Sizes in Sizes in Nos. Decimals Nos. Decimals Nos. Decimals Nos. Decimals of 1 inch. of 1 inch. of 1 inch. of 1 inch. I 0.228 16 O.177 31 O. I20 46 O.080 2 0.22I 17 O.I73 3 2 O. Il6 47 O.079 3 O.213 18 O. 170 33 O.I13 48 O.076 4 O.209 l 9 O. 166 34 O. Ill 49 O.073 5 O.206 20 O. l6l 35 O. no 50 O.070 6 O.204 21 0-I59 36 O. I06 5 1 O.067 7 0.20I 22 O.156 37 O. IO4 52 O.064 8 O.199 23 O.I54 3S O. IOI 53 O.060 9 O. 196 24 O.152 39 0. IOO 54 O.054 10 O.194 2 5 O.15O 4° 0.098 55 O.052 11 O. 191 26 O. 148 4 1 0.096 56 O.O47 12 0.1SS 27 0-H5 42 0.094 57 O.O44 13 0.185 28 O. 141 43 0.089 58 O.O42 14 0. 182 29 O.136 44 0.086 59 O.O4I 15 0. 180 30 O. 129 45 . 082 60 O.O4O 34 en o 1— 1 M) c ~o c o © C/5 ■o C/D (8 o> X s_. t- Ql E ■•J o o fc o CD ■» £ ■"■ ~c^ o "o C3 CO 00 0) -f— ' £ (0 w CI O © C/5 £ n v r a a rt O, c . 02 c W4 Iffl oa . c a JJ03 a . u O o| >- O^ Z* *£ 66 • - 6 i 5 I a CO 0. co^ fcQ a, CO q£ a. CO P£ a. CO 7.814 a. CO 7.818 7.841 7.844 • 003 7.831 -.013 7.826 -.018 7.823 -.021 -.030 7.82a 7.824 -.005 7.806 -.018 7 849 .025 7.830 .006 7 .8n -.013 7-791 7.824 7.829 .005 7.812 -.017 7.808 -.021 7.780 -.049 7-7«4 -°35 7.789 7.818 7.825 .007 7.790 -°35 7-773 -.042 7-75 8 -.067 7 • 755 -.070 7-752 7.807 7.826 .019 7.812 -.014 7-7«Q -037 7-755 -.071 7-749 -.077 7-744 7.805 7.825 .020 7. 811 -.014 7.79a -.027 7.769 -.056 7-741 -.084 7.690 026 °33 040 073 082 '35 6 5 4 3 2 1 Not heated. Low red heat. Red hot. High red. Yellow hot. Nearly white Scintillating. The twelve ingots under consideration were hammered to 1 % inch square bars at one end, and these bars were rolled to .625 diameter round bars. Six of these bars, Nos. 3, 4, 6, 8, 10, 12, were selected for specific gravity tests ; bar No. 2 was lost, or it would have been used instead of No. 3. WHY DOES STEEL HARDEN? 45 Six nicks were made around each bar at one end at inter- vals of about half an inch. The six pieces were numbered from 1 at the end to 6. Each notched bar was then heated until piece No. 1 was scintillating or nearly white hot; No. 2 was yellow hot; No. 3, high red hot; No. 4, red hot; No. 5, barely showing any red, or very low red hot ; No. 6, black. This heating was done in each case as slowly and as care- fully as possible. The results show the inevitable irregulari- ties attending only one such experiment, yet there is enough of regularity to teach us a great deal. As soon as the heats were obtained the bars were quenched in water. The pieces, carefully numbered, both with the ingot num- bers and with the numbers giving their order on the bars, were then broken off and sent to Prof. Langley to have the specific gravities determined. In the table the left-hand column gives the ingot numbers. The other columns give the specific gravities of the ingots, the bars, No. 6 pieces, and of the other five hardened pieces in order, as numbered in the sketch and explained before. The differences are, first, the difference between the Sp. Gr. of the ingots and the bars; second, the difference be- tween the Sp. Gr. of the bar, or piece No. 6, and each piece successively. The differences of Sp. Gr. are given in preference to the actual differences in volume, because the differences in vol- ume run into the infinitesimals, and the mode adopted answers as well for purposes of comparison. On comparing the ingot and bar we see a decided increase in the Sp. Gr. of the bar in every case except one, that of 46 THE TREATMENT OF STEEL. No. 4. We have not discovered the reason of this anomaly. The increase in the other cases is due to hot working; this will be shown by Table IV. It will be observed that the Sp. Gr. of the bars, except in No. 3, is nearly uniform. This seemed very strange at first, but it is capable of a very simple explanation. The hardness of steel and its re- sistance to change of form increase very rapidly with an increase of carbon, and as these bars were all reduced from 3 in. square ingots to S-g in. round bars, it is obvious that it required much more work to reduce No. 12 than No. 4 or No. 6 ; therefore, as hot working increases Sp. Gr., the greater amount of work produced the greater increase in the Sp. Gr. of No. 12. If the Sp. Gr. of the right-hand column pieces No. 1 be compared to the Sp. Gr. of the ingots, it will be seen that the relation between the numbers is entirely restored by the high heat to which the No. 1 pieces were subjected. If the Sp. Gr. of pieces Nos. 5, 4, 3 be examined care- fully, sufficient irregularities in the difference columns will be observed to show that the heating was not accomplished in regular gradations in each case, and if it were desired to lay down an exact law of variation due to differences of tem- perature, it would be necessary to take the mean of a great many experiments. Nevertheless, several general laws are indicated in this table. 1. The Sp. Gr. of the ingot varies directly with the quan- tity of iron present. 2. The greater the quantity of carbon present, the greater is the amount of work necessary to produce change of form. WHY DOES STEEL HARDEN? 47 3. The greater the quantity of carbon present, the greater is the change in volume due to a change of temperature. As, for example, in No. 3 the change in Sp. Gr. from the ingot to the bar is only .003, and from the same bar to the piece No. 1 the change is .026. While in No. 12 the change in Sp. Gr. from the ingot to the bar is .020, or about seven times that in No. 3, and the change from the bar to the piece No. 1 is .135, or about five times the change in No. 3. This is perhaps the most important observation that can be made on this series of experiments, as it shows us why it is that high steel is so much more liable to crack and break in manipulation than low steel. We generally say one is brittle and the other is ductile; we now know that the rate of expansion per degree of tem- perature is much less in low steel than in high steel. There- fore, low steel is much less liable to injurious internal strains than high steel. TABLE III. Ingot Numbers. Sp. Gr. of bars. No. 5. Sp. Gr. of burned pieces. Annealed, No. i. Difference. 7-844 7.824 7.829 7-825 7.826 7-825 7-857 7.846 7-835 7.828 7.824 7.822 + .013 + .022 6 + .006 8 + 003 — .002 10 —.003 In order to settle the question of restoring " burned steel," so called, and also to determine the reverse action due to annealing, Prof. Langley took the six pieces No. 1 of Table II, and heated them all to a high yellow heat. He then allowed them to cool very slowly. 4 8 THE TREATMENT OF STEEL. This raised a heavy scale on the pieces, which was re- moved by touching them on an emery wheel. The specific gravities of these pieces were then taken, and the results are given in the table. The restoration to the Sp. Gr. of the bar is complete, as the differences are only such as might be due to the scale on the original bars and the removal of the scale from the annealed pieces. This will be shown further in Table IV. TABLE IV. DRILL ROD SAMPLES. Nos. i Sp. Gr. 2 Sp. Gr. Hardened. Sp. Gr. Scaled and not hardnd. Diff. 3-2 Effect of hardening. Diff. 3-4 Effect of scale off in 1 and 2. 7.S068 7-794 7.816 7.787 7.818 7.812 7.79O 7-765 7.829 7.828 7.817 7-7SO — .Oil — .Ol6 — .027 —OI5 + .022 2 + .034 + .OOI — .007 It is well known that cold rolling does not increase the Sp. Gr. of iron or of steel. To ascertain the effect of cold hammering under the best conditions to increase Sp. Gr., namely, by hammering between semi-circular dies, an experi- ment was made, the results of which are recorded in Table IV. A round bar, of carbon about one per cent, was operated upon. The bar, as it came from the rolls, and unannealed, was 0.682 inch in diameter; this is No. 1 in the table. A piece of the same bar, annealed and pickled, was 0.673 inch in diameter; this is No. 2 in the table. The same bar twice hammered cold, after annealing, was reduced to 0.624 inch in diameter; this is No. 3 in the table. WHY DOES STEEL HARDEN? 49 The same bar annealed, and hammered cold four times, was reduced to 0.564 inch in diameter; this is No. 4 in the table. Prof. Langley first took the Sp. Gr. of the four pieces as he received them, 1 and 2 having the roll scale upon them, and 3 and 4 being bright polished, all having been boiled in dilute potash and slowly cooled. The results are given in column No. 1. In this case No. 3 indicates an increase of Sp. Gr. due to the cold hammering. Prof. Langley then thinking that the results might have been affected by scale in the first two pieces, next removed the scale and boiled them all in weak potash, and upon removing them from the boiling liquid cooled them rapidly by plunging them quickly into cold water. Column No. 2 gives the results, and here we have the remarkable fact that sudden cooling from boiling tempera- ture causes a hardening effect, which is shown more particu- larly in Nos. 3 and 4, where there is a decided reduction of Sp. Gr. If subsequent trials prove this deduction to be correct, it is very important. Desiring to fortify himself as to this matter of hardening at such a temperature, Prof. Langley again boiled the pieces and allowed them to cool very slowly, thus annealing them. The results are given in column No. 3. Here is a progressive reduction of Sp. Gr., showing that cold hammering as well as cold rolling reduces Sp. Gr. The restoration of the Sp. Gr. of 3 and 4 to the results in column No. 1 shows that there was a hardening due to quenching from boiling temperature. The column of differences 3 and 2 shows the effect of hardening. The column of differences 3 and 1 shows the effect of removing the scale. This column 5o THE TREATMENT OF STEEL. also accounts for the increase of Sp. Gr. shown in the "re- stored" or annealed pieces of No. i, Table I, recorded in Table III. The results recorded in Table IV have an im- portant bearing on the inquiry into the cause of hardening, which will be shown later. They are also important as show- ing the entirely mercurial or thermometric nature of steel. They also indicate a mode of accurate determination of the variable rate of change of volume in steels of different com- position. It will be remembered that there is such a variable rate of change clearly shown in Table I, and further evidence will be given in Table V. Now, by operating upon different samples by boiling, and sudden cooling in water of uniform temperature, we can get results which will range between certain uniform and known temperatures for each experiment. TABLE V. CHANGES IN VOLUME BY REPEATED HARDENING. No. of times hardened. I . 2. 3- 4- 5- 6. Total change. Nos. 6 and 7, Table I. C = about .848 Contraction of hole. .OOI72 .OOI72 .OO6SS .O06SS .0068S .OOOOO 30044 crack'd .02752 No. 4, Table I. C =.649. Expansion. Contrac- tion. .00172 not cracked .OO257 .OO086 .OO482 .OO172 .OOOOO .OO771 No. 3, Table I. C =.520. ! Con> Expansion. ltraction . .OOI72 .OO0S6 .OO0S6 not cracked .00000 ckdoS6 .00172 .00000 .000S6 .00000 Hole was originally .75 diameter. WHY DOES STEEL HARDEN? 5 1 This is an experiment to find by measurement the effect of repeated hardening upon three pieces of steel containing different amounts of carbon. A hole about .75 inch in diameter was drilled in the mid- dle of each piece. The measurements were taken by means of a tapered plug, the difference in the distance to which it entered in each case, after the first and subsequent harden- ing, was measured by micrometer. The left-hand columns give the numbers of the successive hardenings. The other columns show the changes in the diameter of the hole. The first piece, of carbon .848, showed contraction of the hole every time it was hardened except the sixth, and the piece cracked at the seventh hardening. The operator supposes the sixth hardening was accidentally omitted. The second piece, of carbon .649, showed contraction three times, then no change. Then an expansion of the hole followed by a contraction, and the seventh time there was no change. This piece did not crack. The third piece, of carbon .529, showed two contractions, then no change, followed by three expansions, and seventh a contraction. This piece did not crack. The total changes are quite marked : Showing for carbon 84S = .02752 inch. " " " 649 = .0077 1 " " " " 529 = .00000 This shows in another way that steel of high carbon changes more in volume per degree of temperature than steel of low carbon. The high steel cracked, the low did not. All the pieces were of the same quality. The experiment recorded in table No. V forms no part of the investigation by Prof. Langley and ourselves. It was 52 THE TREATMENT OF STEEL. made rather crudely for a practical purpose, and the results obtained in practice confirm the figures in the table. This ends our record of facts and brings us to — 4. A statement of some of the theories which have been given as the cause of hardening. Perhaps the oldest, one of the most plausible, and possibly the true reason, is that unhardened steel contains carbon in graphitic and uncombined form, and hardened steel has its carbon all combined. For proof it is stated that when unhardened steel is dis- solved, the insoluble residue contains flocculent graphitic carbon ; and when hardened steel is dissolved it leaves no residue of carbon ; therefore, the carbon has been combined in the hardening. To answer the objection to this, that it is impossible for iron and carbon to combine in all proportions, one writer states that there is formed a definite carbide Fe. C 4 . That this carbide is excessively hard, and that it acts as a cement or glue, and therefore the high carbon steel becomes so much harder than the low carbon steel. This will be conclusive after the carbide has been sepa- rated and thoroughly examined. Meantime, the hardening from boiling temperature is a little puzzling. Another writer states that solution of the carbon takes place when the steel is heated, and that a great compression caused by the sudden contraction in cooling is the cause of hardness. If this be so, and our experiments are correct, then carbon dissolves in steel at the temperature of boiling water. One writer hastens to inform us that steel hardens because part of the carbon is burnt out in heating, and the rest of the mass is compressed by the sudden cooling. It might afford WHY DOES STEEL HARDEN? 53 amusement to demolish this theory, if it would not be a waste of time. A steel-maker of twenty years' practice says hardening is caused by the carbon assuming the diamond form, in very minute crystals. He gives, as a proof, that the hot steel decomposes water, or the cooling mixture, which always contains hydrogen. The hydrogen combines with the carbon to form diamonds, and this is proved by the fact that the diamond and hard- ened steel both refract light. In case water is the cooling medium, the hydrogen pene- trates the steel to form diamonds, while the freed oxygen, conveniently inert, stays on the outside to form a thin film of oxide. As it is well known that mercury is one of the very best cooling liquids, giving extreme hardness to steel, it is neces- sary to this theory to show that mercury contains hydrogen. Again, if steel really hardens upon being quenched from boiling temperature, then water must be decomposed by that temperature. This diamond theory is very attractive, and has received much consideration in our minds, but we are not prepared to consider it proven. Another writer states that hardening is due to the sudden arresting of the molecular motion that exists in the heated steel, thus causing great tension and resulting hardness. He offers, as proof, that hardened steel is weaker and more brittle than unhardened steel, and cites as a very happy illustration the case of hardened glass which is known to be in a high state of tension. This theory tallies with all the facts better than any we have seen. First, it covers all conditions, from the boiling tempera- 54 THE TREATMENT OF STEEL. ture up to the high yellow heat which causes intense hardness and the brittleness of glass. Again, it is certain that the higher the heat the greater the molecular motion. Also it is certain that from the highest heat we get the greatest hardness, and the greatest brittle- ness. Finally, the restoration of grain, and of Sp. Gr. by anneal- ing, agree well with the idea of tension in one case, and the relief from tension in the other. It is possible, even if this tension theory be accepted as correct, that there may be, in connection with it, a change to diamond form, or from graphitic to combined carbon, and the formation of a definite carbide. Some one of these changes, taking place at a given temperature, may be just what is needed to expla : that very remarkable phenomenon known as "refining." In mentioning a few practical considerations to be drawn from what has been said, it is hardly necessary to address the unfortunate smith and temperer; they, poor fellows, have heard so much of uniform heating and low heating, that they may well feel heart-sick, and determined to do as they please anyhow. Let them do as they will, they will never be allowed to forget that same old cry — " too much heat" — " irregular heat" — and so on. Let that cry continue; it has its uses; and let us look at the engineer's side. As steel advances with irresistible steps into the field of construction, the engineer naturally asks — What am I to do with it ? Can it be worked safely ? Is it reliable ? WHY DOES STEEL HARDEN? 55 Shall I use high steel or low? How is it to be worked ? Is it safe to use the apparent advantages of great strength to be had in high steel? Is it necessary to anneal finished work ? etc. etc. We think it has been clearly shown — 1. That a good soft heat is safe to use if steel be immedi- ately and thoroughly worked. It is a fact that good steel will endure more pounding than any iron. 2. If steel be left long in the fire it will lose its steely nature and grain, and partake of the nature of cast iron. Steel should never be kept hot any longer than is neces- sary to the work to be done. 3. Steel is entirely mercurial under the action of heat, and a careful study of the tables will show that there must of necessity be an injurious internal strain created whenever two or more parts of the same piece are subjected to different temperatures. 4. It follows that when steel has been subjected to heat not absolutely uniform over the whole mass, careful annealing should be resorted to. 5. As the change of volume due to a degree of heat in- creases directly and rapidly with the quantity of carbon present, therefore high steel is more liable to dangerous in- ternal strains than low steel, and great care should be exer- cised in the use of high steel. 6. Hot steel should always be put in a perfectly dry place of even temperature while cooling. A wet place in the floor might be sufficient to cause serious injury. 7. Never let any one fool you with the statement that his steel possesses a peculiar property which enables it to be 56 THE TREATMENT OF STEEL. "restored" after being "burned;" no more should you waste any money on nostrums for restoring burned steel. We have shown how to restore " overheated" steel. For " burned" steel, which is oxydized steel, there is only one way of restoration, and that is through the knobbling fire or the blast furnace. " Overheating" and " restoring" should only be tried for purposes of experiment. The process is one of disintegra- tion, and is always injurious. 8. Be careful not to overdo the annealing process ; if carried too far it does great harm ; and it is one of the com- monest modes of destruction which the steel-maker meets in his daily troubles. It is hard to induce the average worker in steel to believe that very little annealing is necessary, and that a very little is really more efficacious than a great deal. Finally, it is obvious that as steel is governed by certain and invariable laws in all of the changes mentioned, which laws are not yet as clearly defined as they should be, nor as they will be ; nevertheless, the fact that there are such laws should give us confidence in the use of the material, because we may be sure of reaching reliable results by the proper ob- servance of the laws, therefore there is no good reason why en- gineers should be afraid to use steel if they manipulate it in- telligently. Now, if we have wandered over a wide range in answer to the simple question — " Why does steel harden ?" — it was necessary to have looked at many facts before we could have an intelligent opinion of many theories ; and if any are in doubt as to what is the correct answer to this momentous question, we can only say that we are all in the same boat, for if you do not know, neither do we. Overheated steel tells its own story. A high heat opens the grain of steel and prevents refining. Cast steel properly hardened is invariably refined thereby. The temper of steel is regulated by percentage of carbon. The iron used in making steel determines its quality. Consumers of steel should be guided by makers' advices. Sulphur is an enemy to steel. Good fuel is essential to best results in working steel. Charcoal, from sound wood, is the king of fuels. Avoid exposing hot steel to draughts of air. Pure water is a good hardening medium. Cherry red is a safe heat for steel. Nick cold steel with sharp chisels. A good softening heat, for forging, can be safely used if proper precautions are observed. 57 58 THE TREATMENT OK STEEL. The lead screws of lathes are often responsible for inac- curacies in the threads of taps. Large pieces may often be protected in heating by being covered with a coating of dry clay. No annealing is better than over-annealing. Be sparing of the blast. Time and care are necessary in the treatment of steel. Burnt steel is disintegrated steel. Properly constructed furnaces will pay for themselves in the value of tools which will be saved by their use. '"Soaking," or long continued heats, even if low, are in- jurious to steel. It requires a high order of talent to treat high grade steel successfully. In heating for hardening, great care should be taken with irregular shapes that no part of the piece be too hot. Steel which has been annealed at a high heat will not harden on the surface. A low red heat will anneal steel thoroughly. Do not try to harden steel that has been annealed, before turning off the surface. He is a good steel worker who never spoilt a piece of steel. Do not be deceived by nostrums for restoring burnt steel. The peculiarities and proper treatment of steel are stud- ies for a lifetime. SPARKS. 59 Benjamin Huntsman is said to have invented the cast steel process in 1770. Crucible cast steel is recognized the world over as the best steel. Low priced steel may be very dear steel. Steel is mercurial and delicately responsive to heat; its records appear in its own structure. The last age : the age of steel. Dirty, slack fires put dirty sulphurous oxides on steel. The hardening heat varies with each temper of steel, and the only safe course is to harden at the lowest heat that, on trial, is found to give the required hardness. Hardening cracks are more often the result of uneven heating than of any defect in steel. FINE STEEL OUR SPECIALTY. Established 1865. Reuben Miller, Wm. Metcalf, Chas. Parkin, General Partners, Geo. W. Barr. Special Partner. CRESCENT STEEL WORKS, Miller, Metcalf & Parkin OFFICE, 81 WOOD STREET, PITTSBURGH, PA. BRANCH OFFICE, 22 W. LAKE ST., CHICAGO, ILL. Capacity 20 to 25 Tons of Crucible Steel Daily. 62 crescent steel works. Crescent Tool Steel, For all purposes requiring steel of standard quality, such as Axes and Edge Tools, Sets, Track Tools, Smith Tools, Chisels and Mining Drills, etc. Crescent Extra Tool Steel, For all purposes requiring steel of extra fine quality, such as Lathe Tools, Cutters, Reamers, Chisels, Drills, Dies, Mill Picks, Taps, Punches, Nailers' Dies and Knives, etc. Crescent Special Tool Steel, A special steel for difficult and expensive tools. Made only to order. Very fine. Soft Center Tap Steel. Single and Double Shear Steel. MACHINERY STEEL, Crucible and Bessemer. Crucible Cast Spring Steel. Crucible Cast Tire Steel. Crucible Cast Hammer Steel. Crucible. Bessemer So — * ^^— . and HEET OTEEL, - - Crescent. 2d Qual. 3d Qual. Open Hearth. Hot Rolled, Cold Rolled, Planished, in Tool Grade and " Common." STEEL FOR DIFFICULT TOOLS A SPECIALTY. crescent steel works. 63 Polished Compressed Drill Wire and Rods, Accurately finished to micrometric gauge, all number and letter sizes from No. 60 gauge to \\ inch. Needle Wire. Steel for Surgical Instruments and Fine Cutlery, Jewelers' Rolls, Cutting and Drop Dies, Nail Knives and Dies, Rolling Mill Knives, Chilled Roll Lathe Tools, etc. Patent Straight Flanged and Spirally Twisted Drill Steel, for Rock Drilling Machines. Mining Drill Steel, For extra hard rock. Railway Steel, Shop and Track Tool. Locomotive and Car Axles. Frog Points and Plates. Frog Side Bars. Crank Pins. Lathe Spindles. Piston Rods, plain and forged. Slide Bars. Best, Crucible Cast, Locomotive and Car Spring Steel. Spiral Car Springs, Patent Edge Rolled. Patent Equal Bar. "Hebbard" Flat Coil. Round Coil Nest. Spiral Group. Volute, etc. Drop Forged Spring Plates. 64 CRESCENT STEEL WORKS. Sheet Steel, Clock Spring. Watch Spring. Band Saw. Mulay, Billet Web and Hand Saw. Corset, Key Blank and Planished Hoe. Sewing Machine Blanks, etc. "High Tensile" Homogeneous Steel, For Connecting Rods, etc. Miscellaneous, Auger and Auger Bit. Axle Steel for Carriages and Wagons. Pick and Mattock, plain and beveled. Pike and Cant Hook. Coal and Granite Wedge. Fork and Rake. Hoe. Roller. Spindle. Trap Spring Steel. Sleigh Shoe. Swedish Toe Calk. " Crescent " Brands are sold by Dealers in all sections. I ■ ■