/' Ll^u dSolut on 2100 2000 1800 1800 1700 1600 ft 160C 1400 1300 1200 1100 1000 900 Solid * Iron^, \ \ \y + Jraphit ^ Gm. a Solid Solutio n-1-Gr.j bit V 5X ol n.)^ | V / 6 mmJ onFe t c a iron I SoM so . Jphalr utF., P Alpha I ron-t-Fe 2 123 Percent Carbon FIG. 44. Effect of temperature on carbon. Probably Bessemer steel would show lower figures than this and crucible or electric steel higher figures. This is doubtless due to the fact that some processes of steel making remove the oxides and occluded gases better, or to a greater extent, than others, and it has been pretty well demonstrated that these are injurious to the strength and life of steel. INGREDIENTS OF AND MATERIALS USED IN STEEL 71 alter these figures, but with all other conditions equal they will probably hold good. That the maximum strength is placed at about 1% carbon is probably due, to a great extent, to the fact that the crystalline constituents form an intimate mixture near the eutectoid proportions, and hence the crys- tallization is very small comparatively. With more carbon (cementite) present the pearlite grains are surrounded with a network of cementite, while with less the pearlite grains are surrounded with a network of ferrite, and both of these decrease the cohesive force inherent in the metal. One of the oldest theories as to what made high-carbon or tool steel harden was that the carbon in unhardened steel was partially in the graphitic and uncombined form, and when it was hardened all the carbon assumed the combined form. The most generally accepted theory, how- ever, assumes that the hardness is due partly to the presence of the solid solution of carbon in iron, and partly to the iron being in the gamma or beta forms. This solid solution of carbon in gamma or beta iron is exceed- ingly hard and it is preserved in the steel by quenching from above the critical temperature. MANGANESE Manganese occurs in nature principally in the form of manganese dioxide (Mn0 2 ), which is commonly called black oxide of manganese, but occasionally it is found in other compounds, such as braunite, man- ganite, carbonate, etc. Some of its compounds with oxygen and hydrogen are distinctly acids while others are distinctly basic, and it is in connec- tion with the base-forming elements that it is of interest in steel making. For use in steel making the dioxide is separated from its oxygen, in the presence of charcoal or coke, either in the blast furnace or in an electric fur- nace. It looks like cast iron, is brittle and hard, and is combined with iron to form ferro-manganese. Sometimes silicon is added to form ferro- silicon-manganese. Manganese is an element that is always found in steel, but its true properties and effects were not known until about twenty years ago, when they were discovered by R. A. Hadfield, a metallurgist and steel maker of Sheffield, England. Its effect when added to steel up to 2% with various percentages of carbon is best shown by Fig. 45, the actual mode of existence of the carbon in the steel being very important. When more than 2% and less than 6% of manganese is added, with the carbon less than 0.5%, it makes steel very brittle, so that it can be powdered under a hand hammer. From 6% of manganese up, this brittle- ness gradually disappears until 12% is reached, when the former strength returns and reaches its maximum at about 14%. After this a decrease in toughness, but not in transverse strength, takes place, until 20% is 72 COMPOSITION AND HEAT-TREATMENT OF STEEL reached, after which a rapid decrease again takes place. Manganese may affect the tensile strength and ductility of steel, either indirectly .4 .6 .7 A .9 1.0 Percentage of Cartxw FIG. 45. Effect of manganese below 2 per cent. , by retarding the formation of blow-holes, or directly by entering into chemical combination with the metal. 10 11 12 13 14 15 16 17 18 19 20 Percentage of Manganese FIG. 46. Effect of Manganese above 6 per cent. Fig. 46 shows the effect of more than 6% of manganese on the tensile strength and elongation. INGREDIENTS OF AND MATERIALS USED IN STEEL 73 Steel with from 10 to 15% of manganese and less than 0.50% of car- bon is very hard and cannot be machined or drilled in the ordinary way; yet it is so tough that it can be twisted and bent into peculiar shapes without breaking. This makes a steel that is only suitable for casting into the desired shape. A process has recently been patented however, for casting this steel into ingots, and then subjecting them to a heat treat- ment that enables them to be mechanically worked; that is rolled, forged, etc., and this might possibly be extended to machining operation. Manganese in the form of a ferro-alloy containing about 80% of man- ganese is added to a heat of steel at the time of tapping, so that it may seize the oxygen which is dissolved in the bath and transfer it to the slag as oxide of manganese. Manganese prevents the coarse crystallization that the impurities would otherwise induce, and steels low in phosphorus and sulphur require less manganese than those having comparatively high and percentages. Manganese has a greater affinity than iron for both sulphur and oxygen, and is therefore used in steel making as a deoxidizer and to neutralize the sulphur. Manganese oxide (MnO) and manganese sulphide (MnS) are formed, the first of which passes almost entirely into the slag and the sec- ond of which will pass partly into the slag if time is allowed. About four times as much manganese is needed as there is sulphur present, as it does not always catch all of the sulphur; thus if any great amount of sulphur is present a considerable amount of manganese is desired to counteract its effect. If the bath is kept liquid enough and enough manganese is present, but little oxygen or sulphur will be found combined with the iron, which is desirable as they are very injurious to the metal. The length of time and the care required, however, make it commercially impractical to reduce the oxygen and sulphur to a trace in this way. Therefore man- ganese is used to reduce them to commercial percentages, and other mate- rials are used to still further remove them for the finer grades of steel. Manganese sulphide weakens steel greatly if segregated together with phosphide of iron, especially if the metal is rolled, as this magnifies the sulphide by spreading it out during rolling. Manganese is not only useful to cleanse the bath of impurities, but it has other properties that aid in making steel better. The amount that can be left in the steel varies with the amount of various other ingre- dients that are added to the metal, and this is especially so of carbon. In effect it behaves in practically the same manner as carbon, as also does nickel. With a given carbon content the introduction and increase of manganese causes a series of structural changes similar to those that occur in carbon steels, that only contain small percentages of manganese. While the action of these three elements upon iron is of the same kind, 74 COMPOSITION AND HEAT-TREATMENT OF STEEL it is not of the same strength, as the equivalent of 1% of total carbon, that contains the maximum amount of hardening carbon, is 7.25% of manganese and 17.55% of nickel. All three of these cause a structural change in the metal from pearlite, that includes the sorbitic, to marten- site, that includes the troostitic, and then to the polyhedral structure, and with none of them is a special carbide formed. Chromium has an analogous effect, but not as complete, as a double carbide of iron and chromium forms and this is not maintained in solution in the iron without tempering. The critical temperature to which it is safe to heat steel is raised by manganese, owing to its resisting the separation of the crystals in cooling from liquid, and conferring the quality of hot ductility. It also assists" in producing more uniform alloys, and tends to make steel crystals smaller by making the metal plastic, and thus counteracting the tendency toward crystallization that phosphorus causes, although the metal is more liable to crack when heating or suddenly cooling it from a red heat. The good qualities more than offset the bad, and it is a very useful factor in steel making if the proper percentages are used. It atones for many evils in steel by healing it up and producing a smoothly rolled surface. In the ordinary steels this percentage is usually from 0.70 to 1.00%, while in many of the special alloys it runs from 0.30 to 0.50%. In the high-speed steels the manganese content is from 0.10 to 0.30%, and in steels for carbonizing this should be kept below 0.20%. When in very large amounts (from 6.0 to 15.0%) it reverses the effects of rate of cool- ing upon the ductility of steel; slow cooling making manganese steel brittle, while quick cooling makes it extremely ductile. Magnetic qualities are not materially effected when the manganese is kept below 7.50%. When 8%, or more, is present, however, the mag- netic attraction becomes nil. Manganese decreases the electric con- ductivity in greater proportions than any other element, except nickel. Thus third rails, or similar steels, must be given their hardness by materials other than nickel or manganese. A peculiar fact that was brought out by some experiments was that a pure nickel-iron alloy that contained from 12 to 13% of nickel was highly magnetic, but by the addition of 5% of manganese this metal became as non-magnetic as brass. While manganese steels are known to be non-magnetic, it was not known that manganese would have this effect upon nickel, which also makes a non-magnetic steel when added in certain proportion. To sum up, manganese alloys with iron in all ratios, it being reduced from its oxides at a white heat by carbon. Its presence increases the power of carbon to combine with iron at a very high temperature (about 2550 F.), and almost entirely prevents its separation into graphitic, carbon at the lower temperatures, Manganese permits a INGREDIENTS OF AND MATERIALS USED IN STEEL 75 higher total carbon by raising the saturation point, and it is easily separated from iron by oxidation, as it is even oxidized by silica. While it does not counteract the cold shortness caused by phosphorus, it does prevent to some extent the red and yellow hot shortness caused by sulphur. Manganese retards the formation of blow-holes, though not to the extent that silicon does, by preventing the oxidation of carbon, and thus the formation of carbonic oxide. It also increases the solubility of the gases in the steel while solidifying. It probably raises the elastic limit and slightly increases the tensile strength; adds fluidity to the metal; increases hardness; increases fusibility when present in considerable quan- tity, and gives greater plasticity and mobility to the metal at forging heats. Some recent investigations, however, make it doubtful that it diminishes ductility to any extent. SILICON Silicon is the second most important element in the solid part of the earth's crust, oxygen being first, and forms 27.21% of it. It is never found in the free state of nature, but, having a powerful affinity for oxygen, it occurs chiefly as silicon dioxide (Si02), which is commonly called silica, and in the form of silicates in combination with oxygen and such metallic elements as sodium, potassium, aluminum, and calcium. Silica will neutralize any base with which it comes in contact when molten, and all metallurgical slags are silicates thus formed. The silicon used in steel making has to be separated from the oxygen of the silica and united with iron to make ferro-silicon. Sometimes manganese is added to this to form ferro-silicon-manganese. Many contradictory statements have been made as to the effect of silicon on steel. When the silicon is high in Bessemer steelyt is an indica- tion that the metal has been blown too hot, and the metal is apt to be brittle. The percentage varies considerably, according to the heat of the charge, and this causes irregularities which may account for the differ- ence of opinion. The melting point and specific weight of pig iron are governed chiefly by the silicon and the carbon, which are the principal elements. To obtain strength and density the silicon and carbon should both be low. The hardness should be controlled by a careful adjustment of the sulphur, manganese, and phosphorus, together with a study of their effect on the final condition of the carbon. In making castings high and low silicon irons should never be mixed. To get the best results in the steel the silicon should be eliminated as much as pos- sible from the iron and a definite quantity added in the form of high percentage ferro-silicon, or ferro-manganese-silicon. This gives a very different effect from that of silicon left in the process of manufacture. 76 COMPOSITION AND HEAT-TREATMENT OF STEEL If silicon is added to steel in such a manner as to cause it to enter into solution as silicide, it confers upon the metal valuable properties; but if it forms a silicate it is injurious in many ways, even to the point of being dangerous. This latter seldom occurs, or at least occurs only to a slight 'degree, as the silicates of iron, manganese, etc., dissolve into each other very readily and form a slag; although manganese silicate probably occurs more frequently and causes more failures in steel than is generally supposed. Silicon, having a great affinity for oxygen, it seizes this wherever found, and carries it off into the slag, whether in the form of gases, oxides, or dissolved oxygen. This prevents the formation of blow-holes, and makes the steel harder and tougher. Thus it is better able to withstand wear or crushing from continual pounding. This is only so, however, when the silicon has been eliminated as far as possible from the pig iron and is again added to the steel bath in the form of ferro-silicon or silicon spiegel. Otherwise the steel is liable to show brittleness and irregularity of per- centage. One steel maker found that if the percentage of manganese plus 5.2 times the percentage of silicon were made to equal 2.05, the metal would be entirely free from blow-holes, but the pipe would be large; if the total was made to equal 1.66%, the pipe would be smaller and numerous minute blow-holes would appear, but not enough to harm the steel for the use to which it was to be put. He also found that 0.0184% of aluminum would give the same result as the 1.66% of manganese and silicon. In the Bessemer converters the silicon increases the temperature of the bath. Thus the lower the percentage of the silicon in the pig iron the shorter will be the blow. At the end of the blow, 0.2% of silicon is added to rid the bath of the gases. Thus the percentage of silicon is usually under 0.2 in Bessemer steel, and for steel rails many engineers are limiting it to 0.1%. During the " killing" in the crucible process the steel absorbs silicon from the crucible and thus becomes sound by throwing off the gases. The graphite crucibles used in this country give up more silicon than the clay crucibles used in Europe, and consequently allowances have to be made when charging. Too long " killing" makes the steel harsh, brittle, and weak, owing to its absorbing too much silicon. Crucible steels nearly always contain more than 0.2% } of silicon. The influence of silicon on the results of quenching is similar to that of carbon in many ways. It is also dependent upon the coexisting amount of carbon and manganese. It neutralizes the injurious tendency of man- ganese to some extent. An increase in the percentage of silicon slightly raises the tensile strength and lowers the elongation and reduction of area. Up to a INGREDIENTS OF AND MATERIALS USED IN STEEL 77 content of 4%, silicon increases the tensile strength about 80 pounds per square inch for every 0.01%. Beyond this amount a weakening of the metal seems to ta*ke place. Without a considerable percentage of manganese, silicon steels show very low shock resistance, whether annealed or quenched. With 0.20% of silicon the tensile strength is increased about one-third more than 0.01% of carbon would increase it. Beyond a content of 5.0% silicon steels are but little used for any purposes. Steels containing a little less than 1% of carbon and from 1 to 2% of silicon have been used quite successfully for hard-tool steels. Below a content of 1% silicon ceases to have an influence on quenching and the metal may be classed as a special carbon steel. Some makers of steel try to keep the silicon as low as possible, but many of the best steels con- tain from 0.20 to 0.80%. With the carbon content low the silicon may be raised to a fairly high figure, but with the carbon high the silicon should be kept low. It should also be kept low when the phosphorus is high. Silicon steels are extremely fibrous with a remarkable resistance to shock in the direction of lamination, but practically no resistance in a direction perpendicular thereto. This quality makes them especially adapted for leaf springs. Ferro-silicon, as now made in the electric furnace, with a silicon con- tent between 30 and 60%, is very brittle and liable % to disintegrate spon- taneously, even though made of comparatively pure material. With the silicon in any percentage from 30 to 40 and 47 to 65, it gives off quite large quantities of hydrogen phosphide gas, especially when attacked by moisture in any form. This is generated from calcium phosphide, which in turn is formed from the calcium phosphate, that is present in the quartz and anthracite, when it is submitted to the high temperature of the elec- tric furnace. Smaller amounts of hydrogen arsenide are also evolved, and both of these are highly poisonous. When the ferro-silicon disinte- grates, the amount evolved is greater, owing to the largely increased surface that is exposed. Several fatal accidents by explosions and poisoning have been caused from these gases since ferro-silicon has been manufactured in the elec- tric furnace. Most of these have occurred when shipping it on boats, as there is then more moisture to attack it. When the silicon content is below 30 or above 65% these gases do not appear to evolve in amounts that are dangerous. As it is not really necessary to use the alloys between these percentages for the manufacture of any of the iron products, unless it be for basic furnace steel, their use, if not their manufacture, should be prohibited. Where absolutely necessary to use them, the ferro-silicon should be broken up into usable sizes and completely exposed to the air for at least one month before shipping. It should then be stored in a 78 COMPOSITION AND HEAT-TREATMENT OF STEEL place where there is plenty of ventilation to carry off the gases. Ferro- silicon made in a blast furnace, however, does not give off any of these gases, and there is a movement started in Europe by the electric furnace ferro-silicon makers to abandon the manufacture of this alloy in the dangerous percentages. PHOSPHORUS Phosphorus always occurs in nature in the combined condition, in the form of phosphates, derived from orthophosphoric acid H 3 PO 4 , or in the form of organic compounds. It unites with metals to form phos- phides. It forms two oxides, namely, P 2 O 3 and P2O 5 , and also forms compounds of the same character and analogous composition to arsenic, antimony, and bismuth. It is consequently placed in the same chemical group as these. It is always encountered in reducing iron ores, and is a very difficult element to remove entirely from the finished iron and steel product. In these materials it must be reduced to as low a percentage as possible, as phosphorus is without doubt the most injurious element that is found in steel, notwithstanding the fact that in the past many experiments have been carried on that apparently proved that phosphorus, up to about 0.12%, strengthened steel. When these same steels were put into actual use, however, failures occurred, and the cause was nearly always traceable to the phosphorus. In the rolling mills phosphorus does not show any bad effect, as the heat under which the steel is worked seems to overcome this, but when the metal has become cooled and is subjected to sudden shock or to vibra- tional stresses, it breaks very easily. The lower the temperature and the higher the atmosphere the easier will the breaks occur. This has led to the term, " cold-shortness," as applied to the effect of phosphorus on steel. Phosphorus diminishes the ductility of steel under gradually applied load, as shown by the reduction of area, elongation, and elastic ratio when specimens are pulled apart in the ordinary static strength testing machines. But when the steel is tested in rotary or alternating vibration testing machines, as well as with a pendulum impact machine, the decrease in ductility and toughness is shown to a greater degree. Phosphorus also reduces deflection, and the rigidity thus imparted might be considered an advantage for structural purposes except for the metal's weakness at low temperatures and when subjected to shocks. Phosphorus steels are so capricious that they may show a reasonably high static ductility and still show very brittle when shock tests are applied. Therefore the safest rule to apply is to have the phosphorus in all steel products as low as possible. It is a very poor steel that contains 0.10% INGREDIENTS OF AND MATERIALS USED IN STEEL 79 of phosphorus. The ordinary grades contain as much as 0.08%, and the high-grade steels should have less than 0.04%, while in the very best steels it should be even lower than this. In fact, this has been reduced to below 0.01% in some of the electric furnace steels, and occasionally a mere trace is all that is left in the finished product. Phosphorus gets into the metal by entering the blast furnace with the ores in the form of metallic phosphates, the form in which it is usually found in nature, and mainly as phosphate of lime, which occurs as a natural mineral named apatite. Many metallic oxides unite with it to form salts, especially iron and magnesium oxides and lime, but in the presence of silica, which is a stronger acid, it is driven out of the slags and returned to the iron until the silica has been satisfied with bases. In steel, phosphorus has a tendency to cause coarse crystals to form, and this tendency is increased with each percentage of carbon. It forms the phosphide, Fe 3 P, and this forms a series of alloys with iron. The eutectic of this series contains 64% of this phosphide, which equals 10.24% of phosphorus. A certain percentage of phosphorus will dissolve in pure iron and no eutectic will form to produce brittleness, but when carbon is added, each increase in percentage exerts an influence on the phosphorus that causes it to precipitate from the solid ferrite solution and take the eutectic form. Therefore the more pure the iron and the less cementite that is in the steel, the less will be the brittleness that is caused by phos- phorus, while each increase in the percentage of carbon -increases the tendency of the eutectic to form and the steel to assume a coarser crystal- lization, which makes it both weaker and more brittle. Phosphorus is removed from steel to a different degree by the different processes of manufacture. Thus Bessemer steel usually contains the highest percentage of phosphorus, while the other steels contain gradually decreasing percentages in the order in which they are named: acid open- hearth; basic open-hearth; crucible; electric. The acid open-hearth fur- nace requires materials low in phosphorus, while in the basic it is removed by adding a sufficient amount of lime to the slag, and in electric furnaces it is removed by using the proper flux. One of the latest methods consists of using an oxidizing slag in such a way that it will combine with the phosphorus and form a phosphate, and then adding a reducing material to the slag that will convert this phosphate into a phosphide. The reducing material is usually ground coke that floats on top of the slag and reduces the phosphate without interfering with the molten metal below. Owing to the strong com- bination of the phosphide, the phosphorus cannot be separated out by the iron, without first being changed back to phosphate, and this is impos- sible in a reducing atmosphere. 80 COMPOSITION AND HEAT-TREATMENT OF STEEL One charge that was dephosphorized in the Heroult electric furnace, when taken from a Bessemer converter, analyzed: phosphorus, 0.10%, sulphur, 0.16%, manganese, 0.10%, carbon, 0.07%, and silicon, traces. A 15-ton charge of this was put in the Heroult furnace and a black slag composed of 400 pounds each of mill scale and lime was added. This made an oxidizing slag that became fluid when the molten metal below was thoroughly oxidized, and all the phosphorus passed into the slag as phosphate of lime (CaO). Ground coke was then added to the top of the slag, and this reduced the phosphates there into calcium phosphide (P 2 Ca 3 ). Without removing the slag the required amounts of carbon, silicon, manganese, etc., were added. That there was no return of phos- phorus to the steel is shown by the analysis of the final product, which was, phosphorus 0.005%, sulphur, 0.005%, with the carbon ranging from 0.05 to 1.50%, and the manganese and silicon as desired. A high phosphorus steel is sometimes used for the third rail in an electric railway, as phosphorus will increase hardness without decreasing electric conductivity as much as other ingredients would, and it also decreases the purity of the iron less than any other material. This gives the rails the necessary hardness and purity to withstand the abrasive wear caused by the contact shoes, without greatly lessening the conductivity. Phosphorus in cast-iron reduces the melting point, makes the metal more fluid, and prolongs the period of solidification. This is made useful in such work as art castings, where a detail of figure is of more importance than strength, as the metal fills every minute crevice in the molds. By keeping the metal in a pasty state for a long time, or retarding solidifica- tion, the phosphorus allows the graphite to be expelled from the solid solution and occupy spaces between the particles of iron. This action causes the metal to expand and press into every tiny cavity in the mold, and the higher the percentage of phosphorus the longer will the solidifi- cation be delayed. Certain chemical conditions caused by too much phos- phorus, too little silicon, etc., might overcome this by exerting a tendency to keep the carbon in the combined form. A decreased shrinkage because of this expansion may also be caused when the phosphorus separates from its solution in the ferrite and forms a eutectic. Phosphorus also increases the tendency toward segregation. SULPHUR Sulphur is one of the elements of the earth that is found in large quan- tities in the free state, especially in volcanic regions, as well as combined with metals in the form of sulphides. It is given off from the fuels used in reducing the iron ores and refining steel, and at the higher temperatures it combines with the oxygen of the air to form a dioxide, S0 2 . Part of INGREDIENTS OF AND MATERIALS USED IN STEEL 81 this is liable to be trapped in the metal unless precautions are taken or slags used to remove the sulphur. When in steel in the form of sulphide, it causes the metal to crack, tear, and check in rolling, forging, heat treating, or hot working, and, therefore, the term of " hot-shortness " has been applied to its effect on steel. This is the opposite of the effect of phosphorus. Its effect on the properties of steel when cold has not been accurately determined, but it seems certain that the effect is not detrimental to any extent. When steel is heated beyond a dull red, sulphur in the sulphide form is said to cause a crystallization to take place, and when high temper- atures are reached the grain becomes very coarse, as the sulphur is dis- sociated and forms into a gas that diffuses between the iron crystals, thus separating them arid preventing perfect cohesion. When contrac- tion by cooling takes place this may cause microscopic cracks, or even cracks large enough to be seen by the naked eye. These, of course, weaken the metal. Sulphur and phosphorus increase the tendency toward segre- gation. Sulphur takes two forms in steel, one of which is sulphide of iron, and the other sulphide of manganese. Iron sulphide (FeS) usually forms when the sulphur is high and the manganese low, as sulphur has a greater affinity for manganese than for iron. Until the manganese is satisfied, sulphide of iron is not liable to occur, and this latter form does not often occur in commercial steels. It is more brittle than manganese sulphide, and at the proper temperatures for rolling steels is in a liquid state, so that there is no cohesion between it and the molecules of steel. Instead of coming together in drops, as manganese sulphide does, it spreads out in webs or sheets, which are very pale in color and usually completely sur- round the manganese sulphide. These cover a comparatively large area, and the effect of iron sulphide is thus very injurious to steel, as it is very weak and liable to break along these webs or sheets. Owing to its liquid state iron sulphide is very liable to cause trouble at the rolling tempera- tures, whether this temperature be used for rolling or when forging, weld- ing, or heat-treating the steel. Sulphide of manganese (MnS) is formed by the uniting of manganese and sulphur, and it is invariably found in steel; this being the form that sulphur takes in all the good grades of steel, and if there is enough man- ganese present all of the sulphur in the metal will assume this form. It usually forms in globular spots, but when the metal is rolled or hammered, these generally elongate and under the microscope they show a pale state or dove gray color. Opinion differs as to the injurious effect of manganese sulphide upon steel, but, however this may be, it is not as injurious as iron sulphide. It has been melted in coke-fired assay furnaces that would not melt mild 82 COMPOSITION AND HEAT-TREATMENT OF STEEL steel, which would indicate that it was injurious when steel was heated to comparatively high temperatures. It frequently occurs with man- ganese silicate (slag) and it segregates together with phosphide of iron in the form of ghosts. In this case it may be very injurious to steel, and especially so where the sulphide is spread out into threads or ribbons by rolling the metal. Sulphur, when added to soft cast iron that is low in sulphur, increases the strength of the metal, partly by closing the grain and partly by in- creasing the combined carbon. Owing to this tendency to increase the combined carbon and form an iron carbide, it has a hardening effect on the metal. Its effect on the tensile strength of steel has not been definitely settled, but up to 0.10% it does not alter the elastic ratio, elongation, or reduction of area to any extent. The actual percentage of sulphur at which steel ceases to be malleable or weldable varies with other ingredients. Each increment of manganese raises it, and it is lowered if the steel ingots are cast too hot. Attention is being turned to the effect of sulphur, noted in the pre- ceding paragraph, and the old theory that sulphur should be reduced to a mere trace in steel is beginning to be doubted, as some of these effects could be made beneficial if the injurious effects could be overcome. Some recent investigations have led to the belief that the oxides are the real source of weakness and failures in steel, and if these can be removed, the injurious effects of sulphur can at least be nullified, with the probability of its being made beneficial. According to the old theory, 0.08% of sulphur made crucible steel absolutely worthless for welding, forging, rolling, etc., but I have recently seen samples of crucible steel that had the oxides reduced to a minimum, and the sulphur at 0.08%, that were forged under the steam hammer without any signs of checks. A piece of this same steel which con- tained 0.60% of carbon was welded onto machinery steel to form the cutting edge of an axe, and apparently the weld was perfect, as there was no signs of a crack when it was ground to shape. This axe was stood on an anvil with the cutting edge up, and given 20 blows with a heavy sledge before the edge broke, and even then the weld was not harmed. Another test was to drift a hole 4 inches in diameter in stock 1 J inches thick and 4 inches wide without destroying the drift. In still another test a f-inch set stood 200 blows from a 12-pound sledge without breaking. This same set was then used in the daily work at the mill until it was worn out, and it outlasted two sets made from stock steel. The tensile strength was a little better than the ordinary in this high sulphur steel. The sulphur content was carried still higher in later tests and it was found that with sulphur up to 0.13 per cent, no injurious effects were apparent in the steel and the metal did not develop the " hot-shortness " INGREDIENTS OF AND MATERIALS USED IN STEEL 83 that every one heretofore has attributed to sulphur. Above a sulphur content of 0.13% the metal began to show signs of brittleness and was clearly injured. With steels as ordinarily made at present the sulphur should not exceed 0.10% for any use, but for tool making or other uses where the metal has to be repeatedly heated and cooled this should not be over 0.03%, and preferably as much lower as possible. Steel as now made would be much better for nearly all kinds of work if the sulphur could be reduced to a trace. OXYGEN, HYDROGEN, AND NITROGEN Of all the elements that enter into the composition of the earth's crust, oxygen forms nearly one-half, or, to be more explicit, 47.29%. It comprises eight-ninths of water and about one-fifth of the air. It occurs also in combination with carbon and hydrogen, and with carbon, hydrogen, and nitrogen. Besides this it forms a part of most manufac- tured chemical products. The iron ores that are chiefly used for making iron are combinations of ferrite and oxygen. At the higher tempera- tures it has a greater or lesser affinity for and unites with every other elemental substance known, except fluorine, helium, neon, argon, krypton, and xenon, and it acts readily upon a large number of compounds. At the ordinary temperatures oxygen does not act readily upon most things. Its simple compounds are called oxides, and these usually form with the production of heat. One of the elements that combine with it at low temperatures is iron, and this is coated with an oxide when heated to about 400 F., or at nearly any temperature in the presence of moisture. Hydrogen is the lightest substance known, and like oxygen is a gas that is colorless, tasteless, and odorless. It has a high chemical affinity for oxygen, and is a good reducing agent. It forms with carbon something like 200 combinations, known as hydrocarbons. At a red heat it pene- trates iron readily, probably forming a compound with it. Nitrogen, owing to its inactivity, acts principally as a dilutent of oxygen. These three gases readily dissolve in iron or steel when it is molten, but as it solidifies comes out of the state of solution, and then much the larger part passes away. A portion, however, is usually entrapped, and this portion if segregated in large bodies causes blow-holes, gas bubbles, etc. Carbon monoxide gas (CO), which may be generated during the solidification period by a reaction of the oxide of iron with the carbon when carburizing, is also a cause of blow-holes. These blow-holes are usually removed by the use of the deoxidizers, such as manganese, silicon, aluminum, etc. Another portion of these gases, however, is liable to remain in the steel in the form of occluded gases and oxides that are just 84 COMPOSITION AND HEAT-TREATMENT OF STEEL beginning to be recognized as among the most harmful things in steel; oxygen probably being the most weakening element that can be left in steel, with hydrogen and nitrogen closely following. As evidence of this, Bessemer steel, which is purified by blowing air through it, is the poorest and weakest of steels; while open-hearth steel, which is purified without this blast of air, but is not protected from the air striking the surface of the bath, comes next; and crucible steel being protected from air by the melting process taking place in a closed pot, is the strongest and finest grained of all the steels, except those made in the electric furnace, and this is also protected from the air. Another proof is the added static and dynamic strength, wearing qualities, etc., given to steels, by such elements as vanadium, titanium, etc., when they are used to cleanse the metal of these gases. Oxide occurs in very small black specks throughout the metal and can only be seen when the surface has been perfectly polished and mag- nified at least one thousand times. These are invariably found in steels that produce blisters when pickling, and this leads to the conclusion that the blisters were formed by the reduction of oxide by the nascent hydro- gen evolved during the pickling process. High-carbon steel rods that con- tain the same impurity occasionally fracture in the pickling bath and doubtless the same pressure that blows a blister in mild steel will cause a rupture in hard steel. Owing to the gaseous nature of oxygen, and the fact that the drillings must be very fine, it is difficult to analyze steel for the oxygen content. A series of tests, however, was carried out by E. F. Law, of London, by cutting a piece from each of eleven bars of acid and basic Bessemer steel that contained from 0.10 to 0.18% of carbon, and only a trace of silicon. Each piece was then rolled into 24 sheets which were pickled and annealed by the usual process. An adjacent piece of the bar was analyzed, exam- ined with a microscope, and the oxygen determined. The result of these tests was as follows: (See table on page 85.) An examination of the table will show that as the oxygen content increased the number of blistered sheets increased, while the percentage of sulphur seemed to have no effect on the blistering; the set containing 19 blistered sheets only showing 0.071% of sulphur, while the set of sheets that did not blister at all contained 0.076, 0.069, and 0.061% of sulphur, respectively. By way of comparison a piece of basic Bessemer steel was analyzed just before the ferro-manganese was added, and this showed 0.062% of oxygen. The results shown here seem to forcibly confirm the oxide theory. It might appear at first sight that the quantities present are extremely small, but in making comparisons we should not consider alone the amount of the elements present, but also the combinations of these elements that INGREDIENTS OF AND MATERIALS USED IN STEEL 85 TABLE SHOWING EFFECT OF OXYGEN ON BLISTERING Kind of Analysis Microscopical Sheets in 24 that Percentages Steel s P Mn Appearance Blistered Oxygen Acid .061 .049 .340 Very good .021 Basic . . . 069 .034 .385 Good .021 Acid 076 070 .350 Good o .022 Basic .101 .126 .475 Fair 4 .025 Basic . . .... .080 .066 .430 Moderate 6 .026 Acid 106 .188 .320 Bad 7 .026 Basic 079 098 .440 Bad 7 027 Basic .045 .075 .473 Bad 8 .034 Acid 061 .081 350 Bad 9 032 Basic .080 .068 .450 Bad 12 .030 Basic .071 .090 480 Very bad 19 046 influence the quality of the steel. Thus, we speak of 0.05% of sulphur, when in reality it is 0.13% of manganese sulphide that affects the quality of the steel. Oxygen has only half the atomic weight of sulphur, and is capable of forming larger quantities of compounds, therefore it exerts a greater influence. Thus, where 0.05% of sulphur corresponds to 0.13% of manganese sulphide, 0.05% of oxygen corresponds to 0.22% of ferrous oxide. Another fact brought out in these tests is that the amount of oxide visible under the microscope was much less than would be expected from the amount actually found by chemical analysis, and this might be ac- counted for on the theory that a considerable quantity of oxide was in solution in the steel surrounding the black oxide spots. The oxide show- ing on the surface of a polished piece was also reduced by the aid of hydro- gen and an electric current, and the pits thus formed occupied a much larger area than the spots of oxide seen by the microscope. Steels containing oxides also apparently rust much quicker than those free from them, and with two pieces placed side by side the oxide steel will show rusting long before the other, while in dilute acid solutions steels containing oxides corrode more easily and much faster than those free from oxides. The same is true regarding the other impurities in steel and this has led to the production of a metal called " Ingot iron," in which the total impurities, except carbon, have been reduced to from 0.05 to 0.08% and the carbon content to 0.02%. A typical analysis showed carbon 0.02%; manganese, 0.01%; sulphur, 0.02%; oxygen, 0.03%, and phosphorus and silicon a trace. In the making of this metal the theory that ferro-manganese was needed to produce a workable metal in the hot condition was doubted, and the 86 COMPOSITION AND HEAT-TREATMENT OF STEEL usual ferro-manganese decarburizer was omitted. Open-hearth furnaces are worked entirely on cold pig iron low in silicon and sulphur, and with the phosphorus limited to the content for Bessemer working. An active basic slag is maintained that is composed of limestone and fluorspur, with a comparatively large amount of the latter flux to prevent the phosphorus from returning to the metal at the high temperature of 3000 to 3100 F., that is maintained toward the end of the process. A fairly large propor- tion of scrap is charged in the form of open-hearth mill scrap and low- carbon steel turnings, the larger part being of the latter. When sufficient mill scale can be obtained it is substituted for ore in the charge. i The removal of oxygen, probably in the metal in the form of oxides, is most important, and instead of manganese, ferro-silicon, or an equiv- alent material, is added to the bath to remove the oxides, while the other gases are removed by adding below 0.10% of granular aluminum in the ladle. The time consumed for each charge is about 10 hours, and the boiling is carried to a high temperature to thoroughly oxidize the impuri- ties. This brings the temperature very high in the final stages, owing to the higher melting point of the purer materials. If there is too much oxygen in the steel it is liable to cause it to crack on the edges when roll- ing, owing to its creating a red-shortness. COPPER Copper is a widely distributed element of the earth's crust, and occurs in large quantities; sometimes in the uncombined condition, such as the native copper of the Lake Superior regions. It is very malleable and tenacious. In most of the copper ores used, sulphur and iron occurs, and in some of the iron ores used for making steel, copper occurs. A few con- tain as high as 1% of copper, and some of the Bessemer and open-hearth steels contain from 0.30 to 0.50% of this metal. That copper alloys perfectly with all steels and does not segregate until above 4% has been added is a well-established fact. Copper can be alloyed in all proportions, with iron containing 0.15% of carbon, and with 0.09% of sulphur added to this no segregation will occur until 7.70% of copper has been added. With the sulphur low and the carbon at 0.20% no pronounced segregation appears until a copper content of 40% is reached, while with 0.40% of carbon it occurs with a copper content of about 30%.; with carbon 0.60%, at 20% copper; with carbon 0.80%, at 12% copper; and with the carbon at 1%, copper segregation is liable to occur when the copper is 8%. This, however, is only a general rule, and it may be varied greatly by the various other ingredients and methods of making the steel. As the best results seem to be obtained when copper is kept below 5%, segregation will not be much of a factor in copper steel. INGREDIENTS OF AND MATERIALS USED IN STEEL 87 Hard and soft steels with a percentage of copper as one of the ingre- dients have been used for many purposes with the usual number of fail- ures, but these failures have always been traced to other ingredients and none to the copper contents. Crank-shafts for the United States* battleships and gun tubes for 6-inch guns, have been made out of steel containing 0.57% copper, and they stood successfully all of the tests required by the Government. Commercially steels containing over 4% of copper cannot be rolled and forged unless the percentage of carbon is very low, owing to its hard- ening effect and the consequent brittleness it gives to the metal. With percentages up to 4, the copper all goes into solution in the iron, but above that, saturation begins to occur. The point at which saturation begins appears to be between 4 and 8%; it being lowered as the carbon content is increased. When the copper content is increased to above 8% free copper occurs; in a fibrous form in the soft or semi-soft steels, and in nodules in the higher carbon steel. When there is enough sulphur in the steel, it will form with the copper a copper sulphide, according to the formula (Cu 2 S), but if there is an excess of copper it will combine with the iron. Steels containing copper and copper sulphide have an irregular structure, as regards the size and join- ing together of the ferrite crystals, as these imbricate with one another with curved junctions. This gives the metal a higher strength than that of steel without copper. Copper and copper-sulphide principally distribute themselves between the crystals of ferrite, which they envelop. They also cause the quantity of pearlite to increase and the grains of this to assume a finer structure and permeate the metal more and more with each increase in copper. In fact, the structure so closely approaches the martensitic form that it has been mistaken for this in some instances, and in a 7% copper steel threads of cementite and of pearlite appeared. In this way they inten- sify the iron carbide and give to the metal a greater hardness as well as enable it to be hardened more easily when heating and quenching. Copper also lowers the recalescent point from 100 to 150 P F. below that of ordinary steels, but it never brings this below 800 F. In this it about equals high-carbon steel. The 1 to 5% copper steels that are liable to become commercially successful should be quenched in water from about 1325 F. or in oil. It is possible to find traces of copper sulphide in metal that contains only 7% of iron and 0.025% of sulphur. As a small amount of iron in solution in copper makes copper harder, this might suggest the idea of strengthening copper or copper alloys with iron. Copper increases the hardness of steel, as the copper content increases. 88 COMPOSITION AND HEAT-TREATMENT OF STEEL When the carbon content is low it has a greater -effect than when it is high; in some cases almost doubling the Brinnell hardness, and it reached its maximum increase in one series of tests at from 10 to 15% copper. It does not give any color to steel until 8% has been passed. With the carbon content high, copper steel is difficult to work mechan- ically, but it can be easily cast into the shapes desired. If, however, the carbon is kept below 0.50%, steels containing as high as 4% of copper can be easily and successfully rolled and forged, and the heat treatment made a less delicate operation. Such steels seem to have a future as they have a greater tensile strength and elastic limit than the same steel without copper; a better elongation and contraction; more resiliency; a greater resistance to shock and torsional strains; a greater hardness without loss of ductility and a finer grain. The copper steels closely resemble nickel or chromium steel, and follow the same laws as to their increases of strength for each increase of percentage, but they are said to possess a higher elastic limit and maximum strength than nickel steel, as well as greater dynamic strengths. Copper has a more active influence on steel than nickel or manganese, and nearly approximating chromium, molybdenum, and vanadium, and it is a cheaper alloying material than these. Copper steels as rolled show greater tensile strength with each increase of copper, and this is more manifest with the lower carbon per- centages, but it is not dependable in this state. Annealing corrects this to a large extent, but does not leave the metal much if any stronger than the ordinary steel. Hardening and tempering after this, how- ever, more than doubles the tensile strength and elastic limit, and brings the latter up close to the former with a good percentage of con- traction. This would seem to indicate that if copper steels were well made they would be able to withstand shock, torsional, alternating, or vibrational strains, as well as the high-grade steels of the present day, and, owing to the comparative cheapness of copper, they could be pro- duced cheaper. Some corrosion tests were carried on that showed that corrosion was lower by something like 100% in copper steel than in steels that contained no copper. The electrical resistance is also increased in steels containing copper, and reaches its maximum in a 0.15% carbon steel at 2% of copper; in a 0.7% carbon steel at 0.5% copper, and in a 1.7% carbon steel at 0.35% copper. ARSENIC AND THE ANALOGOUS ELEMENTS ANTIMONY AND BISMUTH Phosphorus, arsenic, antimony, and bismuth all belong to the same chemical group, and in general form compounds of the same character INGREDIENTS OF AND MATERIALS USED IN STEEL 89 and of similar composition. Like nitrogen they unite with metals to form binary compounds, called phosphides, arsenides, and antimonides. They all form two oxides, which contain 2 atoms of the above-named elements to 3 atoms and 5 atoms of oxygen. Of these elements phos- phorus occurs most abundantly in nature: arsenic and antimony next, arid bismuth last. The last three occur sometimes in the uricombined state, but phosphorus always occurs in combination with other elements. Many steels contain an appreciable percentage of arsenic, as it com- bines with iron in forms that are similar to the sulphide which it fre- quently accompanies. The arsenides, which are its compounds with metals, occur very widely distributed, and often accompany the sulphides to which they are similar. The most common compound of this kind has the composition FeAsS, and may, therefore, be regarded as iron pyrites (FeS 2 ), in which one atom of arsenic has been substituted for one atom of sulphur. Simple compounds of pyrite and arsenic occur that are anal- ogous to the sulphide FeS 2 , and combinations of sulphur and arsenic form into sulphides. When steel contains an appreciable percentage of arsenic it will give off an odor similar to garlic when heated to a red heat, and this odor may become very intense at a welding or forging heat. As an element it is not poisonous, but when oxidized it may become extremely so and it is easily oxidized. If the arsenic in commercial steel does not exceed 0.20% it does not have any material effect upon the mechanical properties, as the elongation and reduction of area are not changed and the tenacity is but slightly increased. This leaves the bending properties unchanged at ordinary temperature. Above 0.20% the strength of steel is increased and the toughness decreased with each increase in the percentage of arsenic until 4% is reached, when the elongation and reduction of area become nil and the steel becomes very brittle. Even with 4%, however, it does not affect the hot working of the metal, and it can be alloyed with iron in proportions as high as 56% under certain conditions of mixing. These conditions, however, are difficult to fulfil. By ordinary methods attempts have been made to produce alloys in various proportions up to 10% of arsenic, but when analyzed the sample showed that the maximum of arsenic taken up and retained by the iron was about 4%, this appearing to be about the largest amount that could be commercially added to steel. While steels with the higher percentage of arsenic are brittle, no special difficulty is met with in machining them with any percentage of arsenic. Owing to the fact that arsenic, when present in acid pickling solutions, causes a marked reduction in the rate of attack by the acid, it was thought that if the arsenic was added to the iron it might resist the attacks of 90 COMPOSITION AND HEAT-TREATMENT OF STEEL corrosion and become more durable. Numerous tests that were made, however, show no appreciable difference in the non-corrosive qualities of iron and steel that contained arsenic and those of the ordinary brand. Any benefits derived from alloys of arsenic with iron or steel will probably be in connection with their magnetic properties, as some very interesting results have been obtained along this line. It alloys with iron practically in proportions of the solid mixtures, up to an arsenic content of 4%. With each increase of arsenic in steel up to 5%, the mag- netic qualities of iron are made better and the arsenic alloys are on an equality with the best electrolytic material known in respect to mag- netic permeability. When the metal is heated to 1250 F. and slowly cooled, so as to allow the grain to become normal and the forging or roll- ing strains to be removed, the metal shows a decided improvement. A second heating to 1800 F., with slow cooling, improves the quality in the lower ranges of the magnetic forces, but there is a falling off in the upper ranges of the curve. Quenching from 1650 F. shows no harden- ing and but slight changes in the magnetization curves. Arsenic added to iron imparts to the alloy magnetic qualities excelling those of the purest iron, and at least equaling those of the best material from which data is obtainable. ANTIMONY may be added to iron in quite large percentages, but above a content of 1% the metal is not forgeable, and only then with difficulty. It renders the metal brittle so that it is practically worthless, and it is of a lower grade magnetically than the ordinary electrolytic iron. Thus while antimony is in the same chemical group as arsenic, it makes iron products that are difficult to work and have no apparent value as a mag- netic material. Antimony is useful in the non-ferrous alloys for the hardening effect it gives, and that it expands when solidifying makes it valuable for such uses as type casting. These same properties make it detrimental to iron and steel products, and luckily it does not appear in the crude materials used for making these. BISMUTH, like antimony, does not occur in combination with iron or in the products used for producing the iron ore when refining it into steel, conseqently it does not have to be removed as an impurity. To a greater degree than antimony it has the property of expansion when passing from the liquid to the solid state, and therefore it is useful in non-ferrous alloys. When 2% of bismuth, the most diamagnetic element known, was added to iron, it improved the already high magnetic quality of the pure iron. The density values reached exceed those obtained from any of several hundred other different alloys that have been tested. How much bismuth remained in the metal after adding the 2%, however, was not INGREDIENTS OF AND MATERIALS USED IN STEEL 91 known. With bismuth alloys there is but little increase in electrical resistance. Arsenic and antimony, however, give a decided increase in resistance to iron, and in some cases this was from 62 to 67%. BORON In nature boron chiefly occurs in the form of boric acid, or as salts of this acid, such as borax, a sodium salt, or two calcium salts. It belongs to the same chemical family as aluminum, and is very similar to it in the composition of its compounds, but its oxide is acidic, while the oxide of aluminum is usually basic. In some respects it resembles the members of the family to which nitrogen and phosphorus belong. It has a strong affinity for nitrogen, especially at the higher temperatures, and also com- bines readily with sulphur and chlorine. Some boron crystals con- tain carbon and aluminum, which seem to be in combination with the boron. Ferro-boron can be prepared from borate of lime, in the electric fur- nace, without any special difficulty, and the above data would suggest that boron might have some qualities that would be beneficial to steel, but very little in the way of investigation has so far been done. What little has been done would indicate that boron acts like carbon in many respects, especially in adding hardness to the metal. In some recent tests which were made on steel containing 0.20% of carbon and 0.20, 0.50, 0.80, 1, and 2% of boron, the Brinell hardness of the samples tested and quenched at 1460 F. was three times that of the annealed pieces, and equal to that of high-carbon steel similarly treated. Notwithstanding this the hardened samples could be easily filed, sawed, or machined, while 0.87% carbon steel, similarly treated, could not be scratched except with an emery wheel. This is adding evidence to the statement that has been made several times, but disputed by some, namely : that hardness is not the same thing as the ease or difficulty with which steel can be machined with cutting tools. The tests also show that boron confers upon steel the property of temper- ing; but a tempering that is very different from that conferred upon the metal by carbon, in that it increases the tensile strength and elastic limit, without materially increasing the toughness or hardness to machine. On the other hand, the ability to withstand shock tests was doubled by quenching, and the elastic limit was brought up close to the tensile strength. In heating boron steels they show a definite emission of heat at 2100 F., which resembles the recalescent point in high-carbon steel. Slightly marked critical points appear at 1900, 1525, 1350, and 1225 F. The three latter are about the temperature of the points Ar3, Ar2, and Arl of mild steel. The point at 1240 F. is definitely shown in carbon steel, 92 COMPOSITION AND HEAT-TREATMENT OF STEEL but when boron is added and the steel heated, this point almost entirely disappears, and is replaced by the point at 2100 F. Boron may be said to give steel a hardness that increases its strength, up to a content of 2% of boron, providing the carbon is kept below 0.2%, but beyond a content of 2% boron or 0.2% carbon, the metal becomes so brittle that it is weakened and easily powdered under a hammer. Other elements might be found, with further investigations, that would over- come this brittleness and make boron more useful for special alloys of steel. Microscopical examinations show intense black spots in boron steels that are polished and etched, first with picric acid and then with picrate of sodium. These increase in quantity with each increase in the -per- centage of boron. These spots may be a combination of boron-iron; a solid solution of boron-iron containing a very low percentage of boron; a borocarbide of iron, or a boride of carbon. In specimens thus treated the ferrite appears white, the pearlite grayish, and the special constituent very black. On annealing, the volume of pearlite increases and the special con- stituent disappears by forming a eutectic with the ferrite that at times is strongly marked. By annealing in the presence of oxide of iron, so as to decarburize the metal, the pearlite is first caused to disappear and then the special constituent. In carbonizing the special constituent is not increased by case-harden- ing, although at the edges a layer of pearlite is found and this is thinner if the metal does not contain boron. This would indicate that the pene- tration of carbon is delayed by boron, and that the amount of the special constituent depends upon the percentage of boron, and is independent of the carbon content. In the quenched steels, the special constituent was hardly discernible when the percentage of boron was below 0.50, but large quantities appeared in the steels with the higher percentages of boron. This was not altered even when the quenching was carried to 2200 F. The percentage of carbon increases the solubility of the special constituent, and the higher the percentage of boron the less easily does it dissolve. The above data probably indicates that the black spots were a boro- carbide of iron, and its percentage of carbon very low; otherwise a phenom- ena would occur similar to that brought out in the investigations of the vanadium steels, i.e., as the boron increased the pearlite would diminish; but in these steels the special constituent continues to increase. Boron steels are very weak and brittle in the normal state, and, if heated to a very high temperature, crumble when forged or rolled. But if heated to a dull red they can easily be forged, rolled, or otherwise mechan- ically worked, as they act much like soft steel. This will make them use- INGREDIENTS OF AND MATERIALS USED IN STEEL 93 less in the raw state, but after quenching they possess a high tensile strength, a very high elastic limit, and are not any more brittle than the special steels that are in actual use at present. Borax is a sodium salt from which amorphous boron, in almost pure form, can be obtained by heating with magnesium powder. It has been used by many misinformed people as part of a mixture for carbonizing steel, or in a special compound for hardening it, but they have never given any good reason for its use or shown any results that were obtained thereby. It, like boron, retards the penetration of carbon, but when used in a quenching bath may aid in producing a greater hardness, or prevent- ing the metal from cracking or checking. Common table salt (NaCl), however, gives much better results, and is easier obtained and cheaper. Therefore borax is not useful here ; its chief value is as a flux in welding. TANTALUM Tantalum is one of the rare elements. It is never found free in nature, but occurs in combination in the minerals columbite and tantalite, accom- panied by niobium. In chemistry it is grouped with vanadium, niobium, and didymium, all of which are rare. Its rareness, and consequent cost, has prohibited it from being experimented with to any extent, but one series of tests that was conducted appeared to prove that it had a harden- ing effect upon steel, similar to that exerted by tungsten and molybdenum, and to a certain extent gave promise of being beneficial for high-speed steel tools. In all of the eight tests made, the tantalum which varied from 0.42 to 1.69% increased the tensile strength, elastic limit, elongation, and reduc- tion of area over that of the same steel without tantalum, but when nickel or chromium was added in place of the tantalum, the same strengths were obtained and in one case 1.10% of chromium gave about 10% greater strength than 0.43% of tantalum. The greatest increases in strength were obtained with the smallest percentages of tantalum. Under the microscope a dark constituent appeared that was greater in quantity as the percentage of tantalum increased, and this occurred in a finely granular matrix that in the hardened specimens seemed to be martensitic and more or less homogeneous. From the results obtained and its similarity to vanadium the sugges- tion occurs that it acts on steel as a scavenger similar to this, and the best results would be obtained in the quaternary steels, but no evidence has been submitted to prove that it is any better, or even as good as the alloying materials already in use, and which are much cheaper. It is also very difficult to separate it from niobium, with which it is always combined, and this element is liable to cause erratic results in steel. 94 COMPOSITION AND HEAT-TREATMENT OF STEEL PLATINUM Platinum occurs in nature associated with five other elements, more rare than itself. They are divided into two chemical sub-groups commonly called the platinum metals. These nearly always occur in an alloy in which the platinum is from 50 to 80%, while the other five compose the balance. It forms two oxides and two sulphides. It is very ductile and is a grayish-white metal that looks like steel. It can be welded at a white heat. An alloy of platinum and silicon can be formed by bringing it in contact with red-hot charcoal and silicon dioxide. Nitric, hydro- chloric, or sulphuric acid will not dissolve it. Platinum, when finely divided, has an extraordinary power of condensing gases upon its surface; for instance, it absorbs 200 times its own volume of oxygen, also other gases similarly. The oxygen is then in the active condition, and oxidiz- able materials are easily oxidized when brought into contact with it. Thus when sulphur dioxide and oxygen flow together over spongy plati- num, or even the compact metal, they form sulphur trioxide by a unity of the two gases, or when hydrogen flows against the spongy platinum it takes fire. Iridium belongs to the same chemical group and, when this is alloyed with platinum in the proportions of 1 to 9 respectively, it reduces the malleability of platinum, which can be easily drawn into very fine wire; makes the alloy harder; more difficult to fuse; as elastic as steel; unchange- able in the air, and capable of taking a high polish. While platinum is but little cheaper than gold, the above properties have led to its being investigated as an alloying material for iron, but as yet the experiments have been very few, and limited in their scope. Platinum has no transformation points, and it consequently reduces those of iron when mixed with it. Up to 10% of platinum, two transforma- tion or recalescent points occur, while with the platinum from 10 to 40% but one point is produced. The melting-point diagram shows consider- able analogy to that of the nickel-iron alloys, but this is stronger when the alloys are rich in iron than when they are rich in nickel or platinum. The hardness of the platinum alloys decreases from to 5% of platinum, and then gradually increases from there to a platinum content of 40%, after which it remains fairly constant until 90% of platinum is reached, after which it declines again. At 50% of platinum the greatest brittleness occurs. From to 90% of platinum all the alloys are magnetic, and this dimin- ishes in the same ratio as the iron in percentages of from 80 to 20 of that metal. Alloys with the platinum from 10 to 50% lose their magnetic power when heated to from 1475 to 1200 F., and it returns at a much lower temperature when cooling. Alloys with the platinum from 60 to 90% regain their magnetic power at a temperature even lower than this. INGREDIENTS OF AND MATERIALS USED IN STEEL 95 NICKEL The chemical sub-group in which nickel belongs is composed of iron, cobalt, and nickel, and in many respects they are very similar. It occurs native in meteorites, and the iron meteorites always contain nickel. The principal minerals that contain it are nickeliferous pyrites and garnier- ite. Large deposits of minerals containing both nickel and copper have been found. The metals are reduced together and put on the market under the name of monel metal. Nickel, however, is separated in the pure form for many uses, and one of the most important of these is as an alloying material in the man- ufacture of special steels. It is a white metal with a slight yellow cast, and is very hard and capable of being highly polished. It is very brittle in its ordinary condition, but when deoxidized by magnesium becomes very malleable. Nickel reduces the size of the crystalline structure and increases the toughness of steel. It brings the elastic limit closer to the tensile strength, and microscopic cracks, that are liable to develop into larger cracks and produce rupture, do not appear as quickly in steels containing nickel as those without it. In certain proportions it also makes steel more resilient or springy, increases the hardness, raises the tensile strength, and segregates only slightly. Nickel was first added to steel for the purpose of overcoming the property of "sudden rupture," which is inherent in all carbon steels. This it does to a large extent, making steel better able to withstand severe shock and torsional stresses, as well as compressive stresses. This is not due to hardening, as soft steel cannot be made hard by the addition of nickel, except in large quantities, and it is considered that 17.55% of nickel is the equivalent of only 1% of carbon. The properties of nickel steel depend as much upon the carbon content as on the nickel. The fact that a 2 or 3.5% nickel steel is used means nothing unless the carbon content is right for the use to which the steel is to be put. To illustrate, a steel containing 2% nickel and 0.12% car- bon has a good tensile strength with a great elongation, and is useful for some purposes, while a steel that is equally useful for another purpose may contain 2% nickel and 0.9% carbon, and this would give it a high tensile strength with very little elongation. With a high carbon content nickel steel is difficult to harden, especially locally, as fissures and cracks tend to develop in quenching. It also has more tendency to warp in quenching than other steels and may be decarbonized by heating. These tendencies may be overcome to a great extent if the metal is thoroughly 96 COMPOSITION AND HEAT-TREATMENT OF STEEL annealed before it is machined to size, in order to relieve all of the internal strains. Then, when quenched, the piece should be immersed in the bath so that the liquid can cover the greatest possible surface at the instant it strikes the bath, and it should be agitated while cooling. FIG. 47. Cutting test bars. Nickel also gives steel a tendency to show laminations, and makes it weaker at right angles to, than in line with, the direction in which it is rolled. The higher the nickel content the greater will be the contrast between the strength in these two directions. This is best shown by 40 50 Percentage of Nickel FIG. 48. Effect of nickel in different percentages. tests which were made on test bars 1 and 2, cut from a piece of 3.5% nickel steel as indicated in Fig. 47. Test bar 1 showed an elongation of 12% and a reduction of area of 17%. Test bar 2 gave an elongation of 25% and a reduction of area of 65%. The good qualities which nickel INGREDIENTS OF AND MATERIALS USED IN STEEL 97 gives to steel offset these bad qualities to such an extent that it makes a much better steel for gears, crank-shafts and pieces which have similar work to perform than the ordinary carbon steel. Nickel greatly reduces the tendency of steel to be damaged by over- heating, and also increases the effect of hardening in raising the strength of the metal. One series of tests which were made showed a tensile strength of 88,000 pounds per square inch, an elastic limit of 60,000 pounds per square inch, an elongation of 28% and a reduction of area of 58% when in the annealed state. These figures were changed by hardening to a tensile strength of 225,000 pounds, an elastic limit of 224,500 pounds, an elongation of 8%, and a reduction of area of 19%. A good quality of carbon steel might give the same results in the annealed state, but they could not be increased to nearly the same extent by means of the ordinary hardening. Nickel has one peculiarity in its influence on steel which is best shown by Fig. 48. It increases tensile strength and elastic limit, but steels containing 8 to 15% of nickel are so brittle that they can be powdered under a hand-hammer; at 15% of nickel the toughness begins to be restored; from 20 to 25% the elongation rapidly increases, and from there on to 50% a gradual increase is shown. Steel with percentages of nickel from 30 to 35 gives good results for valves on internal-combustion engines, as the nickel makes the steel wear better and it is not as good a conductor of "heat as other metals. Nickel steel can be purchased in the open market in nearly all percentages of nickel from 1 up to 35%, and with varying percentages of carbon. In Fig. 49 is shown the actual results that were obtained from a series of twenty tests, in which the nickel varied from to 20%, and the other ingredients remained fairly constant. Ten of the tests were with forged steel and ten with cast steel. They give a good idea of the strengths that can be expected in nickel steels, although, as has been said many times, nickel steel in the annealed or natural state is but little better than carbon steel, but if properly heat-treated it will greatly exceed carbon steel for static and dynamic strengths, wearing qualities, etc. COBALT The principal minerals containing cobalt are smaltite and cobaltite, and in each of these iron and nickel take the place of a part of the cobalt. It, like nickel, forms compounds that are analogous to ferrous compounds, and also a few that are analogous to ferric compounds. In the latter case, its power is greater than that of nickel. Cobalt is harder than iron, melts at a slightly lower temperature, and has a silver-white color with a tinge of red. 98 COMPOSITION AND HEAT-TREATMENT OF STEEL In the matter of cobalt-iron alloys, investigations up to a cobalt per- centage of 60 have been made. The mechanical properties were but little modified in these, but the breaking strength and the elastic limit increase slowly, while the elongation and the reduction in cross-section are !uaraoH auiWAI uj sptrao j tpui o t2 m qoui axetiDs .tad suoi tiOTjo'BJjuoo pins uorji?3u