Search Documents

By Henry Seltz, Cyril Wells, Hugh J. MacDonald

Several investigators have measured the heat capacity of cementite, using different methods of attack, but the agreement between the values obtained cannot be considered good. Naeserl has made measurements of "average heat capacities" of iron carbide between 1026" and 85" K., using a water calorimeter. His computed true specific heat at 78.1" K. is 11.22 cal. per gram-formula mass, which, if correct, is much too high for any reasonably accurate extrapolation to 0" K. The cementitc used by Naeser was obtained from a steel containing 0.9 per cent carbon, by electrolytically dissolving out the iron in a neutral bath of ferrous chloride. The cementite remained as a coarse, crystalline, gray powder. Andes2 has measured the specific heat of iron carbide, from 102" to 323" K., as it exists in steel. The specific heats of several slugs of steel of known iron carbide content were determined, and the specific heat of the iron carbide (Fe3C) calculated by plotting specific heat versus percentage of carbon, and extrapolating to pure Fe3C (6.7 per cent C). The heat-capacity values for temperatures below 100" K. were obtained by comparing the temperature-specific heat curve for Fe3C to that of Fe. Since the two were very similar in the range examined experimentally and approach more closely at low temperatures, it was assumed that near the absolute zero the form of the curve for Fe3C would be similar to that of Fe. Andes calculates for the entropy of Fe3 C,S298 = 23.55 e.u. The heat of formation of Fe3C has been studied by several workers. Kelley3 has reviewed these measurements and calculations, and selects those of Roth* and Naeser6 as worthy of most consideration. Roth has given AH = 5400 cal. + 30 per cent as a result of his own and other work. Naeser has determined the heat of formation from Fe and different forms of carbon, and for graphite (0) he finds AH = 5600 cal. Kelley accepts the rounded figure AH = 5500 as the heat of formation at 298.1" K., but this value is quite uncertain.

Jan 1, 1940

By Cyril Wells, Robert F. Mehl

The diffusion of carbon in gamma iron plays an essential role in many metallurgical processes. In carburizing, in graphitizing, in homogenizing, in the formation of pearlite from austenite, and in other processes, migration of carbon in gamma iron occurs.1'2 A full understanding of the rate at which these processes occur obviously requires a knowledge of the rate of diffusion of carbon. In carburizing, for example, the depth of carbon pentration is determined by the carbon concentration in solid solution at the surface, and by the rate of diffusion inward of the carbon. Until reliable data on the rates of diffusion of carbon are available it is unlikely that a full quantitative description of this process will be obtained. In the formation of pearlite from austenite, the segregation of carbon and alpha iron from the austenite obviously requires the diffusion of carbon; the rate at which pearlite nodules grow is evidently controlled at least in part by the rate of diffusion of carbon; it is unlikely that we shall understand the factors that determine the rate of growth unless data on rates of diffusion of carbon, and the variation of these with composition, grain size, and temperature are available. Accordingly it was decided: (1) to determine, more accurately than hitherto, the diffusion coefficients of carbon in austenite over a wide temperature and a wide concentration range, (2) to study the influence of grain size and impurities on this diffusion coefficient, and (3) to study the effect of manganese and nickel on the rate of diffusion of carbon in the austenite of manganese and nickel steels, respectively; the influence of carbon on the rate of diffusion of manganese and of nickel will be made the subject of later papers. The rate of diffusion of carbon in gamma iron has been studied by Runge3 by Tammann and Schonert4, by Bramley and co-authors,5-14 and

Jan 1, 1940

By Alden B. Greninger, Alexander R. Troiano

Metallurgists have long believed that martensite in steel forms as plates along the octahedral {111} planes of austenite. Much has been written about mechanisms whereby units of the austenite lattice are pictured as transforming to martensite by shearing movements initiated along the octahedral planes of austenite, supposedly resulting in plates of martensite lying parallel to these planes of shear. A critical inspection of two or three photomicrographs of quenched hypereutectoid steel led us to the conclusion that the orientation habit of martensite, at least in hypereutectoid austenite, could not possibly be octahedral. Inasmuch as recent studies,30,31 of the martensite-type reaction in nonferrous alloy systems have established the irrational orientation habit as characteristic of this type of lattice transformation, the need for a thorough restudy of the crystallography of martensite formation in steel was evident. The present paper reports the results of such a study, extended to include both cementite and some of the products of constant-temperature decomposition of austenite. Also reported are a few of the results obtained from a microscopic and X-ray study of the austempering process in very high-carbon steel. I. The Martensite Transformation Previous Work Microstructure and Crystal Structure of Martensite.—Among the best series of published photomicrographs of quenched steels are those of Whitely,1 Robertson,2 Hanemann and Schrader,3 and Lucas.4 The crystal structure of martensite is body-centered tetragonal, with axial ratio varying from 1 near 0.0 per cent carbon to 1.08 for 1.75 per cent

Jan 1, 1940

By Norman P. Goss

It is generally agreed that cold-working mechanically refines the grains into smaller fragments and with continued working these are oriented with certain crystallographic directions bearing a relation to the direction in which the metal is worked. Some of the energy expended on the metal during plastic deformation appears in the form of heat, but a very small portion remains in the metal as stored energy. The entropy, specific heat capacity, temperature at which recrystallization will just begin, density, ductility, elasticity, tensile strength, and resistance to corrosion are affected. These properties have been studied extensively and some of the changes in the physical properties due to cold-working were attributed to lattice distortion.l-8 The broadening of the Ka lines on the X-ray diffraction diagrams of plastically deformed metals is believed to indicate lattice distortion, and consequently the existence of internal stresses. The evidence supporting this theory rests solely upon the observed experimental fact that the diffraction lines are broadened and the Ka doublets are completely diffused. Elam9 has reviewed the matter of lattice distortion in relationship to the hardening of metals by plastic deformation. The X-ray diagrams of plastically deformed metals differ in general from annealed metals. The hardening of a metal was believed due to lattice distortion.9 The resolution of the doublet is lost when a metal is cold-worked and the Laue spots are elongated. At one time this elongation was thought to indicate the elastic bending of the crystal planes. However, more careful work showed that the asterism was due to new orientations developed within the fragmented grains during plastic deformation. Some theories attributed the hardening to fragmentation of the grains into smaller units. Jeffries and Archer,ll after carefully considering the problem of work-hardening from many angles, came to the conclusion that cold-working mechanically refines the grains into smaller crystal fragments, some of which at least differ in orientation from the parent grain. The increase in

Jan 1, 1940

By J. T. Burwell

The crystal orientation in silicon iron that has been given a particular treatment described by Goss,l has been studied by Goss, by Bozorth2 and by Sixtus,3 but their results do not agree and are almost entirely qualitative. The following pages give a description of the orientation of the crystals in the material as determined by means of X-rays and expressed graphically in a pole figure. The treatment consists of hot-rolling the material into strip and then giving it two cycles of cold reduction followed by recrystallization. This produces remarkably good properties in the rolling direction, the product having a maximum permeability in this direction of more than 25,000 and a permeability at 16,000 Gauss of more than 2200. It is now generally agreed that Goss's original conclusion1 that the crystals in the final product are randomly oriented was incorrect, since Bozorth2 and Sixtus,3 using different methods, have shown that there is a decided preference in orientation. Those writers, however, expressed their results only in terms of one or more ideal orientations with deviations therefrom. X-ray Data A strip of the material containing 3.1 per cent Si, prepared by Goss,† 1.2 by 0.0125 in., was etched rapidly in warm dilute hydrochloric acid until it was 0.005 in. thick, in order to eliminate the effect of the two surface layers. The specimen was then mounted in an integrating camera similar to one described in the literature.4 The purpose of this camera is to translate the specimen during exposure, so that a statistically large number of grains contribute to the pattern photographed. The camera also allows the plane of the specimen to make various angles with the X-ray beam. An area 5 by 1.2 in. was scanned by a beam of unfiltered radiation from a molybdenum target tube operating at 37 kv., and the measurements were made principally on the peak of the white radiation. X-ray photographs were taken at intervals of 15" in a plane

Jan 1, 1940

By George Sachs, J. Spretnak

The process of recrystallization has not as yet been explained satisfactorily. Some definite conclusions could be drawn from recent investigations, such as, for example, that recrystallization is a process of nucleus formation and crystal growth1 and that the orientations of the nuclei correspond to some fragments of the original crystals, which have been considerably distorted during deformation.2 There are, however, a number of metals and alloys crystallizing in the face-centered cubic system, which form a particularly simple orientation if recrystallized as a sheet.3 Such a shect is almost a single crystal with the cubic axes parallel to the rolling, transverse, and normal directions of the sheet. This preferred cubic or (100) orientation of the recrystallized sheet has been considered to be very different from the rolling structure of the same metals,4 but there are two reasons why this last supposition remains open to controversial discussion. The first reason is academic: In practically all cases except the one mentioned above, close correlations and similarities between rolling and annealing orientations have been observed; for instance, in iron and steel sheet.6 Secondly, no deformed metal actually forms a simple orientation. In any thoroughly investigated case a number of different orientations and considerable scattering or deviations from the ideal orientations have been found.6 It is generally true that investigations on single crystals have shown that a definite orientation is approached in stretching or in compression of an originally undisturbed crystal.7 In the special cases, however, where the initial orientation was close to an important crystallographic direction that differed from the final "ideal" orientation, the resulting orientation changed very little by deformation, staying somewhere in the vicinity of the initial orientation.8 Such deviations

Jan 1, 1940

By John Wulff, R. Buehl, H. Hollomon

Although the main features of the transformations occurring in 18-8 have been published already, certain conclusions merit questioning and discussion. The questions may be summarized as follows : 1. If ferrite is formed by cold-working austenite, what happens to the carbon with which the ferrite and austenitc are both presumably supersaturated? 2. What is the nature of the alpha-ferrite formed at grain boundarices by chromium-carbide precipitation at these boundaries?

Jan 1, 1940

By H. H. Uhlig

In our study of the mechanism of and the prevention of corrosion in stainless steels, it was considered of fundamental importance to obtain knowledge of the surface structure of such alloys. If corrosion resistance is due to a self-healing oxide layer, which like a paint film protects the underlying metal, it would be obviously relevant were the effort made to make the film less permeable and more adherent; for example, the surface might be expected to respond to an anodizing process, which is found successful in increasing the film thickness and corrosion resistance of aluminum. However, if passivity and corrosion resistance of stainless steels are due to another cause, the attack of the problem would follow another path. One of the most effective techniques for studying passivity and corrosion tendency is the measurement of the electrochemical potential. Not only can one determine by potential measurements whether a metal is active or passive, but also one can determine the degree of passivity and rate of transformation of passivity with time. Chemical media that build up or destroy passivity when in contact with the steel can also be classified by potential measurements. Something can be learned, too, of the effect of alloy additions such as molybdenum to 18-8, known to decrease corrosion of the alloy in chloride solutions. A study of alloys in previous investigations has sometimes included data on the electrochemical potentials. Tammann and Sotterl measured the electrochenlical potentials of alloys of iron-molybdenum and iron-chromium and reperted the critical concentration of molybdenum or chromium at which passivity initiates. Strauss2 made similar measurements for the iron-chromium system. The electrochemical potential3 has also been used to indicate carbide precipitated 18-8. In 18-8, heat-treated to precipitate carbide, active metal exists at the region of the grain boundaries, contrasted with passive metal within the grains, the active metal accounting for a resultant active potential. Several

Jan 1, 1940

By H. H. Uhlig

Soon after general use of stainless steels began, it was observed in practice that certain combinations of factors tended to induce corrosion by pitting. For most applications this was a serious drawback, not SO much because the structural properties of the metal were impaired but because of the rapidity with which heavy sections of the alloys were penetrated by small elongated pits. These pits seemed to initiate readily in oxidizing metal salt solutions like ferric chloride and in chloride solutions like sea water. It was observed that in sea water pitting was more frequent in stagnant or polluted waters than in highly aerated, moving waters. The steels in these chloride solutions were observed to be sensitive to long contact with material such as sand grains or attached barnacles. At points of attachment or contact it was not uncommon to discover initiation of corrosion. Pits once begun at contact faces were observed to grow rapidly, usually elongating phenomenally in the direction of gravity until the metal was perforated. Hydrochloric acid and other nonoxidizing acids, on the other hand, were observed to corrode stainless steels generally; the weight losses often were higher, but the penetration less than for the pitting type of corrosion. The causes of pitting, a type of corrosion recognized as characteristic not only of stainless steels but of many additional metals and alloys, are important in corrosion studies. A possible explanation bears practical significance, for once it is known why pits begin where they do and continue to grow after initiation, intelligent direction can be given to diminishing corrosion. This can be done either by some alteration in the composition and treatment or' the alloy or by some alteration in the environment. It appears foremost in importance to discover the nature of pit foci, for if initiation of pits can be discouraged their growth will be no longer a factor to be considered. In a multi-component alloy containing many impurities normal to production of such an alloy, each of which may play a part in pitting, and whose corrosion resistance is sensitive to heat-

Jan 1, 1940

By Russell Franks, W. O. Binder

In recent years a considerable tonnage of the 18 per cent chromium and 8 per cent nickel steels has been used in lightweight high-strength structures that must be resistant to deterioration under atmospheric conditions. These steels are austenitic in character and their strength can best be increased by cold-working. Although a marked increase in tensile strength is obtained, claims have been made that the cold-worked steels do not possess fully satisfactory elastic propertics. Under certain conditions, however, this has caused no great concern, because, as explained by Ragsdale,1,2 he cold-rolled steels have been widely applied for construction of lightweight, high-strength trains using spot welding for joining purposes. Ffield3 agrees that the ease with which thin sections of the cold-rolled 18-8 steels can be joined by spot welding is greatly in their favor, but that they would be more satisfactory if higher elastic properties and a higher modulus of elasticity could be obtained. He states that, as far as the mechanical properties are concerned, the straight chromium stainless steel is superior to the cold-worked 18-8 steel, and in order to illustrate, he prcsented the data of Fig. 1, which also give a stress-strain curve for a low-carbon 13 per cent Cr steel. Ffield† proposed that columbium-bearing or titanium-bearing cold-rolled 18-8 steels be given a heat-treatment (preferably for 1 hr.) in the range 400" to 600°C. to precipitate carbides in order to obtain better mechanical properties. Discussions by Sutton,4,5 de Ganahl,6,7,8 Krivobok and Lincoln,' Knerr,l0 and Thaden11 have given further information on the properties of the cold-rolled 18-8 steels compared with other materials used in lightweight, high-strength structures. These investigators and others point out that the ultimate tensile strength of the cold-rolled 18-8 steel can be increased to 200,000 lb. per sq. in. or higher; Krivobok et al. indicate that

Jan 1, 1940

By Charles R. Austin, Carl H. Samans

In an earlier serics of studies on 18-8 stainless steel at 600º, 700º, and 800ºC.l one pretreatment was used throughout; viz., 15 min. in hydrogen at 950°C., followed by an air quench. Several manufacturers of this alloy since have expressed doubt as to the suitability of the temperature employed, recommending instead the more customary Strauss treatment as promising superior results. In this subsequent investigation on the effects of the temperature of pretreatment several interesting points have arisen, which are described in this paper. Previous work on our equipment had shown a marked difference between alloys of "similar chemical composition" that had been manufactured by different companies. Accordingly, one mill-bar length was selected from which to prepare all specimens. It could then be assumed that any marked differences in creep characteristics would be due to effects resulting from variations in the pretreatment of the alloy. The material was furnished by the Allegheny-Ludlum Steel Co. in the form of 5/16-in. dia. rod, finished by air cooling from a mill anneal at 1950°F. (1066°C.). It was analyzed only for the elements shown: carbon, 0.07 per cent; chromium, 18.73; nickel, 9.46; manganese, 0.40. Three pretreating temperatures were chosen, the specimens being held for 15 min. in hydrogen at either: (1) 1150°C., to conform to the recommended Strauss treatment; (2) 950°C., to compare with the results of previous tests; or (3) 75Q°C., to initiate precipitation of carbides in the material before testing. As the test cross sections were only 1/8-in. dia., air cooling followed the annealing treatment. These treatments permitted two direct comparisons: (1) between the 1150" and 950°C. treatments to indicate the effect of grain size when all carbides were probably in solution at the start of the test; and (2) between the 950" and 750°C. treatments to indicate the effect of carbide "nucleation" for similar grain sizes.

Jan 1, 1940

By H. T. Clark, H. K. Work

The instrument for the rapid determination of carbon in steel, described in this paper, was developed originally to facilitate the operation of the pilot-sized open-hearth furnace in the Jones & Laughlin Research and Eevelopment Laboratory. In this small furnace, with its shallow bath, the drop in carbon is accelerated, and a rapid, reliable test for carbon content is indispensable. This problem becomes rather acute when it is desired to finish heats at a carbon content higher than can be estimated with any degree of accuracy by the fracture test. Several independent methods for determining carbon were investigated, therefore, and it was found that a relatively simple magnetic device could be made to give reproducible results. A laboratory instrument was designed and set up for the use of the furnace operators. It was tested exhaustively in the laboratory and was used with considerable success for following the carbon content in the furnace bath. As soon as sufficient data had been compiled to indicate that the instrument might have a wider application, a more rugged model was constructed and placed in the mill to be tested on commercial heats. Finally, about a year ago, instruments were furnished to the plant open-hearth shops, and these have been in use since that time. As a matter of convenience, the instrument has been named the "Carbanalyzer," and will be referred to by this designation in the following pages. $ Eaklier Investigations While by far the greatest number of magnetic investigations in which the carbon content of ferromagnetic materials was considered have had to do with the evaluation of the magnetic properties of steels of known

Jan 1, 1940

By C. F. Quest, T. S. Washburn

The steel industry has a definite need in its tonnage mills for a formula relating expected tensile strength to the analysis of the steel. The McWilliam-Barnes1 and other less comprehensive formulas are used for this purpose, but these formulas were designed for an analysis range that does not include the medium manganese and silicon-bearing steels. This type of steel is being produced to meet the demand for tensile strength slightly higher than that of ordinary structural steel but not high enough to justify expensive alloy steels. The purpose of this paper is to present a somewhat more comprehensive formula based on data accumulated over a number of years by the Inland Steel Co., and to discuss the method followed in obtaining a formula to fit the data. The resultant formula admittedly is still weak in some respects, but for the expanded analysis range appears to be an improvement over the other formulas. It is hoped that with the collection of more data and possible collaboration with other mills studying this problem it will be possible to work out a more nearly perfect formula. Several publications have been issued on this subject. In some, the equation type of relation was evolved while others made a combination of tables and an equation. The latter include work by Campbel12,3 and Webster.4,5 Campbell classified the steels as acid and basic and applied a different formula to each. For basic steels, he used the equation: Teosile etrength = 41500 + 770C + 1000P + yMn [1 y was a variable depending on the carbon and manganese content of the steel and was listed in tables. Webster investigated acid Bessemer steel and developed the equation: Tensile strength = 38000 + 800C + xTP+ yMn [2]

Jan 1, 1940

By J. W. Halley

Copper has frequently been recommended as an alloying element for steel and the precipitation-hardening of steels containing from 1 to 2 per cent copper has been studied by a number of investigators. The straight copper steels, however, have not been used extensively and little advantage has been taken of the precipitation-hardening of complex steels containing appreciable amounts of copper. It is difficult to produce a satisfactory surface on the straight copper steels and undoubtedly this has discouraged their use. During recent years a number of steels containing enough copper to make them susceptible to precipitation-hardening have been put on the market. These steels belong to the group usually referred to as "low-alloy high-tensile " steels. They were designed for use in the " hot-rolled" condition; transportation equipment and large structures that cannot be readily heat-treated have been their chief application. Little emphasis has been placed on the possibility of strengthening those containing substantial amounts of copper by precipitation-hardening. The present investigation was undertaken to determine the precipitation-hardening characteristic of one of the "low-alloy steels." Previous Work The first complete study of the effect of time and temperature on the precipitation-hardening of plain copper steels was made by Nehll in 1930. At about the same time Bucholtz and Koster2 studied copper-chromium steels but did not determine the relation between time and temperature to produce maximum hardening. Later, Smith and Palmer3 determined the rate of hardening at various temperatures for two I per cent copper steels containing different amounts of carbon. Their tests indicated that the higher carbon content decreased the rate of hardening. They made tests on many steels containing other alloying elements in addition to copper, but did not determine the time-temperature relationship for maximum hardening for any of the complex steels. Norton,4 in studying the limit of solid solubility of copper in iron, determined the time for maximum hardening of a 2.06 per cent Cu iron

Jan 1, 1940

Jan 1, 1940

Page Foreword. By Edmund M. Wise..................... 3 A.I.M.E. Officers and Dircctors...................... 4 Institute of Metals Division Officers and Committees............ 7 Institute of Metals Division Annual Award Certificate............ 8 Institute of Metals Division Lectures and Lecturers............. 9 Photograph of E. H. Dix, Jr., Institute of Metals Division Lecturer...... 10 PAPERS Acceleration of the Rate of Corrosion by High Constant Stresses. By E. H. Dm, JR. (Institute of Metals Division Lecture, T. P. 1204).......... 11 Variation of Internal Friction with Grain Size. By Clarence Zener and R. H. Randall. (T.P. 1146, with discussion) .............41 Changes in Damping Capacity during Annealing of Alpha Brass. By John T. NoRton. (T.P. 1168, with discussion)................49 An Electron Diffraction Study of Anodic Films. By R. A. HarrinQton and H. R. Nelson. (T.P. 1158).....................62 Determination of Orientations by Etch Pits. By Charles S. Barrett .and L. H. Levenson. (T.P. 1171, with discussion).............76 Process of Precipitation from Solid Solution, I—A Crystallographic Mechanism for the Aluminum-copper Alloys. By Carl H. Samans. (T.P. 1186). . . 85 Age-hardening of Aluminum Alloys, IV — Discussion of the Theory. By WilTJam L. Fink and Dana W. Smith. (T.P. 1083, with discussion). . . 95 Structure of Aluminum after Compression. By Charles S. BaRREtt and L. H. evensonON. (T.P. 1104, with discussion)................112 Recrystallization Texture of Aluminum after Compression. By ChaRlEs S. BarRett. (T.P. 1141, with discussion)................128 Plastic Deformation and Recrystallization of Aluminum Single Crystals. By J. A. Collins and C. H. Mathewson. (T.P. 1145, with discussion) . . . 150 Recovery and Recrystailization in Long-time Annealing of 70-30 Brass. By S. E. Maddigan and A. I. Blank. . (T.P, 1166, with disonssion).....170 Correlation of Deformation and Recrystallization Textures of Rolled 70-30 Brass. By R. M. Brick. (T.P. 1144, with discussion).........193 Tensile Deformation of Critically Oriented Brass Crystals. By H. L. Burghoff. (T.P. 1149).............................214 Recrystallization of Lead. By Paul A. Beck. (T.P. 1101, with discussion). . 222 Simple Method of Thermal Analysis Permitting Quantitative Measurements of Specific and Latent Heats. By Cyril Stanley Smith. (T.P. 1100, with discussion).............................236 Metallographic Study of Internal Oxidation in the Alpha Solid Solutions of Copper. By Frederick N. Rhines. (T.P. 1162, with discussion). . . . 246

Jan 1, 1940

Jan 1, 1940

Jan 1, 1940

Jan 1, 1940

Jan 1, 1940