Search Documents

By G. Bhat, J. F. Libsch

lsoembrittlement curves depicting the influence of time and temperature in the range 800' to 1260°F (425' to 680°C) on the development of embrittlement in a commercial chromium alloy steel and a commercial molybdenum alloy steel are presented. Two distinct regions of embrittlement occur in the chromium alloy steel: I—at 800' to 1000°F (425' to 540°C) and 2—in the region just below the lower critical temperature. Embrittlement is most pronounced at 800' to 1000°F, decreasing very rapidly with increasing temperature above this region, only to increase again as the lower critical temperature is approached. The data suggest two distinct modes of embrittlement with possible superposition of the two modes at extended embrittling times in the temperature range 1100° to 1150°F (590' to 620°C). While the molybdenum alloy steel shows little susceptibility to embrittlement at 800' to 1000°F (425' to 540°C), considerable embrittlement may occur just below the lower critical temperature. THE subject of temper embrittlement in alloy steels has received considerable attention in the last few years. Points of view on the mechanism of embrittlement differ, however, resulting in part from the incompleteness of the data developed and in part from the speculation regarding the susceptibility of plain carbon steel to temper embrittlement. Libsch, Powers, and Bhat1 carried out short-time embrittling treatments on an AISI 1050 steel and demonstrated that hardened plain carbon steels are quite susceptible to embrittlement when tempered in the range from 850°F (455°C) to the lower critical temperature. The isoembrittlement diagram,' representing the embrittling characteristics of this steel, is reproduced in Fig. 1. It is evident from the shape of the curves shown that embrittlement in plain carbon steel increases progressively with both temperature and time in the embrittling range. A comparison of the isoembrittlement diagram for AISI 1050 steel with that presented by Jaffe and Buffum' for an SAE 3140 steel shows that up to 930°F (500°C) the isoembrittlement characteristics of the plain carbon steel are similar to those of SAE 3140 steel, although the embrittlement is much more severe in the latter steel. Above 930°F (500°C), the rate of embrittlement in the plain carbon steel increases continuously with increasing temperature; whereas, in the SAE 3140 steel, the embrittlement rapidly decreases. The influence of alloying elements upon embrittlement during tempering thus appears to cause a decrease in embrittlement above the region of maximum embrittlement, i.e., 850" to 1000°F. The question naturally arises as to what effect individual alloying elements have upon the embrittling characteristics of the plain carbon steel. Current knowledge on the influence of alloying elements on temper brittleness may be found in the review papers of Hollomon" and Woodfine. Hollo-mon," from the results of other investigators, has shown that, in general, the amount of embrittlement increases with increasing alloy content (except for molybdenum and possibly tungsten and columbium). Jaffe and Buffum," by a comparison of the embrittlement in a plain carbon steel with that of a SAE 3140 steel postulated that the presence of alloying elements in moderate amounts tends to retard the development of temper brittleness. It is difficult to determine what effect chromium has upon temper brittleness, since most of the information available has been based on the combined effect of other elements with chromium, particularly nickel and manganese. However, Wilten, and recently Jolivet and Vidal,' Vida1, and Woodfine have reported that chromium steels are temper brittle, that the embrittlement is reversible with a maximum rate of embrittlement at approximately 975°F (525"C)," and that the susceptibility increases with increasing amounts of chromium. Taber, Thorlin, and Wallacel" have found a large embrittling effect with increasing chromium content in a medium C-Mn-Ni steel. But Hultgren and Chang," from their experiments conducted on synthetically prepared ternary Fe-C-Cr alloys, could not conclude that these alloys are susceptible to temper embrittlement. However, on addition of manganese or phosphorus, these Fe-C-Cr alloys became susceptible, from which fact they concluded that the embrittlement developed in chromium-bearing Fe-C alloys is due chiefly to the presence of these elements. Considerable data are available to show that molybdenum decreases the susceptibility of steel to temper embrittlement. However, its effectiveness in preventing or decreasing embrittlement appears limited to its presence in small amounts. Vidal" has shown that a plain 2 pct Mo steel was susceptible. Hultgren and Chang" also have shown that molybdenum additions in excess of 2 pct to synthetically prepared Ni-Cr steels did not prevent embrittlement. Jolivet and Vidal' and Lea and Arnold found that molybdenum reduced temper brittleness. Lea and Arnold further stated that molybdenum decreased the rate of embrittlement rather than the total amount of embrittlement, whereas Preece and Carter" have shown that the presence of molybdenum greatly reduces the equilibrium extent of the change at a given temperature but does not appear to influence the rate of embrittlement. There appears to be very little information as to how molybdenum by itself affects the temper brittleness susceptibility of a plain carbon steel.

Jan 1, 1956

By S. C. Das. Gupta, B. S. Lement

isothermal formation of martensite occurs in the range of —65o to —197oC and is always preceded by some athermal transformation. By rapid cooling the isothermal, but not the athermal, component of transformation can be suppressed. Stabilization against isothermal transformation can be induced by cycling from below M. to room temperature. A LTHOUGH the transformation of austenite to XV martensite is generally believed to occur only on cooling, recent evidence indicates that this reaction can also occur at constant temperature. Isothermal formation of martensite was observed in tool steel by Fletcher, Averbach, and Cohen.'. They found that the hardening reaction that occurs during the quenching of a plain carbon or low carbon alloy tool steel does not stop immediately when room temperature is reached; instead austenite continues S. C. DAS GUPTA is Research Fellow, Department of Metallurgy, University of Notre Dame, Notre Dame, Ind. B. S. LEMENT, Associate Member AIME, is on the Research Staff, Department of Metallurgy, Massachusetts Institute of Technology, Cambridge, Mass. Discussion on this paper, TP 3123E, may be sent, 2 copies, to AIME by Dec. 1, 1951. Manuscript, April 16, 1951. Detroit Meeting, October 1951. This paper represents part of a thesis by S. C. Das Gupta submitted in partial fulfillment of requirements for the degree of Doctor of Philosophy in Metallurgy to the University of Notre Dame. to transform at room temperature although at a decreasing rate. Since only a small amount of austenite transforms in this fashion, isothermal formation of martensite was considered to be merely a "tailing off" effect of the main hardening reaction. Another point of view was advanced by Kurd-jumov and Maksimova"' * who reported that the austenite-martensite reaction could be suppressed by rapid cooling of a 6 pct Mn-0.6 pct C steel in liquid nitrogen and that as much as 25 pct martensite could be formed by reheating to and holding at subzero temperatures above that of liquid nitrogen. These investigators came to the conclusion that martensite is thermally nucleated at certain special regions in the austenite and subsequently grows to observable dimensions. Isothermal transformation of austenite to martensite was cited by Fisher, Hollomon, and Turnbull" as evidence supporting another nucleation and growth theory of martensite formation. According

Jan 1, 1952

By G. R. Speich, S. A. Kulin

The isothermal formation of martensite at subzero temperatures has been studied in an austenitic stainless steel. The amount of martensite formed isothermally in a given time was found to follow a C-curve behavior with decreasing temperature. Since isothermal rnartensitic transformation occurs at temperatures above the true M. temperature of the olloy, the observed M. temperature is dependent on the cooling rate. It is believed that a general martensite theory must include the basic concepts of the strain embryo theory, and also the important role of thermal fluctuations. UNTIL recently the isothermal component of the martensitic transformation had been considered as a trivial effect in theoretical considerations of this type of transformation. One of the postulates used in defining a martensitic transformation was that the reaction was temperature dependent but not time dependent. Cohen and his associates1'2 were the first to report the isothermal formation of mar-tensite. They obsel-ved the time dependency of the martensite reaction at room temperature in plain carbon steels and certain tool steels. They found that a small amount of additional transformation occurred after the specimen reached the quenching temperature and that the rate of this isothermal transformation decreased rapidly with time. These authors considered this phenomenon to be a "tailing-off" of the main athermal reaction. Kurdjumov and MaksimovaL 5 studied the isothermal formation of martensite at subatmospheric temperatures in steels containing 0.6 pct C-6 pct Mn, 1.6 pct C, and in an iron-base alloy containing 23 pct Ni and 3.4 pct Mn. In each case it was found that both the initial rate and the total amount of isothermal transformation exhibited sharp maxima with decreasing temperature. They stated that it was possible to completely suppress the martensitic transformation in the 0.6 pct C-6 pct Mn steel by drastic quenching in liquid nitrogen. They interpreted their results in terms of a thermally activated process. The maximum in the rate of nucleation vs. temperature curve was assumed to be a result of the temperature dependence of the rate of growth of the martensite plates. It has been demonstrated," however, that martensite plates can form with great rapidity even at temperatures approaching absolute zero. In addition," attempts to confirm the reported suppression of the martensitic transformation by rapid cooling of a 0.6 pct C-6 pct Mn steel have proved unsuccessful. On this basis, and on other theoretical grounds, the validity of some of Kurd-jumov's results and conclusions is open to serious question. Cohen, Machlin, and Paranjpe' have proposed a theory of martensite formation based on the nucleation of strain embryos and a growth mechanism involving cooperative atomic displacements. According to this theory thermal fluctuations can make a strain embryo supercritical and thereby produce martensite isothermally. Das Gupta and Lement' have investigated the isothermal formation of martensite in a 17 pct Cr-0.7 pct C steel at temperatures below the ambient. They found that partial suppression of the martensitic reaction occurred on quenching into liquid nitrogen. That part of the reaction which can be suppressed is taken to correspond to the isothermal rather than the athermal part of the rnartensitic transformation. The initial rate of isothermal transformation for this steel was found to go through a maximum with decreasing temperature. This is in conformity with the findings of Kurdjumov and Maksimova for the steels used in their investigation. Das Gupta and Lement conclude that isothermal martensite formation occurs by the nucleation of new plates and is always preceded by some athermal transformation. An iron-base alloy containing 14 pct Cr-9 pct Ni was used in this investigation to study further the kinetics of the austenite to martensite transforma-

Jan 1, 1953

By G. R. Speich, S. A. Kulin

R. C. Shnay (Dept. of Mines and Technical Surveys, Ottawa, Ont., Canada)-—The authors are to be congratulated on an interesting and informative paper. However, several questions arise when the isothermal formation of martensite .is considered in relation to the formation of bainite at temperatures near the M, point. The reactions in an alloy and in a plain carbon steel would appear to be comparable; and in the course of a recent investigation of the austenite-bainite transformation in a plain carbon steel,Ia the writer has obtained data at temperatures above and below the M, point. These data are in some respects similar to those of the authors and a comparison would seem justified. The steel used in this investigation had the nominal composition: C, 0.80 pct; Mn, 0.20; and Si, 0.20. The course of the transformation was followed by the use of a magnetic balance. The magnetic moment of the specimen as determined by this instrument is directly proportional to the percentage of the total volume of the specimen which has transformed to bainite. ~lf specimens were austenitized at 1850°F for 4 hr, quenched to the transformation temperature, and quickly transferred to the magnetic balance. Readings were taken at regular intervals until the transformation had stopped. A typical transformation curve is shown in Fig. 9. The preliminary curve preceding the S-shaped curve was found at all temperatures above the M, to around 600°F. Both the extent of this preliminary curve and the time interval between it and the S-shaped curve were found to decrease with increasing temperature as was suggested by Davenport and Bain.14 Therefore,

Jan 1, 1953

By R. B. G. Yeo

Pronounced isothermal martensite formation at room temperature was measured dilatometrically in a steel containing 0.01 pct C, 24.9 pct Ni, 0.26 pctAl, 2.58 pct Ti and 0.25 pct Cb. It is shown that martensite will form isothermally if stabilization of the austenite-martensite transformation is eliminated by removal of carbon. The decarburization of two iron-nickel alloys allows isothermal transformation to martensite to occur at temperatures above their athermal Ms temperatures. In an Fe-22.4 pct Ni alloy, at the 0.008 pct C level, the Ms temperature during air cooling is 85°F higher than that during- water quenching-. At the 0.15 pct C level this difference is reduced to only 5°F. The introduction of carbon causes stabilization during air cooling which lowers the Ms temperature virtually to the same level as determined during mate?. quenching. Thus in the 0.15 pct C alloy, the Ms temperature is almost independent of cooling rate. It is suggested that rival theories of martensite formation should be reexamined in alloys of sufficiently low carbon and nitrogen content to eliminate the complication of stabilization. The Ms temperature of a steel containing 0.01 pct C, 24.9 pct Ni, 0.26 pct Al, 1.58 pct Ti and 0.15 pct Cb, solution treated at 1500°F and air cooled as 1/8-in. diam specimens, was found to be 57 °F. 1 However, when held for several hours at room temperature the steel hardened slightly and became appreciably ferromagnetic. Isothermal transformation to martensite was first revealed by Kurdjumov and Maksimova2 in an iron-base alloy containing 0.6 pct C, 6.0 pct Mn and 2 pct Cu. Most of the other studies of isothermal martensite have also been confined to highly alloyed steels which transform at subzero temperatures; for instance, DasGupta and Lement, 3 Cech and Hollomon, 4 Machlin and Cohen,5 Shih, Averbach, and Cohen.6 Averbach and cohen7 noted a rapidly decaying volume increase at room temperature in a steel containing 1 pct C, 1.5 pct Cr and 0.2 pct V. cina8 and Marshall, Perry, and Harpster9 have described the isothermal transformation to martensite at room and subzero temperature in stainless steels. Kurdjumov10 has recently reviewed the subject, noting that the isothermal formation of martensite can be observed only in the temperature range somewhat below Ms or below -50°C. The isothermal formation of martensite may be characterized as follows: 1) It occurs in steels of widely different compositions, the main prerequisite being a low transformation temperature. 2) The transformation occurs by the formation of new plates and not by the growth of old ones. 3) The isothermal transformation is suppresed by stabilization during slow cooling or holding at temperatures near room temperature. However, the factors governing this mode of transformation are still not fully understood. The formation of martensite isothermally suggests an activated process. The early theories of martensite formation did, in fact, utilize classical nucleation concepts. Kurdjumov 11 first proposed the presence of thermally activated nuclei of different sizes and compositions which would grow if they reached critical size during cooling. This treatment was enlarged and developed quantitatively by Fisher, Holloman, and Turnbull12,13 and was later reviewed by the latter two authors.14 Fisher15 associates Ms with a nucleation rate of one per cc per sec. At slightly higher temperatures the nucleation rate is so low that isothermal formation of martensite is not observed in reasonable times. At slightly lower temperatures the nucleation rate is so high that it could not be suppressed by rapid cooling. By judicious use of parameters Fisher 16 was able to obtain good agreement between classical nucleation theory and the incubation times of martensite formation.4 However the application of these principles to martensite formation in iron-nickel alloys15 predicts that an alloy containing about 30 pct (31 at. pet) Ni would not transform. The presence of an Ms temperature at -223° C in an alloy containing 34.1 pct (33 at. pct) Ni led Kaufman and cohen17 to reject the hypothesis of homogeneous nucleation in favor of the reaction path theory originally proposed by Cohen, Machlin and Paranjpe.18 This theory conveniently explains the formation of martensite athermally but becomes labored when dealing with isothermal formation. Kaufman and cohen19 did explain the isothermal activation of embryos by assuming it to be a result of the expansion of their boundary dislocation loops but this treatment predicts an upper temperature limit of isothermal martensite formation.

Jan 1, 1962

By E. S. Machlin, Morris Cohen

The isothermal formation of martensite in a 71 pct Fe, 29 pct Ni alloy is found to take place mainly by the nucleation of new plates rather than by the growth of existing ones, and is dependent on the temperature and time of the isothermal hold, the amount of atherrnal martensite present, the state of internal strain, and the application of tensile stress. The conclusion is reached that the isothermal nucleation is activated by thermal fluctuations superimposed on localized regions of very high strain. The reaction-path concept, involving martensite nucleation at strain embryos by athermal activation on cooling, or by thermal fluctuations on holding, reconciles the athermal and isothermal modes of transformation, and offers a unified picture of the kinetics of martensite formation. THE phenomenon of isothermal martensite formation has been the subject of considerable study lately,'-' and its reality has been thoroughly established. The existence of such isothermal transformation has been claimed to confirm the nucleation and growth hypothesis as applied to the martensitic reaction.1 " However, the isothermal formation of a martensitic phase is also compatible with the reaction-pathw escription of the transformation. The purpose of this investigation is to examine the isothermal mode of the martensitic reaction, both theoretically and experimentally, in order to test the likelihood of the above two mechanisms. Material and Methods The alloy used in this study contained about 71 pct Fe and 29 pct Ni. The complete analysis was 29.5 ± 0.2 pct Ni, 0.036 pct C, 0.02 pct N, 0.19 pct Mn, 0.09 pct Si, 0.008 pct P, and 0.006 pct S. The amount of isothermal transformation was determined by electrical resistance measurements using a Kelvin double bridge. The specimens were in the form of rods 1/16 in. diam x 4 in. long. Aus-tenitizing was performed at 1095°C for % hr either in a pre-purified nitrogen atmosphere or in an evacuated Vycor tube, followed by oil quenching to room temperature. After this treatment, the samples were completely austenitic, and the transformation studies were conducted at subatmospheric tempera- tures. Lineal analysis and X-ray determinations of retained austenite were used to calibrate the electrical resistance changes in terms of the percentage of martensite. The martensite range curve for this alloy is given in Fig. 9 of ref. 6. The M, temperature occurs at about -18°C (255°K). Reaction-Path Theory The nucleation and growth description of the martensitic transformation that occurs isothermally has already been given by Kurdjumow.','2 However, no comparable treatment exists using the reaction-path concept." Consequently this section presents the reaction-path description of the isothermal mode of the martensitic transformation. According to the latter theory, there is a reaction path (a strain) which describes the relative motion of the atoms during their transference from the austenitic to the martensitic state. The possibility exists, therefore, that the nucleation of a martensitic plate is accomplished by an embryo that comprises a state intermediate between that of austenite and martensite. This state may be attained within a finite volume by means of a homogeneous strain of the austenitic matrix. There is a free-energy barrier (F.) along the reaction path, corresponding to a critical strain in a critical volume. When activation is achieved, a spontaneous increase in strain and volume of the embryo takes place to generate the full-size martensitic plate. This mode of nucleation is to be distinguished from the more conventional mode involving embryos of the martensitic phase that merely grow in size by an interface motion. There are two ways whereby embryos in the austenitic phase can achieve the critical state of strain for isothermal nucleation: 1—through the formation of internal strains by plastic deformation or by stress; and 2—through thermal fluctuations, preferably superimposed upon existing internal strains. The activation free energy F,, must be supplied either mechanically by internal strains or thermally by energy fluctuations, or by a combination of both.

Jan 1, 1953

By David J. Mack, A. H. Kasberg

The transformation characteristics are found to resemble a similar binary alloy. The differences are due to the alpha iron particles. While strength properties of the isothermally transformed alloys are high, ductility is low, resulting in high notch sensitivity. The elastic modulus can be varied widely and correlates with strength. Abnormal grain growth sometimes occurs. A CONSIDERABLE amount of work has been done on the isothermal transformation of high purity aluminum bronzes using the methods pioneered by Davenport and Bain.' The first isothermal study of these alloys was made by Smith and Lind-lief,' but more recently Mack", " and Klier and Grymko5 have continued the work. In commercial aluminum bronzes, one or more elements are always added to refine the ß grain size and to improve the mechanical properties. The effect of these elements on the isothermal transformation has not been reported in the literature. No systematic study of the properties of isother-mally transformed aluminum bronzes has apparently been reported, although adequate information is available on their properties after more conventional heat treatments. The fact that a eutectoid alloy will reject face-centered cubic a during isothermal transformation at appropriate temperatures2'" warrants investigation of the properties of such structures. Another cogent reason for determination of properties is the fact that the modulus of elasticity of the eutectoid alloy varies widely with the nature of the heat treatment." Preliminary experiments7 indicated that a high purity eutectoid alloy would not give reproducible results in property determinations because of ex- cessive ß grain growth during heat treatment, a single ß grain frequently occupying the entire cross-section of a standard test bar. For this reason a commercial aluminum bronze containing iron as a grain growth inhibitor was used. The isothermal transformation diagram of the alloy was first determined. On this basis, the heat treatments necessary to produce representative structures were selected and the properties determined. Material A commercial alloy (Ampco Grade 20) of the following composition was used: Cu, 83.68 pct; Al, 12.31 pct; Fe, 3.68 pct; Ni, 0.28 pct; and Mn, 0.03 pct. This alloy proved to be very close to the eutectoid composition and may be compared with the accepted value of 11.8 pct A1 in the binary Cu-A1 system. Comparison is valid only because of the "negative replacement effect" of the iron? Since traces of excess y² are present, the alloy is slightly hyper-eutectoid. The material was received in the form of 3/4 in. round, hot extruded bars. Specimens of approximately ? in. thickness were cut from this stock and drilled to allow suspension in the salt bath. The structure of the as-received material varied from 5 to 40 pct a + y2* in a matrix of ß. The amount of a + y² present was dependent upon the ß grain size of the extruded material. The coarse grained portions did not form as much a + Y, as did

Jan 1, 1952

By C. W. Phillips, D. N. Frey

A commercial Ti-Fe-Cr alloy, Ti-150, exhibits a martensitic transformation on cooling and two nucleation and growth reactions, one above and one below the Mg-Mf region, on isothermal holding below the single-phase ß temperature range. The reactions were followed by metallographic, X-ray diffraction, and microhardness methods. THE rapidly expanding engineering application of titanium and its alloys makes increasingly urgent an adequate knowledge of the effect of heat treatment on this material. The work herein reported is an attempt to determine certain aspects of the behavior of a commercially available alloy, Ti-150, during heat treatment. The alloy, with a nominal composition of 2.7 pct Cr, 1.3 pct Fe, 0.25 pct 0, 0.02 pct N, and 0.02 pct C, was received in the as-rolled condition as a 3/4 in. round. Slugs 0.11 to 0.12 in. thick were cut under a coolant using a silicon carbide wheel. Preliminary work established that the alloy consists of two phases, a (hexagonal close-packed) and /3 (body-centered cubic), at temperatures below 1750" to 1775"F, and of only /3 phase at higher temperatures. The slugs were heated for 1 hr in air at 1800°F to form a single-phase /3 matrix, then transferred to molten tin (1250° to 1650°F) or lead (1150°F and below) for isothermal holding. Holding temperatures ranged from 700° to 1650°F, times from 0.001 to 23 hr. The treated slugs were then sectioned, one part for metallographic examination and diamond pyramid hardness, the other, after deep etching to remove the high oxygen surface,' for comparative lattice parameter determination and X-ray phase identification. Hardness measurements were taken on the lightly etched surfaces of the metallographic specimens; a 1000 g load on a 136° diamond pyramid was used. Five impressions on each sample were made and the average of the hardness calculated. The use of microhardness measurements to follow the course of an isothermal transformation reaction is undesirable because of the possibility of wide differences in microconstituent hardness.' They were used in this case, however, because of the small amount of material available. At the higher transformation temperatures, where the above-mentioned objection was apparent, the readings were proportioned to correspond to the estimated ratio of transformed and untransformed constituents present. At lower temperatures, resulting in much finer microconstituents, the microhardness measurements were more uniform. The standard deviation of the mean of five impressions calculated from these latter values was 4.2 DPH. Employing 2a limits (probability 0.95) the experimental error of the hardness measurements averages was 8.4 DPH. The etchant used was composed of one part glyc-erol and one part 48 pct hydrofluoric acid." Of a number of etches tried, this one appeared most satisfactory. All micrographs were taken at 500X magnification and with oblique lighting to permit easier identification of the acicular structure resulting from rapid cooling. No advantage was gained by using polarized light for metallographic examination. Phase identification and lattice parameter measurements were made with a North American Philips high angle spectrometer in conjunction with a recording potentiometer. Considerable difficulty was encountered with some of the lattice parameter measurements because of the extreme broadness of the few diffraction lines obtained from samples consisting mainly of strained, untempered, quenched structures.

Jan 1, 1953

By C. S. Roberts

Microstructural studies of a Mg-10.3 pct Al alloy showed that discontinuous precipitation during aging multiplies the grain boundary area available for easy deformation in elevated temperature creep. An increase of strain rate for a 6.2 pct Al alloy at 400°F, where precipitation and creep are concurrent, caused an increase in the volume percentage discontinuously precipitated at the completion of the process. The relatively poor elevated temperature creep resistance of heat treated alloys of the Mg-AI-Zn type was explained in these structural terms. C resistance of the heat treatable Mg-A1 based alloys is poor compared to that of Mg-Ce based alloys. While the latter show a continuous precipitate, Mg-A1 alloys prefer a discontinuous precipitate at grain boundaries. '.:' It has been concluded that the excellent creep resistance of Mg-Ce alloys was primarily due to the continuous precipitation which especially localized at the grain boundaries.' The purpose of the present investigation was to obtain a parallel correlation, if such exists, between the poor creep resistance and the state of precipitation in Mg-A1 alloys. Discontinuous precipitation produces about three times as many true grain boundaries as originally existed. This microstructure would be expected to have a marked influence in creep, where quantity of grain boundary area is important. Experimental Procedure The two experimental alloys contained 6.2 and 10.3 pct Al. They were made from electrolytic magnesium by alloying under flux and casting in 3 in. diam permanent molds. The ingots were extruded into 1 1/4 x 1/8 in. flat stock under the following conditions: billet preheat 700°F (2 hr), container and die temperature 700°F, speed 3 fpm, and area reduction ratio 45:1. The creep specimens were machined from blanks which had been solution heat treated 24 hr at 770°F and aged 16 hr at 400°F. Attempts were made to control the grain size as closely as possible for extruded stock. The grain size after solution heat treatment was about two thousandths of an inch for the 6.2 and five thousandths for the 10.3 pct Al alloy.

Jan 1, 1957

By Gareth Thomas

The analysis of Kikuchi pattersns of exct ovientalions from single cryslals and paired Kikuchi lines from single and overlapping crystals is shown to be useful and quanlitalve and is applied to Phase transfovnzcitions including ordering, spinodals, mavten-silic, and nuclealion and growth pyocesses. 112 pinciple, the analysis of exacl orientations enables the crystal system and the Bravais lattice of a crystal to he determined. The advantages of the davk-field imaging technigure for detecting vevy small precipitates are also described. ALTHOUGH Kikuchi electron-diffraction patterns were first observed nearly 40 years ago,' little systematic application seems to have been made of them, until fairly recently during electron microscopy, when their usefulness in contrast experimen and for determining exact orientations8'9 has been pointed out. With the availability of gonio-metric specimen tilting stages it has now become possible to make much wider use of diffraction patterns, particularly the Kikuchi pattern, which is the main subject of this paper. The treatment presented here is not exhaustive but it is hoped that it will stimulate more use of Kikuchi patterns during electron-microscopy investigations. The Kikuchi pattern is formed as a result of Bragg diffraction of the inelastically scattered electrons produced during the interaction of the beam and thick specimens, The important feature of these patterns is that they give an accurate representation of the symmetry of the crystal being investigated, so that it is possible to identify crystal systems and even the Bravais lattice. This means that new structures, e.g., formed in phase transformations, may be identified during normal electron microscopy, so that Kikuchi-pattern analyses considerably extend the uses of the electron microscope. Recent work has also shown that dark-field images are more informative than bright-field images, particularly, for observing small precipitates. The second part of this paper discusses some applications and advantages of dark-field imaging in studies of two-phase systems. 1) DIFFRACTION PATTERNS 1.1) Spot Patterns. Electron-diffraction spot patterns have their limitations because of the importance of the form factor on the intensities and shape of the reciprocal lattice points. Because of the extension of these points into relrods, reflections are possible over a large angular range (+5 deg) and the patterns from thin regions can be complicated because second-layer relrods intersect the reflecting sphere. Spot patterns can thus give only an approximate idea of the crystallography unless the foil is tilted into exact orientation. In this case the spot pattern is symmetrical, with equal numbers of spots on the positive and negative zone directions about the origin. Such cases are necessary for structural analyses and have been used recently to determine the crystal structure of the ordered Ta64C phase." Exact orientation means that the plane of the reciprocal lattice lies exactly normal to the incident beam as shown in Fig. 1(b). It should be noted that in exact orientations the angle of reflection is less than the exact Bragg angle 9 so that the reciprocal lattice points lie to the outside of the sphere, Fig. 1(b). This deviation is denoted by the parameter s and in the usual convention4 s is negative for exact orientations and, of course, zero at the exact reflecting position shown in Fig. l(a). In order to avoid secondary reflections from thickness relrods it is advantageous to work in thicker regions of foils. Diffraction from thick regions may also produce Kikuchi patterns and as dis-

Jan 1, 1965

By A. Spilners, M. Simnad

The rates of formation of the different oxides on molybdenum in pure oxygen at 1 atm pressure have been determined in the temperature range 500° to 770°C. The rate of vaporization of MOO, is linear with time, and the energy of activation for its vaporization is 53,000 cal per mol below 650°C and 89,600 cal per mol at temperatures above 650°C. The ratio Mo03(vapor.lzing)/MoOS3(suriace) increases in a complicated manner with time and temperature. There is a maximum in the total rate of oxidation at 6W°C. At temperatures below 600°C, an activation energy of 48,900 cal per mol for the formation of total MOO, on molybdenum has been evaluated. The suboxide Moo2 does not increase beyond a very small critical thickness. At temperatures above 725°C, catastrophic oxidation of an autocatalytic nature was encountered. Pronounced pitting of the metal was found to occur in the temperature range 550° to 650°C. Marker movement experiments indicate that the oxides on molybdenum grow almost entirely by diffusion of oxygen anions. USEFUL life of molybdenum in air at elevated temperatures is limited by the unprotective nature of its oxide which begins to volatilize at moderate temperatures. Although the oxide/metal volume ratio is greater than one, the protective nature of the oxide film is very limited. Gulbransen and Hickman' have shown, by means of electron diffraction studies, that the oxides formed during the oxidation of molybdenum are MOO, and MOO,. The dioxide is the one present next to the metal surface and the trioxide is formed by the oxidation of the dioxide. Molybdenum dioxide is a brownish-black oxide which can be reduced by hydrogen at about 500°C. Molybdenum trioxide has a colorless transparent rhombic crystal structure when sublimed, but on the metal surface it has a yellowish-white fibrous structure. It is reported to be volatile at temperatures above 500" and melts at 795°C. It is soluble in ammonia, which does not affect the dioxide or the metal. In his extensive and classic investigations of the oxidation of metals, Gulbransen2 has studied the formation of thin oxide films on molybdenum in the temperature range 250" to 523°C. These experiments were carried out in a vacuum microbalance, and the effect of pressure (in the range 10-6 yo 76 mm Hg), surface preparation, concentration of inert gas in the lattice, cycling procedures in temperature, and vacuum effect were studied. The oxidation was found to follow the parabolic law from 250" to 450°C and deviations started to occur at 450 °C. The rates of evaporation of a thick oxide film were also studied at temperatures of 474" to 523°C. In vacua of the order of 10- km Hg and at elevated temperatures, an oxidation process was observed, since the oxide that formed at these low pressures consisted of MOO, which has a protective action to further reaction in vacua at temperatures up to 1000°C. Electron diffraction studies showed that, as the film thickened in the low temperature range, MOO8 became predominant on the surface. Above 400°C MOO, was no longer observed, MOO, being the only oxide detected. The failure to detect MOO, on the surface of the film formed at the higher temperatures does not militate against the formation of this oxide, since according to free energy data MOO3, is stable up to much higher temperatures. At the low pressures employed, this oxide would volatilize off as soon as it was formed. Its vapor pressure is relatively high and is given by the equations" log p(mm iig) = -16,140 T-1 -5.53 log T + 30.69 (25°C—melting point) log p(mm He) = -14,560 T-1 -7.04 log T+1 + 34.07 (melting-boiling point). Lustman4 has reported some results on the scaling of molybdenum in air which indicate a discontinuity at the melting point of MOO, (795°C). Above the melting point of MOO,, oxidation is accompanied by loss of weight, since the oxide formed flows off the surface as soon as it is formed.5,6 Qathenau and Meijering7 point out that the eutectic MOO2-MOO3 melts at 778C, and they ascribe the catastrophic oxidation of alloys of high molybdenum content to the formation of low melting point eutectics of MOO3 with the oxides of the melts present. Fontana and Leslie -explain the same phenomenon in terms of the volatility of MOO,, which leads to the formation of a porous scale. Recent unpublished work by Speiser9 n the oxidation of molybdenum in air at temperatures between 480" and 960°C shows that the rate of weight change of molybdenum is controlled by the relationship between the rates of formation and evaporation of MOO,. They have measured the rates of evaporation of Moo3 in air at different temperatures and estimated an activation energy of 46,900 cal. This compares with the value of 50,800 cal per mol obtained by Gulbransen for the rate of sublimation of MOO, into a vacuum.

Jan 1, 1956

By F. D. Rosi, C. A. Dube, B. H. Alexander

WHEN a deformed polpcrystalline metal is heated to a sufficiently high temperature, a recrystallized structure develops which consists of small, essentially stress-free grains. This transformation is referred to as primary recrystallization. Under certain conditions, on prolonged annealing of the specimens, a type of grain coarsening occurs in which a small number of grains commence to grow at various points by devouring the neighboring finegrained primary structure, which, in itself, exhibits a marked stability toward normal grain growth. This grain coarsening process has been referred to by Burgers' as "secondary recrystallization" or "exaggerated grain growth." Although much work has been done on recrystallization and growth processes, present knowledge of the factors which determine the occurrence and growth of secondary grains is inadequate for the advancement of a theory which accounts for the observed characteristics of this recrystallization phenomenon. In this respect, it would be important to make measurements of the velocity of growth of these large secondary grains. The early work of Feitnecht2 on aluminum showed that the size of secondary grains increased with temperature and time, suggesting that such measurements would be possible. More recently, Burgers1 found that the growth velocity of these grains in aluminum was constant over a large time interval, as is generally found for the growth velocity of primary grains in a uniformly deformed test piece.3-5 It is reasonable to expect that the determination of the activation energy of the secondary growth process might lead to a better understanding of the underlying atomic mechanism. This investigation, therefore, is one of a series primarily designed to measure the growth velocity of secondary grains at various temperatures for silver, copper, and aluminum. It is hoped that these studies along with those of orientation relationships, will establish a sound basis for theorizing as to the origin and nature of the secondary transformation. Experimental Procedure The metal used in the present investigation was high-purity silver (999.9 fine), which was supplied in the form of cold-rolled plates 0.250 in. thick. These plates were annealed at 500°C for 1 hr and then straight-rolled to a thickness of 0.04 in. in three reductions of approximately 45 pct, followed in each case by an anneal at 450°C for 1/2 hr. The 0.04 in. sheets were given a final reduction of approximately 50 pct, which resulted in a thickness of 0.02 in., thereby insuring two-dimensional crystal growth. Individual specimens of this material about 1x3 cm were used for all heat treatments, unless otherwise noted. The annealing was carried out under an atmosphere of argon in an Inconel tube furnace whose temperature was controlled to within -+2°C. Metallographic examination of the secondary transformation was accomplished by mechanically polishing the specimens on one side to approximately one half their thickness through 2/0 metallographic emery paper, and then electrolytically polishing in a solution of potassium ferrocyanide and sodium cyanide. With a current density of approximately 20 amp per sq dm, the time required to obtain a satisfactory surface varied from 3 to 5 min. Standard potassium bichromate was used to etch the electrolytically polished surfaces. A heavy etch with this solution produced a sharp contrast between the secondary grains and the fine-grained primary structure, darkening the latter while leaving the large secondary grains bright. All area measurements of the secondary crystals were made with a planimeter on enlarged tracings of the grains.

Jan 1, 1953

By J. W. Rutter, K. T. Aust

The migration of individual, large-angle grain boundaries has been studied as a function of tempereature and solute concentration in specimens of zone i.e filled lead containig very small additions of silver and of gold. Tile results are compared with various the-ories of grain boundary migration and with observations made prev.iorlsly of grain boundary migration in similar specimens of zone-refined lead containing tin additions. A previous investigation by the authors dealt with [he temperature dependence of grain boundary migration in bicrystals of zone-refined lead containing small additions of tin.' It was shown that tin additions as low as a few parts per million cause a large decrease in the grain boundary migration rate at any given temperature, as well as a marked increase in the temperature dependence of the migration rate. It was found that existing theories of grain boundary migration. based on the motion of dislocations. or upon the concept of atom transfer in groups across the boundary (group process theory). or upon the control of grain boundary motion by volume diffusion of impurity atonls along with the boundary. are incapable of accounting for the observations. The single process theory of grain boundary migration. which is an absolute reaction rate calculation based on the transfer ui atoms singly across the moving boundary, was found to predict the migration rate reasonably well for a number of boundaries whose motion was shown to be very little influenced by impurities, but not for boundaries whose illation was influenced markedly by impurities. It was concluded that the elementary process of grain boundary migration involves the activation of single atoms during transfer across the boundary. and that inadequate knowledge is available to permit the influence of impurities to be properly taken into account. The present study was initiated to check the validity of the above conclusions with other alloy systems, namely high-purity lead with small additions of silver and of gold. Both silver and gold diffuse faster. and with a lower activation energy of volume diffusion. than does tin in lead;' consequently, a study of the effects of silver and gold on grain boundary migration in high-purity lead offered a means of testing theories of boundary migration based on bulk diffusion of the solute (eg. ref. 3). In addition. it was hoped that the present work, in comparison with the results for tin in lead, would provide information concerning which factors are important in determin- ing the interaction between solute atoms and a grain boundary. EXPERIMENTAL PROCEDURE The preparation of bicrystals of zone-refined lead, with various silver or gold additions, was identical to that previously described for the lead-tin alloys.''4 Each bicrystal consisted of a striated crystal which was grown from the melt. and an adjacent striation-free crystal which was introduced by artificial nucleation and growth.''4 The striation or lineage substructure in the melt-grown crystal provided the driving force for grain boundary migration. During the preparation of striated single crystals by growth from the melt, it was found that silver or gold concentrations as low as 2 or 3 ppm by atoms were sufficient to cause formation of the hexagonal cell structure. which is due to the presence of impurity, during freezing. This structure is revealed on the solid-liquid interface by decanting the liquid during freezing. The hexagonal cell structure was observed previously4 in zone-refined lead crystals with tin contents above approximately 200 ppm by atoms. These concentrations of silver, gold, or tin are in agreement with the predicted amounts required for cell formation in lead,5'6 under the present conditions of freezing.4 The absence of cell structure at decanted interfaces, therefore, served as a useful indication that the silver or gold contents were less than 2 or 3 ppm by atoms in the specimens as grown. It was found that grain boundary migration occurred only very slowly when the solute content approached that necessary for cell formation. As a result, the present experiments were conducted with silver or gold additions less than 1 ppm by atoms. This impurity level is well within the solid solubility limits for silver and gold in lead.7 The annealing treatments, measurements of grain boundary velocities, and orientation determinations were carried out as described previously.' However. each bicrystal was also chemically polished in a solution consisting of 8 parts glacial acetic acid and 2

Jan 1, 1961

By M. E. Wadsworth, C. H. Pitt

Nickel corrosion in sulfuric acid solutions at elevated temperatures and oxygen wer-pressures was imestigated. Weight loss was 1inear with time and varied directly with oxygen concentration. Independent variation of the possible active species in solution indicated that the undissociated sulfuric acid molecule plays an important part in the corrosion mechanism. The rate determining step was considered to be the adsorption of oxygen on a site containing adsorbed H2SO4. THE tendency for a metal to go into solution depends on a number of factors some of which are: 1) the position of the metal in the electrochemical series; 2) the type and concentration of corroding species in solution; 3) the concentration of oxygen in solution; 4) the type of film formed at the liquid-metal interface; and 5) the physical state of the metal. The corrosion of a metal in an aqueous solution may be considered to be a heterogeneous reaction that occurs at the liquid-metal interface. The reaction is generally considered to occur as a result of electrochemical potential differences within the system. Little work has been published in the literature concerning the corrosion of pure nickel. A great deal has been publistled dealing with the experimental and theoretical aspects of the over-all corrosion process. Evans,1 Uhlig,' and recently Todt 3 and Tomashoff,4 have published books which review the results of experiments which have been conducted on corrosion, and have developed some of the basic ideas of the more widely accepted theories of corrosion. Friend5 states that. nickel does not readily discharge hydrogen from any of the common nonoxidiz-ing acids and that a supply of some oxidizing agent is necessary for appreciable rates of corrosion. He further notes that at temperatures between atmospheric and close to boiling, corrosion rates tend to vary with concentration, temperature, and aeration in acid concentrations up to 25 pct. Whitman and Russel6 studied the effect of dissolved oxygen on a number of metals in acid solutions and found that oxygen accelerates the rate of corrosion of nickel in sulfuric acid. Flowers and Kelley7 in studying the action of dissolved CO, on electroplated nickel postulated that nickel dissolves as NiHCO3 in the presence of CO, and air. It was found that oxygen was necessary for corrosion to take place and that the rate increased with increasing amounts of oxygen in solution. Flowers and Kelley attributed the oxygen effect to cathodic depolarization. MacGillavry, Rosenbaum, and wensson' determined metal solution potentials of nickel in various electrolytes. They found that nickel has no tendency to corrode in alkaline solutions in the presence or absence of air. In acid solutions in the presence of air, metal solution potential increases in the order of 1/2 v were observed. This increase was attributed to probable chemisorption of oxygen on the nickel surface and the corrosion mechanism suggested was the interaction of H+ ions with the nickel oxide formed at the surface. Delahay9 observed the reduction of oxygen on various metals. A polarograph was used to measure the course of the reaction. With some metals hydrogen peroxide was produced as an intermediate compound. However, nickel was not considered to be a hdrogen peroxide producer. Van Rysselberghe 10,11 and associates attributed the accelerating effect of carbon dioxide on the corrosion process to the presence of percarbonic acid produced in solution reacting with the hydrogen peroxide. The percarbonic acid can be reduced at less negative potentials than hydrogen peroxide and as a result accelerates the reaction. EXPERIMENTAL Since a complete description of the apparatus used can be obtained elsewhere,12'13 no attempt will be made here to describe the equipment other than to note that it consisted essentially of an autoclave unit from which samples could be withdrawn periodically without disturbing the interior state of the autoclave. A titanium autoclave liner was used to protect the autoclave from the solution.

Jan 1, 1961

By R. G. Davies

Isothermal annealing of quenched Fe3Al reveals that the superlattice forms by the nucleation and growth of ordered domains. The activation energy for isothermal ordering and initial domain growth is -29 keal per mole, which is thought to he the activation energy for vacancy migration. Domain growth, in the absence of excess vacancies, has an activation energy of -100 keal per mole. A maximum in the hardness is observed when the structure consists Of ordered domains, -35-4 in size, growing into a short-range ordered matrix and, also as a function of degree of order, S, with an infinite domain size at S = 0.4 to 0.5. It is proposed that both maxima are due to a transition from deformation by unit dislocations to deformation by superdislocntions. FE-AL alloys containing more than -23 at. pet Al, when quenched from above the critical temperature for the Fe3A1 type of order, always contain the maximum possible amount of FeAl (B2) type of order' (see Fig. 1). Thus when the Fe3A1 superlattice forms, in a 25 at, pet A1 alloy, it does so from an imperfect FeAl structure. The formation of the Fe3Al superlattice has been the subject of several recent X-ray,1, 2 electrical-resistivity,3"6 and dilatometric3 investigations. McQueen and Kuczynski3 concluded that the isotherma1 ordering of Fe3Al, after quenching from 800°C (T,, the critical temperature for order, is 55O°C), takes place by the nucleation and growth of ordered domains. However, Feder and Cahn4 found that the isothermal ordering was complex but more like a homogeneous nucleation process than a nucleation and growth process. Lawley and Cahn2 reported that the degree of order is a continuous function of temperature which is a characteristic of a homogeneous nucleation process. Thus, the evidence is conflicting as to the nature of the Fe3Al ordering reaction. Lipson7 was the first to suggest that ordered domains should exist in the Fe3A1 structure since, as

Jan 1, 1964

By G. Borelius

ABOUT the turn of the century, Gibbs' thermo-dynamic theory of heterogeneous equilibrium, on the one hand, and the experimental methods of thermal and microscopic analysis, on the other, gave to the physical metallurgist his first scientific tool, the equilibrium diagram. The classical equilibrium diagram of a binary alloy system shows the boundaries between ranges of homogeneous and heterogeneous equilibrium in their dependence of concentration and temperature. A homogeneous solid sohtion which on cooling passes such a boundary is assumed to precipitate, forming a mixture of two phases with different concentrations. The equilibriunl diagram and the equilibrium theory, however, give no information about the time scheme of the process or the intermediate states passed during precipitation. For this reason it satisfies neither the practical need of the metallurgist nor the curiosity of the physicist. As a matter of fact, in the heat treatment of alloys for technical use the objective very seldom is the equilibrium state. Thus good mechanical properties of construction material are connected, for the most part, with some intermediate state. As these intermediate states are thermodynamically unstable, there is, from a theoretical point of view, always to be expected a decay of the good properties with time; and it is a matter also of practical interest to know whether this natural life time of a material is of the order of, say, ten or thousands of years. Thus, for many reasons, there is a current demand to complete our knowledge of equilibrium through knowledge of the kinetics of the precipitation phenomena. From the point of view of the physicist, the most interesting question in this case is whether there are any general laws governing the kinetics. According to a generally accepted view, precipitation is ruled by two more or less independent phenomena, the formation of nuclei of a new phase and the growth of these nuclei. It is also commonly accepted that there is a tendency for the velocity of growth to increase with increasing temperature because of the increasing mobility of the atoms. There is also a tendency for the velocity of growth to decrease in the neighborhood of the two-phase boundaries. So far, however, very little is known quantitatively about this fundamental phenomenon in the case of solid metallic systems. In our laboratory attention has been directed especially toward the nucleation phenomena, and a series of measurements have been carried out with the guidance of a work- ing hypothesis (based on experiences from previous work on order-disorder transformations in alloys) about the influence on the nucleation of thermo-dynamic potential barriers. However, before discussing the experiments, the theoretical ideas will be considered. In a binary solid solution the arrangement of atoms on the lattice points approaches with increasing temperature a state of full randomness, as illustrated by the ball model of Fig. 1, that might represent a [111] plane of a face-centered alloy with 30 pct "black" and 70 pct "white" atoms. In reality the atoms are changing places continually with their neighbors so that the picture should rightly have been a moving one. On account of this thermal motion the concentration of black atoms within a certain group of, say, a hundred or a thousand lattice points fluctuates with time around the bulk concentration of 30 pct in a manner governed by statistical laws. With decreasing temperature two independent changes in this state grow more and more important. First, the mobility of the atoms decreases, and second, the forces between the atoms will have an increased influence on the fluctuations. In alloys with a tendency for precipitation, which are the concern of this lecture, the distribution function of concentration fluctuations will broaden, so that the relative probability of great local variations from the bulk concentration increases. Fig. 2 gives an example of such a fluctuation. When the alloy is supercooled below the solubility limit into the range of two-phase equilibrium, the fluctuations will now and then at some point give rise to a state that resembles the equilibrium state and thus will form a stable nucleus that is capable of growing by diffusion processes. In discussions with colleagues and in the literature, I have often encountered the idea that three or four atoms of the dissolved metal could form a nucleus of the new phase. A look at the ball model might be enough to indicate that this cannot be true. If it were true, there should be nothing but nuclei, whereas we know from experiment that nucleation must be a rather rare occurrence. In fact we have, as will be mentioned later, certain reasons to believe that the nuclei are formed by fluctuations containing some hundreds of atoms, which should be the order of the number of black balls in the fluctuating group of the figure, if it were extended into three dimensions. As a working hypothesis we have assumed that the fluctuations producing nuclei, though large and rare, still are ruled by the distribution laws of fluctuations of the supercooled solid solution in its initial state. Thus the probability of nucleation will be connected to the thermodynamic properties of the solid

Jan 1, 1952

By E. T. Turkdogan, P. Grieveson

Experimental results are presented for the rate of solution of nitrogen in a iron in the temperature range 750° to 873°C and in 6 iron in the temperature range 1410° to 1470°C. It is shown that the rate controlling process is diffusion of nitrogen into the iron. The diffusiting of nitrogen in a and 6 iron is derived from the results, and the temperature dependence of the diffusivity is represented by the equation D = 7.8 x e- 18,900/RT sq cm per sec. The solubility of nitrogen in a and 6 iron, in equilibrium with 1 atm pressure of nitrogen, has been measured. Using these results and other available data, it is found that the variation of the logarithm of nitrogen solubility with the reciprocal of absolute temperature is nonlinear. In an Appendix, some results of Darken and Smith are presented for the rate of solution of nitrogen in iron using ammonia + hyidrogen mixtures. These data also support the view that diffudsion of nitrogen in iron is the rate-controlling process when ammonia + hydrogen mixtures are used. A considerable effort has been made to obtain data on the solubility1-5 and diffusivity of nitrogen in a iron6-l2 because an understanding of the effect of nitrogen on the properties of steel must be based on an accurate knowledge of the properties of nitrogen in pure iron. However, no information is available concerning the kinetics of solution of nitrogen in a and 6 iron. Recently the authors13 have investigated the rate-controlling mechanism operating in the kinetics of solution of nitrogen in y iron. This study was directed to determine the rate-controlling processes for similar reactions with a and 6 iron as well as to establish values for the solubility of nitrogen in equilibrium with nitrogen gas in a and a iron. EXPERIMENTAL The procedure used in experiments to determine the rate of solution in cylindrical iron rods was the same as that described in a previous communication.13 Ferrovac E grade iron used in all experiments contained the following impurities in weight percent: C, 0.005; Mn, 0.001; P, 0.002; S, 0.006; Si, 0.006; Ni, 0.025; Cr, 0.002; V, 0.004; W, 0.02; Mo, 0.01; Cu, 0.001; Co, 0.01; 0, 0.007. After cleaning the surface, the iron rods were treated in an atmosphere of purified hydrogen for 17 hr before the reacting gas was introduced for known experimental times. After quenching, the samples were sectioned radially and analyzed for nitrogen. In addition to experiments using rods, iron foils were used in the measurements of solution rates of nitrogen in a iron. The foils of two different thicknesses were prepared by cold rolling Ferrovac E grade iron cylindrical rod to 0.051 and 0.152 cm. Foil samples were used in a rectangular form 5 cm long and 1.25 cm wide. The specimens were thoroughly cleaned of surface oxide with fine emery cloth and degreased with carbon tetrachloride immediately before entry into the furnace. The experimental procedure was the same as that used in the study with rods. At the completion of an experiment, the foil samples of the nitrogenized iron were analyzed for nitrogen after discarding 0.3 cm from the perimeter of the specimen. Iron foils were nitrogenized and denitrogenized in the a and 6 range with a gas mixture of 95 pct N and 5 pct H for times varying from 5 min to 2 hr. Results obtained for the average composition of nitrogen in iron for these experiments are presented in Fig. 1. Prior to the denitrogenization experiments, the samples were saturated with nitrogen at 1000°C and 0.67 atm N, giving a uniform nitrogen concentration of 0.0204 pct. According to the known a-y phase boundary in the Fe-N system,14 this composition lies within the ferrite region at temperatures 750" to 850°C. Use of this initial nitrogen content insured that reaction occurred between the gas and a single solid phase, a iron. Examples of the results for the mean concentrations of nitrogen in cylindrical iron rods, 0.356 cm radius for both the a and 6 ranges are given in Fig. 2. Typical examples of the results obtained for the radial distributions of nitrogen in rods are presented in Fig. 3. It appears that the results for radial distributions can be extrapolated to constant surface compositions which agree with the equi-

Jan 1, 1964

By R. Ruka, E. A. Gulbransen

ONE of the interesting questions in the understanding of the reaction of iron with oxygen is the kinetics and the mechanism of the crystal structure changes occurring in the formation and breakdown of the oxide film. In a recent work' the electron diffraction method was applied to the study of the solid phase reactions occurring internally in the oxide film on iron above 570°C and under vacuum conditions. The higher oxides which form under oxidizing conditions are reduced internally by the diffusion of iron into the oxide leaving the bulk of the oxide FeO (wüstite). This paper will ask and consider three questions. These are: (1) What is the nature of the depressed transformation temperature for the Fe3O4 + Fe ? 4Fe0 reaction in very thin oxide films?; (2) What is the mechanism of the forward transformation of Fe3O4 to FeO at or near 570°C?; and (3) What is the nature and mechanism of the slow reverse transformation of FeO to Fe3O4 and Fe below 570°C?. The composition of the oxide as well as its electrical, magnetic, mechanical, and chemical properties depends upon the kinetics of these reactions. Literature Survey The composition of the oxide film on pure iron at temperatures up to 570°C has been studied extensively by the electron diffraction method. Nelson,' Jackson and Quarrell,3 and Gulbransen and Hick-man4 have shown that Fe3O4 and a-Fe2O8 are formed in the temperature range of 225" to 570°C, while y-Fe,O, is most likely formed below 225°C for long oxidation times. In very thin films FeO (wüstite) is found to form at temperatures as low as 400°C, and its existence depends essentially upon the extent of initial oxidation. FeO is not found in the formation of thick films below 570°C. On the other hand X-ray5 and micrographic8 studies show that Fe3O4 may exist as the main com-ponent in thick films up to 650°C, well above the equilibrium temperature for the reaction Fe3O4 + Fe ? 4FeO. The kinetics of the forward and reverse trans-formations of the reaction Fe3O4 + Fe ? 4FeO have not been studied in oxide films. However, analysis of the data of Gulbransen and Hickman4 on the heat-ing and cooling of iron oxide films of known thick-nesses in high vacua shows that the forward reac-tion proceeds rapidly even for a 4000 A film at temperatures of 575" to 600°C. The reverse reaction or decomposition of wüstite (FeO) occurs at tem-peratures below 450°C. Thus, there appears to be a slow process in the decomposition which is in marked contrast to the rapid forward reaction. The decomposition of a bulk sample of FeO free from the metal has been studied extensively by Chaudron, Benard and coworkers using the micro-graphic, X-ray diffraction, and thermomagnetic methods. Chaudron7,8 first observed that the de-composition of FeO (wüstite) below 570°C occurs by the reaction 4FeO ? Fe3O4 + Fe. Later work by Chaudron and Forestier5 showed that the rate of de-

Jan 1, 1951

By D. Turnbull, J. H. Hollomon, J. C. Fisher

Application of the concepts of nu-cleation and growth to the analysis of experimental transformation data has led to valuable descriptions of phase transformations, an outstanding example being the transformation austenite —* pearlite which has been examined with particular care by Mehl and co-workers.&apos;-5 In addition to the pearlite transformation, the proeutectoid fer-rite and proeutectoid carbide transformations are known to proceed by nucleation and growth. Martensite, on the contrary, until recently was thought to form by a mechanism involving neither nucleation nor growth; however, extension of standard nucleation theory6 suggests that martensite, bain-ite, and the other products of austenite decomposition all grow from nuclei in the parent phase. The theory that martensite forms by nucleation and growth is strongly supported by recent experimental work of Kurdjumov and Maksimova.7 The concepts of nucleatioli and growth have been fruitful also in providing a sound basis for quantitative theoretical treatments of the kinetics of phase transformations. For example, Volmer and Weber8 and Becker and Döring9 developed the theory of nucleation from fundamental considerations to a point where excellent agreement was obtained with the results of experiments on the condensation of supercooled vapors. As a result of their analysis, the kinetics of vapor-liquid transformations now can be predicted. It seems probable that application of the theories of nucleation and growth to a quantitative study of austenite decomposition similarly will clarify the nature of the individual transfor: mations involved, and will enable the calculation of austenite transformation kinetics. In the present paper, the theories of nucleation and growth are applied to the austenite ? martensite transformation in steels. The analysis begins with a discussion of nucleation in single component systems. Martensite appears to be coherent with the parent austenite, hence the nucleation theory is modified to include the effects of elastic distortion. Nucleation in the two component iron-carbon system then is discussed, for most steels are primarily alloys of these two elements. Finally, M. temperatures and martensite transformation curves are calculated for each of several alloy steels of varying carbon and chromium content, and are compared with those determined experimentally by Lyman and Troiano10 and Harris and Cohen.11 Nucleation Theory NUCLEATION IN SINGLE COMPONENT SYSTEMS6,12-14 The work required for reversible formation of a region of phase within the parent a phase is expressed conveniently as the sum of two terms: W1 = Aa, the product of the area of the interface and the interfacial free energy, and W2 = VAf, the product of the volume of the region and the free energy increase per unit volume associated with the transformation. The total work is therefore W = Aa + VAf. When a is more stable than ß, Af is positive and W increases without limit as the volume increases. The transformation a ?ß does not occur. It is nevertheless true that small regions of phase ß enjoy temporary existence here and there in the a. The equilibrium number of ß regions of given size is proportional to exp(— W/kT) per unit volume of a, assuring that larger (ß regions occur with diminishing probability. When a is less stable than ß, Af is negative and W passes through a maximum as V increases. Assuming for simplicity that regions of ß are spherical, as is true when the interfacial tension is isotropic and there are no elastic strains, W = 4r2a + (4/3)*r3Af The maximum value of W is W* = 16iro3/3Af2 [1] for regions with radius r* = -2o/Af. [2] For single component condensed systems it has been shown14 that the steady rate of nucleation of 0 per unit volume of untransformed a is nearly proportional to exp[- (W* + Q)/kT] where Q is the activation energy for the unit processes of adding or removing one atom from an embryo or nucleus. If To is the temperature at which a and ß are in equilibrium, the rate of nucleation is a maximum at a temperature 0 < T < To where (W* + Q)/kT is a minimum. P regions smaller than critical size are called embryos; they tend to grow smaller and disappear, only exceptionally growing larger. Regions equal to or larger than critical size are called nuclei. A critical size nucleus may grow indefinitely large or may shrink back to a, either process decreasing the free energy of the region.

Jan 1, 1950

By B. N. Bose, M. F. Hawk

SINCE Davenport and Bain&apos; introduced the isothermal transformation technique for the study of austenite decomposition in steels, a new field of investigation has opened up. Extensive research has been carried out on many different steels, but relatively little work has been done on eutectoid transformations in analogous alloy systems. The researches of Smith and Lindlief,² Mack,³ and Klier and Grymko²0 on the beta-phase transformation in copper-aluminum alloys, of Hibbard, Eichelrnan, and Saunders4 and Smith5 on the kappa-phase transformation in copper-silicon alloys, of Chaudron and Forester6 on the decomposition of "FeO," and of Wasserman7 on similarities between the iron-carbon, copper-aluminum, copper-tin, and copper-beryllium systems essentially comprise the field. The iron-nitrogen system, in view of its remarkable similarity to the iron-carbon system, is particularly attractive for the study of eutectoid decomposition. The present investigation is concerned with the study of isothermal transformation, proceeding by the familiar nucleation and growth processes, and of martensitic transformation characteristics of pure iron-nitrogen alloys. The kinetics of transformation and the morphology of the transformation products have been investigated, for these must be the factors determining the properties and behavior of the final product. The Iron-Nitrogen System he binary iron-nitrogen phase diagram has been a matter of considerable disagreement among vari01.1s investigators. However, the phase diagram according to Eisenhut and Kaupp8 and Lehrer10 is now considered essentially correct as reproduced in fig. 1. The important values, 2.35 pct nitrogen and 591°C, for the co-ordinates of the eutectoid point (obtained from this diagram and used in the re- search to be described) were confirmed in a concurrent study of Paranjpe, Cohen, Bever and Floe." The eutectoid decomposition involves the transformation of the ? face-centered cubic interstitial solid solution of nitrogen in iron to the a body-centered cubic interstitial solid solution of nitrogen in iron and the ?&apos; nitride of iron. This nitride is essentially the compound Fe,N; it also possesses a face-centered cubic lattice but a considerably larger one than the ?; the nitrogen atoms in ?&apos; are arranged in an ordered way in the centers of the unit cubes.8,10,11 A martensitic structure has also been reported8,12 in iron-nitrogen alloys and was studied in the research to be described here. Materials and Methods The specimens used in this investigation were prepared by nitriding "pure" iron. This iron was supplied by the Westinghouse Research Laboratories and is the product bearing their trade name "Puron"; it contains about 0.04 wt pct total impurities, the major portion of which is oxygen. Specimens were ribbons several inches long, a tenth of an inch wide, and 0.125 mm thick; this was the

Jan 1, 1951