Search Documents

By J. S. Kasper, R. M. Waterstrat, B. F. Decker

THE existence of v phase in the Mn-Mo system was discovered in this laboratory several years ago,' but neither the conditions of its formation nor its composition were known, and preliminary attempts to reproduce the phase were unsuccessful. We wish now to specify the information appropriate to phase formation on the basis of experiments performed since that time. The starting materials were in every case electrolytic manganese and high purity molybdenum. Two methods of preparation were employed: l—arc-melting varying combinations of the elements in an inert atmosphere and subsequent heat treating, and 2—sintering compressed mixtures of finely divided powders of the elements. In the latter method, the powders were first passed through 200 mesh screen and then compressed into cylinders of % in. diameter and % in. length, which were sintered at elevated temperatures in an atmosphere of dry hydrogen. For some samples, the sintering temperature was 1200°C (for 1 hr), and these were then heat- treated at lower temperatures for times up to 24 hr. Eventually it was found, for the proper composition, that essentially pure a phase could be produced by sintering at 1125°C for 3 hr. Quenching to room temperature was effected by means of a stream of cold, dry hydrogen in a water-cooled chamber. The arc-melting procedure was found to be unsatisfactory after an investigation of many samples at different compositions and temperatures. Large quantities of manganese were vaporized during melting and none of the products were homogeneous or contained extended regions of pure phase. Nonetheless, these experiments established that no formation occurs in the Mn-Mo system below 1115oC, and indicated, at least roughly, the composition range of interest for further exploration. The more definite results obtained from samples produced by the second method are given in Table I. A portion of sample 2 heated for 16 hr at 1000°C in a Vycor tube transformed to molybdenum plus unidentified phases of complex X-ray pattern. From these few results it appears that the region of phase is quite narrow in composition. It is of interest that the composition is such as to make for difficulties in correlating this structure on an electron atom basis with other a structures of iron, cobalt, and nickel. In that respect Mn-Mo shows the same anomaly as Mn-Cr and Mn-V systems.' A second feature is that in common with the two other known molybdenum (Fe-Mo and Co-Mo), Mn-Mo v exists only at elevated temperatures. The X-ray powder pattern of the pure a phase of Mn-Mo is given in Table 11. It agrees well with the patterns established for other structures, especially with that for Fe-Mo as given by Goldschmidt.An attempt is being made to establish whether an ordering of the respective atoms occurs and such information will be reported in a future publication. Acknowledgment The authors acknowledge gratefully the assistance of Mrs. J. R. Belanger in some portions of this work. References 'J. S. Kasper, B. F. Decker, and J. R. Belanger: Journal of Applied Physics (1951) 22, p. 361. 2 A. H. Sully: Journal Inst. Metals (1951) 80, p. 173. H. J. Goldschmidt: Research (1949) 2, p. 343.

Jan 1, 1954

By A. E. Nehrenberg

IN an earlier publication' it was shown that the.. shape assumed by a volume of growing austenite is inherited from the prior structure. The matrix grains of pearlitic microstructures are equiaxed, and austenite grows in an equiaxed manner in these prior structures. Martensite and bainite, on the other hand, have an acicular grain structure, and the austenite produced in such structures grows in an acicular manner. In the discussion of this paper' it was pointed out that in cold-rolled tempered martensite, austenite grows in an equiaxed, rather than an acicular manner. The results of an experiment carried out to arrive at an understanding of this behavior are illustrated by the micrographs. Fig. 1 shows the acicular microstructure produced by heating tempered martensite to a temperature within the Ac1-Ac3 range, 1350°F. When tempered martensite is cold rolled, then heated to the same temperature, the equiaxed structure shown in Fig. 2 is produced. A sample of cold-rolled tempered martensite heated at 1350°F just long enough for the formation of a small amount of austenite is shown by Fig. 3 to have an equiaxed (recrystallized) ferrite grain structure. The small particles of. austenite which formed at 1350°F in the equiaxed ferrite matrix are likewise equiaxed. Thus, the equiaxed growth of austenite in cold-worked tempered martensite is not a contradiction of the principles described in the previous paper, for the growth pattern reflects the matrix grain structure existing at the time of austenite formation. References A. E. Nehrenberg: The Growth of Austenite as Related to Prior structure. Trans. AIME (1950) 188, pp. 162-174; Journal of Metals (January 1950). 2 J. R. Vilella and R. A. Grange: Discussion of ref. 1. Trans. AIME (1950) 188, pp. 1378-1381; Journal of Metals (November 1950).

Jan 1, 1953

By R. C. Hall

DEVELOPED to grow single crystals of alloys of high hardness and high melting points is a simple but effective apparatus described in this note. Growth from the melt, that is, by the Bridgman technique, hay be applied to metals or alloys which are physically hard as long as no phase transformation occurs on cooling to room temperature. The present apparatus was employed to grow single crystals of the ferromagnetic AI-Fe alloys above 1 wt pct Al and the Si-Fe alloys above 3 wt pct Si. Single Crystal Growth In a simple form, the method for growth of single crystals from the melt requires a molten metal charge in a pointed crucible which is slowly low-ered through a heated zone into a colder zone. Solidification starts at the bottom point and continues up the charge as the crucible is lowered through the heated zone. If a single crystal nucleates at the bottom, one crystal may grow throughout the charge under ideal conditions of slow growth.4,5 The presently developed method of growth was designed so that the chances that more than one crystal would grow were small. The single crystals were growth by the apparatus illustrated in Fig. 1. The charge was in the form

Jan 1, 1958

By Howard Martens, Pol Duwez

XPERIMENTS were performed on a Cr-Fc alloy Econtaining 44.7 pet Cr in order to determine the heat evolved during the transformation of the a solid solution into the s phase, and the change in volume accompanying the transformation. Since previous measurements on the rate of formation of a from s indicated that the reaction was quite fast as soon as the equilibrium temperature of 820°C was exceeded by some 50°, standard thermal analysis experiments were performed on heating specimens of s at the rate of approximately 40°C per sec. A definite thermal arrest was observed at 870"C, Fig. 1. This result indicates that the reverse seaction of into a is exothermic. From the heating curve of Fig. 1, taking 15°C per sec as the heating rate in the neighborhood of the transformation, 2.4 sec as the time for transformation, and assuming that the specific heat of the alloy at 850°C is an average between iron and chromium, namely 9.5 cal per mol per "C, the latent heat of transformation is approximately 342 cal per mol. The order of magnitude of this latent heat was also obtained by comparing the heating curve of Fig. 1 with a similar curve obtained with pure iron under identical experimental conditions. It was found that the length of the thermal arrest for s was about 1.8 times as long as that for iron. Since the latent heat of the a to 7 iron transformation is 203 cal per mol1, 2 the latent heat of it to a would be 365 cal per mol, which is in relatively good agreement with the value determined from the heating curve. For the determination of volume change, thermal expansion curves were recorded using a specimen containing 44.7 pet Cr and completely transformed to cr by previous treatment. The curves were recorded at a rate of heating of about 4°C per min and at approximately 840°C a sharr, increase in length of about 0.08 pet took place. Fig. 2. It is therefore established that the specific volume of a near the transformation temperature is about 0.24 pet smaller than that of a, and consequently, when the s phase forms in an a matrix, it would tend to shrink away from the matrix. Since the s phase is known to be extremely brittle, these results offer an explanation for the microscopic cracks often observed in the a phase under the microscope. The results of this study are in qualitative agreement with previously published data on stainless steel 19-9DLa and on a Cr-Fe alloy containing 45 pet Cr.' The increase in density due to the transformation from n to s given in ref. 4 is 0.90 pet. These measurements, however, were made at room temperature. whereas the presently reported 0.24 pet is the change at the temperature at which the transformation takes place during heating. In the graph of Fig. 2, the percentage change in length that would be observed at room temperature is given by the length X, which is about 0.27 pet. This would correspond to 0.81 pet in volume change. Acknowledgment This work was done at the Jet Propulsion Laboratory, California Institute of Technology, under Contract No. DA-04-495-ORD-18 with the Army Ordnance Corps, Washington, D. C. The authors wish to thank this agency for permission to publish the results of this investigation. References I H. Klinkhardt: Measurements of True Specific Heats at High Temperatures. Annalen der Physik (1927) 84, pp. 167-200. - H. Esser and W. Runghardt: Influence of the Method, of Dctcrmination on the Measured Heat Content of Iron. Archtvo ffur Dct Eisenhiittenwesen 11934) 8, pp. 37-38. J. J. Gilman, P. K. Koh, and 0. Zmeskal. Delta Ferrite Formation and its influence on the Formation of Sigma in a Wrought Heat Resisting Steel. Trans. ASM 1 1949) 11, pp. 1371-1388. 1 J. M. Thomas: Discussion of ref. 3. Trans. ASM (1949) 41, pp. 3 390-1392.

Jan 1, 1957

By L. D. Jaffe

THE data recently published1-" on heat treatment of titanium alloys have not been readily systematized into an overall picture. When the ß titanium phase transforms isothermally at low temperatures, the hardness follows a typical age-hardening pattern, going through a peak as schematized by the "hardness" vs "time of isothermal transformation" coordinates of Fig. 1. As the temperature is raised, the peak will decrease in height and occur at shorter times; slowing must however take place as the /3 transus is approached. The brittle microconstituent in the region of peak hardness has been termed /3'. Its structure is still undetermined but it is clearly a transition stage from soft /3 to the £ + a that is found at the right of the peak. On continuous cooling, high cooling rates correspond to short times on Fig. 1; abscissa labeling has been added accordingly. Extremely rapid cooling will retain soft /3; very slow cooling will precipitate soft a; intermediate rates give hard 8'. Increasing distance from the water-quenched end of a Jominy hardenability bar corresponds to slower cooling (right on Fig. I), but the range of cooling rates covered by a Jominy bar is only a small portion of that shown in the figure. The p-stabilizing alloying elements retard the transformation process, moving the material to the left on Fig. 1. A lean alloy is harder water-quenched (P') than air-cooled (/3 +a); a rich alloy is harder air-cooled (/3') than water-quenched (/3). Examples of the position of various compositions with respect to the curve, based on limited data in the literature, are given at the top of the figure. Oxygen and carbon accelerate the transformation process, moving the material to the right on Fig. 1. If an alloy quenched quickly enough to retain soft /3 is reheated below the £ transus (tempered), it ages, going through a hardness maximum as time increases (moving right on Fig. 1). Material which after cooling from the /3 range lies on or right of the hardness peak, will move still further right on tempering, and so soften. Since the rate of the tempering (aging) reaction increases with temperature, the figure also may be used to represent the variation in properties with tempering temperature, at short tempering times. The simple picture given here neglects martensite transformation, as well as differences between decomposition products of /3 formed by direct isothermal transformation and those formed by quenching and tempering. It may nevertheless be of use as a guide for practical heat treatment and for further research.

Jan 1, 1955

By L. Thomassen, N. Spooner, J. M. Thomas

NI-CR and some Ni-Cr-Fc alloys, when used as electrical resistance heating elements in reducing atmospheres, at times suffer rapid breakdowns due to so-called "green rot." These reducing atmospheres are most frequently used in bright annealing and heat-treating furnaces which are kept for long periods of time at 1500' to 1800°F. The green rot is a preferential oxidation of the chromium in the alloy to such an extent that the remaining metal frequently becomes ferromagnetic. The Curie point (magnetic transition point) at room temperature for a pure Ni-Cr alloy is known to be at about 7 pct Cr.¹ This represents a severe loss in chromium for one popular grade of resistance alloy, which is nominally 80 pct Ni-20 pct Cr. The formation of oxides along the grain boundaries makes the ribbon brittle and gives the fracture a green, earthy appearance—hence the name. The authors have studied this phenomenon, simulating the industrial conditions by heating resistance ribbons in various reducing atmospheres, such as moist hydrogen or the atmosphere specified in the ASTM test.² This atmosphere contains 16 pct H2, 10 pct CO, 4 pct C02, 1 pct CH1, 1.5 to 2.5 pct H2O, balance N2. Little success was had in producing severe corrosion with the commercial 80 pct Ni-20 pct Cr alloys. It had been noticed, however, that 90 pct Ni-10 pct Cr alloy wires deteriorated very rapidly when placed in the bottom of narrow thermocouple protection tubes. The bottoms of the tubes with the wires were heated in air. This took place both in porcelain tubes and in Inconel* tubes either with the cold end open or loosely stoppered with asbestos. On examination the wires showed microstructures which would be classified as green-rot corrosion. Subsequent investigations with various alloy wires confirmed the observation that corrosion occurred in some wires, including 80 pct Ni-20 pct Cr alloys, much more rapidly in such tubes than had occurred previously in test atmospheres mentioned above. When 90 pct Ni-10 pct Cr alloys are oxidized with an abundant air supply, the oxide coat consists of both nickel and chromium oxides. In this case thermodynamic equilibrium between the oxides is not established. However, when oxidized 90 pct Ni-10 pct Cr alloy specimens were heated to 1500" to 1800°F inside small diameter tubes, the oxide coat transformed to a bright metallic outer layer of nickel or nickel-rich alloy, beneath which appeared a green oxide, followed by intergranularly attacked metal. The obvious explanation is that in the narrow tubes, thermodynamic equilibrium is being established according to the equation NiO + Cr (in alloy) ? Ni + Cr2O3 This reaction is favored due to the fact that the oxygen pressure over nickel oxide is more than 10"' times the oxygen pressure over the green chromium oxide. The initial oxygen in the tubes is depleted very rapidly by oxidation of the wire and of the tube, if it is metallic. A confirmation of the occurrence of the reaction, reduction of NiO by the chromium in the alloy, was obtained by putting oxidized nickel foil into an evacuated quartz tube along with a piece of bright 90 pct Ni-10 pct Cr alloy ribbon, and heating the tube at 1820°F for a number of hours. The nickel oxide on the foil was completely reduced to pure nickel, leaving a bright foil. The 90 pct Ni-10 pct Cr alloy became strongly magnetic and showed the typical green-rot structure. Weights of the samples before and after testing showed the weight loss of the foil to approximately equal the weight gain of the ribbon. Evidently the bottom of long, narrow protection tubes, stoppered or not, may under certain conditions act just as the sealed quartz tubes in respect to being a confined space in which oxygen is not freely replenished. This condition of oxygen depletion can be prevented by introducing a small amount of air into the bottom of the tube. Many conditions becloud the green-rot phenomenon, such as carburization, sulphidization, and the presence of other corrosive agents. However, these experiments have shown that the basic reaction is a case of internal oxidation which can occur simply by adjusting the oxygen pressure to a low enough level so that the atmosphere will leave the nickel intact and oxidize the chromium. The action of the other agents in commercial atmospheres can then be taken up as individual cases. From these and numerous other tests completed to date or still in progress, it is believed that a more critical evaluation of electrical resistance alloys for their resistance to green-rot corrosion can be made. A much better understanding of the mechanism for this type of corrosion should also result. A more complete report on these experiments will be published later. In the May 1953 issue: TP 35213 Discussion: Solubility of Carbon and Oxygen in Molybdenum by W. E. Few and G. K. Manning, discussion by N. A. Gokcen: p. 747, first column, the fifth, sixth, and seventh lines should be the first, second, and third lines, followed by the present first, second, third, fourth, eighth, ninth, and tenth lines.

Jan 1, 1954

By W. M. Baldwin, C. R. Johns

Cobalt is reported1,2 to scale in accordance with the Pilling and Bed-worth3 parabolic law: where w = weight increase per unit surface area K = constant l = time The reported values of the parabolic oxidation constant, K, in the temperature range 600 to 1100°C follow an Arrhenius type of equation as inspection of Fig 1 shows. According to Valensi's theory5 on the scaling of metals where multiple-layered scales (for example, MeX, MeX2, and others) are possible, this fact could be construed as indicating that a single oxide only is formed on cobalt in this temperature range. Since the single straight line extends above 920°C where only CoO is stable in air (see Fig 2) it might be inferred that this is the only oxide formed on scaling. Arkharov and Voroshilova10 have studied the scales formed when cobalt is heated in air in the temperature range 385 to 800°C and report that CoO is indeed found in the scale but that CoO4 is also found as an overlay and in increasing amounts in the lower temperature ranges. No quantitative data regarding the relative proportions of the two oxides were reported. Again in the light of Valensi's theory this would indicate that the straight line in Fig 1 would be expected to undergo an inflection at some point in the temperature range 385 to 800°C. Chauvenet2 has studied the oxidation rates of CoO to Co3O4 and has found this reaction to proceed parabolically with time but reports parabolic scaling rates which are considerably (about ten times) larger than the scaling rates of cobalt metal itself. This is difficult to explain in terms of Valensi's theory although a full discussion of the possible discrepancy cannot be made in view of the lack of sufficient data. These points made an investigation of the scaling behavior of cobalt in air an attractive study. Discussion of Results Relatively large (approximately 95 sq cm of surface area) samples of 0.020 in. thick electrolytic cobalt strip,* sanded with 2/0 paper, were heated in air in a cyclone type furnace. The samples were weighed before and after the scaling treatment. The weight increases so observed obeyed the Pilling and Bedworth parabolic law at all temperatures studied. The parabolic scaling constants when plotted on a logarithmic scale as a function of absolute temperature are reproduced

Jan 1, 1950

By W. D. Bennett

AN interesting effect has been observed in the vacuum etching of titanium in the high temperature ß phase. Using a high vacuum annealing furnace, operating at less than 2x10-6 mm with a tubular titanium liner, it has been found possible to etch polished metallographic specimens at different temperatures between room temperature and 1000°C without contamination or increase in hardness.' Heat etching of titanium has been reported by other workers,' but the following example is of interest as an experiment under high vacuum in the absence of any contamination. Two polished specimens of iodide-processed titanium are placed at different positions along the axis of the furnace corresponding to temperatures of 900" and 800°C above and below the a-ß transus. The specimens are then raised to these temperatures and annealed simultaneously for 3 hr under the same high vacuum conditions. After cooling and withdrawing the specimens, their surfaces are photographed without further treatment. Figs. 1 and 2 show the results for the 800" and 900°C specimens, respectively. Fig. 1, corresponding to a vacuum etch in the a region, shows a typical equiaxed structure with what appears to be a slight shift of grain boundaries. Fig. 2, corresponding to a vacuum etch in the ß region, shows an equiaxed type of structure with the growth of striations in certain grains. The striations are only superficial, as a light polish and chemical etch restores the clean equiaxed structure. The vacuum etch, occurring as it does above the a-ß transus, records the state of the metal in the ß phase by means of the surface relief produced by the selective evaporation of metal from different grains; the metal itself reverts to the equiaxed a on cooling. The nature of the ß etch" (Fig. 2) suggests that, in transforming from the close-packed-hexagonal a to the body-centered-cubic ß lenticular-shaped laminae are formed parallel to common planes of the two structures. This would explain why some grains appear to be unaffected by the etch. The technique has interesting possibilities in the investigation of the mechanics of high temperature transformation. Acknowledgment This note is published by permission of the Deputy Minister, Dept. of Mines and Technical Surveys, Ottawa, Ont., Canada.

Jan 1, 1956

By J. B. Gayle

PRESENT flotation processes depend almost entirely on the buoyant properties of air bubbles to effect separations of mineral and gangue, but there is no convenient method for measuring aeration in flotation cells. Consequently the relation between aeration and cell performance has not been definitely established. In connection with a comparative study of kerosene and froth flotation processes, a modification of the displacement method' was developed in which air escaping from the cell surface is measured with a wet-test meter. The arrangement shown in Fig. 1 is satisfactory for use with a number of cells and is not critical, except that care must be taken to use connections and tubing large enough to avoid excessive back pressures. Attempts to use smaller tubing resulted in back pressures that caused displacement of the liquid level in the receiver, so that some air was lost around the sides, an occurrence readily visible when measurements were taken with only water passing through the cell. With the arrangement shown, the manometer indicated an average reading of about 1/2-in. water pressure. No loss of air was evident. Fig. 2 presents typical data for a standard Denver sub-A 24x24-in. unit cell operated at impeller speeds from 0 to 710 rpm. Each plotted point represents the average for five readings distributed over the entire cell surface. Readings were taken with only water passing through the cell. Results of this type have been used in a study of the influence of operating variables on cell performance and in detecting inherent differences in the operating characteristics of cells of various kinds. Although the information presented in Fig. 2 resulted only with water passing through the cell, the apparatus is suitable for use during flotation of coal when various reagents are present. Possible modi- fications that might prove convenient under special circumstances include varying the size and shape of the receiver and substituting a calibrated, inverted, water-filled carboy or other gas-measuring device for the wet-test meter. If such modifications result in excessive back pressures, some form of draft should be applied at the discharge end of the measuring system.

Jan 1, 1959

By T. lkeda, R. S. Montgomery. A. Hara

Actual test specimens used in the previous laboratory wear investigation,' were fully characterized and the data is given in Table 1. Most of the carbides tested could be placed into groups of similar mean grain size. When the wear rates of the members of these groups were plotted as functions of their hardnesses, a family of curves resulted; curves for groups containing data from larger grained composites were at lower wear rates at the same hardnesses. Most specimens tested showed only A-1 porosity (ASTM Standard designa-

Jan 1, 1971

By A. J. de Witte

It has been observed that upon invasion of a sand containing oil or gas and connate water by mud filtrate the hydrocarbons are more rapidly flushed by the filtrate than is the connate water.' In time, it appears that the following displacement pattern emerges: oil (or gas) is being swept ahead by connate water which, in turn, is pushed by invading fluid. Eventually the connate water may have "banked up" sufficiently to form a zone of appreciable thickness leading the invading front. In the extreme case, Fig. 1 shows how the situation will develop.' Immediately adjacent to the borehole (radius r) there is a zone, the invaded zone proper, where the connate water has been flushed out completely and which contains only residual oil or gas and mud filtrate. This zone (extending out to a radius ri) grades more or less abruptly into the next (thickness A), which contains only residual hydrocarbon and connate water. Beyond it, again more or less sharply bounded, extends the virgin formation with the original interstitial water saturation Sw. The various saturations are indicated in Fig. 1. The three zones generally will be marked by resistivity contrasts owing to their different fluid contents. As the connate water is usually more saline than the invading fluid, the second zone having a high connate water saturation forms a concentric cylindrical ring or annulus of low resistivity R, around the borehole.',' It will be referred to as the "low zone." Fig. 1 shows schematically the resistivity profile. It is clear that the phenomenon of a low zone could not occur in invaded water sands. The presence of a low zone, therefore, would be a qualitative indication of a hydrocarbon-bearing formation. Granted that the phenomenon is real, if it were pronounced enough to be detected, one might thereby have a means of locating oil or gas in the ground. This is the aim of the "displacement logging" method.' In any case, the presence of a low resistivity zone will affect the reading of electric logs. Using conventional log interpretation techniques, one must be aware of this and, if necessary, correct for it. Whether correction or detection should be the goal will depend primarily on the magnitude of the effect. The type of log which is likely to be affected most is the induction log. The current pattern of an induction log is concentric with the borehole. Any concentric ring of low resistivity, therefore, will tend to become crowded by current lines. This zone of maximum current density will obscure the relative contributions to the current conduction by other portions of a bed making, for instance, the contribution of the virgin formation less significant and thus detracting from the value or the induction log as an R, reading device. The present note is intended to ascertain the maximum possible effect of a low zone in the extreme case portrayed by Fig. 1 on three commercially available types of induction logs, sc., 5 FF 27, 5 FF 40 and 3 F 60. Type 5 FF 27 or 5 PF 40 is currently run in conjunction with a 16 in. normal and SP log. This combination is gradually replacing the regular electric survey (two normals and one lateral) as a standard log in many areas. On the basis of the scheme of Fig. 1 (sharp boundaries between successive zones), the thickness of the low zone may be computed for various amounts of infiltration measured by the ratio of invaded zone diameter to hole diameter D,/d. Referring to Fig. 1, we can set up a material balance for the water displaced from the invaded zone proper and the water present in the low zone. The volume of connate water displaced from invaded zone upon flushing by mud filtrate is: V1 = e p (r2 - r2) ? Sw . where c = bed thickness ? = porosity Su, = original connate water satu-ration. The volume of water finally accu-mulated in the low zone is: Vz = e p [(r1 -?)2 - r2 4] ? Sw . where SW1 = 1 — Sor = "water" saturation in the invaded and low zones both. The volume of water originally present in the low zone is:

Jan 1, 1958

By E. S. Machlin, W. R. Yankee

IN a previous study1 of the effect of heating com-mercial titanium in air on its subsequent friction coefficient against other metals, as well as itself, it was found that the friction coefficient markedly decreased from a value of about 0.7 to about 0.3. A tentative explanation was given that surfaces normally produced at room temperature are not contaminated sufficiently to prevent seizing or welding of the titanium to the softer mating metals. The latter tend to cleanse themselves during rubbing over the harder titanium. It was thought that the lack of a contaminant protective film on the titanium was due to the high solubility of titanium for oxygen and nitrogen and hence an inability to form a contaminant oxide or nitride. This explanation requires the ratio of the surface absorption rate to the diffusion rate to become much lower at room temperature than it is at high temperatures. In order to check the phenomenon further, commercial titanium specimens were nitrided or oxidized at 800°C for 20 hr in flows of prepurified N2 and 01, respectively, at about 1/2 in. H2O above atmospheric pressure. Friction runs were made in argon using a freshly cut copper hemisphere (cut in argon) on surfaces cut successively into the diffusion layers in the titanium (cut in argon) using the techniques described in a previous publication.' DPH values (100 gram load) were made as a function of depth into the diffusion layer using a Tukon tester. Also, micrographs were taken at separate cross sections to indicate the diffusion layers. The results obtained are presented in Figs. 1 and 2, which show the "static" friction coefficient vs hardness for the nitrided and oxidized specimens, respectively. A separate measurement of the friction coefficient of clean copper vs iodide titanium also was made. From results reported in the literature' giving the oxygen and nitrogen contents as functions of the hardness, cross plots were made showing the friction coefficients as functions of the amount of interstitial solute. These plots are given in Figs. 3 and 4. From micrographs of the diffusion layers and the phase diagrams, it was deduced that the data in Figs. l through 4 correspond to the single phase a region. The points observed on the compound regions have been excluded from the figures. It is apparent that nitrogen or oxygen in solution in the a titanium markedly affects the friction coefficient against a softer mating metal. Discussion of Results These results are extremely interesting from both a practical and theoretical viewpoint. The theoretical implications will be discussed first. According to Bowden, the friction coefficient should be given to a good approximation by the relation where a is the fraction of contact area that has welded, a is the shear strength of weaker component, and H is the hardness of softer component. Using this relation alone, it is difficult to understand the results because none of the terms should be affected by a variation in the oxygen or nitrogen content of the harder and stronger metal, titanium. Even if the ratio of S/H for titanium is used in Eq. 1, the ratio has been shown to be independent of oxy-gen or nitrogen content.' If a more rigorous equa-tion is used combining Eq. 1 and a result given pre-viously" for the case where welding is absent, then the relation obtained is µ = a S/H + (1-a) a W aß/H where a is constant and Waß is the work of adhesion between the two metals comprising the friction couple. This relation states that if a is less than 1/2 or so, variation in the work of adhesion Waß between copper and the titanium should affect the friction coefficient markedly. It is reasonable to expect that the work of adhesion will depend on the oxygen or nitrogen content of the titanium. Available data4 show that clean metals and oxides have much lower works of adhesion than the same metals against the

Jan 1, 1955

By C. C. Clark, J. R. Mihalisin, K. G. Carroll

OBSERVATIONS made in this paper were for the purpose of exploring the anomalous behavior of a particular heat of Type 310 stainless steel. The alloy, of nominal composition 25 pct Cr, 20 pct Ni, was observed to form a grain boundary precipitate at exceedingly high rates. A sheet of this metal, 0.060 in. thick, after cooling in air from 2000°F showed the presence of the grain boundary material, whereas none was found after water-quenching from the same temperature. X-ray techniques were applied without success and, in view of the ex-

Jan 1, 1958

By Emma Smith, R. W. Guard

INVESTIGATION of the nickel-rich end of the Ni-Zr system has been prompted by an interest in the effect of small amounts of zirconium in high temperature alloys. Hansenl presents a hypothetical diagram based on the work of Allibone and Sykes,2 who investigated the system up to 44 atomic pct (55 wt pct) Zr by metallographic means. Since the alloys they used were melted in Alundum crucibles they were probably contaminated with appreciable amounts of aluminum. They found two compounds which they called Ni3Zr and Ni4Zr. Hayes, Roberson, and Paasche3 identified a compound NiZr (50 atomic pct Zr) with a melting point of 1470°C. Such a high melting compound is not compatible with the observation of Allibone and Sykes2 hat a eutectic is formed near 50 atomic pct. The alloys containing up to 60 atomic pct Zr were melted in an inert atmosphere in a noncon-sumable electrode arc furnace. Analysis of most of the alloys indicated negligible loss of zirconium. The alloys were examined in the as-cast condition and also after annealing for 6 hr at 1175°C (or 1000°C for the alloys between 25 atomic pct and 45 atomic pct). Metallographic specimens were etched using mixtures of HF-HNO3-H2O or acetic acid- HN0,-H,O. The specimens for the solidus determinations were wrapped in molybdenum foil and sealed in evacuated quartz tubes. Specimens to be heated above 1300°C were enclosed in zirconia tubes inside the quartz to prevent reaction with molybdenum. The solidus determinations were carried out by heating the specimens at various temperatures for a few minutes and then examining the interior structure metallographically to determine at what temperature melting began. The 150 mesh powders for X-ray analysis were annealed in an evacuated quartz tube for 2 hr at 540 °C to remove cold work. Alloys with 22.5 atomic pct Zr and 44.4 atomic pct Zr required annealing at 700 °C for complete recovery. The grains of both of these alloys were soft and rather ductile. Both Debye-Scherrer film patterns and spectrometer traces were made for each alloy using filtered copper radiation. Results and Discussion The presence of a substantial amount of second phase in an alloy with 1.1 atomic pct Zr indicated that the solid solubility of zirconium in nickel is somewhat less than 1 atomic pct. The a0 value for the face-centered-cubic matrix in the same alloy was equal to that of pure nickel, so that the solubility must be very small. This observation is as expected from the large atom size difference. Annealed specimens of alloys with 15.0, 16.7, 22.5, 28.6, 44.4, and 59.4 atomic pct Zr showed. micro-structures that were essentially single phase. When

Jan 1, 1958

By M. J. Sinnott, H. B. Probst

AN extensive survey of the factors which affect austenite grain growth has already been made.' These factors are temperature, time at temperature, rate of heating, initial grain size, hot-working, alloy content, ofheating,initialand rate of cooling from the liquidus-solidus temperature. In the present work, a vacuum-melted temperature.electrolytic iron was used and the variables studies were temperature, time at temperature, and prior ferrite grain size. Other factors were maintained constant. The iron used in this study was vacuum-melted electrolytic iron of nominal composition of impurities of 0.07 wt pct. It was supplied as a ½ in. round cold-drawn bar. This iron was tested in three conditions: as-received, annealed 6 hr at 1200°F, and annealed 6 hr at 1600°F. Samples were ? in. disks cut from the bar. The prior anneals were carried out in vacuum and the isothermal treatments were carried out in vacuum-sealed Vycor tubing. The thermal etch technique was employed to determine the austenite grain size. Prior to sealing the test specimens, one surface of the sample was polished metallographically. This surface, after heating, was examined to determine the austenite grain size, since the austenite boundaries are revealed by thermal etching. This is essentially the only technique available for measuring the austenite grain size of low carbon steels or pure irons without altering the composition. It has been shown to yield results that are in agreement with other methods used for determining austenite grain sizes.' The specimen size was quite large compared to the grain size measured, so inhibition of growth due to size effects is probably negligible. After vacuum sealing, each sample was placed into a furnace at temperature and at the completion of the run was quenched into a mercury bath. The growth temperatures used were 1700°, 1800°, 1900°, and 2000°F controlled to -~10"F. Growth times were varied from 10 to 240 hr. The long times were used in order to eliminate the nucleation and growth effects occurring during the initial transformation. Time was measured from the introduction of the capsule into the hot furnace to the time of quench. Grain-size measurements were made with the use of a grain-size eyepiece of a microscope. By determining the number of grains per square millimeter at X100 and taking the square root of the reciprocal of this number, the average linear dimension of the grains was determined. Figs. 1 and 2 are plots of these data as a function of time and temperature for the various conditions investigated. The variation of D, the linear dimension of the grains, was assumed to follow the equation3 D = A tn. The curves of Fig. 1 were obtained from the data by the use of the least-squares method of analysis. Fig. 1 is for the growth of the as-received stock and Fig. 2 is for growth after prior treatments. Differentiating the foregoing equation gives an expression for the rate of growth dD/dt = G = nAtn-1 = nD/t. Both D and G as functions of t are given in Table I. It should be noted that G is a function of time; the growth rate is rapid at early stages and decreases with increasing time. Since increasing temperature increases the growth rate, it has been common practice to use the empirical relationship G = Go e-Q/RT to relate temperature to growth rate. The growth rate customarily has been taken at constant values of D on the basis that the rate of growth is related to the boundary surface tension and this is measured by the curvature of the boundary. At constant D values, the growth rate is a function of time and temperature. The growth rate can be related however to temperature at constant time, and this has the advantage that under these conditions the growth rate is a function only of temperature. Obviously the Q values, activation energies, obtained for each assumption will not be the same and the question of which is the more correct is a moot one, since the assumed exponential relationship in either case has no particular theoretical significance. By plotting G, at constant grain size, vs 1/T, the activation energy over the temperature range of 1800" to 2000°F is found to vary from 30,000 cal per mol at the smaller grain sizes to 50,000 cal per mol at the larger grain sizes. The 1700°F data do not correlate with the data at higher temperatures. The activation energies for the 1200" and 1600°F prior annealed materials were calculated as 50,000 and 62,000 cal per mol, respectively, using the reciprocal time to a given grain size as a measure of the growth rate. Plotting G, at constant times, vs 1/T yields an activation energy of 12,300 cal per mol for the tem-

Jan 1, 1956

By L. D. Jaffe, D. C. Buffum, F. L. Carr, M. Goldman

IN an earlier investigation by Jaffe and Buffum,' a time-temperature transformation diagram for isothermal temper embrittlement of an SAE 3140 steel showed some suggestion of two overlapping "noses." These authors mentioned, as a possible explanation, that their specimens embrittled at 525°C were slightly softer than expected, perhaps because of some slight unknown variation in tempering treatment. Because a double nose would be of considerable importance in elucidating the mechanism of the temper brittleness reaction,' it appeared worthwhile to redetermine the portion of the diagram near 525°C. The same heat of SAE 3140 steel was used as in the previous work, and experimental procedures were the same.' Blanks were austenitized 1 hr at 900°C, water-quenched, tempered 1 hr at 675°C, and water-quenched. Groups were then embrittled isothermally at the temperatures and for the times shown in Table I, and water-quenched. For comparison, one series was tempered various times at 675°C, water-quenched, and given no further treatment. V-notched Charpy impact specimens machined from the blanks were tested over a range of temperatures. Energy absorption and percentage of fibrous fracture were plotted vs test temperature for each group, utilizing both the new data and those for the same treatment obtained in the earlier work,' except that the earlier data at 525°C were discarded.* are given in Table I. After a small correction (-0.28°C/BHN) for the variation of hardness" from Rockwell C24, the transition temperatures were plotted against embrittling time for each embrittling temperature, using the data of Table I for the range of treatments covered there, and using previous dataL for other treatments. By cross-plotting, Fig. 1 was finally obtained. It appears from this figure that if there are two noses to the time-temperature-transformation diagram for isothermal temper embrittlement of SAE 3140 steel, the upper nose must lie above 575°C. As interest had been primarily in the region near 525°C, a thorough study of embrittlement above 575°C was not made. Fig. 1 shows that such a study is needed, especially since preliminary measurements at the tempering temperature of 675°C reveal much more rapid embrittlement than would be expected from the data at lower temperatures. The transition temperature increased, approximately, as the logarithm of the tempering time (about 81/2°C, after correction for hardness, for each doubling of the time).& Since three tensile specimens tempered 240 hr at 675°C and water-quenched had an average yield-tensile ratio of 0.844, with 70.7 pct reduction of area, it is assumed1 that the tempering temperature is not above Ae1. The omission of a quench-and-reheat operation below Ae1, for the groups tempered at 675°C, as compared to groups embrittled at lower temperatures, should not affect results." More detailed investigation of embrittlement above 575°C is now under way.

Jan 1, 1954

By T. S. Liu, E. A. Peretti

A LITERATURE survey of the In-Sb system, which was made before a study of the binary diagram was undertaken, revealed that the intermediate phase InSb existed and that its crystal structure is face-centered cubic of the zinc blend type. In 1929 Goldschmidt' determined the lattice constant for InSb and obtained a value of 6.452A. A more recent value, 6.461A was reported by Iandelli2 in 1941. The purity of the indium used was not very high, however. The indium used by Iandelli contained 2 pct Ge. In view of this fact it was decided to remeasure the lattice parameter using high-purity indium (99.97+ pct In) and spectrographically standardized antimony of the highest purity available (99.91 pct Sb). Specimens corresponding to the InSb composition were made by carefully weighing the ingredients on an analytical balance and heating them in vacuo in a quartz tube at 700°C. As melting losses were negligible, the alloys were not analyzed chemically, but the intended analysis was taken as being accu- rate. Powder samples were prepared by either filing or crushing of the brittle alloy, followed by screening at 325 mesh. The powders were then annealed at 400°C in evacuated capsules for about an hour. X-ray photograms were taken with a Debye and a back-reflection focusing camera, using characteristic cobalt K. radiation. Lattice parameters were calculated by Cohen's³ method. The results of a Debye pattern and of a back-reflection pattern at 25°C are given in Tables I and 11. The lattice parameter at 25 °C is 6.47602A.

Jan 1, 1952

By L. L. Hirsch, W. M. Parris, P. D. Frost

IT has been established that titanium alloys con- taining sufficient amounts of ß-stabilizing elements, such as iron, chromium, or manganese, can be age hardened.' Adenstedt, Pequignot, and Raymer2 showed that age hardening occurred in a Ti-15 pct V alloy at room temperature after it was quenched from a temperature in the ß-phase region. During a recent investigation of the heat-treating characteristics of binary Ti-Mn and Ti-Cr alloys at Battelle, further evidence of low temperature aging was found. Appreciable hardening occurred in some of these alloys at aging temperatures as low as 212 OF (100oC). Indeed, the brief heating period employed for mounting metallographic specimens in bakelite resulted in hardening. It was felt that further information on this low temperature aging phenomenon would be of general interest in view of the large amount of work currently being done on titanium alloys. The alloys examined in this investigation were arc-melted in a water-cooled copper crucible as 3/4-lb ingots. Du Pont Process A sponge was used as a base material. The arc-melted ingots were forged at 1700" to 1800°F (927" to 982°C) and hot rolled at 1450°F (788°C) to 14 gage (0.064 in.) sheet. Small specimens were heated in argon at 1742°F (950°C) for 30 min and quenched in various liquid media or air cooled. They were then mounted in a room-temperature-setting resin and given a metallographic

Jan 1, 1954

By D. H. Polonis, J. G. Parr

IN a recent paper on titanium-rich Ti-Fe alloys,' the hardness of quenched powder specimens was given, together with the amounts of martensitic a they contained. The values disagreed in two respects with previous determinations that Worner² made on lump specimens: 1—Some martensitic a' is formed in powder specimens of alloys containing up to 12 atomic pct Fe. On the other hand, Worner reports that, although his results on quenched lumps are not conclusive, alloys containing more than about 4 pct Fe "appear to consist of retained ß'' although they "may not be strictly unaltered solid solutions." 2—According to Worner, quenched lump specimens show a maximum hardness at a composition about 3.5 pct Fe. Earlier -experiments by the authors showed a maximum hardness of martensitic alloys in powder form at 10 pct Fe. Accordingly, more experiments have been conducted in order to confirm these differences and to determine with more accuracy the minimum composition of quenched lump specimens that consist entirely of retained ß. Lump samples?x?x1/20 in. were quenched from 1000°C in helium and, on the whole, showed similar hardness values to the specimens that Worner water-quenched from 950°C. The hardness curve the authors obtained shows a more pronounced peak, but in other respects it is very similar to Worner's curve (Fig. 1). While it was possible to make fairly reliable measurements of the percentage of martensitic a in quenched powders (by Geiger spectrometer), the solid samples allowed only approximate estimates to be made from visual observation of line intensities on film. Retained ? begins to appear in quenched lumps containing 0.2 pct Fe; at 2.2 pct Fe there are approximately equal quantities of ß and a'; in a 4.7 pct Fe alloy only a trace of a' appears. These results are represented as a curve on Fig. 1. The hardness and constitution of powder samples were verified by repeating both types of quenching experiment described in the earlier paper.' Identical results were obtained by the two methods, and they agreed with the figures published previously (Fig. 1). The appearance of the martensitic structure under the microscope is shown in Figs. 2 and 3. The mar-tensite needles are coarser in the lump; otherwise the structures are similar. Three conclusions are made at this juncture: 1— The properties of quenched lumps reported by Worner are confirmed, on the whole. The differences between them and the properties of powders are probably due to an effect of specimen size. Differences of behavior in various sizes of specimen have been observed previously in martensite reactions in Fe-Mn alloys3 and in Cu-A1 alloys. 2—The maximum hardness of both quenched lumps and quenched powders appears at a composition where martensite formation has all but ceased. In this respect, powders and lumps behave similarly. 3—Martensitic a' has essentially the same microscopic appearance both in powders and lumps. Acknowledgment This work is part of a program sponsored by the Defegce Research Board of Canada, Project No. 425.

Jan 1, 1956

By G. D. Hobson, I. Mrosovsky

Material balance relationships for expansion-type reservoirs above the bubble point have been discussed in recent years by H. N. Hall and M. F. Hawkins, Jr. The former drew attention to the effects of rock compressibility, and the latter to the effects of the compressibility of any interstitial water. Hawkins derived a general relationship which took note of oil, water and rock compressibilities, and then by assuming first, the absence of interstitial water and secondly, no effects due to interstitial water or rock compressibility, obtained relationships symbolically of the same form as those previously derived directly by Hall. Hawkins defined his terms clearly, and noted an approximation, due to the omission of a term involving the product of two compressibilities, made in reaching his final equation. There is, however, possible ambiguity with regard to Hall's written definition of the oil production term. One interpretation would mean that the produced oil and initial oil-in-place, if both required as stock tank barrels, would be obtained by using different formation volume factors in the conversion in order to satisfy the conditions implicit in Hall's formulae. The other would involve an approximation which, if used, was not stated. In both Hall's and Hawkins' studies the effect of the approximation or inconsistency is very slight. Nevertheless, it may be justifiable to give a somewhat more consistent derivation, since the complexity of the final relationships is not thereby greatly increased. Let the initial pore volume be Vi and the volume after a pressure drop of Ap, V. Then V = V, (1 - ?p . cr), where c, is the rock compressibility expressed as change in pore volume/unit pore volume/unit pressure change, and is defined with the initial pressure as base. Suppose that the interstitial water saturation under the initial conditions is Sw. The total volume of interstitial water under those conditions will be ViSw. If the volume of the water after dropping the pressure by ?p is V., this will be given by Vw = ViSw (1 + ?p.c,w), the compressibility of the water being c, which is ex- pressed using the volume at the initial pressure as base. If Boi is the reservoir volume factor of the crude oil for the initial reservoir conditions, and B. the equivalent factor when the pressure has dropped by ?p, with N the initial amount of oil expressed as stock tank barrels. N.Boi = Vi - V,Sw If n is the amount of oil produced (in terms of stock tank barrels) as a result of the pressure drop ?p, the remaining subsurface volume of oil at the lower pressure will be Bw, and Bc can be related to the compressibi:ity of oil and the pressure drop, and if c, is the compressibility factor based on the initial conditions, The relationship derived by Hawkins was case to a standard at the lower pressure, while a term (Ap) had been dropped, because of its smallness arising from its dependence on the products of two compressibility factors. When there is no interstitial water and rock compressibility can be ignored, Eq. 1 reduces to:

Jan 1, 1957