Search Documents

By D. H. Woodard, G. W. Geil, W. D. Jenkins

In a series of papers, the authors and their associates have shown the influence of four important factors on the technical cohesion limit.3-l6 By "technical cohesion limitv is meant the technically determin-able resistance to fracture. The four importarlt factors are the stress system, plastic deformation, temperature, and the strain rate, The influence of any one of these factors may be represented by a curve of cohesion limits, although only one point on the curve may represent actual fracture. As the term "fracture stress" has come into general use with the same significance as technical cohesion limit, the authors are here using the shorter term with the understanding that it does not always refer to actual fracture. The previously mentioned series of papers contains ample evidence that plastic extension increases the fracture stress of any polycrystalline metal, In a recent paper of the series,l5 moreover, it has been shown that the effect of plastic deformation on the fracture stress cannot be attributed to reorientation of internal flaws but is a work-strengthening effect, similar to the effect on the flow stress, $ Plastic deformation of steels, however, may not only cause work hardening, but may result in additional hardness because of strain aging. Some evidence that strain aging affects not only the flow stress but also the fracture stress, has been presented in a recent paper.14 In that paper, a study is made of the influence Of Plastic deformstion at room temperature on the fracture stress of ferritic steels at -188°C. The study is based on the results Of two-stage tests. In the first stage each specimen was extended a predetermined amount at room temperature; in the second stage the specimen was tested to fracture in liquid air. In most Of these tests the specimen remained in the testing machine during the interval of about one hour between the first and second stages, it was kept under a tensile load of about 500 lb and maintained at liquid air temperature. The effects of various amounts of .plastic deformation on the fracture stress were thus determined and curves were obtained to represent the influence of plastic deformation at room temperature on the stress at brittle fracture in liquid air.14 Two specimens Of ingot iron, however, were removed from the testing machine at the end of the first stage and several days elapsed before the specimens were tested to fracture in liquid air, As shown in Fig 8 of that paper," the points representing fracture of these specimens are far above the curve representing the results of two-stage tests without a rest interval between the stages. The evidence thus indicated that the fracture stresses of these two specimens were increased not

Jan 1, 1949

By R. W. Hyde, S. E. Eaton, B. S. Old

One of the most important problems facing the steel industry at the present time is that of maintaining at a minimum the sulphur content of many grades of steel where sulphur is known to have harmful effects On both quality and yield. The present outlook in this struggle to reduce the amount of sulphur in steel is rather gloomy since the sulphur content of the raw materials available in the United States is gradually rising, with little relief in sight. A flow sheet of sulphur in the steel process is shown in Fig I. The values used here were obtained by averaging coke oven and blast furnace data at No. 5 furnace in Cleveland and are considered typical for high top pressure practice. The sulphur evolved in the blast furnace top gas was obtained by difference and since the amount is relatively small, little significance should be attached to this particular value. From Fig I it is apparent, however, that coals used in the production of metallurgical coke represent by far the largest single source of sulphur in steel-making. As an indication of the decrease in quality of these coals as regards sulphur content the weighed average annual sulphur analysis of the coals supplied to two different steel plants, since 1925, is plotted in Fig 2} The sharp rise since 1940 is strikingly apparent. Another indication of this trend is found in estimates of coking coal reserves. Thus in the case of Pittsburgh seam coals available in Allegheny, Fayette, Greene, Washington, and Westmoreland Counties in Pa. as of Jan. 1937 only 25 pct of the coal reserves would contain less than 1.50 pct sulphur after washing.2.3 Many of these lower sulphur reserves have been seriously depleted in recent years because of accelerated war demands. still another demonstration of the increasing sulphur content of coals on a country-wide basis is found in a report by the Bureau of Mines4 which shows a steady increase in the sulphur content of coke made in the United States. It should be noted that this increase continues in spite of the wider use of coal washing in recent years. Sulphur occurs in coal in three forms, pyritic (FeS2), organic sulphur, and a small amount of sulphate sulphur (CaS04). The sulphur is subsequently found in the coke in combination with carbonaceous material of is composition and as calcium and iron sulphides' In considering the general problem of sulphur elimination, the Republic Steel Corp. was interested in determining which, if either, of the two major forms of sulphur in coal contributes the greater portion of sulphur to the coke. Since the ratio of pyritic to organic sulphur varies widely in different coals, it was thought that if it could be shown definitely that the major part of the sulphur occurring in coke originates from one of the two ""jar forms of sulphur in the coal, it would then be possible to select with confidence coal which, although high .in total sulphur, would yield coke of fairly low sulphur content. Using the new

Jan 1, 1949

By G. L. Kehl, Emilo Jaraiz F., J. S. Brett

The transformation of eutectoid beryllium bronze has been quantitatively investigated at 550° and 592°C, the temperature region where the products of p decomposition are a fine lamellar eutectoid and a coarse lamellar aggregate which replaces the fine eutectoid with prolonged annealing. Metallographic evidence indicates that the coarse lamellar aggregate forms preferentially in the most densely lamellated regions of the fine eutectoid and favors the hypothesis that the development of the coarse lamellar aggregate in the fine eutectoid is due to the occurence of a cellular segregation reaction. WHEN eutectoid beryllium bronze (Cu-94 pct, Be-6 pct) is isothermally transformed at high temperatures in the and y region after quenching from the fl field, the principal initial decomposition process is G one of growth of approximately spherical nodules composed of a number of lamellar colonies of the a and y phases. After the initial reaction has proceeded substantially or completely, depending on the transformation temperature, a second reaction occurs wherein the fine lamellar eutectoid is replaced by a coarser lamellar aggregate composed of the same oand yphases.' This reaction is not simply a spheroidization of the initial fine eutectoid but the nucleation and growth of a coarser, structurally different, and more stable distribution of the @and yphases. Fig. 1 illustrates the replacement of the initial fine lamellar structure by a coarse lamellar aggregate. The a phase is the dark mottled constituent; the light constituent the y phase. The phase diagram for the copper-rich end of the Cu-Be system2 is shown in Fig. 2. The a phase has the face-centered cubic structure characteristic of copper-rich solid solutions. The B phase is a disordered body-centered cubic structure, while the y phase is body-centered cubic, ordered (CsC1, B2 type). A. viswanathanS investigating Debye patterns of Cu-6 pct Be discovered that a transitory phase appeared on very rapid quenching from the /3 region which was absent when the quenching velocity was moderate. The phase could be assigned a body-centtred cubic structure with a lattice parameter of 2.876. Tempering at 100°C for 2 hr after quenching resulted in complete disappearance of the phase. The work of Fillnow and Mack' is the only previously published study of the isothermal transformation of eutectoid beryllium bronze. Their study elucidated the nature of decomposition at various subcritical temperatures in an exploratory fashion. The purpose of the research reported in this paper was principally to obtain quantitative information on the isothermal decomposition of the fl phase and to clarify the relationships existing between the fine lamellar eutectoid and the coarse lamellar aggregate. EXPERIMENTAL PROCEDURE The material used in this investigation was an almost eutectoid Cu-Be alloy bar furnished by D.J. Mack. The alloy had been melted under a protective carbon cover and centrigugally cast into a graphite mold preheated to 490 "C. Segregation was minimized by violent agitation during melting and pouring. The analysis of the casting was 93.85 pct Cu, 5.60 pet Be with traces of Si, Mg, Mn, Co, and Al. To determine the minimum soaking time necessary to insure specimen homogeneity, a number of specimens were held in the P region at 788OC for times fron 1 min to 12 hr and isothermally transformed at 550°C. Specimens betatized less than 12 min exhibited partial transformation at times shorter

Jan 1, 1961

By L. W. Pommier, F. B. Brien

In order to determine the ability of cyclones with heavy liquids to preconcentrate ores, a study of the variables which govern the behavior of the solids in the apparatus has been made. To assess the test results, the cyclone products have been examined by sink and float analysis at various specific gravities and the results examined by the Tromp type curves. This analysis has allowed the development of the SG50 unit — specific gravity of the particles, 50% of which goes to overflow and 50% to underflow — rather than the d50 (size) unit used in classification. The results of tests using the closed circuit laboratory cyclone assembly with dry feed samples show the validity of the equation: Results indicate that 83% of the feed may be rejected containing 0.1% Sn. Recoveries in excess of 80% are obtained at a grade of 1.7% Sn from feed samples assaying 0.4% Sn. As water-wetted particles may comprise the plant feed, separate tests were performed to examine the effect of water in the feed material. Results indicate that the entrained water divides into two fractions, one of which is defined as free water and is rejected to the cyclone overflow, the other fraction forms an adhering envelope of water on the ore particles, altering their effective size and density. The presence of the envelope of water on the particles increases the ease with which the heavy liquid can be recovered but decreases the concentration efficiency. It is proposed that the detrimental effect of water in the cyclone may be minimized by using a higher feed pressure cx by the use of surfactants. A higher feed pressure, by increasing shearing forces, would reduce the thickness of the water envelope on the particles. The surfactants, by adsorbing on the particles and producing hydrophobic surfaces, would displace the water envelope and allow the heavy liquids to wet the modified surfaces. In both cases a larger proportion of the entrained water would report to the cyclone overflow as free water, thus minimizing the detrimental effect of water on the concentration efficiency. Additional testing is required on the behavior of water in the system; to determine the optimum particle size for the preconcentration step; to determine final particle size for optimum liberation and subsequent treatment of the products in present plants and to determine the efficiency with which the heavy liquids may be recovered from the products. For many years tin mining in Bolivia has been very important to the economy of the country. In recent years the tin content of the ore reserves has been steadily declining. An improvement in concentration efficiency is highly desirable in order to increase productivity from the present low-grade mine ores and to allow retreatment of the many millions of tons of tailings from past operations, some of which equal the present ores in tin content. A number of investigators have been or at present are working on various phases or approaches to the problem. Gravity concentration is used in Boliva today as it has been in the past to recover tin minerals from the mine ores. An objective of the work reported in this paper was to examine the possibility of treating present low-grade ores and sink-float tailings products, at a relatively coarse grind, in order to produce a higher grade feed material to the present gravity concentration plants. The possibility of using heavy liquids in cyclones to treat large tonnages with a minimum of outlay presented itself and is of particular interest since heavy liquids are now available with physical properties which lend themselves to simple recovery methods. Tests were carried out using a closed circuit laboratory cyclone assembly. It was desirable to determine the value of using the SG50 concept and Tromp curves for analyzing the cyclone concentration test results. The effects of entrained water on the behavior of ore particles in the cyclone were examined. THEORY AND TECHNICAL CONSIDERATIONS Three types of heavy liquids are available: aqueous solutions of very soluble salts, salts or mixtures of

Jan 1, 1964

By G. W. Thomas

A simplified mathematical model of underground conduction heating in a system of limited permeability is presented. The model applies to underground retorting of oil shale, or to reservoirs containing extremely heavy oils. We assume that heat is introduced at a constant rate into a horizontal fracture which communicates between wells. The radial temperature distribution along the fractured surface is approximated by a step-function. Heat transfer away from the fracture is assumed to be by vertical conduction, and all convection effects are neglected. The model also takes into account the possible temperature dependence of thermal conductivity. A general expression for calculating the growth of the step-function temperature distribution with time is derived. The use of this expression and solutions to the one-dimensional heat equation make it possible to construct isotherms. Expressions for calculating .oil recovery, well spacing and heat efficiency are also given. An example calculation is presented for the conduction heating process in oil shale. Finally, the effect of the heat transfer coefficient between the gas and the fracture boundaries is investigated. INTRODUCTION Thermal recovery processes of oil recovery fall into the four general areas of hot fluid injection, forward combustion, reverse combustion and conduction heating. The first three of* these processes have been rather extensively studied in the past decade from both the experimental and theoretical points of view. As a result, it is possible to make reasonable engineering predictions and analyses of these processes. Little attention has been devoted, however, to the conduction heating process other than to note its possible utility.I To a certain extent, conduction heating cannot be divorced from the other regimes cited above, insofar as these provide the source of heat energy. In the conduction heating process, heat is introduced (either by combustion in the forward or reverse mode or by hot fluid injection) into a small fraction of the total reservoir thickness. This fraction may be either a streak of high permeability or an interwell fracture. Heat penetrates by conduction into the adjacent, less permeable regions of the oil-bearing rock, where the direction of conduction is essentially perpendicular to the streak or fracture. The heated product then drains by gravity or is gas driven to production wells. Conduction heating is probably most applicable to systems containing immobile bitumens such as tar sands and oil shale deposits and perhaps to low-permeability reservoirs containing highly viscous crudes. The mechanism also acts in combination with other thermal processes where fingering or overriding of a bed occurs. It seems probable that in at least one field test, conduction heating of this type was very influential.2 In this presentation we give a first approximation to some of the quantitative aspects of conduction heating. Marx and Langenheim3 treated a similar problem where they focused attention upon the injection interval, which spanned the total reservoir thickness. In their model, conduction heat losses to the bounding media imposed a practical limit on the calculated heated area. In the present study, however, we shall confine the injection interval to a small fraction of the reservoir thickness and assume it has no heat capacity. We therefore direct our attention to regions outside the injection interval into which the conduction of heat is beneficial. In particular, we will endeavor to locate specific isotherms in the media bounding the injection interval. Furthermore, we will construct our model to allow the thermal conductivity to vary arbitrarily with temperature. Thus the model will be applicable to underground retorting of shale where variations in thermal conductivity may be important. THEORY We consider the rock matrix to have no or, at best, a very limited permeability; consequently a

Jan 1, 1965

By B. W. Levinger

THE system Ti-CI- has been studied by several investigators." ' Though titanium and chromium are completely miscible at high temperatures, an intermediate phase of the approximate atomic proportions TiCr, has been identified at various compositions between 60 and 68 wt pct Cr. The structure of this phase was established as the cubic C15 (MgCu2) type with 24 atoms per unit cell and a lattice parameter 6.929 kX.2 During a study of the ternary system, Ti-Mo-Cr,5 the existence of a high temperature modification of the TiCr, phase was discovered. In a series of alloys annealed at 1300°C and containing from 60 to 70 wt pct Cr and 5 pct Mo, a new phase was present which did not correspond to either ß solid solution or TiCr,. By trial, the new phase was identified to be of the hexagonal C14 (MgZn2) structure with 12 atoms per unit cell. Further work showed that the hexagonal structure was. in fact, also present in binary Ti-Cr alloys quenched from high temperatures. To make the identification more certain, the intensities of the various reflections were calculated assuming a composition TiCr2 and C14 structure. Table I lists reflection planes, interplanar distances, calculated intensities, and a comparison of these values with the lines of a binary Ti-Cr alloy (66 pct Cr) annealed for 20 min at 1385°C and water quenched. Since high angle lines were more diffuse, the correspondence between visually observed and calculated intensities was not as good at these angles. Lattice parameters were computed to be a = 4.922 kX, c = 7.945, c/a :- 1.614. The corresponding interatomic distances are: titanium to titanium distance, 3.00:, kx, chromium to chromium distance, 2.43, kX, titanium to chromium distance, 2.88,, kX. These compare to 3.00, 2.45, and 2.87 kX in the cubic phase, respectively.' The atomic positions in the C14 structure are as follows: Titanium 1/3, 2/3, Z; 1/3, 2/3, 1/2: 2/3, 1/3, i; 2/3, 1/3, (1/2-z); Chromium 0, 0, 0: 0, 0, 1/2; 2x, 2, 1/4; 232, x, %; X, 22, 1/4; X, 2x, 3/4; x. x, 1/4; x, x, 3/4; The values of x and z were calculated for the alloy mentioned and found as follows: z — 0.061, x = 0.165. The values assumed in the intensity calculations were those for an ideal close-packed lattice or z = 1/16, x = 1/6. Work in the Ti-Mo-Cr system indicates that y2 the hexagonal modification of TiCr2, is the stable phase above 1300°C, while y1 with the cubic structure is stable below 1000°C. X-ray examination of a number of alloys annealed at intermediate temperatures showed that the hexagonal and cubic phases can coexist. Alloys quenched from 1000°C and annealed at 1100" and 1200°C showed small amounts of y2 at the higher temperature. A considerable amount of y, seemed present, however, in cast alloys annealed at 1100°C. It further appears that there is a small but definite solubility range for both phases in the ternary system. Acknowledgment The results of this paper represent work carried out in conjunction with a project sponsored by Watertown Arsenal, Contract No. DA-11-022-ORD-272. References 'M. K. McQuillan: Journal Inst. Metals (1951) 79, p. 379. 2 P. Duwez and J. L. Taylor: Trans. ASM (1952) 44, p. 495. W. J. Van Thyne,. H.. D. Kessler, and M. Hansen: Trans. ASM (1952) 44, p. 974. 1 F. B. Cuff, N. J. Grant, and C. F. Floe: Trans. AIME (1952) 194, p. 848; Journal of Metals (August 1952). Unpublished Research, Armour Research Foundation. "H. J. Wallbaum: Ztsch. Krist. 103, p. 391.

Jan 1, 1954

By E. N. Aqua, C. M. Owens

The microstructure in cold-rolled Zircaloy-4 sheet has been characterized by transmission electron microscopy and X-ray diffraction line-broadening techniques. The rate of work hardening is measured in terms of microstrains, particle sizes, micro hardness, the tensile yield strength. The tensile yield strength of Prestrained (less than 10 pet) Zircaloy-4 sheet is found to be a linear function of the square root of the dislocation density, us predicted by work-hardening theories. ZIRCALOY is one of the most important structural alloys for nuclear reactor applications.''2 In order to understand the effects of the nuclear environment on Zircaloy structural members, the preirradiation microstructure must be characterized. This report will describe the substructure existing in cold-rolled Zircaloy-4 sheet as observed by electron microscopy and by X-ray diffractometry. I) EXPERIMENTAL PROCEDURE Commercial Zircaloy-4 strip was cold-rolled and annealed in argon at 760°C for 1 hr. The annealed sheet was again cold-rolled to a maximum of 50 pct reduction in thickness with segments of the strip being retained after 2.5, 5, 10, and 20 pct reductions. One-inch-square portions were chemically polished (50 parts HNO3, 50 parts H2O, 10 parts HF) and mounted onto lucite blocks using double-sided tape to prevent bending during handling. Thin foils suitable for transmission electron microscopy were prepared from each of the sheet samples. The sheets were mechanically polished through 600-grit paper to a 3 to 4 mil thickness and were chemically polished by immersion for 1 to 2 min in a solution containing: 50 parts HNO3, 50 parts H2O, and 10 parts HF, and rinsed in distilled water. The final metallographic procedure was to electropolish for less than 10 sec with an electrolyte of 10 parts perchloric acid and 90 parts glacial acetic acid, at 20 v, at room temperature using stainless-steel point cathodes and a vigorous stirring action. These foils were examined in a Philips EM 200 electron microscope using a 100-kv accelerating voltage.

Jan 1, 1968

By N. H. Fisher

GEOGRAPHICAL DISTRIBUTION THE most important known deposits in Australia of what are commonly referred to as the beach-sand minerals are along the most easterly part of the Australian coast, between Southport, 17 miles north of the Queensland-New South Wales border, and Ballina, 50 miles south of the border, and most of the production has come from this area (Fig 1). Smaller deposits are known to occur farther to the south at intervals for several hundred miles, and beaches have been worked at Yamba, 96 miles south and Woolgoolga, 150 miles south of the border and at Swansea, 60 miles north of Sydney. Important deposits have been shown to exist on north Stradbroke Island and others have been located farther to the north, as far as Tin Can Bay at the south end of Frazer Island. (Fig 2). Table I gives the list of operators with the location of their deposits and workings, their approximate maximum monthly capacity expressed in tons of heavy mineral produced by their concentrating plants before separation into individual mineral concentrates, their methods of mining and concentrating the minerals, and the products obtained. This table applies as of June 1948, but all the operators are remodeling or improving their plants and plant practices will be modified accordingly. In addition, active boring campaigns have been carried out by Alluvial Gold, Ltd., of Sydney, in the Cudgen-Cudgera area, and by Zinc Corporation. Ltd., of Melbourne, on Stradbroke Island, and in other places. PHYSIOGRAPHY Submergence of the order of 100 to 200 ft at the close of the Pleistocene period left the pre-existing hills as promontories and the valleys as deep inlets. Wave action has built bay-bars and sandspits in a northerly direction from the headlands, and lakes and marshes have formed between these bars and the initial post-submergence coast line. These lakes and swamps have been or are being filled in by river-borne sediments or wind-blown sand. Beasley1 has presented evidence that suggests a recent emergence of the coast line of about 10 ft. The evidence for this belief is the presence of a black sand seam 1/3 mile inland and 17 ft above sea level (Fig 3), and although it is not entirely certain that this seam is of beach formation and owes nothing to windblown concentration, other evidence supports the suggestion of recent emergence. Such emergence would undoubtedly aid the formation of bars and sandspits and the easterly progress of the beach front, leaving a series of parallel dunes (Fig 4). This belt of north-south coastal dunes ranges up to 1/2 mile in width and as many as 15 lines of dune have been counted (Fig 3). The foredune is generally the highest, up to 25 ft high, and there is in many places a flat platform in front of the dune known as the berm. The dunes behind the foredune are lower and less well defined

Jan 1, 1947

By R. L. Rickett, N. C. Fick

Despite the large volume of literature on alloys of iron and silicon,' there is little published information dealing specifically with the constitution, at various temperatures, of the alloys containing less than 0.10 per cent carl~on and from 0.5 to 4.5 per cent silicon that are used extensively in the electrical industry. Because such information was needed by those engaged in the production of these alloys in sheet and strip form, the constitution nas determined of a series of commercial silicon steels containing 0.05 to 0.08 per cent carbon and 0.20 to 4.1 per cent silicon, over the temperature range 1300° to 1500°F. These alloys were all from regular production heats and of commercial quality. The results of the tests are given in Fig. I and for comparison the constitution diagram of iron-silicon alloys with 0.01 to 0.02 per cent carbon, taken from the literature, is also presented Material and Procedure All steels were hot-rolled to a thickness of 0.060 to 0.080 in. and used in this condition. Their chemical compositions, including over-all check analysis, and "center" analysis for carbon and silicon, are given in Table I. The so-called center analysis was determined on samples pickled to remove the outer surface and leave only the center third of the gauge thickness—the portion of the sample on which metallographic ratings were based. In most, but not all, of the alloys, slightly higher carbon was found at the center than when the whole strip was sampled; the two methods of sampling show relatively little difference in silicon. Two of the steels, A and E, contain relatively high phosphorus, while I and K are somewhat higher in manganese than the others. The procedure consisted essentially of heating specimens (about 1/16 by 3/8 by 3/4 in.), protected from oxidation or change in carbon content, at a temperature in the desired range and for a selected period, after which they were quenched in iced brine; the degree of transformation that had then occurred was determined by metallographic examination. The soaking periods were selected after preliminary investigation had shown that a con-

Jan 1, 1947

By Walter Craft, John L. Lamont

Adequate hardenability has long been recognized as one of the first requirements for producing desired mechanical properties in a heat-treated steel. Since the introduction of the Jominy end-quench test' and the development of correlative information to expand its use, industry has had an effective measure of hardenability. Prediction and control of the hardenability have been made possible by Grossmann's principle2 for calculating hardenability from chemical composition; and Field's conversiona from depth of half-martensite hardening to Rockwell C hardness has made the relation more quantitative. These developments have facilitated the design of new steels and the control of standard grades to the extent that it has been possible to establish tentative hardenability bands for some of the more commonly used alloy steels.4 As the depth of hardening of the Jominy test specimen is usually measured by Rockwell C hardness, it was considered that it should be possible to predict the hardness by a more direct calculation. It has been found that this can be accomplished by the addition of Rockwell C units proportional to the carbon and alloy content, grain size, and position in the Jominy test specimen. The calculation has been found to be fairly accurate, and because of its additive character is relatively simple to use. In principle the method of calculation is somewhat similar to the expressions developed by Herty, McBride, and Hollen-back,5 Burns, Moore, and Archer,= and Burns and Riege1.l The calculation is started from a base that includes the effects of carbon content and position in the Jominy test speimen. Rockwell C units are added to the base in proportion to the alloy content and grain size. This sum represents the Rockwell C hardness up to the level at which a disproportionate increase of hardness is caused by the formation of martensite, and above this level an increment for martensite hardening is added. The effects of alloys are directly proportional to the amounts present and are independent of each other, the carbon content, and the position in the specimen. Furthermore, the factors for determining the martensite increment are dependcnt only on carbon content and are independent of cooling rate and alloy content. It is this absence of interrelation between the different factors that makes it possible to calculate the hardness and makes the method of calculation consistent with the principle that the character of the hardening reaction is governed primarily by carbon and that alloys exert their effect chiefly through control of the critical cooling rate. The principle of the calculation, and the effects of alloys on the critical cooling rate,

Jan 1, 1947

By R. P. Aikman, F. G. Miller, J. M. Podgursky

As part of a larger study of the mechanism of froth flotation of coal, the rates at which coal particles of various size and specific gravity react to froth flotation were investigated. Results of this study show that in flotation: I) recovery rate increases, but selectivity decreases with a decrease in particle size down to 100 mesh, and 2) recovery rate is constant for all sizes between 100M and 325M. n response to the growing demand for low-sulfur coking coal for iron making, the coal industry is continually striving to perfect methods for cleaning coal to acceptable ash and sulfur levels while maintaining maximum product yield. By present standards the industry has adequate processes for cleaning coarse coal, i.e., + 28 mesh, at almost any desired specific gravity. However, in the case of fine coal, i.e., —28 mesh, neither of the two existing processes - froth flotation or hydrocycloning — in its present state of development yields products that are comparable in ash and/or sulfur content to products normally obtained from coarse coal processes. The work outlined in this paper is part of a larger study directed toward maximum exploitation of froth flotation for fine coal processing. The concepts and data presented are now serving as a basis for designing fine-coal circuits in which improved efficiency is obtained by optimum use of froth flotation with other fine-coal processing methods. The paper includes: 1) A discussion of the value of a detailed wash-ability study as a method of assessing the problems associated with fine coal processing, 2) A general discussion of hydrocyclone principles, included to serve as a basis for understanding the mechanism of coal flotation, and 3) A detailed discussion of the mechanism of coal flotation, as well as the research leading to the optimum use of froth flotation in combination with other fine-coal cleaning methods to provide efficient fine-coal processing. Our study of the rates at which coal particles of various size and specific gravity react to flotation shows that in flotation: 1) recovery rate increases but selectivity decreases with a decrease in particle size down to 100 mesh, 2) recovery rate is constant for all sizes between 100 mesh and 325 mesh, and 3) the flotation process effectively rejects only those extremely fine particles which do not contain surface carbon. (Because of this feature, froth flotation is an excellent method for rejecting clays - the only extreme fines in - 325-mesh coal that are always carbon free.) Washability studies of flotation products are used to assess flotation efficiency. These studies show that: 1) the increase in percent sulfur in flotation products above the theoretical sulfur content as defined by the washability study is due almost entirely to flotation of fine high-specific-gravity particles that are part coal and part pyrite, and 2) in a given size range particle floatability decreases with an increase in particle specific gravity. Error curves are explained and are used to compare the mechanisms of coal flotation and hydrocycloning. These comparisons show that hydrocyclone performance is affected primarily by particle mass, i.e., particle specific gravity and size, whereas coal flotation performance is dependent on both the particle mass and the coal surface available for bubble attachment. Since the error curve is a basic tool for rating the efficiency of any coal processing system, it will be useful to quickly review the principles of error curve construction. RATING COAL-PROCESSING EFFICIENCY BY THE ERROR CURVE The organic component of coal is of lower specific gravity than its impurities, and the most effective principle for coal concentration is therefore separation on the basis of specific gravity. However, since there is no sharp dividing line between pure coal and pure ash components, specific gravity must be closely controlled if the product is to have acceptable quality at maximum yields. In turn, the basis for close control in any coal processing system is the reference standard provided by precise laboratory washability studies. In these studies, coal is carefully separated into a series of progressively higher specific gravity fractions to establish a relationship

Jan 1, 1968

By Nicholas J. Grant, Italo S. Servi

THE creep and stress rupture properties of three grades of aluminum have been reported in a previous paper.' It was found that the stress coefficient of the creep rate and of the rupture time changes at the transition from low temperature-type to high temperature-type deformation. The transition corresponds to the equicohesion for grains and grain boundaries. In order to support the conclusions obtained from the examination of creep and stress rupture data, the deformed specimens were observed metallographically and a few special tests were conducted. The results of these qualitative observations are given in the present paper. Experimental Procedure and Results Table I gives the types of aluminum specimens.' The creep and stress rupture tests were conducted in simple tension at constant stress. Further details on the experimental techniques have been given.' The test bars were electropolished in a perchloric-acetic mixture before testing and were observed after testing without any further preparation unless otherwise specified. Observations of specimen cross-sections were made after electrolytic polishing and etching. The samples were extended along a direction parallel to the longer side of the micrographs of Figs. 12 through 27. The log-log plots of stress vs. minimum creep rate for the three grades of aluminum appear in Figs. 1 and 2. The reference numbers on each curve help relate the photographs to the conditions of testing, indicating where high temperature-type deformation and failure prevailed and where low temperature behavior prevailed. Deformation of Single Crystals at Very High Temperature: An annealed, high purity aluminum specimen was slightly strained at one end and re-annealed. A large single crystal was formed at the strained end of the specimen during re-annealing; the other end remained polycrystalline (see Fig. 3). A stress of 50 psi was applied to the cold specimen to avoid loading at 1100°F. The temperature was then raised to 1100°F, kept at 1100°F for 3 hr, and lowered again to room temperature. Fig. 4 shows the appearance of the specimen after testing. The single crystal deformed by gliding along parallel planes, while the polycrystalline part deformed mainly by grain boundary flow. Some grain growth occurred in the polycrystalline part. Another high purity specimen, which had two large single crystals at the two ends and a polycrystalline zone in the middle portion, was tested at 1000 °F under a stress of 100 psi. The appearance of the sample after test is shown in Fig. 5. Even though considerable slip occurred in the single crystals, the fracture occurred in the weaker polycrystalline portion. Evidence of Grain Boundary Migration during Creep Testing: A high purity, coarse grained specimen, which had a large crystal in the middle portion (see Fig. 6) was tested at 700°F under a constant stress of 175 psi for 30 hr. The appearance of the specimen after testing (Fig. 7) indicates that the large crystal deformed by gliding along parallel planes, while the rest of the specimen deformed mainly by grain boundary flow. Microscopic examination showed clear evidence of grain boundary flow and grain boundary migration, an example of

Jan 1, 1952

By G. C. Rauch, T. H. Courtney, J. Wulff

The superconducting behavior of Nb-Ti alloys containing 40 and higher wt pct of Nb and variable oxygen content was studied as a function of thermomechanical processing. Critical current density (Jc) us applied transverse magnetic field (H) data were obtained for 0.010 in. diam wires at 4.2° Kin steady fields up to 120 kOe. The precipitate responsible for increase in Jc in each of the alloys studied was related to niobium and oxygen content. In alloys containing 40 to 60 wt pct Nb with less than about 1600 ppm of oxygen, w precipitation in the range 350° to 450°C appears to be responsible for the maximum value of Jc observed. In alloys of higher oxygen content (-2500 ppm), the oxygen-enriched a phase which precipitates in the temperature range 450° to 550°C is a more effective fluxoid pinner. Maximum Jc in low oxygen (350 ppm) 80 wt pct Nb alloys was achieved by 600°C aging heat treatments. Measurements of resistive critical field also reflect the changes in composition accompanying precipitation in these alloys. It is well known that the superconducting critical current density, Jc, of solid solution superconducting alloys can be increased by precipitation heat treatments.1-4 The precipitate, to serve as an effective superconzlucting fluxoid pinner and increase Jc, must be of proper size and spacing.' Its electronic properties must also be different from those of the matrix.6 As in age-hardenable systems, optimum temperature and time of aging are dictated by the composition and history of the particular solid-solution alloy under consideration. The alloy diagram of Ti-Nb7-9 shown in Fig. 1 is typical of ß-stabilized Ti alloys. Precipitation to achieve high Jc in Ti-Nb and Ti-Ta alloys is usually carried out in what is believed to be a two-phase a + ß region.1,2,4,10 Previous work on Ti-Nb4,11 and on other ß-stabilized systems such as Ti-V, Ti-Mo,12-14 has shown, however, that the first precipitate to appear is not necessarily (Y. Under specific conditions the meta-stable w phase is precipitated instead. A further complication arises in such alloys from the strong influence of contained oxygen on the phase stability and on the kinetics of precipitation.15,16 The superconducting critical current density of Nb-Ti alloy containing less than 36 wt pct Nb has been studied by Vetrano and Boom1 and Kramer and Rhodes.4 The most effective aging temperature was found to be about 425°C; fluxoid pinning was attributed to the precipita- tion of a by Vetrano and Boom and of w by Kramer and Rhodes. Although the latter workers used recrystal-lized material and the former cold-worked, studies carried out in our laboratory lead us to believe that the type of precipitate responsible for increased Jc is unaffected by prior cold-work.17 That other types of precipitates may also be effective, especially in higher oxygen-containing alloys, has been demonstrated by Reuter, et al.2 in this laboratory with the aid of electron microscopy. Reuter et al. found that at an aging temperature of 600°C, fee-TiO is precipitated, resulting in high values of Jc. Even higher values of Jc were observed after heat treatment at 500°C, but the precipitate could not be indexed as fee-TiO. Whether the precipitate was oxygen-rich w or a could not be clearly distinguished. The work nevertheless indicates that appreciable amounts of oxygen alter the nature of the phases precipitated at different temperatures, at least in alloys containing more than 36 wt pct Nb. Since such compositions are more readily worked into suitable wire than lesser niobium content alloys, it was decided to study the response to aging of the alloy compositions reported in the present paper with the hope of clarifying the effect of composition on the precipitation processes responsible for high values of Jc. I. EXPERIMENTAL Nb-Ti alloys containing 40, 50, 60, and 80 wt pct Nb were prepared by are-melting 35 g ingots of iodide crystal bar titanium and electron-beam melted niobium.

Jan 1, 1969

By M. R. Geer, H. F. Yancey

WITH the exception of pneumatic processes and ii few special beneficiation methods of comparatively limited application, all mechanical coal-cleaning and mineral- dressing processes involve the admixture of solid particles with a liquid, generally water. Upon completion of the beneficiation process the solids must be reclaimed from the water. With coarse particles separation of solids from water is accomplished readily by the simple expedient of screening, but recovery of fine solids presents a much more difficult problem. Thickening and dewatering are practiced so generally, both within the mineral industry and else- where, that they have been the subject of study through the years and a number of processes for making the solid-liquid separation have been developed. Dorr-type thickeners have been standard equipment for many years, and more recently driers employing centrifugal force have gained in popularity. 1 A cyclone device employed to thicken the loess suspension is a novel feature of the loess heavy-medium coal-cleaning process developed by the Netherlands State Mines in the thirties. The cyclone thickener was first mentioned by Driessen2 in 1939 in a paper describing the loess process and has been the subject of two more recent reports by the same author.3 The cyclone does not differ greatly from cyclone dust collectors used in this country for years, except that it is operated wet rather than dry. Advantages of the cyclone over other thickening devices are its simplicity of construction, high capacity, and the fact that, having no moving parts, it is simple and inexpensive to operate. The Bureau of Mines became interested in the cyclone after one of the present authors observed it in operation in the Netherlands in 1945. The first experimental work with the cyclone conducted by the Bureau of Mines was an investigation of the unit functioning as a heavy-medium cleaner for fine coal-a use for the cyclone that was conceived long after it had been perfected as a thickener. The results of this investigation were published last year.4 The present investigation deals exclusively with operation of the cyclone as a thickener. OBJECT AND SCOPE OF INVESTIGATION Although the cyclone thickener has been dealt with in three reports by Driessen, its use as a thickener was not the principal subject of any of these reports and therefore neither the cyclone itself nor its operation was described completely. The design of the cyclone was described only in a very general way, only limited operating results were given, and virtually nothing was disclosed about the influence of the various possible operating variables. Moreover, the Netherlanders apparently had employed the cyclone exclusively for thickening loess and had not employed it on coal slurries. One English colliery has used a cyclone to thicken coal slurry,5 and as it appeared that

Jan 1, 1948

By A. Kaufman

More than 3 trillion gal of water are used annually by the mineral industry. Of this, approximately 21/2 trillion gal are recirculated, the rest constituting intake water. The major users are natural gas processing plants and phosphate rock, sand and gravel, and iron ore producers. Water was used by the mineral industry for mining (6%), processing (64%), cooling and condensing (27%), and miscellaneous uses such as boiler feed and sanitary purposes (3%). Whereas total water use is dependent on the quantity of material processed and on the particular process requirements of an industry, recirculation is dependent on processing, as well as cooling and condensing requirements, quality of new water intake, and the necessity for treating new and discharged water. Consumed water, on the other hand, is dependent on the quantity of water recirculated, and temperature and humidity in the area. Based on this analysis, an increase in water use by the mineral industry of 21/2 times by 1985 is forecast. Wster intake, however, will only rise 62%, because of a substantial increase in recirculation. In one report of a special series concerning the water resources of the United States, the Senate Select Committee on National Water Resources, 87th Congress, estimated that water demand would double by 1980 and triple by 2000.' In view of the possible water deficiencies that might result from such expanded usage and the need for research guidance, the Bureau of Mines organized and carried out a statistical canvass of water use in the mineral industry for calendar year 1962. The data used in this paper, unless otherwise noted, are derived from that canvass.' The efforts in this paper are devoted toward summarization of the canvass and analysis of the data. SOME DEFINITIONS Intake: Water introduced from an external source for the first time into a given mine or plant regardless of quality. Intake water is also called new water, water withdrawn, or makeup water. Fresh Water: Water suitable for cooking and drinking. Saline Water: Water containing more than 1000 parts per million of dissolved solids. Contaminated Water: Water not suitable for domestic use, but excluding saline water. Recirculated Water: Water reused to conserve intake water. Solutions that are recycled primarily because of fixed metallurgical practices, such as copper leaching solutions containing sulfuric acid, are excluded. Gross Water Used: Recirculated water plus intake water. Also called total water used. Consumed Water: Water that is lost by evaporation, as well as water lost in product. Seepage and transferred water are not considered consumed. The use of either intake or recirculated water may result in consumption. However, because of difficulties in measuring consulmption, consumed water is defined as the residual between intake and water discharged from the mine or plant. Mineral Industry: For the purposes of this paper, mineral industry includes all metal and nonmetal surface and underground mines and their associated processing plants, as well as custom mills, coal washing plants and associated mines, petroleum and natural gas well drillers, natural gas processing plants, and secondary recovery operations. WATER USE IN THE MINERAL INDUSTRY Water used by the mineral industry constitutes a relatively minor fraction of the water withdrawn by all users. Data compiled by the U.S. Geological Survey in 1960 indicate that water withdrawals, exclusive of that used to generate hydropower, approximate 99 trillion gal annually.3 Our data indicate that water withdrawals by the mineral industry comprise only 1% of this total, or 2% of the water withdrawn by industrial users. Use by Industry: Water used by various mineral industries is shown in Table I. The table indicates that close to one-half of the gross water used by the mineral industry in 1962 was used by natural gas processing plants, followed by sand and gravel, phosphate rock, and iron ore producers. The largest aggregate users of water are also the largest users per dollar of product. For example, Fig. 1 shows that the natural gas processing and phosphate rock industries are very large users of water per dollar of product. Their recirculation per

Jan 1, 1968

By D. A. Dahlstrom

WITHIN the past decade, certain economic, legal, and technological changes have intensified the necessity for an efficient classifier or deslimer of particles at and below the 200-mesh point. First, mechanization of mining has increased the percentage of clays, silts, refuse, and high-ash material in the run-of-mine coal. Second, many washing plants must rely on smaller percentages of fresh make-up water because of shortages, economic factors, or legal restrictions on waste water disposal. Finally, the crushing of middlings and coarse refuse plus the trends toward the finer sizes in sales has further increased the amount of fines produced per ton of mine-run coal. A joint consequence of these circumstances has been the augmented percentage and build-up of —200-mesh material in the finer coal fractions and tipple circulating water. This results in the following disadvantages to the operator: (1) Higher ash percentages in the fine coal caused in large measure by the greater ash content of the —200-mesh material, (2) more difficult or expensive recovery of the fine coal from plant slurries, (3) lower filtration, drainage, and drying rates caused chiefly by lower permeability and the greater surface exposed per unit mass of the finer particles, (4) greater difficulty of beneficiatiing the fine coal because of lowered equipment efficiency and the partial "coating-out" of the —200-mesh fraction on the coarser coal particles. Obviously some cheap but effective fine-size classifier must be developed to aid in solving the above problems. To eliminate the generally unwanted —200-mesh fraction from coal slurries, several methods are in use. Probably the most popular with coal operators is the settling cone wherein the larger particles settle downward to the apex of an inverted cone while the slimes and fines are carried upward by vertical fluid currents caused by the overflow of the bulk of the water at the outer rim. The main advantage of the unit is the low initial and operating cost since no moving parts are involved. Disadvantages are primarily low capacity per square foot or floor space and the loss of good coal in the overflow above the 100-mesh size due to localized, high-velocity gradients and nonuniform, fluid-flow patterns, especially at high loadings. Also, when high slime concentrations are present, desliming efficiency is severely injured.', ' Other methods employed are the drag and spiral classifiers and the hydroseparator.' These involve mechanical separations achieved by moving parts and naturally result in higher initial costs. As they are all based on the settling velocity of the particles at the mesh of separation under operating conditions, area requirements are again large per gallon of throughput if they are to classify at the 200-mesh point. Usually the selection is made to operate at a higher throughput and thus sacrifice some coarser coal in the slime overflow. Cyclone Possibilities in Desliming To overcome the disadvantages of present deslirning methods, it is necessary to use a device that exhibits a very stable and uniform flow pattern while greatly increasing the settling velocity of

Jan 1, 1952

By Samuel S. Adams

Potash, the generic term for a variety of potassium-bearing minerals, ores, and refined products (Table I), owes its importance as an industrial mineral to the potassium requirement of growing plants. The term "potash" originally referred to potassium carbonate produced by the leaching of wood ashes. This process, for which the first U.S. patent was issued in 1790, was the basis for the largest United States chemical industry in the early nineteenth century. The agricultural use of potash soon eclipsed the chemical applications following the German discovery in 1840 that potassium is essential to plant growth. The initial discoveries of large deposits of soluble potassium minerals, essential for agriculture applications, were made in 1857 in Germany and in 1914 in the United States. More than 90% of the 20.4 million short tons K2O produced in the world in 1970 (Eilertsen, 1971 ) was derived from marine evaporite deposits and approximately 90% was used in fertilizers. The potassium content of an ore, mineral, or product is customarily expressed as % K,O, a unit of measure which is neither a natural nor synthetic compound of potassium (1 .O% K is equivalent to 1.2046% K,O) . Geology Potassium, the seventh most abundant element in the earth's crust, is present in silicate minerals in igneous, sedimentary and meta- morphic rocks. Potassium is also a major constituent of many surface and subsurface brines (Table 2). Economic potash deposits are essentially restricted, however, to widespread, thick accumulations of marine chloride evaporite deposits. These deposits yield high grade, large tonnage ore bodies, many of which are amenable to low cost mining and beneficiation. The products from these deposits are, moreover, ideal for use in fertilizer because of the high relative solubility of the potassium chloride and sulfate evaporite minerals. World potash production and reserves for such de- posits so vastly overshadow other sources of potassium that the following discussion will be limited largely to potash deposits in marine evaporites with brief reference to certain potash-rich brines. Approximately 85% of current domestic production is from bedded salt deposits and much of the balance is from brines. Marine evaporite deposits are accumulations of salt minerals deposited in structural sedimentary basins through the evaporation of sea-water or mixtures of seawater and other brines. The salt minerals are precipitated in assemblages and sequences of assemblages as determined by the compositions of the brines and the solubility relations in the salt systems (D'Ans, 1933). The evaporation of seawater normally yields carbonate minerals as the first precipitates (Table 3). Further evaporation and concentration of the brine leads to the precipitation, in succession, of sulfate, chloride, and supersaline mineral assemblages including the salts of potassium and magnesium. The common minerals, ores, and products of potassium deposits are listed in [Table 1]. Many potash deposits contain accessory evaporite minerals together with a variety of detrital and authigenic silicates (Braitsch, 1971).

Jan 1, 1975

By W. B. Morrison

The principal factors influencing the total percent elongation of a lead-tin eutectic and several low-alloy steels which exhibit superplasticity were investigated. These factors are: a) Strain-rate sensitivity of the stress (m). b) Specimen geometry; in particular, diameter-to-length ratio (d0/l0), and surface irregularity (e.g. notches). A feature of the tensile deformation of a superplastic alloy is the early appearance of a neck which continues to extend throughout the test. As a result the total elongation is directly related to the diameter-to-gage-length ratio. The strain-rate sensitivity of the stress varies during extension, mainly because of microstruc-tural changes. A satisfactory correlation exists between the minimum strain-rate sensitivity and total elongation, as indicated in the equation shown below. Notches cause a considerable reduction in total elongation and their effect is greatest at high values of strain-rate sensitivity. An initial notch depth exists below which there is no appreciable effect on elongation. An equation relating elongation to strain-rate sensitivity and specimen geometry is proposed, THE total percent elongation of a tensile specimen is usually regarded as a measure of superplasticity. However, little attention has been given to the fact that total elongation is also a function of specimen diameter and gage length, even though the importance of geometry has long been recognized in the tensile testing of materials in which elongations are relatively small.* In the present study, the effect on superplas- tic extension of the ratio of specimen diameter to length was investigated and an analysis was made of strain distribution along the gage length. Superplastic alloys are characterized by having a flow stress sensitive to strain rate but relatively insensitive to strain.2"' In an earlier study,5 it was found that the strain-rate sensitivity of several low-alloy steels changed during the test and that the minimum value correlated best with total elongation. In the present study, the variation of strain-rate sensitivity during the straining of a lead-tin alloy was investigated. EXPERIMENTAL PROCEDURE Lead-tin eutectic alloy (32 pct Pb 68 pct Sn), made from high-purity lead and tin, was cast in the form of 1-in. diam rod. The rod was cold-rolled in stages to 0.25 in. diam and then swaged to 0.125 in. diam. Samples were cut from the rod at various stages of the mechanical working process. Tensile test specimens were then made by attaching threaded steel ends to the rods by an epoxy-base thermosetting adhesive which required a setting treatment of 2 days at room temperature. The rods were seated in the steel ends in holes approximately 0.01 in. oversize. The test pieces thus made were of various diameters and gage lengths. These were stored at a temperature of about -70°C (-94°F) until required for testing. The distribution of strain along the gage length of a specimen was determined at various stages of the tensile test. The lead-tin alloy was superplastic at room temperature and direct observations during testing were simple. The gage length was divided by reference marks into five initially equal units. These reference marks were made with ink and were reinforced during the test when the initial marks became faint. Photographs were taken, at intervals, of the deforming specimen and measurements of the elongation in each of the gage units were made on these photographs. Tests were also done on several superplastic low-alloy steels whose properties were described earlier.5 The chemical compositions of these steels are given in Table I. A diagram of the notched tensile specimen used in the steel tests is shown in Fig. 1. The specimens were tested at 900°C (1650°F) in a furnace through which a mixture of He + 2 pct H was continuously circulated.5 RESULTS Strain Distribution. Approximately 20 tests on lead-tin alloy specimens were conducted to determine the strain distribution during the test. An example is shown in Fig. 2 of the strain distribution in the gage

Jan 1, 1969

By W. W. Mullins

The development, by the mechanism of volume diffusion, of a grain boundary groove on an interface separating a solid phase and a saturated fluid phase is calculated under the following assumptions: 1) isotropy of the interfacial free energy 's, 2) applicability of the Gibbs-Thompson formula relating curvature and chemical potential, 3) a nearly planar groove surface. The groove profile is found to have a .fixed shape and linear dimensions that are proportional to (time)1/3. The constant of proportionality is evaluated and involves ?s in such a way that the latter my be evaluated from grooving kinetics. A grain boundary that intersects an interphase interface will tend to produce a groove along the line of intersection (Fig. 1). The ultimate motivation for the formation of the groove is the reduction in interfacial free energ that occurs as the grain boundary contracts. Smith has discussed the way in which the free energy of the interfaces determines their dihedral angles at the groove root and has given a detailed analysis of the effect of the dihedral angles upon the nature the microstructures developed in polyphase alloys. Herring2 has extended the considerations to the case in which the free energy of an interface depends on its orientation and has thus added torque terms to the condition for interface equilibrium. Although the grain boundary and the interphase interface may quickly achieve the proper equilibrium angles where they intersect, as determined by the balance of equivalent tensions and torques,"2 the groove will generally continue to grow (though with decreasing speed). Thermodynamically, this continued growth reflects the fact that the lowest free energy cannot be achieved until the entire grain boundary has been eliminated; mechanistically, the growth occurs in response to inhomogeneities of chemical potential that are caused by the constraint of fixed dihedral angles at the groove root. The latter point is illustrated in Fig. 1, which shows that atoms* near the groove are on the convex side of a cylindrically curved surface and therefore have a higher chemical potential than atoms on the flat interface beyond. Thus atoms tend to leave the curved groove shoulders and force the groove to deepen. When both phases are solid additional chemical potential gradient may arise from whatever strain fields are necessary to insure continuity of material across the interface. In the discussion that follows, this complication due to strain will not occur since we always assume one phase to be a fluid and the other a solid. The transport processes that may operate to permit growth of the groove are surface (interface)* diffusion, volume diffusion in the phase on either side of the interface, and evaporation-condensation when one phase is gaseous. (Also a melting-freezing process is responsible for groove enlargement on a solid-melt interface as discussed in Sec. VI). The theory of grain boundary grooving has been discussed in detail for the processes of surface diffusion and of one type of evaporation-condensation.3'4 The theoretical predictions for the mechanism of surface diffusion have been confirmed in an experimental study of grooving kinetics in copper.5 In this paper, the theory of groove development by volume diffusion will be presented. Volume diffusion can occur either by atomic migration in the solid S or by diffusion of S atoms in the fluid phase, or solvent, adjacent to the groove surface. As discussed in Sec. V, volume diffusion in the solid should be less important than surface diffusion in forming a groove. Apart from verifying this statement then, the main interest of the present analysis centers on diffusion in an external solvent. The solvent may be either a liquid (e.g., water, or-

Jan 1, 1961

By R. Fawley, M. A. Quader, R. A. Dodd

Stacking fault energies and annealing twin frequencies have been measured for austenitic stainless steels having a) constant 20 wt pct Cr and varying nickel contents, and b) constant 20 wt pct Ni and varying chromium contents. The effect of nitrogen has also been studied and, to a lesser extent, so has the effect of carbon. All measured stacking fault energies plot linearly against the corresponding annealing twin frequencies, suggesting that, for constant prior deformation and grain size, twinning frequency can be used to estimate stacking fault energy. Compositions low in nickel possess the lowest fault energy, while the maximum fault energy within the investigated composition range occurs mound 20 pct Cr-20 pct Ni. Nitrogen has practically no effect on stacking fault energy when increased from 0.005 to 0.05 wt pct, although Planar dislocation glide is associated with higher nitrogen contents; this may be a short-range order effect. Carbon may raise the stacking fault energy a little, and causes pronounced dislocation tangling in deformed alloys. The basically different dislocation structures promoted by nitrogen and carbon, respectively, are not associated with large variations in stacking fault energy, and the reason for the difference remains unknown. INTEREST in the dislocation configurations developed in plastically deformed alloys arises principally from a desire for an understanding of mechanisms of plastic deformation and fracture. Comparatively little basic work has been concerned with austenitic stainless steels, although considerable interest has developed in the fundamental aspects of deformation of these alloys since the suggestion that those compositions deforming essentially by planar glide are particularly prone to transgranular stress corrosion cracking.' Certain facts are well-established. An 18 Cr-8 Ni steel (type 304) has a low stacking fault energy which is increased by higher nickel contents, and the structure of these steels when moderately plastically deformed changes concomitantly from planar to less planar to distinctly tangled dislocation configurations. These changes are thought to relate to changes in stacking fault energy, and considerable effort has been expended in measuring stacking fault energies of austenitic stainless steels through the determination of node curvatures2"4 even though the method has limitations due to formula approximations,5 an upper limit of measurement of 50 erg cm-2 at the most,5 and the necessity of an extremely meticulous experimental technique. Therefore, it is not surprising that only qualitative agreement between different investigators is generally to be found.'-' At high energies where node radius measurements are inappropriate, some use has been made of annealing twin frequencies to estimate relative stacking fault energies, although a recent critical evaluation of the method suggests that it should be used cautionsly with due regard to effects of grain size and prior deformation.6 Finally, there is a question as to whether the structures of some deformed steels are due to ordering effects. In particular, the role of nitrogen in promoting planar dislocation motion has been suggested to be due to ordering,' but similar suggestions have been made for steels having low interstitial contents.4 EXPERIMENTAL Small ingots of austenitic stainless steels (-50 g) were prepared from components exceeding 99.99 pct purity, the nickel and chromium being Johnson Matthey "Specpure" metals. Carbon was added as pure graphite, and nitrogen as CrN. The components were accurately weighed, and it was assumed that the final composition was the same as the nominal composition

Jan 1, 1969