Search Documents

By H. H. Smith, G. J. Danek, M. R. Achter

Nickel, Type 316 stainless steel, and copper show decreases in reversed bending fatigue strength with increases in air and oxygen pressure. The leveling off of the log N us log p curve for nickel at 300" and 816°C is discussed in terms of a model in which surface saturation gives rise to a critical pressure. A metallographic study indicated that the effect of environment on crack initiation and growth rates is a function of temperature. It is shown that small amounts of impurities in inert gases may substantially reduce the fatigue strength of nickel. 1n a previous investigation1 of high-temperature bending fatigue in air and in vacuum, reversals in the effect of environment with changes in test conditions were atrributed to a mechanism consisting of two processes acting in competition. One of them, oxidation strengthening, has been demonstrated in a straightforward manner by measuring the rupture strength of a sintered nickel-oxide bond.' The detailed operation of the other process, facilitation of crack propagation by surface absorption of oxygen, is uncertain. Thompson et UZ advanced the possibility that the formation of an oxide layer prevented rewelding of cracks during the compressive part of the cycle, while Wads worth and Hutchings4 further proposed that the reaction of oxygen with atoms at the tip of the crack helped the metal to break there, a possibility which has also been advanced to explain crack formation in creep.5 A difficulty in arriving at a mechanism for this lowering of fatigue strength in air is the inconsistency of some of the data. For example, Wadsworth and Hutchings4 found an approximately linear decrease of log cycles-to-fracture (N) with an increase of log pressure-of-air (p) for copper and aluminum at room temperature, while snowden,' although finding a similar linear dependence for aluminum, observed a stepped curve for lead. Between l0-' and 10"1 torr there is a sharp decrease in the fatigue life of lead with increasing air pressure, but plateaus, signifying no dependence on pressure, are found at higher and lower pressures. If, as discussed elsewhere: these environmental effects arise from the adsorption of gas on crack surfaces, it would be expected that as the pressure is increased a saturation point would be observed, beyond which there is no further effect. To gain further information about the form of the log N vs logp curve, it was decided in the present investigation to repeat this type of measurement at elevated temperature on nickel, copper, and Type 316 stainless steel. An additional object of this program was

Jan 1, 1963

By R. H. Landshof, Reynold Q. Shotts, James A. Redding

GENERAL INTRODUCTION by R. Q. Shotts The sizing of coal particles is one of the most important beneficiation operations performed from the time coal is broken at the face until it is delivered to the customer. Indeed, in many cases it may be virtually the only such operation performed in that interval, although it is most often accompanied by some comminution. Sizing, as the term implies, is a separation of the broken coal particles into arbitrary groups, each group containing particles within restricted mass, volume or size limits. Size itself is not a rigidly defined property, but is actually determined more by the operating principles of the sizing device than by unique properties of the particles themselves. Particle size cannot be uniquely defined independently of particle shape. Taggart l says, "In fact, it is not possible to state the size of a single irregular body by means of a single dimension; such possibility exists only for regular polyhedra and their limit, the sphere." In view of the scarcity of regular polyhedra among coal particles, even though most appear to approximate a cube, we actually define size in terms of a surface opening through which the particle barely will or will not pass, or of two openings, the smaller of which retains all particles of the group under study and the larger of which passes all of them. Alternatively, size is expressed in terms of rate of settling of particles in a fluid under closely controlled conditions. Historical Sketch2-Taken from Coal Preparation, 2nd Ed. "Sizing is the process of separating mixed particles into groups of particles all of the same size or into groups in which all particles range between certain definite maximum and minimum sizes. In coal preparation, sizing is generally accomplished by passing the coal over screens. Separation by differential settling in air or water currents has been adapted to the special field of separating very fine sizes, and these processes are in commercial use to a relatively small extent in the coal industry. Separation with screens is a very old industrial practice. Agricola described the use of screens for ore treatment in the middle ages. In the literature of the British coal industry references were made to the screening of coal as early as 1589, but the practice of extensive sizing for the market did not become general until the last half of the nineteenth century. "In America, the practice of sizing developed first in the Pennsylvania anthracite region, beginning with the use of rakes in the mine working places to recover coarse coal. The mining procedure left the fines in the mine.

Jan 1, 1968

By A. Cochardt, D. H. Feisel

HE success of the sintered aluminum powder material (SAP)' in raising the maximum useful temperature of aluminum alloys has stimulated effort in investigating similar systems in which the metallic matrixes have higher melting points. Examples of such systems which have been studied are nickel-alumina,2, 3copper-silica,4 copper-alumina,4 titanium-cerium oxide,5 and 80/20 nickel-chromium,8 cobalt,= and molybdenum7 matrixes with various hardeners. Conventional precipitation-hardened high-temperature alloys are often thermally unstable in service. The original fine particles that are responsible for good strength properties grow in size at elevated operating temperatures, a process termed "over-aging." This growth of precipitated particles causes a gradual loss in strength properties. A previous paperb describes the factors that govern the process of overaging, and estimates the time required for overaging of a nickel-alloy matrix dispersion-hardened by various precipitates such as are found in current high-temperature alloys, and by alumina. These calculations indicate that nickel-base alloys hardened with Ni3Al, Tic, or TiN are quite unstable at 1340" F (727°F). Such alloys should overage within a few hours at that temperature. A nickel-base alloy hardened with alumina, on the other hand, should behave differently. It should not overage appreciably within a thousand years. It was further shown8 that the thermal stability of the oxide-hardened alloy is mainly due to the low solubility of oxygen in equilibrium with alumina in the nickel-base matrix. Since the oxygen solubility is small, little oxygen can diffuse, and structural changes require very long times. This paper illustrates the effects of time and temperature on the structural stability of nickel alloy matrixes hardened by dispersions of Ni3A1, Ni3Ti, Tic, and alumina, and of nickel-chromium matrixes hardened by alumina. Hardness measurements and metallography have been used as a measure of structural stability. EXPERIMENTAL PROCEDURE Using the best information from published phase diagrams,'-l4 compositions of the Ni-Al, Ni-Ti, and Ni-Ti-C alloys were determined so as to have 5 vol pct of precipitate at 1350° F(732°C), the temperature selected to illustrate the effect of time on structural stability. Table I lists the chemical compositions of these alloys. Standard laboratory melting and alloying techniques were used in preparing ingots of the F-Series of alloys. The cast ingots, 1 3/4 in. in diam, were forged to 3/8-in. diam rods.

Jan 1, 1960

By Alexander Coucoulas

FABRICATION of thin-film circuits deposited on glass substrates requires the formation of permanent electrical contacts. Since thin films are in the order of a few thousand angstroms thick and are deposited on nonconductive surfaces such as glass, new methods of lead attachment are being pursued in the microelectronics industry. Ultrasonic welding is being considered in thin-film technology since joining can be accomplished between a wide variety of dissimilar materials.l,' Ultrasonic welding is accomplished by a transducer-coupling-tip system which converts high-frequency electrical energy into mechanical vibrations and delivers it into the weld region. During the process the tip oscillates in a direction parallel to the surface of the weldments, inducing shear stresses in the work pieces which are clamped between the tip and anvil. Though the basic bonding mechanism is not clearly understood, it has been described as closely akin to friction welding. In the course of an investigation concerned with ultrasonic we!ding of aluminum wires to tantalum thin films (1400A) deposited on glass, low weld strengths were related to glass failure. Upon investigating this condition, an improved method of bonding was developed where the aluminum was held at a temperature where a minimum amount of strain hardening occurred during the introduction of ultrasonic energy. Experimental Procedure. Aluminum wires (0.015 in,) were bonded to tantalum thin films with an optimized set of welding parameters (power, time, clamping force) employing a commercial 100-w 60-kc ultrasonic welder. Forty-five degree shear-peel tests, where the lead is pulled at a 45-deg angle with respect to the substrate at 0.2 in. per min, produced a large variation from zero to the strength of the wire, Table I. Inspection of the weld region generally revealed: glass cracking before and after testing; the removal of glass nuggets from the substrate by the lead during testing, Table I; thin-film penetration. In an effort to improve the shear-peel strength, the aluminum between the tip and substrate was heated above ambient temperature through an argon gas media prior to and during welding with the same parameters employed above. With the addition of heat: wire deformation was increased; shear-peel tests all failed at the weld-wire junction, leaving the bonded portion undisturbed; no evidence of glass cracking or film penetration prior to or after testing was observed. In order to prevent failure at the weld-wire junction due to excessive wire deformation, 0.035-in. aluminum wires were welded to tantalum thin films with an optimized set of welding and heating parameters. As

Jan 1, 1967

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 John L. Lamont, Walter Craft

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 Ralph Hultgren, Bernard York, David W. Mitchell

It is a well-known fact that magnesium-alloy castings are apt to be coarse grained if the melt is not superheated several hundred degrees above the melting point before casting. (The casting temperature varies according to the shape of the mold. A typical casting temperature is perhaps 1400°F.) British practice is to heat to 900°C. (1650°°F.) for 15 min., cool rapidly to casting temperature, and cast.1 In Germany superheating is carried out at 850° to 900°C. (1560° to 1650°F.) depending on the amount of scrap used. It is stated that further refinement will be obtained by prolonging the period of superheating or by increasing the superheat temperature to 950°C. (1740°F.).2 In the United States superheating is carried out at i600° to 1700°F. (871° to 927°C.), perhaps even higher in some places. It is stated that superheating effects are found as low as 1500°F. (816°C.), but several hours are required.3 These commercial treatments lead to grain sizes in sand castings that may vary from 4 to 8 grains per sq. mm.§ in heavy sections and from 60 to l00 in 3-in sections to perhaps 380 in thin sections. It is also generally known that the refining effect of superheating is lost if the melt is held for some time at lower temperatures before casting. Aside from these very qualitative statements, there is little to be found in the literature as to the effect of temperature and time of superheat on the final grain size. Because of the uncertainty of these factors, the Office of Production Research and Development of the War Production Board placed a research contract with the University of California, supervised by the War Metallurgy Committee. This research, which was done under the "Restricted" Project NRC-550, is the basis of this paper, which has been released for publication by the O.P.R.D. The material studied was an alloy prepared from carbothermic magnesium. The temperatures required for grain refinement proved to be widely different from those used in practice, as described in the statements in the first paragraph. This difference is attributed to the fact that the magnesium in the specimens was of carbothermic origin instead of the electrolytic magnesium in common use. MATERIALS AND METHODS The magnesium alloy studied was one of the most popular sand-casting alloys, with a nominal composition of 9 per cent aluminum, 2 per cent zinc, and 0.r per cent (minimum) manganese and the balance magnesium. This is designated

Jan 1, 1945

By F. J. Meringer, C. W. MacGregor

Various investigations have been carried out to determine the effects of cold-rolling on the common physical properties as represented by the yield strength, tensile strength, percentage of elongation and ordinary reduction of area. Very little information is available in the literature with reference to the effects of cold-rolling on the true stress-strain properties. Since the ordinary physical properties base both the stresses and the strains on the original dimensions of the test piece, which are greatly changed during the test, they are essentially empirical values and do not represent physically either true stresses or true strains. Consequently, they do not give an accurate picture of the effects of different conditions, both mechanical and metallurgical, on the strength and ductility of the material. It is not, however, the purpose of this paper to review the various advantages accruing from the use of true stress and true strain values, since these have been shown elsewhere,1-l2 but to utilize them in representing the various effects considered. The true stress S and true strain (e = q') values mentioned above for the tension test are defined by1 S =1/A and where P, A, Ao are the axial load, the instantaneous area and the original area of cross section, respectively. From simultaneous readings of the axial load and the dimension or dimensions determining the instantaneous area of cross section, the true stress-strain curve is constructed by using the relations given in Eq. I. It is then the object of this paper to discuss the general effects of cold-rolling on the true stress and strain values for a low-carbon steel together with the effects of age-hardening and directionality on these properties. Preliminary Investigations The material considered throughout the paper is a low-carbon steel of the following chemical analysis: C, 0.19 per cent; S, 0.118; P, 0.092; Mn, 0.85; Si, 0.01. The stock was received in bars, 7 by 3/4 in., and was first annealed for one hour at 1650°F. followed by a furnace cool. Photomicrographs of this material showed considerable directionality even in the annealed condition. In view of the possible effects of the nonhomogeneity and directionality in this material, it was considered important to investigate their influence on the true stress-strain diagrams. In addition, the effect of strain-aging in this steel was also made the subject of a preliminary study. Effect of Aging on True Stress-strain Properties Since a certain time element would elapse between the cold-rolling operation and the final tension tests, it was determined to investigate the aging character-

Jan 1, 1945

History of Reverberatory Smelting in Montana, 1879-1933. By FREDERICK LAIST (Trans, vol 106 31,800 words ) The development of the art of reverbera¬tory smelting from 1879, when smelting first started in Butte, up to the present time is described. The paper traces the development of the reverberatory furnace, as regards both constructing and operating, from the initial 10-ton furnaces to the present-day 600-ton furnaces and the changes in operating practice which accom¬panied the gradual enlargement of the furnaces and the reasons therefor are described and explained The results obtained from an investigation (made in the course of the Carson litigation) into the various methods of charging reverberatory furnaces are given These investigations proved that the method of charging a reverberatory had very little to do with its performance It was also found, at the same time, that the depth and extent of the bath maintained in the furnace had very little to do with its performance Numerous operating results are included and the article is illustrated Development of Copper Converting at Butte and Anaconda. BY WILLIAM KELLY AND FREDERICK LAIST (Trans , vol 106 4400 words.) This paper is based on recollections of one of the authors, W M Kelly, who has been identified with the development of smelting and converting at Butte and Anaconda since early times The article starts with a description of the circumstances surrounding the first intro¬duction of copper converting at the Parrot Works in Butte under the Manhes patents It briefly recounts the difficulties which were encountered and the steps taken to overcome them Among these were the absence of means for punching tuyeres in the original converter This difficulty was overcome crudely by drilling holes in the windbox opposite the tuyeres, said holes being closed by wooden plugs Later these plugs were replaced by steel slides and still later by the Dibblie ball valve, which has remained standard ever since The development of the converter from both struc¬tural and operating standpoints is traced from the early beginning to the present day Developments in Lead and Copper Converting at Tooele. BY B L SACKETT (Trans, vol 106. 3900 words) This paper shows the development at this Utah lead and copper smelter from separate converting of the two kinds of matte produced to a more efficient and economical treatment of the combined mattes This was accompanied by the use of relatively large amounts of silicious ore in the converting of lead matte as well as copper matte, which greatly improved the quality of converter slag produced, both from the standpoint of its copper content and its desirability for

Jan 1, 1935

By A. L. Johnson, S. C. Lyons

A FEW years ago a leading American builder of centrifuges said, "No one uses a centrifuge if the job can be done any other way." This statement was essentially true at that time, not because the basic functional and economic principles involved did not frequently favor centrifugal methods, but rather because so many of the highly specialized features of the mechanical and functional peculiarities of continuous centrifuges had not been fully developed. The idea of using centrifuges in mineral dressing is definitely not new. There are literally hundreds of patents on centrifugal machines and processes for the mining industry. Many of them cover principles that are basic to the functioning of most centrifuges in operation today. In spite of this, until 1935, there were practically no successful applications of continuous centrifuges in the United States in the mineral industry, nor for that matter were there more than a mere handful of such machines in operation in North America in any industry. This was true despite the obvious desirability of machines that could furnish the advantages of high centrifugal settling forces in a continuous operation. The reasons for this are simple, even if not obvious: I. Finely divided solids, when thrown out of fluid suspension under high centrifugal forces, frequently exhibit physical properties (particularly rheological or flow properties) that are markedly different from those they possess when settled out by gravity alone, and machines that are not designed in careful detail to cope with these peculiarities seldom function satisfactorily. 2. Most continuous centrifuges embody a solids-conveying element in the form of a very special screw conveyor. The rotation of this member under conditions of operation calls for drive designs that were not satisfactorily provided until the introduction of the Bird gear unit about 1934. These matters will be touched on later. Continuous centrifuges are now operating successfully in the mineral industry. Some units have singly handled more than 500,000 tons of solids during the past ten years. Generally their operation has been characterized by large unit capacities, complete elimination of labor, minimum supervision, moderate maintenance and power costs. TYPICAL CENTRIFUGE APPLICATIONS Most materials classified by centrifuges are of subsieve particle size. Practically speaking, this means finer than l00-mesh. A unit of measure other than screens is needed in order to consider these products properly. Such a unit is the micron, which is equal to one one-millionth of a meter; roughly 1/25,000 of an inch. The opening in a standard 200-mesh screen is about 75 microns while a 325-mesh opening is roughly 43 microns.

Jan 1, 1947

By Pierre R. Hines

The first progress in American ore dressing practice was made in the mills of the Mother Lode in Calif., the Comstock Lode in Nev., and Gilpin County in Colo., during the years 1861 to 1870, when the ancient wooden stamp battery, with square wooden stems, was converted into the 'California' stamp made of wrought and cast iron, with round stems which rotated. Eli Whitney Blake invented the jaw crusher in 1858 for stone crushing, and it was soon used for crushing ore. F. W. Gates' gyratory crusher followed twenty years later. S. R. Krom, Thomas A. Edison, and others improved the Cornish rolls by increasing their size and spring pressures; the Symons Brothers made horizontal and vertical disc crushers; W. Ball and E. T. Leavitt designed powerful steam stamps for Lake Superior native copper ores; and jigs were mechanized. William B. Frue, a mill superintendent in the Lake Superior copper district, invented the vanner in 1874. Arthur F. Wilfley built his riffled concentrating table in 1895, just after Dr. Robert H. Richards commenced his Ore Dressing. The Wilfley table so changed concentrator flow sheets that Dr. Richards was compelled to revise his text, and its publication was delayed until 1903. The value of sizing or classifying before concentration was recognized. Dr. Richards had persuaded the Anaconda Copper Mining Co. to install round tables with special surfaces to catch the 'extremely fine free mineral.'l Gravity concentration had achieved almost every refinement for recovering or separating minerals by the difference in their specific gravities. However, one serious problem remained unsolved, i.e., how to retrieve the enormous loss of the extremely fine mineral in the slime portion of the ore. Slime, then, as now, was not a precise term due to the variation in the mineral composition of the ores and the different requirements in crushing, grinding, sizing, and classifying. At that time, slime was considered to be fine sand and colloidal material passing 100 mesh.2 Most sulfide ore minerals are more friable than their associated gangue minerals, so that crushing and grinding increased the proportion of the extremely fine ore minerals in the slime. Sparkling specks of galena, chalcopyrite, argentite, or ruby silvers, according to the ore, floating away on the tailing stream from vanners and slime tables, were familiar sights in the old gravity concentrators. Gravity concentration practice was aimed at recovering the ore minerals at the coarsest possible size because the recovery was higher and unnecessary

Jan 1, 1962

By R. I. Jaffee, R. W. Wood, H. R. Ogden, D. N. Williams

D. W. Morgan, D. H. Polonis,and R Taggart (Uuivevsity of Washington)—Dispersion hardening in titanium-copper alloys is of particular interest to us in view of our current research activities in phase transformations in this system. We should like to raise questions concerning several points in the paper. The flow sheet for fabrication and heat treatment (Authors' Fig. 2) indicates that 0.045 in sheet is treated in air for a period of 30 min at 1625' F. Were any surface coatings used to reduce oxygen contamination at this stage? One might ask the same question about the 8-hr annealing at 1400F. Contamination during fabrication and heat treatment of such thin sections could be considerable without some protection. Although the authors state that no analyses were made, they note that contamination masks the distinction between base strengths of the different alloys. In this connection, what fabrication and heat-treatment procedures were used in obtaining the comparison specimens of unalloyed titanium? It would seem that oxygen contamination must cast doubt on the reliability of the proposed strength-mean free path relationship deduced from the data (Authors' Fig. 5). It is usual to evaluate dispersion hardening on the basis of yield strength rather than ultimate tensile strength, and accordingly we have replotted the authors' Fig. 1 to include yield strength. The yield strength data do not justify a discontinuity of 3 p at either 70" or 1000 OF. Can the authors substantiate their conclusion 1) with additional evidence? Otherwise it seems that they have succeeded only in pointing out the possibility of a linear relationship between yield strength (or Uts) and the logarithm of the nleall free path, comparable with that demonstrated for iron-carbon alloys (Authors' Ref. 2). Work in our own laboratories has shown that the hardness of the phase Ti,Cu is 235 Vpn, and that of a 1.5 pct (at.) (2 pct wt) supersaturated solid solution of copper in titanium is 190 Vpn. This small difference in hardness would suggest that Ti2Cu should not be a very effective dispersion-hardening phase. D. N. Williams, R. W. Wood, H. R Ogden, RI. L Jaffee (authors' reply)— This discussion brings out some interesting points with respect to the interpretation of our results, and also raises some questions regarding the experimental technique. We shall consider the latter questions first. Contamination during heat treatment was prevented by performing all heat treatments in argon-filled Vycor capsules. Fabrication was done in air, with the usually observed scaling. Scale was removed

Jan 1, 1962

JESSE L. JONES,* Pittsburgh, Pa. (written discussion?).-The most important criterion on which judgment of the quality of manganese bronze can be based is its freedom from lead. Not over 0.03 per cent. of lead should be present in a good grade of manganese bronze. While it is not questioned that manganese bronze with a high tensile strength and considerable ductility can be made .from materials that will give a lead content of possibly 2 to 3 per cent., there is no doubt that such material, while it may machine readily, will be found to withstand the impact test very poorly, and hence be unsuited for high-grade engineering work. When lead is present in considerable amounts, it forms in small pools in the alloy and the material has little strength at these points; in fact, such spots in the metal are equivalent to nicks in a test bar as far as their effect on tensile strength is concerned. A further objection to a high lead content is the ease with which it oxidizes and its consequent tendency to produce dross. A second characteristic of a high-grade manganese bronze is a low content of manganese. This should not exceed 0.05 per cent.; in fact, very many of the best grades of manganese bronze show, on analysis, only a trace of manganese or no manganese whatever. Manganese is a deoxidizer and its ready oxidation protects the other metals of the alloy; but if it is present in large amounts it will produce considerable manganese dioxide, which does not separate readily from the molten alloy and may produce black specks or drossy spots and render the obtaining of clean castings difficult. Aluminum should be present in a good grade of manganese bronze in only very small amounts. A content of 0.10 per cent. aluminum is ample. Larger amounts have a tendency to increase the shrinkage of the alloy, to give the ingots and castings a white, scruffy appearance, and to prevent the beautiful, golden, oxidization color characteristic of high-grade manganese bronze and a most excellent indication of quality. The work that the author of the paper has done is valuable from a con- servation standpoint, and it will supplement, in a marked degree, the advances that have been made by a number of manufacturers of manganese bronze in utilizing the turnings from this material. A number of firms have succeeded in making a very good grade of ingot metal from turnings and other light scrap, obtaining a tensile strength and elongation

Jan 4, 1919

By J. B. Cheatham

The influence of friction on the force required for an idealized bit tooth to penetrate a "plastic" rock is analyzed. The rock is assumed to obey the Coulomb-Mobr yield criterion and the tooth is represented by a two-dimensional sharp wedge. At the tooth-rock interface the stress field not only satisfies the yield condition but also is such that the ratio of the shearing stress to the normal stress is equal to the coefficient of friction. The solution relates the force on the tooth to the depth of penetration, the tooth angle, the coefficient of friction between the rock and bit tooth and the angle of internal friction and compressive strength of the rock. INTRODUCTION Most rocks are brittle at atmospheric pressure but exhibit ductility and increased strength as the confining pressure is increased. In a previous analysisl the force required for a bit tooth to penetrate a rigid-plastic rock was computed for two limiting conditions: (1) a perfectly lubricated tooth-rock interface which should correspond to a lower limit for the penetrating force, and (2) a perfectly rough interface which should provide an upper limit. Experimental results2 for rock indentation under confining pressure have confirmed that the analysis generally predicts upper and lower limits for the penetrating force as a function of rock properties and tooth angle. The present analysis extends the previous work to account for the friction between the rock and the tooth. ANALYSIS The force required for incipient plastic flow of an idealized model of a rock under a bit tooth is computed subjected to the following assumptions. BIT TOOTH 1. sharp two-dimensional wedge (end effects neglected) 2. rigid tooth (elastic deformation of tooth neglected) 3. displacement of tooth is normal to rock surface 4. friction exists between tooth and rock ROCK 1. semi-infinite, isotropic and homogeneous 2. rigid-plastic (elastic deformation of rock neglected) 3. non-work hardening (horizontal stress-strain curve) 4. Coulomb yield condition* (linear Mohr envelope) 5. smooth, flat surface According to the Mohr theory of failure it is assumed that failure occurs when the state of stress is such that the Mohr stress circle becomes tangent to an experimentally determined envelope. The Mohr failure curve is frequent1y approximated by a straight line and referred to as the Coulomb-Mohr failure criterion. If the rock is under sufficient pressure to be ductile we can use the linear envelope as a yield condition. The Coulomb-Mohr yield criterion is illustrated by Fig. 1 and is expressed by the following equation:l

By W. N. Carr, J. R. Biard, B. S. Reed

An analysis of the semiconductor injection laser is presented which is based on a phenomenological model using device and material parameters. The intent of the laser threshold analysis is not to predict from theory the actual threshold current density but rather to provide a logical means of interpreting experimental results. The device and material parameters have been selected such that they describe both the threshold conditions and quantum efficiency of the lasev. A technique is proposed whereby the current required to obtain a population inversion is predicted from device 1-V characteristics. Experimental data are presented which support the analysis. THIS investigation is concerned with the characteristics of GaAs lasers fabricated in this laboratory. However, the assumptions used in these calculations should not limit their application to GaAs alone. It is felt that the analysis is sufficiently general to be useful in describing the operation of other semiconductor-laser materials. ANALYSIS heavily doped semiconductors to be considered. The absorption and/or generation of optical radiation may be completely specified in terms of q(E) and the probability functions f,,(E) and f,(E) which define the occupancy of the valence-band and conduction-band levels, respectively. The symbols used in this analysis are listed at the end. Following the approach of all' it can be shown that the effective gain coefficient, g, in the plane of a forward-biased P -N junction is g(E)=-ai@)[1-fv(E)-fc(E)] [1I In particular Eq. [I] holds for the photon energy at which lasing occurs, Ep. Throughout the following discussion Eq. [I] is considered only for the lasing transition E = Ep. For simplicity the functional energy dependence of the various factors is omitted in subsequent equations. Unless otherwise stated the symbols used represent the values at Ep. If the analysis is limited to the P-type side of the laser junction, the function q(l -f,,) is recognized as the band-to-band absorption in the P-type material at thermal equilibrium. Under forward-bias conditions the function f, is determined by the electron quasi Fermi level in P-type material. For sufficiently large forward bias the second term in Eq. [3] will become dominant and net gain will result; i.e., g will become positive. In order to simplify the subsequent analysis several assumptions are in order at this point. a) No conductivity modulation occurs. /„ is independent of f, and therefore go is a constant. b) The P-type material is heavily doped such that the degeneracy of the valence band permits popu-

Jan 1, 1964

By M. C. Lavine, H. C. Gatos, R. S. Mroczkowski

The structure of GaAs-Ge heterojunctions prepared by a back-melting process was studied by X-rav diffraction, melallographic, and electron-micro-analyzer techniques. The boundary region between the GaAs and the germanium was found to consist of a four-layered band. A discontinuity between the two inner bands was brought about by solidification occurring simultaneously from the GaAs and the germanium side. The junction material OH either side of the central discotinuity was found to have grown as a single Crystal with the orientation of the parent substrate. It was found essential to determine the GaAs-Ge phase diagram. This was accomplished by a series of alloys containing 0 to 40 wt pet GaAs. Over this range the diagram is a simple eutectic with the eutectic point at 74 wt pet Ge and 845 C. The limiting- solubility of Coils in germanium was found to bc less than 5 wt pet. The Kossel divergent-beam X-ray technique was used to study orientation relationships across the boundary lager. A great deal of effort has been recently expended on the electrical properties of heterojunctions. Little attention, however, has been given to the metallurgical variables which may be of importance in the formation of the junctions. The present work was undertaken to investigate some of the more pertinent metallurgical aspects of interface-alloyed GaAs-Ge heterojunctions. Such parameters as solidification processes, composition gradients, orientation effects, and the accommodation of crystal structures have been the objects of study in this program. EXPERIMENTAL PROCEDURES Phase Diagram. The phase diagram of the GaAs-Ge pseudobinary system was determined by means of cooling curves. A Bridgeman furnace was employed with the center section maintained at 1200°C and the upper section at 800°C. Melts varying from 0 to 40 at. pct in GaAs content were used. Each charge was incapsulated under a two-thirds atmosphere of argon in a carbon-coated quartz tube. Five min in the furnace hot zone provided rapid melting and homogenization. The capsule was then removed to the cooler portion of the furnace. Cooling times were of the order of 3 min. The resulting homogeneous microstructure indicated that equilibrium conditions were obtained by this procedure. Formation of Heterojunctions. Rectangular parallelepiped single-crystal samples were cut with the large faces of either the (l00), (110), or (111) orientation. The remaining four surfaces were also cut to specific orientations to permit accurate relative orientation of the GaAs-Ge pairs. The

Jan 1, 1965

By R. J. Allio, C. H. Randall

An automatic system for the quantitative analysis of microstrulcture has been developed. The method is based upon a statistical model of microstructure and employs phase intercept fractions as fundamental random variables. Structure-sensitive properties can be related explicitly to parameters derived from the distribution of phase intercept fractions in a matevial. THE properties of mechanical dispersions and of multiphase alloys art: intimately related to the amount, distribution, and intrinsic properties of the dispersed material. In many instances, the amount and distribution of the dispersed substances have greater influence on the over-all system properties than do the properties of the dispersed substance-Such properties as average particle size, particle size distribution, homogeneity of dispersion, volume per cent of dispersed phase, and mean free path between dispersed particles thus become critical in any evaluation of the properties of a dispersion. Unfortunately, techniques for evaluating quantitatively the microstructural characteristics of dis~ersions rapidly and accurately have not been available. If necessary, qualitative or semiquantitative devices can be utilized. Some such devices are based merely upon comparison of the specimens in question with arbitrary standards. Other techniques involve superposition of grids and tabulation of particles observed in different portions of the grid. Still other methods based upon lineal analysis entail counting lines or points at random upon the specimen to be examined; the number of intersections of the line with a given phase, the proportion of line or number of points which are covered by given phase are variously recorded. The mathematical basis for most of the techniques of quantitative metallography have been amply reviewed in the literature.'l2 without exctption, however, the standard techniques for quantitative metallography are either imprecise or are exceedingly tedious to apply. For example, tranverse integrating stage for rapid manual areal fraction measurement has been designed by Smith.3 Howerer, as in the case of any device based upon visual eslimation, operator bias may cause appreciable error. Various flying spot devices, on the other hand register particle size distributions in 2-dimensions, from which it is difficult to derive other properties of the dispersion. Thus, e.g., the 3-dimensional distribution cannot be inferred unless certain rather broad assumptions are made concerning the randomness of distribution and geometrical configuration of dispersed phase. Many limitations to the application of quantitative structural analysis can be overcome by the use of the microdensitometer, which is essentially an automated areal analysis device. Data obtained in the form of a distribution of phase intercept values are

Jan 1, 1962

By W. D. Robertson

The effect of a tensile stress in accelerating the corrosion-cracking of certain alloys of aluminum, magnesium and iron is widely recognized. The literature is extensive and it is only necessary to cite the recent ASTM—AIME Symposium,l There are, however, no experimental data conclusively showing the effect of an applied compres-sive stress. It has been inferred, from the generally accepted mechanism postulated for the acceleration of corrosion by an applied tensile stress, that a compressive stress would inhibit corrosion. Furthermore, it is observed that corrosion cracks invariably start on the tension side of a specimen bent as a simple beam. This observation has led to the suggestion that shot-peening of the surface would inhibit or eliminate stress-corrosion cracking in magnesium alloys and aluminum alloys. However, the degree of inhibition produced by shot-peening does not appear in the literature except for the above qualitative references which are unsupported by experimental data. Moreover, while peening undoubtedly imposes a surface compressive stress, the alteration of the surface by the severe cold-working involved complicates the interpretation of the results with respect to the effect of a compressive stress alone. If corrosion proceeds along preferential paths produced by precipitation from solid solution, it is easily understood how a component of tensile stress normal to these paths will accelerate the corrosion and lead to ultimate failure. On the other hand, if the metallurgical conditions are such as to create preferential corrosion paths, then it is difficult to understand how a compressive Stress, produced by shot-~eening Or any other method, can inhibit corrosion except in the sense that the accelerating factor is removed. Thus, if the basic corrosion mechanism remains unaltered, there should be little or no inhibition resulting from the application of a compressive stress if inhibition is defined as the reduction of the degree of corrosion below that experienced in the unstressed state. The experiment described below was performed to isolate and evaluate the effect of a compressive stress on the corrosion and stress-corrosion cracking of a high strength aluminum alloy (24-S); and, by a Comparison of the corrosion rate of 'pecimens subjected to an applied tensile stress, compressive stress and in the unstressed state, to clarify the role of stress in intergranular corrosion. Experimental Procedure The experiment was conducted with the aluminum-copper-magnesium alloy known as Duralumin 24-S.* Specimens, 0.064-in. thick, in the form of I-in. wide strips sheared parallel to the rolling direction,

Jan 1, 1949

By E. S. Wheeler

The conversion plant of the Clirnax Molybdenum Co. is at Langeloth, Washington . County, Pennsylvania, approximately 30 miles west of Pittsburgh. The molybdenite concentrates converted originate in the company's mine and mill at Climax, Co1orado.l They are delivered to the Langeloth plant by rail packed in 175-lb. bags. War production of concentrates at the mill started in 1918 but it was not until about 1920 that any peacetime uses of the metal began to develop. Early conversion of Climax concentrates was mainly into ferromolybden um, produced for the company by a custom smelter. Chronology of Plant Development In 1924 a small hand-rabbled roaster was built at Langeloth, the roasted concentrates being used principally in the production of calcium molybdate by a conventional wet method. The size of the original plant is indicated by the fact that during 1925 about 400,000 lb. of molybdenum was converted. In 1925-1926, a multiple-hearth roaster was erected (12-ft. 0.d. 8-hearth Nichols-Herreshoff); also the first experimental ferromolybdenum was produced at Langeloth. At this time sa.1cium molybdate was produced in the furnace by a process to be described. With continued interest in and increased use of molybdenum, it was necessary to arrange for enlargement of plant facilities. Although only one 12-hearth 16-ft. 0.d. roaster was erected at this time (1929), plans were prepared for a duplicate unit, which was completed in 193 5, giving a potential roasting capacity of some 16,000 lb. Mo per day under the then prevailing roasting practice. Shortly after this second unit had gone into production, increased demand overtaxed the plant, and as a result two 12-hearth 18-ft. 0.d. roasters were placed in operation in 1936 and 1937. This gave the plant a potential roasting capacity of 40,000 to 50,000 lb. Mo per day, and necessitated major changes in auxiliary handling equipment as well as improved facilities for producing ferromolybdenum. Following the introduction (1938) of a new product—molybdic oxide briquets— and because of the constantly growing demand for molybdenum, plant facilities were again enlarged in 1940 and 1941 by the erection of two 16-hearth 18-ft. 0.d. roasters. These two new units, along with a change in roasting practice, brought the potential roasting capacity to 150,ooo to 190,000 Ib. Mo per day, based on about 15 lb. per sq. ft. of hearth area. Other departments of the plant have necessarily grown, until today the plant is modern throughout and completely mechanized, with equipment geared to produce marketable products as follows:

Jan 1, 1944

By Gerald Edmunds

In a paper1 before this Institute in 1940, the writer reported that the surface orientation texture of zinc and cadmium differed from the texture existing within the casting, in that basal planes were parallel to the surface. He was unable to rationalize the , origin of this surface texture. It is conceivable that the surface orientation texture is influenced by the atomic force fields, which are different at the surface and within the body of metal. It was considered that a radical changc in the surface against which the metal is cast might alter the surface texture. In trial of this, zinc has been cast against a surface of molten lead, but the same result has been obtained—i.e., (001) parallel, ± 2oo, to the surface—as was obtained when it was cast against air, Transite or iron.' Into the bottom of an open cast-iron tray, 7 by 14 in. by 2 in. deep, preheated to about 275°C., about 1/4 in. of lead, then 1/2 in. of zinc was poured, and the whole was covered with an electric heater to enforce freezing from the bottom upward. After a lapse of time sufficient for the zinc but not the lead to solidify, the zinc slab was raised free of the molten lead. The bottom surface of the zinc casting was coated with a thin layer of lead. A specimen of the slab 136 in. square was metallographically polished to flatten it and remove the slight film of lead, and the orientation was determined by the method used before,' with the result stated above. Some determinations have also been made on other metals, as follows: Surface orientation determinations were made on the 3/8-in. diameter grip section of two 1/4-in. diameter aluminum-alloy die-cast tensile bars and on a 1/8 by 1/2-in. flat aluminum-alloy die-cast tensile bar. None of these three aluminum die castings exhibited any preferred orientation at the surface. A brass die casting (nominal composition: 60 Cu, 37.75 Zn, 1.00 Sn, 1.00 Pb, 0.10 Al, 0.15 Si) furnished by C. S. Smith was also examined. The surface showed a random orientation of alpha and beta. Summary A zinc casting solidified against a molten lead surface was found to have the same surface orientation texture, (001) parallel to the surface, as other zinc and cadmium castings. Aluminum and alpha-beta brass die castings were found to have random grain orientation textures at the surface. Acknowledgment The X-ray determinations were conducted by M. L. Fuller.

Jan 1, 1945