Search Documents

By S. A. Herres, A. R. Elsea

Temper brittleness refers to the loss in the notched-bar impact resistance encountered in most medium- or low-alloy steels when they are tempered within the temperature range of 700 to ll00°F or slowly cooled from a higher tempering temperature. Temper brittleness is important because it impairs the service properties of otherwise suitable steels over the range of hardnesses obtained at medium tempering temperatures and because it is impractical to cool pieces of large section rapidly enough from higher tempering temperatures to prevent temper brittleness. The problem of temper brittleness has always been closely associated with ordnance manufacture because armor plate and gun steels require good shock resistance and are often heat treated in heavy section. However, it is now realized that many brittle failures in structural alloy steel parts are attributable to temper brittleness, and there is little question that it contributes to the low impact resistance of alloy steels used in the hot-rolled or normalized condition which are slowly cooled from the hot-rolling or normalizing temperature. Recently Hollomonl has presented a comprehensive summary of the literature on the subject and an excellent description of the manner in which temper brittleness affects the notched-bar impact resistance and other mechanical properties of steel. Carpenter and Robertson2 have given a good analysis of the probable mechanism of temper brittleness, which is similar in many respects to the precipitation-hardening phenomenon. Fig 1 is their diagram illustrating the response to tempering of a susceptible steel. Three changes are recognized as the temperature of tempering a previously quenched steel is raised: (1) pre- cipitation occurs more rapidly and to a greater extent and the notched-bar impact value falls to a minimum; (2) coalescence of the precipitated particles begins to be effective and the impact values for both rapidly cooled and slowly cooled specimens increase; (3) the precipitate goes back into solution at higher tempering temperatures and the impact values of specimens cooled rapidly enough to retain the solution increase, while reprecipitation occurs in slowly cooled specimens and their impact values decrease again. No change must proceed to completion before the following one begins and there is overlapping in temperature ranges. Although there have been a large number of experimental investigations of temper brittleness, insufficient evidence has been accumulated to permit definite identification of the precipitate wnich causes temper brittleness nor have the specific effects on it of various alloy elements and impurities in steel been definitely established. This lack of evidence is largely a result of the lack of a suitable measure for temper brittleness. Temper brittleness is evidenced by an increase in the temperature of brittle failure of notched-bar impact specimens, but most of the data are in the form of an arbitrary susceptibility ratio of the room-temperature impact value for specimens quenched from a high tempering temper to that of similar specimens furnace cooled from the temper. It is obvious that this numerical ratio is dependent upon the type of specimen used and the specific conditions of testing rather than the characteristics of the steel. In general, manganese, chromium, and nickel alloy steels are most susceptible to temper brittleness and molybdenum has heen used to decrease

Jan 1, 1950

By D. J. Freeh, W. E. Horst, R. C. Kellner

This paper is concerned with the description of grinding characteristics in a batch grinding system. A mathematical model was developed and used for simulating the system on an analog computer. A general equation describing the change of selected size fractions with respect to time is presented. The empirical verification of the model is discussed. In order to understand the performance of a grinding system, it would be desirable to have a mathematical description of the system's behavior. The development of a mathematical model to describe the performance of a batch dry-grinding operation was investigated. This is the first step toward reaching the objective of a practical mathematical description of a grinding system that could be employed for process control and optimization of a simple or complex grinding circuit. Although this article is concerned only with batch grinding performance, experimental work is being conducted to extend the model to continuous grinding systems. BASIC PROCESS CONSIDERATIONS In a paper dealing with the probability theory of wet ball milling, Roberts1 correlated the rate of change of oversize material, C, with energy input, E, as follows where the energy input is expressed in power/unit mass and k is a constant for each particle size. Experimental grinding studies were made by Bowdish2 to extend the previous work of Roberts. The work of Bowdish also showed that the rate of breaking of oversize particles in a ball mill is proportional to the concentration of oversize particles present. He further demonstrated that breaking of oversize particles is proportional to the surface area of the balls used for grinding. Rate equations were presented and it was concluded that grinding could be considered as a first order phenomenon according to the rate equation [2] di which is in accordance with Eq. 1. In Eq. 2, the effect of ball area is included in the constant k,. Other rate equations were reported to show the effect of a change in ball area and removal of fines during grinding. Correlation studies made by Arbiter and BhranY indicate a first-order rate equation for quartz; however, their data showed less than first-order rates existed for some of the other materials evaluated. More recently Kelsall4 has shown that the rate of breakage of any narrow size range of quartz could be described by a first-order rate equation. Several other investigators5-7 have described grinding behavior in terms of rate equations or kinetics. In general, the rate equations have been described for each particle size; however, the interactions between sizes and the overall relationships that exist have not been considered, at least not in the form of solving a group of simultaneous differential equations. Extensive work has been conducted by various investigators using a probabilistic approach which has expanded stein's' earlier work. Selection and breakage matrices are incorporated in the probabilistic model as contrasted with the deterministic method presented here. A batch grinding operation may be considered as an unsteady-state process from the standpoint that the conditions within the mill are a function of time. Therefore, a plot of mass fraction of the i-th size versus time would give a set of curves as shown in Fig. 1. In these curves X , represents the mass fraction of large size material, X2 intermediate size, and X3 small size material. An examination of these curves shows a strong similarity to the curves that would be obtained from a chemical reaction of the type9 A?B?C, [3] where A would be comparable to the largest size material being broken to an intermediate size, 3, and

Jan 1, 1968

By Linn M. Farish

The reported production of petroleum in the Dominion of Canada for 1930 was 1,555,199 bbl., an increase of 455,373 bbl. over 1929. Alberta accounted for nearly all the production with Ontario and New Brunswick, the other two productive provinces, contributing less than 1 per cent, of the total. The completion of about 25 producing wells in Alberta was more than sufficient to offset the decline of the older wells in that province and the small decline in Ontario. The production of about 5000 bbl. in New Brunswick represents a restricted output to meet the market demand. Table 1 shows the production by provinces for the past 10 years. The consumption of natural gas for all purposes in Canada reached a new record of practically 40,000,000,000 cu. ft. in 1930, which was a considerable increase over 1929. The estimated value of the gas output was $10,500,000, against a little over $5,000,000 for the total petroleum production. Gas wells in Alberta accounted for over 75 per cent. of the total, with Ontario second and New Brunswick a poor third. The outlook for 1931 is for very little change in production of either gas or oil in Ontario and New Brunswick but for a substantial increase in both in Alberta.

Jan 1, 1931

By L. A. Weakly

The Viburnum ore trend is the largest, single known ore body of lead in the world. Four well-known mining companies operate in the trend, which is approximately 64 km (40 miles) long and up to 610 m (2,000 ft) wide, located some 121 km (75 miles) southwest of St. Louis, MO. Total production from the mines for 1979 was 33 kt/d (36,000 tons) of ore or 472 kt/a (520,350 tons) of lead and coproduct metal. The district is the number one producer of lead, providing 90% of the United States' output with its coproducts, zinc ranking second, and copper and silver both ranking sixth in total US output. Mining is done in a stratiform deposit approximately 305 m (1,000 ft) in depth with the host rock principally dolomite. Ore is mined in large open stopes by the method commonly referred to as room-and-pillar mining. Mining methods and equipment are fairly similar, with the mines classified as trackless, all mining equipment operates on rubber tires. This paper will present a description of the different methods and equipment used to move ore from the active heading underground to the ore shaft for hoisting to the surface. Basically, three methods are incorporated: front-end loader and truck haulage combination; front-end loader and truck, with main line track haulage combination; and front-end loader and truck, with conveyor haulage combination. The logistics of each system, such as method of use, equipment specifications, ore handling rate, and those cost figures that could be made public, will be included.

Jan 1, 1983

By Orville R. Lyons

THE Russellton preparation plant of Republic Steel Corp., located north of Pittsburgh, Pa., prepares 395x0-in. Thick-Freeport coal by means of a 13-ft 6-in. diam Chance cone and 16 No. 7 Deister tables. The plant feed ranges between 370 and 410 tons per hr, of which approximately 31 pct is 1/4xO-in. size, and the plant reject ranges between 14 and 16 pct of the feed.

Jan 6, 1953

By B. J. Kochanowsky

TO improve blasting methods it is necessary to know how the explosive force acts and how rock resists this force. Because of the tremendous power developed within milliseconds and the great number of other factors directly affecting the technical and economic results, an analysis of the fundamentals of blasting theory is difficult. But since the rules used for layout design and for calculations of size of explosive charges are based on theoretical assumptions, complete knowledge of blasting theory has great practical importance in mining. Analysis of Blasting Theory: It is interesting to note the opinion of blasting experts with respect to contemporary blasting theories. F. Stussi; Professor of the University of Zurich, stated: "We do not have enough experience yet to change our army engineering regulations in blasting and base it on new fundamentals. It is our duty to collect more practical data and to do more research in blasting to close this gap." K. H. Fraenkel,2 editor of the Manual on Rock Blasting published in 1953 in Sweden and written by well-known Swedish, German, Swiss, and French blasting and explosive experts, said: "To the best of our knowledge no suitable formulas for civil blasting work are to be found in the American, French or German literature." Present blasting theory is based upon two assumptions. 1) The blasting force of explosive acts in concentrical and spherical form. 2) Rock resistance against the explosive force is directly proportional to the strength characteristics of the rock. The first classical formula based on theoretical fundamental in blasting theory for explosive charge calculation was introduced by Vauban, a military engineer who lived 300 years ago. It was Vauban who proposed the famous formula L = w3 q, where L is the explosive charge, w = line of least resistance, and q = specific explosive consumption proportional to the weight of rock. Later engineers used q as proportional to the strength of the rock. Since Vauban's time different suggestions concerning blasting theory have been proposed. However, the principles stated at that time so affected the thinking of later generations that his formula is still in use and practically unchanged. The first controversy concerned the form of crater. It was found that geological features of rock affected its form. The factor q was analyzed thoroughly by Lares3 and later by Ohnesorge," Weichelt,5 Bendel,6 and others, but the assumption remained that resistance against explosive force is directly proportional to the strength of the rock blasted. The greatest controversy, which has not yet been settled, concerned w. It was noted that w3 is more appropriate for long lines of resistance and w2 for lines of resistance less than 15 ft. Based on the assumption that the explosive force acts concentrically and spherically, spacings between charges were limited to distances not greater than the length of line of least resistance. Sometimes larger spacing is recommended, but this is due to the advantageous geological and physical properties of rock and not to the action of an explosive force as such. In addition to the classical formula, empirical formulas are used widely. These state that the explosive charge is directly proportional to the volume of blasted rock in cubic yards, and the amounts of explosive required are usually expressed in pounds of explosive per cubic yard of rock. Empirical and classical formulas are contradictory. In the empirical formula, but not in the classical formula, explosive charge is taken proportional to all three space axes: line of least resistance, spacing, and bench height. In spite of this contradiction, both formulas give good results. This is possible because as now practiced the explosive charge calculation for heavy burdens need not be highly accurate. Each, open pit or quarry, usually works with a certain relation between bench height and line of least resistance and between charge spacing and line of least resistance. When these relations are changed, however, the specific explosive consumption q changes greatly. This is one of the reasons why the principles on which the formulas are based appear to be incorrect. In addition to the formulas discussed, others exist and are based more or less on the same theoretical

Jan 1, 1956

By Claude H. P. Lupis, John F. Elliott

The classic Siez~erts' technique has been employed to determine the solubility of oxygen in liquid pure silver and the effect on the oxygen solubility of solute metal additions to liquid silver. The results indicate that oxygen deviates slightly from Sieverts' law in pure silver as = 3.9. Gold, platinum, md palladium decrease the solubility of oxygen in silver. First-order free energy, enthalpy, md entropy interaction coefficients are determined from the data. An important aspect in the establishment of the formalisms to be applied to represent the thermody-namic behavior of components of solutions is the accuracy of the experimentally determined thermody-namic properties. Wagner developed the first-order free energy interaction coefficient.' Recent papers2'3 extend the treatment formally to other thermodynamic properties and define the second- and higher-order interaction coefficients. A current communication4 considers in some depth the requirements and limitations imposed because of experimental error on the use of several formalisms now under active consideration in the field. This communication reports the results of an experimental study of the behavior of oxygen in liquid silver alloys that was carried out concurrently with the development of Refs. 2 and 3 with the hope of providing sufficiently precise information so that enthalpy and entropy interaction coefficients could be determined. THE EXPERIMENTS The Sieverts technique was employed using an improved design of the reaction bulb over that developed originally by Humbert5 and subsequently improved by pehlke6 and Weinstein.7 The Vycor reaction bulb as shown in Fig. 1 was air-cooled and contained a port for the thermocouple angling off to one side and a vertical port for connections to the gas supply and

Jan 1, 1969

By J. E. Gosline

IN any theory of a hydrodynamic nature dealing with the vertical flow of gas-liquid mixtures in pipes, the two factors that present the greatest difficulty are the relative motion between the phases and the energy loss resulting from turbulence work. The former has generally been referred to in the literature as "slippage" and the latter as "fric-tion." A third factor of importance is the nature of admixture between liquid and gas for both the relative motion between the phases as well as the turbulence work are functions of the type of mixture. Although much work of a general nature has been performed on the vertical flow problem, there have been very few attempts made to study experimentally the factors involved in the phenomenon individually. Of these, perhaps the most notable is the work of Moore and Wilde1 on slippage. Some writers on the subject have assumed that over a wide range of operating conditions the controlling type of mixture, is one in which gas bubbles of uniform size are distributed throughout the liquid, and that the data obtained from experiments on the motion of single bubbles through various liquids may be used in evaluating the slippage. Ver-sluys2 has further assumed another controlling type of mixture in which the liquid occurs as a dispersed phase within the gas and applies data on the fall of drops of water in air to calculating the relative velocity. It was for the purpose of studying the applicability of data on single bubble motion to the slippage factor and to further observe the nature of mixtures that a research program was conducted at the University of California during 1932 and 1933. These experiments formed a basis for

Jan 1, 1935

By J. B. Rather

The use of sulfuric acid in refining illuminating oils antedates the beginning of the petroleum industry in America by many years. It was used as early as 1792 by Tower in refining "coal oil" in the British Isles, and by 1860 there were 53 refineries in the United States manufacturing "coal oil," most of them using sulfuric acid as a refining agent. Following the completion of the Drake well in 1859, many of the "coal oil" refiners began to use petroleum for manufacturing illuminating oil, and their methods and equipment were adapted so far as possible to this new raw material. The value of sulfuric acid in refining L'coal oil" having been established, this reagent was immediately tried and found to be satisfactory in refining petroleum. Notwithstanding the tremendous advances in petroleum technology since that time, sulfuric acid remains today the most valuable and most widely used means for refining petroleum products. Early Methods of Disposal of Acid Sludge With the almost immediate expansion of the petrolrum industry came the problem of disposing of the acid sludge produced in the refining operations. While in 1859, the year of the discovery of the Drake well, a patent was granted for sludge separation and separated acid concrntration, most of the earlier refiners, and rnany of their successors, felt that separation and recovery of spent acid was uneconomical. Accordingly, many of then) disposed of the sludge in any way that seemed expedient nt the time. It was huricd in the refinery grounds, harged and tlurnped at sea, run into neighboring streams, or dumped into sewers. Some optimists carted it to open pits in the hope that it would eithcr sink into the ground or evaporate. Not so many years ago, the superintendent of nn eastern refinery, who was using the open-pit method of disposal, lost, a sludge cart, with horse attached, in such a pit. The horse was recovred almost immediately, but its availability for further use was reduced to zero by the loss of its hair.

Jan 1, 1928

By E. Keller

Some years ago the writer's attention was called to the fact that rolling-mill scales from auriferous copper do not have the gold content proportional to the gold contained in the copper from which they are derived.' So, later, he made a few simple experiments, which are described herewith. In order to test the gold movement on the changing surface of a copper plate under the influence of oxidation, plates 6 in. by 8 in. by 1 in. (15.2 cm. by 20.3 cm. by 2.5 cm.) of converter copper, refined, of 99.29 per cent. purity, and leaded copper with a content of 95.65 per cent, copper, were obtained. Ten cuts were made through the thickness of the plates along the longitudinal center and the silver and gold content ascertained in each of the 10 samples in both plates, as shown in the accompanying illustration. The assay results, recorded in Tables 1 and 2, show that the gold distribution is relatively uniform in both plates, while silver has greater variations, especially in the leaded plate. The plates subjected to oxidation were placed in a red-hot muffle for about 2 hr., two small scarifiers being used as supports. Arranged alongside each other, they nearly filled the width of the muffle, and so were practically under equal conditions of heat and draft. The accurate observance of the time was not possible, however, because of other work, but the time of the exposure of each plate in a pair was identical. At the end of each period, each plate was plunged into water and completely freed of scale, which was carefully collected, dried, weighed, and analyzed. Samples of the plates had also been taken of the molten metal and duly analyzed. Table 3 contains the analytical data for the two plates and their several sheets, or layers, of oxide scales. These figures show that in the oxidations of the converter copper the silver is fairly uniform, while the gold is highly irregular and especially low in the first layer. In the oxidations of the leaded copper plate, the silver and gold contents are variable; however, this irregularity is due to two different causes. The

Jan 1, 1919

By Wolfram Ruff, H. A. Schwartz

THE spheroidal form of the temper carbon nodules in malleable cast iron and of the graphite mottles of "mottled" cast iron suggests that in both all the graphite in a given mottle or nodule grew from a single center of crystallization or nucleus. ‡ There is, however, a distinct metallographic difference between the mottle and the nodule; the area of mother metal impoverished in combined carbon by the crystallization of graphite in a mottle is practically coextensive with the ramifications of the graphite, while the impoverished zone of metal corresponding to a given nodule is always larger, and usually much larger, than the approximately spherical mass of graphite crystals forming the nodules. Furthermore, the mottle probably is formed during freezing and the nodule almost always at a considerably lower temperature. Assuming that the nodules or mottles can be used as a measure of the existing nuclei (capable of germination) we propose to record certain observations as to the graphitizing reaction in the light of the number and habits of growth of these nuclei. The present authors, in cooperation with others 1-7, have reported upon various cognate aspects of the problem. The fact that temper carbon is a mass of flakes and not a solid granule has become well recognized since the introduction of an improved polishing technique8,9,10 for cast iron. The problem of deriving from the circular planar intercepts

Jan 1, 1935

By A. J. W. Headlee

RECENT interest in germanium centers around its use as a semi-conductor of electricity in electronic devices. It is generally believed that germanium transistors, diodes, triodes, photocells, and rectifiers will largely replace the present-day vacuum tubes. The germanium electronic devices are much smaller, lighter, and more resistant to shock, and they require only a small fraction of current required for conventional devices. Hence only a small amount of heat is given off by the equipment. This development will require many times the quantity of germanium now being produced, which in 1948 was approximately 1000 lb1 and was obtained almost entirely from zinc ores in Missouri, Kansas, and Oklahoma. No one knows with certainty what the demand will be for germanium by 1956. It may be 40,000 lb.1 Known concentrations in zinc ores are not sufficient to

Jan 10, 1953

By V. C. Allison

The explosibility of a dust is favored by its fineness and content of volatile combustible matter; its explosibility is decreased by its moisture and ash content. The explosibility of a dust can be reduced by admixture of either inert dust or water. The major effect of both the inert dust and water lies in their heat capacity; they absorb the heat of the explosion flame until the temperature of this flame drops below the ignition temperature of the explosive heat. The amount of inert dust necessary to make an explosive dust non-explosive can be calculated from its proximate analysis. FOREWORD LARGE-SCALE testing of the explosibility of coal dust as conducted by the Bureau of Mines in its Experimental Mine involves a large initial investment, and a high charge for maintenance and conduct of the test; in addition the work is necessarily slow because of the time required for the preparation of the tests and the cleaning and repairing of the mine after each test. These tests have been made along practical mining lines that admit of a range of mixtures of dust varying from 5 to 10 per cent. and with gas as low as 0.5 per cent. increments. Because of the time and expense involved, efforts have been made to devise a type of gallery for laboratory use in which the results obtained in the Experimental Mine may be correlated as an aid in a study of dusts, the explosibility of which might be approximately determined quickly and at a minimum of expense. The explosibility of coal dust has been studied by a number of European investors, using small laboratory devices; among these may be mentioned the experiments con¬ducted by Galloway, Vital, Hall and Clark, Abel, Mallard and Chatelier, Thorpe, Bedson and Widdas, Engler, Holwartz and Meyer, and Taffanel, results of whose experiments have been reviewed by the Bureau of Mines., The Bureau has conducted many tests with the Frazer-Clement2 apparatus, which gives the pressures set up when the dust cloud within is, ignited with a platinum coil or tube, heated to different temperatures, but for these tests the sample must be especially prepared as to size (200 mesh) and must be dry. An attempt has been made to design an apparatus that may be used for checking and correlating the experimental tests. The first design was a wooden gallery, 8 by

Jan 7, 1925

By R. E. Gilchrist, R. R. Harvey, M. R. Dean

A simple and rapid three-component analysis procedure has been developed for use in fluid-flow studies. The method is based on refractive index measurements, combined with refractive dispersion measured between the Ha and H wave lengths. The procedure requires. calibration curves for the three-fluid system. The only restriction on the fluids employed is that only one be an aromatzc or olefinic compound. Individual sample measurements require only a few minutes. Composition results within the limits of accuracy required for fluid-flow analog studies are obtained. INTRODUCTION The simultaneous analysis of three-component mixtures is a problem often encountered in fluid-flow analog systems. In petroleum production research, the problem is usually encountered in experiments on the displacement of fluid from a porous medium. In the case of immiscible displacement, simple volumetric measurements of the produced fluids are usually sufficient, the analog fluids often being the actual reservoir fluids themselves — oil, water and gas. However, modern secondary recovery of petroleum sometimes involves production under miscible conditions, in which the oil is displaced by liquefied petroleum gas which, in turn, is displaced by dry gas. Since this type of displacement is done under conditions of pressure and temperature that are not easily reproduced in the laboratory, studies are often performed with analog fluids at essentially atmospheric pressure. Since the number of samples taken during one of these experiments is usually quite large, the time required for analysis becomes an important factor. In the past, the difficulty of conventional analysis has been avoided by the addition of various tracers to one or more of the components, or by the use of a tracer fluid as one of the components. With these techniques, a relatively large amount of analytical time is still required. Moreover, in the case of added tracers, the tracer material may migrate in part to one of the other fluids, may be adsorbed on the solid surfaces of the model, or may introduce side effects not characteristic of the prototype fluid. The substitution of a tracer fluid for one of the components can severely limit the choice of analog fluids. Optical refractometry provides a rapid and economical method of analysis of two-fluid systems or of monitoring small changes in composition of an essentially one-component system. Refractometric methods are capable of high degrees of accuracy, require only very small samples and can easily be adapted to automatic methods. They are not generally applicable, however, where three or more fluids are present over wide ranges of concentration. The method of three-component analysis described here has three essential characteristics. 1. It is based on refractometric measurements and thus preserves all of these desirable properties. 2. It does not make any stringent demands on the choice of fluids that can be analyzed. 3. It is capable of accuracy sufficient for fluid-flow analog studies. THEORETICAL The difference in the refractive index of a substance at two different wave lengths of light is known as its dispersivity. When the difference is taken between the specific refractions at the two wave lengths, the quantity is referred to as the specific dispersivity. Specific refraction was defined empirically by Gladstone and Dale1 as r = n-1/d....[1] where n is the index of refraction and d is the density. The definition is given more correctly by the expression r =n2-1/n2+1 1/d..........(2) 1References given at end of paper. MICIFTV OF PETROLEUM ENGINEER JOURNAL.

By J. P. Hammond, C. J. McHargue

The wire textures for cold rolled and recrystallized iodide titanium and the sheet textures for this material produced by cold and hot rolling, and recrystallization at a series of temperatures were determined. &apos;The effect of the a + ß transformation on the sheet texture was noted. UNTIL recently it was believed that all hexagonal close-packed metals deformed by slip on the basal plane, (0001), and that rolling should tend to rotate this slip plane into the plane of the rolled sheet. The pole figures of cold rolled magnesium&apos; are satisfactorily explained on this basis. There is a tendency for the <1120> directions to align parallel to the rolling direction, and the principal scatter is in the rolling direction. Zinc% as a rolling texture in which the hexagonal axis is inclined 20" to 25" toward the rolling direction. Twinning is believed to account for the moving of the basal plane away from parallelism with the rolling plane. The texture of beryllium3 places the basal plane parallel to the rolling plane with the [1010] direction parallel to the rolling direction, and the scatter from this orientation is primarily in the transverse direction. Cold rolled textures reported for zirconium&apos; and titanium5 how the [1010] directions to lie parallel to the rolling direction and the (0001) plane tilted by approximately 25" to 30" to the rolling plane in the transverse direction. Rosi has recently reported that the mechanisms for deformation in titanium are distinctly different from those commonly reported for hexagonal close-packed metals. The principal slip plane is the prismatic plane, {1010), with some slip also occurring on the pyramidal planes, (1011). However, there is no evidence for basal slip. The slip direction is reported to be the close-packed digonal axis, [1120]. In addition to the twin plane commonly reported for metals of this class, {1012), Rosi found the twin planes (1122) and {1121), with the dominant twin plane being (1121). Information regarding the recrystallization and hot rolling textures of hexagonal close-packed metals is limited. Barrett and Smigelskas report that rolling beryllium at temperatures up to 800°C and recrystallization at 700°C produce textures not differing from the cold rolled sheet texture.3 McGeary and Lustman find that hot rolling at 850°C produces the same basic texture in zirconium as rolling at room temperature.&apos; These investigators also report that the texture for sheet zirconium recrystallized at 650 °C differs from the cold rolled orientation inasmuch as the [1120] direction, instead of the [1010] direction, is parallel to the rolling direction. In the case of titanium, it is not possible to deduce which direction is preferred in the recrystallized state from the pole figures presented by Clark." The purpose of this paper is to report an extensive investigation of the preferred orientations in iodide titanium. Since the deformation mechanisms for titanium are different from those commonly given for hexagonal close-packed metals, it is not surprising to find distinct differences between the textures of titanium and other metals of this class. Materials and Methods This investigation was carried out on iodide titanium obtained from the New Jersey Zinc Co. with an analysis as follows: N2, 0.002 pct; Mn, 0.004; Fe, 0.0065; A1, 0.0065; Pb, 0.0025; Cu, 0.01; Sn, 0.002; and Ti, remainder. The crystallities of titanium were broken from the as-deposited bar and melted to form 20 g buttons on a water-cooled copper block in a vacuum arc-furnace. Hardness tests conducted on the material before and after melting differed by only two or three Vickers Pyramid Numbers, indicating no or insignificant contamination. The buttons were hot forged, ground, and etched to sizes and shapes suitable for the rolling schedule, and vacuum annealed at 1300°F. Specimens for determination of the wire textures were reduced 91 pct in diameter to 0.027 in. in 24 steps using grooved rolls. In order for the orientation of the central region to be studied, portions of these wires were electrolytically reduced to a diameter of 0.005 in. using the procedure described by Sutcliffe and Reynolds.&apos; The sheet textures were determined on titanium cold rolled 97 pct to a thickness of 0.005 in. A reduction of approximately 10 pct per pass was used, and the rolling direction was changed 180" after each pass. Specimens used for determination of the recrystallized textures were annealed in evacuated quartz tubes at 1000°, 1300°, and 1500°F. The grain size of the 1000°F specimen was sufficiently small to give satisfactory X-ray patterns with the specimen stationary. However, it was necessary to scan the surface of the other recrystallized specimens. The microstructure of each annealed specimen was that of a recrystallized material. The diffraction rings all showed the break-up into spots typical of recrystallized structures.

Jan 1, 1954

By W. A. Backofen, N. E. Paton

DEFORMATION twinning contributes to the plasticity of titanium and has been reported on {10i2}, {1121},{11221}, {1123}, and {1124} planes.&apos; Twinning on {1011} has also been reported in commercially pure titanium as a result of quenching from 1000&apos;, but no reports of such twins occurring as a result of deformation alone are known to the authors. In this study single crystals grown from iodide titanium were compressed parallel to the c axis over the temperature range 25 to 800°C. From it has come evidence that (10il) deformation twins are the most common type between 400" and 800°C. Crystals about 6 mm high with a rectangular cross section measuring about 3 by 4 mm were used in the tests. Chemical analyses showed them to contain 220 to 250 ppm 0 arid 10 to 15 ppm N. The long axes of the crystals were made parallel to the c axis while the sides were parallel to (i100) and (1120), as shown in Fig. 1. The crystals were compressed between polished alumina platens with no lubrication, in a vacuum of 2 x lom5 torr. Between 25&apos; and 300°C the predominant twinning mode was (11221, which is to be expected for this stress state. There was an abrupt change at 400°C, however, and from that temperature up to 800°C twinning took place almost entirely on (10i1 ). Twin planes were determined by the two-surface rnethd. Measurement of the crystal orientation within several (10i1) twins by means of polarized-light4 and transmission electron microscopy showed that the lattice reorientations were the same as those previously reported for the (10i l ) twinning in magnesium.&apos; This reorientation is consistent with The calculated (10il ) twinning shear for titanium is 0.10, while that measured from surface offsets on a (1150) face, which is parallel to the shear direction, was about 0.15. Slip within the twins, or in the matrix adjacent to the twins, probably accounts for the larger measurement. Slip in both locations was judged to have occurred from observations on foils in the electron microscope.

Jan 1, 1970

By E. T. Turkdogan

The activities of SiO2, FeO, and Fe2O3 are calculated from previous experimental data on the activity of oxygen in Fe-Si-O melts at 1550°C. Using the oxide-activity data, the free energy of formation of fayalite from its oxides is evaluated for 1550°C. A number of investigators1-4 contributed to the study of the state of oxidation of iron in SiO,-FeO-Fe,O3 melts and the phase equilibria in the system Fe-Si-O. In a more detailed study, Schuhmann and Ensio5 and Michal and Schuhmann% easured the state of equilibrium in the SiO,-FeO-Fe,O3 melts saturated with y-iron or silica or both under low-oxygen potentials and within the temperature range 1250" to 1350°C. Similar measurements were also made by Turkdogan and Bills7,&apos; over wide ranges of composition and oxygen potentials at 1550"C. Using the experimental data on the oxygen activities in the ternary system, Schuhmann&apos; calculated the activities of silica and ferrous oxide in Si0,-FeO-Fe2O3 melts for 1350°C. Adequate knowledge on the activities of the constituent oxides of melts is essential in the study of reaction-equilibria and reaction-kinetics in oxide-melts. With this in view, the activities of oxides in SiO2-FeO-Fe2O3 melts at 1550°C are computed using the experimental data of Turkdogan and Bills7&apos;8 on the oxygen activities in these melts.

Jan 1, 1962

By A. Kenneth Schellinger

DURING the comminution of a brittle material in the presence of dry air, no known phase change or chemical reaction takes place. The energy changes associated with the comminution are those of the transformation of kinetic energy into heat energy and the energy absorption represented by the actual rupture of molecular and/or ionic bonds. The random nature of the actual grinding process in a tumbling-mill often precludes the effective application of a rupture force on each individual particle of brittle mineral present. The result must be a great deal of useless vibration on an atomic scale in both the brittle mineral and the grinding medium below the elastic limit, which vibration is eventually manifested as heat energy in the mill. The actual rupture of molecular, or ionic, bonds creates new surface on the brittle minerals; each square centimeter of this new surface then must possess a number of potential bonds, each of which could reform theoretically with the release of energy. The sum total of this potential energy of bonding per unit of new surface, minus bond satiation effects of the enclosing gas or liquid, must be the surface energy of the mineral. If the comminution of the brittle mineral were carried out within the chamber of a calorimeter, the energy released as heat could be determined rather accurately. Moreover, the total kinetic energy input to the comminuting mass (mineral and medium) conceivably could be derived by a suitable torquemeter and revolution counter if the whole calorimeter were revolved. The difference between these two energy figures should then represent the net energy (gross energy input—heat enerb output) used in creating new surface, gliding or slip on atomic planes, twinning, and possibly other lattice rearrangements within the mineral. If it be postulated further that with a hard, brittle crystalline material, such as quartz, the energy used to produce gliding, twinning, and other permanent lattice movements is a very small fraction of the total net energy absorbed, then this energy absorption within the grinding calorimeter can be ascribed to the creation of new surface energy alone. Such energy is, from the thermody-namic standpoint, the true effective work of the grinding device, in this case a tumbling-medium mill. The ratio of this effective work (in calories) to the total energy input (in calories) should express the thermodynamic efficiency of the comminution taking place under the conditions used in the grinding calorimeter. This approach to the measurement of the energetics of tumbling-mill grinding was suggested by some previous work by Fahrenwald and coworkers&apos; with a laboratory ball mill and a thermometer. They measured the efficiency of grinding based on the formula: Pct Grinding Efficiency = Energy (calories) — Thermal Energy x 100 1npZ Output (calories) Energy input (calories) Efficiencies thus measured, while grinding quartz sand, ranged from 19.7 pct to 7.4 pct. Cook in 1915,&apos; used this approach to the problem of grinding efficiencies in stamp batteries and obtained the energy input by stamp weights times drop distances and the energy output by means of thermometers and pulp flow rates. He calculated that 80 pct of the input energy to the stamp shoes was wasted as sensible heat in the ore pulp. No other studies of comminution by this method of energy differences came to the author&apos;s attention. Reasonable agreement as to the order-of-mag-nitude of comminution efficiencies in tumbling-media mills is lacking in the literature. There is a widespread impression among technical workers that the grinding efficiency of the industrial ball mill may be on the order of 1 pct, or even a fraction

Jan 1, 1952

By J. E. Clennell

It is a common observation that the improvements introduced in practice since the first announcement of the cyanide process have been almost entirely mechanical. Although .a good deal of study land research has been devoted to the chemical problems [involved, the results obtained appear trifling in comparison. Few of the suggested modifications in the chemical treatment have come into practical use and those only in limited fields. Nevertheless, if we study the recent history of the cyanogen compounds from a somewhat broader standpoint, we shall find that considerable additions have been made, not only to our knowledge of their chemical properties, but also to their applications for industrial purposes. The present review of these advances naturally falls into two divisions: (1) processes involved in the treatment of ores; (2) manufacture of cyanides and other related compounds. Part I. Advances in the Chemistry of Ore Treatment by Cyanide In the attempt to extend the scope of the process, various ores have been encountered which contain gold and silver in such forms that they will not yield satisfactorily to ordinary methods of cyanide treatment. In some cases the difficulty is due to the excessive action on the solution of base metals present in the ore; in others, to the occurrence of minerals such as dyscrasite (the native alloy of antimony and silver), or a supposed oxidized manganese compound containing silver, which are either altogether insoluble in or very slightly attacked by cyanide. A particularly troublesome case is that of the so-called "graphitic" ores, containing carbon in some form which has the property of precipitating gold and silver already dissolved. Other efforts have been directed to improving extraction by electrolysis of working solutions, and by addition of various chemicals with or without electrolysis; to the diminution of cyanide consumption by "re-

Jan 1, 1916

By V. G. Leak, R. A. Swalin

The dilhsion of silver and trace concentrations of tin in liquid silver has been rrzeasured in the temperature range from about 975° to 1350°C. The difBsion dala. fil lhe following equations: fov self-diffusion of silver The ratio of DSn to DAg is found to be about 1.34. The higher diffusivity of tin is interpreted in terms of the coullombic repulsion which results from the fact that tin dissolved in silver has a valence of +3 relative to silver. THE phenomenon of diffusion has been well-described theoretically for hard-sphere gases and solids and found to be in good agreement with experimental data for certain gases and solids. Liquid-diffusion phenomena have not been well described theoretically and there is generally a lack of good experimental data. In this investigation self-diffusion and solute diffusion in liquid silver were studied. Silver was chosen as a solvent for two main reasons. First, silver is a noble metal and the atoms are considered to have a spherically symmetric charge field; hence the liquid may be considered to be a random array of approximate hard spheres. Mercury, gallium, and other lower-melting metals were eliminated from consideration since it appears possible that in their liquid states there is some residual long-range order and directional bonding. Second, silver behaves as a monovalent solvent and, while cadmium, indium, tin, and antimony all have nearly the same atomic size, they have chemical valences relative to that of silver of +1, +2, + 3, and +4, respectively. Slifkin, Lazarus, and coworkers1-5 investigated the diffusion of these solutes in solid silver and found that their diffusion rates increased with an increase of the excess valence of the solute. In the solid state the diffusion-rate increase was calculated to be due to a change in local modulus of the solvent caused by the excess valence of the solute.1"3 For the present investigation it was planned to determine the effect of valence upon solute diffusion in liquid silver. It was deduced that a coulombic repulsion between solute and solvent might be responsible for larger volume fluctuations in the vicinity of the solute thereby enhancing the diffusion of the solute atoms. Tin was the first solute investigated and was chosen for experimental convenience. If an excess-valence diffusion effect exists in the liquid state, the solute tin with its excess valence of + 3 might show a large enough effect to distinguish it from the self-diffusion of silver in silver. The silver self-diffusion data were obviously required as a base line for comparison. In addition silver self-diffusion was investigated with a view toward examining the data in connection with the Sutherland- Einstein: Coheen-Turnbull,7 and swalin8 theories of diffusion in liquids. I) EXPERIMENTAL TECHNIQUES The diffusion coefficients for tin and silver in liquid silver were determined in separate experiments using the capillary-reservoir method of Anderson and saddingtono but following closely the experimental techniques outlined by Ma and Swa1in. The radioactive alloy bath was prepared by plating either tin-113 or silver-110 m isotope* onto 99.999 *Isotopes obtained from Oak Ridge National Laboratory. pct Ag rods.* The silver and isotope were melted *Silver obtained from Cominco Co. and mixed in a graphite crucible under a purified argon atmosphere, then outgassed for capillary filling. Some of the fused-silica capillaries were filled individually as reported by Ma and Swalin, but most were filled in a different manner. A long piece of capillary tubing was sealed off on one end and held in the system so that the open end was just above the surface of the molten alloy. The system was evacuated, the open end submerged into the alloy, and argon was admitted into the system up to atmospheric pressure. The molten alloy was forced up the capillary several inches where it solidified and was subsequently cut into segments of the proper length for diffusion annealing. The apparatus for diffusion annealing was the same as that used for capillary filling except that a Pt—Pt, 10 pct Rh therinocouple was placed in the solvent bath in order to accurately measure the temperature during the run. A diagram of the dif-

Jan 1, 1964