Search Documents

By E. W. Davis

About a year ago several carloads of magnetite ore from New York state were sent to the Mines Experiment Station at the University of Minnesota for grinding and concentration tests. The flowsheet used minus ½-in. ore as feed to a 3 by 3-ft. overflow cylindrical ball mill operating at 37 r.p.m. in closed ciruit with a vibrating screen. The screen oversize was returned to the mill for further grinding and the screen undersize was concentrated magnetically. Several different screens, between 8 and 35 mesh, were used and it was desired to determine the capacity of the ball mill operating in closed circuit with each screen. From earlier tests using this flowsheet and equipment, it was known that when comparatively coarse screens were used it was necessary to increase the grinding dilution in the mill beyond that ordinarily used when grinding in closed circuit with classifiers. It was also known that after the circulating load had passed about 100 per cent of the original feed it tended to increase rapidly in amount, and grinding conditions appeared to become unstable. These conditions may be peculiar to this small ball mill but they were very marked in this series of tests. If conditions were not right, the mill would suddenly "go dead," and it was found desirable to operate so that the occasional rattle of balls against the lining of the mill could be detected. As the test proceeded, it was found that the cure for almost any difficulty that developed in the grinding circuit was to add more water to the mill feed. To begin with, the mill was operated at pulp densities of from 60 to 65 per cent solids but gradually it was found that with higher dilutions grinding conditions were more stable and the ball-mill power per ton of screen undersize decreased. Actually, the most satisfactory operating conditions were secured with a pulp density of about 40 per cent solids. The ore was hard but it was possible to grind ½-in. feed so that all would pass an 8-mesh screen with a power expenditure in the ball mill of less than 2½ kw-hr. per ton and with a circulating load of less than 50 per cent. With so high a dilution in the feed to the mill, it was feared that steel wear would be very high. However, the mill was not noisy, and in order to find exactly the grinding conditions, a special test was made at the end of the main series of tests. After the mill had been in operation for several hours and grinding conditions had become balanced, the power to all the equipment was shut off at a given signal. The ball mill made about one revolution after the power was turned off, and just as soon as the mill stopped, a spout was placed at the discharge trunnion and all of the pulp that flowed out of the mill was caught in tanks. The manhole in the mill was then opened and the whole mill charge dumped into other tanks. Both of these products were weighed, dried, and weighed again. The balls were screened out and weighed, and the dry solids were sized. In this way it was possible to determine the nature of the pulp and the ore within the mill under actual grinding conditions.

Jan 1, 1947

By H. A. Wahl

This paper presents methods of designing hydraulic fracture treatments in formations saturated with slightly compressible liquids. Howard and Fast describe the fluid-loss control resulting from the viscosity and compressibility effects of the reservoir fluid in terms of a fracturing fluid coefficient. However, for horizontal fractures, this coefficient is applicable to infinitely thick reservoirs. These methods of evaluating fluid loss from horizontal fractures during treatment differ from previous methods in that: (I) formation thickness is incorporated in the velocity function describing the rate of fluid loss to the formation; and (2) for a given thickness the relative position of the fracture to the horizontal boundaries of the reservoir or any interrnediate permeability barrier is incorporated in the design procedure. This theory is applied to several actual and hypothetical field cases. The results of these calculations demonstrate that the eflect of formation thickness on fluid loss from horizontal fractures may result in fractures several times larger than predicted by other methods. INTRODUCTION Hydraulic fracturing has been the dominating well-stimulation technique during the last decade. From the time it was commercially available, fracturing has had a very favorable success ratio compared to other stimulation methods. However, numerous wells that were successfully stimulated remained economic failures. Only in the last few years has the oil industry recognized the importance of designing treatments for each well which will yield the maximum profit. This implies that for a given set of treatment conditions the size and general physical characteristics of the fracture can be determined. One of the outstanding contributions in the field of fracture design was presented by Howard and Fast.' In their paper, three flow mechanisms that control fluid loss from the fracture to the formation are defined: (1) viscosity effects of the fracturing fluid, (2) viscosity and compressibility effects of the reservoir fluid, and (3) wall-building effects of a treated fracturing fluid. The rate of fluid loss to the formation adjacent to a fracture is shown to be a function of time. For each flow mechanism the velocity function v(t) is expressed as a constant or fracturing fluid coefficient divided by the square root of time. This form of v(t) can be used in the Carter equation, as developed in the Howard and Fasc paper, to estimate the areal extent of fractures. This equation may be written as In the following sections the effect of formation thickness on fluid loss from horizontal fractures is evaluated when the controlling flow mechanism is the viscosity and compressibility of the reservoir fluids. The major assumptions inherent in the Carter equation are brought out in subsequent mathematical developments. THE EFFECT OF RESERVOIR FLUIDS ON FRACTURE EXTENSION When a fracture treatment is performed using a fracturing fluid identical to the reservoir fluid, the only flow mechanism controlling fluid loss is the viscosity and compressibility effects of the reservoir fluids. In developing a fracturing fluid coefficient for this mechanism, Howard and Fast considered a segment of the formation bounded by a unit area on the fracture face and extending to infinity in the direction of leak-off as shown in Fig. 1(a). The direction of leak-off z is normal to the fracture face. The equation governing the linear isothermal flow of fluid into this segment is where a = Eq. 2 is a form of the diffusivity equation. It is derived from the equation of continuity, Darcy's law and Eq. 3

By F. W. Bloecher

DETERMINATION of the surface area of finely divided minerals is of interest to mineral engineers engaged in flotation, comminution, and de-watering studies. In the industrial minerals field, many materials are classified as to fineness according to their specific surface (surface area per unit weight). A method of surface area determination has been tested and used in the Richards Mineral Engineering Laboratories at M.I.T. It involves low temperature krypton adsorption measurements using the technique described by Beebe.' During the past fifteen years a number of new surface measurement techniques have been developed. Those that are broadly applicable to materials dealt with by mineral engineers include the permeability method in which a gas or liquid is passed through the sample, and the low temperature gas adsorption methods first described by Brunauer, Emmett, and Teller."" Efforts to make surface determinations from size distribution curves have long been attempted"."." but because of the uncertainty of sub-sieve size distribution and inaccuracies in screen analyses, the method is at best only a rough approximation. The most widely used gas adsorption method utilizes nitrogen gas." Nitrogen adsorption, however, is not particularly suited as a tool for most mineral engineers because the method is not applicable for surface determinations on coarse materials, that is, materials coarser than approximately 150 to 200-mesh. Other gases have been used in mineral engineering studies, particularly ethane."' lo The use of ethane gas, however, involves .a technique that is too sensitive for most mineral engineering work. The permeability method of surface area determination has the disadvantage of not being sensitive enough to detect the surface represented by minute cracks, the so-called internal surface of a particle. The low temperature krypton adsorption method described by Beebe1 utilizes apparatus and techniques that are essentially the same as those used in the ethane adsorption method."1° The surface area is calculated by the B.E.T. method."-' The krypton adsorption technique is particularly suited to the needs of the mineral engineer because: (1) The method can be used for making surface area measurements of rather coarse materials (35 to 48-mesh), as well as extremely fine materials, such as clays or colloidal slimes with specific surface that might be as high as 20 to 30 mb er g. (2) The use of krypton gas offers the advantage of low saturation pressures (2-3 mm) at the conveniently obtained liquid nitrogen temperature. Therefore the krypton vacuum system need not be designed to measure or withstand internal gas, pressures above one atmosphere as is the case with the nitrogen gas method. (3) The krypton technique offers the safety feature of employing liquid nitrogen, as contrasted to the ethane technique in which liquid oxygen is required. Furthermore, ethane vapor pressure at liquid oxygen temperatures is low (about 10 microns), and a high sensitivity McLeod gage is required for pressure readings. (4) The measurement technique with krypton gas is much simpler and faster, and the apparatus is less elaborate than that required with either nitrogen or ethane gases. Apparatus A krypton surface measurement apparatus like that used at M.I.T. is shown in Fig. 1. It is constructed of pyrex glass. A vacuum is maintained by two mercury diffusion pumps in series backed by a mechanical vacuum pump."' The vacuum thermocouple gage is used for checking the vacuum within the system. This type of gage is quite sensitive down to a pressure of 1 micron of mercury," but may be used to indicate pressures much lower than that

Jan 1, 1952

By J. J. Cosgrove, J. E. Warren

A general model which approximates the effect of cross- flow has been developed to give a practical method for predicting the waterflood behavior of a stratified reservoir. The model is based on a modification of Dietz's theory and allows for variations in both the permeability and hydrocarbon pore volume. In the particular cases considered, it is assumed that the permeability can be characterized by a log-normal distribution and the hydrocarbon pore volume by a normal distribution. A simple graphical method which enables the practicing engineer to predict the behavior of a stratified system is presented The results obtained by the proposed method are compared with those obtained by the Dykstra-Parsons method. As a result of this study the following conclusions have been drawn: (1) The effect of cross-flow in a stratified system can be appreciable, particularly at very favorable or very unfavorable mobility ratios; (2) Under normal conditions, the effect of variations in the hydrocarbon pore volume can be neglected; and (3) The failure to use all of the available permeability data can lead to large errors in the prediction of the behavior of a stratified reservoir. INTRODUCTION Various methodshavebeenproposedto characterize and to predict the waterflood behavior of a stratified reservoir. Most of the methods assume that the reservoir is composed of discrete homogeneous continuous layers. With this model, the degree of stratification can be measured by several parameters based on core-analysis data. Among these are the Lorenz coefficient1 and the variation of a log-normal permeability distribution? The behavior of stratified systems is usually predicted by the Stiles, Dykstra-Parsons method or some modification of these. In both of these methods the reservoir is divided into discrete homogeneous layers with no cross-flow between layers. In the Stiles method the mobility ratio is assumed to be equal to unity while in the Dykstra-Parsons method it is allowed to take on any value. Other predictive methods have been proposed; method allows for cross-flow between the beds and a method by Schmalz and Rahme6 correlate.s the recovery directly with the Lorenz coefficient. In this new approach, which is essentially a continuous analog of Hiatt's method, the effects of both mobility ratio and cross-flow between the beds have been included. It is assumed that the permeability can be represented by a log-normal distribution and the hydrocarbon pore volume or porosity by a normal distribution. These types of distributions have been observed by several authors, 1,2,4 and most field data seem to confirm their observations; e.g., if these distributions are assumed and there is a 1:l correspondence between porosity and permeability samples, the commonly encountered linear relationship between porosity and the log of permeability is obtained. Since, in many field cases the permeability and poro sity data are truncated or cut-off at predetermined upper and/or lower values, the effect of discarding part of the data was investigated with the proposed method. A better technique for truncating the data is suggested. THEORY In the derivation of the method to be described in this paper the following assumptions are made: 1. Capillary forces are negligible per se; their effects are only manifest in the relative permeability curves. 2. The fluids are immiscible, incompressible and homogeneous. 3. The reservoir is horizontal and uniformly thick; it is initially liquid saturated i.e., connate water and oil. 4. The reservoir is composed of discrete layers, each having its own permeability, porosity and connate water saturation; each layer is homogeneous,

Jan 1, 1965

By R. L. Smith, I. G. Greenfield

The binary system Ga-Sb has been investigated by thermal, X-ray, and metallo-graphic methods. 'The intermetallic compound GaSb melts at 705.9OC and forms a eutectic with antimony at 11.8 atomic pct Ga. This eutectic melts at 589.8OC. A eutectic appears to be formed at 29.8°C but is too close to the composition of pure gallium to be detected. IN the course of an investigation of the semiconducting properties of the intermetallic compound, GaSb, the phase equilibrium between gallium and antimony has been investigated. Recently, the phase diagram of similar systems, such as In-Sb,' In-AS,' and Bi-As3 have been determined. These diagrams have the same general appearance in that an intermetallic compound is formed at 50 atomic pct and a eutectic is formed on both sides of the intermetallic compound. Jenny,& who has been investigating the semiconducting properties of intermetallic compounds, has concluded from thermal data of several alloys that the phase diagram of the Ga-Sb system resembles those mentioned previously. This paper represents a more detailed study of the Ga-Sb equilibrium. Experimental Procedure The gallium that was used to prepare the alloys was obtained from A. D. Mackay Inc. and the Aluminum Co. of America; the latter reported the purity as 99.95 pct. The antimony was obtained from Bradley Mining Corp. who specified a purity of 99.997 pct. Since the intermetallic compound of GaSb was prepared primarily to be used for further studies of the semiconducting properties, additional purification by a standard zone-melting technique under an atmosphere of hydrogen was performed. Hall-effect measurements on this material indicated an impurity content of less than 3 ppm.5 The antimony also was purified by zone melting and used in conjunction with the compound GaSb to make the necessary alloys for the determinations of the antimony-rich side of the phase diagram. The antimony-rich alloys were chill cast in Vycor tubes which were evacuated to a pressure of less than one micron. Since these alloys expand on freezing, precautions were taken to prevent oxidation of the ingot in the event the tube fractured. To insure homogeneity, the brittle alloys were ground and then placed in the furnace for the thermal-arrest determinations. Since the cost of gallium is high, the experimental procedure was complicated somewhat because of the necessarily limited sizes of the melts. In order to economize, the gallium-rich side of the diagram was determined by using a 35 g ingot of pure gallium initially and adding increasing amounts of antimony in order to obtain additional alloys. The furnace used in the determination of the cooling curves is illustrated in Fig. 1. A 2¼ in. quartz tube, sealed at one end, was fitted with a brass flange which was attached with DeKhotinsky cement. The brass collar was water cooled to prevent the cement from softening. An "0" ring was used to provide a seal between the brass collar and the furnace cover. Nichrome ribbon was wound around a 2 ½ in. asbestos-covered copper core which was used to reduce the thermal gradient in the furnace. The insulation consisted of alternate layers of asbestos and aluminum foil surrounded by sil-o-cel refractory and

Jan 1, 1956

By R. W. Olson, A. L. Benham

Viscous fingering was studied as it occurred in an open Hele-Shaw model (1 ft x 4 ft x 1/16 in.); it was also studied in the same model packed with 80-mesh glass beads during miscible displacements under unfavorable viscosity ratio conditions. It was determined that the lengths of 20 elements of the front were distributed normally around the average position of the combined elements under all conditions. Represented by the normal distribution, the length of the viscous fingers grew linearly with the distance traveled from the point of finger formation, increased with displacement velocity, and increased with increasing mobility ratio. Results obtained during the first few inches of displacement were of little or no use in predicting finger growth and/or finger length throughout the 4-ft model since the point of finger initiation cannot be predicted and would often occur several inches from the point of injection with either positive or apparent negative coordinates. The fingering occurring during a miscible slug displacement was much greater than would be predicted based upon the actual mobility ratios between in-place fluid and slug and between slug and following fluid, using the results described above for miscible displacement in the absence of a slug. Many of the experiments in the packed model showed that the rate of growth of the viscous fingers was diminishing toward the end of the displacement in the 4-ft long model, indicating that microscopic mixing, such as diffusion or dispersion, was decreasing the viscous fingering effect. INTRODUCTION Viscous fingering is a manifestation of a finger-shaped interface between displaced and displacing fluids occurring during typical miscible displacement projects for oil recovery. Its cause may be traced to the instability of a viscous fluid being displaced by a more mobile fluid. Viscous fingering takes on important significance in the miscible slug process 1.2 where it may be a dominant factor in determining minimum slug size. Experimental1-3 and theoretical4-7 studies of viscous fingering have been made by other investi- gators. However, an exhaustive study of the variables affecting viscous fingering had not been made. The present study was undertaken in an effort to determine the effects of some of the more obvious variables—such as mobility ratio, displacement velocity, distance displaced, and packing—upon viscous finger length and growth in a small laboratory model. It does not necessarily follow that conclusions reached by studying the results of this model study may be applied to the field. Future studies would have to evaluate the effect of model size on extension of these results. EQUIPMENT AND EXPERIMENTAL PROCEDURES A flow diagram for the equipment used in this study is shown in Fig. 1. Basically, it consisted of (1) a constant-rate pump for the injection of displacing fluid into (2) a model made up of two flat transparent plates spaced a small distance apart and (3) provisions for the production of fluid into a calibrated graduate. The displacing fluid contained a dye which allowed visual observation of finger spacing and length. The pump performed at rates from 2.5 cc/hr to 480 cc/hr, while pumping a light mineral oil into a bladder in a closed glass container holding the displacing fluid. The model was made up of two plates of 1-in. Plexiglass with dimensions 1-ft wide by 4-ft long. These were spaced 1/16 in. apart so that the internal model dimensions were 1-ft wide x 4-ft long x 1/16

By W. A. Stickney, J. W. Town

Tests demonstrated that aqueous solutions of sodiu~n sulfide would dissolve over 95 pct of the cinnabal- in 5 pct Hg flrotation concentrates and 60 to 90 pct of the cinnabav in low-grade ol-es. Double-leach tests showed that potassiutn sulfide would dissolve 90 pct of the avsenic sulfide in Hg-As satnples and only 1 to 10 pct of the Hg; the sodium sulfide would then dissolve 80 to 90 pct of the rernaining Hg and 30 to 70 pct of the vemaining As. 1 HIS paper describes the extraction, by alkaline sulfide solutions, of mercury from cinnabar-bear ing ores and flotation concentrates. Leaching time and solution temperature had minor effects on mercury extraction when compared to the effects of reagent quantity and ore type and grade. Cinnabar, although readily soluble in sodium sulfide, was only partially soluble in potassium sulfide and essentially insoluble in ammonium sulfide. A double leach process was developed to extract selectively the arsenic from Hg-As products with K2S and to treat the residue with sodium sulfide-hydroxide solutions to extract the mercury.' Mercury was then recovered from the pregnant solutions either by precipitation with metallic aluminum or by electrodeposition. The solubility of HgS in alkaline sulfide solutions was first reported by Volhard in 1878.' The first reported commercial application was made in 1915 when Thornhill applied the method to the treatment of an amalgamation tailing.3 Mullhollen, in 1911. described the chemical reaction of HgS with Na,S and alkaline hydrate to be:4 Most investigators have reported using a 4-pct sodium sulfide and a 1-pct sodium hydroxide soluti~n.3-4 An increase in temperature or excessive dilution of the solution causes the reaction to be reversed. Church Holmes described a similar process used for the extraction of antimony from silver concentrates at the Sunshine Mining Company's mill at Kellogg. Idaho.5 More recently. Erspamer and Wells reported the results of leaching studies on a Hg-Sb ore from Alaska.6 EXPERIMENTAL PROCEDURE Leaching tests were made on samples of mercury ore from eight deposits in Oregon, Washington, and Idaho. In addition. tests were made on flotation concentrates from seven of the samples as well as a flotation concentrate from an operating mill. Since previous investigators reported reagent requirements greatly in excess of that required by the equation HgS + Na2S = HgS . Na2S, initial studies were made on chemically pure HgS. Results showed that the reaction actually required about 1 g of Na,S per g of HgS in an aqueous solution rather than 0.338 g of Na,S as indicated by the equation. To check further the matter of reagent requirements, Na2S and HgS were mixed together in equal mole ratios. Water was then added, a drop at a time, until the components reacted; X-ray diffraction analyses showed no HgS nor Na,S present. The yellow hydrate compound which formed would not redissolve until about three additional mole equivalents of Na,S were added to the solution. This indicated that a stable solution would require at least 4 moles of Na2S for each mole of HgS. While these implications may be valid for pure HgS, the possibility of entirely different results caused by the existence of other elements present in ores and concentrates was recognized. Consequently, experiments of the same order were made on samples of a flotation concentrate from a producing mill. The concentrate, containing about 18 pct

Jan 1, 1962

By H. H. Rachford, W. P. Schultz

The use of computing machines to solve physical problems has made it imperative to represent physical data in a form computing machines can use. Although curve-fitting is an old and well-practiced art, there still seems to be some interest in questions regarding what methods can be used and why. Even though there are many good references on this subject, it may be of some value to discuss a particular method which has been found to be useful, and to make available the associated tables for its use in the hope that there can be saved a few of the tedious hours necessary to reduce many of the suggested methods to practice. Although a very brief discussion of the mathematical background of the tables is included immediately below, those interested only in their use can skip directly to the section on Use of the Tables. Direct Fit of a Set of Points A frequently encountered problem is: Given data which can be represented by a smooth curve +(x), to find an easy-to-evaluate function F(x), with only a few coefficients, which is a good approximation of +(x) on the interval 0 5 x 5 x,,. Consider F(x) as a linear combination of m + 1 arbitrary, single-valued, continuous, linearly independent functions, f,(x), Ff.r) = a,f,,(x) + a,f,(x) + . . . -+ aifi(x) + If the value of Q(x) is known or can be read from the curve at the points xj: +(x,), +(x,), . . . +(xj) . . . +(x.), and if n = m, then all the coefficients of F(x) can be determined, and F(x,) — +(x,) = 0 for j =. 0, 1, . . . n. This follows from the solution of the set of linear simultaneous equations. +(x<.) = aofu(x,) + aJI(x0) + . . . + a,,f,,(x,) =a, = f,,(x,) + a,f,(x,) + . . . + a,f,,(x,) +(x,,) = a,,f,(x,,) + a,f,(x,) + ... + a,,f.,(x,,) ...........(2) for the coefficients ai. Although the set a, from the solutions of the set of Equations 2 yields an fix) which gives an exact value of $(x,) for the n + 1 points, there is no restraint on ~(xj in the interval xi < x < xj+,. Indeed, if the points +(x,) are not sufficiently smooth, very serious differences may arise between F(x) in the interval and the smooth curve +(x). Least-Squares Fit It has been suggested&apos; that if n is made to exceed m, and the set a, is found to minimize the function the function F(x) is much less sensitive to small errors in +(xi), and usually provides a much better fit to the smooth function +(x). This is known as the best fit in the sense of the least-squares error. To minimize $, set £ = 0,i = 0,1,... iTi,.....(3) The m + 1 equations from Equation 3 are k—n k=n a.^p /,,(a) f,(xt) + a, X A: = 0 k=O k = n k = n k=0 k=0 +(x.) f,(x,),............(4) one equation being obtained for each i = 0, 1, . . . m. Inasmuch as the arbitrary functions f(x) may be evaluated for x,, k = 0, 1, . . . n, and the +(x,) are known values of the curve to be fitted, the simultaneous Equations 4 may be solved for the set a,. Particular Functions Given the functions fi(x) and the set +(xj), the solutions for require evaluating a number of sums of products as well as the solution of an m + 1 order set

Jan 1, 1956

By M. A. Dayananda

Based on the Kirkendall frame of reference, a method is proposed for the determination of atomic mobilities of all the diffusing species in a multicorn-ponent system with a single diffusion couple. The method is based on an atomic mobility model developed on the concept of negligible interactions among the intrinsic fluxes. It is shown in the model that mobilities can be obtained directly from the cumulative intrinsic fluxes past the Kirkendall interface. The proposed method replaces the need for a prior determination of interdiffusion coefficients far the calculation of mobilities. Relations between intrinsic diffusion coefficients and mobilities are derived and the mobility calculations are illustrated with diffusion data in Cu-Zn-Mn alloys at 850°C. ThE concept of "intrinsic flux" with respect to a fixed lattice was introduced "i Darken1 to explain toa Kirkendall effect2 demonstrated by the motion of inert markers during isothermal diffusion in single-phase binary solid solutions. Based on this concept, intrinsic diffusion studies have been made in various binary metallic systems3,4 employing such a lattice frame of reference, commonly referred to as the Kirkendall interface. Similar approaches, both theoretical and experimental, have been presented5"10 to study intrinsic diffusion in multicomponent systems. Based on the direct integration of Fick&apos;s law extended to multicomponent diffusion, Philibert and Guy8 proposed a method for the determination of intrinsic diffusion coefficients. For an n-component system, this procedure requires an experimental study with (n — 1) diffusion couples, all having their Kirkendall interfaces at the same compositio;, where n(n — 1) intrinsic diffusion coefficients are defined. This technique has been employed for ternary diffusion studies by Dayananda and Grace" and Guy and Leroy.11 A different approach is recommended by Ziebold and Cooper,12 who have derived relations between interdiffusion coefficients and atomic mobilities employing fluxes referred to the local center of volume. Their analysis is based on the simplifying assumption of negligible interactions among intrinsic fluxes. This assumption has been considered satisfactory for dilute solutions.5"7 The use of this analysis for the calculation of mobilities requires an initial determination of interdiffusion coefficients with a set of diffusion couples having intersecting diffusion paths. It is the purpose of this paper to propose a method that would permit a direct determination of mobilities of all components of a multicomponent system from a single diffusion couple. This method is based on a mo- bility model developed with diffusion fluxes referred directly to the Kirkendall interface, generally identified by inert markers. gen erall, like that of Ziebold and Cooper,12 employs the simplifying assumption that the flux of each component relative to a fixed lattice is caused only by the driving force of its own chemical potential gradient. THE KIRKENDALL INTERFACE Inert markers are employed in diffusion couples to locate experimentally the position of the Kirkendall interface. They are initially placed at the weld interface for "sandwich" couples, or at the original vapor-solid interface for vapor-solid couples. Studies of marker motion in binary systems3&apos;4 with both types of couples indicate that markers remain in a plane of constant composition and move with this plane para-bolically with time. Parabolic marker movements were also observed in ternary Cu-Zn-Mn couples by Dayananda and Grace.10 It is reasonable to expect a similar marker behavior even for diffusion couples with a higher number of components.

Jan 1, 1969

Discussion of the paper of DORSET A. LYON and ROBERT M. KEENEY, presented at the Salt Lake meeting, August, 1914, and printed in Bulletin. No. 92, August, 1914, pp. 1791 to 1800. LAWRENCE ADDICKS, Chrome, N. J.-I read this paper for the first time this afternoon, and it is really worthy of a much more careful reply than I can make with such a short time to consider it, but I just want to say that I differ with Dr. Lyon in his premises and conception of what happens in a refining furnace. He cites steel as an instance where the production of special grades justifies the high cost of power in an electric furnace, and goes on to say the same will be true in the case of copper, for the reason that you will get a better quality of product. I don&apos;t think you will get a higher grade of copper,. or a better conductivity, as I will explain in a moment. The cost of fuel in the Eastern plants is somewhere between 30c. and 40c. per ton of copper refined, and, if my memory serves me correctly, in an open-hearth steel furnace it is something like $1.25, so it is evident that steel could be handled to better advantage than copper on the basis of fuel. Now, taking the Great Falls plant to compare with is not quite fair, because that plant operates under peculiar conditions. The furnaces are relatively small; the coal is very bad and is very high in sulphur. No one would think of using a coal with 3.6 per cent. sulphur were anything better available. Due to these conditions, the amount of slag is inordinate, it being quoted as 3.5 per cent., and it is necessary to use lime, as there is a tendency to have troubles from arsenic owing to the very high current density used. I think he has taken about as hard a case as he could pick out. Now as to the conductivity of copper, my own idea about the reason that melted copper does not give as high a conductivity as pieces drawn directly from a cathode, or native copper, is that in the latter case the impurities are present as a mechanical mixture, while after melting they are chemically combined with the copper, which is a very different condition. Further, it is necessary to take the sample from a relatively smooth and consequently pure part of the cathode, as otherwise it would not draw. The same with the mass of copper; you pick out a good solid part of the copper, which is not representative. In fact, the refiners have all found they could not take fair samples of cathodes for a chemical analysis. It is always customary to use wire-bar copper -for the purpose of sampling, and the impurities are several times as great as shown by samples taken from .the cathodes. I think if we examine all of the examples he has given of cases where copper is higher in conductivity, it will fit with that explanation. For example, the Elmore process is a direct electrolytic deposit. Incidentally, at the present time there are a number of ex-

Jan 12, 1914

By C. L. T. Edwards

In the preparation of this paper, the author has drawn upon experience with the operation of a blast furnace on Ioo per cent scrap burden, which he believes was the first operation of its kind in the country. It is the writer&apos;s further belief that much of the controversy attending discussions of this subject may be obviated by the presentation of a group of facts that should be known by blast-furnace men and open-hearth men alike. The following question has been raised: "Is iron made with scrap in the blastfurnace charge as good iron as that made without scrap—made with straight ore?" This question may he answered in part by the fact that the open-hearth men of the Bethlehem Steel Co. who used pig iron from a blast furnace that operated on Ioo per cent steel scrap found no difficulty in the processing of such iron other than that to be expected from the known analysis. Operation with ioo PeR Cent Steel Scrap During World War I, there was accumulated at the Bethlehem plant of the Bethlehem Steel Co. some 22,000 tons of alloy-steel turnings and borings. This scrap material was highly corroded and, like alloy-steel turnings of that period, were as a drug on the market. There was also a small tonnage of scale, rich in valuable metal. If this scrap could be worked up into pig iron, it would become readily available for use in the open hearth. In order that the pig iron should contain the maximum percentage of nickel, permission was obtained from the Management to try a Ioo per cent scrap operation. Accordingly, a blast furnace was swung over on July 25, 1919; the percentage of steel scrap increased more or less irregularly until July 29, when Ioo per cent steel turnings were charged. Except for occasional charges of nickel scale as an "cxtra," the operation continued uninterruptedly until Aug. 27, when the furnace was swung back to the regular ore—ore and scrap—operation. During this period, the furnace production ranged from a low of 568.60 tons to a maximum of 766.07 tons. At the swing-over from Ioo per cent scrap to the regular burden, the charge I hung up rather severely. While the latter condition has nothing to do with the object of the paper, a brief discussion of the experience will appear later because of the interest it may have for blast-furnace operators. In Table I, results of the first and last days of the ore charge are shown, July 25 and July 29. The last column shows a summary for the period of maximum scrap &apos; charge. The furnace employed had a hearth 16 ft. o in., bosh 22 ft. o in., stock line 15 ft. 6 in., bell 11 ft. 6 in., batter in., bosh angle 73&apos; 27&apos; and a Balrer-Neuman top, with gas-engine blowers. Analyses Range of material charged Gravel: SiO2, 64.40 to 70.54 per cent; A1203, 3.52 to 4.58; CaO, 7.27 to 9.49; MgO, 5.09 to 6.14; FeO, 1.67 to 3.34, ignition loss, 9.50 to 13.83. I

Jan 1, 1942

By N. F. Fiore, M. C. Mangalick

DISLOCATION damping in single crystals is orientation dependent.&apos; If dislocations are distributed uniformly over all the slip systems in a crystal, and if each dislocation contributes identically to the damping process, then a damping loss or decrement 6 arises. In fact, however, for a crystal of axial direction (hkl), dislocations on certain slip systems are oriented to make a larger contribution to the damping loss than are dislocations on other slip systems. Thus the measuring damping, 6 (hkl) is proportional to an orientation factor (hkl), which weights the ability of each slip system in the crystal to contribute to the total damping. If it is assumed that dislocations are distributed uniformly on all slip systems, then for a given orientation of an fcc crystal this factor is determined2,3 by the three nonzero elastic constants Cll, C12, and C44 and the density p. In this communication measured values of elastic constants for two Fe-18Cr-Ni alloys are reported, and the dislocation damping results on two different orientations of each alloy are analyzed for orientation dependence. Elastic constants were determined by measuring the ultrasonic velocities of one longitudinal and two shear waves in (110) direction of the fcc crystals.4 Variation of velocities with the change in orientation is very small for (110) orientation,3 and so this direction is ideal for orientation factor and comparative damping measurements. Experimental Procedures. The nominal compositions of the alloys studied are given in Table I. These alloys were vacuum-melted and vacuum-cast from high-purity materials at U.S. Steel Applied Research Laboratories. The single crystals were grown by the modified Bridgeman technique. The specimens were homogenized at 1350°C for 20 hr, cut to shape and size with a spark cutter, and then annealed at 1100°C for 20 hr prior to use. In the KHz frequency range, the Marx composite piezoelectric oscillator was used to obtain the damping. The ultrasonic pulse technique was used to measure the ultrasonic velocities and the attenuation in the MHz frequency range. M.c.mANGALICK,-- StudentMemberAIME,formerly Research Assistant, University of Notre Dame, is now with Casting Laboratory, Chase Brass and Copper Co., Cleveland, Ohio. N. F. FIORE, Associate Member AIME, is Assistant Professor, Department of Metallurgical Engineering and Materials Science. University of Notre Dame, Notre Dame, Ind. This communication is part of a Ph.D. thesis presented by M. C. MANGALICK to the Graduate School of the University of Notre Dame. Manuscript submitted May 20. 1968. IMD Elastic Constants. Table I gives the values of the ultrasonic velocities. Values of C11, C12, and C44 calculated from these velocities are also given in Table I, and it is evident that there is very little variation in C11, C12, and C44 with nickel content. This slight variation may be attributed to inaccuracies in velocity measurements and practical differences in the orientations of the specimens. For example, in B the time interval for vl was 2.4 ± 0.05 sec giving an error in v1 of 0.235 x or 0.005 in. per sec. The elastic constants obtained by pulse technique were tested by calculating E<110>, Young&apos;s modulus for the (110) direction, from the equation: This value was compared with the value computed from the resonance bar measurements in which where fr is the resonance frequency and I, is the length of the specimen. In all cases, the values calculated from the ultrasonic velocity measurements

Jan 1, 1969

By B. Ramaswami, Y. Nakada, R. Schaefer

EARLY investigations1 of the tensile properties of fcc single crystals did not reveal the easy glide region of crystals with axial orientations inside the standard stereographic triangle. With the production of high-purity metal, a region of easy glide became known1 Consequently, modern work of fcc single crystals ordinarily uses high-purity metal. In order to determine whether further purification of aluminum beyond 99.992 pct purity would result in significant alteration of the tensile properties of single-crystal specimens, metal was zone-refined and crystals grown from the purified material were compared to crystals grown from the original material. Some similar studies were made of gold and silver crystals. Aluminum. The material to be zone-refined was cast into bars 1/4 by 1/2 by 8 in. A zone, approximately 1-1/2 in. in length, was passed over each bar eighteen times at a rate of 6 in. per hr. Melting and stirring were carried out by an induction heater under a vacuum of approximately 10"5 mm Hg, the metal being held in a graphite container. he last 3 in. of each processed bar were then cut off and discarded. No quantitative determination of purity was made, but the effectiveness of the zone refining was tested by a method first suggested by Walton, Tiller, Rutter, and winegard3 depending on the relation between cell structure and impurity content. Rutter and Chalmers4 have shown that a cellular substructure is associated with the presence of impurities during solidificatibn, and have shown how solute rejection at the freezing interface can lead to such an association. Thus the observation of cells gives evidence of impurities being present. To look for cells, the last quarter of a zone-refined bar was cut off, remelted, and grown rapidly into a single crystal. It was then heavily electropolished and thermally etched.~ Cells were revealed by the presence of many parallel rows of etch pits. The same process was carried out with material from the first half of the bar, but no cells were detected. These observations give evidence that a significant amount of impurity material was moved to the "dirty" end of the zone-refined bars. Further indication of purification was obtained from recrystallization experiments. It was found that, after cold rolling to a 97 pct reduction in thickness, the material from the pure end of the bar re-crystallized at room temperature, but the material from the "dirty" end did not. Other investigators6 have previously reported that zone-refined aluminum recrystallized at room temperature. The purified material was cast into 1/4 by 1/4 in. bars, and seeded crystals were grown in a (100) orientation and in a single slip orientation where the Schmid factor is m =2. The crystals were then cut to the desired length, annealed for 24 hr at 610°C, furnace-cooled, and etched in Tucker&apos;s etch. The crystals were tested in a floor-model Instron machine, using a gage length of 4 in. and a strain rate of 1/2 pct per min. The resulting curves are shown in Fig. 1. It is seen that the differences between the curves for zone-refined and original-purity materials are small. These differences are within the scatter observed for crystals of the original purity. In particular, the zone refining does not significantly affect the onset of the different "stages" of the curves.

Jan 1, 1964

By C. L. T. Edwards

In the preparation of this paper, the author has drawn upon experience with the operation of a blast furnace on Ioo per cent scrap burden, which he believes was the first operation of its kind in the country. It is the writer&apos;s further belief that much of the controversy attending discussions of this subject may be obviated by the presentation of a group of facts that should be known by blast-furnace men and open-hearth men alike. The following question has been raised: "Is iron made with scrap in the blastfurnace charge as good iron as that made without scrap—made with straight ore?" This question may he answered in part by the fact that the open-hearth men of the Bethlehem Steel Co. who used pig iron from a blast furnace that operated on Ioo per cent steel scrap found no difficulty in the processing of such iron other than that to be expected from the known analysis. Operation with ioo PeR Cent Steel Scrap During World War I, there was accumulated at the Bethlehem plant of the Bethlehem Steel Co. some 22,000 tons of alloy-steel turnings and borings. This scrap material was highly corroded and, like alloy-steel turnings of that period, were as a drug on the market. There was also a small tonnage of scale, rich in valuable metal. If this scrap could be worked up into pig iron, it would become readily available for use in the open hearth. In order that the pig iron should contain the maximum percentage of nickel, permission was obtained from the Management to try a Ioo per cent scrap operation. Accordingly, a blast furnace was swung over on July 25, 1919; the percentage of steel scrap increased more or less irregularly until July 29, when Ioo per cent steel turnings were charged. Except for occasional charges of nickel scale as an "cxtra," the operation continued uninterruptedly until Aug. 27, when the furnace was swung back to the regular ore—ore and scrap—operation. During this period, the furnace production ranged from a low of 568.60 tons to a maximum of 766.07 tons. At the swing-over from Ioo per cent scrap to the regular burden, the charge I hung up rather severely. While the latter condition has nothing to do with the object of the paper, a brief discussion of the experience will appear later because of the interest it may have for blast-furnace operators. In Table I, results of the first and last days of the ore charge are shown, July 25 and July 29. The last column shows a summary for the period of maximum scrap &apos; charge. The furnace employed had a hearth 16 ft. o in., bosh 22 ft. o in., stock line 15 ft. 6 in., bell 11 ft. 6 in., batter in., bosh angle 73&apos; 27&apos; and a Balrer-Neuman top, with gas-engine blowers. Analyses Range of material charged Gravel: SiO2, 64.40 to 70.54 per cent; A1203, 3.52 to 4.58; CaO, 7.27 to 9.49; MgO, 5.09 to 6.14; FeO, 1.67 to 3.34, ignition loss, 9.50 to 13.83. I

Jan 1, 1942

By G. F. Sager, B. J. Nelson

Magnesium alloy forgings offer higher and more uniform mechanical properties than heat treated magnesium alloy castings and are used principally for light weight parts that may be stressed in fatigue or subject to impact loads, or where space limitations do not permit the use of the necessarily bulkier magnesium alloy castings. Compositions and Properties of Magnesium Forging Alloys The nominal compositions and properties of the principal commercial magnesium forging alloys are given in Table 1. The two that are most commonly used are AMC58S and AM65S. The former has the highest strength and is, therefore, employed for heavy duty applications. However, the forging of this alloy and of AMC57S ordinarily requires the use of slow speed hydraulic presses. The forging alloy AM65S combines good, although somewhat lower, mechanical properties with excellent forging characteristics. This alloy can be forged on hammers, upsetters, rolls, and mechanical presses, as well as on hydraulic presses. These characteristics make it particularly useful for many applications. Manganese Segregation in AM65S The alloy AM65S contains 5 pct tin, 3 1/2 pct aluminum, and a small amount of manganese, the manganese being added to insure satisfactory resistance to corrosion. Some manganese segregation was occasionally encountered in this alloy. In many applications, such segregation is of no particular consequence, but when forgings are acid dipped prior to the application of various chemical coatings, the segregated particles of manganese constituent may stand in relief. The resultant surface roughening may impair appearance and is undesirable in forgings employed in the textile industry, where very smooth surfaces are essential to avoid the catching and breaking of fine threads. An effort was therefore made to determine the cause of manganese segregation in this alloy and find a method for its elimination. Radiographic examination of a series of cast ingot sections and extruded samples of the AM65S type having manganese contents ranging from zero to about 0.7 pct revealed no appreciable manganese segregation when the manganese was less than about 0.4 pct. A small amount was observed with a manganese content of 0.46 pct and considerable with manganese contents of 0.6 pct or higher. Attempts to eliminate the manganese segregation by variations in the melting and casting conditions were unsuccessful and it soon became apparent that the trouble was attributable to the fact that the manganese contents that caused segregation were in excess of the amounts that would remain in solution at temperatures encountered in the melting and/or casting operations prior to the complete solidification of the ingots. A knowledge of the amounts of manganese that would be soluble in molten alloys of the AM65S type at various temperatures was obviously desirable. The liquid solubilities of manganese in binary Mg-Mn alloys and in Mg-Al-Mn and Mg-Zn-Mn alloys have been determined by Tiner and others,* but no data on the solubility of manganese in molten magnesium alloys containing tin and aluminum are given in the literature. For this reason, an investigation to determine the solubility of manganese in a magnesium alloy containing 5 pct tin and 3.5 pct aluminum (AM65S type) was undertaken.

Jan 1, 1950

By M. E. Learmont, R. J. Lipp, I. Iwasaki

In the selective flocculation of iron ores, the concentration of alkaline earth ions in pulp solutions plays a critical role. Sodium silicate, alone or with sodium tripolyphosphate, has been used as a dispersant. When Ca+ + concentration exceeds 15 to 20 ppm, however, only a strong complexing agent, such as ethylenediaminetetraacetate (EDTA) was found to be effective (Heerema, Lipp, and Iwasaki, 1979). Some iron ores release Ca+ + in excess of 20 ppm. With ores that contain gypsum, the Ca+ + concentration may reach several hundred ppm. The purpose of this investigation was to compare the effectiveness of EDTA with nitrilotriacetate (NTA) in countering the adverse effects of Ca+ +, particularly at high Ca+ + concentrations. Emphasis was placed on the use of NTA since it is more cost-effective than EDTA. It is also used in detergents in some countries (Malaiyandi, Williams, and O&apos;Grady, 1979). NTA, like EDTA, forms a soluble complex with Ca+ + with a 1:1 molar ratio. Its structure and the value of log Kca are given below together with EDTA (Shibata, 1963). [ ] For experimental materials, an equal weight mixture of Lac Jeannine hematite and a high-purity quartzite from Thompson, MN, both ground to - 30 µm (- 560 mesh) was used. Selective flocculation test conditions and procedures were identical to those used in previous investigations (Heerema, Lipp, and Iwasaki, 1979). The optimum level of starch addition for the above mixture was determined to be at 300 g/t (0.6 lb per st). Initially, a series of selective flocculation tests were performed on the mixture pulped with 10-3M CaC12 (40 ppm Ca + +) at pH 11. With no NTA present, flocculation of the entire suspension resulted. The selectivities achieved in tests with 90, 100, and 150% of the stoichiometric amount of NTA were virtually identical to each other and also to the selectivity of separation obtained in the absence of Ca+ + Such an observation could have been expected since Ca+ + concentrations up to a few ppm may be tolerated in selective flocculation (Krishnan and Iwasaki, 1982). A similar series of tests were performed at 10-2M CaC12 (400 ppm Ca+ +). The results are shown in Fig. 1 as grade¬recovery curves. For comparison, the results obtained in the absence of Ca++ are included. At 90% of the stoichiometric amount of NTA, a large enough concentration of uncomplexed Ca+ + was left in solution to cause indiscriminate flocculation. Note in this figure that NTA is capable of countering the adverse effect of Ca+ + even at this high concentration when it is used in excess of stoichiometric amounts. The free Ca+ + concentrations at 100 and 150% are calculated to be 2.5 and 0 ppm, respectively. [ ] The relative effectiveness of NTA and EDTA was then compared at 100% of the stoichiometric amount using 10-2M and 10-3M CaCI2 (Fig. 2). EDTA appears to be slightly more effective than NTA. The free Ca++ of 0.02 ppm in 10-2M CaC12 with EDTA compared to 2.5 ppm free Ca+ + with NTA might account for the difference. As evident, prohibitively high additions of EDTA and NTA were required to complex high concentrations of calcium ion. The use of both of these agents would result in the accumulation of complexed Ca+ + in recycle water.

Jan 1, 1985

By E. W. Davis

ABOUT a year ago several carloads of magnetite ore from New York state were sent to the Mines Experiment Station at the University of Minnesota for grinding and concentration tests. The flowsheet used minus ½ -in. ore as feed to a 3 by 3-ft. overflow cylindrical ball mill operating at 37 r.p.m. in closed ciruit with a vibrating screen. The screen oversize was returned to the mill for further grinding and the screen undersize was concentrated magnetically. Several different screens, between 8 and 35 mesh, were used and it was desired to determine the capacity of the ball mill operating in closed circuit with each screen. From earlier tests using this flowsheet and equipment, it was known that when comparatively coarse screens were used it was necessary to increase the grinding dilution in the mill beyond that ordinarily used when grinding in closed circuit with classifiers. It was also known that after the circulating load had passed about zoo per cent of the original feed it tended to increase rapidly in amount, and grinding conditions appeared to become unstable. These conditions may be peculiar to this small ball mill but they were very marked in this series of tests. If conditions were not right, the mill would suddenly "go dead," and it was found desirable to operate so that the occasional rattle of balls against the lining of the mill could be detected. As the test proceeded, it was found that the cure for almost any difficulty that developed in the grinding circuit was to add more water to the mill feed. To begin with, the mill was operated at pulp densities of from 60 to 65 per cent solids but gradually it was found that with higher dilutions grinding conditions were more stable and the ball-mill power per ton of screen undersize decreased. Actually, the most satisfactory operating conditions were secured with a pulp density of about 40 per cent solids. The ore was hard but it was possible to grind ½ in. feed so that all would pass an 8-mesh screen with a power expenditure in the ball mill of less than 2 1/2 kw-hr. per ton and with a circulating load of less than 50 per cent. With so high a dilution in the feed to the mill, it was feared that steel wear would be very high. However, the mill was not noisy, and in order to find exactly the grinding conditions, a special test was made at the end of the main series of tests. After the mill had been in operation for several hours and grinding conditions had become balanced, the power to all the equipment was shut off at a given signal. The ball mill made about one revolution after the power was turned off, and just as soon as the mill stopped, a spout was placed at the discharge trunnion and all of the pulp that flowed out of the mill was caught in tanks. The manhole in the mill was then opened and the whole mill charge dumped into other tanks. Both of these products were weighed, dried, and weighed again. The balls were screened out and weighed, and the dry solids were sized. In this way it was possible to determine the nature of the pulp and the ore within the mill under actual grinding conditions.

Jan 1, 1945

By E. S. Lenker, H. J. Van Hook

A complete swies of solid solutions has been found between the compounds InAs and GaAs Below the solidus. The melting relations determined by differential thermal and static quenching technzques, indicate that the liquidus and solidus are of the simple ascendant type. No evidence of unmixing of the homogeneous solid solutions was observed at sub-solidus temperatures. A quenching method for locnting the solidus, using X-ray diffraction techniques, is described in some detail. THE semi-conducting properties of compounds formed by various combinations of the less familiar intermetallic elements have created an increasing interest in the phase relations in systems involving these elements. The complete phase diagram of such a system, containing a compound of special interest, will often provide information as to the most favorable conditions for preparing large crystals from a melt, as well as specifying the temperature stability of the crystalline phase or phases produced. In view of the current importance of gallium arsenide in semiconductor research and of the structural similarities of indium arsenide, it was thought that the complete phase relations of the system InAs-GaAs should be investigated. Previous work on the system Ga-As by Koster and homa&apos; and on In-As by Liu and peretti2 has shown that GaAs and InAs are the only compounds formed, having melting points of 1238° and 942°C, respectively. In a survey of various systems of elements from groups IIIb and Vb, spengler3 also re- Mines Ltd., for his continuing encouragement and for permission to publish the results of this work. ports complete miscibility of gallium, indium, and arsenic in the liquid state. In a partial investigation of the InAs-GaAs system, Woolley and smith4 found extensive solid solution throughout the alloy range. Their subsolidus work, however, was inconclusive at temperatures below 900" and they did not determine the liquidus relationships of the system. PRESENT INVESTIGATION The starting materials used were high-purity gallium, indium, and arsenic. The method of preparation of the compounds follows that devised by M. Shafer and is described in detail in a paper by Shafer and eiser.&apos; The end-member compounds, GaAs and InAs, were first prepared by reaction of the elements in the desired proportions in evacuated silica glass vials. The vials were rotated in a furnace as the temperature was slowly raised to insure rapid and complete reaction at low temperatures. Thereafter the starting materials, GaAs and InAs, were melted by heating the sealed vials to 1300" and 1000°C, respectively; the vials in this case were filled as completely as possible to keep the vapor space small and thereby minimize the loss of arsenic to the vapor phase. Ten binary mixtures of these compounds were then prepared by reaction of the powdered material in the solid state just below solidus temperatures. At 900C preparations containing 3, 5, 8, 10, 25, 35, 50, and 75 mol pct GaAs were found to be single homogeneous phases, indicating a complete series of solid solutions between GaAs and InAs. Moreover, it was discovered that the homogeneity of any crystalline binary preparation could be very accurately checked by X-ray diffraction because of the large differences in unit cell dimensions of the end membe5s. The cubic lattice constant ofor GaAs is a = 5.652A and that of InAs is a = 6.061A

Jan 1, 1963

By A. H. Larson, R. J. McClincy

ALTHOUGH arsenic has a very high vapor pressure and vacuum distillation appears to be a likely method for its removal from lead, the published literature contains virtually no information concerning the vacuum removal of arsenic from lead. Caldwell, Spendlove, and St. Claire1 mention arsenic in their investigation of the removal of volatile metals from lead by vacuum distillation. They do not, however, give any data for arsenic removal or refer to any actual experimental work involving arsenic dissolved in lead. Moriyama, Kozuka, and Miura2 have determined the activity of arsenic in dilute As-Pb alloys in the temperature range from 450" to 600°C using a closed galvanic cell. They found that the activities of both arsenic and lead showed a positive deviation from ideality at 600°C. In view of the lack of experimental data regarding the vaporization of arsenic from As-Pb alloys, a limited vapor pressure study was made on liquid alloys containing from 0.5 to 3.3 wt pct As in lead at 703°C. The pressure of As4(v) above the alloys studied was determined from weight-loss data collected by means of the inert-gas saturation method. The experimental apparatus and procedure used were similar to that discussed by Alcock and Hooper9 with modifications as described by the authors., The technique involved was one of passing a known volume of inert gas over the alloy to be studied in a known amount of time and determining the loss in weight of the sample. Because of the low vapor pressure of lead at 703"C, the observed weight loss could be attributed solely to the vaporization of arsenic as the tetratomic molecule AS43 Measurements of the apparent vapor pressure were repeated at different inert-gas flow rates until from three to five values were obtained which were independent of inert-gas flow rate. This value of the apparent vapor pressure was taken to be the equilibrium vapor pressure of the alloy under consideration. The vapor pressures measured in this investigation have been converted to the activities of arsenic in As-Pb alloys, Table I, by use of the data compiled by Elliott and Gleiser5 for the vapor pressure of pure arsenic. The standard state for arsenic was chosen as the pure, hypothetical, supercooled liquid. The activity of arsenic was calculated from the fourth root of the ratio PAs,/PAS,. The low magnitude of the vapor pressures given in Table I explains the difficulty en- countered by earlier workers in successfully vacuum distilling arsenic from very dilute As-Pb alloys. The data shown in Fig. 1 indicate that the As-Pb system behaves ideally. The apparent outlier represented by the alloy containing 1.43 mol pct As is a result of the fact that the pressure of As, above the alloy was barely detectable by the experimental procedure used in this investigation. Because of this, the apparent vapor pressure determined for this alloy could be in considerable error. Moriyama, Kozuka, and ~iura&apos; have reported that arsenic shows a positive deviation from ideality in dilute solution in lead. However, details of their results have not been made available. This research was made possible by a grant from the National Science Foundation (GK-196) for which the authors are very appreciative.

Jan 1, 1970

By H. R. Stahi

THE ore of the Southeast Missouri lead district consists essentially of galena in dolomitic limestone. The galena usually is in a very finely disseminated condition. The experience of recent years in the concentrating plants of the St. Joseph Lead Co. has shown that dry crushing carried to a point where about 90 per cent of the ore passes a 14-mesh screen liberates the larger portion of the galena, and minimizes further necessary reduction in rod or ball mills. Concentration practice has attempted to carry size reduction to a point where economic unlocking of the galena is obtained, without causing undue sliming of the very friable galena. While 60 per cent of the tonnage of the district is treated by the flotation process, only 40 per cent of the total concentrates is produced by that means, as efforts have always been made to keep the grade and tonnage of flotation feed as low as possible. The milling practice of the district for some years has recognized the generally accepted idea of better selective grinding in rod mills as compared with ball mills, with the attendant decrease in production of fines. The function of rod and ball mills in the various concentrating plants is the grinding of table middlings and coarse tailings that carry lead values too high to be discarded. The simplified flowsheet in Fig. I illustrates the place of grinding mills in the circuit. The arrangement of classifiers, tables and mills constitutes a closed circuit within somewhat flexible limits. In 1937 experiments were conducted in replacing the ball charge in a 6 by 4-ft. ball mill with rods 2 in. in diameter and 51 in. long. This step was taken with some hesitancy, as it was easy to picture a mass of tangled and crossed rods. However, no difficulties developed, and after a short trial period, six more ball mills were converted to rod mills. These mills are of the spur gear and pinion, belt-driven type, using in most cases a 75-hp. motor and having a 9 1/2 -in. open trunnion discharge. As ball mills, they carried a charge of 16,000 to 20,000 lb. of balls, 2 in. in original size. The feed to the mills was dewatered by dewatering wheels and introduced at 70 to 75 per cent solids, with a single scoop feeder. One change in mill design was found advisable when rods were introduced. The existing end liners had an outward flare of 3 ½ in. in 12 in. These were replaced by linings having vertical interior faces, in order to prevent undue end travel of the rods. This change gave an effective working length of 5 2 ¾ in. inside the mill. No alterations were made in the shell liners, which are of the smooth type, made of manganese steel 2 in. in original thickness. The rods at present in use are 51 in. long and 1 3/4in. in diameter. The composition has been varied at times; at present they contain 0.70 per cent carbon and 0.70 per cent manganese. Some breakage and etching occurs as the rods become small, but this is considered preferable to using softer rods, which may kink and cause tangling of the charge. The mill charges are renovated occasionally, broken pieces being removed and the charge brought up

Jan 1, 1945