Search Documents

By H. P. Moore, R. W. Fenn, Harold Margolin, W. R. Hibbard

Deformation lines, also called etch markings or strain markings, are non-effaceable lines developed in individual grains by etching a metal specimen which has been cold worked sufficiently to cause atomic non-conformity or slight orientation differences within the grains. Etching occurs more rapidly at these lines of high chemical reactivity to produce a result similar to that revealing grain boundaries. Wider etching effects forming bands have been found. They are believed to be bands of deformation lines which form areas of measurably different orientation from the crystal matrix. The crystallite rotations causing these bands have been shown by Barrett2 to contribute to deformation textures. Although deformation lines have been observed in many metals, the amount of deformation requireti for their appearance has been established only in general terms. Mathewson and Phillips3 observed the first appearance of deformation lines in cold-rolled alpha brass at about 20 pct reduction in thickness using the standard ammonium hydroxide hydrogen peroxide etchant. Price and Davidson4 noted defor- mation lines in common high brass after 26.5 pct reduction by cold rolling using the same etchant. Johnson6 observed dark thick forklike lines in 99.99 pct pure copper reduced 31.8 pct. Adcock6 observed deformation lines in 80-20 copper-nickel alloy rolled 50 pct. Samans7 found no deformation lines in alpha brass single crystals cold rolled 9.5 pct but they did appear after 25 pct reduction using ammonium hydroxide and hydrogen peroxide etchant. Barrett and Levenson8 observed deformation bands in drawn polycrystalline iron at very small reductions in area and noted that their initial appearance perhaps depends on grain size and other conditions. They reported deformation bands in a single crystal of iron reduced 6 pct by drawing. Brickg noted that deformation lines in 70-30 brass apparently associated with I I I I ) planes appeared after reductions of from 20-50 pct by cold rolling. Brick and Williamson10 noted that in single crystals of 70-30 alpha brass "as in polycrystalline brass, no deformation markings appeared until reductions exceeded 20 pct." Cook and Richards" reported that in Copper cold rolled less than about 50 Pet few strain markings were apparent but for higher reductions many strain markings appeared transverse to the rolled direction. In a review article Brick and Phillips12 noted that previous studies had not shown deformation lines in cold rolled alpha brass below 20 pet reduction and added that 4

Jan 1, 1949

By W. R. Hibbard

Although extensive pole figure studies have been reported by Brick1 and others2,3 showing preferred orientations in rolled and annealed 70-30 alpha brass, and by Hermann and Sachs4 in 70-30 alpha brass cups, comparatively little information has been published on textures in drawn 70-30 alpha brass tubes. Crampton5 reported a (110) , [111] texture in drawn brass tubes based on selected individual X ray photograms following the method of Norton and Hiller,6 but the extent and ease of developing this texture and the annealing texture have not been reported and pole figures of these structures were not found in the literature. The present investigation was further stimulated by the possible influence of textures on the mercury stress-corrosion cracking of brass tubes. According to Snoek,7 iron-nickel alloy sheet with a strong cubic texture was more resistant to corrosion than a randomly oriented sheet. Crampton5 reported that single crystals of 70-30 alpha brass drawn into tubes did not crack after 24 hr in a standard mer-curous nitrate test. Edmunds12 found that single crystals of 70-30 alpha brass did not crack in a standard mercurous nitrate test but did crack when subjected to ammoniacal vapor. Wasserman11 reported a similar failure of a single crystal of brass in ammonia. A tube with a strong texture might approach single crystal behavior and tend to resist mercury cracking if the differences in orientation between adjacent grains were sufficiently slight. Preparation OF Specimens Crampton's5 tubes drawn 28 pct reduction in area developed a slight degree of preferred orientation while in tubes drawn 48 pct it was more marked. Based on this information, in order to obtain specimens with markedly different degrees of preferred orientation but with the same final draw, two sets of 70-30 alpha brass tubes were made following the mill schedules shown in Table I. The final draw involved a large reduction in outside diameter, and a ratio of reduction in area to reduction in outside diameter similar to that reported by Crampton,5, 10 to cause mercury cracking of brass tubes. The brass contained 71.31 pct copper, 0.01 pct lead, 0.01 pct iron and zinc by difference. Determination of Pole Figures A 6-in. length from each tube was rotated rapidly by an electric stirrer in 25 pct nitric acid until the gauge was reduced to 0.003-0.005 in. The tubes were carefully cut longitudinally and the walls gently flattened. They were further etched in 25 pet nitric acid to 0.0015 in. thickness and (111) pole figures were determined using copper K-alpha radiation in the manner described by Brick.' The three general types of pole figures

Jan 1, 1949

By W. R. Hibbard

Although extensive pole figure studies have been reported by Brick1 and others2,3 showing preferred orientations in rolled and annealed 70-30 alpha brass, and by Hermann and Sachs4 in 70-30 alpha brass cups, comparatively little information has been published on textures in drawn 70-30 alpha brass tubes. Crampton5 reported a (110) , [111] texture in drawn brass tubes based on selected individual X ray photograms following the method of Norton and Hiller,6 but the extent and ease of developing this texture and the annealing texture have not been reported and pole figures of these structures were not found in the literature. The present investigation was further stimulated by the possible influence of textures on the mercury stress-corrosion cracking of brass tubes. According to Snoek,7 iron-nickel alloy sheet with a strong cubic texture was more resistant to corrosion than a randomly oriented sheet. Crampton5 reported that single crystals of 70-30 alpha brass drawn into tubes did not crack after 24 hr in a standard mer-curous nitrate test. Edmunds12 found that single crystals of 70-30 alpha brass did not crack in a standard mercurous nitrate test but did crack when subjected to ammoniacal vapor. Wasserman11 reported a similar failure of a single crystal of brass in ammonia. A tube with a strong texture might approach single crystal behavior and tend to resist mercury cracking if the differences in orientation between adjacent grains were sufficiently slight. Preparation OF Specimens Crampton's5 tubes drawn 28 pct reduction in area developed a slight degree of preferred orientation while in tubes drawn 48 pct it was more marked. Based on this information, in order to obtain specimens with markedly different degrees of preferred orientation but with the same final draw, two sets of 70-30 alpha brass tubes were made following the mill schedules shown in Table I. The final draw involved a large reduction in outside diameter, and a ratio of reduction in area to reduction in outside diameter similar to that reported by Crampton,5, 10 to cause mercury cracking of brass tubes. The brass contained 71.31 pct copper, 0.01 pct lead, 0.01 pct iron and zinc by difference. Determination of Pole Figures A 6-in. length from each tube was rotated rapidly by an electric stirrer in 25 pct nitric acid until the gauge was reduced to 0.003-0.005 in. The tubes were carefully cut longitudinally and the walls gently flattened. They were further etched in 25 pet nitric acid to 0.0015 in. thickness and (111) pole figures were determined using copper K-alpha radiation in the manner described by Brick.' The three general types of pole figures

Jan 1, 1949

By W. R. Hibbard

The term lineage was first introduced by Buergerl to denote dendritic branches, grown from a crystal nucleus during solidification from the liquid, with imper- fections in alignment of the order of 10 : to 10-4 cm. These imperfections are associated with distortions of the dendrites caused by the subsequent solidification of the inter-dendritic filling and are illustrated in Fig I. Buerger etched zinc single crystals made by the Bridgman3 method to develop this dendritic pattern and show lineage microscopically. Since maximum crystal growth in face-centered cubic metals occurs along the [IT] directions,2 lineage would appear in copper as small deviations of dendritic branches from these directions. Elam4 and others5"8 reported etching effects resulting from coring along the cube face traces of various copper alloy crystals. These traces were used by Samans8 to orient single crystals. Williamson9 noted a "lined structure (in alpha brass single crystals) consisting of an ordered arrangement of etch pits" which were probably traces of {IOO} planes but somewhat uncertain because they were never straight (that is, apparently lineage). Greningerl1 described the distortion and multiplicity of Laue spots in back reflection X ray photograms of copper single crystals showing lineage. The present investigation -includes metal-lographic, X ray, and property studies of lineage in large grain and single crystal specimens of copper. Preparation of Specimens specilnens 10 in. long 35 1/2 in. diam of oxygen-free copper (o.F.H.c.) (99.985 pet) were' melted vertically in graphite crucibles under an atmosphere of commercial nitrogen purified through a gas train including sulfuric acid, potassium chloride, magnesium chiorate, hot copper chips and phosphorus pentoxide. After a temperature of 1200°C had been maintained for 1/2 hr, the furnace was turned off and the specimens cooled at a rate of about 150°c an hour' This method of direct solidification was used by Greningerlo to produce single crystals with marked lineage and in this investigation produced a fair yield of single crystal specimens in

Jan 1, 1949

By W. R. Hibbard

The term lineage was first introduced by Buergerl to denote dendritic branches, grown from a crystal nucleus during solidification from the liquid, with imper- fections in alignment of the order of 10 : to 10-4 cm. These imperfections are associated with distortions of the dendrites caused by the subsequent solidification of the inter-dendritic filling and are illustrated in Fig I. Buerger etched zinc single crystals made by the Bridgman3 method to develop this dendritic pattern and show lineage microscopically. Since maximum crystal growth in face-centered cubic metals occurs along the [IT] directions,2 lineage would appear in copper as small deviations of dendritic branches from these directions. Elam4 and others5"8 reported etching effects resulting from coring along the cube face traces of various copper alloy crystals. These traces were used by Samans8 to orient single crystals. Williamson9 noted a "lined structure (in alpha brass single crystals) consisting of an ordered arrangement of etch pits" which were probably traces of {IOO} planes but somewhat uncertain because they were never straight (that is, apparently lineage). Greningerl1 described the distortion and multiplicity of Laue spots in back reflection X ray photograms of copper single crystals showing lineage. The present investigation -includes metal-lographic, X ray, and property studies of lineage in large grain and single crystal specimens of copper. Preparation of Specimens specilnens 10 in. long 35 1/2 in. diam of oxygen-free copper (o.F.H.c.) (99.985 pet) were' melted vertically in graphite crucibles under an atmosphere of commercial nitrogen purified through a gas train including sulfuric acid, potassium chloride, magnesium chiorate, hot copper chips and phosphorus pentoxide. After a temperature of 1200°C had been maintained for 1/2 hr, the furnace was turned off and the specimens cooled at a rate of about 150°c an hour' This method of direct solidification was used by Greningerlo to produce single crystals with marked lineage and in this investigation produced a fair yield of single crystal specimens in

Jan 1, 1949

By C. H. Mathewson, W. R. Hibbard, Robert Maddin

A satisfactory mathematical relationship between shear, elongation and change of orientation in the axial straining of a single crystalline rod or wire may be based on a simple model in which blocks, representing glide lamellae, glide over one another and rotate so as to maintain the original fixed position of the axis as the wire elongates.' Obviously such a model gives no information concerning strain-hardening or the atomic adjustments at the sliding surfaces and their influence on the internal structure of the blocks. Moreover such an operation could be conducted with rigid blocks only by using self-aligning grips or a flexible attachment which would permit the adjacent blocks to freely assume their changing orientations. There are many unsolved problems (see Andrade2) concerning the manner in which nature departs from the simple model in permitting the essentials of slip, rotation and elongation to occur along with strain-hardening and the appearance of glide lamellae of various dimensions and special characteristics, not fully appreciated at the present time. This paper describes part of an extensive investigation on the development and characteristics of glide ellipses in single crystalline alpha brass rods and is especially concerned with the observation of cross-slip in crystals oriented wholly within the range of slip on a single system according to the generalizations first presented by Taylor and Elam in 1923.³ Experimental Procedure Half-inch rods of alpha brass (nominally 70 pet Cu, 30 pet Zn)* were converted into single crystals by melting and gradual solidification in a graphite mold according to the Bridgman method.4 In order to remove any coring, the crystals were homogenized at 800°C for 16 hr and tested by etching with a 50 pet nitric acid solution. The back-reflection Laue technique as described by Greninger6 was used for determining the orientations of the crystals. Standard stereographic operations were employed to locate the specimen axis in a triangle of the Taylor and Elam3 projection so that the slip plane and direction could be predicted. The lineage structure reported by Buer-ger6 was present in the crystals used in this investigation. However, only crystals were used in which the internal differences in orientation did not exceed two degrees. Very little is known concerning the effect of the presence of a lineage structure on the detailed mechanism of plastic deformation.

Jan 1, 1949

By C. H. Mathewson, Robert Maddin, W. R. Hibbard

A satisfactory mathematical relationship between shear, elongation and change of orientation in the axial straining of a single crystalline rod or wire may be based on a simple model in which blocks, representing glide lamellae, glide over one another and rotate so as to maintain the original fixed position of the axis as the wire elongates.' Obviously such a model gives no information concerning strain-hardening or the atomic adjustments at the sliding surfaces and their influence on the internal structure of the blocks. Moreover such an operation could be conducted with rigid blocks only by using self-aligning grips or a flexible attachment which would permit the adjacent blocks to freely assume their changing orientations. There are many unsolved problems (see Andrade2) concerning the manner in which nature departs from the simple model in permitting the essentials of slip, rotation and elongation to occur along with strain-hardening and the appearance of glide lamellae of various dimensions and special characteristics, not fully appreciated at the present time. This paper describes part of an extensive investigation on the development and characteristics of glide ellipses in single crystalline alpha brass rods and is especially concerned with the observation of cross-slip in crystals oriented wholly within the range of slip on a single system according to the generalizations first presented by Taylor and Elam in 1923.³ Experimental Procedure Half-inch rods of alpha brass (nominally 70 pet Cu, 30 pet Zn)* were converted into single crystals by melting and gradual solidification in a graphite mold according to the Bridgman method.4 In order to remove any coring, the crystals were homogenized at 800°C for 16 hr and tested by etching with a 50 pet nitric acid solution. The back-reflection Laue technique as described by Greninger6 was used for determining the orientations of the crystals. Standard stereographic operations were employed to locate the specimen axis in a triangle of the Taylor and Elam3 projection so that the slip plane and direction could be predicted. The lineage structure reported by Buer-ger6 was present in the crystals used in this investigation. However, only crystals were used in which the internal differences in orientation did not exceed two degrees. Very little is known concerning the effect of the presence of a lineage structure on the detailed mechanism of plastic deformation.

Jan 1, 1949

By B. J. Nelson, F. N. Rhinos

Studies upon the rates of oxidation of copper alloys containing small quantities of the alloying elementsl,2 have shown that steady growth of the scales at predictable rates is limited to a small concentration range and to oxidation at temperatures above 600°C., whereas unpredictable and erratic oxidation behavior is found with alloys containing larger additions and at low temperatures. The latter effect appears to be associated with modifications in the geometric disposition of the oxides. At low concentrations of the baser elements in copper, simple oxidation behavior is typified by the formation of two distinguishable zones of oxidation: (I) a subscale composed of discrete particles of the oxide of the alloying element embedded in substantially pure copper and (2) an external scale composed of cuprous oxide (with a very thin layer of cupric oxide on the outside) containing particles of the alloying element distributed throughout, but most profusely near the metal surface (see Fig. 2, top). Where this type of structure obtains, the process of oxidation is believed to be implemented, and its rate controlled, by the diffusion of the reacting elements through a' continuous matrix phase; i.e., through the alpha (metallic) phase of the subscale and alloy and through the cuprous oxide of the external scale. Copper is presumed to diffuse outward through the cuprous oxide layer to the external surface, where it reacts to form more cuprous oxide.3 Oxygen is delivered at the oxide-metal interface in the form of cuprous oxide, which may either react with the alloying element to deposit its oxide at the inner limit of the external scale or dissolve in the pure copper that remains at the surface of the metal after it has become depleted in its oxidiza-ble alloying elements. Both oxygen and the alloying element are pictured as diffusing through the metal, the oxygen inward and the alloying element outward. Where the two diffusion streams meet in sufficient concentration the oxide of the subscale is precipitated. If the oxygen pressure in the surrounding atmosphere is just below the pressure necessary to maintain cuprous oxide undecomposed, no cuprous oxide can be formed; some of the oxide of the alloying element may form upon the external surface, if its decomposition pressure is below that of cuprous oxide, and a normal sub-scale appears. This condition is achieved experimentally by enclosing the alloy with cuprous oxide together with an excess of copper. At temperatures from 600°C. downward, the precipitated oxides tend to accumulate at the grain boundaries of the alpha phase where a relatively small volume establishes a partial barrier to the diffusion of the reactants- Similarly, large volumes of Precipitated oxides tend to deposit in various patterns that may be expected to

Jan 1, 1944

By F. N. Rhinos, B. J. Nelson

Studies upon the rates of oxidation of copper alloys containing small quantities of the alloying elementsl,2 have shown that steady growth of the scales at predictable rates is limited to a small concentration range and to oxidation at temperatures above 600°C., whereas unpredictable and erratic oxidation behavior is found with alloys containing larger additions and at low temperatures. The latter effect appears to be associated with modifications in the geometric disposition of the oxides. At low concentrations of the baser elements in copper, simple oxidation behavior is typified by the formation of two distinguishable zones of oxidation: (I) a subscale composed of discrete particles of the oxide of the alloying element embedded in substantially pure copper and (2) an external scale composed of cuprous oxide (with a very thin layer of cupric oxide on the outside) containing particles of the alloying element distributed throughout, but most profusely near the metal surface (see Fig. 2, top). Where this type of structure obtains, the process of oxidation is believed to be implemented, and its rate controlled, by the diffusion of the reacting elements through a' continuous matrix phase; i.e., through the alpha (metallic) phase of the subscale and alloy and through the cuprous oxide of the external scale. Copper is presumed to diffuse outward through the cuprous oxide layer to the external surface, where it reacts to form more cuprous oxide.3 Oxygen is delivered at the oxide-metal interface in the form of cuprous oxide, which may either react with the alloying element to deposit its oxide at the inner limit of the external scale or dissolve in the pure copper that remains at the surface of the metal after it has become depleted in its oxidiza-ble alloying elements. Both oxygen and the alloying element are pictured as diffusing through the metal, the oxygen inward and the alloying element outward. Where the two diffusion streams meet in sufficient concentration the oxide of the subscale is precipitated. If the oxygen pressure in the surrounding atmosphere is just below the pressure necessary to maintain cuprous oxide undecomposed, no cuprous oxide can be formed; some of the oxide of the alloying element may form upon the external surface, if its decomposition pressure is below that of cuprous oxide, and a normal sub-scale appears. This condition is achieved experimentally by enclosing the alloy with cuprous oxide together with an excess of copper. At temperatures from 600°C. downward, the precipitated oxides tend to accumulate at the grain boundaries of the alpha phase where a relatively small volume establishes a partial barrier to the diffusion of the reactants- Similarly, large volumes of Precipitated oxides tend to deposit in various patterns that may be expected to

Jan 1, 1944

By C. Ernest Birchenall, Robert F. Mehl

Application of diffusion laws in the customary form to experimental studies in binary metallic solid solutions has shown the diffusion cocfficient to vary with concentration for all systems investigated. The factors determining this variation have not been ascertained. However, the question of whether the diffusion coefficient would become a constant, independent of concentration throughout a single phase if the concentration gradient in the diffusion equation were replaced by an activity gradient has been an attractive subject for speculation. Wagner¹ derived an expression for the force acting on a diffusing particle in terms of the chemical potential. This was also the approach chosen by Jost2 in his monograph. From a somewhat different point of view, Eyring3 and his co-workers have used activity gradients when dealing with diffusion in concentrated nonideal solutions. The only actual test with experimental data is that given by Darken,4 though this is more qualitative than quantitative. Many others opposed the application in any way of thermodynamic concepts to kinetic phenomena, including diffusion processes. A rigorous examination of this principle requires good data expressing the relationship between activity and concentration for solid solutions for which reliable diffusion work has been done. The literature provides such information only for the interstitial solution of carbon in gamma iron (austenite) and for the substitutional copper-zinc solutions (alpha brass). Activity data for carbon in austenite at 800°C and 1000°C are available from the recent work of Smith5 on the equilibrium of hydrogen-methane and carbon monoxide-carbon dioxide mixtures with gamma iron. Combining these with the diffusion studies of Wells and Mehl6 makes possible a quantitative test for this interstitial solid solution. The vapor-pressure determinations for zinc over brass by Hargreaves7 allowed the calculation of activities in this system over a wide range of temperatures and concentrations. Extensive investigation of diffusion in this substitutional solid solution has been carried out by Rhines and Mehl.8 Activities in Brass9 If the vapor phase of zinc is assumed to behave as an ideal gas, any solid or liquid solutions that are in equilibrium with the same partial pressure of zinc at the same temperature have the same thermodynamic activity of zinc. In order to assign a numerical value to the activity azn, it is necessary to choose some standard reference state to which an activity value of unity is arbitrarily assigned. It is customary to choose the liquid or solid state of the pure material. Then azn is defined numerically by

Jan 1, 1947

By C. Ernest Birchenall, Robert F. Mehl

Application of diffusion laws in the customary form to experimental studies in binary metallic solid solutions has shown the diffusion cocfficient to vary with concentration for all systems investigated. The factors determining this variation have not been ascertained. However, the question of whether the diffusion coefficient would become a constant, independent of concentration throughout a single phase if the concentration gradient in the diffusion equation were replaced by an activity gradient has been an attractive subject for speculation. Wagner¹ derived an expression for the force acting on a diffusing particle in terms of the chemical potential. This was also the approach chosen by Jost2 in his monograph. From a somewhat different point of view, Eyring3 and his co-workers have used activity gradients when dealing with diffusion in concentrated nonideal solutions. The only actual test with experimental data is that given by Darken,4 though this is more qualitative than quantitative. Many others opposed the application in any way of thermodynamic concepts to kinetic phenomena, including diffusion processes. A rigorous examination of this principle requires good data expressing the relationship between activity and concentration for solid solutions for which reliable diffusion work has been done. The literature provides such information only for the interstitial solution of carbon in gamma iron (austenite) and for the substitutional copper-zinc solutions (alpha brass). Activity data for carbon in austenite at 800°C and 1000°C are available from the recent work of Smith5 on the equilibrium of hydrogen-methane and carbon monoxide-carbon dioxide mixtures with gamma iron. Combining these with the diffusion studies of Wells and Mehl6 makes possible a quantitative test for this interstitial solid solution. The vapor-pressure determinations for zinc over brass by Hargreaves7 allowed the calculation of activities in this system over a wide range of temperatures and concentrations. Extensive investigation of diffusion in this substitutional solid solution has been carried out by Rhines and Mehl.8 Activities in Brass9 If the vapor phase of zinc is assumed to behave as an ideal gas, any solid or liquid solutions that are in equilibrium with the same partial pressure of zinc at the same temperature have the same thermodynamic activity of zinc. In order to assign a numerical value to the activity azn, it is necessary to choose some standard reference state to which an activity value of unity is arbitrarily assigned. It is customary to choose the liquid or solid state of the pure material. Then azn is defined numerically by

Jan 1, 1947

By A. D. Smigelskas, E. O. Karenna

This is the third paper in a series on the diffusion of zinc in alpha brass.' At the time of the first paper it was accepted that diffusion in a substitutional type of solid solution depended upon an equal interchange of solute and solvent atoms across the diffusion interface (or past a given lattice reference point). Discrepancies in the size of the areas under the diffusion curves on each side of the interface, if observed, were explained on some other basis than the fact that the solute atoms might cross the diffusion interface without an equal number of solvent atoms diffusing across the interface in the opposite direction. The second paper presented evidence indicating that zinc atoms in alpha brass diffuse across the interface more rapidly in one direction than the copper atoms diffuse across the interface in the opposite direction. This phenomenon forced the interface to move in the opposite direction of the diffusion of zinc atoms to provide space for the additional zinc atoms dissolved in the alpha brass. That the interface and columnar grain structure on one side would remain intact during this mass movement was not expected; how ever, it is gratifying that they did. While the evidence was sufficiently satisfying to the author, it was decided that additional experiments were in order to prove that zinc diffuses more rapidly than copper in alpha brass. It is the purpose of this paper to present the results of these experiments. Procedure A bar of 70-30 wrought brass was selected, to avoid the relatively large volume change in going from beta to alpha brass as in the previous investigation. A bar 7 in. (18-cm.)* long and about 3/4 in. (1.9 cm.) wide was surface ground and polished with abrasive paper on plate glass to 0.4184 0.0002 in. (1.062j + 0.0005 cm.) thick. Six0.00j-in. (0.013-cm.) diameter molybdenum wires, chosen because of their very low solubility in coppcr and alpha brass, were strctched lengthwise along each of the two plane surfaces. Glass tubes placed at right angles to the wires every 1/2 in. held the wires tightly against the brass. The sample was then emersed in an acid copper sulphate plating bath and electroplated to a depth of about 0.020 in. (0.050 cm.) copper on all sides. The glass tubes were no longer necessary to hold the molybdenum wires in place and were removed. The sample was then electroplated to a depth of

Jan 1, 1947

By E. O. Karenna, A. D. Smigelskas

This is the third paper in a series on the diffusion of zinc in alpha brass.' At the time of the first paper it was accepted that diffusion in a substitutional type of solid solution depended upon an equal interchange of solute and solvent atoms across the diffusion interface (or past a given lattice reference point). Discrepancies in the size of the areas under the diffusion curves on each side of the interface, if observed, were explained on some other basis than the fact that the solute atoms might cross the diffusion interface without an equal number of solvent atoms diffusing across the interface in the opposite direction. The second paper presented evidence indicating that zinc atoms in alpha brass diffuse across the interface more rapidly in one direction than the copper atoms diffuse across the interface in the opposite direction. This phenomenon forced the interface to move in the opposite direction of the diffusion of zinc atoms to provide space for the additional zinc atoms dissolved in the alpha brass. That the interface and columnar grain structure on one side would remain intact during this mass movement was not expected; how ever, it is gratifying that they did. While the evidence was sufficiently satisfying to the author, it was decided that additional experiments were in order to prove that zinc diffuses more rapidly than copper in alpha brass. It is the purpose of this paper to present the results of these experiments. Procedure A bar of 70-30 wrought brass was selected, to avoid the relatively large volume change in going from beta to alpha brass as in the previous investigation. A bar 7 in. (18-cm.)* long and about 3/4 in. (1.9 cm.) wide was surface ground and polished with abrasive paper on plate glass to 0.4184 0.0002 in. (1.062j + 0.0005 cm.) thick. Six0.00j-in. (0.013-cm.) diameter molybdenum wires, chosen because of their very low solubility in coppcr and alpha brass, were strctched lengthwise along each of the two plane surfaces. Glass tubes placed at right angles to the wires every 1/2 in. held the wires tightly against the brass. The sample was then emersed in an acid copper sulphate plating bath and electroplated to a depth of about 0.020 in. (0.050 cm.) copper on all sides. The glass tubes were no longer necessary to hold the molybdenum wires in place and were removed. The sample was then electroplated to a depth of

Jan 1, 1947

By G.A. Moore, C. E. Sims

In previous publications of the writers4-7 it has been shown that vacuum extraction of steel can be carried as close to quantitative completion as desired provided the steel is in the austenitic state and sufficiellt time is allowed. For precise work, the method claims preference because the equipmelit for holding and heating the sample can be made simple, and because crucibles, boats, and other unnecessary materials can be eliminated and the blanks thus made very small. It is well known that the evolution rate of hydrogen is smaller just above the trans. formation than immediately below, as would be expected from the increase in solubility when passing into the gamma state. Many attempts have been made to obtain analyses by extraction just below the transition, using times of approxi. mately 2 hr. while it is true that the rate of evolution drops nearly to zero under these conditions, it has repeatedly been shown that the process is incomplete since on raising the temperature into the austenitic range, evolution will be resumed, usually after a, considerable induction period. The low-temperature evolution process appears to be useful in some cases for comparative tests where the residual crror may be more or less constant between samples, For general use, however, there is no guarantee that the error caused by incomplete extraction is constant; hence, the more laborious but more reliable high-temPe'atu'e Process must be used. At high temperature, an original flash evolution, similar to that at lower temperature, is first observed. After not more than an hour, the flash subsides and cvolution settles down to a regular first-order reaction rate for which the half-life* is usually of the order of a few hours for the 5/8-in. diam cylinder at 1050°C. The regular evolution may be followed for whatever period appears justified, after which the residual gas may be estimated as equal to that which was evolved in the last observed half-life period. Tests have shown that the composition of the evolved gas does not vary greatly during the progress of evolution; therefore the residual gas may safely be taken to have the same composition as the portion col-

Jan 1, 1949

By C. E. Sims, G. A. Moore

In previous publications of the writers4-7 it has been shown that vacuum extraction of steel can be carried as close to quantitative completion as desired provided the steel is in the austenitic state and sufficiellt time is allowed. For precise work, the method claims preference because the equipmelit for holding and heating the sample can be made simple, and because crucibles, boats, and other unnecessary materials can be eliminated and the blanks thus made very small. It is well known that the evolution rate of hydrogen is smaller just above the trans. formation than immediately below, as would be expected from the increase in solubility when passing into the gamma state. Many attempts have been made to obtain analyses by extraction just below the transition, using times of approxi. mately 2 hr. while it is true that the rate of evolution drops nearly to zero under these conditions, it has repeatedly been shown that the process is incomplete since on raising the temperature into the austenitic range, evolution will be resumed, usually after a, considerable induction period. The low-temperature evolution process appears to be useful in some cases for comparative tests where the residual crror may be more or less constant between samples, For general use, however, there is no guarantee that the error caused by incomplete extraction is constant; hence, the more laborious but more reliable high-temPe'atu'e Process must be used. At high temperature, an original flash evolution, similar to that at lower temperature, is first observed. After not more than an hour, the flash subsides and cvolution settles down to a regular first-order reaction rate for which the half-life* is usually of the order of a few hours for the 5/8-in. diam cylinder at 1050°C. The regular evolution may be followed for whatever period appears justified, after which the residual gas may be estimated as equal to that which was evolved in the last observed half-life period. Tests have shown that the composition of the evolved gas does not vary greatly during the progress of evolution; therefore the residual gas may safely be taken to have the same composition as the portion col-

Jan 1, 1949

By T. L. Joseph

It is a distinct privilege to participate in this meeting convened to honor the memory of Henry Marion Howe, a distinguished scientist and metallurgist. Many have added to our rapidly growing fund of metallurgical knowledge but few have made the unique contribution of formulating basic concepts essential to an over-all comprehensive view of the subject of metallography. Of those previously selected to deliver this lecture, several have added a personal touch to their discourse because they were able to refer to their associations with Dr. Howe. It was not my good fortune to know him or to enjoy his inspiring influence. Like most of you, my contact has been with his published works, which are outstanding in that they provide lucid, detailed expositions of the nature and behavior of metals and alloys. His deep understanding and clarity of expression have greatly accelerated our progress in metallography and in the use of metals. The degree to which a man's thinking, experimentation and writing contribute toward an understanding of basic laws and phenomena determines whether his life's work is of transitory or permanent value. This occasion appropriately recognizes the fundamental importance of Dr. Howe's pioneer work in the field of metallurgy.. On two former occasions, gray iron was selected as a subject for discussion. The fact that the production of pig iron has not been discussed before does not reflect any lack of interest on the part of Dr. Howe or former lecturers in the process that underlies the production of steel. Although Dr. Howe's work in his later years centered around metallography, his earlier publications show a considerable interest in the production of iron and steel. He recognized that in the ascending spiral of metallurgical advancement we are not winding a single thread but are weaving a complex fabric. Specific evidence of his interest in the production of pig iron may be found in his book' on The Metallography of Cast Iron and Steel. In explaining why blast furnaces have been in constant use for about 600 years, he says: "The process is economical because it utilizes with extraordinary completeness the calorific power of the coke due to the thorough transfer of heat from the gaseous products of combustion to the charge." This emphasis on heat transfer, overlooked in many more extended and more recent discussions of the blast-furnace process, illustrates Professor Howe's ability to see the essentials. Other distinct advantages of the blast furnace, such as the removal of- much of the sulphur and practically all of the oxygen, were mentioned. It was also stated that although the blastfurnace process is complex in nature, its outward form is comparatively simple, thus permitting large-scale production at small labor costs. All who contemplate the production of metallic iron from its ores by any other process might well consider the

Jan 1, 1947

By T. L. Joseph

It is a distinct privilege to participate in this meeting convened to honor the memory of Henry Marion Howe, a distinguished scientist and metallurgist. Many have added to our rapidly growing fund of metallurgical knowledge but few have made the unique contribution of formulating basic concepts essential to an over-all comprehensive view of the subject of metallography. Of those previously selected to deliver this lecture, several have added a personal touch to their discourse because they were able to refer to their associations with Dr. Howe. It was not my good fortune to know him or to enjoy his inspiring influence. Like most of you, my contact has been with his published works, which are outstanding in that they provide lucid, detailed expositions of the nature and behavior of metals and alloys. His deep understanding and clarity of expression have greatly accelerated our progress in metallography and in the use of metals. The degree to which a man's thinking, experimentation and writing contribute toward an understanding of basic laws and phenomena determines whether his life's work is of transitory or permanent value. This occasion appropriately recognizes the fundamental importance of Dr. Howe's pioneer work in the field of metallurgy.. On two former occasions, gray iron was selected as a subject for discussion. The fact that the production of pig iron has not been discussed before does not reflect any lack of interest on the part of Dr. Howe or former lecturers in the process that underlies the production of steel. Although Dr. Howe's work in his later years centered around metallography, his earlier publications show a considerable interest in the production of iron and steel. He recognized that in the ascending spiral of metallurgical advancement we are not winding a single thread but are weaving a complex fabric. Specific evidence of his interest in the production of pig iron may be found in his book' on The Metallography of Cast Iron and Steel. In explaining why blast furnaces have been in constant use for about 600 years, he says: "The process is economical because it utilizes with extraordinary completeness the calorific power of the coke due to the thorough transfer of heat from the gaseous products of combustion to the charge." This emphasis on heat transfer, overlooked in many more extended and more recent discussions of the blast-furnace process, illustrates Professor Howe's ability to see the essentials. Other distinct advantages of the blast furnace, such as the removal of- much of the sulphur and practically all of the oxygen, were mentioned. It was also stated that although the blastfurnace process is complex in nature, its outward form is comparatively simple, thus permitting large-scale production at small labor costs. All who contemplate the production of metallic iron from its ores by any other process might well consider the

Jan 1, 1947

By A. W. Schlechten, W. J. Kroll

As this paper is written, the only method for the commercial production of lithium metal is by the fusion electrolysis of LiC1-KC1 mixtures, as first proposed by Gunkz.2 The details of the industrial process have not been made public hut Osborg8 stated that an efficiency of more than 90 pct is obtained with a lithium metal recovery above 95 pct and that the metal is 99.5 pct pure. Pletenev9 gave a power consumption of 66 kw-hr per kilogram of lithium metal produced with a salt consumption of 9 kg of LiCl and 0.4 kg of KCI. Pletenev's figures indicate a somewhat lower recovery than that given by Osborg. In a recent report on' German lithium plants, Motock7 gives figures that are also less favorable than those reported by Osborg. At the German plants the lithium recovery in electrolysis was only 83.4 pct, with a power consumption of 140 kw-hr per kilogram of lithium. The metal averaged only 97 pct pure, with sodium and potassium as the chief impurities. The presence of SO2, SO3, SiO2, Ba, Ca, and appreciable amounts of Na and Fe2O3 were especially disturbing in the electrolysis, and it was necessary to run the cell for some time to eliminate most of these impurities and condition the bath. It may be noted here that most of these impurities do not interfere with a thermal reduction process. These facts show that the present method of producing lithium is fairly efficient and would suggest that other methods could not compete. However, fusion electrolysis has certain drawbacks, such as the necessity of supplying costly, low-voltage direct current and of collecting and processing the anodic chlorine if it is not to be wasted. An even greater disadvantage is that lithium chloride must be used for an electrolyte. It is one of the most expensive salts of lithium, since it cannot be produced in anhydrous form by precipitation from aqueous solution, as is done with LiF or Li2CO3. The reduction of lithium by a vacuum process similar to the ferrosilicon method for the production of magnesium would have certain advantages. Small units such as Pidgeon retorts, would give flexibility of production to meet demand, and a small plant would require moderate capital outlay. There is a possibility of using idle Government-owned magnesium plants for such purposes. The raw materials needed for vacuum thermal reduction would not have to be extremely pure, Director, with respect to the iron and calcium Content. This paper describes a series of experiments to determine the effectiveness of various agents for the reduction of lithium compounds. The investigation is part of a program of vacuum metallurgy con-

Jan 1, 1949

By R. A. Stauffer

The chief ore of lithium is spodumene, a lithium-aluminum silicate containing up to 3 pct lithium. The preparation of lithium salts from spodumene is costly because of the low concentration of the metal and the difficulty of separating it from the other alkali metals present. Dr. Kroll's and Dr. Schlecten's recent paper1 has reviewed various methods proposed for preparation of lithium. It will be seen from this paper that the methods proposed to date have required one or another of the lithium salts as a raw material. The high cost of these salts has been the chief item of expense in the lithium metal produced. This situation has made it interesting to investigate the possibility of producing the metal directly from spodumene. When spodumene is calcined at 1000ºC in the presence of lime, lithium oxide is replaced in the ore as is evidenced by the fact that it may be leached from the reaction mixture with water.2 Free energy calculations suggested that lithium oxide might be reduced with aluminum or silicon and the lithium produced distilled from the reaction mixture by heating-to about 1000ºC in a vacuum. These facts taken together suggested that it might be possible to prepare lithium metal directly from spodumene by heating a mixture of the ore, lime, and a suitable reducing agent in vacuum and distilling off the lithium metal as formed. The Replacement of Li2O in Spodumene A series of experiments were undertaken to determine the ratio of lime to spodumene necessary to free substantially all the lithium oxide Present in the Ore. The spodumene used in these experiments was supplied by Foote Mineral Co. in the form of powder of about 200 mesh' Analysis of the spodumene was as follows: PCT SiO2......................... 62.7 Al2O3........................ 29.9 Li2o......................... 4.9 Na2O........................ 0.89 CaO......................... 0.75 MgO........................ 0.49 K2O......................... 0.44 Lime (cp —200 mesh) was added in various ratios to spodumene and thoroughly mixed in a ball mill. Two hundred gram samples of spodumene-lime mixtures were charged loosely in fireclay crucibles and held at 1000ºC in a gas-fired furnace for 5 hrs. The product in each case was ground to pass 140 mesh and then placed in a thimble in a soxhlet extractor. Extraction with water was carried on for 6 hrs. After this period the extract was reduced in volume by evaporation in order to precipitate dissolved lime and was then filtered and titrated with 1/4 normal hydrochloric acid. After correcting for residual calcium oxide in solution, the alkali present was calculated as lithium oxide. The results of these experiments are shown in Table I and Fig I. At lime-spodumene ratios less than 1.5, the products were invariably fused. At the ratio of 1.5 and above, the product was loosely sintered with little evidence of fusion. When the ratio of lime to spodumene was three, replacement of alkali oxides in the

Jan 1, 1949

By R. F. Evans, J. M. Avert

Much has been written of the glamour of magnesium from sea water, the Aladdin-like creation of a huge magnesium plant in the Nevada desert using cheap hydroelectric power from Boulder Dam; the marvels of the Hansgirg process for the electrothermal reduction of magnesium oxide by carbon, and the ferrosilicon process for producing magnesium from dolomite. This is the story of the production of magnesium by the electrolytic process, using steam-electric power supplied by a public utility, in a plant at Painesville, Ohio, 30 miles east of Cleveland on the shore of Lake Erie. The story is not without interest, since it involves several unique developments that provide the answer to a question frequently asked: How is it possible to produce magnesium metal economically at such a location as Painesville, Ohio? Basically, the answer lies in making use of facilities already in existence for the production of soda ash, which makes it possible to manufacture magnesium chloride, the principal raw material, cheaply enough to offset the high cost of electric power. To this is coupled the successful conversion of by-product chlorine into a product required for the war program. The source of magnesium for this opera- tion is the enormous surface deposit of high-grade dolomite in the region south of Toledo, within reasonable freight haul of Painesville. This same geological stratum underlies the Painesville plant, but unfortunately at a depth of about 3000 ft.— apparently too great a depth to permit underground mining to compete with the delivered cost of rock from surface quarries in other areas. The cost of dolomite obviously is an important factor, since the rock contains only some 12 per cent magnesium (20 per cent MgO) so that one dollar on the cost of stone means more than $8 in terms of its magnesium content, and more than that in terms of recoverable magnesium metal. Production of Magnesia from Dolomite In the summer of 1941, the Diamond Alkali Co. was well along with construction of a plant intended for the commercial production of refractory and other grades of magnesia from dolomite. The process to be used was designed to make maximum economic advantage of existing plant and operations by tying the production of magnesia in with the manufacture of soda ash in such a way as to make use of the lime [CaO)., value of calcined dolomite. The process finally adopted after a lengthy research and development program has much in common with the old Clerc-Nihoul process, but includes certain innovations found to be necessary or advantageous. Essentially, the process is based upon the following series of reactions:

Jan 1, 1949