Search Documents

By J. R. Long, E. L. Anderson, R. S. Dean, F. S. Wartman

Titanium has been estimated to comprise about 0.65 per cent of the earth's crust and ranks fourth in abundance among the metallic elements suitable for engineering uses. In spite of this, applications of titanium for structural purposes have not as yet been developed. Its commercial exploitation has been largely confined to the chemical industries and the production of paints and pigments; no doubt because of the difficulties that lie in the way of preparing the pure ductile metal. Published reports, however, indicate that ductile titanium can be prepared and possesses some unique properties. This, together with the occurrence in this country of several important titaniferous deposits, has led the Bureau of Mines to investigate processes for the production of metallic titanium, methods required for its fabrication, and the properties that may be produced in this interesting metal. Methods of Preparing Metallic Titanium Many early investigators claimed to have isolated metallic titanium; but in the light of present knowledge, it seems that they really had produced various mixtures of carbides and nitrides that have a metallic appearance but are brittle, hard, and unworkable. The first metallic titanium pure enough to be malleable when cold was produced by Hunter1 in the early part of the twentieth century. Kroll,~ van Arkel,~ and Fast,4 subsequently confirmed the facts that pure titanium is a. ductile metal and that small amounts of impurities were responsible for previous reports of brittleness. When heated, titanium reacts chemically with various gases, such as hydrogen, oxygen and nitrogen; if the volume of gas thus absorbed is too great, the metal becomes brittle and unworkable. The principal difficulty in making ductile titanium is prevention of such contamination, and the success of a given process may be judged by the completeness with which it solves this difficulty in a practical manner. Of the gases mentioned, hydrogen is the least objectionable, since its absorption by the metal is completely reversible and, although dimensional changes occur and absorption of hydrogen embrittles the metal, the gas may be removed readily by heating in a high vacuum for a short time. On the other hand, the absorption of oxygen or nitrogen results in the formation of permanent solid solutions of the oxide and nitride, which cannot be decomposed by the strongest reducing agentsz with sufticient completeness to restore the inherent ductility of the metal. Because of

Jan 1, 1946

By F. R. Hensel

PuRe silver and silver-lead alloys have been studied as to their suitability for bearing~.' A review of the properties of thallium and the silver-thallium constitutional diagram was made by the author to analyze the possibilities of silver-thallium compositions for antifriction materia1s. t The silver-rich end of the diagram as reported in the literatureg-" was found to be sketchy and it was necessary to carry out corisiderable experimental work to arrive at definite conclusions. Some of the results of this work are reported in this paper. Test Materials In the beginning of the work, only fused alloys were investigated. Later, research Table i.—Composition of Fused Silver-thallium Alloys Tested for Antifriction Properties Percentage Alloy No. of Thallium 1533...................... 0.48 1534.................... 1.06 a31 HT................... 2.04 231....................... 2.10 232 HT................... 3.83 Balance 232....................... 4.07 Silver 233 HT................... 5.86 a33....................... 6.38 a34....................... 9.84 a34 HT................... 10.07 was carried out on electroplating methods and diffusion processes. In preparing the fused alloyst care was taken to exhaust the toxic fumes caused by the thaillium content. The compositions of the series of fused alloys are listed in Table I. The silver-thallium alloys were melted in clay-graphite crucibles and cast into preheated steel molds of 3/4-in, diameter. In the cast condition, they showed a cored structure, as indicated by the micrographs of Figs. I to 3. The etching reagent used was a mixture of 2 grams K2Cr207, 8 C.C. H2S04, and IOO C.C. H20. The cored structure is unsatisfactory for the type of corrosion resistance required for bearing applications, and homogenizing experiments were carried out at various temperatures, the results of which are shown in Figs. 4 through 8. The alloys with a lower thallium concentration, which were heated for 2 hr. at 525°C show an almost completely homogenized solid solution type of structure. With higher thallium concentration, the homogenizing temperature was dropped to 475OC. to eliminate the formation of a liquid phase. It is evident that heating for 2 hr. at this temperature did not iesiilt iii complete elimination of the cored structure. As would be expected, cold-working of the cast structure is an expedient in establishing equilibrium conditions. The fine homogeneous grain structure shown in Fig. 7 corresponds to the cold-worked area under the Rockwell ball penetration when the hardness was taken before heat-treatment. A detailed structural study was made on a cast silver-thallium alloy containing 3.67 per cent thallium. Three micrographs (Figs. 9, 10 and 11) show the transition

Jan 1, 1946

By F. R. Hensel

PuRe silver and silver-lead alloys have been studied as to their suitability for bearing~.' A review of the properties of thallium and the silver-thallium constitutional diagram was made by the author to analyze the possibilities of silver-thallium compositions for antifriction materia1s. t The silver-rich end of the diagram as reported in the literatureg-" was found to be sketchy and it was necessary to carry out corisiderable experimental work to arrive at definite conclusions. Some of the results of this work are reported in this paper. Test Materials In the beginning of the work, only fused alloys were investigated. Later, research Table i.—Composition of Fused Silver-thallium Alloys Tested for Antifriction Properties Percentage Alloy No. of Thallium 1533...................... 0.48 1534.................... 1.06 a31 HT................... 2.04 231....................... 2.10 232 HT................... 3.83 Balance 232....................... 4.07 Silver 233 HT................... 5.86 a33....................... 6.38 a34....................... 9.84 a34 HT................... 10.07 was carried out on electroplating methods and diffusion processes. In preparing the fused alloyst care was taken to exhaust the toxic fumes caused by the thaillium content. The compositions of the series of fused alloys are listed in Table I. The silver-thallium alloys were melted in clay-graphite crucibles and cast into preheated steel molds of 3/4-in, diameter. In the cast condition, they showed a cored structure, as indicated by the micrographs of Figs. I to 3. The etching reagent used was a mixture of 2 grams K2Cr207, 8 C.C. H2S04, and IOO C.C. H20. The cored structure is unsatisfactory for the type of corrosion resistance required for bearing applications, and homogenizing experiments were carried out at various temperatures, the results of which are shown in Figs. 4 through 8. The alloys with a lower thallium concentration, which were heated for 2 hr. at 525°C show an almost completely homogenized solid solution type of structure. With higher thallium concentration, the homogenizing temperature was dropped to 475OC. to eliminate the formation of a liquid phase. It is evident that heating for 2 hr. at this temperature did not iesiilt iii complete elimination of the cored structure. As would be expected, cold-working of the cast structure is an expedient in establishing equilibrium conditions. The fine homogeneous grain structure shown in Fig. 7 corresponds to the cold-worked area under the Rockwell ball penetration when the hardness was taken before heat-treatment. A detailed structural study was made on a cast silver-thallium alloy containing 3.67 per cent thallium. Three micrographs (Figs. 9, 10 and 11) show the transition

Jan 1, 1946

By J. R. Long, E. T. Hayes, D. C. Root, R. S. Dean

The U. S. Bureau of Mines has recently reported on the development of a process for preparing pure ductile titanium in substantial quantities1 and on the physical properties that may be attained in the metal, when consolidated by powder metallurgy methods and fabricated by certain hot and cold-working procedures.2 The high physical properties (Table I), combined with excellent corrosion resistance and light weight, will lead no doubt to appropriatc applications of the metal. In any application, methods of joining will be of considerable importance. The relatively high activity of the metal toward the common gases raises the question of weldability and conditions that might be required for successful welding. Since hydrogen, oxygen, and nitrogen embrittle the metal, it should be protected from these gases at elevated temperatures. This was demonstrated by a few preliminary trials of arc, gas, and resistance welding. In these trials only the resistance method proved to be immediately- applicable. While atomic: hydrogen welding of titanium may be successful in some instances, a considerable amount of development work will be required to prove its worth. In addition, it would be suited only to small assemblies that could be heated subsequently in a high vacuum to remove embrittling hydrogen. Heli-are methods likewise would be limited to small assemblies, which could be completely protected by a helium atmosphere. It was quite apparent that titanium required more protection than was afforded by the normal flow of helium gas used for welding magnesium and aluminum. The hot metal is attacked by oxygen and nitrogen, and the rapid diffusion of the oxide into the metal prevents the formation of a protective coating and requires the helium atmosphere to be maintained until the weld has cooled nearly to room temperature. The preliminary work indicated that titanium could be successfully spot-welded without deterioration of the metal and that the microstructures of the welds were normal, exhibiting good bonding, and no oxidation. The high resistivity of titanium (53 microhms per cu. cm.) suggested that it would behave somewhat like stainless steel and allow considerable latitude in the several welding variables. The results reported here are intended to give a general picture of the effects of varying welding current, welding time, and tip pressure on single spot welds made on several different thicknesses of titanium sheet. The work was undertaken to demonstrate the feasibility of welding titanium rather than as an exhaustive study of spot welding. There are, therefore, a number of factors that will require further investigation to establish the specific controls required for actual applications.

Jan 1, 1947

By J. R. Long, R. S. Dean, D. C. Root, E. T. Hayes

The U. S. Bureau of Mines has recently reported on the development of a process for preparing pure ductile titanium in substantial quantities1 and on the physical properties that may be attained in the metal, when consolidated by powder metallurgy methods and fabricated by certain hot and cold-working procedures.2 The high physical properties (Table I), combined with excellent corrosion resistance and light weight, will lead no doubt to appropriatc applications of the metal. In any application, methods of joining will be of considerable importance. The relatively high activity of the metal toward the common gases raises the question of weldability and conditions that might be required for successful welding. Since hydrogen, oxygen, and nitrogen embrittle the metal, it should be protected from these gases at elevated temperatures. This was demonstrated by a few preliminary trials of arc, gas, and resistance welding. In these trials only the resistance method proved to be immediately- applicable. While atomic: hydrogen welding of titanium may be successful in some instances, a considerable amount of development work will be required to prove its worth. In addition, it would be suited only to small assemblies that could be heated subsequently in a high vacuum to remove embrittling hydrogen. Heli-are methods likewise would be limited to small assemblies, which could be completely protected by a helium atmosphere. It was quite apparent that titanium required more protection than was afforded by the normal flow of helium gas used for welding magnesium and aluminum. The hot metal is attacked by oxygen and nitrogen, and the rapid diffusion of the oxide into the metal prevents the formation of a protective coating and requires the helium atmosphere to be maintained until the weld has cooled nearly to room temperature. The preliminary work indicated that titanium could be successfully spot-welded without deterioration of the metal and that the microstructures of the welds were normal, exhibiting good bonding, and no oxidation. The high resistivity of titanium (53 microhms per cu. cm.) suggested that it would behave somewhat like stainless steel and allow considerable latitude in the several welding variables. The results reported here are intended to give a general picture of the effects of varying welding current, welding time, and tip pressure on single spot welds made on several different thicknesses of titanium sheet. The work was undertaken to demonstrate the feasibility of welding titanium rather than as an exhaustive study of spot welding. There are, therefore, a number of factors that will require further investigation to establish the specific controls required for actual applications.

Jan 1, 1947

By Otto H. Henry, Edward L. Badwick

Up to the present time, according to Hansenl and Haughton,² the constitution of the bismuth-indium system has not yet been published. The generally accepted value for the melting point of indium, as listed in most scientific journals and textbooks, is 155°C. Recently, however, H. M. Davis3 has determined a new melting point for indium of 157.3²C O.l°C. Since bismuth has been used in preponderance as an alloying element in the majority of low-melting alloys, and indium has been used with considerable success in lowering the melting points of low-melting alloys, 4,6 an interest has been developed in the application of these two metals, making a complete knowledge of the constitution of its alloys desirable. In this investigation, the melting point of indium reported by H. M. Davis has been found to be substantially correct. A value of 157. ²°C ± O.5°C was determined. The two eutectic compositions have been placed at 34 and 50 pct bismuth with their respective eutectic temperatures as 72.4°C ± 0.5°C and 90°C ± 0.5°C. The bismuth-indium phase diagram has been determined by thermal and microscopic analysis and is shown in Fig I. Thermal Analysis Granulated bismuth and indium shot, having a guaranteed purity of 99.99 and 99.95 pct, respectively, were melted in a carbon crucible by means of a laboratory heater. To minimize oxidation of the alloys, a heavy mineral oil (Nujol) having a high- flash point was used. A platinum-platinum rhodium thermocouple of fine wire encased in a thin-walled Pyrex glass tube, about 1/8 in. in diameter, was connected to a type K potentiometer for temperature measurements. The thermocouple was calibrated against the melting points of tin, bismuth, and the silver-copper eutectic and the boiling point of water. Since the temperature readings could be reproduced to 0.²°Cl and the accuracy of the temperature standards in this range is slightly better than 0.5°C, it scems reasonable to assume that the temperature readings are reliable to approximately 0.5°C. The cooling rate could be controlled by varying the amount of insulation surrounding the crucible. Three or four layers of asbestos paper, 1/16 in. thick, sufficed to give a cooling rate of from I° to ²°C per minute. Undercooling was slightly noticeable. This condition was minimized by vigorous stirring, since the advantage of using slower rates was too small to justify the greater expenditure

Jan 1, 1947

By Otto H. Henry, Edward L. Badwick

Up to the present time, according to Hansenl and Haughton,² the constitution of the bismuth-indium system has not yet been published. The generally accepted value for the melting point of indium, as listed in most scientific journals and textbooks, is 155°C. Recently, however, H. M. Davis3 has determined a new melting point for indium of 157.3²C O.l°C. Since bismuth has been used in preponderance as an alloying element in the majority of low-melting alloys, and indium has been used with considerable success in lowering the melting points of low-melting alloys, 4,6 an interest has been developed in the application of these two metals, making a complete knowledge of the constitution of its alloys desirable. In this investigation, the melting point of indium reported by H. M. Davis has been found to be substantially correct. A value of 157. ²°C ± O.5°C was determined. The two eutectic compositions have been placed at 34 and 50 pct bismuth with their respective eutectic temperatures as 72.4°C ± 0.5°C and 90°C ± 0.5°C. The bismuth-indium phase diagram has been determined by thermal and microscopic analysis and is shown in Fig I. Thermal Analysis Granulated bismuth and indium shot, having a guaranteed purity of 99.99 and 99.95 pct, respectively, were melted in a carbon crucible by means of a laboratory heater. To minimize oxidation of the alloys, a heavy mineral oil (Nujol) having a high- flash point was used. A platinum-platinum rhodium thermocouple of fine wire encased in a thin-walled Pyrex glass tube, about 1/8 in. in diameter, was connected to a type K potentiometer for temperature measurements. The thermocouple was calibrated against the melting points of tin, bismuth, and the silver-copper eutectic and the boiling point of water. Since the temperature readings could be reproduced to 0.²°Cl and the accuracy of the temperature standards in this range is slightly better than 0.5°C, it scems reasonable to assume that the temperature readings are reliable to approximately 0.5°C. The cooling rate could be controlled by varying the amount of insulation surrounding the crucible. Three or four layers of asbestos paper, 1/16 in. thick, sufficed to give a cooling rate of from I° to ²°C per minute. Undercooling was slightly noticeable. This condition was minimized by vigorous stirring, since the advantage of using slower rates was too small to justify the greater expenditure

Jan 1, 1947

By John L. Ham, Robert M. Parke

The melting point of molybdenum is 2625° + 50°C. Heretofore the metal has been considered too refractory to be melted in commercial quantities; hence, it has been formed into rod, wire, and sheet by the methods of powder metallurgy, wherein the temperatures do not exceed 90 per cent of the melting temperature. The manipulation of liquid metals at temperatures of about 2000°C. and above presents some difficult problems in heat transfer, in purification, and in container materials. For metals melting at such high temperatures, it is generally regarded as expedient to circumvent the more common melting and casting process by resorting to powder metallurgy. This circumvention is not accomplished without some penalties, the more notable of which are the relatively high cost of the powder metallurgy method and particularly the size limitations of its product. In an effort to overcome the inherent disadvantages of the powder metallurgy method, especially in respect to the limit of size, a project was begun in 1943 at the Climax Molybdenum Co. laboratory to develop a method of melting and casting molybdenum capable of producing larger bodies of molybdenum than were then obtainable. Later, from Feb. I, 1944, to Aug. 31, 1945, the project was under the supervision of the National Research Committee, Division One. In peacetime, too, there are needs for molybdenurn, which make consideration of another process desirable; for example, large parts for heat engines generally, the omnipresent demand for metal of higher quality and lower cost, and the development of molybdenum-base alloys. This paper describes methods for melting and casting molybdenum, with details of the special process needed to make molybdenum castings ductile, and reports some of the properties of what may be regarded as a new, or at least modified, form of the metal. First Process for Melting and Casting Molybdenum The maintenance of a temperature greater than 2625°C. in a 'pace large enough to melt several pounds of molybdenum may involve the loss of heat at a Very considerable rate. For melting molybdenum, therefore, the source of heat must be of high potential and of fairly high power. Evidently, also, the source of heat must not impair the purity of the product. The electric arc operating in vacuum appeared to meet these requirements and was accordingly the first choice as a source of energy. A number of arc devices for melting molybdenum were examined, and several were tried. In the first experiment, a 20-kw., 60-cycle-per-second alternating-current arc was operated between two horizontal and opposed electrodes of molybdenum- In this arrangement the falling droplets of molybdenum solidified below the arc in an irregular mass. This experiment did not produce a useful casting, but the data obtained indicated that

Jan 1, 1947

By John L. Ham, Robert M. Parke

The melting point of molybdenum is 2625° + 50°C. Heretofore the metal has been considered too refractory to be melted in commercial quantities; hence, it has been formed into rod, wire, and sheet by the methods of powder metallurgy, wherein the temperatures do not exceed 90 per cent of the melting temperature. The manipulation of liquid metals at temperatures of about 2000°C. and above presents some difficult problems in heat transfer, in purification, and in container materials. For metals melting at such high temperatures, it is generally regarded as expedient to circumvent the more common melting and casting process by resorting to powder metallurgy. This circumvention is not accomplished without some penalties, the more notable of which are the relatively high cost of the powder metallurgy method and particularly the size limitations of its product. In an effort to overcome the inherent disadvantages of the powder metallurgy method, especially in respect to the limit of size, a project was begun in 1943 at the Climax Molybdenum Co. laboratory to develop a method of melting and casting molybdenum capable of producing larger bodies of molybdenum than were then obtainable. Later, from Feb. I, 1944, to Aug. 31, 1945, the project was under the supervision of the National Research Committee, Division One. In peacetime, too, there are needs for molybdenurn, which make consideration of another process desirable; for example, large parts for heat engines generally, the omnipresent demand for metal of higher quality and lower cost, and the development of molybdenum-base alloys. This paper describes methods for melting and casting molybdenum, with details of the special process needed to make molybdenum castings ductile, and reports some of the properties of what may be regarded as a new, or at least modified, form of the metal. First Process for Melting and Casting Molybdenum The maintenance of a temperature greater than 2625°C. in a 'pace large enough to melt several pounds of molybdenum may involve the loss of heat at a Very considerable rate. For melting molybdenum, therefore, the source of heat must be of high potential and of fairly high power. Evidently, also, the source of heat must not impair the purity of the product. The electric arc operating in vacuum appeared to meet these requirements and was accordingly the first choice as a source of energy. A number of arc devices for melting molybdenum were examined, and several were tried. In the first experiment, a 20-kw., 60-cycle-per-second alternating-current arc was operated between two horizontal and opposed electrodes of molybdenum- In this arrangement the falling droplets of molybdenum solidified below the arc in an irregular mass. This experiment did not produce a useful casting, but the data obtained indicated that

Jan 1, 1947

By E. A. Nesbitt

The important permanent-magnet alloys 15 years ago contained carbon and depended upon it for their permanent-magnet properties. In recent years great advances have been made in a number of new alloys substantially free of carbon. These include Remalloy,' an alloy of iron, cobalt, and molybdenum, the various iron-nickel-aluminum alloys of the Mishima2 type, and the more recent Alnico V3 in which the best properties are developed by heat-treating in a magnetic field. The Mishima and the Alnico types of alloys are very brittle and after casting must be ground to size. This article describes a permanent-magnet material4 that can be machined as cast, and when handled with some care can be rolled to thin sheet and drawn to fine wire. Optimum magnetic properties are developed by a simple heat-treatment, and at this time the ductility of the alloy disappears. If the cast alloy is heat-treated without cold-working the mag netic energy product is about 1.0 X 106. This material is called Vicalloy I (pronounced vicalloy, V for vanadium, I for iron, and C for cobalt). When the alloy is first hot-worked and then considerably reduced in area by cold-swaging, drawing or grooved rolling, it is found that after heat-treatment the magnetic energy product measured in the direction of previous extension can be increased to 2.0 to 3.5 X 106. These higher values have been obtained only after reductions in area of 75 to 95 per cent, and since the material hardens considerably during working, the reduction ordinarily can be accomplished only in fine wires, although larger sizes can be reduced by the use of powerful machinery. When the material has been cold-worked in such a manner as to improve its energy product substantially, it is called Vicalloy 11. Commercial use of Vicalloy has been made in the recording of speech. Tape 0.002 by 0.050 in. has been found superior in quality to all others tried, and is now used in the Western Electric Mirrorphone 5 In a variety of forms Vicalloy I or I1 is useful in many laboratory instruments, and in experiments in which the sizes and shapes required cannot be obtained with the other more brittle permanent-magnet materials. In 1936, during a study of other age-hardening magnetic systems, it occurred to the author that some of the iron-cobalt-vanadium alloys might be suitable for permanent magnets. The proportions of iron, cobalt, and vanadium were varied and a range of alloys of particular interest was found, as shown by the indicated area in Fig. I. The limits of this range of composition are 30 to 52 per cent iron, 36 to 62 per cent cobalt, and 4 to 16 per cent vanadium. The compositions indicated by points outside this area were also tested. Preparation and Properties OF Vicalloy I The alloys were usually cast in graphite molds into bars 3/4 in. in diameter and

Jan 1, 1946

By E. A. Nesbitt

The important permanent-magnet alloys 15 years ago contained carbon and depended upon it for their permanent-magnet properties. In recent years great advances have been made in a number of new alloys substantially free of carbon. These include Remalloy,' an alloy of iron, cobalt, and molybdenum, the various iron-nickel-aluminum alloys of the Mishima2 type, and the more recent Alnico V3 in which the best properties are developed by heat-treating in a magnetic field. The Mishima and the Alnico types of alloys are very brittle and after casting must be ground to size. This article describes a permanent-magnet material4 that can be machined as cast, and when handled with some care can be rolled to thin sheet and drawn to fine wire. Optimum magnetic properties are developed by a simple heat-treatment, and at this time the ductility of the alloy disappears. If the cast alloy is heat-treated without cold-working the mag netic energy product is about 1.0 X 106. This material is called Vicalloy I (pronounced vicalloy, V for vanadium, I for iron, and C for cobalt). When the alloy is first hot-worked and then considerably reduced in area by cold-swaging, drawing or grooved rolling, it is found that after heat-treatment the magnetic energy product measured in the direction of previous extension can be increased to 2.0 to 3.5 X 106. These higher values have been obtained only after reductions in area of 75 to 95 per cent, and since the material hardens considerably during working, the reduction ordinarily can be accomplished only in fine wires, although larger sizes can be reduced by the use of powerful machinery. When the material has been cold-worked in such a manner as to improve its energy product substantially, it is called Vicalloy 11. Commercial use of Vicalloy has been made in the recording of speech. Tape 0.002 by 0.050 in. has been found superior in quality to all others tried, and is now used in the Western Electric Mirrorphone 5 In a variety of forms Vicalloy I or I1 is useful in many laboratory instruments, and in experiments in which the sizes and shapes required cannot be obtained with the other more brittle permanent-magnet materials. In 1936, during a study of other age-hardening magnetic systems, it occurred to the author that some of the iron-cobalt-vanadium alloys might be suitable for permanent magnets. The proportions of iron, cobalt, and vanadium were varied and a range of alloys of particular interest was found, as shown by the indicated area in Fig. I. The limits of this range of composition are 30 to 52 per cent iron, 36 to 62 per cent cobalt, and 4 to 16 per cent vanadium. The compositions indicated by points outside this area were also tested. Preparation and Properties OF Vicalloy I The alloys were usually cast in graphite molds into bars 3/4 in. in diameter and

Jan 1, 1946

By G. Derge, Lo Ching Chang

The chemical and physical propertties of slag systems are of special interest to metallurgists, for nearly all metals are in contact with molten slags during the primary reduction from their ores and then during the subsequent refining and alloying processes that determine their ultimate use. These processes are carried out at high temperatures and involve such complex chemical systems that funda-mental scientific knowledge of the Properties and behavior of these slags is inadequate. Various hypotheses have been Proposed to explain known slag characteristics. The commonest of these regard the molten slag as constituted of compounds that are dissociated in varying degrees into their constituent oxides, as illustrated by Schenckl and the recent Papers of Chipman and co-authors. Another proposal is that these compounds dissociate by com-plete6 or partial ionization, the evidence for which has been reviewed by Martin and Derge.? whatever hypothesis is favored, the exact formulation of these compounds is incomplete. It is based on three principal types of information: 1. Studies of solidified 51% samples by X ray diffraction, petrography, and similar methods that can be used at room temperature. The correlation of this information with the high-temperature properties of the molten slag is uncertain. 2. Studies of slag-metal equilibria, which have the advantage of providing observations on the molten slag at high temperature. However, even these data present problems of interpretation because they require a complete knowledge of the metal phase. 3. Direct measurements of properties of molten slag at the temperature of greatest metallurgical interest. Viscosity, electrical conductivity, specific heats, and similar properties have been measured, but the data of this type are widely scattered and frequently difficult to interpret. Additional direct measurements of properties of molten slag that can be interpreted in terms of chemical behavior are required, and this paper shows that the standard physical-chemical measurement of reversible electrode potentials, which measure the activity of specific constituents, is a fruitful source of such information. The purpose of this investigation, therefore, was to find an electrode combination that would yield reversible electromotive force measurements when dipped into the molten slag to be studied. Slags of the systems cao-sio2 and cao-120,-SiO2 were chosen for study because graphite containers could be used and the data could be correlated with those for viscosity, conductivity, and other proper-

Jan 1, 1948

By G. Derge

The chemical and physical propertties of slag systems are of special interest to metallurgists, for nearly all metals are in contact with molten slags during the primary reduction from their ores and then during the subsequent refining and alloying processes that determine their ultimate use. These processes are carried out at high temperatures and involve such complex chemical systems that funda-mental scientific knowledge of the Properties and behavior of these slags is inadequate. Various hypotheses have been Proposed to explain known slag characteristics. The commonest of these regard the molten slag as constituted of compounds that are dissociated in varying degrees into their constituent oxides, as illustrated by Schenckl and the recent Papers of Chipman and co-authors. Another proposal is that these compounds dissociate by com-plete6 or partial ionization, the evidence for which has been reviewed by Martin and Derge.? whatever hypothesis is favored, the exact formulation of these compounds is incomplete. It is based on three principal types of information: 1. Studies of solidified 51% samples by X ray diffraction, petrography, and similar methods that can be used at room temperature. The correlation of this information with the high-temperature properties of the molten slag is uncertain. 2. Studies of slag-metal equilibria, which have the advantage of providing observations on the molten slag at high temperature. However, even these data present problems of interpretation because they require a complete knowledge of the metal phase. 3. Direct measurements of properties of molten slag at the temperature of greatest metallurgical interest. Viscosity, electrical conductivity, specific heats, and similar properties have been measured, but the data of this type are widely scattered and frequently difficult to interpret. Additional direct measurements of properties of molten slag that can be interpreted in terms of chemical behavior are required, and this paper shows that the standard physical-chemical measurement of reversible electrode potentials, which measure the activity of specific constituents, is a fruitful source of such information. The purpose of this investigation, therefore, was to find an electrode combination that would yield reversible electromotive force measurements when dipped into the molten slag to be studied. Slags of the systems cao-sio2 and cao-120,-SiO2 were chosen for study because graphite containers could be used and the data could be correlated with those for viscosity, conductivity, and other proper-

Jan 1, 1948

By R. B. Snow

This paper describes a technique of polishing and etching specimens of open-hearth furnace slags or hearth aggregates for identification of the crystalline constituents —lime (CaO), tricalcium silicate (3CaO.SiO2), dicalcium silicate (2CaO.-SiO2), rnonticellite (CaO.MgO.SiO2), or forsterile (2MgO.SiO2), with especial em-phasis on the mineral merwinite (3CaO.-MgO.2SiO2). With proper standardization, this identification does not require the use of the petrographic microscope. The composition of basic open-hearth slags and furnace bottoms falls, almost without exception, within systems containing CaO, MgO, 'IFeO,,, MnO and SiOz, in which the number of basic molecules so greatly exceeds the orthosilicate ratio (two molecules of base to one of silica) that free basic oxides, and combinations between them such as alumi-nates or ferrites, are present in cooled specimens. Orthosilicates of (CaO + NgO) are the most common in such specimens, since in nearly all cases, except premelt slags, the molecular ratio of (CaO + MgO) to SiO, is more than 2 to I. When sufficient lime is available it combines with the silica to form dicalcium silicate (2Ca0.Si02), which contains little, if any, IvfgO, FeO or MnO in solid solution whereas the latter oxides combine to form the oxide solid solution known as periclase. If the lime present is insufficient to form dicalcium silicate (2Ca0.Si02) it combines with Mgo to form either merwillite or moIlticellite (SiOz); these minerals take little if any FeO or MnO into solid solution and the remaining MgO, FeO and hInO combine as periclase. This generalization seems to be valid for basic slags and furnace bottoms, since minerals such as Ca0.MnOSi02 and CaO.FeO.SiOz are found only in slags in which the lime-silica ratio is less than 2 and are not observed in 'pecimens from furnace bottoms. The identification of crystalline constituents in such materials, especially of fine crystals in the groundmass, is difficult under the petrographic microscope. They are often masked by their neighbors because of their small size in relation to the thickness of the thin section and because of the presence of Opaque Or colored constituents. The indices of refraction and the optical sign of the mineral are sometimes difficult to determine because of the small size or because Of twinning or of inclusions within the crystal. Moreover, the positive identification of merwinite (3Ca0.Mg0.2Si02) from its optical properties is usually difficult in the presence of dicalcium silicate (zCaO.SiO2). CaO, MgO, jCa0.Si02 and 2Ca0.SiOz in open-hearth 'lags have been identified for a number of years in the U.S. Steel Corporation Laboratory by the usual

Jan 1, 1948

By R. B. Snow

This paper describes a technique of polishing and etching specimens of open-hearth furnace slags or hearth aggregates for identification of the crystalline constituents —lime (CaO), tricalcium silicate (3CaO.SiO2), dicalcium silicate (2CaO.-SiO2), rnonticellite (CaO.MgO.SiO2), or forsterile (2MgO.SiO2), with especial em-phasis on the mineral merwinite (3CaO.-MgO.2SiO2). With proper standardization, this identification does not require the use of the petrographic microscope. The composition of basic open-hearth slags and furnace bottoms falls, almost without exception, within systems containing CaO, MgO, 'IFeO,,, MnO and SiOz, in which the number of basic molecules so greatly exceeds the orthosilicate ratio (two molecules of base to one of silica) that free basic oxides, and combinations between them such as alumi-nates or ferrites, are present in cooled specimens. Orthosilicates of (CaO + NgO) are the most common in such specimens, since in nearly all cases, except premelt slags, the molecular ratio of (CaO + MgO) to SiO, is more than 2 to I. When sufficient lime is available it combines with the silica to form dicalcium silicate (2Ca0.Si02), which contains little, if any, IvfgO, FeO or MnO in solid solution whereas the latter oxides combine to form the oxide solid solution known as periclase. If the lime present is insufficient to form dicalcium silicate (2Ca0.Si02) it combines with Mgo to form either merwillite or moIlticellite (SiOz); these minerals take little if any FeO or MnO into solid solution and the remaining MgO, FeO and hInO combine as periclase. This generalization seems to be valid for basic slags and furnace bottoms, since minerals such as Ca0.MnOSi02 and CaO.FeO.SiOz are found only in slags in which the lime-silica ratio is less than 2 and are not observed in 'pecimens from furnace bottoms. The identification of crystalline constituents in such materials, especially of fine crystals in the groundmass, is difficult under the petrographic microscope. They are often masked by their neighbors because of their small size in relation to the thickness of the thin section and because of the presence of Opaque Or colored constituents. The indices of refraction and the optical sign of the mineral are sometimes difficult to determine because of the small size or because Of twinning or of inclusions within the crystal. Moreover, the positive identification of merwinite (3Ca0.Mg0.2Si02) from its optical properties is usually difficult in the presence of dicalcium silicate (zCaO.SiO2). CaO, MgO, jCa0.Si02 and 2Ca0.SiOz in open-hearth 'lags have been identified for a number of years in the U.S. Steel Corporation Laboratory by the usual

Jan 1, 1948

By J. R. Lewis

During the past few years there has been considerable interest in the sizing and the preparation of the iron ore fed into blast furnaces. Furnacemen know that proper sizing of ore tends to increase the efficiency of the furnace. It has been shown by Joseph1 and his associates that there is a direct relationship between the porosity of an ore particle and the time required for its reduction in the blast furnace. This relationship may be stated as follows: All other factors remainig constant, the time for reduction of a given sample of ore is inversely propor-tiofid to its porosity. As a consequence, large pieces of very porous ore, say 3 or 4 in. in cross section, will be completely reduced by the time they reach the hot zone just above the tuyeres, while large pieces of dense, nonporous ore may reach the hot zone with their centers unreduced. This means that for the latter complete reduction will be accomplished by the so-called "direct method;" i,e., where carbon reacts directly with the ore. Some recent work by philbrook2 seems to contradict this rule. He found that of two unlike ores he studied, the one with the lower porosity reduced more readily. He explained his results as "possibly due to an apparently greater tendency of the rA, ore (lower porosity) to crack during preliminary dehydration and during reduction." This brings his results in line with the rule stated above. It is not the author's intention to enter into a discussion of the merits of "direct" versus "indirect" reduction. It is enough to Say that, in general, indirect reduction is desirable. For the greatest efficiency, Gruner's theorem3 requires that all reduction in the blast furnace be indirect. Joseph seems to believe this also, because in his recent article on Some Comparisons between Iron Ore, Sinter and Nodules as Blast Furnace Feed4 he concludes that: "In genera', the carbon in blastfurnace coke is used most efficiently when the largest amount of the carbon charged reaches the tuyeres." However, P. V. hlartin,3 in an excellent paper, has shown that some direct reduction may actually increase the efficiency of the furnace; that is, increase its rate of production while maintaining maximum fuel economy. For nonporous ores there is more opportunity to have an excess of direct reduction unless the factors that make for indirect reduction are favored; for instance; the Proper sizing of the Ore particles' Another factor that has an important bearing on the reduction rate, as it is affected by particle size, is the flow of heat into the particle. The ore particle

Jan 1, 1948

By J. R. Lewis

During the past few years there has been considerable interest in the sizing and the preparation of the iron ore fed into blast furnaces. Furnacemen know that proper sizing of ore tends to increase the efficiency of the furnace. It has been shown by Joseph1 and his associates that there is a direct relationship between the porosity of an ore particle and the time required for its reduction in the blast furnace. This relationship may be stated as follows: All other factors remainig constant, the time for reduction of a given sample of ore is inversely propor-tiofid to its porosity. As a consequence, large pieces of very porous ore, say 3 or 4 in. in cross section, will be completely reduced by the time they reach the hot zone just above the tuyeres, while large pieces of dense, nonporous ore may reach the hot zone with their centers unreduced. This means that for the latter complete reduction will be accomplished by the so-called "direct method;" i,e., where carbon reacts directly with the ore. Some recent work by philbrook2 seems to contradict this rule. He found that of two unlike ores he studied, the one with the lower porosity reduced more readily. He explained his results as "possibly due to an apparently greater tendency of the rA, ore (lower porosity) to crack during preliminary dehydration and during reduction." This brings his results in line with the rule stated above. It is not the author's intention to enter into a discussion of the merits of "direct" versus "indirect" reduction. It is enough to Say that, in general, indirect reduction is desirable. For the greatest efficiency, Gruner's theorem3 requires that all reduction in the blast furnace be indirect. Joseph seems to believe this also, because in his recent article on Some Comparisons between Iron Ore, Sinter and Nodules as Blast Furnace Feed4 he concludes that: "In genera', the carbon in blastfurnace coke is used most efficiently when the largest amount of the carbon charged reaches the tuyeres." However, P. V. hlartin,3 in an excellent paper, has shown that some direct reduction may actually increase the efficiency of the furnace; that is, increase its rate of production while maintaining maximum fuel economy. For nonporous ores there is more opportunity to have an excess of direct reduction unless the factors that make for indirect reduction are favored; for instance; the Proper sizing of the Ore particles' Another factor that has an important bearing on the reduction rate, as it is affected by particle size, is the flow of heat into the particle. The ore particle

Jan 1, 1948

By A. H. Geisler, F. Keller

Some of the important changes that take lace in the structure of aluminum alloys are largely submicroscopic in character. This is especially true of the changes that accompany age-hardening and recrys-tallization. Although improved metal-lographic practices have been helpful in indicating some of these changes indirectly, results are limited by the resolving power of optical microscope lenses. The electron microscope, however, provides a new and important means for investigating the fine structure of metals in a range not possible heretofore. It is anticipated that this microscope will yield new and useful information to metallurgists when suitable techniques are developed to utilize its very high resolving power. Several unique methods have been devised to permit the examination of the structure of opaque samples by the electron microscope. These are all based 011 the principle of producing a very thin film that represents the prepared surface of the opaque metal sample, which can be examined in the transmission microscope; however, the available methods fall into three categories, depending on the manner in which the film is obtained. With the oxide-film method developed by Mah13-l7 a very thin surface layer of the metallographic sample is converted into oxide by a thermal, chcmical or anodic treatment. .After removal from the metal sample by one of the special techniques described by Evans,l9 this oxide film is examined in the electron microscope, and is found to portray the structure of the original metal surface. Films formed on aluminum,8,l7 nicke1,7 iron3,8,9 and a nickel-beryllium alloyl1 by heating; films formed on iron8,9 by chemical action; and films formed on aluminum3-'" by anodic oxidation, have been examined. The other two are more recent methods developed principally in this country; they consist in making the thin film in the form of a mold or replica of the surface contour of an etched sample. With the negative replica method the thin film is made directly on the sample; with the positive replica method it is made on a negative reproduction of the sample. With both methods, variations in the contour of the metal sample, caused mainly by etching, produce variations in the thickness of the surface replica. Because of the direct dependence of the amount of scattering of the illuminating electrons upon the thickness of the film penetrated, only the variations in thickness of the replica are responsible for tones in the electron image. Negative replicas are made by forming a thin film of some suitable material directly on the polished and etched metal sample, and derive their name from the fact that after removal from the sample they have the high and low points in the sample reversed, as shown in Fig. I. Nega-

Jan 1, 1944

By F. Keller, A. H. Geisler

Some of the important changes that take lace in the structure of aluminum alloys are largely submicroscopic in character. This is especially true of the changes that accompany age-hardening and recrys-tallization. Although improved metal-lographic practices have been helpful in indicating some of these changes indirectly, results are limited by the resolving power of optical microscope lenses. The electron microscope, however, provides a new and important means for investigating the fine structure of metals in a range not possible heretofore. It is anticipated that this microscope will yield new and useful information to metallurgists when suitable techniques are developed to utilize its very high resolving power. Several unique methods have been devised to permit the examination of the structure of opaque samples by the electron microscope. These are all based 011 the principle of producing a very thin film that represents the prepared surface of the opaque metal sample, which can be examined in the transmission microscope; however, the available methods fall into three categories, depending on the manner in which the film is obtained. With the oxide-film method developed by Mah13-l7 a very thin surface layer of the metallographic sample is converted into oxide by a thermal, chcmical or anodic treatment. .After removal from the metal sample by one of the special techniques described by Evans,l9 this oxide film is examined in the electron microscope, and is found to portray the structure of the original metal surface. Films formed on aluminum,8,l7 nicke1,7 iron3,8,9 and a nickel-beryllium alloyl1 by heating; films formed on iron8,9 by chemical action; and films formed on aluminum3-'" by anodic oxidation, have been examined. The other two are more recent methods developed principally in this country; they consist in making the thin film in the form of a mold or replica of the surface contour of an etched sample. With the negative replica method the thin film is made directly on the sample; with the positive replica method it is made on a negative reproduction of the sample. With both methods, variations in the contour of the metal sample, caused mainly by etching, produce variations in the thickness of the surface replica. Because of the direct dependence of the amount of scattering of the illuminating electrons upon the thickness of the film penetrated, only the variations in thickness of the replica are responsible for tones in the electron image. Negative replicas are made by forming a thin film of some suitable material directly on the polished and etched metal sample, and derive their name from the fact that after removal from the sample they have the high and low points in the sample reversed, as shown in Fig. I. Nega-

Jan 1, 1944

By William A. Johnson

The mathematical analysis of diffusion curves in solid metals is carried out ordinarily by analogy with the flow of heat in a continuous medium and no account is taken of the fact that the materials involved are composed of discrete particles that move discrete distances at irregular intervals. It is possible to treat the problem on the basis of simple assumed atomic mechanisms, in which case, because of the small scale of the elementary process compared to the gross effects observed experimentally, the two procedures lead to the same expression for the variation of concentration with distance. Thus, for engineering purposes, the usual methods are quite satisfactory. If, on the other hand, our purpose is to go beyond the mere collection of data, however desirable this may be, it is necessary to recognize and understand the factors that are of fundamental importance. One factor that has not hitherto been considered when dealing with solid metals is the difference in atomic weights of the isotopes of a single metal. While all the atoms of a given element are alike chemically, their weights may vary by a small but significant amount, and atoms of different weights may have different diffusion rates. With gases this variation in isotopic weight produces a change in the rate of diffusion by a factor inversely proportional to the square root of mass. While, except for the lightest gases, this is not a large effect, it is, for example, sufficient to permit a considerable concentration of one isotope at the expense of others. If metals exhibit a similar phenomenon, some modification of the analysis of the diffusion process will be required, since the different isotopes of a given metal will have different probabilities of moving and the heat-flow analogy will contain a fundamental error. This situation cannot be rectified by using a diffusion coefficient based on the average probability of moving, and it can be shown that even when the diffusion coefficient is independent of composition, the diffusion curve will not agree with the simple theory except in the rare cases where only one isotope exists; e.g., gold. If, however, the mass effect in the solid state were no larger than in the gaseous, the errors arising therefrom in the analysis of ordinary diffusion curves would be quite negligible. Nevertheless, our desire to obtain all possible information concerning the details of the diffusion process, and the possibility of diffusion in the solid state as a means for separating isotopes are cogent reasons for undertaking an experimental study to determine whether isotopes of different masses diffuse at different rates. Experimental Procedure If isotopes of different masses diffuse at different rates, the ratio of the concentration of one relative to another will vary with the distance from the diffusion inter-

Jan 1, 1946