Search Documents

By R. E. Reed, C. J. McHargue

Poly crystalline aluminum specimens having random and diffuse (001) initial orientations were reduced 90 pct in area at room temperature by drawing, extruding, and swaging. Although each procedure produced the duplex (001)- (111) texture, there were marked dqferences in the relative amounts of each component. This also produced variations in the tensile properties of the aluminum rods. It is generally agreed that the fiber texture found in aluminum rods consists of a strong (111) and a weak (001) component after reductions in area of 90 pct or more. An examination of data summarized in recent reviews'-3 shows much spread in the values for the relative amounts of the two components reported by different investigators. That the initial orientation has a marked influence on the final texture even after reductions in area greater than 90 pct is well-documented.4-7 Another factor which may influence the relative amounts of the two components is the method of fabricating the wires or rods. Barrett and ~assalski' state that the center of wires formed by rolling, swaging, and drawing have the same textures. The data cited in support of this conclusion came from qualitative X-ray techniques. In recent years there has been renewed interest in correlating the type of preferred orientation with such parameters as stacking fault energy and onset of mechanical twinning.'' Since these analyses have often used many sources of data, it is essential that the data be evaluated for the effect of fabrication variables. In this study we have produced aluminum rods by drawing, swaging, and extruding and have quantitatively determined their textures. All specimens had the same purity and fabrication temperature, and the amount of deformation and the starting textures were determined. We have compared our results with published studies on similar materials and fabricating parameters. We have also compared results from the same three fabricating procedures for single-crystal specimens in three instances. EXPERIMENTAL PROCEDURE The aluminum was 99.99+ pct with the only detected impurities being (by weight) 0.002 pct Cu, 0.001 pct Si, and 0.001 pct Mg. Cylinders a in. in diam by 1 in. long with a fine-grain random texture were made by the following method. Cubes 1+ in. on a side were machined from 2-in.-diam ingots. These cubes were slowly pressed about 30 pct in each of the cube directions. The material was recrystallized by annealing 1 hr at 350°C. Three deformation-anneal cycles produced the desired material. Polycrystalline specimens with a diffuse (001) texture were made by casting into a graphite tube 1 in. ID, li in. OD, by 8 in. long which had been preheated to 900°C. One end of the tube was placed on a water-

Jan 1, 1969

By S. E. Bronisz, J. O. Brittain

Experiments have been carried out to study the effect of an applied stress on strain aging in steel. It was found that the stress-induced strain aging reaction proceeds rapidly in the temperature range of 0° to 60°C. The variation of stress-induced strain aging with plastic strain, aging time, and temperatures is reported. The activation energy for the stress-induced strain aging was found to be 10,000 cal per mol for the steel used in this investigation. VARIOUS observations1-5 have been made of the influence of an applied stress upon the yielding of iron and steel. However, there have been only cursory studies of the effect of an applied tensile stress upon strain aging in iron and steel1,2 The following represents the results of a preliminary study of strain aging as influenced by the application of an external tensile load during the aging process. For the most part, previous experimental work on strain aging can be classified into two parts. First, the experimental technique involved a direct determination of mechanical properties of a suitable specimen that has been plastically deformed to a constant strain and aged generally at a temperature of the order of 200°C, with the aging times varying from about 1/10 to 2 hr.6-8 The second classification of strain aging experiments involves an indirect determination of strain aging by means of physical property such as electrical resistivity or internal friction as a function of the aging time at a constant temperature and strain.9-12 The latter results have

Jan 1, 1961

By A. Mitchell, M. Etienne

VARIOUS'-3 reports on the electroslag remelting process refer to the "thin skin of slag" existing around the slag pool and the solidifying ingot, but none considers the various mechanisms by which this layer forms and thus influences the surface quality of the ingot. A vertical section4,5 of the Cao-CaF2-Al2O3 diagram is shown in Fig. 1 between CaF2 and 6Al2.CaO3 From this, it is seen that a typical "70:30" CaF2:A1a3 slag will solidify to produce 6Al2O3.CaO primary phase in a eutectic matrix. The commercial slags contain sufficient CaO to produce* this composition by both the

Jan 1, 1969

By K. P. Stricker, H. Keddeinis, K. Grebe

Recent investigations by Mannesmann Research have shown that disintegration of sinter with widely varying quality levels is subject to a much more complex set of influences than was previously assumed. It was confirmed that disintegration at temperatures between 300° and 600°C (570°and 1,110°F) is triggered by the degradation of hematite. The specific tendency of hematite to disintegrate is, however, not constant: sinters containing the same amount of hematite can show widely differing disintegration tendencies, depending on structural homogeneity, porosity, and gangue quantity. Sinters with heterogeneous structures transmit stress to the hematite phase and this favors disintegration. Another factor is the brittleness of the glassy silicate phase, which can be influenced even by small modifications of minor gangue constituents.

Jan 1, 1984

By John M. Rayburn

When plans were made for the new mine of the Allegheny-Pittsburgh Coal Co., Springdale, Pa., it was decided to operate on the room and pillar system, but not to have any extensive room development, the pillar work being done altogether on the retreat. On each room heading five rooms were to he driven advancing, while five pillars were being removed, so timed as to form a continuous pillar line, with all approaching

Jan 1, 1930

Discussion of the paper of WALTER 0. SNELLING, presented at the New York meeting, February, 1915, and printed in Bulletin No. 100, April, 1915, pp. 695 to 704. A. F. LUCAS, Washington, D. C.-Are the results of the experiments in practical operation? WALTER 0. SNELLING.-There have been no large-scale operations up to the present. The necessary conditions of temperature are easily arranged for, but the required condition of pressure, 800 lb., is not so simple a matter. The company has been working on this problem, and at the refinery in Oklahoma work of this kind is now being carried on. The way in which an ordinary still will leak at 50 lb., and more at 100 lb., leaves some doubt in the mind of many as to what can be clone at a continuous working pressure of 800 lb. On the other hand, the fact that they are making ammonia abroad by a method which employs higher pressures and temperatures than anything involved in this method, leads me to believe that the difficulties, which will undoubtedly arise when this is attempted on a commercial scale, can be overcome. WILLIAM N. BEST, New York, N. Y.-Is there any change in the calorific value of these oils?

Jan 5, 1915

Jan 1, 1926

By Charles A. Morgan

The proms of mire drawing depends upon the property which certain metals possess, termed ductility, which is defined in Brande's Dictionary of Science as a property inconhequcnce of which metals may be drawn out at length wilhotit interrupting the continuity of their constituent parts. As ticfincd thy M. Tresca, it is the property of flowing when subjected to pressure, and ie termed by him the "flow of solids," in his paper, read before the AcaiXemy of Sciences in Paris, November, 1866 * In the simple language of the wire shop, it is that quality in certain metals that thcm to be drawn (or etretched) out througl~ a tapering hole in a metal block or die, the die being more dense and harder than the metal to be redud. It is a fact well known to tllose slrillet1 in the ai*, that to eu~xeed in drawing iron or steel wire, it is of' the utmost, importance to have tborough lubritatiun, as a lack of it woullii udu-se abrasion of the wire, or of the walls of the tapering hole in the dia or plate, and quickly put an end to the operation. In drawing coarse or large sizes of pire, say & inoh diameter, the pressure upon the sides of the tay~dng hole, where the wire comes i.u contact with the die, so great that an ordina,y labricaot is equeezed out when the nsoal rc~luctions iu size are matie, and abrasion takea place. In order to meet this diffic:ulty, and emure proper lubrication, it has been a ccmmon prtictice for yazrs to apply a paste made of rye or wheaten flour, or lime, to the surface of the wire to be drawn, and when the pnstc is dry, to smwr tl~e wire with tallow or grease; the wire is then ready to be draw, the dried pa serving fo prevent the lubricant from being presetf out in the prows. Some years since, when Bessemer metal came iargely into use material for wire, 'the writer fourid that its dmwing requid so much more power than iron wire, that he was led to institute a. series of trials to ascertain the difference in power required. It was fotind that Bessemer wire requircd from 100 to 200 per cent. more power to makc the same reduction in sizes than soft iron wire, the difference depending npou the amount of carbon and other chemical constitu-

Jan 1, 1881

The Pennsylvania State College, Mineral Industries Experiment Station, State College, Pa For publications address the Director The Pennsylvania State College through the Minerals Research Experiment Station has issued 10 Bulletins and three Technical Papers, from which the following have been selected as of greatest interest to mining engineers and geologists: Bulletin 1, An investigation of the properties of smitliing coals (1922), free; 2, Petrography of the Mica Pendotite Dike at Dixonville (1926), free; 4, The Helderberg Group from cen¬tral Pennsylvania to southwestern Virginia (1929), 25 cents, 5, Bentonite in Pennsylvania (1929), 10 cents, 6, Production data on Appalachian oil fields (1930), free; 8, Oil field waters of Pennsyl¬vania (1930), free; Technical Paper 2, Condition of water in coals of various ranks.

Jan 1, 1933

By Nicholas J. Grant, Hanna Ibach, Bill C. Giessen

The system Ta-Rh was investigated over the entire comnposition range using metallogvaphic and X-ray techniques as well as thermal analysis. Terminal solubility limits, solidus temperatures, and the crystal structures of two new intermediate phases were determined. There are five intermediate phases: o, tetragonal, isostructural with o(Fe-Cr); a1, orthorhomnbic : a2, orthorhombic, isostructural with CozSi; 0 3, a high -temperature phase of unknozun structure; and a TaRh3, cubic, AuCu3 structure. a, aL, a,, and a melt peritectically; and a TaRh, has a maximum melting point. Five peritectic and one eutectic reactions occur. THE Ta-Rh phase diagram has not been fully treated in the literature; however, two inter metallic phases have been described.&apos;-&apos; This study attempts to work out the diagram covering the details of phase boundaries as well as the crystallography of the intermediate phases. EXPERIMENTAL METHODS Starting materials were tantalum powder of 99.8 pct purity, supplied by National Research Corp., Cambridge, Mass., and rhodium powder of 99.9 pct purity, supplied by Bishop Co., Malvern, Pa. Analyses as given by the manufacturers are presented in Table I. In addition to the spectroscopic analysis, an oxygen analysis has been made of the tantalum powder. After preliminary trials, the standard procedure employed in making the alloys was as follows (see Ref. 4 for further details). The powders were weighed, mixed manually, compacted under 16,000 psi, and arc-melted at not more than 500 amp under thoroughly titanium gettered welding-grade argon of 400 mm Hg pressure. The gettering process was repeated after each melt. It had been shown previously5 that under these conditions no oxygen pick-up exceeding 0.001 pct in a hafnium button could be detected. The buttons, ranging from 5 to 10 g in weight, were inverted or, if possible, crushed and remelted three times. Melting presented no difficulty, except for tantalum-rich al- loys, which showed considerable spashing during melting. The average weight loss of 1 wt pct rose to 2.5 wt pct in these cases. Subsequent melting, however, resulted in losses of less than 0.1 wt pct. The melts obtained frequently showed small in-homogeneities, mostly in the tantalum-rich part of the system below 25 at. pct Rh. This was not due to insufficient blending of the elements, but to crystallization taking place from the cooled bottom. This condition was not improved by further re-melting; therefore the following precautions were taken. Cross sections of as-cast buttons were made. For the investigation only buttons were used where these inhomogeneities did not exceed 1 at. pct (estimated from the amounts of haSeS resent). In addition, only the center portions of such buttdns were utilized to minimize errors. Smaller buttons showed smaller concentration gradients and were used as a check. Finally, all alloys were homogenized to eliminate coring (see below). After this, the error in alloy composition due to inhomogeneity was cl at. pct. In order to confirm the concentrations of the alloys and to provide an impurity check, wet-chemical analyses and oxygen-fusion analyses of four key alloys were carried out. The results, together with the concentrations as calculated with the weight balance method, are presented in Table 11. It was found that for three alloys the agreement is very good (< 0.2

Jan 1, 1964

By Morris Cohen

Within the past two years, a number of publications have called attention to the double peaks, or stages, that appear in the hardness and strength curves of certain aging alloys. The author has shown the existence of double hardness peaks during the aging of a high-purity silver-copper alloy;&apos; and on the basis of electrical resistance, microscopic, dilation and lattice-parameter measurements, hc concluded that the initial hardness peak occurs prior to precipitation while the secondary hardening is due to precipitation. It was suggested that the general mechanism of age-hardening involves two consecutive processes, the first consisting of nucleus or knot2 formation, and the second consisting of actual precipitation and particle growth. It was further held that the two stages are interdependent, but because of nonuniform precipitation they may overlap each other to an extent that increases as the aging temperature is raised. At high aging temperatures, the effect of the first stage cannot be detected because of the rapid occurrernce of precipitation. The general theory of age-hardening recently presented by Dr. M. L. V. Gayler3 is in close agreement with the picture outlined above. With reference to work done on duralumin, copper-beryllium and copper-silver alloys, Dr. Gayler also concluded that age-hardening takes place in two consecutive steps: first, diffusion (in which the solute atoms migrate to certain crystallographic planes), and then precipitation. However, W. L. Fink and D. W. Smith4 have shown that the double aging peaks that appear in the strength curves of high-purity aluminum-copper alloys may be explained without recourse to pre-precipitation phenomena. By means of very careful metallographic technique, Fink and Smith demonstrated that during aging the precipitation takes place at an accelerated rate along the grain boundaries and slip planes as a result of the plastic deformation caused by the drastic quench from the solution heat-treatment temperature. The general precipitation occurs at a

Jan 1, 1939

By Morris Cohen

Within the past two years, a number of publications have called attention to the double peaks, or stages, that appear in the hardness and strength curves of certain aging alloys. The author has shown the existence of double hardness peaks during the aging of a high-purity silver-copper alloy;&apos; and on the basis of electrical resistance, microscopic, dilation and lattice-parameter measurements, hc concluded that the initial hardness peak occurs prior to precipitation while the secondary hardening is due to precipitation. It was suggested that the general mechanism of age-hardening involves two consecutive processes, the first consisting of nucleus or knot2 formation, and the second consisting of actual precipitation and particle growth. It was further held that the two stages are interdependent, but because of nonuniform precipitation they may overlap each other to an extent that increases as the aging temperature is raised. At high aging temperatures, the effect of the first stage cannot be detected because of the rapid occurrernce of precipitation. The general theory of age-hardening recently presented by Dr. M. L. V. Gayler3 is in close agreement with the picture outlined above. With reference to work done on duralumin, copper-beryllium and copper-silver alloys, Dr. Gayler also concluded that age-hardening takes place in two consecutive steps: first, diffusion (in which the solute atoms migrate to certain crystallographic planes), and then precipitation. However, W. L. Fink and D. W. Smith4 have shown that the double aging peaks that appear in the strength curves of high-purity aluminum-copper alloys may be explained without recourse to pre-precipitation phenomena. By means of very careful metallographic technique, Fink and Smith demonstrated that during aging the precipitation takes place at an accelerated rate along the grain boundaries and slip planes as a result of the plastic deformation caused by the drastic quench from the solution heat-treatment temperature. The general precipitation occurs at a

Jan 1, 1939

By R. W. Raymond

The tidings of Mr. Hewitt&apos;s death, cabled to me at Rome, Italy, brought me a great shock as well as a sincere sorrow. I had left him, a few weeks before, not indeed in vigorous health, but still quite able to command recognition as a continued power in public affairs; and I thought him destined to enjoy for many years that final stage of a long, active and honored career, in which, having passed beyond the struggles and rivalries of business and of politics, his life and character could be justly appreciated by all men, like some noble work of art, already crowned, and marked henceforward "hors de concours." Moreover, he was, and would be to the end, not merely the passive recipient of admiration and esteem, but the source of iuspiring appeals, effective warnings and wise counsel. His rich and varied experience he had placed at the service of his fellows. His lightest word was weighty, and those who contradiEted him only injured their own cause thereby. Great trusts (true " trusts," in the old and proper sense of the term), like the Cooper Union, the Carnegie Institution, and others, still commanded his disinterested services. Great principles of liberty and of justice continued to be reinforced in perilous crises by his powerful advocacy; great measures were daily receiving from him a decisive impetus. No wonder it seemed to. us that this happy period of reward so well earned, and of beneacence so welcome to innumerable recipients of its blessing, must continue yet awhile. We did not expect the night instailtly to engulf the full-orbed sun. Nor, indeed, has this been our experience; in this instance, the sunset has left an afterglow more luminous than many a noonday. For the principal data contained in the following brief sketch of Mr. Hewitt&apos;s life, I rely upon memoranda personally approved by hirn, some months before his death, for my use in

Jan 1, 1904

By D. L. Feucht, R. L. Longini

The semiconductor heterojunction is considered in terms of simple models which may lead to an understanding of move complex heterojunctions. Metallurgical and electrical properties of hetero-junctions aye discussed including the interface structure, energy -band diagram, and carrier transbovt across the interface. It is found that in a heterojunction all mechanisms such as injection, tunneling, and junction recombination found in simple junctions play modified voles. INTERFACES between materials (grain boundaries, the electrical junction between two differently doped materials in a single crystal, the oxide-metal interface, or metal-metal junctions) are of considerable importance in many situations. These various interfaces all have one very fundamental thing in common. Quantum mechanically speaking, the wave functions of the electrons in one material may penetrate the other material but, in general, only to the extent of angstroms. From an electrical point of view the conduction mechanism changes as a current passes through such junctions. In some cases the change is tremendous, in others almost negligible. The interface, then, is the locus of a change of conduction mechanisms. Some of these, particularly in semiconductors, are well-understood. The ordinary p-n junction in a single crystal can be the locus of an injection mechanism or a tunneling process, depending on conditions. The mechanisms are probably best understood in semiconductors because of the possible simplified view of particlelike conduction. The bands are either nearly filled or nearly empty and band overlap is seldom involved. The same fundamentals are probably important in other situations too but they are very difficult to look at naively. Although the simple look at the semiconductor case only gives us a relatively rough picture which must then be refined, the other systems, which involve a more complex situation, immediately are in many ways too difficult. There are too many initial choices of complex systems and therefore it is not possible to be even reasonably certain of any one model. Because of the relative simplicity of semiconductors, their good and controllable structure, and because of the ability to make many measurements on them not normally available to either metals or insulators! they are probably the best understood materials. It is therefore desirable to use them as a tool to further the understanding of interfaces in general. Semiconductor-heterojunction concepts were first proposed by kroemer1 in 1957. This was followed several years later by reports on the fabrication and experimental characteristics of heterojunction structures by Anderson2 and Diedrich and jotten.3 I) THE HETEROJUNCTION STRUCTURE To get down to hardware, when we refer to a semiconductor heterojunction we imply that there exists an intimate contact between different semiconductor materials. We could put two pieces of material together, complete with oxide layers, we could remove the oxides, or we could even melt the interface and hopefully get wetting and a good "bond" on solidifying. In fact we could by some means grow a crystal of one material using the other as a seed. Essentially we are interested only in the last two because they are the simplest to look at analytically. The degree of perfection of fit varies greatly and is reflected somewhat in the arc welder&apos;s joint strength. The lattice match of the two materials, their orientation, and so forth. is obviously necessary for a good bond but so is the continuity of any polar bonds which are involved such as in the III-V semiconductors. The mechanical misfit between two similar lattices can be described in terms of edge dislocations. The edge-type dislocations must be very close together for the usual misfit and there must be dislocations for each of several different Burger&apos;s vectors in order to produce a lattice match. The .&apos;dangling bonds&apos;&apos; resulting will be involved in producing interface charge. Order of magnitude estimates of the charge density extrapolated from low densities of dislocations in homogeneous materials give 5 x 1013 cm-2 Ge-Si and 1 X 1012 cm-2 Ge-GaAs electronic charges. Edge dislocations also act as very active recombination centers between holes and electrons. One lattice "matching" difficulty usually exists even if two structures have essentially the same lattice constants as they will have different coefficients of therma1 expansion. Thus, on cooling from the usually high temperature of fabrication to room temperature, dislocations are produced, a good fit not existing at both temperatures. In brittle materials this shrinkage may even result in cracking. For the Ge-Si interface the mismatch is about 2 x 10 -6 per degree whereas it is less than 10"7 per degree between germanium and GaAs. The exact effect of the misfit is dependent on the thickness of the materials involved. For a very

Jan 1, 1965

By Charles H. Morgan

THE process of wire drawing depends upon the property which certain metals possess, termed ductility, &apos;which is defined in Brande&apos;s Dictionary of Science as a property in consequence of which metals may be drawn out at length without interrupting the continuity of their constituent parts. As defined by M. Tresca, it is the property of flowing when subjected to pressure, and is termed by him the "flow of solids," in his paper, read before the Academy of Sciences in Paris, November, 1865 * In the simple language of the wire shop, it is that quality in certain metals that permit, them to be drawn (or stretched) out through a tapering, hole in a metal block or die, the die being more dense and harder than the metal to be reduced. It is a fact well known to those skilled in the art, that to succeed in drawing iron or steel wire, it is of the utmost importance to have thorough lubrication, as a lack of it would caws abrasion of the wire, or of the walls of the tapering hole in the die or plate, and quickly put an end to the operation. In drawing coarse or large sizes of wire, say 2/10 inch diameter, the pressure upon the sides of the tapering hole, where the wire comes in contact with the die, is so great that an ordinary lubricant is squeezed out when the usual reductions in size are made, and abrasion takes place. In order to meet this difficulty, and secure proper lubrication, it has been&apos; a common practice for years to apply a paste made of rye or wheaten flour, or lime, to the surface of the wire to be drawn, and when the paste is dry, to smear the wire with tallow or grease; the wire is then ready to be drawn, the dried paste serving to prevent the lubricant from being pressed out in the. process. Some years since, when Bessemer metal came largely into use as material for wire, "the writer found that its drawing required so much more power than iron wire, that he was led to institute a series of trials to ascertain the difference in power required. It was found that Bessemer wire required from 100 to 200 per cent. more power to make the same reduction in sizes than soft iron wire, the difference depending upon the amount of carbon and other chemical constitu- * Interesting experiments in the flow of metals have been made, by Messrs. Hoopes Townsend, of Philadelphia, in cold-punching iron, published and illustrated in the Journal of the Franklin Institute, of March, 1878, by Mr. David Townsend.

Jan 1, 1881

By Gustav Egloff

THE treatment of emulsified crude oils has been a problem for years in the petroleum industry. Until comparatively recently various methods of settling with and without chemicals have been used with indifferent success. Since wells have been acidized, emulsified crudes have been produced, which acid considerably to the need for improved methods of treatment. The presence of large quantities of salts in the oils from acid-ized wells has presented several new problems to the industry. The use of hydrochloric acid in well treatments has caused the formation of sodium, magnesium, and calcium chlorides by interaction with the dolomitic limestones. These salts dissolve readily in the water normally present in the oil, and they are therefore incorporated in the emulsified oil. The crystalline salts have also been found in the oil, prevented from settling by protective films such as asphalt, resins and petrolatum. In, some instances, free hydrochloric acid has been found. The salts may be present in crude oil in amounts ranging from a few pounds to over 1000 lb. per 1000 bbl. of oil. The transportation of these crude oils through pipe lines has led to erosive and corrosive effects, which are a tremendous cost to the petroleum industry. During 1937, more than 1,200,000,000 bbl. of crude oil was transported through pipe lines. Of this quantity approximately 10,000,000 bbl. was water, and considerably more than 50,000,000 lb. was salts. If the crude oils were treated at the producing wells to remove the salts and water, an appreciable saving would be possible. The treatment would not be difficult, since it has been found that these oils are more easily treated at the time they are produced than after the aging and more stable emulsification that occur before they reach the refinery. Since great costs are involved in the use of these salty crudes at the refinery, treated oils essentially free of salts would naturally command a premium.

Jan 1, 1938

By R. V. Wheeler, W. Francis

Of the methods that have been used for studying the chemical composition of coal, attack by reagents has not, in general, yielded much information. Most of the reagents used have been strong oxidants such as nitric acid or Schulze&apos;s solution. The use of a milder oxidizing agent, such as air at low temperatures, enables progressive changes in the character af the coal substance to be studied and thus helps toward an understanding of its constitution. British bituminous coals frequently have a banded structure. These bands (of four types) can be separated one from another and have been found to possess markedly different chemical characteristics, even when taken from contiguous portions of the same lump coal. It has been found convenient to give each of the ingredients of banded coal a distinctive name, more or less descriptive of its appearance, and the study of the chemistry of bituminous "coal" in England has resolved itself into a study of each ingredient separately. The coal chosen for this work was taken from the Top Hard seam at East Kirkby colliery, Nottinghamshire, a detailed chemical analysis of which has been recorded by Baranov and Francis.&apos; The banded structure was well marked and the bands could readily be separated. The research was confined to the brilliant vitreous bands (vitrain), the dull hard bands (durain) and the "sooty partings" main); for previous work had shown that the fourth type, to which the name clarain has been given, is intermediate in character, chemically, between vitrain and durain. The method of experiment was, briefly, to draw a slow current of moist air through weighed samples (50 gm.) of the three ingredients, vitrain, durain, and fusain, in the form of powder (through a 60-mesh and on a 90-mesh sieve) contained in glass tubes which could be maintained at a constant temperature during prolonged periods. After passing through the coal, the air bubbled into a measured volume of

Jan 1, 1925

By Lawrence H. Van Vlack

The energy of the y-iron grain boundary was determined to be 850 ergs per cm2 at 1105°C. The a/a and the a/y boundaries possess somewhat less energy. The microstructures of several iron alloys are discussed in terms of the interfacial energy relationships. THE dependence of the shape of liquid interfaces upon their energies has been studied for some time.1"3 Recently, Smith&apos; concluded that many of the microstructures observed in metals may be attributed to the energies of the interfaces between the grains and phases. While the energies of liquid surfaces have been measured by numerous different methods: only two attempts have been made to determine experimentally the energy of surfaces involving crystalline solids. Udin, Shaler, and Wulff8 used a procedure introduced by Berggren7 to measure the surface tension of solid copper. It consisted of determining the load required on a thin copper wire so that the surface tension was counteracted, and no net strain resulted. Bailey and Watkins8 sed a different technique and were able to determine both the surface tension of copper and the tension of the copper grain boundary. They measured the contact angle which liquid lead makes against a copper surface. They then calculated the above energies from the surface energy of liquid lead by using the Pb vs. Cu/CuY dihedral angle and the surface groove angle formed during thermal etching. They obtained a value of 1800 dynes per cm for the surface tension of clean copper and approximately 640 dynes per cm for the tension of copper grain boundaries. The former value was somewhat higher than Udin&apos;s value of 1500 dynes per cm. This paper includes a determination of the energy of the y-iron grain boundary. In addition, the micro-structural relationships of several other metallic and nonmetallic phases with iron were examined and an attempt was made to interpret the observed structures in terms of interfacial energy relationships. Theoretical Considerations In a three-phase material containing a solid and two liquids, there may be two solid/liquid interfaces, one liquid/liquid interface and one inter- granular interface. Liquid/liquid interfacial energies can be measured directly. Therefore, by determining the energy of such an interface and by measuring the required interfacial angles: the energies of the interfaces involving solids may be obtained.10 They may be shown schematically as in Fig. 1. The above procedure requires the following assumptions: 1—The interfacial energies are, in general, unaffected by boundary orientation, and 2— the shapes of the interfaces are unaltered between annealing and sectioning for microscopic examination. The first assumption is good to a first approximation as shown by Dunn and Lionetti11 except for a few critically orientated boundaries. Two conditions may alter the shape of a boundary upon cooling: 1—Precipitation of dissolved material upon one side of the interface, and 2—a change in volume during the transition from one phase modification to another. The former must be detected by microscopic observation. The latter may become appreciable in systems containing an interface between two liquids which solidify after annealing. However, possible movement of the two-liquid interface may be checked in the system diagrammed in Fig. 1 by measuring 8, for example, and comparing it with its value as calculated from the other interfacial angles by: tan8, C11 05 COS- o- + cos 8, cos- The choice of a suitable pair of liquids is limited by several factors. The relative values of the several

Jan 1, 1952

By Jim Douglas, H. H. Rachford, D. W. Peaceman

The problem of unsteady-state gas flow through porous media has been solved numerically only for the case of linear or radially symmetric reservoirs. A recently introduced numerical method for solving the unsteady-state heat flow equation in two dimensions is applied to the calculation of the depletion of a square region containing a perfect gas. Solutions are presented in graphical form for various values of dimensionless parameters. The solutions are compared with published .solutions for radial reservoirs. I NTR ODUCTION The problem of unsteady-state flow of gas through porous media gives rise to a second-order non-linear partial differential equation for which no analytical solution has been found. Numerical approximations to solutions of the gas flow problem have been obtained by the stepwise solution of an associated difference equation1,2. However, the methods so far developed have required that the reservoir be either linear or radially symmetric. This restriction in shape has been necessary so that only two independent variables be considered, namely, one distance variable and time. In order to deal with reservoirs having more realistic shapes, it is necessary to develop numerical procedures for the solution of the gas flow problem involving two distance variables. A numerical procedure, denoted as the alternating-direction implicit method, for the solution of the heat-flow problem in two dimensions has recently been introduced" By the use of this procedure, approximate solutions have been obtained for heat-flow problems in a square8, and in regions having various non-rectangular boundariesa. Because of the similarity between the equation for heat conduction in solids and the equation for gas flow in porous media, it is reasonable to expect that the alternating-direction implicit method should also be useful for solving the gas flow problem in various two-dimensional regions. Solutions have been calculated for the simplest two-dimensional region, a square reservoir with a single well in the center. While reservoirs of such simple geometry seldom, if ever, exist, the solution of this problem is of some practical importance because it is also the solution of another problem, that of determining the depletion history about an individual well in an infinite uniform reservoir containing wells spaced throughout on a square lattice and producing equally from each well. Finally, it is of interest to compare solutions for a square reservoir with those for a circular reservoir and to determine the effect of the shape of the boundary for that particular case. METHOD OF CALCULATION Basic Differential Equation By combining the equation of continuity, the perfect gas law,

Jan 1, 1956

By G. E. Pellissier, P. H. Salmon Cox, B. G. Reisdorf

Banding that occurred in plates rolled from the early production heats of 18Ni(250) maraging steel is described and related to the segvegation of certain alloying elements (nickel, molybdenum, titanium), the extent of which was quantitatively evaluated by means of electron-microprobe analysis. The effect of banding on mechanical properties is discussed, with particular reference to observed directional differences in plane-strain fracture toughness of plates. It is shown that banding originates as interdendritic segvegation during ingot solidification and persists in some degree through normal soaking and hot reduction to plate. The results of the study showed that heating sections of small laboratory-cast ingots at 2200°F for 4 hr was sufficient to markedly reduce microsegregation and to considerably improve mechanical properties. Hot rolling of 7-in.-thick ingot sections to 1/2-in.-thick plate effected a similar reduction of microsegregation, but resulted in even greater increases in ductility and toughness than that obtained by homogenization treatment alone. DURING the past few years, considerable attention has been directed towards the low-carbon, high-alloy maraging steels and in particular towards the 18Ni-8Co-5Mo-0.4Ti alloy. The steels of this group, having an excellent combination of high strength and toughness, have a number of advantages over their more conventional medium-carbon low-alloy, quenched-and-tempered counterparts. In the annealed condition, the maraging steels are in the form of a ductile marten-site; aging at a relatively low temperature, typically 900°F for 3 hr, increases greatly the strength through the precipitation of intermetallic compounds. One problem in the early production heats of maraging steel was that the finished plate frequently displayed a banded structure. Previous work on other steels1-&apos; had established that banding in wrought products is either a direct or an indirect consequence of chemical segregation, which occurs during solidification and persists to some extent through normal thermal and mechanical treatments. For example, Smith and others: in a study of low-alloy steel, were able to correlate the severity of banding in the wrought product with the degree of interdendritic segregation of nickel and chromium in the as-cast ingot. The effect of banding on the mechanical properties of steels is usually considered to be detrimental, although there is only limited evidence to suggest that a marked improvement in properties can be obtained with less heterogeneous structures. Comparison of the longitudinal and transverse tensile properties of banded and of homogenized 4340 steel showed that only the transverse ductility was improved by homogenization, but even then the improvement was not commercially significant.&apos; Conversely, homogenization of through-the-thickness tension specimens of quenched-and-tempered steel plate, containing 1.47 pct Mn, increased the strength by as much as 10 pct and the tensile ductility by at least a factor of twos5 This improvement was related to the elimination of manganese-rich bands, which also are one of the factors responsible for cold cracking in the heat-affected zone of metal-arc welds.7 In the present study the nature and severity of banding in early commercial 18Ni(250) maraging steel plate and in laboratory-melted 18Ni(250) maraging steel plate was determined. The effects of banding on plane-strain fracture toughness and the effects of thermal homogenization treatments on the strength, tensile ductility, and toughness of 18Ni(250) maraging-steel as-cast ingots and rolled plate were evaluated. In addition, the effects of hot deformation by rolling on the mechanical properties of ingots were determined. 1) STUDIES OF BANDING IN EARLY PRODUCTION PLATE The chemical composition of the steel (A) used in this part of the investigation is shown in Table I. Banding was not clearly evident in either as-rolled or annealed* plate, but annealed and agedc** plate had a banded structure. The typical banded condition, Fig. 1, consists of layers of unetched austenite (white) and dark-etching martensite in a light-etching martensitic matrix. X-ray diffraction measurements showed that this steel contained more than 6 pct austenite. An electron-probe X-ray microanalyzer (using a focused beam of electrons) was used to determine the composition of the bands and of the material between the bands with respect to the main alloying elements— nickel, molybdenum, titanium, and cobalt. The recorded X-ray intensities were converted to concentration values with the use of a standard of similar composition. To facilitate probe positioning, all analyses were conducted on specimens that had been given a light etch. The influence of this etching on the analytical results was negligible; analyses made on the identical area before and after etching yielded essentially the same concentration values. The results of the electron-microprobe analyses at selected points revealed that the layers of austenite and adjacent dark-etching martensite contained greater amounts of nickel, molybdenum, and titanium than did the surrounding matrix, Table 11. The austenite layers

Jan 1, 1968