Search Documents

By Brahm Prakash, R. Schuhmann

Chemical conditions for flotation and nonflotation of quartz with oleic acid as collector and barium, calcium, aluminum, iron, and tin as activators were studied using a simple vacuum-flotation technique in glass-stoppered graduates. The detailed study of barium activation led to an interpretation based on ideal Langmuirian chemi-sorption. FLOTATION of quartz is of practical importance as something to be avoided in soap-floating many types of ores. Clean, unactivated quartz is not floated with fatty acids and soaps, such as oleic acid and sodium oleate, in the quantities normally used for flotation. However, data in the literature indicate that almost any multivalent cation will activate quartz if given an opportunity. Thus, a common problem is to prevent activation of quartz by the various inorganic cations inevitably present in flotation pulps. Wark and his coworkers1 have demonstrated the reversibility of the chemical reactions and adsorptions involved in the activation, depression, and collection of the common sulphide minerals. The procedure in much of their work was to bring a mineral surface to equilibrium with solutions of known pH, collector concentration, and activator concentration, and then to test the floatability of the mineral by contact-angle measurement. From the data, graphs were constructed with pH and reagent concentrations as coordinates. These graphs show fields of flotation and fields of nonflotation, separated by narrow transition regions whose locations are shown by so-called contact curves. From the shapes and locations of the contact curves, which roughly separate fields of flotation from fields of nonflotation, a quantitative understanding of the interaction of the reagents with each other and with the minerals often can be deduced. The study of quartz flotation to be described in this paper follows in broad lines the approach of Wark and coworkers. That is, pH, activator concentration, and collector concentration were varied to find equilibrium conditions of flotation and non- flotation, and the results are presented graphically by means of contact curves. However, instead of testing for floatability by measuring the contact angle on a polished surface, a simple vacuum flotation technique was developed and used. Purified oleic acid was the collector and terpineol the frother. Barium activation was studied in some detail, and exploratory studies were made of activation with calcium, aluminum, ferric iron, and stannic tin. Preparation of Materials Quartz: Large lumps of high-grade vein quartz were crushed dry in a cone crusher and rolls. The —20, +28-mesh portion was screened out and used in the subsequent steps. This material was passed through a high-intensity magnetic separator to discard iron, then leached twice with hot concentrated HCl and washed repeatedly with distilled water. The cleaned sand was then wet ground with porcelain balls in a porcelain pebble mill, deslimed repeatedly by settling and decantation to discard —800-mesh material, and again washed with hot HCl followed by distilled water. The resulting stock of quartz was stored under water. Chemical analysis gave 99.8 pct SiO2. Table I gives the size analysis of the quartz used for flotation tests. Calculations from these data, using shape factors given by Gaudin and Hukki9 indicate a specific surface of about 500 cm2 per g. Blank flotation tests in distilled water, and in water with added frother, showed the prepared quartz to be completely nonfloatable and thus indicated the absence of organic contamination. Oleic Acid: The preparation of oleic acid was based on fractional vacuum distillation of methyl oleate2,3 followed by regeneration of oleic acid, and finally fractional crystallization of oleic acid from acetone solutions at low temperatures." The pure oleic acid was stored in a refrigerator. The iodine number of the oleic acid was found to be 90.0 (theoretical 89.93). Oleic acid was used in the form of a dilute water solution of sodium oleate, after preliminary flotation tests showed no effects of form of addition and order of addition of reagents when an adequate conditioning time (that is, 30 min) was provided. Other Reagents: Sodium hydroxide solutions low in carbonate were prepared by first making 1:1

Jan 1, 1951

By J. S. Aronofsky

The first step in a quantitative analysis ot the mechanism of oil displacement by water in a fractured reservoir is usually conceded to be the solution of the differential equation describing the saturation distribution of two immiscible fluids flowing in a porous medium, where the capillary pressure is taken into account. In such a system the production mechanism may consist of displacement of oil both by the flow of water due to natural or artificially imposed pressure gradients and by imbibition, which implies a flow of water not due to external pressure gradients. Owing to the presence of the two oil displacement mechanisms, the mathematical model given by the differential equation intended to describe the system may not properly represent the behavior of the physical system. In fact, in the reservoir the rate of water advance may be very slow, and in the case of a fractured reservoir with a great number of large fractures, the pressure difference determining the flow of water through the matrix may be much less than 1 lb/'psi over lengths of a few feet. In such a case, imbi-bition (the exchange between oil in the matrix and water in the fractures resulting from capillary forces) may become, with time. a significant element of the production mechanism. It occurred Lo the authors, however, that without going into a physical analysis of the process of production, it might be possible by means of simple abstract reasoning to throw some light on the variation of recovery with time under conditions occurring in a highly fractured oil reservoir with rising water table. The object of this paper is to present both the reasoning and its application to a reservoir of the highly fractured type. Specifically, the analysis given here was undertaken to try to explain the increase of recovery (as defined later) with time as observed in this reservoir, without having to assume unlikely variations in the reservoir parameters with depth. This attempt has been successful as will become clear upon comparison of the computed recoveries with the actual field data. ABSTRACT MODEL Let us consider a small volume. of porous matrix saturated with oil at time, t = 0. Let the process of oil displacement by water start at time, t = t,,. At some time, t, the process will have terminated. Then a volume of oil equal to or smaller than the original oil contained in the matrix will have been produced. The first basic assumption that describes the model and guides the forthcoming reasoning is that the oil production from the small volume. dv, is a continuous monotonic function of time and that it converges to a finite limit. Such an assumption is not inconsistent with the results of laboratory waterflood tests as well as with results of imbibition tests where this is, in fact, observed. The second basic assumption is that none of the properties which determine the rate of convergence change sufficiently during the process to affect this rate or the limit. Let it be assumed that the form of the function of time relative to production from the matrix volume. dv, is given by where V,(t) is the volume of oil produced up to that time t. R is the limit toward which the recovery converges, A is a constant giving the rate of convergence, and V,(t) is the volume of oil originally in place in the volume, dv. It should be noted that recovery at time I will be understood here to he It follows that r. tends to R as t tends to infinity. CONSTRUCTION OF RESERVOIR FROM ABSTRACT MODEL Let the reservoir consist of a series of identical blocks of porous matrix stacked vertically and separated by fractures. Let each of these blocks satisfy the conditions of our abstract model. These conditions are: (1) that recovery is a continuous mono-tonic function of time converging to finite limit, and (2) none of the properties that determine the rate of convergence change sufficiently to affect the rate or the limit. Let water be rising in fractures so that oil production from any part of the block starts when the water comes in conLact with it. TOTAL RECOVERY FROM THE RESERVOIR COMPARED TO RECOVERY FROM A SINGLE MODEL ELEMENT As stipulated in Eq. 1, in the case of the abstract model,

By P. W. Bennett, E. M. Elkin

A new technique for the recovery of palladium and platinum and sludge from go12 electrolyte eliminates many of the drawbacks of the zinc-dust cementation process. In the electrolytic refining of gold by the Wohwill method, the electrolyte contains about 100 gpl gold as AuCl, and 100 gpl free hydrochloric acid. Any palladium and platinum contained in impure gold dissolve and accumulate in the electrolyte. When their concentration approaches that of gold, it becomes desirable to discard the electrolyte and to recover the gold and the palladium and platinum. As practiced at Canadian Copper Refiners until recently, discarded electrolyte was treated with sulfur dioxide to precipitate the gold. 2AuCl3 + 6H2O + 3SO2 —2Au + 6HC1 + 3H2SO4 The gold precipitate was filtered off, washed with water, and melted with scrap anodes. The wash water was added to the gold-free filtrate and treated with zinc dust to cement the palladium and platinum. PdCl2 + Zn — Pd + ZnCl2 PtCl4 + 2Zn—Pt + 2ZnCl2 The precipitate was separated and sold as Pd-Pt concentrate, and the barren filtrate and wash water were discarded. The receipts of palladium and platinum at this company were too small to warrant their refining. The process suffered from a number of disadvantages. Handling the gold-free solution was disagreeable, as it was laden with sulfur dioxide even after blowing with air overnight. Subsequent treatment with zinc dust was also troublesome. A large excess of zinc dust was required. Spray, acid mist, and vigorous evolution of hydrogen accompanied the reaction. The Pd-Pt precipitate was grossly contaminated with occluded unreacted zinc and with yellow granules of sulfur, thought to be formed by the reduction of sulfur dioxide and its derivatives. Repeated boiling of the precipitate with hydrochloric acid and washing with water decreased somewhat, but did not eliminate, the zinc. After drying overnight, the precipitate hardened to a clinkerlike mass that was difficult to grind. Grinding, handling, and packing created heavy dusting. Even weighing in an analytical balance generated dust, indicating that the powder was easily charged with static electricity. The total palladium and platinum content of the powder varied from 56 to 72 pct, with an average of 65 pct. The powder dissolved in aqua regia with difficulty, leaving a small amount of insoluble residue. However, the main drawback of the process was the strongly hygroscopic Pd-Pt product. The powder showed a marked though erratic increase in weight and a corresponding decrease in assay on exposure to air, even during the course of laboratory analysis. Such changes were the source of disagreement between Canadian Copper Refiners and the buyers of concentrate. The cause of the change in weight is believed to be hydrogen, liberated by the action of zinc dust on hydrochloric acid. Some of this hydrogen was adsorbed so strongly by the precipitate that it could not be driven off during the drying. On exposure to air it oxidized catalytically, forming water. Experiments soon showed that the entire process of treatment of discarded electrolyte then in use would have to be changed. A small quantity of apparatus was moved into the gold room, and experimentation was held to a minimum compatible with the development of a new process. PALLADIUM AND PLATINUM PRECIPITATION It was thought that zinc could be replaced with a reagent other than a hydrogen-evolving metal. Semiquantitative experiments showed that platinum and palladium could be precipitated completely from the gold-free solutions with such well-known reagents as sodium formate, formic acid, formaldehyde, and paraformaldehyde. Sodium formate was the preferred reagent, since the other three are volatile and somewhat toxic. The gold-free solutions used were those saturated with sulfur dioxide, partly freed of sulfur dioxide by aerating overnight, completely freed by boiling, as well as strongly and weakly acidic, neutral, and alkaline solutions. The reactions are complex and can be represented by these over-all equations: PdC1, + HCOONa — Pd + NaCl + HC1 + Co2 PtCl4 + 2HCOONa — Pt + 2NaCl + 2HC1 + 2CCX, Strongly acidic solutions required as much as five times the theoretical amount of sodium formate as did solutions neutralized to pH 4.5. Increasing the pH decreased the consumption of

Jan 1, 1965

By R. V. Hughes, J. N. Breston

Conclusions drawn by previous research workers with reSPect to the relation between Pressure gradients and/or velocity and oil recovery obtained by laboratory water flood tests have been in disagreement probably due to variable procedures and unnatural conditions and materials. The Bradford Laboratory of the Pennsylvania Grade Crude Oil Association as part of its secondary recovery research program has conducted nineteen water floods on two long cores of widely differing characteristics in an attempt to clarify this relationship and make it an aid in predicting flooding pressures in the field. Unlike previous research procedures the present experiments were conducted with the aim of duplicating field conditions as closely as possible by using long unextracted consolidated cores, a live crude, and natural brines for both flooding and, connate water content. Also, the pressure gradients and flooding velocities were representative of field conditions where similar sands were being flooded. Eleven floods on one core and eight floods on the other core showed increased recoveries and lower residual oil saturation with increased flood pressure gradients and flood velocities. A marked decrease in recovery was obtained from both cores at very low flood velocities. This pressure versus recovery relationship is shown to hold up to the point of water breakthrough and also up to the 100 and 1 produced water to oil ratio point. INTRODUCTION The possibility of water flooding oil sands was suggested by Carl1 of the Pennsylvania Geological Survey in 1880. It is not known when the practice was tried intentionally for the first time, but its beneficial effects were noted in the annual production rate of the Bradford field as early as 1907. The practice was illegal in Pennsylvania until 1921. Early water floods in the Bradford field usually consisted of shooting or splitting the casing secretly to permit subsurface waters to enter the producing sands under hydrostatic head. As it was noted that the benefits of water flooding seemed to be proportional to the quantity of water dumped into the well many also began to utilize surface sources after the practice became legal. It was probably during the middle 20's before many producers realized that the pressurehead of the water upon the producing sand determined the rate and quantity of water that would enter the sand. Hence, rate and quantity of production appeared to be a direct function of input pressures. By 1927 a few producers had ventured the installation of pressure pumps in order to increase water-input rates and production through the combination of hydrostatic and hydraulic pressures. The adoption of pressure flooding and the "five-spot" drilling pattern in the Bradford field were essentially simultaneous. Water-input pressures in 1930 seldom exceeded 300 p.s.i. at the well head or 1100 p.s.i. at the sand face. Since that time, water-flood producers in the Bradford-Allegany fields have gone to higher and higher pressure until today 600 p.s.i. at the well heads is called a low pressure flood. Many high pressure floods now operate at 1300-1400 p.s.i. at the well head. The limiting and advisable pressure at the sand face has been pronounced as that pressure just under what is required to lift the overburden or to cause formation parting: According to this rule any water flood operation utilizing well-head pressures nearly equal in pounds per sq. in. to 1.1 times the average depth in feet to the top of the producing sand would be considered as a high-pressure flood. Despite the higher pressure trends in the Bradford-Allegany field operations and the results of early laboratory water flooding research, the desirability and benefits of high input pressures are still questioned by many operators, particularly in midwestern water flood operations. The present paper recounts a series of 19 laboratory water floods using two long, consolidated cores of widely differing characteristics saturated with live crude and flooded with oil field brines in an effort to simulate field conditions under • various pressure gradients and flooding velocities. For both cores, higher recoveries and lower residual oil saturations were obtained at higher pressure gradients and flood velocities. This relationship is shown to hold up to the water breakthrough point and also up to the 100 to 1 produced water to oil ratio point.

Jan 1, 1949

By Joseph L. Gillson

IT is a somewhat curious circumstance that the newest of the Institute's several awards should be conferred in the oldest of our several professional fields—for there is little question that geology and mining have antedated metallurgy and petroleum. Perhaps so did geophysics. After all, one geophysical instrument, the magnetic compass, has been successfully employed for many centuries. Young in years the Jackling Award may be, but in distinction—whether in terms of the title it bears or in terms of the first three medalists (Reno Sales, E. D. Gardner, and the late Father MaceIwane)—the Jackling Award is already one of the ranking awards of the Institute. The name of the 1957 medalist will bring added luster to an already illustrious roll. J. L. Gillson was born and brought up in Evan-ston, Ill., in an area characterized by singularly flat and uninspiring topography, and devoid of ore deposits and even of bedrock—a most unlikely environment from which to expect a future Jackling medalist to emerge! On graduating from high school he enrolled as a chemistry major at Northwestern University. At the end of his frkshman year his scholastic prowess had won him a prize of two chemistry books. But these went unread, for about this time he came under the spell of U. S. Grant, that distinguished and enthusiastic geologist who for many years was head of Northwestern's geology department. From thence forward Joe was a dedicated geologist. After a Navy interlude in World War I, he went to MIT for graduate work, substituted for the late Professor Palache for one year at Harvard in mineralogy, received his doctorate at MIT and stayed on as a meniber of the MIT faculty, where the geology department was headed by that great dean of all economic geologists, Waldemar Lindgren. In 1926 Joe spent a summer looking into barite deposits for the du Pont Co., and since 1928 he has been du Pont's chief geologist. Do men who lead such active lives in our exploration organizations generally have permission to publish their observations? Do they generally find time to publish, even if they have permission? Alas, no. But in Joe Gillson's case, yes. His bibliography runs to well over two score papers, including one of the very first papers in this country on the applications of petrography to problems of Portland cement. He has written on such topics as the origin of talc and the genesis of ilmenite, fluorspar, and alkaline rocks. Each year for a number of years his annual reviews of industrial minerals in the Mining Congress Journal have been eagerly awaited. - Do men who do so much to advance their science have anything left to give to the advancement of their profession? In Joe Gillson's case, yes. He has been a vice president of the Mineralogical Society of America and has been active in the councils of the Society of Economic Geology, particularly in their research programs. This year he is president of the American Geological Institute, currently perhaps the most challenging post to which a geologist may be elected. His contributions to AIME scarcely need documentation, so I shall mention only three—his chairmanship of the Committee on Democratization of the Institute, a committee whose work and recommendations brought a new and hopeful leaven to AIME at a time when this was sorely needed; his chairmanship, a few years ago, of the Industrial Minerals Division; and his current place on the Board of Directors as Vice President of the Institute. Such' a catalogue of contributions and achievement, even as condensed as I have had to make it, certainly merits the Jackling Award. Yet, implicit although not often expressed in the conferring of most awards is the hope that besides recognizing a distinguished career, it will provide stimulus for further effort. In the present instance, such an expectation is already well on the way to fulfillment. Only recently Joe Gillson has undertaken the editorship of a new and greatly revised edition of the Institute's Seeley Mudd volume, Industrial Minerals and Rocks. And so, Mr. Chairman, it is with pleasure and enthusiasm that I present Dr. J. L. Gillson for the 1957 Jackling Award.

Jan 1, 1958

By W. W. Stephens, W. J. Kroll, H. P. Holmes

THE only two methods for producing commercial quantities of malleable zirconium, up to now, have been using magnesium reduction of the anhydrous chloride under a neutral gas, and using purification of raw zirconium by the iodide dissociation method on a hot filament. The purity of the metal obtained by the latter process is about the same as that resulting from the chloride reduction, with the exception of the oxygen content, which may be about 0.05 pct lower in the iodide metal. Even traces of oxygen have a strong hardening effect on zirconium. Since, in the iodide process, the impurities, Si, Al, Fe, Ni, and Mg in the raw zirconium used, are carried over, it would appear uneconomical to use this method of refining solely to eliminate the traces of oxygen, unless very soft metal is desired. One important step to improve the iodide method, namely, dissociating pure iodide made from carbide and recycling the iodine, which would change this refining process into a method for extracting zirconium from the ore, has not yet been made. The Bureau of Mines, by using the chloride reduction, wanted to produce quantities of a reasonably pure and malleable metal at low cost for large-scale industrial use. This aim has been achieved, and a pilot plant permitting production of 600 lb of malleable zirconium per week will be described (fig. 1). Briefly, the Bureau of Mines process consists of reducing purified gaseous anhydrous zirconium chloride with fused magnesium under a noble-gas atmosphere. The reaction product, magnesium chloride, and any excess magnesium metal present are removed from the sponge zirconium in the next step by fusion and evaporation in a good vacuum at about 925°C. In this process, care is taken to keep air out of contact with the hot metal, which otherwise would contaminate the product with oxygen and nitrogen. Both these impurities embrittle zirconium even when present in an amount as low as 0.1 pct. A number of previous reports1-3 describe the method and the various improvements that finally led to a workable process. The major steps made were the production of large quantities of carbide, which is the raw material for producing the chloride, its chlorination with reasonable efficiencies, and the elimination of iron trichloride contained in the raw chloride, as well as that of the zirconium oxide which is always present in this product. The physical nature of the anhydrous chloride, which sublimes at 331°C at atmospheric pressure and melts at 420°C under a pressure of 25 atm, made it necessary to provide a special safety device on the purification and reduction vessels in the form of a lead-alloy seal, permitting escape of excess gases if pressure tended to build up. The main difficulties encountered in the reduction and in the following step, salt separation in a vacuum at elevated temperature, were caused by the material from which the reaction pot was constructed—iron. This metal reacts with zirconium at about 940°C with formation of a eutectic. The last step, salt separation by evaporation of the magnesium and magnesium chloride in a vacuum, is quite unconventional and has not been used before on such a scale. It is only due to the improved construction of high-vacuum pumps, especially oil-diffusion pumps, that such a method could be used. The various steps of the Bureau of Mines process, as described in the general outline above, may be summarized as follows: (1) production of the carbide, (2) chlorination of the carbide, (3) purification of the raw chloride, (4) reduction of the pure chloride with fused magnesium, (5) separation of excess magnesium and magnesium chloride in vacuum at elevated temperature, and (6) melting of the sponge obtained. Production of Carbide The arc furnace, described in a previous report,' was improved both to reduce the graphite consumption caused by the burning of the bottom plates and to concentrate the heat in order to obtain better power efficiency. This was achieved by providing the bottom with graphite plates entirely embedded in refractory bricks, the contact with the current being made with three water-cooled copper plugs. Figs. 2 and 3 show the arrangement. The crucible, made

Jan 1, 1951

By C. H. Li, R. D. Carnahan, R. J. Stokes, T. L. Johnston

When silver chloride deforms by pencil glide at temperatures of 26ºand 72°C, grain size has no effect upon the proportional limit and the material necks down to a knife edge under tension. At -196ºC, deformation takes place on fewer slip systems to produce straight slip traces, the flow stress becomes sensitive to grain size and fracture occurs by cleavage from an intergranular source without any reduction in area. THERE are significant differences between the appearance of the slip bands formed in silver chloride and in the alkali halides at room temperature. The bands in sodium chloride, e.g., are fine and geometrically straight, whereas in silver chloride they are coarse and wavy.' Such differences in the slip behavior prompted us to inquire what the effects of grain boundaries might be on the mechanical properties of these solids. As an introduction to a general study of ionic poly-crystal behavior, we have investigated the effect of grain size on the stress-strain relationships in silver chloride at different temperatures. We have chosen silver chloride chiefly because polycrystals can be easily prepared free from cracks and macroscopic voids by conventional techniques. Moreover, as we shall see, the slip behavior can be readily modified by changing the temperature. It will be shown that the effect of grain size on the yield stress of silver chloride is strikingly dependent upon the temperature of deformation and that this temperature effect comes about through a change in the number of slip systems which operate at the onset of plastic flow. EXPERIMENTAL PROCEDURES Polycrystalline tensile specimens were made by recrystallizing cold-rolled AgCl sheet. Thirty-gram quantities of AR grade silver chloride powder were melted in a low heat capacity furnace, cast into cylindrical forms 3/4 in. in dim and approximately 1 in. in length and then extruded with a 16:l reduction ratio into rods 3/16 in. in dim. Extrusion temperatures were varied from room temperature to 350°C. In order to impart various degrees of cold'work to the extruded rod. All extrusions were then cross rolled to sheet 0.040 in. in thickness at room temperature and machined to form tensile specimens having a gage length of 0.750 in. with a width of either 0.250 or 0.145 in.* Annealing was carried out in an air oven at 200°C for periods up to 20 hr. specimens having average grain diameters ranging from 0.024 to 0.40 mm were prepared by varying the extrusion temperature and annealing time in the conventional manner. To obtain specimens with grain sizes larger than 0.40mm, lightly deformed thin prisims of silver chloride single crystals (supplied by Harshaw Chemical Co.) were recrystallized. Grain size determinations were made by an intercept method and revealed that the grains were equi-axed. Laue back-reflection X-ray patterns showed little tendency to form a preferred orientation texture. Specimens which were to be used for the slip band studies were polished by immersion in fresh boiling concentrated NH,OH for 60 sec. They were then etch polished for 10 sec on each side by hand lapping on a stationary velveteen lap wetted with a dilute Hypo solution (5 g of Kodak Acid Fixer in 100 cc H20). Finally the specimens were rinsed in warm water and alcohol, blown dry and carefully mounted in the tensile machine for testing. Tensile tests were performed at three temperatures, namely, 26º, -72º, and -196°C. A screw-driven hard tensile machine was used at a constant strain rate of 0.006 in. per in. per min. Loads were determined with a light- proof ring fitted with A-18 strain gages and extensions were measured with a 0.0001 in. dial gage. Constant level baths of alcohol with dry ice and liquid nitrogen were used to maintain the two lower test temperatures. EXPERIMENTAL RESULTS Reproducibility—In view of the light sensitivity and reactive nature of silver chloride, it was necessary to determine whether reproducibility depended upon either exposure to light or to the condition of the surface. However, during the course of the work it became clear that variation of laboratory lighting

Jan 1, 1962

By R. I. Jaffee, F. C. Holden, H. R. Ogden

The effects of grain size and shape on alloys of titanium with nitrogen and aluminum have been determined. Increasing a grain size decreases strength and hardness and increases impact resistance. Quenching from the ß field produces subgrain markings delineating a plates in Ti-N alloys but not in Ti-AI alloys. This suggests a precipitation from the high nitrogen alloys. ALPHA titanium alloys are single-phase alloys with the hexagonal-close-packed structure of a titanium. If these alloys are worked and annealed within the a field, equiaxed structures result in which strength is derived from the solid solution of the alloying elements present. Strain hardening can be used to increase further the strength of a titanium alloys. This is the only other mode of strengthening the a type of alloy. Quenching from the ß field has been noted by many investigators to have little effect on the strength of iodide titanium and on commercial titanium. Ti-A1 alloys were found to be little different in hardness, whether annealed in the a field or quenched from the ß field.' These facts illustrate a characteristic of a titanium alloys: they are insensitive to heat treatments from the ß field. Underlying this heat-treatment insensitivity are the facts that the a-ß field is generally narrow in a titanium alloys, the transformation-temperature range increases with a-stabilizing content, and the a solubilities of the alloying elements are greater than the ß solubilities. When quenching is done from the ß field, there is no supersaturation of the alloying elements in the acicular transformation structure produced. Welds in a titanium alloys generally are ductile, because no transformation hardening occurs during cooling.1,2 Other desirable features of a alloys are: no thermal-stability problem, excellent toughness, and retention of strength at elevated temperatures better than a-ß or ß alloys of comparable room-temperature strength. The prime factors governing the properties of a titanium alloys are the amount and kind of solute present. This can change the a phase from a soft, tough material into a hard, brittle material with no apparent change in microstructure. Moreover, brit-tleness can exist over a wider alloy-content range, without change of phase, than in most other metals. The solutes for a titanium may be divided into interstitial and substitutional types. Oxygen and nitrogen dissolve interstitially in rather large amounts, extending well through the brittle range, and the transformation-temverature ranges are raised in the process.:' , carbon has a maximum interstitial solubility of about 0.5 pct at the peritectoid temperature, while at lower temperatures the solubility decreases to about 0.2 pct.3,5 Substitutional a solutes include aluminum and tin. Aluminum has a high a solubility of 25 pct Al, and increases the transformation-temperature range markedly.6,7 Tin has a high a solubility of about 20 pct, but has a relatively innocuous effect on the transformation-temperature range.' The choice between interstitial and substitutional solutes generally has been conceded to the substitutional solutes because of a belief that they do not have as adverse an effect on ductility and notch toughness. However, a recent investigation8 on the effect of hydrogen on titanium has shown that this element, which is not under compositional control, has a powerful detrimental effect on notch toughness. Hydrogen was shown to form a hydride which is practically insoluble (about 0.002 pct) in a titanium at room temperature, although the solubility at the eutectoid temperature of about 300°C is relatively high, about 0.16 pct. Limited data were presented on the effect of hydrogen on a alloys, where the same impairment of toughness as in unalloyed titanium is found. The work reported here describes the effects of interstitial and substitutional solutes on 1—the mechanical properties of high purity titanium free of titanium hydride and 2—the influence of structural variables on these properties. Nitrogen was selected as an example of an interstitial solute with high solubility and aluminum as a substitutional solute with high solubility. All compositions were well within the ductile range, since it was not intended to study the brittleness problem here, but to study only the factors influencing optimum mechanical properties. Procedures The alloys were prepared from iodide titanium as 1/2 Ib ingots double melted to insure homogeneity.

Jan 1, 1955

By G. A. Moore

A brief review is given of methods designed to produce metallic iron of high purity, and typical results are listed. A recent method, utilized at the National Bureau of Standards, consists of the extraction of ferric chloride by ether, reduction of this ferric chloride to ferrous chloride, further purification of this chloride, and the subsequent electrolytic deposition of metallic iron. iron produced by this procedure apparently is softer than, and otherwise different in properties from, any iron previously prepared and contains appreciably smaller amounts of impurities. THE history of attempts to produce "pure" iron reaches to antiquity and it may be presumed that each ancient armorer who succeeded in making a better steel concluded, correctly, that he had done a better job of removing the "base metals," and incorrectly, that he now at last had a "pure" metal. Early metallurgical papers mentioned use of "pure iron" in making alloys—this "pure" iron in most cases being inferior to some commercial stocks of the present time. Improvement has been continuous, and usually at a sufficient rate to convince each succeeding group of workers that they, at last, were using the really pure metal (until the analysts also improved their techniques to again discover the impurities). These adventures were reviewed in some detail by Cleaves and Thompson.' Although the ores of a metal may be abundant, difficulties in extracting it may make the pure metal very rare. When impurities are restricted to a total of a few parts per million, nearly all pure metals become rarities. Lead, copper, gold, mercury, silver, zinc, aluminum, bismuth, and the six platinum metals are claimed to be available with total impurities ranging from 2 to 50 ppm. The present small and scattered world supply of so-called "pure" iron holds an unimpressive place in another group of 16 metals having approximately 100 ppm of foreign material. Of about 20 less rare metals, only the platinum metals are more costly to prepare. While the production of such rare varieties of iron may appear insignificant in the presence of thousand-ton operations with 95 to 99 pct metal, it must be emphasized that all researches on commercially interesting irons and steels are in fact studies of the modifications of the properties of iron by additional materials. Until the properties of high purity iron are directly measured, all ferrous research must operate without known base values. Traces of impurities may affect the properties of a metal in many ways. Infinitesimal traces of solutes, by disturbing the electronic configuration, greatly change the electrical properties of transistors and semiconductors2-3 and slightly larger traces might alter these quantities in iron. Soluble impurities which disturb the perfection of lattice arrangement not only may alter the magnetic constants and electric properties, but by their close association with dislocation phenomena probably control the very existence of the "yield point"; determine the value of yield stress; and perhaps control the selection of slip and cleavage planes. It has been speculated that impurities might even cause the allotropic transformation in iron, but in any case their rearrangement must contribute to the unreliability of heat capacity and other thermodynamic measurements. Impurities which do not remain in solution may cause even greater effects on the properties. Microscopically visible amounts of phases other than ferrite can be found in all high purity irons which have come to my attention. It can be calculated that from 50 to as little as 2 ppm of an insoluble material might be sufficient to completely film all grain boundaries in irons having grain sizes from ASTM Nos. 10 to 1. Should this occur, such films, even though invisible, may be very important in fracture problems, especially at extremes of temperature:' Studies of grain growth and diffusion normally imply consideration of a single-phase system, hence, in the presence of insoluble impurities they can be expected to give ambiguous data." High purity iron is also in demand for use as chemical and spectro-chemical standards; for work in classifying the lines of the iron spectrum; for biological work in nutrition; and for work in nuclear physics. where the presence of some sensitive

Jan 1, 1954

By B. R. Hollebone

The chemical action of molten anhydrous pyridinium chloride (pyridine hydrochloride) on oxy salts and ores of some Group IV and V metals are discussed-in particular zirconium, hafnium, niobium (columbium) , and tantalum. Laboratory-scale experiments are described whose results suggest that this reaction might provide a practicable means of converting ores of these metals to the anhydrous metal chlorides. Experimental results are given which provide some insight into the nature of the reactions and some of the compounds which could be present at intermediate stages. The treatment of ores of the more refractory metals is difficult and expensive and often demands the use of gaseous chlorine, hydrogen fluoride, or other reagents which are, if nothing else, expensive and dangerous to use. This situation is, of course, due to the strong metal-oxygen affinity resulting from the high charge and the small size of the metal ion which combine to produce such high lattice energy in the metal oxide as to defeat standard reduction methods. Thus, the characteristic of these metals which leads to one of their most important properties—corrosion resistance—is the main hurdle in obtaining the metal. Many common methods' for treating ores of zirconium and niobium (columbium), for example, proceed as directly as possible to the preparation of anhydrous halides. These halides are purified by now standard processes of solvent extraction, fractional distillation, and so on. This paper reports on preliminary stages in the development of a method of treating these ores—particularly of niobium—so as to prepare the anhydrous, volatile halides by the use of chemical reagents which are much more easily handled than those used at present. CHEMICAL PRINCIPLES The solvent action of aminium halides has been referred to periodically in the literature over several decades. The chemistry of pyridinium chloride has been discussed particularly by Audreith2 and Starke.3 However, since this is not familiar chemistry in its nonaqueous setting, it is perhaps well to point out some general metallurgical principles and applications. One can use, as point of departure, the fact that HC1 dissolved in pyridine is an acid very much the same as aqueous HC1. It thus gives essentially all the reactions of the HC1 with which we are familiar. However, when the ratio of HC1:pyridine reaches 1:l (a ratio far higher than is possible with H2O) the solution becomes solid C5F5NHC1, pyridinium chloride.* Inter- estingly, the acidic properties of HC1 persist in this medium, although they are manifest only in the molten state of the compound. Some typical reactions of divalent metal compounds which we have carried out in our laboratories are the following: 2PyHCl + Zn — (ZnCl2) + H2 + 2Py 2PyHCl + MnO — (MnCl2) + H2O + 2Py 2PyHCl + CuS — (CuCl2) + H2S + 2Py 2PyHCl + CaCO3 — (CaCl2) + H2O + CO2 + 2Py In these equations, the parentheses indicate a solvated compound which in water would be predominantly a hydra ted cation but which in molten PyHCl will more likely be a complex chloro-anion of the type MC14-. Some important physical properties of pyridinium chloride (PyHCl) are given in Table I. One sees from the data in Table I that at the temperature of molten pyridinium chloride both water and pyridine will boil away. Thus the system will be kept anhydrous. For this reason the reaction 2NbCl5 + 5H2O— Nb2O5 + 10HC1 has no counterpart in molten PyHCl. Instead, the chloride is made even more stable by the formation of a hexachloro complex: Nb2O5 + 12PyHC1 — 2PyHNbC16 + 5H2O + 10Py One interesting property of these complexes is their thermal instability which for the case of niobium can be illustrated as PyHNbCl6 — PyHCl + NbCl5 With a view to possible use of this chemistry in ore treatment, we have attempted to dissolve Pyrochlore (essentially FeO - Nb2O5) and recover the NbC15 from it.

Jan 1, 1968

By J. M. Blank, V. A. Russell

When the nucleation of silicon on a sapphire substrate is accomplished by gradually decreasing the substrate temperature while subjecting it to a constant impingement rate of hydrogen and silicon tetrachloride, the resulting deposit is characterized by widely separated islands of silicon scattered over an "open sea" of sapphire. Analysis shows that nucleation takes place mainly on sites of large adsorption energy, about 85 kcal per mole, and small surface concentration, about 108 cm-2. Crystal growth proceeds rapidly on these nuclei and results in enough depletion of Sicl4 from the ambient gas to suppress further nucleation. The technological importance of single-crystal films on dielectric substrates has prompted considerable work on the deposition of silicon on sapphire. This paper is concerned with the deposition of silicon, by the reduction of silicon tetrachloride with hydrogen, onto heated sapphire substrates. The apparatus consisted of a temperature-controlled container for the silicon tetrachloride along with gas lines and flow meters necessary to convey a hydrogen-silicon tetrachloride mixture to a reaction chamber constructed of quartz tubing. Heat for the substrate was provided by means of an inductively heated graphite susceptor. The observations pertinent to the present discussion were made with an American Optical Co. high-temperature microscope which allowed us to make motion pictures of the deposition process at a magnification of approximately 50 times. Two main categories of experiments were carried out: one on determination of critical condensation temperatures and the other on the deposition of epitaxial silicon films on the sapphire substrates. Critical condensation temperatures were determined in the traditional way by raising the substrate temperature well above that at which condensation could be expected at the impingement rates being studied and then gradually lowering the temperature until the condensation of the first silicon was observed. This was done initially with ellipsometry-type detection but it was soon discovered that the deposits were occurring as widely separated islands and groups of islands scattered over an "open sea" of sapphire. We have called these archipelago formations. They were most easily detected by optical microscopy. Data gathered from experiments of this type have been collected in Table I. Motion pictures of the formation of archipelago patterns show that the nucleation period lasts about 10 sec if the substrate temperature and impingement rate are kept constant. Essentially all of the nuclei are formed in a few seconds and all subsequent silicon deposition merely adds to the size of the established nuclei. After an archipelago formation has matured at a given substrate temperature and impingement rate, it is very difficult to stimulate any further nucleation in the open space between islands. Lowering the substrate temperature will occasionally stimulate nucleation but increasing the impingement rate hardly ever produces any more new nuclei. It should also be emphasized that the nucleation rates appear to be very small, about 104 cm-2 sec-1. They will receive special attention in the analysis section. Another series of experiments was performed employing a systematic variation of substrate temperatures and impingement rates approximating those expected to produce epitaxy. Sapphire substrates with surfaces normal to the 2233 direction were used. In contrast to the procedure that produces archipelago patterns, these films were deposited by raising the substrate temperature to a desired value at which time the silicon tetrachloride and hydrogen flow was begun. Of course, these depositions produced mainly films that were continuous. After removal from the deposition chamber, the substrate and silicon film were examined microscopically to determine the nature of the deposit and by X-ray diffraction to determine the extent to which a single-crystal film had been achieved. Results of these experiments are summarized in Table 11. Of interest for the present analysis is the fact that epitaxy appears to be confined to a fairly narrow temperature range between about 1050" and 1150°C. This also will be discussed in the analysis section. ANALYSIS In this section we shall apply nucleation theory to the data presented above with the objective of testing

Jan 1, 1967

By W. B. Hickman, J. J. Pickell, B. F. Swanson

Many physical properties of the porous media-immiscible liquid system are dependent upon the distribution of fluids within the pores; this in turn, is primarily a function of pore structure, liquid-liquiri interfacial tcnsion and liquid-solid wetting conditions. The cnpillary pressure hysteresis process provides a means of investigating the influence of pore structure upon fluid distribution for consistent sur/acc conditions. Invcstigntiot2s indicate that residual non-wetting-phase saturations /ollozuir,g the imbibition process (i.e.,wetting phase displacing nor2-wctting phase) are dependent upon both pore structure and initial non-raetting phase saturation and suggest that residual fluid is distributed as Discontinuous ,globules, one to a few pore sizes in dimension, thrnugh the entire range 01 pore sizes originally occupied. It clppears that air-mercury capillary pressure dala adequntely rellect the distribution of fluiris in a water-oil system when strong wetting condition preliczil. An oil-air counter-current imbibition technique has also been found to provide a rapid rneans of obtaining residual-initial saturation data. In a majority of cnses, rcsidual saturations detf-rniinecl from the oil-air or air-mercury process reasonably approximate residual oil saturation following water drive of a strongly water-wet medium. INTRODUCTION A reliable estimate of recoverable reserves depends not only on the amount of original oil-in-place but also on pore geometry and distribution of fluids within the pores. A critical parameter determining the recovery from a reservoir under waterflood, for example, is the amount and distribution of residual oil within the various rock types present. The purpose of this paper is to investigate the mechanism of capillary trapping and assess its importance in laboratory measureqents of residual oil saturation. The degree of wettability of a reservoir rock is recognized as an important factor in waterflood or imbibition experiments. In this paper, however, only the water-wet case has been considered. Considerable experimental evidence1 suggests that for water-wet rocks, capillary forces predominate in tile distribution of fluids and that viscous forces in the range normallv of interest in the reservoir have a minimum influence on residual oil saturation. It follows that if the ultimate recovery is controlled by pore geometry, a unique residual non-wetting phase saturation should exist for a given set of initial conditions. Two laboratory procedures.found to be extremely useful in the study of pore structure and degree of fluid interconnection at various saturations are decribed. Although air-nercury capillary injection curves have been used2 previously to characterize the drainage case, the withdrawal or imbibition case can provide valuable supplementary data. The air-mercury process, however, has several disadvantages; it is difficult to run in a sufficiently accurate manner, mercury does not always act as a strongly non-wetting liquid and in the air-mercury process the sample is rendered unsuitable for future analyses. An alternative process is described in which air is the non-wetting phase and naptha, hentane, octane or toluene is the wetting phase. INTERFACIAL TENSION AND CAPILLARY PRESSURE Interfacial tension between immiscible fluids is due to the difference in attraction of like molecules as compared with their attraction to molecules of the neighbouring fluid. This net attraction results in a tension at the interface. To extend the interface; thus, interfacial tension s can also be thought of as free surface energy. Interfacial tension is normally expressed as dynes/cm, and interfacial energy is measured in ergs/cm2 hence, both have dimensions ml A 2 and are numerically equal.

Jan 1, 1967

By J. P. Boisseau, C. N. J. Wagner, E. N. Aqua

An analysis ulas made of powder-paltern peaks from cold-rolled polycrystalline copper and from copper powders, compacted into briquets (1 in. diameter) with pressures up to about 106 psi. Powder-pattern peaks from both contact faces were measured with a diffractometer and CuKa radiation. Laltice-parameter measurements indicate the presence of residual stresses in the bulk specimen of the order of a - -10 kg per sq mm after 93 pct reduction in thickness, and give some evidence for the occurrence of deformation stacking faults (probability a - 0.003) in both bulk and powder samples. Fourier analyses of the line profiles were carried out on the (111) - (222) and (200) - (400) Pairs of reflections. The particle sizes decrease rapidly with increasing deformation and reach asymptotic values of De (111) = 550A and D, (200) = 350A at about 25 pct reduction in thickness or 80,000 psi pressure for compaction. The anisotropy of the particle sizes indicates the presence of microtwins and stacking faults on (111) planes. The rms strains increase also with deformation and reach limiting values of 2 x 10-3 in the [100] direction and 1 x 10-3 in the [Ill] direction at a distance L = 100A normal to the reflecting planes. THE diffraction of X-rays in cold-worked, polycrystalline, fcc metals and alloys yields Debye-Scherrer patterns in which the individual (hkl) lines may show a broadening, a shift, and an asymmetric shape.' These effects can be explained by the fragmentation of the small, coherently diffracting domains, the presence of residual elastic strains in the lattice, and the formation of stacking faults. X-ray studies of stacking faults in fcc metals and alloys have been methodically performed during the last few years, but so far most of the investigators dealt with cold-worked powder specimens obtained by filing.2-8,10-19 In such specimens, the crystals are randomly oriented and the average elastic strains are assumed to be zero, so that the observed peak shifts are due to stacking faults only. The agreement between X-ray diffraction theory1,20 and experiment is generally good. On the contrary, cold-worked, polycrystalline bulk specimens (after wire drawing, rolling, plastic deformation by tension or compression, etc.) contain residual elastic strains, which may not average to zero in the surface of the specimen and show very strong preferred orientation of certain lattice planes (texture) parallel to the specimen surface. Nonzero elastic strains also produce peak shifts, the magnitudes of which depend upon the crystallographic orientation of the reflecting lattice planes in elastically anisotropic materials as well as on the type of plastic deformation and which have to be taken into account when measuring stacking-fault probabilities in deformed bulk specimens.21-23 The purpose of the present investigations was to perform a detailed X-ray analysis on a bulk specimen of a fcc metal deformed by rolling at room temperature and on powder of the same material deformed by compaction, and to compare these results with those previously obtained on cold-worked filings.18 Pure copper was chosen for this study since it is known to produce only a small quantity of stacking faults by filing at room temperature.3,16,17,19 The residual elastic strain in the surface of the bulk copper should be the predominant factor determining

Jan 1, 1965

By G. J. W. Kor, F. D. Richardson

The sulfide capacities a1 1500°C of slags in the systern CaO + CaF, +Al2O3 have been measured by bringing samples into equilibrium with gas mixtures made from CO + CO2 + SO, + N2. The capacities are negligzble both JOY CaF, and CaF, + Al2O3 melts but when lime is present even at 20 wt pct capacilies as high as 30 x lom3 can be obtained. The results have yielded information concerning the liquidus at 1500°C and fluoride-oxide melts in general. The slags used in electroslag refining contain substantial amounts of calcium fluoride. Robinson and Grainger' report that good results can be obtained by using slags of the approximate composition 70 wt pct CaF2 + 30 pct Al2O3. It has also been reported2 that a mixture of 50 pct CaF2 + 30 pct A12O3 + 20 pct CaO makes a good slag for general use. In view of the fact that very low sulfur contents are claimed for ingots produced by the electroslag process. studies have been made of the sulfide capacities [c, = wt pct S (PO2/Ps2)1/2 at 1500°C. of slags consisting of pure calcium fluoride, and of its mixtures with lime and/or alumina. These studies have also provided information on the phase diagrams. The slags were brought into equilibrium with gas mixtures prepared from NZ + CO + CO2 + SO2. Because CaFz is volatile and also easily hydrolyzed. measurements were made of its loss in weight in various flowing gas mixtures. EXPERIMENTAL Procedure. The apparatus and procedure were the same as described previously. Both platinum and iridium crucibles were used (height and inner diameter about 1 cm), but those made from iridium were stronger and able to withstand a greater number of experiments. In each experiment two crucibles were set in an alumina support and 1 g of slag mixture was divided between them. The materials from each crucible were analyzed separately after being brought to equilibrium with the gas, and then quenched in the water-cooled end of the reaction tube. Materials. Alumina and lime were prepared as de-scribed previously.4 Calcium fluoride was obtained by igniting Analar calcium fluoride in a platinum boat at about 800°C in a stream of pure dry nitrogen: the product was cooled in a desiccator. The slags were produced from mixtures of the powdered constituents, these being melted together in the course of each equilibrium experiment. The sulfur analyses and the X-ray diffraction meas- urements on the solidified slags were made as described previously.4 From the latter it appears that the CaF2 contained less than 0.5 mol pct CaO. RESULTS Values of (ps2/po2) for the Gas Mixtures. The re-sults have been represented in terms of (ps, /Po,)"? for the various gas mixtures used. The va1;es of this function at 1500°C are listed in Table I: they were calculated from the data cited previously for calculations at 1350°C 4 Weight Losses with CaF, and CaF, + CaO. The re-sults of the weight loss measurements* at 1500°C for

Jan 1, 1970

By R. G. Gifkins

Mean boundary migration depended linearly upon creep extension under various conditions. Prolonged annealing or air-casting increased the proportion of immobile boundaries at a given stage of tests. Grain growth during annealing followed a power law. The linking of migration and sliding in creep and models for migration are examined; activation energies are calculated from these models. Segregation of oxygen to boundaries appears to be a controlling factor. THE idea that boundary migration and boundary sliding during creep are linked has often been suggested,'-5 especially when evidence has been obtained for an alternation between the two processes. In simple bicrystals sliding can occur without any migration6 but in more complex bicrystals7 it is observed. McLean' and Couling and Roberts 5 made measurements of both sliding and migration; McLean concluded that there was no linking of the two processes whereas Couling and Roberts supported the idea. McLean' also developed a model for migration during creep and annealing. The present work was initiated as part of a study of the recrystallization of lead during creep9 but has been extended to examine some aspects of the mechanism of migration. MATERIALS AND METHODS Specially refined high-purity lead* was used; the total of impurities, excluding oxygen, was 0.0008 wt pct. Oxygen was removed as far as possible by melting in vacuo, treating at 550" to 600°C with hydrogen and then casting at 450°C under a vacuum of 10-2 mm Hg; the oxygen was then believed to be 0.00005 wt pct. Specimens were made from extruded strip and then annealed at the required temperature in an air oven, controlled to 5 lac. They were then polished in a mixture of acetic acid and hydrogen peroxide. At this stage some specimens were etched to give grain contrast and their grain sizes determined by the intercept method. Migration of boundaries during creep was measured on the surface of polished (not etched) specimens, directly loaded at room temperature (22°C) or 50°C (52"C) to initial stresses of 300 to 700 1b per sq in. Twenty six vacuum-cast hydrogen-reduced specimens were used and, in addition, two specimens were from material air cast after hydrogen reduction ("partial" air cast) and two from air-cast material not treated with hydrogen ("full" air cast). Duplicates were made in the majority of tests. At intervals through the creep tests, specimens were unloaded and the migration measured, normal to the direction of migration at the point of measurement, for each boundary along standard traverses. Usually between 110 to 180 boundaries were measured at each of some twelve stages of the tests. Migration was measured on the projected image (X100) using prism illumination. Fig. 1 illustrates the method of measurement and shows the kind of variation of migration found from boundary to boundary and along particular boundaries. Fig. 2 gives an example of more extensive migration. The mean value of distance migrated was generally reproducible to *10 pct but occasional points gave a scatter of *20 pct. The correctibn applied by McLean8 to allow for the contribution sliding makes to the apparent distance of migration could not be used, for the sliding at most boundaries was too small to measure with the equipment available. There were two reasons for supposing the correction unnecessary in the present work;

Jan 1, 1960

By J. E. 214-000-000-010 Warren, M. R. J. Wyllie, T. Meidav, L. Scharon, R. Uhley, A. J. deWitte

By considering the stoichiometry of the underground combustion process, an equrztion was derived relating the point velocity of the combustion front as a function of the air flux, fuel content, efficiency' of oxygen uitilization, hydrogen:carbon ratio, and ratio of carbon dioxide In carbon monoxide produced. In addition, an equation for cnlcrllating the amount of injected air necessary 10 produce the oil from I Cu. ft of reservoir space was derived. This equation has a direct bearing on evaluating compression costs for the process. Values calculated from these equations are in good agreement with experimentnlly obscrvcd values obtained from laborator) tube runs reported in the literature. The equations are also u\eful for atlcrlyzinf: experinlental data for inconsistencie.r.. I NTRODUCTION The injection of energy in the form of heat into an oil-bearing formation as a means of increasing the recovery of oil has been the subject of much discussion and experimentation as far back as 1920. The ideas have ranged from the injection of steam to the direct use of electricity as a means of getting thermal energy into a formation. However, most laboratory and field studies have been confined to the process whereby a portion of the crude in the reservoir is burned to generate the necessary heat to increase the mobility of the oil by reducing its viscosity. A recent paper by W. L. Martin, et al.', described the results of a number of laboratory experiments in which the combustion process was studied in a sandpack. Studies such as this indicate that with the temperatures involved, the oil is partially coked in this process and it is this coke which provides the fuel for combustion. The combustion of this residual coke can be analyzed theoretically by applying simple stoich-iometric principles to the combustion process. It is the purpose of this paper to compare the results reported in these experiments with those obtained from theoretical considerations. The variables studied by Martin, et al, among others, include combustion front velocity, air flux, oxygen utilization, fuel concentration, ratio of carbon dioxide and carbon monoxide in combustion gases, and the hydrogen:carbon ratio of the residual fuel. All these variables come into consideration when the stoichiometry of the combustion process is studied. The experimental conditions and results are tabulated in Tables 3 and 4 of Ref. 1. For purposes of comparison with calculated results, it was necessary to change the units of a number of the variables. UNDERGROUND COMBUSTION STOICHIOMETRY The combustion or oxidation of the carbonaceous residue or "coke" on the sand can be described in terms of the equation: where n is atomic ratio of hydrogen to carbon, m is ratio of moles of CO2 to CO produced, and CH. is taken as the unit molecular weight, realizing that actually the coke has a large range of high molecular weights. Assuming that all the fuel laid down in the sand is burned, the ratio of moles of oxygen consumed to moles of coke can be written as follows. Moles 0MolesMolescoke In terms of moles O2 per pound of coke, where 12 + n is the molecular weight in pounds per mole. If the 0, consumed is measured in standard cubic feet (60°F and 1 atm) and the coke in pounds per cubic foot of formation, 2, then:

By M. A. Torcaso, P. Raimondi

To study mass transport in systems simulating oil recovery processes, different porous media were saturated with a mobile (carrier phase) and a stationary phase. Slugs of carrier phase containing a small amount of solute were displaced with pure carrier phase. By analogy to the chromato-graphic processes, the velocity of the solute can be predicted from a knowledge of the partition coefficient and the saturation provided that equilibrium between the two phases exists. Equilibrium was found to exist for different porous media, solutes and rates. The conditions were varied over the range normally encountered in the laboratory and in the field. The longitudinal dispersion of a solute undergoing interphase mass transfer was also investigated. INTRODUCTION The production of hydrocarbons by gas cycling, enriched gas drive and C02 or alcohol displacement involves, among other factors, relative motion between two phases and compounds, hereafter called solute, which are soluble in both phases. The solute is carried forward by the faster flowing phase at a lower velocity than the average velocity of that phase. Retardation of the solute is caused by chromatographic absorption and desorption in the slower flowing phase and by the degree of departure from equilibrium. At equilibrium the concentration of solute in the two phases can be related by the equation* CSw = K Cso ,............(1) where C, and C are the concentration of solute in the aqueous and oleic phases respectively and K is the equilibrium ratio, or partition coefficient. Displacement theories must contain an explicit assumption with regard to equilibrium, i.e., whether the compositions can be related by Eq. 1. The existance of equilibrium depends, in general, on the relative velocity between the phases. Unfortunately, other factors such as gravity segregation and viscous fingering, also depend on velocity. For this reason, whenever effects of rate on displacement were observed,2-4 it was practically impossible to discern what caused them — lack of equilibrium or the factors mentioned above. Equilibrium between phases has been the subject of extensive studies in fields such as extraction or chroma to graphy. It has received only small attention in flow through the type of porous media encountered in oil production.5 For this reason a method was developed which makes it possible to study the movement of a solute as it is affected by rate, type of porous media, partition coefficient and carrier phase, but in the absence of segregation or fingering. The information obtained enables one to determine when the assumption of equilibrium can be made. Briefly, the method consists of (1) saturating the core with a mobile and an immobile phase, (2) injecting a slug made up of the same fluid as the mobile phase and a small concentration of mutually soluble solute, (3) measuring the lag and the peak height of the slug at arrival and (4) correlating these variables with fluid properties such as partition coefficient and mixing constants of the medium. PROPOSED MECHANISM The principles of chromatography are combined with the equation of longitudinal mixing to predict the velocity of a solute slug compared to the bulk velocity and the peak height of a slug.6-9 The equation so obtained is valid under equilibrium conditions only. Therefore, a comparison between experimental and predicted results will give a measure of departure from equilibrium. This work was done with either the oleic or the aqueous phase being immobile. For simplicity, the following development is based on the case where the oleic phase is immobile. However, the treatment is the same in either case.

Jan 1, 1966

By Dwight K. Smith, Greg Carter

Studies have been conducted on The properties of many deep well cementing compositions to determine their Strength behavior over curing periods to 180 days at elevated tem-peratures and 3,000 psi pressure. This pressure results in essentially the .vat, compressive strength as that obtained with much higher pressures in ac-cordance with the findings of an API subcommittee ' These different compositions include API C1asses A, E, and F cements containig additives such as retarders, berrtorlite, heavy weight materials and pozzolans. All compositions except mixtures of Poz-zolrrrts and hydrated lime show from mild to severe degrees of strength lass over a temperature rang from -730 to 320" F after setting. Pozzolatl-time compositions actually gain in strength with time at these high temperatures. Other chemical find physical properties evaluated to observe their inter-relation.ship with Strength loss show very little correlation except in per-meability. As retrogression in strength increases, the permeability of set cement increases. X-ray difraction patterns on the set products indicate the formation of certain compounds having little or no strength value in those cementing compositions where strength loss was more severe. INTRODUCTION As a result of the growing trend toward deep well completions, a study has been made on the properties of the various cementing slurries presently in use at elevated temperatures and pressure. The strength of these compositions is of prime importance in selecting the most suitable material for use at high temperatures, to determine WOC time and the proper time to perforate with a minimum of shattering. Of primary interest in this investigation was the strength behavior of various compositions currently in use, and the effect of additives on cement after long periods of curing under severe conditions. Earlier investigators have pointed out that the strength of some cements will increase with increasing curing temperature to about 220 to 340" F, but at higher temperatures a loss in strength occurs at extended time intervals. Studies by the API Mid-Continent District Study Committee on Oil Well Cements' outlined a testing procedure whereby field conditions of temperature and pressure could be simulated in the laboratory.' It was observed that retarded cements undergo changes at elevated temperatures and some lose as much as 50 per cent of their early strength when cured at high temperatures. The scope of these tests was limited to curing periods of 1 to 28 days; additives were not covered in this study. Saunders and Walker' found that the use of additives influenced the strength behavior of cements at high temperatures, but studies were conducted only to a seven-day curing period. Ludwig and Pence' conducted an investigation of two types of cements under elevated pressures and temperatures in an effort to explain the causes of this loss in strength phenomenon. It was observed that with retrogression in strength there were other physical and chemical changes which took place as the cement aged. Today there are a number of cementing compositions available for use at well depths below 10,000 ft. Since all these earlier investigations were limited in curing time as well as the number and type of compositions studied, it was believed that a more extensive study should be conducted on the properties of cements at elevated temperatures, and this work should include those compositions actually in general field use at this time.

By N. J. Grant, H. Brunner

This paber is concerned with the determination of equations relating elongation to the amount of shear taking place both along grain boundaries and in slip planes of poly crystalline aggregates during creep. Certain discrepancies with previously used equations are pointed out. Grain boundaries were observed to become serrated during creep, but it is shown that this does not necessarily prevent further grain boundary sliding. Direct obser-vations of subgrain boundary ingration lead to the conclusion that this migration facilitates sliding along serrated grain boundaries. It is bointed out that the state of stress at triple boints is still unknown. MOST of the mechanisms of deformation contributing to elongation are shear processes, such as slip, kinking, grain boundary sliding, and twinning. For quantitative investigations of individual deformation mechanism it is necessary to know the equation which permits the calculation of speci-tnen elongation from measured amounts of shear. This calculated elongation must then be compared with the observed total elongation. The authors' special interest concerned grain boundary sliding during creep. Although this subject has already been investigated quantitatively by McLean,' Rachinger,' and Fazan, Sherby, and Dorn, calculations indicated that none of the equations used thus far to relate shear with elongation were sufficiently well founded. A basic study of this relationship was thus made and the results of this study (which do not entirely agree with those previously used) are presented. Most of the ideas set forth apply not only to grain boundary sliding but to other mechanisms of shear as well. THE SPEAR-ELONGATION RELATIONSHIP Basic Models—The relationship between shear and elongation is trivial in the simplest case of shear, which is shown in Fig. l(a). The vector A; designates the displacement across the shear interface, and Au, and Aut are its components parallel and normal to the tensile direction, respectively. It is evident that the specimen elongation due to shear is equal to the longitudinal component of the shear vector, and it should be noted that the normal component does not contribute to the elongation of the specimen (there are, of course, two normal components in a three-dimensional model). The model shown in Fig. l(b) may be applied to mechanical twinning and to subgrain boundary tnigration. In this case the width (h) of an already existing shear band increases by the migration of an interface, but the shear angle y (angle of mis-orientation in the case of a subgrain boundary) remains constant. The geometrical relationships indicated in Fig. l(b) permit this case to be reduced to the one shown in Fig. l(a). For the resulting elongation A6 we find: where AV/V = fraction of the volume of the specimen swept by migrating interfaces , and 1 = length of specimen. Thus, assuming for example that the angle of mis-orientation is 1 deg, and that P = 45 deg, we find that the specimen elongates 0.87 pct of its length if the entire volume of the specimen is swept once by such a subgrain boundary. An example of deformation by migration of subgrain boundaries is shown in Fig. 2. The specimen shown was repolished after 9.8 pct creep strain and the test then continued for another \ pct. Since the shear vector has, in general, a component normal to the surface, subgrain boundary migration results in the formation of an inclined band delineating the region swept by the subgrain boundary. These inclined bands become clearly visible when observed with oblique illumina-

Jan 1, 1960

By S. J. Burns, W. W. Webb

The dislocation arrangements formed during unsteady propagation of cleavage fractures on (010) planes in LiF have been investigated by high-resolution etch-pit techniques and by X-ray diffraction topogvaphy. In impurity-hardened crystals, cracks propagating in any direction on (010) planes at velocities of about 103 cm per sec are usually accompanied by light plastic deformation which is enhanced wherever the crack is allowed to hesitate. Dislocations with [011] type Burger's vectors are often formed looped around the crack tip. They are dragged along with the crack tips leaving edge dislocations lying nearly parallel with the clearage surface about 5µ from it. These dislocations are easily overlooked since they may initially intersect with the cleavage surface infrequently and later may disappear from the crystals due to image forces during aging at room temperature. Although the plastic work associated with these dislocations is only comparable with the reversible (010) surface energy, an instability in crack propagation appears due to a limit of the velocity with which dislocations can be dragged along with the crack. Transition to a fully brittle nonplastic fracture mode occurs when the crack velocity exceeds the velocity with which dislocations can be moved by the stress field of the crack. TRANSITIONS between brittle and ductile modes of fracture appear to be common processes in the behavior of materials in which the flow stress is close to the stress at which brittle cracks can propagate. High-strength engineering alloys fit into this category and the transition from ductile to brittle modes of crack propagation may be associated with catastrophic failures in these important materials. The methods of theoretical mechanics indicate the geometrical conditions in which these processes may occur but the detailed mechanisms controlling the generation of dislocations in the initiation of the plastic mode and the process by which the loss of plasticity occurs on transition to the brittle mode are not yet clear. Since these mechanisms are not amenable to direct observation in engineering alloys we have chosen a semi-ductile alkali halide LiF as a model system in which to observe transition mechanisms between plastic and brittle modes of crack propagation. Cleavage cracks in this system are known to undergo a plastic-elastic transition at velocities between 103 to 104 cm per secl and dislocation configurations in LiF are amenable to study by optical-microscopic observation of dislocation etch pits2 and by X-ray diffraction topography.3 We have been able to identify a dislocation mechanism responsible for the transition from brittle to slightly plastic modes of crack propagation in LiF. A model based on this mechanism permits calculation of the critical crack-propagation velocity for this transition and indicates the importance of the stress dependence of dislocation velocity in determining the critical crack velocity. Of course, this model cannot be carried directly over to more complex materials but it is quite possible that its qualitative features are generally applicable. Dislocations intersecting fresh cleavage surfaces in LiF were reported by Gilman,4 by Forty,5 and by Wash-burn et a1.6 Such dislocations were construed to have been nucleated where the cleavage crack hesitated or stopped, near stress concentrators, such as cleavage steps, on the crystal face. The loci of such dislocations deposited On the cleavage plane in bands were called "stop lines" or "deformation zones". Gilman, Knudsen, and walshl measured the velocity of cleavage cracks in lithium fluoride by evaporating thin resistive strips on the crystal side and recording

Jan 1, 1967