Search Documents

By Charles S. Barrett

THIS paper is a progress report covering metallographic applications of the electron microscope that have been made during the past year at Carnegie Institute of Technology. An account is presented of the techniques that have been most satisfactory, the difficulties encountered, the resolution obtained, and the peculiarities that have been noted in the microstructures. About 12m exposures have been made during the year on the instrument, which is an RCA type B2 transmission microscope,1 and a number of these have been selected to illustrate the results obtained. Prints are also included showing the common types of falsities-to borrow a term from the biological field, the "artifacts." In brief, the paper is a record of our attempt to determine the value of the electron microscope in the study of the structure of metals and alloys. n addition to comments on techniques and the quality of photomicrographs obtainable with them, the paper discusses the nature of the etch attack on a few metals, and the possible relation between the fine structure revealed by etching and the mosaic block structure of the grains. Since numerous papers and reviews are available on the subject of imperfection in crystals, the comments here may be very brief; in fact, are restricted almost entirely to points having to do with recent theories relating the strength of metals to imperfections and to spacings of slip lines. SPECIMEN PREPARATION AND PHOTOGRAPHS Mechanical polishing by hand was used throughout this work. Precautions that must be taken in microscopy at high magnification, which have been well explained by Vilella in his book on Metallographic Technique for Steel, are likewise important with the electron microscope. This applies particularly to the need for repeated polishing and etching (each successive polish becoming less severe and each etch lighter), since the final etch should in general be no heavier than is used for oil-immersion optical microscopy and in many instances should be much lighter (compare Figs. 6a and 7a, 6d and 7b). An extremely light etch, of course, is incapable of cutting through a flowed layer of appreciable thickness. Unquestionably it would be very advantageous in each instance to compare the structures obtained by electrolytic polishing with those obtained by mechanical polishing, but difficulties with the electrolytic method have prevented this in the following collection of prints. The choice of etchant is important, of course; for example, picral proved superior to nital for tempered martensite. With the exception of Figs. Ib and IC, the replicas used were of silica, deposited up0n polystyrene and floated from the polystyrene in ethyl bromide according to polystyrene in ethyl bromide according to the method of Heidenreich and Peck.2 The samples were mounted in Bakelite, polished and etched, then the polished Bakelite

Jan 1, 1943

By A. M. Sarem

For the performance prediction of multiphase oil recovery processes such as steam stimulation, there is an acute need for three-phase relative permeability data. No fast and simple experimental technique, such as the unsteady state method proposed by Welge for two-phase flow, is available for the three-phase flow. In this paper, an unsteady - state method is presented for obtaining three-phase relative permeability data; this method is as fast and easy as Welge's method for two - phase flow. Analytical expressions are derived by extension of the Buckley-Leverett theory to three-phase flow to express the saturation at the outflow face for all three phases in terms of the known parameters. It is assumed that the fractional flow and relative permeability of each phase are a function of the saturation of that phase. Other simplifying assumptions made include the neglect of capillary and gravity effects. The effect of saturation history upon relative permeability is acknowledged and attainment of similar saturation history in laboratory and field is recommended. The required experimental work and computations are simple to perform. The test core is presaturated with oil and water, then subjected to gas drive. During the test, required data are the rates of oil, water, and gas production, together with pressure drop and temperature. The ordinary gas-oil unsteady-state relative permeability apparatus can be readily modified to measure the required data. The proposed technique was applied to samples of a Berea and a reservoir core. The effect of saturation history upon relative permeability was studied on one Berea core. It was found that increase in initial water saturation has a similar effect upon three-pbase relative permeability as it does in two-phase flow. INTRODUCTION In the light of increasing demand for three-phase relative permeability data for predicting the performance of thermal and other multiphase-flow recovery processes, a simple and accurate method of experimental determination of such data is extremely desirable. Leverett and Lewis 1 described the simultaneous flow method of obtaining three-phase relative permeability data. However, Caudle et al.2 reported that this method is very time consuming and cumbersome. Corey3 proposed calculating the three-phase relative permeability from measured krg data. Corey's theory is based on simplified capillary pressure curves,4 assuming a straight line relationship between I/Pc2 and saturation. Also, Corey's method assumes a preferentially water-wet system. The simplest and quickest method of obtaining three-phase relative permeability data is the unsteady-state method where, for instance, oil and water are displaced by gas. However, in such a test the correlation of average saturation with relative permeability does not give a valid relationship because the rates of oil, water and gas flow in the sample change continuously from the upstream to downstream end. This difficulty in calculating valid relationships was solved by Welge for two-phase flow by deriving an expression for determining the outflow face saturation from Buckley and Leverett frontal advance equations.5,6 In this paper, relations are established to determine the outflow face saturation and relative permeability to all phases in a three-phase flow displacement experiment. PROPOSED METHOD The fundamentals established by Buckley and Leverett5 for two-phase flow were extended to three - phase flow and used as a basis for the derivation of saturation equations. This approach is comparable to Welge's6 use of Buckley and Leverett theory in arriving at expressions to determine the outflow face saturation of the displacing fluid in a two-phase flow system.

Jan 1, 1967

By J. Chipman, N. J. Grant, J. H. Walsh, T. B. King

WITH the development of adequate sampling and analysis techniques, much information has been obtained concerning the behavior of hydrogen in the steel bath during the course of steelmak-ing operations. From studies of this type it has become clear that the slag has an important role to play in determining the hydrogen content found in commercial products. A knowledge of the behavior of hydrogen in slags is necessary to round out our knowledge of hydrogen in steelmaking practice; with this information we are better able to assess practical methods of control. In this paper a method developed for analyzing slags for hydrogen content, studies of the solubility of hydrogen in synthetic slags, and studies of the variation of hydrogen content in slag and steel in industrial heats will be described. Previous Work—While very little has been done in the past to determine the hydrogen content of metallurgical slags, measurements have been made on such slag-like materials as geologic magmas and glasses. Goranson as made extensive studies of the solubility of water in natural and synthetic granites. Many workers have investigated the solubility of gases in synthetic and commercial glasses, but the recent work of Russell" is the most extensive. Russell found that the solubility of water in glass was proportional to the square root of the water vapor pressure in water vapor-oxygen mixtures. It was found that the solubility could either increase or decrease with temperature depending on the glass composition. Wentrup, Fucke, and Rief,' and Piper, Hagedorn, and Backes' have extracted hydrogen by holding a slag at 1300" and 800°C respectively in vacuum. Wentrup et al. reported hydrogen and water vapor as evolved together while Piper et al. passed the evolved gases from the hot extraction treatment over hot ferromanganese to reduce the evolved water vapor to hydrogen and, therefore, reported the analyses in terms of total dissolved hydrogen. Yavoiskii has reported hydrogen and nitrogen analyses on steelmaking slags. Dobrokhotov, Povol-skii, and Khan have studied the hydrogen content of basic open hearth steel under slags of varying viscosity. These authors reported that the hydrogen content of the steel bath increased as the fluidity of the slag was increased by the addition of bauxite. They recommended that a viscous slag be used and the steel tapped without deoxidation in the furnace to ensure a low final hydrogen content. Analysis of Slags for Hydrogen Development of Apparatus—The vacuum fusion analysis equipment described by Shields, Chipman, and Grant" was modified for the analysis of slags." This method depends upon the reduction of water in the slag to hydrogen, which is readily determined by the method used for hydrogen in steel. Preliminary experiments were made with a eu-tectic ferrosilicon melt as the reducing agent. Other experiments included the use of graphite and molybdenum crucibles. The ferrosilicon alloy bath was found to produce a thick coating on the furnace tube which could adsorb hydrogen. Graphite produced CO by the reduction of FeO, MnO, and SiO2 in such quantities that analysis for hydrogen was virtually impossible. Molybdenum, while giving a low blank value, could reduce the slag sufficiently to produce a volatile oxide which again was found to adsorb hydrogen. The arrangement finally adopted is shown in Fig. 1. The molybdenum sheet served as the induction-heated susceptor, and was separated from the escaping gases by alumina guide tubes. An outer jacket of alumina protected the quartz furnace tube, and alumina and quartz stools completed the assembly. The falling sample is shown wrapped in aluminum foil, 0.002 in. thickness, which acted as the reducing agent' This foil was careful'~ washed in acetone and an d and then dried in hot air before wrapping the weighed slag sample for charging into the apparatus. Investigation of the hydrogen content of the aluminum foil itself showed that a small quantity of hydrogen was released in vacuum, but had no relation to the weight of the aluminum charged. This quantity of hydrogen was

Jan 1, 1957

By N. A. D. Parlee

The diffusion rate of hydrogen in liquid iron has been measured by a gas-liquid metal diffusion cell technique. The diffusion cell was formed by immersing an alumina tube containing hydrogen gas at 1 atm in a bath of stagnant liquid iron. The change in the composition of the melt in the cell was determined by measuring the rate of absorption of the gas in the cell. The appropriate solution to Fick's second law was used to examine the data and calculate diffusivi-ties. The absorption of hydrogen in stagnant pure liquid iron has been found to be diffusion-controlled. The results show that the chemical diffusion coefficient, D, of hydrogen in pure iron in the range of 1547" to 1726°C can be represented by the following Arrhenius relation: D(sq cnz per sec) = 3.2 x X exp(- 3300 i 1800/RT) where the uncertainty in the activation energy corresponds to the YO pct confidence level. Oxygen in the melt (above 0.015 pct 2) increased the apparent rate of absorption of hydrogen. The importance of diffusion data on liquid metals for predicting the rates of certain metallurgical processes has been recognized for a long time. Moreover, these data are much needed to test and develop theory for diffusion in liquid metals. Despite this practical and theoretical interest, however, relatively little reliable information about diffusion in liquid metals is available in the literature. This is particularly true for gas components such as hydrogen, oxygen, and nitrogen in liquid metals, where almost no data on chemical diffusion coefficients are to be found. This is probably due to a multitude of experimental difficulties particularly associated with high-temperature melts. In an effort to fill this gap in information, a research program was undertaken to study the diffusivities and rates of solution of gases in liquid metals. This paper presents the results of a study of the diffusion of hydrogen in liquid iron. EXPERIMENTAL METHOD Two methods for the study of the kinetics of dissolution of gases in liquid metals are being employed in this laboratory. Both involve the measurement of the volume of gas absorbed by the melt as a function of time and as such both avoid the uncertainties involved in chemical analyses of quenched samples for relatively small amounts of gas. In the first method, the gas dissolves in an inductively stirred melt and, in the absence of a slow surface reaction, the results are often interpreted in terms of mass transport across a liquid "boundary layer" between the homogeneous gas phase and well-stirred part of the melt. Other interpretations of the results of such experiments have also been described in the literature.1'5 In the second method a gas-liquid metal diffusion cell is used.' The gas dissolves in a cylindrical column of stagnant liquid metal and, in the absence of a slow surface reaction, the results are interpreted in terms of a non-steady-state diffusion solution to Fick's second law. The weakness of the first method is that while it gives information on the mechanism of absorption by stirred melts it yields an overall rate constant which even in the simplest cases depends on the nature and the thickness of the "mass transport layer" or "boundary layer". It yields no values of diffusion coefficients. The second method was used in this research because in many cases it is possible to determine the diffusion coefficient of the gas component in the liquid metal. In this research it has been utilized to measure diffusion coefficients of hydrogen in liquid iron. The apparatus used was essentially the same as that described by Mizikar, Grace, and par lee but certain modifications have been introduced to meet the elevated temperatures and special conditions of this research. Fig. 1 is a schematic drawing of the apparatus and Table I gives the identification of various parts in this figure. The diffusion cell, shown in detail in Fig. 2, was formed by immersing an impervious alumina tube (hereafter called absorption tube) in a bath of pure liquid iron contained in an alumina crucible. Two types of tubes were used, Morganite triangle RR and McDanel AP35. The crucible was contained in a vertical impervious alumina combustion tube (32 mm ID by 914 mm long) which was closed at both ends by water-cooled brass heads employing O-ring compression seals, Fig. 1. A protection tube enclosing a Pt, 5 pct Rh-Pt, 20 pct Rh thermocouple was introduced through the lower end of the combustion tube

Jan 1, 1968

By M. Mills, M. W. Clegg

This Paper Presents an approximate analytical solution to the problem of the pressure distributions arising from the production of a compressible liquid in a partially penetrating well. The limits of validity of the approximation are discussed and it is shown that the basic approximation is the same as that used in previous numerical of the problem. The variation in pressure at the production well with time for different penetration ratios is investigated and is shown to agree with previous work. The solutiOns are used to assess the effects of partial penetration at large distances from the well and at the wellbore itself. In the latter case the effects of anisotropy are also studied for constant, but unequal, horizontal and vertical permeabilities. It is 'uggested that measurements of the Pressure drawdown below the producing interval in a well should give some indication of a thickness-vertical permeability product for the formation; this would chmacterize the effectiveness of the formation below the producing interval. INTRODUCTION In many oil and gas reservoirs the producing wells are completed as partially penetrating wells; that is, only a portion of the pay zone is perforated. This may be done for a variety of reasons, but the most common one is to prevent or delay the intrusion of unwanted fluids into the wellbore. This paper studies the effects of partial penetration on the pressure distribution resulting from the production of a compressible liquid at a constant rate. To simplify the mathematical analysis the case of a single well located in an infinite reservoir of constant thickness is considered, and it is assumed that this well is perforated over an interval adjacent to the upper (or lower) impermeable boundary of the formation. A well producing from an arbitrary interval within the bulk of the formation could be treated in a similar way, but this action would merely add to the complexities of the solution and contribute little to the appreciation of the over-all effects of partial penetration. In a recently published paper by Odeh,10 the more general problem of producing from an arbitrary interval within the pay zone was studied for the quasi-steady state flow of a compressible liquid, This is the case in which there is a a potential at the outer boundary and a constant flux across the same boundary. This contrasts with the problem studied here, which is concerned with the nonsteady-state behavior in an infinite reservoir. The exact solution of the partial penetration problem presents great analytical problems because the boundary conditions that solutions of the partial differential equations must satisfy are mixed; i.e., on one of the boundaries the pressure is specified on one portion and the flux on the other. This difficulty occurs at the wellbore, for the flux over the nonproductive section of the well is zero, and the potential over the perforated interval must be constant. In the case of constant rate production from the well, this uniform potential is time dependent and unknown, and the additional condition that where hl is the thickness of the perforated interval, must also b'e satisfied. This problem may be overcome in the case of constant rate production by making the assumption that the flux into the well is uniform over the entire perforated interval, so that on the wellbore the flux is specified over the total formation thickness. This approximation naturally leads to an error in the solution since the potential (pressure) will not

Jan 1, 1970

By Robert Maes, Robert de Strycker

The Fe-Ni-As phase diagram has been established by the study of about a hundred alloys, by microscopic observation, and by thermal analysis, with arsenic contents up to 50 pct. The iron and nickel arsenides present extensive solid-solution fields, owing to the substitution of nickel by iron and vice versa; the extent of the solubility field of each compound has been determined with an accuracy of ±1 pct. In the investigated range of compositions, the solidification reactions were established, and the temperatures of the invariant reaclions detevmined with a preciston of iZ°C in the most favorable and ±5°C in the least favorable cases. The isothermal lines of the liquidus surface have also been drazum, with an accuracy estimated at i5°C. Reactions in the solid state, which take place for the formation or the decomposition of certain phases , were investigated in detail. ThE constitution diagram of the Fe-Ni-As system is fundamental for the understanding of the properties of nickel speiss; these by-products of the extraction of certain nonferrous metals indeed often contain the three elements iron, nickel, and arsenic as main constituents. The Fe-Ni-As system has already been the object of earlier investigations. In 1932, Guertler and Savels-berg' have presented some elements of the ternary phase diagram, for arsenic contents up to 55 pct (all percentages in this paper are given in weight percent), including the vertical section FezAs-Ni&s2. This investigation is, however, incomplete and certain anomalies suggested the necessity to verify the results published by these authors: the section Fe2As-Ni5AsZ, for example, is presented as quasi-binary, with a field of complete miscibility in the solid state, even though these compounds do not have the same structure. Recently, ~useck' established an isothermal section at 800°C in the Fe-Ni-As system by X-ray diffraction and microscopic examination of water-quenched alloys. These techniques are sometimes inaccurate for the determination of the fields where alloys are liquid at the investigated temperature, and they may lead to erroneous conclusions when a compound exists at the investigated temperature but is not stable at room temperature and cannot be maintained by quenching. LIMITING BINARY PHASE DIAGRAMS The Fe-As constitution diagram has been the object of several investigations which have been reviewed by Hansen and Anderko.3 For the Ni-As phase diagram, a recent study has been effected by Yund.4 In this system, Heyding and calvert5 have determined the existence of a compound of unidentified structure at arsenic contents slightly higher than those corresponding to Ni5As2 and at temperatures lower than about 200°C; by analogy with iron and cobalt arsenides, this compound could correspond to the formula Ni2As, as suggested by Kulle-rud; although this is not definitely established. In the region of arsenic contents from 35 to 55 pct, Fried-rich7 detected anomalies in the solidification reactions; an interpretation of these anomalies was given by Hansen ,8 which assumed that the solidification reactions in practice are not in equilibrium, but are meta-stable. The main features of the Fe-Ni constitution diagram are the existence of a complete miscibility field, at least at high temperatures, for the fcc phase (which will be designated by My in the remainder of this work) and of a limited solubility field for the bcc phase (designated by Ma). EXPERIMENTAL METHODS The fundamental technique used in this investigation was microscopic observation, which allowed the determination of the reactions occurring during solidification of the alloys, and possible reactions in the solid state.

Jan 1, 1968

By G. M. Gay, R. L. Perrine

This paper describes a method of numerical computation for three-dimensional, unstable, miscible displacement behavior useful for heterogeneous systems, as well as for more ideal conditions. In the method, flow equations are first linearized by a perturbation approach. The basic flow process is separated and a solution for its behavior readily obtained. The remaining problem of deviations from the basic flow caused by non-ideal conditions is then subjected to numerical analysis. Results obtained from use of the method are also presented. Although conditions assumed in the test cal-culations were severe, results show the type of dispersing flow expected and appear quite satisfactory. The method has eliminated or reduced in importance problems of oscillating values near steep fronts, excessive computer smoothing, etc. A unique advantage of the method is that the source of variations from ideal behavior can be observed. ,The one serious drawback results from the algebraic complexity of the perturbation approach, and the need for second-order terms to be retained in calculations of interest. Fewer array points are available and more computer time is required than would be desired. However, these difficulties are also experienced with other approaches to the solution of three-dimensional dispIacement problems. INTRODUCTION Prediction of behavior of the miscible displacement process within a porous medium for any system of engineering importance is plagued by a number of difficulties. With typical fluid properties the displacing fluid has the lower viscosity, and there is a natural tendency toward flow instability. The problem of predicting instability is compounded by the fact that every real system is heterogeneous. Permeability will vary from point to point — not entirely systematically, and yet not in a random fashion leading to a readily defined average. Furthermore, permeability is unlikely to be isotropic. These permeability properties, which are characteristic of any real porous medium, accentuate the effects of flow instability. Another factor to be considered is that flow dispersion accompanies the displacement process in a porous medium. Mechanistically, dispersion is due to the fact that flow between any two points in the medium follows multiple tortuous paths, each characterized by slightly different flow properties. The fact that the coefficient characterizing dispersive properties of the real medium JS a tensor with variable elements complicates matters. Further difficulties arise because the parabolic influence, while not negligible, is small. Thus, there is a tendency toward steep and uneven displacement fronts along which dispersion smoothing must be represented accurately. Yet frequently used numerical analysis schemes may tend toward either instability, oscillation or excessive smoothing, none of which gives the desired accurate picture of flow behavior. Thus, while many experimental and analytical studies of the process have been made, predictions of actual performance are still subject to considerable uncertainty and possible improvement. This paper reports on part of a study which attempts to develop improved methods for solution of the flow equations describing the miscible displacement process. Of necessity, the calculations were performed on a large digital computer. A three-dimensional system is represented, and the coefficients defining dispersion and permeability can be varied in a manner representative of a real system. Because the differential equations are nonlinear, a first step is to linearize them using a perturbation approach. Objective of this approach has been to separate the basic flow which would occur in an ideal system and for which predictions are readily made. Then the numerical analysis, which is necessarily complex, is directed toward solution of the perturbation equations. These equations state how the basic flow is modified

Jan 1, 1967

By R. I. Jaffee, E. S. Bartlett, D. N. Williams

The oxidation resistance and fabricability of molybdenum powder-metallurgy alloys containing up to 20 wt pct nickel in the form of a fine dispersion were studied. The effects of dispersion distribution and of amount and type (pure nickel or nickel compounds) of additive were considered. A definite improvement in oxidation resistance was observed, but the amount of improvement fell far short of requirements for an oxidation-resistant alloy. Fabricability by hot-rolling was limited. INVESTIGATIONS at Ohio State university1 and Battelle Memorial institute2 have indicated that the attainment of a truly oxidation-resistant molybdenum-base alloy for high-temperature service, with the coincident retention of high-temperature strength and toughness, will be a difficult, if not impossible, feat. Among the most promising alloy systems studied was the molybdenum-nickel system shown in Fig. 1. Alloys in this system show good oxidation resistance at temperatures well above the normal melting point of molybdenum trioxide because of the regenerative formation of protective nickel-molyb-date (NiMoO4) on the surface of the alloy during exposure at high temperatures. Unfortunately, the required alloy content to obtain oxidation resistant molybdenum-nickel alloys far exceeds the limits of the narrow terminal solid-solution boundary, and closely approaches the peri-tectic composition involving the brittle intermetal-lic compound, MoNi. Because of the peritectic reaction, fusion alloys are characteristically brittle, since the MoNi phase, being the last to form, tends to occur as a continuous grain-boundary network around each molybdenum grain. An added problem with this system is that the protective nickel-molybdate layer formed at high temperatures is unstable, undergoing one or more allotropic modifications upon cooling. The volume changes associated with this modification during cooling are such as to cause violent expulsion of the protective molybdate coating from the surface of the piece. This is, of course, quite detrimental in material exposed to cyclic temperature during service. The research reported here resulted from an attempt to overcome the aforementioned problems associated with the molybdenum-nickel alloys primarily by controlling the amount, size, and location of the MoNi phase. By using powder-metallurgy techniques of alloy preparation and sintering the blended molybdenum-nickel powder mixtures at a temperature below the peritectic reaction temperature, molybdenum-nickel alloys with a randomly dispersed MoNi phase were produced. MATERIALS The molybdenum powder used in the preparation of alloys for this study was commercial hydrogen reduced, minus 325-mesh (Tyler) material. Master alloys, by which nickel was added, consisted of elemental carbonyl nickel powder (12 average particle size), an alloy of 50 wt pct Ni-50 wt pct Cr, and compounds of MoNi, Ni,Si, and Ni2B. The nickel-chromium powder was prepared by filing a vacuum-arc-melted ingot, separating impurities magnetically, and screening to minus 200-mesh. The MoNi powder was prepared by pulverizing a hydrogen-sintered powder-metallurgy compact to minus 325-mesh material. Further sizing of this powder for study of effects of dispersion distribution was accomplished by air classification of the minus 325-mesh powder. The Ni2Si and Ni2B powders were prepared by crushing vacuum-sintered powder-metallurgy compacts to minus 200-mesh material. EXPERIMENTAL PROCEDURE Specimen Preparation—The general procedure followed in the preparation of alloys for testing is described as follows: 1) Blending. Either hand-blending of the powders in a ceramic mortar and pestle for X 1/2 or mechanical blending was used. Hand-blending seemed preferable for optimum homogeneity. 2) Compacting. Most of the blended powders were pressed to green compacts at pressures of 15 tons per sq in. in a single-action die, although some were pressed at 80 tons per sq in. A 1 X 11/4 X 1/4-in. green compact was the standard size for oxidation test material, and 7 X 1/2 X 1/4-in. compacts were used for fabricability test specimens. 3) Sintering. All green compacts were sinteredat

Jan 1, 1959

By J. A. Elias, A. J. Heckler

A simplified method is described for the prepauatior~ of comnposite samples for determination of one quadrant of an X-ray pole figure of a sheet material by the Schulz reflection method. Other special conzposite sampling methods were used to investigate specific areas of the pole figures. Plotting nets for these sampling plans were drawn by a computer. Step-scanning techniques were used to obtain detailed information about specific regions of a pole figure. Satisfactory pole figures were obtained for large-grained (0.5 to 1 in. diam) sheet materials. COMPOSITE sampling techniques for preparing one quadrant of an X-ray pole figure have been reported by a number of authors.'-5 All of these methods are similar in that they use a composite sample prepared by stacking and holding together coupons cut from a sheet material. The schulze reflection technique is applicable to the composite sample so that only minor corrections are needed for defocusing and absorption effects. Also, a relatively large area of a sheet is sampled through the use of multiple cross sectional pieces. In general, sampling methods necessary to obtain a composite-type specimen have been formidable enough to discourage their use in routine pole figure determinations. This paper presents a versatile sampling technique which is simpler than those previously described along with an accurate method for obtaining stereographic plotting nets for any orientation of a composite specimen surface. A step-scanning technique is used to obtain more quantitative determinations of specific regions in pole figures, and the composite technique is satisfactorily applied to determine pole figures of very large grained sheet materials. REVIEW OF PREVIOUS TECHNIQUES Mueller and ~nott' first proposed a technique for obtaining one complete quadrant of an X-ray pole figure by using a sheet sample and six different composite samples in which the individual coupons were aligned and held together by two types of jigs. Data from the seven different samples were overlapped to obtain a single quadrant. ~eieran~ and Lopata and ~ula~ obtained a complete quadrant of a pole figure from a single sample prepared by forming a cube from coupons cut from a sheet. These coupons were bonded together using an epoxy cement, and a cut was made at 54.7 deg to the normal, transverse, and rolling directions of the sheet, Fig. I. The normal direction of the cut face projects into one quadrant of a stereographic projection at equal angles to the three principal orthogonal reference directions mentioned above. X-ray reflection data obtained along a spiral path in physical space could then be plotted on a stereographic net offset 54.7 deg from the normal, as shown in Fig. 2. A rotation about an axis representing the surface normal is designated as a, and a rotation about an axis lying in the plane of the surface is 4. From Fig. 2 it is apparent that a tilting angle of no more than 55 deg @ is necessary to obtain the complete quadrant. ~eber~ devised an improved sampling technique by cutting coupons 45 deg to the rolling direction of a sheet, stacking and bonding these with an epoxy cement at an angle of 54.7 deg to a reference plane, and finally grinding and lapping a flat specimen surface parallel to this plane. An important consideration of this technique and that of Lopata and ~ula~ is variation in the diffracting power of the specimen surface due to the bonding medium. While any composite-type sample will have some such variation due to boundaries between coupons, the problem can be minimized by elimination of the bonding medium and by avoiding prolonged etching times which could lead to significant inter lame llar attack of the coupon faces. EXPERIMENTAL TECHNIQUE We were working concurrently though independently on a technique similar to ~eber's,' but which does not

Jan 1, 1968

By L. D. Mullins, E. B. Elfrink, W. L. Wahl

In the past few years several articles and papers presenting results of solution gas-drive depletion calculations have appeared in the lit-erature. Such calculations are of interest to the oil industry, for investment decisions must often be made before much is known about a reservoir. At other times, an estimate of the possible benefits to be realized from alternate production methods is desirable, and theoretical depletion calculations can serve as a floor or reference level from which to work. In any case, an estimate of ultimate oil recovery based upon engineering data is commonly required. An engineer confronted with the problem of obtaining, for a specific reservoir system, an estimate of ultimate oil recovery by solution gas-drive depletion usually will be forced to perform the calculations himself. This is despite the quantity of data in the literature. Rarely will either experience or the literature provide results from a reservoir system similar in all important respects to the one under consideration, and calculated results are not so plentiful that satisfactory interpolation procedures can be devised. Performing the calculations, however, is a tedious, time-consuming task unless an electronic computer is available, and, in practice, time and manpower are not always available for this purpose. A quick, simple, consistent method was needed for reducing the uncer- tainty in estimated oil recovery from solution gas-drive reservoirs when only minimum information about the reservoir system is available. PROCEDURE Method of Calculation The usual requisite assumptions were made so that the material balance equation could be used to calculate data for the charts. The following assumptions were made: (1) the reservoir is homogeneous and isotropic; (2) oil recovery is due entirely to solution gas drive and neither a gas cap nor a water drive nor gravity drainage is present; (3) the initial reservoir pressure is the bubble-point pressure of the reservoir fluid; (4) initial total liquid saturation is 100 per cent of pore space; (5) interstitial water saturation remains at the initial value as the reservoir pressure declines from the bubble-point pressure to atmospheric pressure; (6) equilibrium gas saturation is 5 per cent of pore space; and (7) oil and gas saturations are uniformly distributed throughout the reservoir at all times. There are no saturation gradients due to a wellbore, nor is the geometry of the reservoir system considered. The material balance equation was written in the form of a differential equation' which was integrated to determine the change in oil saturation for an assigned pressure drop. Formal integration was not possible, so recourse was made to the Runge-Kutta method' of numerical integration. All computations were performed on IBM equipment. Numerical integration yielded the change in oil saturation within the reservoir as the pressure. declined from the bubble-point pressure to atmospheric pressure. The initial oil saturation minus the change in oil saturation yielded the oil saturation at atmospheric pressure. The oil originally in place was obtained by dividing the initial oil saturation by the initial formation volume factor (differential liberation) while the oil in place at atmospheric pressure was obtained by dividing the final oil saturation by the formation volume factor at atmospheric pressure. U1timate oil recovery, expressed as a percentage of the initial oil in place, was obtained by dividing the difference between the oil initially in place and the oil in place at atmospheric pressure by the oil initially in place and multiplying by 100. PVT Data Charts were based upon 135 solutions to the material balance equation. PVT properties of the reservoir fluids were variables in this equation and had to be known as functions of pressure. The required PVT data might have been obtained from either actual reservoir fluid systems or correlated data. However, correlated PVT data were developed and employed in the calculations for the following reasons: (1) it is doubtful that 135 sets of PVT data could have been obtained for the values of variables investigated in this study, and (2) results of recovery calculations from randomly obtained PVT data could not be correlated as well as rssults from selected PVT data. The PVT data used to develop the

By Morris Muskat

A theory is presented for calculating the performance history of complete water-drive systems producing from idealized stratified formations. The general equations are applied to systems where the permeability stratification is either of the exponential or linear type. Calculations were carried through for different degrees of permeability stratification, but with special emphasis on the effect of the mobility ratio between the produced oil and the invading water on the resultant performance. These results are also expressed graphically as curves for the initial water breakthrough recovery, for the different degrees of stratification, as a function of the mobility ratio, and of the composition of the produced fluid stream as a function of the cumulative oil recovery. For several typical cases the latter has also been plotted as a function of the cumulative oil and water throughflow. The general result is that when the mobility of the oil is lower than that of the invading water the channelling tendency resulting from the permeability stratification becomes aggravated as the higher permeability zones become flooded out. Situations of this type would obtain when producing low gravity or highly viscous oils. Conversely, if the mobility of the oil is high compared to that of the invading water, the flooding of the high permeability zones will lead to a retarding and choking effect, and the gross bypassing phenomena will be partially suppressed. These conditions would correspond to those of flooding high gravity or low viscosity oils. A discussion is given of the various basic assumptions made in the analysis, including that of ignoring the stripping phase of the production history as implied by relative permeability concepts. INTRODUCTION The physical ultimate recoveries from oil reservoirs are basically determined and limited by the physical oil displacement processes associated with the reservoir producing mechanism. In practice, however, the economic ultimate recoveries are further limited by the mobility of the reservoir fluids and the uniformity and continuity of the producing formation. In fact, it is the differential depletion between the component parts of the composite reservoir which ultimately determines the total recovery at the time of field abandonment. While this observation applies to both the solution gas drive and gravity drainage mechanisms, in which use is made only of the energy contained within the original oil-bearing reservoir, it is of even more paramount importance under operations wherein the energy associated with extraneous fluids provides the ultimate oil expulsion mechanism. Whether the invading fluid is the water from an edgewater drive, water injected for pressure maintenance, gas injected for pressure maintenance, or gas returned to the formation in a cycling program, it is often the continuity and uniformity of the producing section which will control the economic efficiency of the operations. The importance of the problem of reservoir non-uniformity does not, of course, lessen its complexity or the difficulties of its solution. In fact, these are inherently such that the concept of a "general" solution is virtually meaningless. About all that can be reasonably hoped for is the analysis of specific and well-defined types of non-uniformity which may give some degree of approximation to actual reservoir conditions. Since variations in the nature of the reservoir which depend only on the position along the streamlines will not lead directly to major differential depletion development within the reservoir, the types of non-uniformity considered thus far have involved stratification assumptions. That is, the producing section has been replaced by a multi-layer "sandwich," each uniform areally, and differing from the others only in its basic physical constants as to thickness, porosity and permeability. The fluid motion in the composite system is thus approximated by a parallel superposition of the independent fluid movements in the individual strata. For the specific application to cycling operations the theory of the effect of permeability stratification has been developed for both discontinuous' and continuous types of permeability stratification. Among the latter, treatments have been given of systems in which permeability distributions are governed by exponential, linear' or probability3 functions. In all these studies complete dynamical equivalence was assumed between the injected dry gas and the displaced wet gas. The overall effective permeability of each stratum was therefore considered as constant and independent of the degree of invasion of the injected fluid. In the case of the displacement of oil by water, the assumption of dynamical equivalence between the water and oil will be strictly valid only by accident. Even if the oil viscosity should be the same as that of the water, the effective permeability to the oil in the presence of the connate water will in general be quite different from that of the water behind the water-oil interface flowing past the trapped residual oil. As a result the effective permeability for the stratum as a whole and rate of water invasion will change with time as the intrusion continues. The differential fluid motion in the individual strata will thus also vary with time. Qualitatively, it is easy to predict the resultant effects. If the permeability to viscosity ratio of the invading fluid exceeds that of the fluid displaced, the stratification and bypassing effect of the perme-

Jan 1, 1950

By B. W. Christ, P. M. Giles

Mossbauer effect measurewents have been made on I-mil-thick foils of commercial 1 wt pct C steel and Fe-2 wt pct C alloy. The experimental method required about 3 to 5 vol pct of a phase in the nzultiphase steel sample for detection. Room-temperature Md'ssbauer patterns obtained on austenitized atid quenched samples exhibit fifteen, and possibly twenty-one, lines. A sharp parama&tetic singlet and a quadrupole doublet, poorly resolued from the singlet, are attributed to austenite. Remaining lines are due to tnartensite. Accurate evaluation of austenite line paranzeters is not feasible if significant amounts of other phases such as carbides or martensite occur simultaneously with austenite. This is demonstrated by comparison of hyperfine interactions determined for austenite in multiphase high-carbon samples with those reported for Fe-C austenite in a nearly 100 pct austenitic sanple. Lines from carbides are incompletely resolced from austenite lines, as demonstrated by comparison of austenite line positions with carbide line positions calculated frow published values of hyperfine interactions. One martensite line overlaps an austenite line in the pattern for commercial 1 wt pct C steel. Results of this study suggest that the usefulness of e M6ssbauer pectroscopy for quantitatizle analysis of austenite in bulk samples of quenched and tempered high-carbon steels is restricted by poor resolution. Use of Mossbauer spectroscopy for phase identification and for evaluation of atomic and electronic structures appears quite feasible, however, The Mijssbauer effect has been widely discssed,'-and e Mossbauer effect measurements have been reported on materials of metallurgical interest.7"20 In particular, it has been proposed that e Mossbauer patterns of commercial steels and laboratory-made Fe-C alloys, in the quenched condition, are composed of lines originating in two phases, Fe-C austenite and Fe-C martenite.-' Evidence accumulated in this study demonstrates that three absorption lines found in the central region of the e Mossbauer pattern obtained on quenched steels are attributable to retained austenite. This interpretation is supported by parallel decreases in the intensity of these three lines caused by subambient cooling of commercial 1 wt pct C steel samples after water quenching to room temperature. A second result of this study is to clarify effects of line resolution and sensitivity in the Mossbauer patterns of multiphase steels on the accu- rate determination of austenite line parameters. Experimental line widths (full width at half height) are generally 1.5 to 3 times larger than the natural line width of 0.19 mm per sec. At least two lines, and sometimes more, from a single phase such as cementite (Fe3C), other carbides, martensite, and austenite fall in the energy band, i0.85 mm per sec. hhis band width is employed simply for convenient reference. It represents approximately the energy interval between the + 112 to 112 transitions in ferrite and is expressed as the velocity needed to Doppler shift a 14.4kev 7 ray to the aforementioned ferrite energy levels. This energy band is referred to as "the central region of the Mossbauer atttern:: in this paper. Hence, due to the large number of lines from different phases in a multiphase steel falling in a relatively narrow energy band, absorption lines from the different phases may overlap. Analysis of available data, presented below, indicates that this occurs to a significant extent for phases which commonly occur in quenched and quenched and tempered high-carbon steels. One consequence of limited resolution in the Mdssbauer patterns from multiphase steels is difficulty in accurate determination of such line parameters as position, width, and intensity. In fact, it appears that quantitative analysis for retained austenite in quenched and tempered high-carbon steels is not practical with the present experimental method. Line resolution is influenced to some extent by sensitivity. We point out below that atom or volume fractions of less than about 3 to 5 pct are not detected by the present experimental method. Thus, the presence of a multiplicity of phases does not always lead to impaired resolution. Finally, we report in this paper room-temperature MGssbauer parameters determined for austenite in a freshly quenched, commercial 1 wt pct C steel and in a freshly quenched laboratory heat of an Fe-2 wt pct C alloy. These parameters are compared with others reported in the literature. Three types of hyperfine interactions are detectable in a Mossbauer effect measurement: isomer shift, quadrupole interaction, and magnetic dipole interaction. These interactions are evidenced by one, two, and six line patterns, respectively.'-4 More than one type of interaction has been reported in certain metallurgical phases thus far studied by this method. Isomer shift is the experimentally measured displacement of line position from some arbitrarily defined reference position. In the case of a multiline pattern, isomer shift is given by the displacement of the centroid (center of gravity) of that pattern from the reference position. All isomer shifts measured at finite temperatures contain a second-order Doppler effect characteristic of that temperature. The isomer shift is related to the total s electron density at the nucleus, becoming more negative with increasing s electron density. The first-order quadrupole effect arises from the interaction between the nuclear quadrupole moment and any axially symmetric electric field gradient in

Jan 1, 1969

By G. R. Leverant, C. P. Sullivan

To evaluate the effect of a dispersion of nondeform-able, incoherent, second-phase particles on high-strain cyclic deformation and fracture, recrystallized TD-nickel (Ni-2ThO2) and a commercially pure nickel, Ni-200, were fatigued under strain control at total strain ranges varying from 0.009 to 0.036. Relative to the Ni-200, the slip at the surface of the TD-nickel was more wavy and discontinuous due to the presence of the thoria particles. This made crevice formation (incipient cracking) within slip bands more difficult in TD-nickel than in Ni-200. Both materials cyclically hardened to a constant (saturation) flow stress which increased with increasing plastic strain amplitude. Cellular substructures were developed in both materials during cycling. The cell size in TD-nickel was controlled by the thoria particle distribution and was independent of plastic strain amplitude over the range investigated. The cell size in Ni-ZOO was larger than that in TD-nickel at similar plastic strain amplitudes and was a function of plastic strain amplitude. These results, together with the cyclic stress-strain curves for both materials, are discussed in terms of a model for fatigue strain accommodation at saturation recently proposed by Feltner and Laird. NUMEROUS fatigue investigations have considered the interrelation of slip character, dislocation substructure, and cracking in pure metals and solid-solution alloys. However, except for the studies of the low-strain fatigue of internally oxidized copper alloys1 and cast, dispersion-strengthened lead,&apos; little is known about the effects which small, incoherent, nondeform-able, second-phase particles have on cyclic deformation and cracking processes. Effects due to the particles alone are often obscured by a dislocation substructure introduced during thermomechanical processing of dispersion-strengthened metals. In the present study, recrystallized TD-nickel and a commercially pure nickel, Ni-200, were employed to evaluate the effect of a thoria dispersion on high-strai fatigue deformation and cracking at room temperature. I) MATERIAL AND EXPERIMENTAL PROCEDURE The TD-nickel was supplied by DuPont as a 5/8-in.-thick stress-relieved plate which had been subjected to a proprietary schedule of thermomechanical treatments, and the Ni-200 as 3/4-in. bar which was subsequently annealed for 2 hr at 850°C in argon resulting in an average grain diameter of 0.05 mm. The compositions of these materials are given in Table I. The microstructure of the TD-nickel consisted of elongated grains parallel to the primary working direction with an average width of 0.16 mm, Fig. l(a). Many fine annealing twins were present indicating that the starting material was in a recrystallized condition; this supposition was confirmed by the absence of of any extensive dislocation substructure, Fig. l(b). Sheetlike stringers parallel to the rolling direction were occasionally seen both within grains and at grain boundaries. Some approximately spherical particles about 2 in diam, which may correspond to exceptionally large thoria particle aggregates, were also present. The average Young&apos;s modulus of the plate material in the rolling direction was 21.8 X 106 psi which is consistent with a {100}<001>recrystalliza-tion texture3&apos;* being prominent. In transmission microscopy, the 2.3 vol pct of thoria particles generally appeared to be uniformly distributed although some clusters, 0.1 to 0.3 in diam, of larger particles were observed as previously reported for TD-nickel sheet,5 and stringering of particles was present in some areas as welt. The average diameter of the thoria particles was 450A with a calculated mean planar center-to-center spacing of 2100A, as determined by quantitative metallographic analysis.= The 0.2 pct offset yield stress was 36,000 psi which agrees with the value predicted by the modified Orowan relation7 for edge dislocations bowing between thoria particles of the size and spacing observed in the present investigation. Fig. 2 illustrates the specimen design employed for the axial high-strain fatigue testing. Adapters were screwed onto the threaded portions of each specimen so that testing could be performed in the same manner as that reported for buttonhead specimens.8 Stressing was coincident with the working direction for both materials. The gage section of each specimen was electropolished and lightly etched prior to testing. The total strain was controlled, being varied between zero and a maximum tensile strain ranging from 0.009 to 0.036. In addition to these tests, a circum-ferentially notched TD-nickel specimen was cycled over a total strain range of 0.0075. The same strain

Jan 1, 1969

By A. Lubinski

In drilling operations, attention generally is given to hole angles rather than to changes of angle, in spite of the fact that the latter are responsible for drilling and production troubles. The paper presents means for specifying maximum permissible changes of hole angle to insure a trouble-free hole, using a minimum amount of surveys. It is expected that the paper will result in a decrease of drilling costs, not only by avoiding troubles, but also by removing the fear of such troubles. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Excessive dog-legs result in such troubles as fatigue failures of drill pipe, fatigue failures of drill-collar connections, worn tool joints and drill pipe, key seats, grooved casing, etc. Most of these detrimental effects greatly increase with the amount of tension to which drill pipe is subjected in the dog-leg. Therefore, the closer a dog-leg is to the total anticipated depth, the greater becomes its acceptable severity. Very large collar-to-hole clearances will cause fatigue of drill-collar connections and shorten their life, even in very mild dog-legs. Another finding regarding fatiguing of collar connections in dog-legs is that rotating with the bit off bottom sometimes may be worse than drilling with the full weight of drill collars on the bit, mainly in highly inclined holes when the inclination decreases with depth in the dog-leg. Means are given for specifying maximum dog-legs compatible with trouble-free holes. An inexpensive technique proposed is to take inclinometer or directional surveys far apart; then, if an excessive dog-leg is detected in some interval, intermediate close-spaced surveys are run in this interval. The application of the findings should result in a decrease of drilling costs, not only by avoiding troubles, but mainly by removing the fear of such troubles. The result would be much more frequent drilling with heavy weights on bit, regardless of hole deviation. Because of errors inherent to their use, presently available surveys are not very suitable for detecting dog-legs. There is a need for instruments especially adapted to dog-leg surveys. Crooked hole drilling rules should fall into two distinct categories—(1) those whose purpose is to bottom the hole as desired, and (2) those whose purpose is to insure a trouble-free hole. Three kinds of first-category rules in usage today are as follows. 1. A means to bottom the hole as desired is to prevent the bottom of the hole from being horizontally too far from the surface location; this may be achieved by keeping the hole inclination below some maximum permissible value such as, for instance, 5. 2. Another means to achieve the same goal is to limit the rate at which the inclination is allowed to increase with depth. A frequently used rate is 1/1,000 ft. In other words, a maximum deviation of l° is allowed at 1,000 ft, 2 at 2,000 ft, 3 at 3,000 ft, etc. 3. Whenever application of the first two means precludes carrying the full weight on bit required for most economical drilling, then the best course is to take advantage of the natural tendency of the hole to drift updip, displace the surface location accordingly and impose a target area within which the hole should be bottomed. This method has already been successfully applied,&apos;.&apos; and its usage probably will become more frequent in the future. Means for calculating the amount of necessary surface location displacement are avail-able.3&apos;5&apos;6 If in high-dip formations the full weight on bit should result in unreasonably great deviations, the situation could be remedied by increasing the size of collars and (if needed) the size of both hole and collars,351 or in some cases by using several stabilizers. Rules which would fall into the second category (i.e., rules whose purpose is to insure a trouble-free hole) are seldom specified today. It is vaguely believed that following Rules 1 and 2 of the first category will automatically prevent troubles. Actually, this is not true. If at some depth the only specified rule is that the hole inclination must be less than 4", the hole may be lost if the deviation suddenly drops from 4 to 2, or if the direction of the drift changes, etc. Rule 3 of the first category is generally used in conjunction with a rule belonging to the second category, namely, that the hole curvature&apos; (dog-leg severity) must not exceed the arbitrarily chosen value of 1½ /100 ft. Moreover, when using this rule, the industry is not clear over what depth intervals the hole curvature should be measured. All this results in a frequent fear

By D. A. Munro, G. P. Contractor

Fourteen 50-lb laboratory melts were investigated to determine the effect of uranium on the tenpering characteristics of loo-carbon (0.06 to 0.1 pct C) steels. It was found that uranium additions, particularly in the range 0.30 to 0.45 pct, enhanced the hardness and both ultimate and yield strength of the experivzental steels in the quenched and tempered condition. The structural and morphological chazges indicated that uranium retarded tempering of the tnartensite, thereby hindering the normal formation of polygonal ferrite formed in the late stages of tempering. The effect of this was to make possible the re-tension of the acicilar ferritic structure in the uranium-bearing&apos; steels. The iraniuin-bearing steels also showed IVidnzanstatten-type growth of ferrite plates and had large prior austenite grains containing assenzblies of fine ferrite grains, mainly acicular in geometry. The fine-grained ferrite structure and the presence of more numerous and apparently smaller precipitates in the uranium-bearing steels are thought to he principally responsible for the itnproved tensile strength and hardness of the experinzental uranium-bearing steels. At ternperirzg temperatures above 455% (850&apos;F) the ferrite in the higher-uraniun steels nzaintained acicularity and, hence, its strength and resistance to tempering. Uranium did not produce a secondary hardening peak. However, it retarded softening during the third stage of tempering because of its effect of inhibiting the grouth of cementite particles and of retaining the acicularity of ferrite plates. The resistance to coalescence accounted for the slow grocth of the ferrite grains in the uranium-modified steels and, hence, fov the persistence of the acicular ferrite structure. IT had been found previously1 that uranium additions up to about 0.45 pct had no significant effect on the tensile properties of low-carbon steel (0.06 to 0.10 pct C) in the as-rolled and normalized conditions, Fig. 1. On the other hand, it was observed that uranium in excess of about 0.30 pct had an embrittling effect as revealed by Charpy V-notch impact results. It was also noted that, as the uranium content increased, the morphology of pearlite changed from lamellar to feathery and the ferrite grains showed an etching effect resembling striated or dashed markings, suggestive of precipitation. The sharp drop in the impact properties shown in Fig. 2 warranted an assumption that the uranium content of about 0.30 to 0.45 pct may produce some secondary hardening reaction on tempering, analogous to that associated with a Cr-Mo-V steel, which shows very poor CVN toughness at the secondary hardness peak in the tempering curve.1&apos; With this background and the reported findings of Hasegawa and noda that low-carbon uranium-treated steel showed signs of secondary hardening, the present investigation was undertaken to determine the effect of uranium additions on the mechanical properties of 0.10 pct C steels. No attempts were made to investigate in detail the mechanisms of hardening, although some suggestions based on the experiments are made. MATERIALS AND PROCEDURES A series of 50-lb induction-furnace melts was made using AISI 1008 rimming steel billets as the melting stock. The melting, forging, and rolling techniques proven satisfactory in previous projects&apos;-3 were employed as a guide for this investigation. The steel was deoxidized with aluminum (2 lb per ton) prior to the addition of high-purity uranium. The analysis of each melt is given in Table I. Properties were evaluated as a function of heat treatment and are presented in terms of hardness and tensile strength vs tempering temperatures. The variation of hardness with the tempering temperature was studied on the quenched and tempered specimens, some of which measured 0.50 by 0.25 in. diam and the others 0.40-in. cubes. Before quenching, the specimens were vacuum-sealed in glass tubes and normalized at 900°C (1650°F) for 20 min. Following this treatment, the sealed specimens were hardened by austenitizing at 955°C (1750°F) for 20 min and water quenching, and then tempered for 1 hr in the range 150 to 730°C

Jan 1, 1967

By W. S. McCain, B. A. Wilcox, A. H. Clauer

The creep and creep-fracture behavior of a Ni-200-2Th0, alloy has been studied over the temperature range 816° to 1038°C and stress range 4000 to 19,000 psi Specimens having their axes either parallel or transverse to the final sheet-rolling direction were tested in air, and the apparent activation energy for creep was found to be 92 kcal per mole, regardless of specinien orientation. The stress dependence 01- the steady-state creep rate, was satisfied by where n depended on orientation; n = 9.7 for rrzaterial transverse to the rolling direction and n = 24.3 for material parallel to the rolling direction. In the latter case, the relation was also valid. The rupture strength of material parallel to the rolling direction was -1.5 times gveater than the transverse rupture strength. Several tests in vacuum revealed that here the creep rate was higher and rupture life lower than for comparable tests in air. THERE is an increased interest today in producing improved dispersion-hardened alloys for high-temperature structural applications. This had led various materials designers to modify existing alloys, such as thoriated nickel. One system that has received considerable attention is based on chromium additions to Ni-ThO2, with the obvious motive of improving the oxidation resistance. Another effect of chromium is that of lowering the stacking-fault energy, y, of nickel. For example, the addition of 30 wt pct Cr decreases y from -225 erg per sq cm for pure nickel1 to -70 erg per sq cm.&apos; This decrease in y could have several effects on the deformation behavior. There is some evidence that cross slip of dislocations around Tho2 particles is a mechanism of tensile3 and creep4 deformation in TD Nickel (Ni + 2 vol pct ThO2) at temperatures below 0.5 the absolute melting point of nickel, T,. A decrease in ? by chromium additions should increase the difficulty of cross slip. Also it has been suggested that the high-temperature (T > 0.5Tm) rate-controlling creep process in "recrystallized" Ni-Tho2 alloys is the climb of edge dislocations over Tho2 particles.5 According to weertman6 a decrease in the stacking-fault energy should decrease the dislocation climb velocity resulting in a decreased steady-state creep rate. Thus, chromium additions to recrystallized Ni-Tho2 alloys would be expected to decrease the creep rate, all other conditions being constant. However, a difficulty arises when the stacking-fault concept is applied to high-temperature creep of dispersion-strengthened metals which are produced on a commercial or semicommercial scale. This occurs because the materials are processed so that they have a high-stored energy of cold work, and the other substructural effects could obscure any influence of stacking-fault energy per se. An example of an alloy with a high stored energy of cold work is TD Nickel bar. This material has a fibrous grain structure (1 I diam by 10 to 15 p long), and the structure is stabilized by the ThO2 particles. Annealing at high temperatures, e.g., 3 hr at 1300°C, causes no recrystallization or grain growth.4 Previous work by the authors4 showed that the high -temperature creep behavior of TD Nickel could not be related to any simple dislocation theory of creep, and this probably was associated with the complex initial structure of the material.

Jan 1, 1968

By E. Buehler, H. J. Levinstein

The temperature ranges in which the three inter-metallic phases in the Nb-Sn system form have been determined and the composition and structure of two of the three phases has been established. The kinetics of the formation of Nb3Sn in cored wire samples has been studied in the temperature range of 800° to 1050°C. From 800°to 950°C the rate of formation increases by four orders of magnitude. The rate-controlling step for the formation process in this temperature range appears to be the diffilsion of tin through NbSn. At higher temperatu~es a change occurs in the mechanism of the formation process such that up to a temperature of 1050°C the rate of formation of Nb3Sn does not increase above the rate observed at 950°C. For temperatures helow 950°C the current-carrying capacity of the wire increases with increased percent reaction reaching a maximum value when the formation process is 90 to 95 pct complete. The maximum current-carrying capacity obtainable in this temperature range is independent of the temperature. Above 950°C tlze current-carrying capacity obtainable in the wire decreases with increasing temperature of formation. A model is proposed which accounts for the ohserved behavior. RECENTLY, Buehler et a1.l reported the results of an investigation of the process variables which influence the superconducting properties of Nb3Sn-cored wire. These results indicated that at least four variables affect the properties of the manufactured wire. These include composition, particle size of the starting powder mix, temperature of heat treatment, and time of heat treatment. In order to understand completely the role of these variables, it is necessary to have an accurate knowledge of the phase equilibria in the Nb-Sn system. At the present time, phase-equilibrium diagrams for the Nb-Sn system have been published by a number of investigators.2-5 The diagrams differ as to the number of phases present, the composition of the phases, and the temperature range of stability of the phases. The present investigation was undertaken in order to resolve these differences. Since the investigation of Buehler et al. demon- strated that the length of time at the temperature of heat treatment affected the superconducting properties of Nb3Sn, it is apparent that it is necessary to understand the kinetics of the formation process as well as the equilibrium conditions before a complete understanding of the system is possible. As a result, the kinetics of formation of the various phases in the system were also studied in this investigation. EXPEFUMENTAL PROCEDURE Diffusion couples and sintered powdered compacts were employed in the phase-diagram investigation. The diffusion couples were made by filling 1/8-in.-ID monel-sheathed niobium tubes with tin. The monel sheath was employed to facilitate drawing.&apos; The tubes were then drawn to a tin-core diameter of 32 mils. Samples approximately 3 in. long were then cut from the drawn composite. The tin was drilled out of the ends to a depth of 1/4 in. and niobium-wire plugs were inserted into the ends and peened over. The monel was removed by etching in concentrated nitric acid, after which the samples were sealed in evacuated quartz bulbs and heat-treated in a resistance-wound tube furnace. The samples were quenched into ice water upon removal from the furnace. The diffusion couple samples were examined metallographically employing a chemical etching solution consisting of 10 ml of saturated chromic acid per g of NaF. In addition, two anodizing solutions were used for phase-identification purposes. The first was the picklesimer7 solution; the second consisted of equal parts by volume of 30 pct H2O2 and concentrated NH4OH to which 1 g of NaF was added per 25 ml of solution. The anodizing conditions for the second solution were 2 v and 100 ma with a tin cathode. The powdered compacts were made by pressing previously mixed powders of 99.9 pct pure Sn and 99.6 pct pure Nb supplied by the United Mineral Co. into cylinders 3/8 in. in diameter by 1/2 in. long. The cylinders were then sealed in quartz tubes and heat-treated in the same manner as the diffusion couples. The samples were examined metallographically and by X-ray diffraction techniques. Since it was desirable to be able to correlate the kinetic data with current-carrying capacity, the type of specimen chosen for this part of the investigation had to be a compromise between the optimum system for studying kinetics and one which was suitable for making current-carrying capacity

Jan 1, 1964

By D. S. Flett, D. W. West

The solvent extraction of chromium 111 has been studied for the system Cr 111, H,SO., H,O/RNH/RNH., xylene, where the primary amine used was Primene JMT. Rate studies have shown that extremely long equilibrium times are required, ranging from 1 hr at 80°C to 20 days at room temperature. Heating the solution prior to extraction increases the rate of extraction. The variation in the amount of Cr 111 extracted is an inverse function of the acidity of the aqueous phase. Thus, the slow rates of extraction appear to be connected with the hydrolysis of the Cr I11 species. Extraction isotherms for the extraction of Cr 111 have been obtained for two sets of experimental conditions, namely at 60°C and for a heat-treated solution cooled to room temperature. The separation of Fe 111 from Cr 111 and Cr 111 from Cu 11 in sulfate solution by extraction with Primene JMT has been studied and shown to be feasible. A survey of the literature relating to the solvent extraction of chromium showed that, although many systems exist for extraction of Cr VI, only a very few reagents have been found to extract Cr 111. The extraction of Cr III by di-(2-ethyl hexyl) phosphoric acid has been reported by Kimura.&apos; A straight-line dependence of slope —2 was observed between log D,, and the log mineral acid concentration at constant extractant concentration. Since the slope of this plot reflects the charge on the ion extracted, it must be concluded that a hydrolyzed species of Cr III is being extracted. Carboxylic acids generally do not form extractable complexes with Cr III but di-isopropyl salicylic acie does extract Cr 111. Simple acid backwashing of the organic phase, however, failed to remove the chromium. Similar difficulty in backwashing was found by Hellwege and Schweitzer8 in the extraction of Cr I11 with acetyl-acetone in chloroform. The extraction of Cr 111 from chloride solutions by alkyl amines has been reported4-&apos; but the maximum amount of extraction achieved in these studies did not exceed 10%: From sulfate solutions, however, Ishimori" has shown that appreciable amounts of Cr I11 were extracted by amines. The amines used were tri-iso-octyl amine, Amberlite LA-1 (a secondary amine, Rohm & Haas) and Primene JMT (primary amine, Rohm & Haas). The efficiency of extraction with regard to amine type was primary>secondary> tertiary. Appreciable extraction of Cr I11 was recorded for Primene JMT as the aqueous phase acidity tended to zero. The major difficulty with Cr I11 in solvent extraction systems stems from the nonlabile nature of the ion in complex formation. This accounts for the slow rate of extraction generally experienced and the difficulty encountered in backwashing the Cr I11 from the organic phase in the case of liquid cation exchangers. Consequently, the possibility of extraction of Cr I11 as a complex anion is attractive since the backwashing problems should be minimized in this way. From published data, it appeared that the extraction of chromium from sulfate solutions of low acidity by primary amines afforded the best chance of success for a useful solvent extraction system for Cr iii This paper presents the results of a study of the extraction of Cr I11 from sulfate solution by Primene JMT and examines the application of such an extraction procedure for the recovery of chromium from liquors containing iron and copper. Experimental Chromium solutions were prepared from chrome alum in sulfuric acid and sodium sulfate so as to maintain a constant concentration of sulfate ion of 1.5 molar. Solutions of Primene JMT were prepared in xylene and the amine equilibrated with sulfuric acid/sodium sul-fate solutions, of the same acidity as the chromium solution, until there was no change in acidity between the initial and final aqueous phases. The solutions of Primene JMT conditioned in this way were then used for the equilibration experiments. Equilibrations at 25°C were carried out in stoppered conical flasks shaken in a thermostat; equilibrations at all other temperatures were carried out in stirred flasks in a thermostat. After equilibration, the phases were separated and analyzed for chromium. In the tests on the rate of extraction, small samples of equal volume of both phases were withdrawn from time to time and the chromium distribution determined. The chromium analyses were carried out either coloi-imetrically using diphenyl carbazide, or volu-metrically using addition of excess standard ferrous ammonium sulfate and back titration of the excess iron with potassium dichromate. The oxidation of Cr 111 to Cr VI in the case of the raffinate solution was effected by boiling with potassium persulfate in the presence of silver nitrate and, for the backwash solution, by boiling with sodium hydroxide and hydrogen peroxide. Results Preliminary experiments indicated that extraction results were effected by the age of the chromium solution, higher distribution coefficients being obtained with solutions which had been allowed to stand for some time. Consequently a stock solution of chrome alum, 10 m moles per 1 Cr I11 in 1.4 M Na,SO,/O.l M &SO,,

Jan 1, 1971

By B. C. H. Steele, C. B. Alcock

In choosing solid oxide electrolytes for use in the measurement of thermodynamic quantities at high temperatures, the two most important criteria are the values of the partial ionic and electronic conduc tizities. Measurements of these conducticities have been made for some Group IVA oxides and solid solutions of these oxides with CaO, Y2O3, and La2O3. The total conductivities as a function of temperature. oxygen partial pressure, and composition were determined, as well as the electromotive forces of cells in which the oxygen partial Pressures of the electrodes were between 10-7 and 10-29 atm at 1000°C. Additional factors that influence the experimental arrangements, such as the operating temperature, the appropviate cell atmosphere, and electrode kinetics, are also discussed. IT has long been recognized that galvanic cells incorporating solid electrolytes can possess many advantages for the measurement of high-temperature thermodynamic data. Unfortunately, the electrical-transport properties have only been established for a few solid electrolytes, although attempts have been made, and are still being made, to use such materials as porcelain&apos; and magnesia.&apos; Kuikkola and Wagner have discussed3 and demonstrated4 various applications of solid electrolytes, and, in particular, their work stimulated the electrochemical measurement of oxygen activities using cells incorporating the solid oxide electrolyte, Zr0.85Ca0.15O1.85 The excellent characteristics of this oxide electrolyte have been confirmed by the investigations of Kingery et al.5 (oxygen ion diffusion), Carter and Rhodes6 (zirconium and calcium ion diffusion), and Weissbart and Ruka7 (electronic transference number). However the observations of Peters and Mobius,8 later confirmed by Schmalzried,9 indicated that the electronic component of zirconia-based electrolytes could become significant at low oxygen chemical potentials. Under these conditions, thoria-based electrolytes have obvious advantages arising from their superior thermodynamic stability. It was known that certain thoria solid solutions possessed defect fluorite structures containing anionic vacancies (cf. Zro.85Cao.l50O1. 85), although data on their electrical properties were conflicting. Kiukkola and wagner4 concluded that thoria-based electrolytes exhibited appreciable electronic conductivity, whereas Peters and Mobius8 mentioned that the electronic conductivity only became important at high oxygen partial pressures, but gave no further details. In addition, both Kiukkola and wagner4 and Peters and Mann&apos;" had determined the oxygen chemical potentials associated with the Fe-FeO, Co-COO, Ni-NiO, and Cu-Cu2O equilibria, and it was not known whether discrepancies in their results could be attributed to the different cell design, or the thoria-based electrolyte used by Peters and Mann. These uncertainties have been resolved during the present investigation into the measurement of low oxygen chemical potentials, and transport properties of the relevant oxide electrolytes are presented and discussed. Other experimental difficulties associated with low oxygen chemical potential measurements are also described, with particular reference to the Nb-O system. 1. THEORETICAL CONSIDERATIONS Oxide systems of potential use as solid electrolytes possess a large energy gap between the valence and conduction bands. Little is known about the defects present in these wide-gap metal oxides as a function of impurity content and deviations from stoichiom-etry, or the mechanisms by which electronic charges move in these solids. For these materials it is convenient to represent the measured conductivity as the the sum of three terms: where a, is the conductivity due to positive hole conduction (p type), and oe is the electronic conductivity due to electron conduction (n type). U there is a large concentration of ionic vacancies fixed by composition then the ionic contribution will not be a function of oxygen pressure. Although the possibility of a pressure-independent electronic component also cannot be ignored, it is likely that such a term would only become significant at very high temperatures, and in general the positive hole conduction will increase with increasing oxygen pressure, and the electron conduction will increase with decreasing oxygen pressure. It follows that the total conductivity is related to oxygen partial pressure as follows:

Jan 1, 1965

By K. F. Veith, R. L. Marovelli, T. S. Chen

An analytical study is made of thermal stress distribution in a thin circular disc subjected to a peripheral thermal shock at various rates of heat transfer. The problem is of importance in predicting the thermal shock response of a rock body of finite size. The theoretical analysis is based on radial heat flow by conduction in the disc and heat exchange by convection between the disc and the surroundings. The case of constant properties and plane stress is treated. Solutions of the stress distribution are presented for both cooling and heating shocks and an average stress theory is formulated. Preliminary experimental verification was obtained from the results of shock tests on thin rock discs insulated on both flat end faces so that heat exchange took place through the exposed peripheral surface. Physical properties vital to the analysis are Young&apos;s modulus, tensile strength, coefficient of linear thermal expansion, thermal diffusivity and thermal conductivity. Plots of these properties are presented. Historically, thermal energy has been used for primary rock removal throughout the ages. Although it is not widely known, the fire-setting method of thermal rock fragmentation was still in use 100 years ago in systematic underground mining operations in Scandinavia.&apos; There, the availability of cheap labor and abundant wood fuel made the method competitive with the early blasting powders. By 1880, as high explosives replaced powder, use of the method declined and it was finally abandoned about 1885. Only 40 years ago, some experimental fire-setting methods were used for rock removal underground at Pribram and Zinnwald in Central Europe.2 Compressed air-oil burners replaced the earlier wood fuels. About 15 years ago, gaseous oxygen-oil burners were introduced in the United States and gained acceptance for difficult blasthole drilling and some quarry opera- tions. At present, there is renewed interest in the compressed air-oil burner designs. The literature included information on thermal secondary processes for rock weakening or particle liberation3,4,5,6,7,8 on thermal spallation of rock 9,10,11,12,13 and on the thermal fracture of materials other than rock. The 1955 Symposium on Thermal Fracture14 and the later work of Hassel-man, cover recent developments in thermal shock investigations on brittle ceramics. Although the number of published analytical and experimental investigations conducted on ceramic and refractory materials is large, the literature reveals that there is a lack of information on both the theoretical and actual response of rock to thermal shock. Many of the thermal shock studies on ceramic and refractory bodies are based on the case of heat transfer at the solid&apos;s surface by convection in either liquid or air. A similar heat transfer situation was adopted for a theoretical analysis of the thermal stresses in a thin circular disc subjected to a peripheral thermal shock. Experimental thermal shock tests were then conducted on rock specimens to check the analytical results. Specially insulated quartzite, basalt and taconite discs were shocked in either liquid or still air bath at various rates of heat transfer. The work was done in order to resolve the question as to whether theoretically determined tensile stresses are actually realized in a finite rock body undergoing thermal shock. Specifically, if theory predicts that the tensile strength of a rock should be exceeded in a particular type of controlled thermal shock, will fracture occur? This information is needed by Bureau researchers working on thermal and electrical rock fragmentation methods. The object of this paper is to present the main features of the theory and procedure being used in the authors&apos; approach to quantitative prediction of rock response to a thermal shock. THEORETICAL ANALYSIS In the analysis to follow, simple geometry of a thin circular disc is considered. The disc is assumed to be isotropic, homogeneous and perfectly elastic, and to have constant thermal and mechanical properties. The effect of radiation is neglected so that heat is transferred by conduction inside the disc and by convection with its surrounding. In addition, the end

Jan 1, 1967