Search Documents

By P. Chiotti, G. R. Kilp

Thermal, metallographic, vapor pressure, and X-ray data were obtained to establish the phase diagram for the zinc-zzrconiz~m system. Five compounds corresponding to the stoi-chiometric formulas ZrZn, ZrZn,, ZrZn,, ZrZn,, and ZrZn14 were observed. All these compounds, with the exception of ZrZn2, which melts congruently at 1180°C under constrained zinc-vapor conditions, undergo pexitectic reactians. The temperature at which the zinc vapor pressure is I atm for a series of alloys was determined from vapor-pressure measurements. The data obtained are summarized in the construction of a I-atm-pressure phase diagram and a phase diagram corresponding to a pressure of less than 10 atm. THE purpose of this investigation was to establish the phase diagram for the zinc-zirconium system. Thermal, metallographic, vapor pressure, and X-ray data were employed in determining the phase regions. Partial investigations of this system have been conducted by Gebhardt1 and Carlson and Borders.' Carlson and Borders studied the high-zirconium region and established the existence of a eutectic at 69 wt pct Zr with a melting point of 1015°C. The terminal phases of the eutectic horizontal were shown to be an intermetallic compound ZrZn and a solid solution of ß zirconium containing 21 wt pct Zn. The ß solid solution decomposes into ZrZn and a zirconium at 750°C. The eutectoid composition is given as 15 wt pct Zn, and the solubility of zinc in a zirconium at temperatures below 750°C is indicated to be negligible. Gebhardt studied the zinc-rich region and observed a lowering of the melting point of zinc from 419.5" to 416°C and temperature horizontals at 545" and970°C. Some preliminary observations by Chiotti, Ratliff, and Kilp were reported by Hayes.2 pietrokowsky3 has reported the compound ZrZn2 to have a cubic MgCu2 structure with ao = 7.396A. MATERIALS AND EXPERIMENTAL PROCEDURES The metals employed in the preparation of alloys were Bunker Hill slab zinc or Baker analyzed reagent granulated zinc, both 99.99 pct pure and hafnium-free iodide-process crystal bar zirconium obtained from the Westinghouse Electric Corp. The zirconium contained 200 ppm Fe, 200 ppm Si, 100 ppm C, and minor amounts of other impurities. The zirconium was milled or machined into thin chips or shavings. These were cleaned with a nitric-hydrofluoric acid solution, rinsed with water, and acetone, and dried just prior to their use in alloy preparation. The granulated zinc was similarly cleaned using dilute nitric or hydrochloric acid. Weighed quantities of these materials, 20 to 30 g total, were mixed and pressed at 20,000 to 70,000 psi to give relatively dense compacts. During the early part of this investigation the pressed compacts were placed in MgO-15 wt pct MgF, crucibles which were then sealed inside of quartz ampules. The compacts were given various prolonged heat treatments prior to their use for thermal analyses, or vapor-pressure measurements. Because of expansion of the compacts and the relatively high zinc vapor pressure it was difficult to heat to the melting temperatures of the alloys without failure of the quartz ampules. Homogenization at temperatures below the melting temperature gave brittle, porous alloys unsuitable for metallographic examination. It was also difficult to prevent condensation and segregation of zinc on the colder parts of the quartz ampules during heating and cooling operations. These problems were eliminated to a great extent by the use of tantalum crucibles. Tantalum proved to be a satisfactory container with little or no reaction between the alloys and the tantalum. Small tantalum thermocouple wells were successfully welded in the bottom of these crucibles. Pressed compacts were sealed inside the tantalum crucibles by welding on preformed caps under an argon atmosphere. Heat treating and differential thermal analysis were combined into a single operation. The experimental sample assembly is shown in Fig. 1. This assembly was enclosed inside a stainless-steel tube heating chamber which could be evacuated and filled with an inert gas. The thermocouple leads were brought out of the heating chamber between two rubber gaskets used to provide a vacuum seal for the water-cooled head. Most of the compounds in this system undergo peritectic decomposition. After heating above the temperature of a particular peritectic horizontal the sample was cooled to just below the peritectic temperature and held at temperature for several hours. The sample was then reheated through the peritectic temperature and the size of the thermal arrest, if still present, compared with the one previously obtained. If the thermal arrest was not characteristic for the alloy composition being investigated its magnitude diminished and repeated cycling and annealing eventually eliminated it. The peritectic thermal arrests characteristic of a particular composition were established in this manner.

Jan 1, 1960

By G. W. King, A. Bogrow, L. F. Marek

The purpose of this paper is to present the data and some interpretation of the results of a laboratory study of the reduction of iron ore and the deposition of carbon from the reducing gas mixtures in such reductions. One major object of the study was to determine the effect of gas pressures on these chemical processes. Effort was made to approximate the conditions of temperature, gas composition, linear and mass gas velocities, pressure range, and extent of reduction prevailing in commercial blast furnaces. The results show that the reduction of Fe203 proceeds by two steps, largely consecutive under the conditions used herein: reduction of Fe203 to FeO, and reduction of FeO to Fe. Unless the gas has a CO!CO, ratio greater than 2.1 only the first step occurs under these conditions (850°C) but the following generalizations apply to either step or to the combination. The rate of reduction of relatively fine ore particles .at any time is proportional to the mass rate of supply of CO (provided some of it is thermodynamically available): the rate therefore is proportional to the linear velocity at constant pressure, or to the pressure at constant linear velocity. Thus, the rate of reduction may be increased without increase in linear velocity of the reducing gas by increasing the mass velocity of the gas and increasing the pressure correspondingly to avoid any increase in the linear velocity. The rate "constants"—i.e., the efficiency of conversion of the "available" CO supplied— is not actually constant but decreases as the reduction proceeds, and is a function only of the state of the ore; i.e., of the amount of reduction already accomplished and hence of the amount of CO passed. The particle size of ore used in nearly all the experiments was relatively small (8 to 14-mesh). The rate of reduction in run 16 with 3 to 8-mesh ore particles is within experimental error the same as with the finer ore. Although this would indicate an absence of effect due to particle size in this range of ore particles, it is desirable to have more data on other ore sizes and on ore of different densities before arriving at any generalizations on this point. Although it is possible to extend the analysis of these results to consideration of the more detailed kinetics and mechanism of iron oxide reduction, such extension is beyond the scope of the present paper. From the experiments on carbon deposition, it was evident that if carbon deposition is increased at all by the application of pressures up to 10 psi ga., such increase appears to be less than the experimental and analytical error. It appears safe to conclude that at top pressures on the order

Jan 1, 1948

By L. F. Marek, G. W. King, A. Bogrow

The purpose of this paper is to present the data and some interpretation of the results of a laboratory study of the reduction of iron ore and the deposition of carbon from the reducing gas mixtures in such reductions. One major object of the study was to determine the effect of gas pressures on these chemical processes. Effort was made to approximate the conditions of temperature, gas composition, linear and mass gas velocities, pressure range, and extent of reduction prevailing in commercial blast furnaces. The results show that the reduction of Fe203 proceeds by two steps, largely consecutive under the conditions used herein: reduction of Fe203 to FeO, and reduction of FeO to Fe. Unless the gas has a CO!CO, ratio greater than 2.1 only the first step occurs under these conditions (850°C) but the following generalizations apply to either step or to the combination. The rate of reduction of relatively fine ore particles .at any time is proportional to the mass rate of supply of CO (provided some of it is thermodynamically available): the rate therefore is proportional to the linear velocity at constant pressure, or to the pressure at constant linear velocity. Thus, the rate of reduction may be increased without increase in linear velocity of the reducing gas by increasing the mass velocity of the gas and increasing the pressure correspondingly to avoid any increase in the linear velocity. The rate "constants"—i.e., the efficiency of conversion of the "available" CO supplied— is not actually constant but decreases as the reduction proceeds, and is a function only of the state of the ore; i.e., of the amount of reduction already accomplished and hence of the amount of CO passed. The particle size of ore used in nearly all the experiments was relatively small (8 to 14-mesh). The rate of reduction in run 16 with 3 to 8-mesh ore particles is within experimental error the same as with the finer ore. Although this would indicate an absence of effect due to particle size in this range of ore particles, it is desirable to have more data on other ore sizes and on ore of different densities before arriving at any generalizations on this point. Although it is possible to extend the analysis of these results to consideration of the more detailed kinetics and mechanism of iron oxide reduction, such extension is beyond the scope of the present paper. From the experiments on carbon deposition, it was evident that if carbon deposition is increased at all by the application of pressures up to 10 psi ga., such increase appears to be less than the experimental and analytical error. It appears safe to conclude that at top pressures on the order

Jan 1, 1948

By L. D. Jaffe, J. H. Hollomon

The effect of tempering temperature and time upon the properties of quenched steel is clearly a subject of great practical importance, as well as of considerable theoretical interest. It would be very desirable to be able to predict the properties of a quenched steel when given any selected tempering cycle, with a minimum of experimental work upon that heat or even type of steel. Two problems are involved: one consists of relating the structure obtained to the physical properties, and will not be discussed here; the other, with which this paper deals, involves the relations of time, temperature and composition with the structure obtained 011 tempering. Indentation hardness is perhaps the most convenient quantitative measure of the tempered structure, or of what might be called the "degree of tempering." Undoubtedly, it is not an exact measure, and in some cases (as when temper-brittleness occurs) pieces of the same quenched and tempered steel may have widely differing properties at the same hardness. However, hardness has the advantage of being simple to determine and is certainly the most widely used indication of the degree of tempering. Moreover, for fully quenched steels that are not temper-brittle and do not contain retained austenite when quenched, it appears that, in general, if the hardness of the steel is the same, the tensile and impact properties will also be identical, whether a high or a low tempering temperature is used. The tempering charts available in the metallurgical literature usually give the hardness as a function of tempering temperature for only one tempering time. However, in practice, tempering times frequently vary, and a method of converting tempering curves for one time to tempering curves for some other time would be very valuable. It is also advantageous, in determining the tempering treatment necessary for a given part, not to have to temper samples at several temperatures for the time used for the commercial tempering operation. A more satisfactory procedure would be to employ times of the order of 10 to 30 min., rather than 5 or 10 hr., and if possible to use only one temperature. Furthermore, it is frequently desirable, for reasons such as temper-brittleness,* to obtain a given hardness with short times at relatively high temperatures, and it would be convenient to be able to compute without tests the time required to produce this hardness. Remarkably few quantitative studies of the hardness of quenched steel tempered over wide ranges of temperature and time

Jan 1, 1945

By Howard Scott

Invar is the preeminent low-expansion metal by virtue of the fact that it can be prepared with a zero coefficient of expansion at atmospheric temperature. This fact suggests that there is little room for improvement in the expansion properties of the low-expansion nickel steels. The inference is true, however, only so far as the expansion at and near atmospheric temperature is concerned, because the expansivity of invar starts to increase rapidly on heating above 100" C. and soon reaches the value of copper. Guillaumel met this situation by increasing the nickel content above that of invar. By that expedient, alloys having low expansivity under higher temperatures were obtained, but not without a large increase in minimum expansivity. The increase in minimum expansivity with increasing nickel content severely limits the technical applications of these alloys at high temperatures. Prospects of improving the expansion characteristics of the nickel steels is offered by alloying. Search for an advantageous alloying addition by analogy with other alloy systems, however, is not productive because the low-expansion characteristic is unique with that system. The rule of mixtures usually gives a fair value of the expansivity of other solid-solution alloys, but in the iron-nickel systems gives a value of 18 X 10-6 per deg. C. whereas the actual value is 1 X 10-6 per deg. C. Apparently suitable additions are not to be deduced from such ingenuous considerations. Recognition of the fact that the temperature range of low expansivity in nickel steels is terminated with loss of magnetism affords a more promising basis for the selection of an advantageous element. One may expect from that fact that additions of nonmagnetic elements to the iron-nickel alloys are detrimental to the low-expansion characteristic. Indeed, carbon, manganese, silicon, copper and chromium are detrimental to the expansion properties and no nonmagnetic element is reported to be beneficial. From the same viewpoint, the addition of a magnetic metal offers prospects of improvement. As two of the three ferromagnetic elements are already present in the

Jan 1, 1930

By J. B. Cheatham, C. E. Murphey

Presently, computer control of Borobolic direction cannot be obtained during drilling, and most straight-holc drilling methods attempt to resist hole deviation rather than control direction. Many of the theories advanced as possible explanations of the cause of hole deviation are Summarized berein. A new correlation of physical partables is introduced to indicate bow factors such iis drill collar stiffness clearance and hit weight influence borehole deviation, .A methord is proposed for predicting the rate of change of hole angle when drilling conditions are changed. INTRODUCTION Control of borehole direction during drilling can be difficult and costly. Unintentional crooked holes are often lrilleti in dipping formations and many times directional drilling is required when the surface location is not directly above the target area —- for example, at offshore and mountainous locations. Prilllng progress can be greatly hindered in either air or liquid drilling when it be comes necessary to use low bit weight to prevent excessive hole angle build-up. If hole inclination becomes too great, drill pipe drag becomes excessive and fishing risks are increased, logging is more difficult and problems of differential sticking, key seating and fatigue failures may be encountered. Dog-legs and key seats are more serious problems than steep inclination angles: therefore, reducing rate of direction change is preferred to attempting to maintain absolutely vertical holes. Consequently, a straight inclineti hole is preferable to a nearly vertical crooked hole containing numerous dog-legs. In this paper, theories of the cause of hole deviation and analyses of drill string behavior under down-hole conclitions are summarized. Methods for computing hole deviation are presented und systems for resisting deviation as well as neans for provic!ing control of holt- direction are iliscusseI. A new correlation of physical variables is introduced tu indicate how factors such as drill collar stiffness, clearance and bit weight influence borehole deviation. A method is oroposeti for predicting the rate of change of hole anFrle when drilling conditions are clianged. IIEVIEU' OF PREVIOUS WORK ON HOIaE DEVIATION Significant progress in the theoretical analysis of hole deviation problems has been made in the past 1 5 years. The pioneering work has been primarily a result of the efforts of Lubinski and Woods.l-5 In 1950, Lubinski 1 considered the buckling of a drill string in a straig:rt vertical hole, a problem also considered by Willers6 in 1941. It wns concluded that very low bit weights must be used to prevent hole deviation resulting from drill collar buckling. The use of conventional stabilizers was proposed2 in 1951 by Mac Donald and Lubinski as a method for permitting greater Sit weights to he carried without drill collar buckling. These authors pointed out that a 2° nearly vertical spiral hole can cause severe key seatinp, and drill pine near, wilereas n 3 straight inclined hole with deviation all in one direction, while not vertical, will not result in serious drillirig or producing problems. Studies were continued with an investigation of straight inclined holes by Lubinski and Woods3 in 1953. In this paper they concluded that perfectlv vertical holes cannot be drilled even in isotropic forriations unless extremely low bit P-eights are used. .l'l~ey postulated that constant drilling conditions produce holes of constant inclination angle and varying conditions cause the hole to drill at a neiv equilibrium angle. This analysis was not concerned with drill string buckling since it was based on an equilibrium solution in which the ~irill string Was presumed to lie along the lower side of the hole abcrve the point of tangency. Weight of the drill collars below the point of tangency tends to force the hole toyard the vertical, whereas the weight on hit tends to force the hole aurav from the vertica l. The concept of an anisotropic formation Lvas introduced as an empirical method for explaining actual drilling data and as a means for extrapolating known deviation data to otiier conditions of hit

Jan 1, 1967

By H. W. Hockin, D. r. Brandt, R. H. Walsh, P. L. Dietz, P. R. Girardot

New Jersey, Florida, and Canadian ilmenites were reduced with hydrogen or coke under various experimental conditions and the phase changes occurring in the ilmenite upon reduction have been studied by microscopic examination of polished sections and by X-ray diffraction. The products formed were dependent upon the type of ilmenite, temperature, time and reducing agent. Of the reducing agents, hydrogen was the more effective at lower temperatures. 1 HE possible utilization of ilmenite as a source of both iron and titanium has resulted in the development of a number of methods for the separation of the iron and titanium content. Slagging processes as currently used in Canada are typical of such methods. Somewhat less attention has been given industrially to the reduction of the iron content at less than the slagging temperature. Although references maybe found to such work, in general only one type of ilmenite, either natural or synthetic, has been studied by each author. We have attempted to draw some relationship between the results of experimental reduction and the type of deposit from which the ilmenite was derived, as evidenced by phases occurring in the ores and in the reduction products. Ilmenite ores having from 27 to 61 pct titanium dioxide were included in this study and in each case reduction was carried out at such temperatures that only limited coalescence of the particulate iron product occurred within the ilmenite grains. The history of the individual ilmenite samples and the temperature of reduction were found to determine the occurrence of various phases and the mode of distribution of the iron. REVIEW OF EARLIER WORK ON REDUCTION One of the earliest references to the reduction of ilmenite at less than the slagging temperature was made in 1917.' The metallic iron produced was_____ leached out by the action of dilute sulfuric or hydrochloric acid, to leave a product high in titanium dioxide. Subsequent to this, numerous patents2-'? and other references13"21 have appeared concerning reduction of ilmenite by carbon, hydrogen, carbon monoxide, methane, coal gas, water gas, or the like in the absence of fluxing agents. Mineragraphic examinations were not reported, and in the main, the only examinations made were chemical analyses. However, in two cases3,13, titanium suboxides were reported in the products and in one case2' rutile was reported, in addition to metallic iron. Both were identified by X-ray diffraction. While reductions in the absence of fluxing agents were generally followed by either a wet or dry chemical process for removal of the metallic iron, a Canadian source17 reported removal of the metallic iron by magnetic separation. By the addition of fluxing agents during reduction, the metallic iron has been coalesced into "pearls." Each reference to such a process23-28 has shown that the coalesced iron could be removed by magnetic or gravity means after the product was crushed. In the absence of fluxes, the coalescence did not generally occur. The major part of the present study has been limited to reductions without fluxing agents in order to determine the primary reactions of naturally occurring ilmenites. Titaniferous iron ores have also been studied,29-33 17, l9 their reduction having been used as the basis of a commercial process for beneficiation of such ores in Norway29 and more recently considered in the United States34,35. The ease of reduction of titaniferous iron ores relative to that of hematite or magnetite has been referred to,33 with the conclusion that the more

Jan 1, 1961

By G. W. Healy, N. R. Griffing

The activity coefficient of sulfur in Cr-Fe-S melts was determined by measuring the values of Ph3Rh,in equilibrium with such melts. The results showed that chromium has a pronounced negative effect on the activity coefficient of sulfur, which up to about 20 pct Cr may be represented by the equation: log f; = - 0.019 pct Cr It was also found that the influence of chromium was not altered by the presence of appreciable amounts of carbon or silicon. The influence of alloying elements on the activity of sulfur in molten iron has been the subject of many investigations; however the effect of one important alloying element, chromium, has not been reported, and the object of the present investigation was to determine its effect on the activity coefficient of sulfur in liquid Fe-Cr-S alloys both in the absence and in the presence of carbon and/or silicon. EXPERIMENTAL PROCEDURE The general method employed consisted of passing mixtures of hydrogen and hydrogen sulfide over the sulfur-containing melt of interest and analyzing the outgoing gas to determine the equilibrium value of which resulted from the reaction shown in Eq. 111. H, (gas) + S (dissolved in melt) = H,S (gas) 111 The activity coefficient of sulfur, fs, was then calculated with Eq. [2] in which K, is the known equilibrium constant of Eq. [I] for dilute solutions of sulfur in iron. Raw materials used in the preparation of the charges included commercial grades of electrolytic iron and chromium and carbon-reduced, acid-treated silicon of 99.9 pct purity. Oxygen in the former presented no problem since it was removed by hydrogen in the normal course of an experiment. Carbon was added in the form of premelted iron-carbon and chromium-carbon master alloys. Sulfur was charged as iron sulfide prepared by first heating powdered iron and sulfur in air in an alumina dish until reaction occurred, and then deoxidizing by melting in a hydrogen atmosphere. Commercially bottled hydrogen and hydrogen sulfide were used after being passed through purification systems containing palladium, "Drierite," and magnesium perchlorate in the case of hydrogen, and "Drierite" alone in the case of hydrogen sulfide. Lamp-grade cylinder argon served as a flushing agent without further purification. Charges weighing from 300 to 500 g were placed in a high-purity alumina crucible of 0250-ml capacity and heated in an argon atmosphere in the induction-heated carbon-tube furnace illustrated in Fig. 1. During heating, a dry hydrogen atmosphere was substituted for argon in order that deoxidation of the charge and flushing out of argon would be accomplished sooner. After the charge melted, a Pt-Pt 10 pct Rh thermocouple contained in an alumina protection tube was immersed in the melt, and the desired melt temperature (usually 1600°C) was established. Higher temperatures up to 1760°C were required for some chromium-rich alloys because of their higher melting points, and in these cases a Pt 6 pct Rh-Pt 30 pct Rh thermocouple was used. The thermocouple output was fed to an automatic controller which provided straight line temperature

Jan 1, 1961

By David B. Reger

The expansion of the previously discovered gas pools and an intensified search for additional pools in the Oriskany and deeper sands were the principal petroleum activities in West Virginia during 1941. One small oil pool was discovered and another was considerably extended. One small gas pool was discovered, several isolated well discoveries were expanded into definite pools, and various recognized pools were substantially increased. The proved oil territory of the state was increased about 600 acres and the proved gas territory about 45,000 acres. The new discoveries of oil did not equal withdrawals but gas reserves added were greater than gas withdrawals. Oil activity, as in the past several years, was slight because of unfavorable crude prices. Most operators prefer to leave oil in the ground rather than to drill for it without a better return. Gas activity increased greatly because of better demand, even though the price of gas did not improve. The account of operations, as gathered from trade journals and other reporting services, shows that 838 new wells were drilled, resulting in 114 new oil wells with 2606 bbl. of daily new production; 558 new gas wells with 949,226,000 cu. ft. of daily open flow; and 166 dry holes. Also, 86 old wells were drilled to deeper sands with 9 bbl. and 56,030,000 cu. ft. of added production. On the new wells the oil average was 22.86 bbl. per well per day; and the gas average was 1,701,000 cu. ft. per well per day. On new wells the ratio of dry holes to completions was 13.61 per cent. On deeper drilling to other sands the ratio of failures was 15.12 per cent. From an exploratory standpoint one strictly new oil pool (Higginbotham Run) was opened in Kanawha County and one strictly new gas pool (Lost Run) in Taylor County. One entirely isolated wildcat gave commercial Oriskany-sand gas production and may open a pool. Various other wildcats greatly extended the boundaries of known Oriskany-sand gas pools and various others may have opened new pools in the shallow sands. On the other hand, various Oriskany tests in Harrison, Monongalia, Roane and Wood Counties were dry and all five tests to the White Clinton sand (Sil.) in Boone, Harrison, Kanawha and Wood Counties failed to find commercial production. The principal oil wells completed, by counties, were: Calhoun, 24; Clay, 7; Gilmer, 6; Hancock, 9; Kanawha, 9; Pleasants, 10; and Ritchie, 14. Production for the year was estimated by the Oil Weekly as 3,421,000 bbl., compared with U. S. Bureau of Mines final figures of 3,444,000 bbl. in 1940. The leading counties in gas-well completions were: Boone, 32; Braxton, 19; Cabell, 16; Calhoun, 46; Clay, 20; Gilmer, 37; Jackson, 94; Kanawha, 82; Lincoln, 17; Putnam, 46; Ritchie, 33; Upshur, 20; and Wayne, 33. Production for the year is estimated by the author as 210,000,000,000 cu. ft., as compared to the U. S. Bureau of Mines' final figures of 188,751,000,000 cu. ft.* in 1940.

Jan 1, 1942

By David B. Reger

The expansion of the previously discovered gas pools and an intensified search for additional pools in the Oriskany and deeper sands were the principal petroleum activities in West Virginia during 1941. One small oil pool was discovered and another was considerably extended. One small gas pool was discovered, several isolated well discoveries were expanded into definite pools, and various recognized pools were substantially increased. The proved oil territory of the state was increased about 600 acres and the proved gas territory about 45,000 acres. The new discoveries of oil did not equal withdrawals but gas reserves added were greater than gas withdrawals. Oil activity, as in the past several years, was slight because of unfavorable crude prices. Most operators prefer to leave oil in the ground rather than to drill for it without a better return. Gas activity increased greatly because of better demand, even though the price of gas did not improve. The account of operations, as gathered from trade journals and other reporting services, shows that 838 new wells were drilled, resulting in 114 new oil wells with 2606 bbl. of daily new production; 558 new gas wells with 949,226,000 cu. ft. of daily open flow; and 166 dry holes. Also, 86 old wells were drilled to deeper sands with 9 bbl. and 56,030,000 cu. ft. of added production. On the new wells the oil average was 22.86 bbl. per well per day; and the gas average was 1,701,000 cu. ft. per well per day. On new wells the ratio of dry holes to completions was 13.61 per cent. On deeper drilling to other sands the ratio of failures was 15.12 per cent. From an exploratory standpoint one strictly new oil pool (Higginbotham Run) was opened in Kanawha County and one strictly new gas pool (Lost Run) in Taylor County. One entirely isolated wildcat gave commercial Oriskany-sand gas production and may open a pool. Various other wildcats greatly extended the boundaries of known Oriskany-sand gas pools and various others may have opened new pools in the shallow sands. On the other hand, various Oriskany tests in Harrison, Monongalia, Roane and Wood Counties were dry and all five tests to the White Clinton sand (Sil.) in Boone, Harrison, Kanawha and Wood Counties failed to find commercial production. The principal oil wells completed, by counties, were: Calhoun, 24; Clay, 7; Gilmer, 6; Hancock, 9; Kanawha, 9; Pleasants, 10; and Ritchie, 14. Production for the year was estimated by the Oil Weekly as 3,421,000 bbl., compared with U. S. Bureau of Mines final figures of 3,444,000 bbl. in 1940. The leading counties in gas-well completions were: Boone, 32; Braxton, 19; Cabell, 16; Calhoun, 46; Clay, 20; Gilmer, 37; Jackson, 94; Kanawha, 82; Lincoln, 17; Putnam, 46; Ritchie, 33; Upshur, 20; and Wayne, 33. Production for the year is estimated by the author as 210,000,000,000 cu. ft., as compared to the U. S. Bureau of Mines' final figures of 188,751,000,000 cu. ft.* in 1940.

Jan 1, 1942

By R. W. Snyder, H. J. Ramey

This paper presents the results of applying the Buckley-Leverett' displacement theory to petroleum reservoirs consisting of a finite number of layers. The layers are assumed to communicate only in the wellbores, and the reservoir may be represented as a linear system. Most previous investigations of this nature were limited by assumptions and by inconsistent calculation techniques. This study improves on previous work by applying the Buckley-Lev-erett displacement theory to a noncommunicating layered system where permeability, porosity, initial saturation, residual saturation and relative permeability vary from layer to layer in a logical and consistent manner. Gravity and capillary-pressure effects are neglected. A modification of the Higgins-Leighton calculation method was used in this study. Waterflood predictions were made with all properties varying, and then with only permeability varying using several inability ratios. These results were compared with the Stiles and Dykstra-Parsons predictions. It is shown that the latter methods generally give poor values for the breakthrough recovery and pessimistic predictions for the performance after breakthrough. Similar results were obtained for a gas-displacement case. lNTRODUCTION Field experience with immiscible displacement usually shows constant producing conditions until breakthrough of the displacing fluid. Then oil production continues at increasing displacing-to-displaced fluid ratios until the economic limit is reached. Three different ideal mechanisms are known that will produce this behavior: (1) relative permeability effects as described by Buckley-Leverett frontal advance theory,' (2) vertical stratification as considered by Stiles,2 Dykstra and Parsons5 and others and (3) different path lengths involved in areal (two-dimensional) flow between wells as described by Dyes et al.4 Without question, a combination of these factors modified by formation heterogeneity and other known and unknown factors actually does control the behavior of real systems. This paper presents results of an investigation of certain factors that should affect performance but which have received little attention to date. In 1944, Law5 demonstrated that porosity and perme- ability are often found to have normal and logarithmic-normal distributions, respectively. throughout cored intervals in natural formations. This led to the concept of the noncommunicating, multilayered reservoir model for immiscible displacement. This model assumes that the reservoir is composed of a number of layers that communicate only at the wellbores. Each layer is individually homogeneous, but may be different from every other layer. Stiles' presented one of the earliest applications of this model to waterflood performance. In addition, Stiles assumed that the initial saturations and relative permeabilities were the same for each layer, porosity was the same. displacement was piston-like, fluids were incompressible and injection into each layer was proportional to that layer's permeability capacity (permeability-thickness product). The last assumption would be true if the mobility ratio for the displacement were unity.21 Dykstra and Parsons" used the same model as Stiles, but rigorously included mobility ratios other than unity for piston-like displacement. Dykstra and Parsons used their general result to produce charts for log-normal permeability distributions between layers. Similarly, Muskat6 Pub1ished analytical solutions for linear and exponential permeability distributions. In 1959, Roberts' described a scheme for calculating water-drive performance for the noncommunicating, layered reservoir model which considered two-phase flow in the displaced region. Roberts used the same model and assumed that the injection rate into a layer was proportional to that layer's permeability capacity, but that flood front locations could be evaluated from the Dykstra-Parsons results. These assumptions are inconsistent, and a material balance cannot be maintained except for a mobility ratio of unity. At the same time, Kufus and Lynch8 coupled Buckley-Leverett displacement theory with the layered model to provide an improvement of the Dykstra-Parsons method that was consistent. In 1960, Higgins and Leighton9 resented a numerical method for calculating waterflood performance also considering two-phase flow in the displaced region. The result was used to investigate variation in absolute permeability and oil viscosity. An excellent, detailed history of using the noncomrnunicating, layered reservoir model was presented by Nielsen.'" The preceding techniques (and many related ones) were similar in that differences in initial saturations, residual saturations and relative permeabilities from layer to layer were neglected. It is well known that the irreducible water saturation is an important function of absolute permeability. Calhoun11 showed that the irreducible water saturation

By John Brownson

Ge-Si N-N heterodiodes hare been built recently which show promise as high-speed logic devices. Low-resistivity germanium is deposited on silicon substrates held at temperatures above the germanium melting point and an alloyed heterojunction formed. The diodes are fabricated from this material using conventional mesa techniques. Over-all device qualify has been greatly improved by the use of epitaxial silicon substrates, that is, a subsirale consisting of a film of thin, high -resistivity silicon on a thicker Piece of low-resistirily silicon crystal. A semimpirical device design equation has been devised which predicts switching performance fairly well within the range of resistivitics and geometries employed. Switching times 01 0.9 ns and PIV 's of —20 v are typical of the better 10-ma diodes. Switching times as low as 0.5 ns have been observed for 10-ma diodes and as low as 2.8 ns for 200-ma diodes. With further development it should be possible to improve switching speed by a factor of four. IN our laboratory, Ge-Si heterodiodes have been built recently which show promise as high-speed logic devices. The diodes have been built using a process reported in some detail in an earlier publication.' The process in short involves alloying a thin film of low-resistivity germanium into a silicon substrate. the germanium having been transported and deposited by the well-known Theuerer halide reduction process.2 This process contrasts with that reported earlier by Oldham who used an essentially conventional direct epitaxial deposition technique.3 Oldham's process, unlike ours, requires cleavage of the silicon sample inside the reactor furnace the instant prior to deposition. Our process has the advantage of yielding large-area macro-scopically plane heterojunctions. Heterodiodes in general have several interesting properties. Without attempting to summarize the theory of heterojunctions.4 two of these properties which have direct relevance should be mentioned. First, N-.V and P-P heterojunction diodes rectify alternating current. Provided that either the band gap or the electron affinity of the two materials is different (which in general is the case), electronic barriers will exist in the junction band structure. When the device is sufficiently forward-biased, the barriers shift so as to permit current flow, and when it is reverse-biased. the barriers block current flow. Second, for N-N and P-P heterodiodes. the foward-conduction process involves majority carriers only. Consequently. when such a device is turned off. there is no minority-carrier storage and hence the diodes switch on and off quite rapidly. As we have reported earlier. N-P and N-N Ge-Si heterodiodes have been built in this laboratory. As one would predict from theory, the N-P heterodiodes were rather slow in turning off while the N-N's were quite fast. Similarly, the turn-off time of the N-P's increased rapidly with forward current while that of the N-N's was independent of forward current. Both N-N's and N-P's were moderately photosensitive. Photocurrents resulting from normal room illumination were as high as 1 µa at 1 v for small-area devices. The most disappointing parameter of the early N-N devices was reverse leakage.' The reverse characteristic was invariably "soft" and leakage at only several volts reverse bias was intolerable. The reverse characteristic was greatly improved by using higher-resistivity silicon substrates. Leakage becomes large at voltages about one third of the avalanche voltage one would expect of a P-N homo-junction formed on silicon of the same resistivity. Thus leakage could be reduced to reasonable limits by using unconventionally high-resistivity silicon. Medium-power diodes formed on 30 ohm-cm substrates leaked typically 15 to 50 µa at 100 v reverse bias. The use of high-resistivity silicon substrate material, however, created a new problem. In order to have an acceptable forward-conductance characteristic. the area of the diodes had to be increased. With this increase in area, the capacitance increased. Switching time for N-N heterodiodes is essentially a linear function of capacitance. Thus when the resistivity was raised. so was the switching time. A SEMIEMPIRICAL HETERODIODE DESIGN EQUATION As a result of our experiments with a limited range of diode geometries and resistivities. a semi-

Jan 1, 1965

By Rudolph G. Wuerker

MUCH descriptive matter has appeared on the subject of suspension roof supports, or roof bolting, as it is more commonly called. The widespread introduction of roof bolting into coal mines and metal mines is truly phenomenal. Mine operators were quick to recognize the advantages of supporting wide openings without hindrance to machine maneuverability and ventilation. Although suspension roof support has long been installed at St. Joseph Lead Co. mines in southeast Missouri,'" its application to coal mining presented new problems, such as proper anchorage and bearing for the bolts, bolt diameter, and spacing of bolts. After continuous testing and experimenting at the mines, standard roof-bolting materials were determined.'!' The study reported in this paper is not concerned with such details as bolt diameter, which may be considered already solved in practice. In the tests discussed here, small models patterned on actual bolts were found to function in the same way and as satisfactorily as their prototypes. The aim of these tests was rather to investigate the influence of roof-bolting systems on the stress distribution around mine openings and to study the fracture patterns obtained in actual testing. Little was found about this in the literature, as testing of suspension roof methods and quantitative measurements are only now coming to the fore. Several suggestions and actual measurements have been made to evaluate critically the functioning of roof bolting systems, single roof bolts, and parts thereof. Outstanding among them is Bucky's outline of structural model tests.'" Since none of the suggested testing equipment was available, however, for the experiments discussed below, a different approach was chosen. The response of a mine roof under stress has often been compared to that of a beam. The slow coming down and bending through of beam or plate-like banks of shale, sandstone, or top coal is a familiar occurrence, extensively cited in the literature." It was felt that testing of roof-bolt systems installed in a concrete beam which was loaded in bending would be a fair approximation of the behavior of a mine roof underground. Another school of thought considers the roof behavior over an underground opening in connection with the stress distribution all around a circular or rectangular opening. This is more accurate, and leads to the concept of a dome-shaped zone of material destroyed under tensile stress. This is likewise a common sight in unsupported roadways where the continuous fall of roof results in what has been called the natural outline of roof fracture. This theory could not be tested and is treated separately in Appendix B. It is important to note that according to both assumptions the immediate roof fails in tension; the use of a beam in these tests, therefore, should give information valid for either of the two theories. With the testing equipment at hand it was possible to load concrete beams 6xlx0.5 ft under two-point loading, giving an equal bending moment over the center part in which the model bolts were installed. A comparison was made of the ultimate loads needed to break plain beams and beams in which roof bolts were installed. Arrangements were made with: 1—plain beams; 2—bolts with plate washers, some with holes drilled at 90" angles and others with holes drilled at 45" angles; 3—bolts with channel irons underneath; 4—bolts in holes filled afterward with cement; and 5—bolts anchored in a stronger stratum. The foregoing arrangement is made in order of increasing strength, as assumed from the theory of reinforced concrete. Likewise, laminated beams with wooden model bolts and with combinations of the foregoing set-ups were tested. All in all, 21 experiments were made out of the much greater number of combinations possible. There were, too, some trial tests. Enough observations from this limited number were made to interpret the behavior of mine roof, supported by various types of suspension bolts, at fracture. In present-day concepts, which have been proved by mathematical derivations and stress analyses, any opening driven underground will change the distribution and magnitude of the stresses existing around it. It does not matter whether the stresses become visible, as in rocks whose strength is less than the forces acting upon them, or whether they are invisible, as in the gangways lacking evidence of rock pressure. In this latter case the rocks can withstand changes in stress-distribution. To consider the mine roof as a beam, there are, with transversal loading, tensile stresses in the lower fiber and compressive stresses in the upper layers above the neutral axis of the beam. Beams of brittle material such as rock and concrete fail exactly as shown in Fig. 1. Nearly all model beams showed the same fracture pattern as that of a tension crack. The influence of support, by roof bolting or conventional

Jan 1, 1954

By J. G. Patchett, R. W. Rausch

Fresh waters and the presence of clay in many Rocky Mountain and West Coast sands require special methods of log analysis. Archie's saturation equation requires addition of a shale correction term, and the SP equation must also be modified to account for clays. Suitable equations were developed several years ago, but have not been widely used due to the algebraic complexity. A computer-oriented method has now been developed to overcome this problem. The basic shaly sand equations are rearranged in four different ways to permit solution for various sets of available input data. Essential to application of the method is the correction of observed SP values to those that would be observed if the resistivity of the formation waters were exactly interchangeable with the activity. A graphic method for doing this is given. Where conditions require consideration of the effect of clay in the sands, the method presented has been found to improve the accuracy of water-saturation determinations. INTRODUCTION Log interpretation in many Rocky Mountain and West Coast basins is complicated by rapid vertical and lateral changes in water resistivity. Calculation of formation water resistivity from the SP curve becomes difficult in zones that contain clay, since changes in SP deflection may be due to changes in either clay content or water salinity. In hydrocarbon-producing reservoirs, the problem is further complicated because hydrocarbon saturation also reduces the SP.1 A log interpretation system using computers has been developed to provide a solution to this problem, based on equations proposed by de Witte.2 Four different simultaneous solutions of de Witte's equations have been made. Each solution method uses a different set of input data as independent variables. Thus, a choice of solution method is possible, depending upon the logs run and the availability of other data. Two of the solutions do not require a knowledge of water resistivity. This system is intended to be used primarily in multiple sandstone-shale sequences of low and moderate resistivities where the principal contaminant in the sandstones is clay. However, where sufficient regional data are available, interpretation in single-zone sandstone reservoirs can also be improved by using the method. THEORY AND HISTORY OF SHALY SAND ANALYSIS The log interpretation formula originally proposed by Archie3 in 1941 is applicable only to rock-fluid systems wherein the rock has negligible electrical conductivity. In 1949, Patnode and Wyllie4 showed that if the rock itself can be considered conductive due to the presence of clay, a different calculation approach is necessary. During the following years, this problem was investigated at great length, as was the related problem of the effect of rock conductivity on the SP.5-11 These investigations established functional relationships between SP, resistivity, water saturation and water resistivity for such a formation. Refs. 2 and 12 provide summaries of these studies. Unfortunately, practical use of these relationships required that water resistivity be known independently from the SP. Although log interpretation methods for rock systems containing clay were proposed at that time,' they were not generally accepted for routine use. There are three principal reasons for this. First, in many field situations involving high-salinity water, rock conductivity may be neglected (even if present) without introducing appreciable error. This may be seen by considering the following expression for waier-saturated rock.' 1/R2=1/R1+1/FRn....(1) where 1/R, is conductivity due to clay. As Rw becomes small, I/FRw becomes much greater than 1/R, which may be neglected. Where 1/R, may be neglected, the sandstone is called clean. If the term may not be neglected, the sandstone is termed dirty or shaly. For resistivity purposes, the classification between clean and shaly sands then depends not only upon the conductivity due to shale in the sand, but also upon the resistivity of the associated water (shale is used here to mean surface condition due to disseminated clay). A sand of given conductivity might safely be treated as clean in association with high-salinity water, but would require shaly sand methods if associated with fresher waters. Shaly sand methods are not required in many areas having saline waters; but in Rocky Mountain and West Coast sands having relatively fresh waters (often more than 0.3 ohm-m resistivity at formation conditions), the shaly sand methods are needed. Errors Rw calculations from the SP due to the presence of shale are likewise related to water salinity. In saline water formations drilled with fresh mud, the ratio of mud filtrate resistivity to water resistivity is high, the SP is large and the presence of shale can introduce large errors in water resistivity calculated by the conventional method. When the resistivity ratio is low, the errors are smaller. At zero SP, no error would result from shale. Thus, from the SP viewpoint, a given rock could be shaly if associated with a saline water, and clean in association with a fresh water, which is the opposite of the resistivity-oriented definition above.

By F. George, W. Tice, M. Salkind

It was found that Al-A13Ni could be readily cold rolled perpendicular to but not parallel to the whiskers. Reductions of more than 98 pct were achieved without cracking by rolling perpendicular to the whiskers, whereas extensive edge cracking was noted after only 15 pct reduction when rolling parallel to the whiskers. The longitudinal and transverse tensile strengths were nearly doubled, and the longitudinal yield strength more than tripled by cold rolling 50 pct in a direction perpendicular to the whiskers. The whiskers exhibited some waviness (elastic bending) as a result of cold rolling, but at very high reductions (greater than 75 pct) whisker fracture and misalignment became significant. A fine dislocation substructure in the matrix consisting of cells attached to the whiskers was pro -duced by cold rolling. Most of- the substructure was readily removed by a 1-hr anneal at 500°C. Cold rolling was found to substantially reduce the thermal stability of the microstructure at 610°C but did not affect the stability at 500°C. FIBER and whisker reinforced composite materials promise significant improvements in properties over conventional materials. Before they find wide use, however, it will be necessary to understand the response of these highly anisotropic materials to common metalworking processes. Most of the nonmetal-lic fiber reinforced materials have very low elongations (a few pct or less) in the direction of fiber alignment. Thus, metalworking techniques such as rolling and forging would not be as broadly applicable to these materials. This investigation was initiated to determine how a composite system consisting of Al3Ni whisker reinforced aluminum responded to rolling, what changes in the microstructure occurred, and the effect of deformation on the mechanical properties. The composite material studied was produced by unidirectional solidification of the A1-Al3Ni eutectic alloy'-7 and consisted of 10 pct by volume of aligned whiskers of Alai in a matrix of aluminum. It should be pointed out that this system is not representative of all composite materials, and the results will therefore not be universally applicable. The A1-Al3Ni system is characterized by: 1) A strong fiber-matrix interfacial bond 2) A ductile matrix 3) A sufficiently low fiber content to allow significant plastic flow between fibers 4) Strong, completely elastic whiskers (tensile strength 400,000 psi, elastic modulus = 20 X 106 psi.1 These factors allow the material to be readily rolled perpendicular to the fibers. If the fiber-matrix bond were not strong, such a weak interface could fail during rolling. A measure of the ability of a composite to be rolled in the transverse direction can be obtained from noting the transverse tensile behavior. In the case of Al-Al3Ni,2 there is considerable ductility (15 to 30 pct). In the case of boron filament reinforced aluminum, for example, the transverse elongation is less than 1 pct,8 and the material could probably not be cold rolled as readily in that direction. EXPERIMENTAL PROCEDURE 3-in. diam ingots of A1-A13Ni eutectic were unidi-rectionally solidified in graphite crucibles. The starting materials consisted of 99.99+ pct pure nickel and aluminum, and the pure eutectic ingots were made with 6.2 wt pct Ni. The unidirectional solidification process (described in detail elsewhere1-3) consists of preparing a master heat of eutectic composition, remelting, and withdrawing the ingot vertically downward through the heat source at a controlled rate so that plane front solidification proceeds upward at a constant velocity. The resulting microstructure consists of 10 pct by volume of whiskers of very high aspect (length to diameter) ratio. The fiber lengths have not been measured because of the difficulty of detecting fiber ends9 but exceeds 104. There is some possibility that the fibers may be continuous within one grain. Flat sheet specimens 2¾ in. sq and approximately 0.2 in. thick containing whiskers parallel to the plane of the sheet and to one edge were used for this study. A1-A13Ni exhibits either a rod-like (high solidification rates) or a blade-like (low solidification rates) whisker morphology,1,3 and both types were studied. Rolling was accomplished using a two-high rolling mill at a speed of approximately 10 fpm. The rolling direction was either parallel to or perpendicular to the direction of growth (direction of whisker alignment). Reductions of from 0.002 to 0.03 in. per pass were used with the most common value being 0.005 in. per pass. Cold rolling of Al-Al3Ni to more than 98 pct reduction in thickness was accomplished with no intermediate anneals. In addition. a series of speci-mens was cold rolled 97 pct with a 1-hr, 500°C anneal in air after each 50 pet reduction. Tensile testing was accomplished using a Tinius-lsen four screw testing machine. Flat sheet specimens + in. wide and between 2 and 2; in. long with the thickness dependent upon rolling reduction, were used. The gage section was in. wide and 1 in. long. Strain was measured using a clip-on LVDT extensome-

Jan 1, 1970

By R. Bakish, C. S. Barnett

IN experiments on tantalum strained in tension, Bechtold did not observe deformation-twinning even at a temperature as low as that of liquid air.' This is an unexpected behavior for a metal of body-centered-cubic structure. Therefore, it is desirable to determine whether the lack of twinning is an inherent characteristic of this particular metal, or whether the reluctance to twin can be overcome under suitable conditions. The following experiments were performed under conditions favorable to twinning and cleavage in their competition with slip, namely, with impact loading on large-grained sheet at low temperatures. The crystallography of cleavage in tantalum sheet was also investigated carefully because of the remarkable ductility of tantalum at low temperatures and because recent tests on oxygen-embrittled tantalum had disclosed unusual cleavage behavior for a metal of this structure: cleavage primarily on (110) planes and only occasionally on {100}.2 The tantalum, supplied by Fansteel, was of 99.9 pet purity, with the chief impurities being iron (0.03 pet) and carbon (0.03 pet). Coarse grains were produced by the strain-anneal method, the anneal being 4 hr at 2000°C in a vacuum of 10-3 mm Hg, followed by furnace cooling. Impact deformation was applied in various ways: a) by hammer blows on an anvil, the specimen and anvil being immersed in liquid nitrogen ( —196°C), b) by driving a precooled center-punch into the sheet, and c) by bending around a crystallographically determined axis with a hammer blow. The impact-deformation produced bands and slip lines on the surface, Fig. 1. The bands have every appearance of being deformation twins, a) They are tilted with respect to the untwinned surface. b) They have the appearance of Neumann bands in iron, thin lens shape, some having serrated edges. c) The orientation in each is different from that of the surrounding matrix and uniform within a given band, see slip lines within them in Fig. 1. d) Upon repolishing and etching, the difference in orientation and the resemblance to Neumann bands is again evident, Fig. 2. (In this micrograph the slight bend in the bands as they cross grain boundaries was noted frequently and results from the high degree of preferred orientation in the sheet.) e) The bands persist after annealing treatments that would have removed them had they been a low-temperature martensitic phase, e.g., after an anneal of 1/2 hr at 900°C. When back-reflection Laue data were plotted for single-surface analysis of several grains, every trace of the bands in each grain could be explained by (112) planes. In addition, two of the bands were located in space by traces on two surfaces, and these also were found to lie on (112) planes within the error of the stereographic plot. Since (1 12) is the anticipated composition plane for deformation twins in body-centered-cubic metals it is felt that the observations listed above are conclusive evidence for the identification of the bands as twins. Similar tests with impact at —77° and at 25°C produced no twinning. It is concluded that the critical shear stress for slip at the rates of straining used was always low enough to prevent the applied stress from reaching the value required for twinning at these temperatures. It was noted that the serrations along the edges of the twin bands exhibited parallel crystal facets. Several similarly oriented notches are visible in Fig. 1, and a remarkable set is shown in Fig. 3. It appears from Fig. 1 and similar areas that notches are frequently associated with the intersection of a slip line with a twin, although not all such intersections are at notches. In some instances a slip line appears to enter the twin, cross it, reflect from the opposite side and recross it, producing a pattern resembling a notch but without producing the notch itself, see Fig. 1. It is suggested that the stress concentration at the end of a slip line is responsible for these details. The strain energy associated with the pile-up of dislocations at a twin boundary may be lessened by untwin-ning or failing to twin a portion of the twin band, or

Jan 1, 1959

By G. H. Kessler

ALTHOUGH considerable attention has been devoted to reaction mechanisms and equilibria of a number of endothermic reactions through which metals or their refractory compounds can be formed upon heated surfaces, only a comparatively small amount of work has been done to develop rate information for these systems. The interestin and very worthwhile work of Holden and Kopelman, Runnalls and Pidgeon,Dijring and Moliere, and Herrick and ICrieble4 is to be noted; however, their work was done under very specific conditions and the rate information which they have developed is not applicable to many systems which have been studied experimentally. That a more generalized approach is needed is shown by the diversity of these systems and of the experimental conditions which they require. These include the preparation of titanium, zirconium, tungsten, molybdenum, chromium, co-lumbium, aluminum, and silicon by decomposition or disproportionation of their halides, in some instances through hydrogen reduction; the preparation of nickel and iron by decomposition of their car-bonyls; and the preparation of copper by decomposition of the acetylacetonate. They may be carried out at temperatures ranging from 100" to over 2000°C and at pressures of from less than one mm Hg to superatmospheric. The physical mechanism of the overall process can be broken down into several steps, any one of which may be rate-determining; and the rate associated with each of these basic physical processes can be predicted. Consequently, by proper application of the present technology of rate processes, it is possible to arrive at a rate of metal deposition in the system of interest and to reach an understanding of the influence of' each major variable upon the deposition rate. It is the purpose of this paper to illustrate how this can be done and to present experimental results for a specific system for comparison with theory. The overall process of deposition of metal upon a heated surface which is immersed in vapors of a compound of the metal can be broken down into the following steps: 1) Transport of the reactant vapors to the surface. 2) Decomposition of the reactant in the immediate vicinity of the surface to a) establish chemical equilibrium if reaction rates are high; or to b) establish a dynamic equilibrium in accordance with the several reaction rates. 3) Diffusion of reaction products away from the vicinity of the surface, with or without simultaneous recombinations to form still other products. The overall process is said to be diffusion controlled if Step 1, or the equivalent Step 3, is slow; it is reaction-rate controlled if Step 2b is slow. When experimental evidence shows the reaction rate to be controlling, the well-known theories of Eyring' and others can be used as a basis for correlating and extrapolating the experimental data. Although these theories give only approximate estimates of reaction rates, they are of considerably greater value in correlating experimental data. Most deposition systems operate at surface temperatures sufficiently high that reaction rates near the surface are very high. Mass-transport rates then become controlling; and it is this regime of diffusion controlled rates which is most often of concern and Which will be considered here. When the system pressure is very low, i.e., in the micron range, then the transport processes are those of molecular movement, and the results of kinetic theory can be applied to specify a rate of arrival of reactant molecules at the heated surface. If the condensation coefficient can be estimated, then the deposition rate will be given by the product (rate of arrival) x (condensation coefficient), Conversely, experimental data can be used to determine a condensation coefficient, as was done by Holden and Kopelman' for the decomposition of zirconium tetraiodide upon a heated molybdenum surface, who found the coefficient for zirconium on molybdenum to be 1.0 at temperatures of 1382°C and above. In this pressure range, the geometry of the system. and in particular the distance between the source of reactant molecules and the heated surface, can exert a considerable effect upon the deposition rate;

Jan 1, 1961

By R. V. Higgins, A. J. Leighton

Interest by petroleum engineers in the flow of three phases—oil, gas and water—in irregularly bounded porous media lies mostly in the performance calculation of water floods of reservoirs that have been partially depleted as the result of expansion of much of the originally dissolved gas. The authors present a method to forecast three-phase flow in complex geometry and explain the details by the use of a specific example of a five-spot water flood of a partially depleted stratified reservoir. In this example, the fluid and rock properties of a field given by Prats, et al,' were used. The computed results fit the field performance more accurately than Prats, et al, and Slider.' The time required for the high-speed digital computer to make the calculations, including the contributions from the different layered zones, is about one minute. INTRODUCTION The declining rate of discoveries of new oil fields in the United States makes the recovery of more oil from known reservoirs more attractive than previously. Water flooding has an excellent proved background for recovering additional oil economically. Accordingly, the main interest of this paper is in this type of recovery. In petroleum engineering studies, commercial interest in three-phase flow is mostly in the water flooding of reservoirs in which the oil has been partially produced by the expansion of dissolved gas. When water is pumped into these reservoirs, three-phase flow takes place. Although this paper is concerned with these conditions, the principles involved could be used for other conditions when and where they come to the fore. Many investigators have made contributions to non-empirical forecast methods using basic scientific engineering principles. Several of these used the oil in place at the start of the water flood and the oil remaining after a large quantity of water has passed through a core for key values in their calculations. Recently, Prats, et al,' and Slider have reduced assumptions by adding the third phase—gas—in their forecasting methods. Slider uses the mobilities in the immediate vicinity of inlet and outlet wells as an aid to simulate the resistance effect in the five-spat pattern of the flow of oil, gas and water. Prats, et al. minimize assumptions by using previously determined laboratory data for determining sweep efficiency in the five-spot pattern. In neither of these papers is the saturation profile continuously affected by permeability-saturation curves. Sheldon and Dougherty- recently described a method that employs continuously changing saturation profiles using permeability curves, has a minimum of assumptions and needs no prior sweep efficiency. The Higgins-Leighton method, described in this paper, has all of these features; however, many of the techniques are different from those of Sheldon and Dougherty. The Higgins-Leighton method, tested in May, 1961, is direct and easy to apply and requires very little computer time to calculate a forecast. The short computer time is especially helpful in the study of a reservoir containing many layers of different relative permeabilities. In the Higgins-Leighton method, the individual pressures do not have to be calculated, as the resistance to flow in each cell in the flow pattern is readily determined without the use of any iterative techniques. The saturation and permeability distributions are readily determined. These data and the shape factor, which is measured only once from the potentiometric model when mobility ratio is one, determine the resistance to flow in each cell. THEORY The authors showed in a previous paper4 that, as an aid to calculating performance, the reservoir can be divided into channels using the streamlines of a potentiometric model as a guide. See Fig. 1. This procedure also was used in this paper. The authors also showed that, by treating conduits as approximately one-dimensional and neglecting pressure gradients transverse to the main flow, the Buckley-Leverett equation may be expressed as The principles expressed by this equation are employed extensively in the three-phase flow, as they were in two-phase flow. In the three-phase flow, the channels (the size and shape of which are taken from a potentiometric model)

By J. C. Fisher

Martensite transformations in steels and other alloys are characterized in part by the absence of composition changes during the growth of a new phase. Transformation occurs rapidly, and there is insufficient time for long range diffusion or partition of alloying elements to take place; martensite reactions in alloys thus are similar to phase transformations in single component systems. A fundamental understanding of martensite transformations in steels is impossible without knowledge of the free energy change upon transforming austenite (face centered cubic iron containing alloying elements) to ferrite (body centered cubic iron containing alloying elements) of the same chemical composition. The present paper assembles the best information available concerning the influence of temperature and composition on this free energy change. Most of the material has been taken from the work of Johansson,' Mehl and Wells,2 Zener3,4 and Smith;5 and indirectly, through these authors, from the work of Austin.6 In agreement with the generally accepted viewpoint, martensite is assumed to be an ordered solution of carbon in ferrite of the same composition as the parent austenite; only at high temperatures and low carbon concentrations is the carbon in ferrite distributed at random. The properties of the disordered solution are estimated by extrapolating the known properties of iron-carbon solid solutions into the range of supersaturation, and the free energy change associated with ordering is estimated from the theory developed by Zener. By incorporating Smith's recent thermodynamic measurements and Zener's theory of ordering, the present analysis modifies previous estimates of the free energy change associ- ated with martensite transformations. Consider a two component system consisting of a solvent A and a solute B. Let Na and Nb represent mol fractions of A and B respectively, let aa = raNa and ab = YbNb represent activities, and let superscripts 1 and 2 refer to phases 1 and 2. The partial molal free energies of A and B in phases 1 and 2 can be summarized as follows: free energy standard state Fa' - Fao1 = RT In an1 pure A' Fa2 - Fa2 = RT In aa2 pure A2 Fb1 - Fbo = RT In ab1 pure B Fs2 - FB2 = RT In aB2 pure B. The free energy of a gram atom of phase 1 is* AF1 = Na'Fa1 + Nb'Fb1 and that of phase 2 is AF2 = NA2FA2 + NB2Fb2. A martensite transformation from phase 1 to phase 2 requires Na1 = Na2 = Na and NB1 = NB2 = Nb, and the free energy change per gram atom accompanying transformation is AFi-2 = NA(Fa2 - Fa1) + Nb(Fb2 - Fb1) = NA[RT In (aA2/aA1) + AFa1?2] + Nb RT In (aB2/aB1) = Na[RT In (ra2/ra1) + AFa1?2] + Nb RT In (TaVTB1). [1] where yn2, yn1, yB2, yB1 are activity coefficients, and where Ma'+* is the free energy change upon transforming a gram atom of pure A from phase 1 to phase 2 at the temperature in question. For martensite transformations in plain carbon steels, A = iron (Fe), B = carbon (C), 1 = austenite (7). 2 = ferrite (a), and Eq 1 is AFr ?a= NFe[RT In (rf.a/rfer) + AFF.r?a] + TVc RT In (eca/rc1)- L2] Nothing is known concerning the values of yFea and yca for carbon concen- , trations in excess of 0.025 pct. However, the approximation rf.7 = 7f." = 1 cannot be appreciably in error for small carbon concentrations, and Eq 2 reduces to Afr-a = NfeFfer?a1 + NC RT In (rca/rcr)- [3] Johanssonl and Zenera have calculated MFfr?a from the specific heat measurements compiled by Austin.6* Their calculated values agree closely, and are summarized in Table 1. The activity coefficients relative to graphite for carbon dissolved in iron vary with temperature according to the relationships d In rcr/d(1/T) = AHcr-R dlnyca/d(l/T) = AHCa/R where AHc? and AHca are the heats of solution of graphite in y and a iron. Assuming the values of AH to be independent of carbon concentration and temperature, In rCr = AHcr/RT + I1 In rca = AHCa/RT + 12

Jan 1, 1950

By R. U. Fitting, A. C. Bulnes

Field experience with limestone and sandstone production indicates the existence of wide differences between the reservoir behavior of these two types of formation. Little attention appears to have been given to the separate study of the flow of fluids and the retention of fluids in limestones. This paper presents data and arguments to demonstrate the existence of a difference between the two types of formation, and urges the separate intensive investigation of limestone reservoirs. AH experimental data presented pertain to dolomitic limestone formations in west Texas and New Mexico. Two kinds of Porous media are recognized— intergranular and intermediate. Intergranular rocks are those in which the porosity and permeability are determined by the geometrical properties and the sorting of the sedimentary units; intermediate rocks, those in which there is no direct relationship between grain properties and the porosity and permeability. Limestones in general are intermediate media. The partial relationship between porosity and permeability of any class of Porous media is represelried by an area of .finite extent and definite shape on the permeability-porosity plane. The horizontal and vertical variations of porosity and of permeability in limestone and sandstone formations are discussed and compared. Comparisons are made of connate-water content, the relative permeability-saturation relationship, and capillary phenomena in the two kinds of rocks. A number of tentative conclusions are drawn relative to the reservoir properties of lime- stones; in particular, (I) dolomitic limestones are oil bearing and apparently are oil productive in zones of permeability less than 0.1 millidarcyj and (2) primary depletion (by gas expansion) oil saturations may be lower in dolomitic limestones than in sandstones. Introduction It is well known that the reservoir performance of limestones displays numerous irregularities when compared with that of sandstones, and that the departures from "normal" behavior of limestones are more frequent and generally more marked than in sandstones. This is particularly true of fractured limestones and those that have undergone considerable development of secondary porosity' A review of the literature of production research published during the past 15 years reveals a startling absence of theoretical and experimental investigations directed specifically toward explaining and predicting the performance of limestone reservoirs, even. though such problems as the prediction of water encroachment and the estimation of ultimate recoveries continue at the "inspired guess" stage. By far the greater part of the experimental data reported pertain to measurements on sandstone core samples, while the theoretical studies, almost without exception, are developed from the assumption of an ideal porous medium and therefore seldom are applicable to limestone formations. The chief cause of this situation appears to be the common belief that, to all practical intents and purposes, sandstones and the majority of producing limestone formations

Jan 1, 1945