Search Documents

By Stephen Krickovic

AS it is commonly understood in the bituminous coal mining industry, pillaring means removal, as completely as is practical, of all pillars formed in the development of headings and rooms on first mining. The degree to which this can be accomplished, termed percent of recovery, is of course dependent generally upon the physical conditions of the bottom and the overburden, the system of min-

Jan 1, 1956

By F. D. Wright, H. D. Fara

The solution to the plastic and elastic stresses around a circular shaft in a hydrostatic stress field that has appeared in two editions of a widely used book on rock mechanics1 is in error. Although this is a relatively minor matter, the correct solution was not found by a literature survey, so it appears worthwhile to present it. For a homogeneous, isotropic, rigid plastic-elastic body in plane strain which obeys Tresca's yield criterion the following equations must be satisfied:'," In the plastic zone,

Jan 1, 1963

By B. L. Averbach, M. Cohen, G. P. Conard

IT has been frequently observed that, when metal crystals are bent and annealed or when they are bent at elevated temperatures, the asterism of the Laue spots is split discontinuously into tiny intensity maxima. This finding indicates that segmented crystallites or subgrains are produced, which are relatively undistorted within themselves but which differ slightly in orientation from their neighbors. The following viewpoints have been put forward concerning the nature of this segmentation process: 1—Crussard' has described the phenomenon as "re-crystallization in situ." According to this mechanism, the deformed lattice recrystallizes on heating into an aggregate of comparatively strain-free subgrains separated from one another by low-angle boundaries and having nearly the same orientation as the original crystal. This process differs from the classical type of recrystallization in that the latter involves the formation of grains having quite different orientations from that of the mother crystal. 2—A similar but more detailed picture has been advanced by Cahn' who adopted the term "polygonization." In this interpretation, the plastic bending of lattice planes results in an excess of dislocations of one sign, but at suitable temperatures these dislocations migrate along slip planes and collect in localized regions to form an array of low-angle boundaries. Concurrently, the lattice segments between the boundaries lose their curvature and elastic strain. 3—The "cell formation" concept of Wood and co-workers: unlike the two hypotheses cited, postulates that the segmentation of the lattice is basically a mode of deformation and is not an annealing phenomenon. Cell formation appears to compete more effectively with respect to the normal slip process as the deformation temperature is increased and as the strain rate is decreased. The present work was performed on 3.2 mm diameter 99.99' pct Zn single crystals, grown from the melt in a gradient furnace. The major impurity was 0.003 pct Fe. The crystal orientations ranged from those with basal plane parallel to the rod axis to those with the basal plane almost perpendicular to the rod axis. The specimens were bent on templates to effective radii varying from 14.2 to 1.3 cm at the temperatures from —196" to 100°C. In each case, the axis of bending was parallel to the basal plane. Subsequent annealing was carried out at temperatures from —196" to 400°C. Back-reflection Laue pictures were taken on the convex surface of the bent specimens, before and after annealing. This procedure was found to yield better resolution of the detail in the diffraction spots than did the previously used Laue transmission technique. Some X-ray pictures of rods bent at —196" and —78°C were taken at these respective temperatures without intermediate warming to room temperature. In addition, bent crystals were acid cut" parallel to the basal plane, and double-crystal spectrometer rocking-curves were obtained from the etched (0001) face before and after increment an-

Jan 1, 1954

By L. J. Klinkenberg

It is pointed out that the current concept of pore size distribution is not unequivocally defined and that different methods aiming at a determination of such a distribution therefore do not lead to the same results. A pore size distribution should not be based on the idea that a core should, under all circumstances, behave as a bundle of straight non-interconnected capillaries of the same pore size distribution. This is illustrated by calculations and experiments on the displacement of miscible liquids. INTRODUCTION The introduction of the concept of pore size distribution has been an attempt to characterize the intricate geometry of the flow channels in porous media. As pores in a porous medium are interconnected it is impossible to describe exactly what is meant by a "pore" and, therefore, the term pore size distribution requires further amplification. In the visual method for the determination of pore size distribution "the pore" is considered to be the hole between the grains as observed on thin sections or on a polished surface of the rock under a microscope. In all other presently known methods,1,2 a displacement process is studied and it is ascertained what as- semblaae of straight cylindrical capillaries would show the same behavior. The distribution of the capillary diameters of this assemblage is said to represent the size distribution of the pores of the actual porous medium. From the foregoing it is already apparent that it is useless to try to decide which process or which method of investigation leads to a "correct" pore size distribution. To illustrate this further a displacement process in which a fluid is displaced by another miscible with it will be discussed and the results will be interpreted in terms of a pore size distribution. DISPLACEMENT EXPERIMENTS WITH MISCIBLE LIQUIDS OF NEARLY IDENTICAL VISCOSITY AND DENSITY Interpretation in the Form of a Pore Size Distribution If distilled water is displaced from a homogeneous unconsolidated sand pack by, for instance, a dilute solution of NaCI, the breakthrough recovery i.e., the amount of distilled water produced at the moment the first trace of NaCl can be detected in the effluent water, is between 95 and 100 per cent of the pore volume of the pack. This result can be understood by following in imagination the course of one particle of the liquid. Imagine a length, ?l, at the end of which a liquid particle has an independent chance of a renewed choice of pores of different diameters. In other words, the particle has a choice of traveling time, at, for any length ?l; instead

Jan 1, 1958

By Jack L. Lytton, Oleg D. Sherby

RESULTS of recent investigations1 on the creep of metals at high temperatures have revealed that the activation energies for creep of pure metals, AH,., about equals the respective activation energies for self-diffusion, hHD. This fact strongly suggests that the rate-controlling process in creep at high temperatures might be that of dislocation climb, as first postulated by Mott.' It would be desirable, however, to relate the creep rate of a metal at a given temperature and stress with actual diffusivity data, rather than with a comparison of the tempera-

Jan 1, 1957

By L. D. Mullins, H. B. Bradley

Numerous articles have appeared in the literature citing the application of the potentiometric analyzer to cycling operations. Specific information obtainable by a potentiometric analysis, as noted by various investigators, includes: 1. Pattern sweep efficiency for various injection and production rate programs. 2. Injected fluid breakthrough time at production wells. 3. Optimum production and injection rates. 4. Prediction of injection and/or production well locations. 5. History of advance of the in-jected-produced fluid interface. 6. Ultimate life of the reservoir. 7. Ultimate reservoir production. This note describes a procedure for predicting pressure distribution in a gas-condensate reservoir by using the potentiometric analyzer to obtain potential distribution data on an analog model of the reservoir and then applying Muskat's' equations. The method relates the volumetric weighted average datum pressure to the mean potential in the model.

Jan 1, 1958

By F. E. Steigert, A. H. Geisler

THE permanent magnet alloy comol, which contains 17 pct Mo, 12 pct Co, balance Fe, exhibits a precipitation reaction analogous to that at the iron end of the binary Fe-Mo system. The equilibrium precipitate, (P,), has a structure isomorphous with Fe,Mo, which is rhombohedra1 with parameters a = 8.97A and a = 3038.6&apos; according to the work of Arnfelt and Westgren.&apos; In the ternary alloy the parameters are a = 8.87A and a = 31°0.5&apos;. Oscillating crystal patterns of a single matrix crystal, M, in which P, was precipitated by slow cooling from 1300°C, showed that the orientation relationship of the two lattices is as follows: (110)P3 // (110)* (111)p3 // (110)m There are crystallites of P, in 12 different orientations in a single crystal of the parent solid solution. Microscopic examination shows that the crystallites are generally plate-like with their lateral dimensions parallel to (110) matrix planes. When precipitation in a supersaturated alloy quenched from 1300°C progressed isothermally at temperatures between 600" to 800°C, the evolution of the P, structure could be traced. Diffuse diffraction effects in Laue patterns provided early evidence of the precipitation process. Analyses of these by a previously described technique2 showed that they corresponded to rod-like elements parallel to <110> matrix directions in reciprocal space. With further aging to promote growth of the precipitate, the rod-like elements contracted to points in reciprocal space for normal Bragg reflection from the precipitate as shown by the left diagram in Fig. 1. Thus the diffuse scattering was attributed to the small thickness of the plate-like particles which was inadequate for sharp diffraction. When the precipitate particles were first thick enough to give normal diffraction their structure was different from that of the equilibrium precipitate, P,. The first structure, P1, presumably is a coherent transition struc-

Jan 1, 1952

By L. M. Howe, J. Gordon Parr, E. Saarema

THE decreasing solid solubility limit at the titanium-rich end of the Ti-Cu constitutional diagram,&apos; Fig. 1, suggests the possibility that titanium-rich alloys may be age-hardenable. However, results obtained by Holden et al. reproduced in Fig. 2 show that after quenching from 790°C the age-hardening of an alloy containing 1.7 pet Cu is very slight, while a 0.8 pet alloy decreases in hardness, during heat treatment at 400°C. It was believed possible that powder samples of alloys might show different results from the lump samples used by Holden et al., since small particles are often more sensitive to diffusion processes than larger samples. Consequently, a 1.9 pet (by weight) Cu alloy was made by the technique of levitation melting," checked for homogeneity, and filings of 48-65 Tyler screen size were cut from it for aging experiments. The filings were helium-quenched (by a technique described previously)&apos; from 790°C, sealed in vacuo in silica tubes and aged at temperatures of 400°, 450°, and 500°C. Microscopic and X-ray examinations failed to reveal any structural changes during the aging process. This is not too surprising since early stages of precipitation do not usually manifest themselves in ways that are readily detectable by metallographic methods, and the extent of precipitation on overaging a 1.9 pet alloy is very small. However, hardness readings (taken on a Bergs-man micro hardness tester, using a 25 g load applied to a diamond indenter) do show a hardness peak at each aging temperature, Fig. 3. About a dozen readings were taken on each heat treated sample; the highest reading and the lowest reading were ignored and an average taken of the remaining values. In

Jan 1, 1957

By Carl E. Johnson

INTRODUCTION A method for predicting water-flood oil recovery was reported by H. Dykstra and R. L. Parsons&apos; in 1950. It is now generally known as the Dykstra-Parsons method and is widely used by petroleum engineers. The method is semiempirical and consists of a correlation of four fundamental variables. These are: V, vertical permeability variation;&apos; a, mobility ratio; Sw initial water saturation; and R, fractional recovery of oil in place at a given producing water-oil ratio. The correlation extends over a wide range in each of these variables, and may be applied to all formations with initial oil saturations of 45 per cent or greater. The method of calculation originally outlined by Dyhtra and Parsons closely follows their derivation. And this is a convenient way of illustrating the development of some rather complicated ideas. However, in actual use it is more cumbersome than necessary. The purpose of this note, therefore, is to provide a simplified method for making Dykstra-Parsons predictions. A SIMPLIFIED GRAPHICAL SOLUTION The correlation between V, a, Sw and R, corresponding to a given producing water-oil ratio, can be shown on a single graph. This is best done by plotting V against a to show lines of constant R(1 - S, . WOR-0.2). Figs. 1, 2, 3, and 4 show these plots

Jan 1, 1957

By C. Goldberg, R. H. Wynne

WHEN using an optical goniometer to determine crystal orientation&apos; it is advantageous to use a preferential etchant so that the etch pits have plane faces which are parallel to crystallographic planes, all of which are of the same "form." "Westinghouse Silver Etch" (WAg), a preferential etch for germanium, is made up as follows: 40 ml of 49.7 pct hydrofluoric acid, 20 ml of 70.7 pct nitric acid, and 40 ml of distilled water containing 2.0 g of silver nitrate. Examination of metallographs of WAg etched surfaces (Fig. 1, for example) indicates that, regardless of crystal orientation, the faces of the etch pits are (111) surfaces. Optical goniometer examination of these WAg etched surfaces also indicates that the faces of the etch pits are (111) surfaces, i.e., all bright reflections are from (111) planes. When using an optical goniometer of the type described by Barrett,&apos; not only (111) reflections from a WAg etched germanium surface but also other reflections of lower intensity will be observed. These latter reflections can cause confusion. We have used an optical goniometer made from a war surplus astrocompass and a galvanometer lamp. The reflections are observed by using the goniometer in a darkened room and letting the reflected light be in- cident upon a piece of white paper. The resulting decrease of goniometer sensitivity enables (111) reflections only to be observed. This in turn permits the crystal orientation to be determined in a matter of minutes, since only two measurements are necessary. The authors wish to acknowledge the help received from R. E. Davis who assisted in making the metallographs and aided the authors in the interpretation of the metallographs.

Jan 1, 1954

By W. D. Roberts

INTERGRANULAR corrosion of metals and alloys has been extensively investigated and the current conclusions are summarized in the Symposium on Stress-Corrosion Cracking of Metals.&apos; It appears, however, that the relative importance of the various factors involved in the mechanism is still obscure. Several recent, independent studies, in combination with available X ray data op preferred orientation, indicate that there may be a direct correlation between relative orientation of contiguous grains and susceptibility to intergranular corrosion. The object of the present note is to show the nature

Jan 1, 1951

By L. K. Jetter, C. J. McHargue

DEVELOPMENT of preferred orientation in uranium is of interest because of the strongly anisotropic properties of the orthorhombic crystal structure. Harris&apos; reported inverse pole figures for rolled uranium rods which showed <010> and <410>* to be parallel to the rolling direction after On 0.19-in. spherical X-ray diffraction specimens taken from the center of the middle section of each rod, pole distribution data were obtained by the spectrometric method described previously.4 The specimen was rotated rapidly about the axis coinciding with the rod axis in order that a large number of grains could be scanned and was rotated slowly about an axis perpendicular to the rod axis. The variation in the diffracted X-ray intensity was recorded as a function of the latter motion (angle 4). Fiber-axis distribution charts (inverse pole figures) were prepared by the method described by Setter, McHargue, and Williams." After having been corrected for background radiation, the pole distribution data (I vs ?) were normalized. A normalizing factor for each plane has been shown to be Iranaom = f I sin 4 ? 4 and was obtained by graphical integration from the data for that plane plotted as I sin ? vs ?. The axis-distribution charts deduced from the normalized pole densities are shown in Figs. 1 to 3. On these charts, the contours indicate angular positions of equal normalized axis density,

Jan 1, 1958

By J. H. Keeler

AN alloy of zirconium containing 18 atomic pct Cb was found to be body-centered-cubic and readily deformable at room temperature. The behavior of this alloy provided an opportunity to determine if

Jan 1, 1957

By J. W. Fredrickson

VERY little information is available in the literature concerning the solubility of magnesium in titanium. Aust and Pidgeon&apos; report the solubility of titanium in magnesium to be 0.0025 pct at 650°C and 0.015 pct at 850°C. X-ray diffraction studies failed to detect any evidence of solid solubility. Based on the fact that the addition of titanium to magnesium increases the c/a ratio of magnesium, Busk&apos; suggested that the Mg-Ti system, like the Mg-Zr, is a peritectic. Eisenreich- eported the solubility of titanium in magnesium to be 0.003 pct at 655°C and 0.064 pct at 760°C. Removal of hydrogen from the melt increased the solubility of titanium to 0.115 pct at 800°C. Eisenreich proposes a peritectic-type diagram for the high magnesium side of the system. Recent work&apos; with Ti-Mg diffusion couples indicate that the solid solubility of titanium in magnesium is extremely small and that there is limited solid solubility of magnesium in titanium. Experimental work on the Ti-Mg system is complicated by the fact that the boiling point of magnesium is considerably below the melting point of titanium. Consequently, the usual alloying methods are impractical. Alloying was accomplished in this investigation by sheath-rolling compressed compacts." The alloys were made from titanium sponge and high purity magnesium. Chemical analysis of the sponge showed 0.04 pct Mg, <0.005 pct Al, <0.001 pct Cu, 0.05 pct Fe, 0.022 pct Mn, <0.0025 pct Ni, 0.007 pct Si, <0.005 pct V, and <0.01 pct Zn. All heat treating was conducted in a protective atmosphere by encapsulating the specimens in fused silica ampules under a partial pressure of helium. Equilibrium was achieved by holding the specimens at temperature for periods of 24 to 120 hr. The alloys were hot rolled, cold rolled, and annealed prior to solution heat treatment. Fig. 1 shows the constitutional diagram obtained for the Ti-Mg system up to 1.5 pct Mg. Oxygen was known to be present in the alloys; therefore, it is necessary to consider the alloys as ternary rather than binary and the diagram should be viewed as a plot of a section through the Ti-Mg-0 system. Other investigatorso have shown that the addition of oxy- gen to titanium stabilizes the a phase. The data presented in Fig. 1 indicate that the addition of magnesium to titanium tends to stabilize the p phase. Armour Research Foundation7 has shown that the addition of magnesium to zirconium lowers the a transus and is classified as a p stabilizer in that system. Magnesium is soluble in solid titanium to the extent of at least 1.5 pct in both the a and /3 phases. No intermetallic compounds or magnesium-rich phases were observed either by light microscopy or X-ray diffraction. However, this does not negate the possibility that such phases exist at higher alloy concentrations.

Jan 1, 1956

By Dilip K. Das, Douglas T. Pitman

ONE of the major uses of barium in metallic form is as a getter material in vacuum tubes. Because of the high chemical reactivity of the metal, Ba-Al alloys are extensively used. Numerous methods for the preparation of Ba-Al alloys have been published, a few of which1-4 are cited here. Most of these methods were found to be quite elaborate, involving the reduction of BaO, and not too well adapted for the close control of the final composition of small amounts of alloys prepared for laboratory use. A simple laboratory method for the preparation of Ba-A1 alloys in small batches starting from pure metals was devised, so that it was possible to control the desired compositions to within 1 pct. The pertinent features of the alloy system Ba-Al5 are 1) an intermediate compound BaAl, with the melting point of 1050°C, and 2) a eutectic between aluminum and BaAl4 at 98 pct Al. The accompanying sketch shows the experimental arrangement for the preparation of the alloys. Weighed amounts of aluminum and barium were placed in an alumina and a stainless steel crucible, respectively. According to the supplier&apos;s specification, the purity of the metals used in the alloys is as follows: a) aluminum rods—99.9 pct Al, and b) barium rods—99.5 pct Ba. The stainless steel crucible, tapered at the bottom and having a 1/16 in. diam hole, rested on top of the alumina crucible. The assembly was placed inside a graphite sleeve which rested on a refractory platform. The platform moved the assembly up and down through the field of a radio frequency coil. A glass bell jar was placed between the crucible assembly and the radio frequency coil to maintain a steady flow of helium around the melt. A small window was cut out on the wall of the alumina crucible to observe the progress of the re- action and to record the temperature with an optical pyrometer. The platform was first raised high enough to move the barium out of the radio frequency coil field in order to allow only the aluminum to melt. The assembly was then lowered so that the barium began to melt and flow out through the small orifice into the molten aluminum. In order to keep the violence of the exothermic reaction under control, the rate of flow of barium was carefully regulated by raising or lowering the crucible assembly. All the samples prepared by this technique were examined by a Norelco X-ray diffractometer using CuKa radiation, The diffraction specimens were prepared by placing the finely powdered samples in flat specimen holders. The Ba-Al alloys prepared with a high barium content were found to consist mainly of BaAl4. The structure of BaAl4 has previously been reported by Alberti and Andress.8 They found that BaAl4 was body-centered-tetragonal with an a0 = b0 = 4.530Å and c0 = 11.14Å. An alloy whose composition was found by chemical analysis to be almost 100 pct BaA1, was used to determine the relative intensities. The d-spacings were obtained from the same alloy to which a small amount of tungsten had been added as a calibrating material. Accurate values for a, and c, were calculated according to the method proposed by Taylor and Floyd.&apos; The calculated values are: a, = b, = 4.566Å and c0 = 11.250Å. The measured d-values for BaAl4 are shown in Table I along with relative peak intensities above background and hkl indices. Acknowledgment The authors are grateful to L. J. Cronin, the head of the Techniques Dept., for suggesting the problem and for his constant interest. References 1Froges and Camargue: German Patent No. 809107, 1951. French Patent No. 935324, 1949. 2E. Bonnier: Annales de physique, 1953, vol. 8, pp. 259-312. 3M. Orman and E. Zemhela: Prac Institute of Metals; 1952, vol. 4, pp. 437-445. 4E. Fujita and H. Yokomizo: Reports Gov. Chemical Industrial Research Institute, Tokyo, 1952, vol. 47, pp. 291-297. 5E. Alberti: Ztsch. fur Metallkunde, 1934, vol. 26, p. 6. 6E Alberti and K. R. Andress: Ztsch. fur Metallkunde, 1935, vol. 27, p. 126. 7 A. Tnslor and R. W. Floyd: Acta Crystallographica, 1950, vol. 3, p. 285.

Jan 1, 1958

By E. R. Brownscombe, Francis Collins

The pressure distribution in a reservoir producing an incompressible fluid by radial flow in a horizontal structure is a simple logarithmic function&apos; used daily by reservoir engineers. The assumption of an incompressible liquid is equivalent to assuming that pressure is maintained constant at the well radius and at some external "radius of drainage," or, in other words, that the entire flow into the well passes across the external radius. In most cases a well is surrounded by other wells draining adjoining areas so that the "radius of drainage" is not a line of constant pressure with flow equal to the flow into the well but rather a contour of zero pressure gradient across which no flow occurs, production being due to the expansion of the liquid within the area. The radial flow steady state pressure distribution for this case does not seem to be extensively used despite the fact that the mathematical analysis has been given in detail&apos; and the results are simple and generally applicable. &apos;The object of this discussion is to develop this result from a physical rather than a mathematical viewpoint, thereby aiding intuition in judging circumstances to which it may be applicable. If the production rate at a well in an unsaturated reservoir is maintained constant while no flow occurs across the outer radius a steady state will soon be reached in which the flow rate or pressure gradient is constant with time at all points in the reservoir. To preserve such a condition it is necessary that the pressure decline at the same rate at all points in the reservoir, that is, the curves of pressure versus radial distance at successive times are all parallel. Then each unit volume of oil is expanding at the same rate and contributing an equal share to the production. The rate of flow q across any circle r is thus proportional to the volume of the reservoir between r and the external radius re, or q = qw (I-r2/r2e) ......(1) where q, is the production from the well and it is assumed that rw is very small compared with re. This equation can be combined with the differential form of Darcy&apos;s law for radial flow, 2phkr dp q =—— ........(2) aµß dr ( where q is not a constant as in the incompressible flow case but is a function of r given by Equation (1). If Equation (1) is substituted in Equation (2) the variables p and r can be separated and the result integrated to obtain the pressure distribution p = pw +--------In------------- 2pkh[ rw 2re2] ......(3) This equation differs from the incompressible case only in the addition of the term proportional to the square of the radius

Jan 1, 1950

By C. G. Dunn

THE present note, which was given in the form of a Research in Progress item,&apos; traces the origin of secondary recrystallization grains produced on annealing a cold rolled bicrystal specimen. The investigation began with the realization that a recently acquired conception for the origin of secondary grains produced on annealing cold rolled single-crystal specimens might require modifications for the case of polycrystalline materials. The stated conception2,3 begins with several conditions to be met by primary recrystallization: 1—a primary re-crystallization structure to consist of many small primaries and relatively few large primaries (diam- eters two or three times the average); 2—a primary recrystallization texture to consist of one strong preferred orientation or component and one or more weak components; and 3—a primary recrystalliza-tion structure-texture combination with some of the large primaries belonging to the weak components of the texture. When these conditions occur, some of the large primaries from the weak components of the texture

Jan 1, 1955

By H. H. Rachford, and J. D. Rice

The effectiveness of digital computing machines in making technical calculations depends on how well the work is arranged to utilize the capability of the machines. This note presents a particularly useful way of calculating hydrocarbon vapor-li.quid equilibrium in the flash vaporization (or condensation) system. The method is well suited to sequence-controlled computing equipment. It is not limited to equilibrium calculations and may be used for solution of most implicit equations in one variable. INTRODUCTION There is increasing interest in the use of electronic digital computers in research and engineering calculations. This is a fortunate and inevitable trend in view of both the increasingly extensive numerical work which is becoming a routine part of many daily production operations and growing demand for the overwhelming amounts of calculations required by newly developed numerical methods for solving heretofore unsolved problems. Machines are in many ways ideally suited to the task but of necessity present certain difficulties. for a particular poblem to be solved must often be formulated quite differently from the way it would be arranged for manual solution. This is done in order to take advantage of the inherent speed and precision of electronic computers and at the same time to limit the need for number storage to the capacity of the machine. Therefore, this note is submitted to present a general and quite powerful method of finding solutions of the frequently encountered implicit equation: F(x,y1,y2,...ya)=0.......(1) where the root, x, is to be found for a given set of y1, ... yn.. The procedure is well suited for use with computing machines, for it usually requires but little storage or programming beyond that necessary to evaluate F. HYDROCARBON EQUILIBRIUM The method is described in terms of the problem it was designed to solve, i.e., the calculation of hydrocarbon vapor-liquid equilibrium in flash vaporization. Given the composition of a hydrocarbon mixture and appropriate values of the equilibrium ratios K, to find the phase ratio and compositions in a closed system: let z, be the mol fraction of the i-th component in the mixture. If K, is the ratio yl/xl, where y, and x,

Jan 1, 1952

By F. G. Stone

A strain anneal process has been developed for the production of iron tensile test specimens, the gauge length consisting substantially of a single crystal. Since the method has been quite successful and the results of general interest, the process will be described in detail. The essential feature of the method is the continuous, slow increase of furnace temperature during the grain growth annealing treatment. Briefly, the procedure is as follows: I. Heat machined bars to 950°C in vacuum, hold I hr, furnace cool. 2. Strain to 3.5-3.8 pct elongation (axial loading). 3. Anneal, continuously raising furnace temperature from 450 to goo°C at a rate of so0 per day. 4. Hold at 900°C for 100 hr, furnace cool. Tensile test bars with a 1/4- X I-in. gauge section have been machined from 35 in. Armco iron plate and from 3/8 in. rounds of Puron, a high purity laboratory iron. The axes of the test bars were parallel to the rolling direction. During the machining operation the bars were polished in the lathe with No. 600 emery paper, and this polished surface proved satisfactory for subsequent macrographic examination. A micrographic examination of the materials showed that both were high in FeO inclusions, those of the Puron being smaller but much more numerous. An initial vacuum anneal of I hr at 950°C followed by furnace cooling resulted in a ferritic grain size of ASTM No. 3 for Armco and No. I for Puron. Special heat treatments which reversed these initial grain sizes produced no marked improvement in single crystal production after subsequent treatments. The strain for subsequent grain growth was determined by measuring the decrease in diameter of the tensile bar during elongation. The change in diameter could be measured with a micrometer to +0.0001 in., corresponding to an accuracy of 0.08 pct for the elongation. The accuracy of the diameter measurement is also dependent on grain size because of surface roughening during elongation, and for this reason, a No. 3 grain size is preferable. Optimum grain growth was obtained for strains between 3.5 and 3.8 pct elongation. Once this critical range had been cstablished, all tests were held within it, and thus only meager data were obtained on grain growth following strains outside this range. In general, grain size decreased with increasing strain to about 5.5 pct elongation. Grain growth was occasionally ohserved with less than 3.5 pct elongation although this may have been the result of error in measurement. The program control device for the grain growth anneal consisted of a I rpm motor and worm gear drive, operating the dial shaft of a pyrometer controller through a pulley. Dial speeds of 33 and 50°C per day were tried, but the results at the slower speed were not sufficiently better to warrant the increased furnace time. The furnace tube was evacuated at intervals during the run and kept sealed the remainder of the time. When the controller reached 900°C, it was stopped and the temperature held for 100 hr. This eliminated most of the numerous minute crystallites of about 0.5 mm dimension usually visible on the etched surface of the single crystal. Since recrystallization occurs at the A3 point, the top temperature of 900°C provided a 10"

Jan 1, 1949

By E. R. Morgan, J. C. Shyne, V. F. Zackay

NITROGEN used as an alloying element in steel has received increased attention in recent years. Its merit as an austenite stabilizer and hot strengthener has long been recojinized, particularly by European investigators.&apos; Recent work&apos; indicates that nitrogen contents of 1/2 pet or more allow the addition of several percent of ferrite-forming hot strengtheners, such as molybdenum, without subsequent austenite decomposition at elevated temperatures. The room temperature properties of such alloys are characterized by high yield and tensile strengths with excellent ductility, even after cold working, as shown in Tables I and 11. The 1350°F creep-rupture properties of Cr-Mn-Mo-N steels are comparable to those of commercial alloys such as Timken&apos;s 16-25-6, containing 16 pct Cr, 25 pct Ni, and 6 pct Mo, as shown in Table 111. Cr-Mn steel containing 1/2 pet N must be made under pressure if ingot gassing is to be avoided. However, there are difficulties attendant to this practice. Pressure melting restricts the ingot size, and the equipment required is costlier and more complicated than comparable air melting facilities. An effort was made, therefore, to determine more economical methods of adding nitrogen. The pressures generated in centrifugal casting usually range from 5 to 10 atm and are sufficient to prevent the gassing of high nitrogen Cr-Mn steels. A centrifugal casting was made of an alloy containing 16 pct Cr, 18 pct Mn, 2 pct Mo, and 0.73 pct N. A risk, 8 in. diam and 11/2 in. thick, was cast at 1000 rpm. It has been calculated that a pressure of 9 atm was developed at the periphery of the disk at this speed. Metallographic inspection of specimens taken from the disk revealed no gas holes or other defects. The elevated temperature properties of radial and circumferential specimens taken from this casting are given in Table IV. The creep-rupture life of this cast nickel-free austenitic steel was equivalent to that of the wrought Cr-Mn-Mo-N steels. Although centrifugal casting appears to lend itself to the manufacture of high nitrogen austenitic steels, its applicability is limited to relatively small sizes. A process applicable to the production of large ingots of high nitrogen steels is suggested by low carbon rimming steel practice. Gas holes formed during ingot solidification are not detrimental so long as they weld shut during hot rolling. An economic advantage of gassing ingots is the increased ingot-to-slab yield, especially if the ingots are mechanically capped. Also, it is probable that higher nitrogen levels can be achieved with mechanically capped ingots than with hot-topped ingots. A series of gassy heats was made under both atmospheric and higher pressures, as indicated in Table V. Several of the air melted ingots were mechanically capped by chilling the molten ingot tops with a steel plate. The ingots were allowed to freeze in steel molds without hot-tops unless otherwise indicated. A comparison of Table 111 with Table V shows no significant difference in the creep-rupture properties of the gassy and the nongassy ingots after hot work- ing nor between the gassy ingots produced under different conditions. References 1 F. Rapatz: Application of Stainless and Heat Resisting Steels Alloyed with Nitrogen. Sinhl and Eisen ( 1941) 61, pp. 1073-1078. &apos;V. I?. Zackay, J. F. Carlson, and P. L. Jackson: High Nitrogen Austenltlc Cr-Mn Steels. Trans. ASM 119551 48, PP. 509-523.

Jan 1, 1957