Search Documents

By Rodney L. Stone, David S. Cahn

As part of an over-all laboratory and plant test program to determine mixing ad other techniques for a more economic use of bentonite as an iron ore pellet binder, a settling test for measuring the bentonite content within a pellet or a batch of pellets has been developed. Through this test, it has been shown that dry fracture load is directly proportional to bentonite content in batches of pellets, but inversely propor tional to bentonite content in individual pellets of the same batch. This apparent paradox is explained by a negative correlation between a pellet's size and its bentonite content and the fact that large pellets can withstand a higher fracture load than small pellets. In past attempts to characterize green and dry iron ore pellets, the authors found large variations in the results of strength tests on closely sized pellets from the same batch.' Reasons advanced for such variations in the quality of pellets in which the binder is bentonite were the presence of internal flaws and the nonuniform distribution of the binder. Even when resulting pellet strength is good, there is still the question of whether the same strength could have been attained through a more effective use of a smaller amount of bentonite. The authors' own and other studies 2-3 have pointed out the strong dependence of pellet strength on the degree of "mixedness" of the iron ore concentrate and the bentonite binder. However, before such a relationship could serve as a guide for the development of better mixing techniques and the associated economy of

Jan 1, 1969

By A. W. Asman

The problem of balanced design in a belt conveyor has been under constant study. Increased use of this method of transport in recent years has stimulated investigations. Conveyor belt, power transmission and other machinery should be adequate to perform the required service. All components of the belt conveyor should work as an efficient unit moving its load continuously between terminals. As part of a long term research project on belt conveyors, Hewitt-Robins established in 1954 a 2-year grant at Pennsylvania State University. Under terms of the grant, certain fundamental phenomena were to be studied. This work evolved a new concept in belt conveyor power determination based on the principle of belt tension. The Penn State contract furnished the foundation for later more detailed studies. From these most recent investigations, more specific factors were defined. In the final analysis, all of the data collected has been worthwhile in belt conveyor design. Units now operate over long distances at speeds of over 800 fpm. Long high lifts are commonplace. Where vertical or inclined shafts were the rule, belt conveyors receive first consideration. Although belt conveyors have moved forward rapidly, the future will see a tremendous gain in their use. The ideal test belt conveyor would be several thousand feet in length and have a maximum tension of at least 70,000 lb., Also, this conveyor should be horizontai and have a variable feed rate of more than one material. Data from such a field belt conveyor combined with laboratory information would permit accurate analysis. Since this type of conveyor was not available, many field locations were visited: An alternate would have been to erect a closed circuit system in the laboratory. Either method had limitations and the first was selected as having the best possibilities. Manufacturers of conveyor belting and machinery basically have been using the same horsepower formula. As more information became available from field installations, factors in the basic formula were modified. Practically all conveyors are designed today by adding together the horsepowers required to move the empty belt, move the load horizontally, and elevate the load. In selecting size of drive motor, the first two elements are relatively unimportant if the last is large. However, it is the combination of

Jan 1, 1961

By Thomas C. Wilder

Jan 1, 1965

By W. H. Boyer, J. W. Gabelman

The concept of metal zoning has long been recognized (Spurr, 1907, 1923; Emmons, 1924, 1926; Lindgren, 1933; Rateman, 1950), although it has been considered indistinct for most epithermal districts. The geometric relation of barren zones of hydrother-mally altered rock to ore deposits was emphasized in the work of Lovering (1941, 1949, 1950), Sales and Meyer (1948, 1949, 19501, Burbank (1941, 1950), Schwartz (1939, 1950), and Kerr (1947). Lovering (op. cit.) and Sales and Meyer (op. cit.) emphasized the consistent presence of barren alteration envelopes enclosing individual ore bodies, particularly of the vein type. Schwartz (1950) noted a difference between the distribution of alteration zones described by Lovering and by Sales and Meyer, as compared with those associated with the porphyry copper deposits of the Southwest. The disseminated deposits of the Southwest usually occur within large bodies of altered rock whereas most veins do not. In the following discussion the mineralization process is conceived of consisting of two distinct phases: pre-ore nonmetallic alteration and metallization. In the present investigation the position, size, and shape of all known hydro thermally altered areas and the position, size, and metal type of all known ore deposits were plotted on base maps of southwestern and west-central Colorado (Figure 1). The maps indicate a generally consistent geometric relation between altered and metallized areas. The addition of alteration to the pattern of metal deposits establishes a temperature zoning pattern of much greater clarity than that formed by the epithermal metal deposits alone, and allows a more precise definition of mineral belts. The center of strongest mineralization is indicated to be the area of greatest barren alteration, and metal deposits are grouped about it both under structural control and crudely in response to temperature zoning. The combination of an area of strong alteration and associated marginal clusters of metal deposits is defined as a mineral district. The strong similarity in individual deposits, in groups of different types of deposits, and in the alteration-metallization relation in districts throughout the region implies that a single metallogenic province is represented in southwestern and west-central Colorado. The occurrence of deposits in Tertiary voi-canics indicates a late Tertiary metallogenic epoci., although an early Tertiary epoch is known locally as at Ouray. Patterns formed by mineral districts can be

Jan 1, 1961

By S. M. M. Safvi, A. Jowett

Small-scale continuous flotation tests are described in which the influence of 1) pulp density and 2) feed rate on the rate of flotation are investigated. The results provide further evidence of first-order kinetics for flotation and indicate that tests at different feed rates provide a particularly simple method of determining rate constants. Flotation kinetics are discussed in general terms. Over a period of many years numerous research workers have studied the rate of recovery of valuable materials from flotation pulps with a view to deter-mining the mechanism of the froth flotation process. Froth flo ation is fundamentally a collision process involviug at some stage collision and then mutual attachment between air bubbles and mineral particles. It is possible to draw analogies therefore between the flotation process involving mainly macroscopic particles and the ordinary processes of chemical reaction kinetics involving collisions between atoms, molecules and ions. By adopting this analogy we may postulate a general equation representing the kinetics of the flotation process: -dC/dt =K.CC ,Aa (1) The rate constant, K, will be a complex function involving reagent concentrations, particle and bubble sizes, induction times and no doubt flotation cell design. However, these are factors which can be kept constant in any series of experiments, permitting investigation of the exponents a and c, which determine the order of the reaction occurring in flotation. The various research investigations carried out so far have all aimed at evaluating a and c and in the majority of cases only since the concentration of air has usually been kept constant in a series of tests, so that a relationship of the following type has been investigated: -dC/dt = K1 Cc (2) The techniques of both batch testing and continuous testing have been used to investigate this relationship. The technique of continuous testing, in which steady-state conditions are established, permits direct investigation of the above equation. In the case of batch testing, C is decreasing continuously

Jan 1, 1961

By B. A. Wilcox

ThERE are several reports in the literature which indicate that both solid-solution hydrogen and hydride precipitates can promote low-temperature em-brittlement of tantalum.1-3 For example, Imgram et al.2 showed that hydrogen additions in the range of 140 to 580 ppm promoted a ductile-to-brittle transition at slightly above room temperature in both wrought and recrystallized tantalum. This contrasts with the observation that tantalum of norma1 purity is very ductile, even at liquid-nitrogen temperature. Imgram et al. also found that hydrogen additions caused a deterioration of notch tensile properties of tantalum. The purpose of this communication is to present some limited evidence that hydrogen can also cause fatigue embrittlement of tantalum. The material used in this investigation was wrought, stress-relieved (1000°C, 1 hr), are-melted tantalum rod (5/8-in. diameter) containing the following impurities in weight percent: 0.005 C, 0.007 02, 0.0005 H2, 0.003 N2, 0.08 Cb, 0.01 W, and 0.01 Fe. Hydrogen charging was done in an atmosphere of flowing dry hydrogen for 3 hr at 600°C, followed by an homogenizing anneal in purified argon for 15 min at 600°C, and cooling in an argon atmosphere. Typical micro-structures of the hydrogen-free and the hydrogen-charged materials are shown in Fig. 1(a) and Fig. 1(b), respectively. Fig. 1(b) shows the appearance of hydride precipitates which, in many cases, formed preferentially parallel to the rod axis. The hydride dispersion was uniform across the diameter of the specimen. The fact that the precipitate was a tantalum hydride was tentatively confirmed by X-ray diffraction on a metallographically prepared surface. After hydrogen charging, chemical analyses by vacuum fusion revealed the hydrogen content to be 0.0295 wt pct. In addition, the oxygen and nitrogen contents were slightly raised to 0.0095 O2 and 0.005 N2.

Jan 1, 1964

By E. H. Jr. Dix

THE use of magnesium as an alloying element in aluminum alloys has been limited, in general, to comparatively small quantities. In duralumin-type, strong-aluminum alloys, magnesium is present to the extent of ½ to ¾ per cent. and various grades of Magnalite and the "Y" alloy of the National Physical Laboratory of England contain up to 1 1/2 per cent. In this country small quantities of aluminum alloy sheet containing 4 per cent. magnesium are produced commercially. All commercial aluminum contains sufficient silicon to convert a portion of the magnesium added to the compound Mg2Si. It is well known that the solubility of both magnesium and magnesium-silicide decrease with decrease in temperature. It is, therefore, impossible to separate the age-hardening effect of the magnesium-silicide from that of magnesium, except by the use of aluminum of such high purity that the silicon content is negligible. The present paper is the fifth1, 2,3,4 of a series from the laboratories of the Aluminum Co. of America reporting the results of investigations of the equilibrium relations in aluminum-rich alloys made from the high purity aluminum manufactured by this company .5

Jan 1, 1929

Jan 1, 1963

By E. J. Dulis, A. Kasak, R. C. Westgren

The potential of uirought tungsten- and tantaluw-rich alloys of the W-Ta-Mo-Cb system was investigated for high-strength structural applications above 2500° F. Appreciable strengthening of tung-sten and tantalum—essentially attributable to solid-solzction effects—was obtained with binary, ternary, and quaternary alloys. Tensile strengths as high as 67,000 psi were achieved at 3000 OF. ThE quest for wrought structural materials which possess good strength properties at temperatures above 2000°F has led to a considerable research effort during the past decade. In addition to exploring the potential of various high-melting-point refractory metals (such as molybdenum, columbium, tantalum, and tungsten) for very high-temperature structural applications, alloy-development programs have been undertaken to achieve greater strengths than could be achieved with unalloyed refractory metals. The methods of strengthening refractory metal alloys have included both solid-solution and dispersion strengthening. General relationships among the relevant properties (melting point, recrystallization temperature, and high-temperature strength) of refractory metals indicate that the highest strengths at temperatures of 2500°F or higher would be obtained with tungsten-or tantalum-base alloys. Therefore, the present investigation was undertaken to determine the high-temperature strength potential of solid-solution strengthened alloys based on the binary system W-Ta; to effect further strengthening, minor additions of molybdenum and columbium (at the 6 and 12 pct levels) were made. The portion of the W-Ta-Mo-Cb quaternary alloy system of interest is shown schematically in Fig. 1. The component metals—W, Ta, Mo, and Cb—have a bcc structure at all temperatures, and their binary systems are mutually completely soluble.' Complete solid solubility in the W-Ta-Mo and W-Ta-Cb alloy systems has been demonstrated,' and, therefore, it was expected that the quarternary alloys of the system would also exhibit complete solid solubility at all temperatures. In the virtual absence of contamination by interstitial elements, all the alloys would be truly solid-solution strengthened. Twenty specific compositions of this investigation included binary, ternary, and quaternary alloys of the W-Ta-Mo-Cb system as well as unalloyed tungsten and tantalum. The compositions are shown in Table I. EXPERIMENTAL PROCEDURE Consumable-electrode vacuum-arc melting with ac power was used to produce small laboratory-sized, cylindrically shaped ingots (1-1/4 in. in diameter and 2 to 3-1/2 in. long) of the twenty compositions. A furnace was specially constructed to melt as many as four electrode rods simultaneously into molds lined with tungsten or tantalum sheet. High-purity rods of tungsten, tantalum, molybdenum, and columbium were used for consumable electrodes; the diameter of each rod was determined by the desired weight percentage of the component metal in a given alloy. The details of the construction and operation of the melting furnace will be discussed in another p~blication.

Jan 1, 1964

By V. Koump, T. F. Perzak, R. G. Olsson

Jan 1, 1965

By H. O. Hofman

At the Lake Superior meeting, September, 1904, Messrs. A. Gibb and R. C. Philp presented a paper entitled " The Constitution of Mattes Produced in Copper-Smelting,"' in which they concluded that ferrous and cuprous sulphides formed the chemical compound, 5 Cu2S, FeS. They arrived at this inference from the observed fact that a matte containing Cu,S, 90.4, and FcS, 9.6 per cent. (or, in round figures, Cu2S, 90, and FeS, 10 per cent.), did not contain any metallic copper, while one with more than 10 per cent. of FeS readily dissolved copper, and with less than this amount retained copper in mechanical suspension. Although they determined the melting-points of five mattes containing from 32.6 to 80.1 per cent. of copper, they did not draw a freezing-point curve, as the necessary apparatus was not available. A microscopical examination of 11 polished samples gave them a grayish-black field of very uniform appearance, except that five of the specimens showed the presence of metallic copper. The object of the present investigation was to supplement the above paper by drawing the freezing-point curve of the ferrous sulphide-cuprous sulphide series, and to see how the constitution could be further revealed by microscopical work. The results obtained lead to conclusions very different from those of Gibb and Philp. While their paper bears the date 1904, it was published only in 1906.2 In the meantime, Bolles3 had presented his paper, " The Concentration of Gold and Silver in Iron-Bottoms." The raw material which formed the basis of the research was a matte containing Fe, 61.68, and Cu, 11.20

Jan 1, 1908

By W. A. Burns

Vertical flow is an important mechanism in many petroleum reservoirs. Yet no adequate method has heretofore been proposed for determining the in-situ vertical permeability in a reasonable length of time. Core measurements of vertical permeability do not necessarily represent effective in-situ values. Cores may be representative of only localized pockets of high or low permeability, and total reliance on cores might cause one to overlook the presence of vertical fractures altogether. Conventional single-well tests, interference tests, and pulse tests commonly are used to determine horizontal transmissibility, khh/u, but generally are not designed to measure vertical trans-missibility or other groups containing vertical permeability, ku. A new well test described here provides a measure of in-situ vertical permeability as well as the previously measurable horizontal permeability. This "vertical test" determines the vertical diffusivity, kv/gcu, and horizontal transmissibility, khd1/u (where dl is the length of injection or production perforations). The test is basically a short-term interference or pulse test normally lasting less than a day. Conventional down-hole equipment is used between two isolated sets of perforations in the same well, as shown in Fig. 1. A "layer-cake" description of a reservoir can be obtained by running two or more vertical tests in the same well, each over different vertical intervals. Each test can be analyzed for the effective vertical and horizontal properties of the interval tested. In addition, a vertical well test can determine the effective- ness of a shale barrier or tight zone by testing across it. Field use to date of vertical testing has confirmed its utility in reservoir description. Mathematical Basis Interpretation of vertical tests for ku and kh is based on comparison of measured pressure response with computer-generated response. Pressure behavior during a vertical test in a homogeneous, anisotropic, infinite-acting reservoir (infinite in radial extent and bounded in both directions vertically) is given by the diffusion equation for slightly compressible fluids with negligible gravity effects.l A schematic diagram of the problem is presented in Fig. 2, and an analytic solution, based on the image

Jan 1, 1970

By R. C. Neavel

One of the most important requirements to be met by any coal producer is delivery of size consists tailored to specific applications. Generally these specifications include a minimum as well as a maximum size limit. The customers of a major Indiana coal producer noticed that some shipments of sized coal contained a large proportion of undersize pieces. Also, in these shipments, the pieces which remained within the specified limits were cracked and, when handled, crumbled readily. The producer, certain that these shipments had been properly sized at the tipple, suspected that degradation had occurred in transit. The Indiana Geological Survey was requested to examine the coal from one of these shipments to attempt to determine the cause of breakage. Two basic facts about coal and coal seams led to the employment of petrographic techniques in this investigation. First, all coal seams consist of accumulations of diverse organic and inorganic constituents; and second, most coal breakage, either desirable or undesirable, can be attributed to the inherent weakness of some of these constituents or to poor adhesion between certain constituents. Analyses of degraded coal Sampling: Samples were secured from 2 stock piles of 2-by 4-inch coal in a dealer's yard. The coal in both piles had received the same treatment from mining to storage, although it had been mined and shipped on two different days. The size consist of one pile, however, appeared smaller than that of the other. The sample from the degraded pile is referred to as the slack sample, and the other is called the unslack sample. Size analyses: In order to determine the actual size consist, the samples were screened, with a minimum of rough handling, on square 2-inch, 1-inch, 1/2-inch, and 12-mesh screens. It is assumed that 100 percent of the 2-by 4-inch coal is retained on a 2-inch round screen at the tipple. Analyses of several loads at the tipple indicate that approximately 26 percent of the 2- by 4-inch coal, 100 percent of which was assumed held on a 2-inch round screen, passed through a 2-inch square screen. Twenty-six percent may be taken as the normal difference between round and square screens. Results of the analyses of stockpile coal are shown in Table 1. Cumulative percent versus size are plotted on a Rosin-Rammler chart as Figure 1. Both samples approximately follow normal size distribution; this indicates that breakage is not size preferential and also that the samples are representative of the stockpiles. Although both samples exhibit the same general size distribution, the median size of the slack sample is considerably smaller than that of the unslack sample. Fifty-five percent of the unslack sample was undersize (29 percent greater than the normal 26 percent difference), and 83 percent of the slack sample was undersize (57 percent greater than the normal 26 percent difference). Both piles contained a greater percentage than normal of undersize coal; this indicates that they were both affected by degradation, but to a different degree. Bedding-plane analyses: Coal, because of its laminated nature, will break preferentially along bedding planes, especially where adjacent constituents are poorly cemented to each other. Subsequent breakage perpendicular to the

Jan 1, 1961

By Robert B. Fulton

The use of industrial minerals by the chemical industry as raw materials for manufacturing and in processing spans a wide assortment of minerals. This chapter aims to supplement rather than duplicate the commodity chapters. Its objective, to describe the uses of industrial minerals in the manufacture of chemicals, is attained by considering minerals used by the chemical industry in terms of the chemical element required and by giving particular emphasis to market factors. The mineral aspect, which is covered in the commodity chapters, is subordinated and attention is focused on the chemical and market aspects. Condensing this broad subject into a few pages permits treating only the most important elements derived from industrial minerals. The elements derived from petroleum, hydrogen and carbon, which quantitatively dominate as raw materials for the chemical industry, are omitted, as are the metallic elements and the minerals covered in other "use" chapters such as phosphorous, potash, and nitrogen for fertilizers and titanium dioxide for pigments. After eliminating on this basis, our definition leaves six elements of major importance: boron, bromine, chlorine, fluorine, sodium, and sulfur. These elements are treated individually under separate headings. Salt brines have particular importance as raw material sources for the chemical industry as shown in Table 1 "Salt Brine Derivatives" (Anon., 1967), which charts the chemical compounds derived from four types of brines: (1) Owens Lake type brines which are sources of boron and sodium compounds; (2) Midland type brines from which bromine, iodine, and chlorides of calcium, magnesium, potassium, and sodium are derived; (3) Searles Lake type yielding boron, bromine, lithium, magnesium, potassium, and sodium compounds; and (4) Silver Peak type produced mainly for lithium. Table 2 (Jones, 1973) affords an overview of the chief industrial minerals, the chemical products derived from them, and end uses of the products.

Jan 1, 1975

By Job Goostray

IN THIS paper, as in so many other discussions of historical nature, there is little chance for original material and much has had to be rewritten from older papers, documents, accounts, reports, and the like, some of which are of doubtful authorship or contain conflicting statistics. Fully aware of these little historical pitfalls, the authors have tried to quote the origin where possible, and also to seek further confirmation when the statement warranted it, but cannot assume responsibility for certain facts and dates. The Encyclopedia Britannica was frequently consulted. The authors wish to acknowledge assistance from MacKintosh-Hemphill Co.-formerly the Fort Pitt Works-and also from officials of the Watertown Arsenal who loaned valuable books and gave much necessary information. They are particularly indebted to Colonel Dixon of that institution. The data on modern guns were furnished by Mr. F. Brauer of the Watertown Arsenal; Mr. H. S. Chaffee assisted by obtaining records and data of work carried on by the Builders' Iron Foundry of Providence, R. I.

Jan 3, 1925

By Curtis Wilson

FROM the results of his investigation of the hardenable copper-nickel-silicon and copper-cobalt-silicon alloys, M. G. Corson1 explained the dispersion-hardening effect by assuming the precipitation of the inter-metallic compounds nickel silicide and cobalt silicide, and on this basis he constructed diagrams for the pseudobinary systems copper-nickel silicide and copper-cobalt silicide. A consideration of these diagrams and of the factors that exert such an important influence upon the rate of precipitation, such as degree of supersaturation, nucleus number and rate of diffusion, leads to the expectation that the rate of precipitation of the silicides on reheating the quenched solid solutions at temperatures below the alpha-phase boundary should increase continuously as the reheating temperature is raised. The present investigation shows that while this rate does increase with the reheating temperature in a general way, it does not do so continuously. Supplements to Corson's diagrams are offered, showing the discontinuities, and slight modifications in the solubility curves are suggested.

Jan 1, 1933

By E. B. Brauer, W. Hurst, J. D. Clark

Because of drilling, completion, and workover practices, the permeability around a wellbore generally is different from the permeability of the formation. The zone with the altered permeability is called "skin" and its effect on the pressure or flow behavior of the well is called "skin effect". Hawkinsl has shown that the radius and the permeability of this zone are related to the skin by: If the permeability in the skin zone is less than that of the formation, the skin is positive; if it is more than that of the formation, the skin is negative. If the two permeabilities are equal, s is zero; that is, there is no skin. Van Everdingen2 and Hurst3 have given mathematical solutions to the case of a zone of reduced permeability around the wellbore. This skin effect is illustrated in Fig. 1. They treat the positive skin as a zone of reduced permeability of infinitesimal thickness around the wellbore. When applied to a well with negative skin, however, their solutions lead to the calculated flowing well pressure which is smaller than the formation pressure, i.e., an injection situation. This is a physical contradiction. In this paper, it is shown that this mathematical difficulty can be overcome by assuming an effective wellbore radius larger than the actual wellbore. Existing solutions2"4 are modified to include the effect of an enlarged wellbore radius, and thus enable the engineer to deal with negative as well as positive skins. Theoretical Development Van Everdingen and Hurst,4 through the use of Laplace transformation, obtained the following expression for transient fluid influx into a wellbore: The values for QtD, reported extensively by van Everdingen and Hurst,4 are obtained from the following integral:

Jan 1, 1970

By Fulton Flick

THE micrography and macrography of aluminum and its alloys present certain difficulties. Many of the difficulties attendant on the micrography have been removed by methods developed during the past few years, notably those of E. H. Dix, Jr.1 The development of grain size in aluminum and its alloys may be equally troublesome, yet comparatively little detailed attention appears to have been given to it. Satisfactory reagents. are available for the development of such structures in ferrous and most non-ferrous metals, but those applied to aluminum and its alloys, i. e., alcoholic or aqueous solutions of sodium hydroxide or hydrofluoric acid are not entirely satisfactory. Hanson and Archbutt,2 in speaking of these two reagents, state that "These two reagents give fairly satisfactory results, but they sometimes produce badly tarnished surfaces from which the surface deposit is difficult to remove." It has been our experience that neither of these reagents is to be entirely relied on for developing structures suitable for study or photography. This means that when the etch fails to produce satisfactory results the, specimen must be repolished, through the 000-paper at least, and then etched with another reagent. This procedure entails a con¬siderable loss of time, and the final results may not be all that might be desired. However, the reagent to be described, although not a panacea for all etching troubles, has improved results so markedly that it is felt that it is worth offering as a grain-size etch generally applicable to aluminum and aluminum alloys. Czochralski3 in 1915 mentioned the application of a solution of 10 to 20 c. c. concentrated hydrofluoric acid in 100 c.c. absolute alcohol, etching in this reagent being followed by dipping in concentrated hydrochloric-acid solution.

Jan 2, 1925

By W. W. Krech, T. E. Perkins

As fractures are propagated through rocks, energy is absorbed near the extending crack tip. Apparent surface energies for several rocks have been measured by cleavage under dynamic con-ditions. At nominal crack velocities from 0.5 to 500 in (min, measurements showed that fractures propagated in discrete jumps. Calculated surface energies and moduli were relatively insensitive to nominal rate of cleavage. in another set of experiments, rocks were cleaved under high confining stresses. The rocks were submerged in low leak-off fluids which formed a filter cake on the freshly cleaved surfaces (similar to the hydraulic lracturing process). Apparent surface energies were increased substantially as the surrounding fluid pressure was increased. Moduli in bending increased significantly upon application of the first 1,000 psi but were insensitive to stress level at greater pressures. INTRODUCTION For almost 20 years, hydraulic fracturing processes have been utilized effectively to stimulate oil and gas wells. During this period, some process improvements have resulted from studies of fracture orientation, mechanics of fracturing, areas generated, conductivities of cracks, etc. Yet many questions remain concerning the conditions and pressures needed during fracture propagation. In this paper we will report additional studies of the mechanics of fracture extension. It was shown previously3 that large rock samples could be cleaved under controlled conditions so as to measure the apparent surface energy (that amount of energy absorbed per unit area of new surface created). In this paper we consider the effects of two additional factors on surface energies, viz.: (1) effect of cleavage rate and (2) effect of confining stress level. THEORY Cleavage experiments were conducted on rock samples similar to that illustrated in Fig. 1. Blocks of rock several inches wide, 2- to 3-in. thick, and up to 3 ft in length were grooved longitudinally with shallow guide slots. A crack was initiated and allowed to extend along the web as the top of the rock specimen was pulled (or pushed) apart. Auxiliary equipment permits the measurement of (1) force applied at the top, (2) separation at the top and (3) crack length. (Further experimental details will be given in the next section.) The rock beams created by the crack are considered ton be cantilever beams. The deflection (or separation of the rock beams) at any point is calculated 4,5 by the beam Eq. 1.

Jan 1, 1967

Jan 1, 1969