Search Documents

By H. R. Brannon, R. J. Robinson, G. W. Kattawar

Accurate prediction of storm wave characteristics are needed for design of offshore structures. Statistical methods of random noise analysis provide techniques for predicting required wave properties. These techniques have been used to analyze and characterize storm wave profiles from Gulf of Mexico recording installations in 34, 65 and .98 ft of water. Correlations based on the results can be used to predict wave crest probabilities and wave shapes for a range of Gulf water depths and storm conditions. Example predictions of wave crest probability as a function of water depth for a particular set of storm conditions are given. INTRODUCTION Accurate predictions of wave crests and wave shapes are needed for design of offshore structures. With predictions of wave crests, platform deck elevations giving adequate protection from exposure to storm waves can be selected. Use of hydro-dynamic wave theories with the wave crest and wave shape predictions permits an estimate of the wave forces an offshore structure must withstand. Observation of ocean waves suggests use of statistical analysis for studying wave characteristics for water depths beyond the breaking wave zone; recent wave generation theories lend support to this approach. l-3 One of the most powerful methods of statistically analyzing wave profiles was introduced by Cartwright and Longuet-Higgins; most of the basic relations used were developed by Rice5.6 in studies of random noise theory. This method predicts wave crest elevation probabilities from two parameters characterizing a wave profile. The statistical approach also provides a means for predicting wave shapes. In the random noise model of a wave profile, surface elevation is represented as an infinite sum of sine waves with closely spaced frequencies and random phase angles. Power density is proportional to the sum of the squares of the amplitudes of sine waves having frequencies within a narrow band, and the distribution of power density as a function of frequency is called the power density spectrum, or power spectrum. The power spectrum characterizes an irregular sea, and hence finds use in motion studies of ships, barges and semisubmersible drilling platforms. From the power spectrum, a wave profile of any duration of time can, in principle, be calculated; this long wave profile depicts the many wave shapes to which a structure may be exposed. Thus, the statistical methods of wave analysis provide an approach to selecting wave shapes as well as wave crest elevations needed for design of offshore structures. For practical application of these techniques, power spectra and related quantities must be predicted from storm properties. Barber and Tucker 7 have reviewed correlations of wave properties and storm conditions. Their review summarizes work of Darbyshire,8 among others, in which a correlation of power spectra with wind intensity of the North Atlantic was developed. These results, plus more recent work by Pierson and Moskowitz and Kitaigorodskii,10 establish feasibility of correlation, but there is no theoretical basis for modifications extending the relations to other areas as remote as the Gulf of Mexico. Hence, to predict characteristics of Gulf of Mexico waves, a study of local observations is needed. This paper presents (1) a practical approach to computing wave profiles hat depict shapes of waves for use in force calculations, (2) a summary of relations for predicting probabilities of wave crest elevations, (3) correlations of parameters needed to apply these methods in the Gulf of Mexico and (4) examples of application of the techniques. THEORETICAL BACKGROUND The following sections summarize relations needed to calculate wave profiles and to estimate wave crest probabilities. These relations also identify quantities that must be related to storm

By Donald Ofte, L. J. Wittenberg

The absolute viscosity coefficients and activation energies for viscous flow of bismuth, lead, and zinc are reported. The viscosities were measured in an oscillating cup viscosimeter from nearly 1000°C to within 1 deg of the melting point of zinc; 2 deg of the melting point of lead; and in the supercooled condition of bismuth, 9 deg below the melting point. Two methods of relating the change in the viscosity coeflicient with temperature were compared. From the plot of the logarithm of viscosity (77) as a function of reciprocal absolute temperature (1/T), the equations of the lines which fit the data are: AS a part of the Reactor Fuels and Materials Research program at Mound Laboratory, precise measurements of the physical properties of liquid metals are of interest because the desirability for a low vapor pressure liquid coolant in high-temperature nuclear power plants has focused attention on the properties of the liquid metals.' Also, a clearer understanding of the liquid metal state can be achieved when better measurements are available. For these reasons, the viscosity values of certain liquid metals have been determined. The viscosities of the liquid metals, bismuth, lead, and zinc are measured in an oscillating cup viscosi-meter. In this apparatus the liquid is sealed in a right-circular cylindrical cup which is attached to a torsion fiber so that the system forms a torsion pendulum. The damping effect of the molten metal upon the normal oscillations of the torsion pendulum is a function of the viscosity of the liquid. This method is advantageous because of the small amount of sample required (8 to 10 cc), the ease of remote operation of the apparatus in a vacuum system, and the containment of the liquid metal in a hermetically sealed, nonreactive, tantalum crucible. With the liquid metal sealed in the crucible, loss of liquid by vaporization and chemical reactions with atmospheric gases are prevented. Additionally, an absolute viscosity determination is obtained which avoids the sources of possible error in a relative measurement. The difficulties in a relative method include the repetitious cup calibrations, and the uncertainty inherent in the practice of calibration of the apparatus at room temperature and extrapolation to the high temperatures necessary for the investigation of most liquid metals. The oscillating cup viscosimeter was first described by Meyer in 1891.2 The first complete mathematical treatment offering an absolute measurement of viscosity with this type of apparatus was proposed by Andrade and chiong3 in 1936 for an oscillating sphere filled with a liquid. This solution was modified to include a right circular cylinder by Hopkins and Toye4 in 1950, and R. Roscoe5 in 1958. Roscoe's treatment for the calculation of the viscosity of a liquid at a particular temperature from the logarithmic decrement and period of oscillation of the torsion pendulum was used in this investigation. A detailed description of the method of calculating the viscosity from these measurements has been reported earlier.' This calculation has been programmed for the IBM 704 and IBM 1620 computers for expedition of the otherwise laborious calculations. APPARATUS The cutaway view of the apparatus in Fig. 1 shows the suspension system within the high vacuum enclosure. The low residual gas pressure. maintained between 1 x 10-4 and 5 X 10-6 mm Hg during a determination, avoided both air drag on the pendulum and

Jan 1, 1963

By Paul A. Beck, John Towers, W. D. Manly

Recent work on grain growth in high purity aluminum and in a solid solution type alloy of aluminum and magnesium1 showed that the isothermal increase of the average grain diameter D with time follows the equation D = K(tg + A)2 Eq 1 Here tg is that portion of the total annealing time 1, which is available for grain growth, that is the annealing time less the time necessary for complete recrystallization. The parameters A and n depend only on the temperature. A has the dimension of time, and is usually of the same order of magnitude as the time for recrystallization: R. In some instances R may be substituted for A in Eq I. This is the case for example at long periods of annealing or at high temperatures, where both A and R become small in comparison with t. In such instances Eq I is simplified to D = K, tn Eq 2 For high purity aluminum the exponent n increases in the ratio of about I to 5 in. the temperature range of 350' to 6w°C. 70-30 brass has been long known to exhibit grain growth of the continuous type, such as occurs in pure metals. For this reason, it was of interest to determine whether grain growth in brass also follows Eq I and 2 as it does in aluminum. Isotherma1 grain growth data for 70-30 brass were published in recent years by R. S. French2 and by H. L. Walker.' Fig I shows a log D vs. log t plot of French's data, together with the 500, 600 and 700°C data by Walker for 42.8 pct reduction by rolling. In such a log D vs. log t plot Eq 2 corresponds to a straight line with a slope of n. As seen in Fig I, the data of French and the 600°C data of Walker can be represented by straight lines. However, the 500 and 7c0°C data of Walker give curved lines and the average slope of his 500°C line is very different from the slope of the line representing French's data for the same temperature. Furthermore the curvature of the 700°C line suggests a tendency for grain growth to stop after 8 hr. If confirmed, this would constitute a definite deviation from Eq 2, not explainable on the basis of the specimen thickness effect found in high purity aluminum.' Walker's specimens were 0.125 in., or 3.18 mm thick, while the deviation occurs at a grain size of approximately 0.4 mm. A further point of divergence arose recently, when J. E. Burke suggested4 on the basis of Walker's data, that, instead of Eq 2, grain growth in 70-30 brass follows the equation D - D, = K . tgn Eq 3 Here D, is the grain size as recrystallized at the temperature in question. Although it was possible to show5 that, for pure aluminum, for which isothermal grain growth has been carefully investigated, Eq 3 is incompatible with the experimental facts, the situation remained somewhat obscure with respect to 70-30 brass. In view of the discrepancies between the published data on brass from different

Jan 1, 1949

By John Towers, Paul A. Beck, W. D. Manly

Recent work on grain growth in high purity aluminum and in a solid solution type alloy of aluminum and magnesium1 showed that the isothermal increase of the average grain diameter D with time follows the equation D = K(tg + A)2 Eq 1 Here tg is that portion of the total annealing time 1, which is available for grain growth, that is the annealing time less the time necessary for complete recrystallization. The parameters A and n depend only on the temperature. A has the dimension of time, and is usually of the same order of magnitude as the time for recrystallization: R. In some instances R may be substituted for A in Eq I. This is the case for example at long periods of annealing or at high temperatures, where both A and R become small in comparison with t. In such instances Eq I is simplified to D = K, tn Eq 2 For high purity aluminum the exponent n increases in the ratio of about I to 5 in. the temperature range of 350' to 6w°C. 70-30 brass has been long known to exhibit grain growth of the continuous type, such as occurs in pure metals. For this reason, it was of interest to determine whether grain growth in brass also follows Eq I and 2 as it does in aluminum. Isotherma1 grain growth data for 70-30 brass were published in recent years by R. S. French2 and by H. L. Walker.' Fig I shows a log D vs. log t plot of French's data, together with the 500, 600 and 700°C data by Walker for 42.8 pct reduction by rolling. In such a log D vs. log t plot Eq 2 corresponds to a straight line with a slope of n. As seen in Fig I, the data of French and the 600°C data of Walker can be represented by straight lines. However, the 500 and 7c0°C data of Walker give curved lines and the average slope of his 500°C line is very different from the slope of the line representing French's data for the same temperature. Furthermore the curvature of the 700°C line suggests a tendency for grain growth to stop after 8 hr. If confirmed, this would constitute a definite deviation from Eq 2, not explainable on the basis of the specimen thickness effect found in high purity aluminum.' Walker's specimens were 0.125 in., or 3.18 mm thick, while the deviation occurs at a grain size of approximately 0.4 mm. A further point of divergence arose recently, when J. E. Burke suggested4 on the basis of Walker's data, that, instead of Eq 2, grain growth in 70-30 brass follows the equation D - D, = K . tgn Eq 3 Here D, is the grain size as recrystallized at the temperature in question. Although it was possible to show5 that, for pure aluminum, for which isothermal grain growth has been carefully investigated, Eq 3 is incompatible with the experimental facts, the situation remained somewhat obscure with respect to 70-30 brass. In view of the discrepancies between the published data on brass from different

Jan 1, 1949

By B. H. Kear, H. G. F. Wilsdorf

After a few percent strain, dislocations in disordered Cu3Au are arranged in groups in ulell-defined slip planes, which contrasts with the more or less random distribution of dislocations in the corresponding ordered alloy. Unusually long and straigkt dislocation segments observed in the ordered alloy are interpreted in terms of cross slip of screw dislocations from their slip planes into cube planes. It is proposed that cross slip, and also dislocation reactions of the Lomer type, play an important role in work hardening of the ordered alloy. ACCORDING to to dislocations in the disordered lattice are only partial dislocations in the corresponding super lattice, so that in the super-lattice dislocations must occur as pairs, with each pair connected by a ribbon of antiphase boundary. Recently, striking confirmation of the theory was obtained by Marcinkowski, Brown, and Fisher3j4 in a study of dislocation configurations in thin foils of the Cu3Au alloy by transmission electron microscopy. In the present investigation, the observations made by these workers have largely been confirmed. Moreover, it has been found that dislocations in the plastically deformed ordered alloys are often unusually long and straight. This paper is mainly concerned with these new observations. A tentative explanation is offered in terms of cross slip of dislocations in the ordered alloys. PREPARATION OF SPECIMENS An alloy close to the stoichiometric composition Cu3Au was prepared by melting together high-purity copper and gold (99.995 pct) under vacuum in a graphite mold. The resulting ingot was rolled down and cut into strips 1 1/2 by 1/8 by 0.005 in. in thickness. All strips were annealed at 900°C for 3 hr and cooled to 500'~. A few of the strips were quenched from ~OO'C, while the remainder were slowly cooled in the temperature range 400" to 150°C at intervals of a few degrees every 24 hr to produce a high degree of order. After this annealing treatment the average grain size of the material was about 0.1 mm. Material in different states of order was obtained by reannealing fully ordered material at temperatures below the critical temperature: followed by quenching. The equilibrium degree of order attained at a particular reannealing temperature was obtained from the paper by Keating and warren.' A simple hand-operated tensile jig was used to give the strips the required amounts of plastic strain. Following straining, a number of specimens were cut from each strip by means of an electrolytic cutting device.= Finally, each specimen was thinned electrolytically using the polishing procedure described by Bakish and Robertson,7 and these thin sections were examined in the Philips EM-100B electron microscope operating at 100 kv. EXPERIMENTAL OBSERVATIONS Antiphase boundaries. The average domain size in the ordered alloys was found by direct observation to be about 0.25 p. Also, it was observed that the grown-in configuration of antiphase boundaries was changed little by plastic deformation, at least up to 6 pct strain. Fig. 1 shows an electron micrograph of a thin foil of partially ordered alloy (S - 0.8) after 3 pct strain. The micrograph shows a few dislocation pairs and a somewhat crystallo-graphic network of antiphase boundaries. The normal to this foil was found by diffraction analysis to be approximately 11121, with principal directions in

Jan 1, 1962

By I. W. Alcorn

The cost of drilling in the past few years of proration and ever decreasing allowables has received increasing thought and study. It seems to parallel the strides made with respect to pumping problems in 1932 and 1933. Low profit margins stimulate the search for economies. Any discussion of slim-hole drilling involves a multitude of abstract values, and the problem can be attacked only by a study of records and rational thinking. In a measure, traditional methods must be studied with a questioning attitude. The words "slim hole" may be somewhat misleading, as the term is entirely relative. For that reason I wish to define it as used in this discussion. On the Gulf Coast, I would define a normal hole as one in which 10 3/4-in. surface casing has been set and below which a 9 7/8-in. hole is dug. With this as a datum, holes of larger diameter below surface would be called "large holes," with 121/4 in. as the "extra large" class and holes of smaller diameter classed in the slim-hole range, 6 1/4 in. representing about the minimum size. The optimum size has not been definitely determined, but the preference is for a program that would call for setting 8 5/8 -in. surface casing and digging a 7 3/4 - in. hole. Different sections of the country might very properly deviate from this practice because of differences in depth, formations, regulations, or other factors. This paper will be confined to a discussion of the drilling of exploratory wells, as the drilling of this class of wells is quite a different problem from that of drilling a proven field, although many of the same considerations apply in both cases, as will be apparent. Although for many years shallow slim-hole and core-test holes of small diameter had been dug, it was not, I believe, until the fall of 1937 that serious consideration was given to its application in drilling wells 7000 ft., 8000 ft. and 9000 ft. deep on the Gulf Coast. This does not mean that nearly all such wells drilled on the Gulf Coast now are of the slim-hole variety, but only that efforts have been aimed in that direction with very encouraging results and there has been a very definite trend to reduce the size of holes, a trend that probably will continue. As to exploration policy, there appear to be two distinct and opposite views. The one usually followed in the past was more or less designed to "shoot the works" on one well. Generally on the Gulf Coast, the depth at which production might be obtained is limited by the drilling equipment and economics. With this thought in mind, wildcat wells are often planned with the idea of carrying the depth to 10,000 It., or even 15,000 ft. All of the possible hazards are anticipated and the general drilling and casing programs are so predicated. Such wells contemplate the drilling of large-diameter holes; the setting of at least one protective string of casing and an extensive coring policy, all of which finally adds up to an expensive well. AS opposed to this general policy is the slim hole. This, on the Gulf Coast, recognizes that new discoveries represent gam-

Jan 1, 1941

By I. W. Alcorn

The cost of drilling in the past few years of proration and ever decreasing allowables has received increasing thought and study. It seems to parallel the strides made with respect to pumping problems in 1932 and 1933. Low profit margins stimulate the search for economies. Any discussion of slim-hole drilling involves a multitude of abstract values, and the problem can be attacked only by a study of records and rational thinking. In a measure, traditional methods must be studied with a questioning attitude. The words "slim hole" may be somewhat misleading, as the term is entirely relative. For that reason I wish to define it as used in this discussion. On the Gulf Coast, I would define a normal hole as one in which 10 3/4-in. surface casing has been set and below which a 9 7/8-in. hole is dug. With this as a datum, holes of larger diameter below surface would be called "large holes," with 121/4 in. as the "extra large" class and holes of smaller diameter classed in the slim-hole range, 6 1/4 in. representing about the minimum size. The optimum size has not been definitely determined, but the preference is for a program that would call for setting 8 5/8 -in. surface casing and digging a 7 3/4 - in. hole. Different sections of the country might very properly deviate from this practice because of differences in depth, formations, regulations, or other factors. This paper will be confined to a discussion of the drilling of exploratory wells, as the drilling of this class of wells is quite a different problem from that of drilling a proven field, although many of the same considerations apply in both cases, as will be apparent. Although for many years shallow slim-hole and core-test holes of small diameter had been dug, it was not, I believe, until the fall of 1937 that serious consideration was given to its application in drilling wells 7000 ft., 8000 ft. and 9000 ft. deep on the Gulf Coast. This does not mean that nearly all such wells drilled on the Gulf Coast now are of the slim-hole variety, but only that efforts have been aimed in that direction with very encouraging results and there has been a very definite trend to reduce the size of holes, a trend that probably will continue. As to exploration policy, there appear to be two distinct and opposite views. The one usually followed in the past was more or less designed to "shoot the works" on one well. Generally on the Gulf Coast, the depth at which production might be obtained is limited by the drilling equipment and economics. With this thought in mind, wildcat wells are often planned with the idea of carrying the depth to 10,000 It., or even 15,000 ft. All of the possible hazards are anticipated and the general drilling and casing programs are so predicated. Such wells contemplate the drilling of large-diameter holes; the setting of at least one protective string of casing and an extensive coring policy, all of which finally adds up to an expensive well. AS opposed to this general policy is the slim hole. This, on the Gulf Coast, recognizes that new discoveries represent gam-

Jan 1, 1941

By J. Geertsma

In order to obtain a better insight into the pressure-volume relationship of reservoir rocks a theory of pore and rock bulk volume variations is presented. The theory is independent of the shape of the pores but is restricted to isotropic porous media built up of continuous homogeneous matrix material. The main conclusion obtained from this theory is that only three elastic constants and three viscous constants are required for describing pore and rock bulk volume variations if the porosity is explicitly introduced into the treatment. In addition, reasonable approximations are introduced for various types of reservoir rock, e.g., sandstones, limestones, and shales, which lead to further simplifications of the basic formulas. In consequence there is then a further reduction in the number of deformation constants which have to be determined experimentally. It is shown how measurements of these remaining deformation constants can be performed most conveniently. Finally the application of the theory to reservoir studies is discussed and the translation of experimental results obtained in the laboratory into reservoir behavior is considered. INTRODUCTION The decline of fluid pressure in connection with the withdrawal of fluid from an underground reservoir gives rise to a change in volume of both reservoir fluids and reservoir rock. The volume variation of the reservoir rock results in a decrease of both the pore volume and the total volume of the fluid-filled formation. Whereas the variation in volume of the reservoir fluids with pressure is usually known from PVT analysis, that in the volume of the porous medium is rarely measured, as it is considered of minor importance in reservoir engineering. Nevertheless, certain experimental results' suggest that in a number of cases the neglection of the variation in pore volume may introduce errors into material balance calculations of reservoirs producing above the bubble point. Only a few experimental results on the compressibility of porous rocks have been published. The compressibility of pure quartz and of calcite can be found in a paper by Adams and Williamson'. Botset and Reed' carried out some experiments on the compressibility of loose sands, while Carpenter and Spencer' measured the compressibility of a number of consolidated sandstones, which data were supplemented by Hughes and Cooke5. Hall' finally reported a number of experiments designed to study the influence of the porosity on the compressibility of oil-bearing formations, from which it appeared that, particularly in the low porosity range, the compressibility of the pore volume is important enough to be taken into account. Difficulties, however, arise in applying these experimental results to field cases. This is partly due to the fragmentary nature of the published data. Yet even if more complete data were available, the need for a theoretical background would be strongly felt. Such a background can be partially found in studies carried out by Biot,6,7 who published a theory of elastic deformations of porous materials and their influence on fluid displacement within the pores, by Lubinski8, who studied the stress distribution in an elastic porous medium neglecting pore compressibility, and by Gass-mann" who mathematically analysed the elasticity of packings of spheres. The contributions of Biot intro-duce a number of deformation constants which are impractical for reservoir rocks from an aspect of experi-mental determination. The theory developed by dass-mann is correct but difficult to read. It appeared, how-

Jan 1, 1958

By C. D. Cochran, F. H. Deily, M. M. White, Victor Horner

A laboratory study of the effects of physical and easily measured rheological properties upon dynamic filtration was carried out. The purpose was to determine the factors affecting dynamic filtration and thus learn how to control this important property. The recommended procedure for determining the amount of fluid loss prescribed by API Code RP 29 "Field Procedure for Testing Drilling Fluids,"' is very useful for mud conditioning purposes, but inadequate for comparing filtration properties of different muds under dynamic conditions. Where dynamic filtration control becomes important for purposes such as logging and coring, other means of evaluating filtration properties are needed. A special device and testing technique were developed to provide reliable data on filtration characteristics of different fluids under dynamic conditions. Liquid saturated rock samples were drilled with a microbit, and the amount of connate liquid displaced by the fluid being investigated was measured. The empirical testing procedure was considered reliable. A large number of laboratory-prepared and field-mixed dritling fluids with a wide range of physical properties were studied. These properties included: (I) shape of consistency curves, (2) apparent viscosity, (3) plastic viscosity, (4) yield value, (5) density, and (6) API filter loss. Test results indicate that very little, if any, trend could be established between changes in these fluid properties and dynamic filtration. INTRODUCTION During the process of drilling, excessive loss of liquid from the mud to the formation is undesirable for many reasons. Filtration can reduce production-zone permeability, cause sloughing of formations, affect electric log curves, and render the fluid data obtained from cores unreliable. Thus, it is very important that we gain a knowledge of the factors that control filtration. It is generally agreed that in any type of filtration (static, dynamic, or combinations thereof), the mechanics of filtering depend in great part upon the composition of the fluid. Moreover, it is also true that the physical and rheological properties are influenced by the composition of the fluid. Indirectly, then, it was hypothesized that a relationship existed between dynamic filtration and physical-rheological properties, two effects of a common cause—fluid composition. This paper discusses an attempt to relate filtration-while-drilling to physical and easily measured rheological properties. It is believed that the results of this study may prove useful to others in the industry who are working on the problem of understanding filtration. The industry accepted practice for determining fluid loss from drilling mud is described in API Code RP 29. The practice is to measure the amount of filtrate that passes during a given interval of time from a quiescent drilling fluid under applied pressure through a piece of filter paper. This is primarily a test of static filtration properties. Other investigators have studied filtration under dynamic conditions. Williams5 and Pro-kop5 studied radial filtration and filter-cake buildup by circulating fluid through a porous cylinder under conditions similar to actual mud flow. Nowak,4 Krueger,3,4 and Vogel3 studied the effect of filtration upon permeability by flowing mud past the face of rock Samples. In their tests, the bit action was simulated by mechanical scrapers that kept the filter cake from building up. Ferguson and Klotz2 studied three types of filtration: static, dynamic, and beneath-the-bit. They investigated each type separately, using a model well in which bottom-hole hydraulics and mechanics were closely approximated.

Jan 1, 1958

By J. F. Gormley

Drill pipe, and the connections used to join the pipe together, have had a long history of development and improvement. With the growing use of high tensile-strength pipe, the rotary shouldered connections used to join the lengths of drill pipe have become more critical. Fatigue-type failures of rotary-shouldered connections, or tool joints, have become a cause of much conern to operators and manufacturers. Most fatigue tests on tool joints have been made on the basis of bending moment applied vs number of cycles, rather than on the basis of stress vs cycles. A desire to learn more about the nature of stress in a tool joint due to bending caused a rather intensive experimental study to determine what actually happens to cause fatigue failure. It was known that a tool joint with insufficient make-up torque was subject to early fatigue failure. Therefore, the establishment of the relationship between make-up torque and allowable bending moment would enable a testing engineer to better correlate the results of various tests. The problem of calculating the stress in a tool joint due to an applied bending moment was complicated by: the effect of unknown stress concentrations; the applicability of the well known formula S = Mc/I; and the question of whether to use the combined cross section of the pin and box in calculating the moment of inertia I. TEST EQUIPMENT The test equipment consisted of the following. 1. A lathe modified to act as a fatigue machine for full-size tool joint or drill-collar specimens. 2. An adapter plate bolted to the headstock of the lathe and into which a short adapter was screwed. 3. Tool-joint pins on sections of drill collars bucked into the adapter. 4. A hydraulic load cylinder which rested on the bed of the lathe and applied bending moments to the joints. 5. Paper-backed strain gauges (FAP-12-12, 1/8-in. gauge length and A-5-1, 1/2-in. gauge length). 6. Eastman 910 contact cement for bonding gauges to joint and Epoxylite 222 waterproofing. 7. A 12-position switchbox, SR-4 bridge balancing unit connected to an SR-4 strain indicator to read the gauge output. 8. A four-channel Sanborn recorder, Model 150, was used as the readout instrument for part of the testing. The fatigue machine was well suited as a loading device. Because of the long lever arm available, the change of moment with distance was small; consequently, the vertical shear stress was small, less than 300 psi. Some difficulty was experienced with the hydraulic load cylinder at high bending moments due to leakage. The strain gauge equipment was simple and quite reliable. A 41/2-in. extra hole (E. H.) joint with "as machined" threads and shoulders and a 2 7/8-in. internal flush (I. F.) joint with copper-plated threads and shoulders were selected for testing because of their common usage in the field. In addition, the 4 1/2-in. E. H. was suitable because it had an inside diameter large enough to permit the insertion and accurate positioning of strain gauges without using special jigs, while the 2 7/8-in. I. F. joint required much lower bending loads to cause separation of the shoulders of the joint when made up to the recommended torque. The joints were made to standard API specifications, except for very small bevels on the box and pin shoulders and cylindrical stress-relief grooves added to the pins. Larger bevels would have

By M. B. Bever, A. B. Michael

THE structure of cast A1-Si alloys is altered profoundly by modifying agents. Sodium, in particular, reduces the size of the silicon particles in the eutectic and tends to change their shapes from plates to nearly rounded grains. This refinement is accompanied by a shift of the eutectic to higher silicon concentrations and lower temperatures. Similar effects can be attained by rapid solidification. Hypotheses advanced to explain modification have involved the formation of a ternary eutectic,1 the removal of oxides by fluxing,' and even a presumed allotropic transformation of silicon. Other hypotheses assume that the modifying agent causes an envelopment of the silicon particles by the aluminum-rich solid solution, but attribute this envelopment to different mechanisms. According to various explanations, the growth of the silicon particles is obstructed by an immiscible sodium-rich liquid," an adsorbed layer of sodium,' or a ternary compound;5 his obstruction has been conceived as purely mechanical or as a colloidal action.' It has also been proposed that sodium reduces the surface tension on the Al-Si interface, thus permitting the aluminum-rich solid to grow faster and to envelop the silicon particles.? Still other hypotheses stress undercooling and the attendant change in the rate of nucleation of silicon, but fail to ascribe any definite function to the sodium addition." '" Procedure High-purity alloys of aluminum and 3 or 7 pct Si were treated with modifying agents containing radioactive sodium tracers. The experiments were intended to investigate the partition of sodium between the primary aluminum-rich solution and the eutectic, and, if possible, to determine whether sodium is concentrated in or around the silicon particles in the eutectic. Radioactive sodium fluoride, alone or mixed with stable sodium chloride, was allowed to react with the alloy for 5 min immediately before solidification in the crucible. Modification could not be attained with additions of less than the equivalent of approximately 0.1 pct Na. Some alloys were not completely modified, but comparison of the structures of treated and untreated specimens indicated a very considerable reduction in particle size of the silicon in all of the former. A treated alloy is shown in Fig. 1. Radioactive sodium fluoride was prepared by bombardment with deuterons having an energy of 14 mev. This bombardment (totaling, for example, 6 microampere-hours in one run) produced radioactive Na24 (half-life 14.8 hr) by the reaction 11Na23 (d,p) 11Na24. The radioisotopes of fluorine also produced have short half-lives and were not detected. A half-life determination by counting confirmed that the measured activity was due to Na24. The distribution of the sodium in the solidified alloys was determined by an autoradiographic technique by which a suitable stripping film is attached to a very thin metallographic specimen. After exposure and development, this emulsion remains superimposed on the specimen and may be examined microscopically or photographed by reflected light. The emulsion of the autoradiograph can be correlated directly with the underlying surface by simple adjustment in focus. The details of this technique have been described elsewhere." Results Fig. 2 is a micrograph of the autoradiographic emulsion ("autoradiograph") lying directly above the area of eutectic in Fig. 1. Fig. 3 shows a similar autoradiograph of a region consisting of primary solid solution. The small black dots (tracks) in these autoradiographs are developed grains in the emul-

Jan 1, 1954

By F. A. Veer, W. G. Burgers, B. H. Kolster

Starting-- from Cu-Si diffusion couples a study of diffusion in the intermetallic compound Cu3Si of the Cu-Si system has been made. Mmker experiments showed copper to be the diffusing element. From the rate of growth of the Cu3Si phase diffusion coefficients were derived. The activation energy for diffusion is 18.6 kcal per mole. The couples show some interesting microscopic features, for which a tentative explanation is given. THE synthesis of dimethyldichlorosilane, (CH3)2SiC12, an important compound in the manufacture of methyl-silicones, takes place at the surface of a thin Cu3Si layer deposited on a mass of silicon particles."' By passing CH3C1 vapor at about 350°C, silicon atoms are consumed from Cu3Si in forming the volatile silane molecules, thus leaving copper atoms at the surface. It is well-known that copper acts as a poison for the above-mentioned reaction by catalyzing the dissociation of methylchloride.' However, no poisonous effect is observed, at least not before most of the silicon from the particles is converted into silane.3 This would be understandable, if either the copper atoms diffuse through the Cu3Si deposit to the underlying silicon or silicon atoms diffuse in the reverse direction to the copper, both processes resulting in the formation of new Cu3Si and so ensuring the proceeding of the reaction. In this paper experiments are described, demonstrating copper to be the diffusing element. Also a very low activation energy for the diffusion is observed. EXPERIMENTAL The silicon used contained 1 ppm Mg and 1 ppm Cu; the copper was spectroscopically pure (Fe: 2 ppm; Ag, Mg, Si: 1 ppm). The copper and silicon samples were prepared in the form of blocks of approximately 10 by 10 by 5 mm with two parallel sides by spark-machining. The blocks were ground on emery paper (grit 600), etched (copper with nitric acid, silicon with a 4:l nitric acid:hydrofluoric acid mixture), and cleaned with alcohol. As the brittleness of the silicon did not allow a welding procedure to prepare the diffusion couples, the copper and silicon blocks were pressed together in a steel clamp. The pressure exerted by the clamp brought about contact between the parts of the couple and so ensured proceeding of the diffusion. The couples were heated in a vacuum furnace (about 10"5 Torr). The temperature was stabilized within 1 pct. After a heat treatment a section parallel to the direction of diffusion was examined in the microscope and in the X-ray scanning microanalyzer. To determine whether copper or silicon is the diffusing element marker experiments were carried out using tungsten wires with a 20-p diam. EXPERIMENTAL RESULTS Figs. 1, 2, 3, and 4 give micrographs of diffusion zones of Cu/Si couples. Analysis by the electron mi-croprobe of the Cu/Si samples after a definite heat treatment revealed the existence of only one intermetallic phase as can be seen from the line scan shown in Fig. 5. From the X-ray diffraction pattern it could be deduced that this phase was CU 3 Si.4 Even when diffusion zones of considerable width (up to 600 p) were obtained, no other phases occurring in the Cu-Si system could be detected, neither by the electron microprobe nor by X-ray diffraction analysis. The optical anisotropy of the Cu3Si phase made it possible to observe the columnar shape of the crystal-

Jan 1, 1969

By I. W. Alcorn

THE cost of drilling in the past few years of proration and ever decreasing allowables has received increasing thought and study. It seems to parallel the strides made with respect to pumping problems in 1932 and 1933. Low profit margins stimulate the search for economies. Any discussion of slim-hole drilling involves a multitude of abstract values, and the problem can be attacked only by a study of records and rational thinking. In a measure, traditional methods must be studied with a questioning attitude. The words "slim hole" may be somewhat misleading, as the term is entirely relative. For that reason I wish to define it as used in this discussion. On the Gulf Coast, I would define a normal hole as one in which 10 3/4-in. surface casing has been set and below which a 9 7/8-in. hole is dug. With this as a datum, holes of larger diameter below surface would be called "large holes," with 12 1/4 in. as the "extra large" class and holes of smaller diameter classed in the slim-hole range, 6 1/4 in. representing about the minimum size. The optimum size has not been definitely determined, but the preference is for a program that would call for setting 8 5/8-in. surface casing and digging a 7 3/4-in. hole. Different sections of the country might very properly deviate from this practice because of differences in depth, formations, regulations, or other factors. This paper will be confined to a discussion of the drilling of exploratory wells, as the drilling of this class of wells is quite a different problem from that of drilling a proven field, although many of the same considerations apply in both cases, as will be apparent. Although for many years shallow slim-hole and core-test holes of small diameter had been dug, it was not, I believe, until the fall of 1937 that serious consideration was given to its application in drilling wells 7000 ft., 8000 ft. and 9000 ft. deep on the Gulf Coast. This does not mean that nearly all such wells drilled on the Gulf Coast now are of the slim-hole variety, but only that efforts have been aimed in that direction with very encouraging results and there has been a very definite trend to reduce the size of holes, a trend that probably will continue. As to exploration policy, there appear to be two distinct and opposite views. The one usually followed in the past was more or less designed to "shoot the works" on one well. Generally on the Gulf Coast, the depth at which production might be obtained is limited by the drilling equipment and economics. With this thought in mind, wildcat wells are often planned with the idea of carrying the depth to 10,000 ft., or even 15,000 ft. All of the possible hazards are anticipated and the general drilling and casing programs are so predicated. Such wells contemplate the drilling of large-diameter holes; the setting of at least one protective string of casing and an extensive coring policy, all of which finally adds up to an expensive well. As opposed to this general policy is the slim hole. This, on the Gulf Coast, recognizes that new discoveries represent gam-

Jan 1, 1941

By R. A. Swalin, L. A. Neimark

The research work presented in this paper had initially a two-fold goal. First, further data concerning the low-temperature recovery of cold work was desired. Recovery phenomena have been extensively studied in face-centered-cubic metals but at the time of initiation of this research, no investigation concerning the low-temperature recovery in body-centered-cubic metals had been made. During the course of the investigation a paper by Martin on the recovery of point defects in molybdenum appeared.&apos; Comparison of our results on tungsten with Martin&apos;s results on molybdenum will be made. Secondly, knowledge concerning the influence of a particular type of impurity addition on the recovery kinetics was desired. It has been known for a long time that extremely small amounts of certain impurity additions exert a profound influence on the recrystallization behavior and the resulting microstructure of tungsten. Small amounts of potassium silicate and aluminum chloride added prior to the reduction of tungsten oxide to the metal have been found by Smithells to raise the recrystallization temperature markedly,2 and yield an ultimate microstructure in which grains several millimeters long are produced in fine wires of say 10 mil diam. If the impurities are added singly, ordinary secondary recrystallization results with none of the spectacular properties of the wire to which special additions were made. Swalin and Geisler3 and Rieck4 have studied the recrystallization behavior of such doped wire in some detail. The former workers found that small amounts of these impurities raised the recrystallization temperature over 400°C and electron microscopic examination revealed no obvious inclusions in the microstructures. In fact. chemical analysis showed that there was <0.001 pct K, <0.002 pct Al, trace of Si, and <0.0003 pct 0 in the wire studied. The possibility that the role of the special impurities is to keep the structure open by void formation and that submicroscopic porosity acts as the inhibiting agent has been discussed.2&apos;3 Meijering and Rieck5 have suggested that the impurities are strung out as second-phase particles in a direction parallel to the drawing direction. It was thought that a kinetic study of recovery could perhaps elucidate the role of the impurities to some extent. When a metal is deformed by cold work three basic types of defects are thought to be introduced into the crystal lattice, namely, interstitials, vacancies, and dislocations. Van Bueren6 has proposed that recovery of cold-worked metal is a four-stage process, each step being the result of the annealing of a primary defect or a combination of defects and having a unique activation energy. The steps in order of increasing activation energy are attributed to a) the diffusion of interstitials recombining with vacancies, b) diffusion of vacancy pairs, c) the diffusion of single vacancies, and d) climb of dislocations. In metals with relatively low- or medium-range melting points the temperatures for recovery are low enough so that the first three steps may occur during room-temperature deformation. For tungsten, because of the high melting point and subsequently high recovery temperature, it was calculated, using Nowick&apos;s approach7 that this recovery would occur in the temperature range 200" to 600°C. It is also known that recovery of X-ray line broadening, presumably due to dislocation climb, occurs at about 600°C.8 EXPERIMENTAL METHOD Two types of cold-rolled 10-mil-diam tungsten wire were used in the investigation; a) pure tungsten, (called undoped) and b) G. E. No. 218 wire (called

Jan 1, 1961

By Donald E. Grimes

In conjunction with Inland Steel&apos;s development of lead-bearing steels possessing improved machin-ability because of tellurium and/or selenium additions, it was decided to determine liquidus and solidus temperatures in the Pb-PbTe-PbSe sub-ternary system. With the exception of alloys in the PbTe-PbSe system, it is believed that no previous study of this subternary has been performed. Thermal analysis was employed in a general survey of the subternary. In addition to PbTe, PbSe, and pure lead, ten ternary compositions were selected at intervals of ten at. pct. Sixty-gram alloys were prepared by sealing an appropriate mixture of the elements in quartz crucibles with a central thermocouple-well under a vacuum of less than 10 p. All three elements were 99.999+ pct pure with only trace impurities detected by spectroscopic examination. After the alloys had been heated at least 20°C above the liquidus in a Kanthal furnace, cooling and heating curves were obtained. Two or more cooling curves were employed both for the determination of each alloy liquidus and for the solidus of those al- loys containing greater than 50 at. pct Pb. Cooling rates were 0.5° to 3.5°C per min during the 5 min immediately prior to the start or end of solidification. For those ternary alloys containing 50 at. pct Pb, the solidus was determined from heating curves. Temperatures were measured with a Pt/Pt-10 pct Rh thermocouple which was calibrated in the same apparatus against the freezing points of zinc (A. D. MacKay—99.999 pct), antimony (A. D. MacKay-99.999 pct), and copper (ASARCO—99.999+ pct) both before and after the experimentation. The only interruptions in the continuous strip-chart recording of the temperature occurred for brief periods at significant points of interest when direct readings were made with a Rubicon Type B potentiometer sensitive to * 1 µv. The average liquidus and solidus temperatures obtained in this investigation are listed in Table I together with probable errors based on the accuracy of the thermocouple calibration and the reproduci-bility of the data. No chemical analyses of the alloys were performed because the synthetic or as-charged compositions were felt to be reliable. The melting points obtained for PbTe (924.7° ± 0.7°C) and PbSe (1084.0° ± 0.6°C) compare favorably with recent findings of other investigators.&apos;-&apos; As indicated in Table I, both the liquidus and solidus temperatures increase as selenium replaces tellurium for a constant atomic percent lead. As a result, the ternary liquidus and solidus temperatures for a given percent lead lie between the corresponding binary temperatures published in the literature. It should be noted that those ternary alloys containing 50 at. pct Pb exhibit a relatively narrow liquid-plus-solid region (<80°C) when compared with alloys containing greater than 50 at. pct Pb. Alloys with 60, 70, or 80 at. pct Pb not only have a much wider liquid-plus-solid region (>500°C) but also exhibit considerably lower and approximately constant solidus temperatures which thus form a virtually isothermal solidus "plane" slightly below the melting point of pure lead (327.3°C). For these alloys

Jan 1, 1965

By Albert G. Anderson

I&apos;m sure many of you have wondered what the term "Cross Flow Filtration" means. I shall explain it so that you more clearly understand it, and I hope you will be able to make it a useful tool for you in the future. In the simplest form of cross flow filtration contaminated fluid. or feed stream, is fed at a high velocity down the center of a porous tube, clear fluid weeps through the porous wall of the tube while the suspended solids concentrate in and are swept away by the feed stream. (See figure A). The high velocity flow of the feed stream tends to keep the inside of the tube free from a build up of contamination. By utilizing this cross flow approach to filtration, we have been able to concentrate such things as paint pigments, titanium dioxide, iron oxide, precipitated metals such as uranium yellowcake or silver, cadmium, some metal hydroxides, as well as gelatinous particles such as yeast cells. You will recognize that cross flow filtration lends itself to a recirculation loop system due to the necessity of maintaining a high velocity on the inside of the tube. This means that we generally have a storage tank from which we are recirculating, and the solids will concentrate in the storage tank. (See figure B). The storage tank design becomes an integral part of the total system and is very important to the total function of the system. The cross flow filter system requires only low pressure, and will work with centrifugal pumps, diaphram pumps or in some cases no pump at all. As an example, if we wanted to filter a 10 GPM (37.8 liters/min.) stream that contained 1% of submicronic particulate, we would flow the 10 GPM (37.8 liters/min.) into the storage tank and recirculate through the filter at a flow of anywhere from 30 GPM to 100 GPM (113.6 to 378.5 liters/min.) depending upon the filterability of the particles from the liquid. As the contaminated fluid recirculates, the solids will build up in the storage tank and loop until they reach a level, which is dependent upon the design of the tank. Then the settling rate in the tank becomes equal to the amount of the solids being fed to the tank, and as a result you reach an equilibrium in the amount of solids being presented to the inside of the porous tubing. We have in effect created a dynamic settling tank or thickener. A flow of 10 GPM (37.8 liters/min.) of clear filtrate comes off for either recycle or disposal with periodic removal of thick mud from the bottom of the tank. The thick mud can be either processed through a centrifuge, a rotary drum filter. a flat bed pressure type filter, or other means. Often the small volume of filtrate from the mud de-watering stream can be recycled back into the recirculation tank so that any fines in the feedback are again concentrated, thus closing the loop on the continuous process 100% removal system. Cross flow filtration will not do efficient molecular separation. We work primarily with discrete solids but we do separate them totally from the liquid carrier fluid. I know some of you will raise your eyebrows when I say that we remove 100% of the particulate, but in measurement tests, we have measured the filtrate cleanliness with instruments that record as low as .05 ppm and the instruments read zero, so that it is less than .05 ppm.

Jan 1, 1980

By C. S. Land

Relative permeability functions are developed for both two- and three-phase systems with the saturation changes in the imbibition direction. An empirical relation between residual nonwetting-phase saturation after water imbibition and initial nonwetting-phase saturations is found from published data. From this empirical relation, expressions are obtained for trapped and mobile nonwetting-phase saturations which are used in connection with established theory relating relative permeability to pore-size distribution. The resulting equations yield relative permeability as a function of saturation having characteristics believed to be representative of real systems. The relative permeability of water - wet rocks for both two- and three - phase systems, with the saturation change in the imbibition direction, may be obtained by this method after properly selecting two rock properties: the residual nonwetting-phase saturation after the complete imbibition cycle, and the capillary pressure curve. INTRODUCTION Relative permeability is a function of saturation history as well as of saturation. This fact was first pointed out for two-phase flow by Geffen et al. 1 and by Osaba et a1.2 Hysteresis in the relative permeability-saturation relation also has been reported for three-phase systems.3 Since saturations may change simultaneously in two directions in a three-phase system, four possible relationships arise between relative permeability and saturation for a water-wet system. The four saturation histories of this system were given by snel14 as 11, ID, DI and DD. I and D refer to the direction of saturation change (imbibition and drainage), with the first letter of the symbol indicating the direction of change of the water phase. As used in this paper, the second letter of the symbol refers to the direction of saturation change of the gas phase, i.e., D and I indicate an increase and decrease, respectively, in gas saturation. Only a few three-phase relative permeability curves have been published. Leverett and Lewis5 published three-phase curves for unconsolidated sand, and Snell4 reported results of several English authors for both drairrage and imbibition three-phase relative permeability of unconsolidated sands. Three-phase relative permeability curves for a consolidated sand were published by Caudle et a1.3 for increasing water and gas saturations (ID). Corey et a1.6 reported drainage (DD) three-phase relative permeability for consolidated sands. Recently, Donaldson and Dean7 and Sarem8 calculated three-phase relative permeability curves from displacement data on consolidated sands, also for saturation changes in the drainage direction. The only published three-phase relative permeability curves for consolidated sands with saturation changes in the imbibition direction (11) are those of Naar and Wygal.9 These curves are based on a theoretical study of the model of Wyllie and Gardner 10 as modified by Naar and Henderson.11 Interest in three-phase relative permeability has increased recently due to the introduction of new recovery methods and refinements in calculation procedures brought about by the use of large-scale digital computers. The scarcity of empirical relations for three-phase flow, and the experimental difficulty encountered in obtaining such data, have made the theoretical approach to this problem attractive. RELATIVE PERMEABILITY AS A FUNCTION OF PORE-SIZE DISTRIBUTION Purcell used pore sizes obtained from mercury-injection capillary pressure data to calculate the permeability of porous solids.12 Burdine extended the theory by developing a relative permeability-pore size distribution relation containing the correct tortuosity term.13 The works of Purcell and Burdine were combined by Corey14 into a form that has

Jan 1, 1969

By J. C. Swartz

New measurements of graphite and cementite solubilities in ferrite have been obtained to resolve disagreements among previous data. To measure graphite solubility, specimens of iron were equilibrated with rarious carburizing atmospheres at temperatures be-twee~z 480" and 1435°C. The carbon activity of each atmosphere relative to graphite was taken from published data. After quenching from each equilibration treatment the concentration of carbon (typically 10 pprn by wt) was determined from the height of the Snoek peak in the internal friction spectrum. A few inclividual analyses by combustion support the reszdts of Smith on the solubility of graphite near the eutectoid temperature. Combination of these combustion and internal friction analyses determines In a, (at 0.01 wt pct) = 24,000/RT - 12.72, where a, is the activtity relative to graphite in ferrite at fixed cornposition. The same specimens were used to determine the solubility of self-stressed cementite precipitates. The results indicate solubilities of 6.3, 23.2, and 68 pprn carbon by wt at respectively 400°, 500". and 600°C. The once large discrepancy between the cementite solubility determined by internal frictiori and that calculated from flee energy data has been almost completely resolved. A recent study1 (hereafter designated I) on the solubility of cementite in a iron did not fully resolve the disagreement among previous data. The remaining disagreement is illustrated by the phase diagram in Fig. L The data of I are represented by the line labeled "stress-free" cementite. They determine a heat of solution of cementite of (14.8 * 0.2) kcal per mole which contrasts with the (20 * 1) kcal per mole derived from Smith&apos;s graphite solubility2 and the formation energy of cementite.3 Toward resolving this disagreement we have obtained new data on the solubilities of graphite and cementite. METHODS As in I the Snoek peak in the internal friction spectrum was used to measure the concentration of carbon dissolved in ferrite. This method relies on the relation C=kQ-1 [1] in which C is the concentration of dissolved carbon, Q-I is the Snoek peak height, and k is a proportionality factor which is a function of the peak temperature, polycrystalline texture, and other parameters characteristic of the ferrite lattice. To determine true heats of solution via Eq. [1], it is necessary to establish that k is independent of the temperature from which the specimen is quenched. Some evidence on this point is contained in a separate report.* Additional evidence is shown by the data points in Fig. 1. These represent the Snoek peaks of two specimens carburized to about 10 ppm C, copper plated, and quenched first from 888º then from 588°C. The copper plate was removed before each Snoek peak measurement. These data show that there is no detectable dependence of k on quenching temperature. The solubility of graphite was determined by repeatedly carburizing various specimens of iron wire with gas mixtures of known carbon activity relative to graphite. To cover a wide temperature range within the a phase boundaries it is desirable to use as low a carbon concentration as is compatible with reasonable precision. With the present internal friction apparatus the best compromise of these desires is a concentration near 10 wt ppm. The dashed line in Fig. 1 was used to approximate the activity of carbon* for

Jan 1, 1970

By R. W. Fountain, W. D. Forgeng

A new constitutional diagram is presented for the cobalt-rich end of the cobalt-titaniurn system. The modifications result from the presence of a new, intermediate, fcc phase, ?, the existence and homogeneity limits of which were established by metallo-graphic and X-ray studies of alloys containing from about 3 to 30 pct Ti. The precipitation of the ? phase from supersaturated solid solution was studied by hardness and electrical resistivity measzsrements, and two distinct stages in the process were observed. COBALT forms the base for a number of precipitation -hardenable alloy systems which may be divided into two distinct categories of practical interest, a) those hardened by intermetallic compounds and b) those hardened by carbide formation. The precipitation of intermetallic compounds from solid solution in cobalt-rich alloys has, however, received very little attention. Although the phase diagrams for some binary systems capable of precipitation have been determined,&apos; there is an almost complete lack of data on the property changes associated with the precipitation, and even less information on the kinetics of the reactions or morphology of the products. More information is available on the precipitation of carbides because of the practical significance in superalloys. A survey of cobalt binary phase diagrams suggested that the cobalt-titanium alloys might provide interesting and useful precipitation-hardenable alloys. The equilibrium diagram as proposed by Wallbaum2 is shown in Fig. 1. Köster and wagner3 have indicated that the maximum solubility of titanium in cobalt is about 10 pct at the eutectic temperature (1135oC), and this decreases to about 7.2 pct at room temperature. The temperature of the allotropic transformation in cobalt is lowered by the addition of titanium, about 5 pct being sufficient to retain the high-temperature fcc modification to room temperature. Wallbaum and Witte 4,5 have indicated that the precipitating phase in alloys containing up to about 29 pct Ti is Co2Ti, a hexagonal Laves phase of the MgNi, type. With slightly higher titanium contents, they also report a cubic modification of Co2Ti of the MgCu2-type Laves phase. Duwez and Taylor6 confirmed the existence of the hexagonal (MgNi2) modification but not the cubic (MgCuz) modification of Co2Ti and suggested that existence of the cubic form may have resulted from impurities in Wallbaum&apos;s alloys. In their work on Laves-type phases, Elliott and Rostoker 7 reported the cubic modification of Co,Ti, but did not confirm the existence of the hexagonal modification. However, Dwight8 in a discussion of the work of Eliott and Rostoker again showed the existence of both modifications of Co2Ti, a result which was confirmed at that time by Elliott and Rostoker. As a result of a study on the iron-cobalt-titanium system, Köster and Gellers suggested the existence of a Co3Ti phase isomorphous with Fe3Ti. Witte and Wallbaum,&apos; however, established the fact that no compound, Fe3Ti, exists in the iron-titanium system, and in a later publication, wallbaum2 stated that Kitster and Geller&apos;s reasoning was not valid, and no compound, Co3Ti, exists. This conclusion was later acknowledged by Köster.10 Preliminary experiments by the present authors to determine the precipitation-hardening characteristics of the cobalt-rich, cobalt-titanium alloys re-

Jan 1, 1960

By A. W. Ruff

In the last ten years large gains have been made in the field of blasting. These gains have been both in the theoretical and in the practical application of explosives. One of the most publicized changes in the practical application of commercial explosives is the now-prevalent use of ammonium nitrate and fuel oil. The new water-compatible mixture of ammonium nitrate and TNT should also have a large field of application. Most of the theoretical gains have been concerned with the actual rock breakage obtained with explosives. In the last ten years, large gains have been made in the field of blasting. These gains have been both in the theoretical and in the practical application of explosives. One of the most publicized changes in the practical application of commercial explosives is the now prevalent use of ammonium nitrate and fuel oil. The new water-compatible mixture of ammonium nitrate and TNT should also have a large field of application. Most of the theoretical gains have been concerned with the actual rock breakage obtained with explosives. EXPANDING-GAS-BUBBLE THEORY There has been an outstanding change in the conception of how an explosive produces rock failure. The early theory which explained rock breakage was correct for some of the first force-producing agents such as quicklime. When a drillhole was filled with quicklime and water added, the material expanded and steam was generated. The expanding material and the steam exerted a slowly increasing pressure on the sides of the drillhole. The force broke the rock by splitting. Because a great quantity of gas was produced by the deflagration of gunpowder, the same process of the gas pressure splitting the rock was assumed to take place. Even after 80 years of use, the rock-breaking action of all high explosives was thought to be caused entirely by the expanding gas bubble which tore and split the rock. Fig. 1 shows a vertical section through an exploding spherical charge of high explosive. It illustrates the early theory which attributed the entire breakage to the expanding gas bubble. The basic idea of this theory is that the breakage started at the explosive charge and progressed toward the free face. REFLECTED-WAVE THEORY me expanding-gas- bubble theory has gradually been replaced by the reflected-wave theory. The reflected-wave theory explains rock failure by explosives in the following manner: The enormously high gas pressure in the drillhole produces fracturing of the rock on the sides of the drillhole. The fracturing around a 1 1/2-in. diam drillhole is usually confined to an area with about a 6-in. diam. In the fractured area, the pressure of the gas is reduced until the force on the walls is no longer capable of fracturing the rock. Although the gas pressure does not continue to fracture the rock, the pressure imposes a high impact stress on the wall rock. In addition, a pressure or seismic pulse is produced in the rock. The pulse is probably entirely a compression pulse when originally produced in the rock. But, with distance, part of the energy of the compression pulse is transformed into a tension pulse which follows the compression pulse. The explosive charges and burdens used in conventional mine blasting are such that the tension pulse of the original wave is probably too small to be considered in a practical evaluation of the theory. The compression pulse is transmitted through the rock with no associated fracturing.

Jan 1, 1961