Search Documents

By Martin S. Maier, H. R. Nelson

THE diffusion of metals plays an important role in many metallurgical processes, 1,2 as, for example, the formation of alloys by the annealing of mixed powders, the homogenization of segregated alloys, and the carburization of steel. Other equally important phenomena, such as recrystallization and creep, are based on the movement of like atoms. Although much is known concerning the effects of diffusion in metals, very little is known concerning the details of the physical processes involved.1,3-5 The data that should give most information concerning the processes involved in diffusion are those on the diffusion of like atoms-self-diffusion. Very little is known concerning the rates of self-diffusion of metals because before the discovery of artificial radio-activity in 1934 there was no means of detecting the movement of the diffusing atoms through the solvent metal except in the elements possessing natural radioactive isotopes-lead, thallium and bismuth. Using the induced radioactivity of the element as an indicator, atoms can now be "tagged" so that the rate of diffusion of the radioactive isotopes of an element (tagged atoms) into the stable isotopes of the same element can be measured. As the radioactive and stable isotopes of an element differ only in nuclear structure, it is quite certain that all isotopes of an element diffuse at the same rate. Two techniques have been used to measure self-diffusion. In one, a thin layer of radioactive metal is deposited on the face of the solvent metal, and the decrease in the surface activity as the radioactive atoms diffuse into the solvent metal is measured. This method has been used to determine the coefficients of self-diffusion of lead6-10 and bismuth10 using natural radioactive isotopes, and of gold11 and copper12 using artificial radioactive isotopes. In the other technique a thin layer of the radioactive metal is deposited on one face of the solvent metal, and after diffusion has occurred the specimen is cut in thin sections to obtain the concentration-penetration curve. This method has been used to determine the coefficients of self-diffusion of gold,13 zinc, 14-16 silver,17 and copper.18 The second method was used for the work described in this paper. MATERIALS The coefficients of self-diffusion were determined in two grades of polycrystalline copper and in single crystals of copper. Commercial OFHC copper was used in the form of disks I in. in diameter by 1/8 in. thick. Samples of high-purity copper described by Phillips and Skinner19 were supplied through the courtesy of the American Smelting and Refining Co. The high-purity copper rod was used without remelting by cross-rolling to a thickness of 0.075 in., from which disks 7/8 in. in diameter were cut. A good grade of bus-bar copper was used to make the single crystals.

Jan 1, 1942

By E. A. Kilinski

THE mineral production of Dutch Guiana has been of little consequence. This is due to several rea-sons. Inaccessibility is probably the greatest factor, for the Guianas are off the main commercial route. More or less closely related causes are the tropi-cal climate and the temperament of the inhabitants. Small quantities of gold have come from Dutch Guiana for many years, and recently bauxite has entered the list of exports. It is likely that other minerals may be found in commercial quantities. The Colony is on the northeast coast of South Amer-ica. It is known by two names, that of Dutch Guiana, and that of Surinam. - British Guiana (Demerara) adjoins it on the west, French Guiana (Cayenne) on the east, and Brazil on the south. It covers an area of approximately 58,000 sq. miles. The population, in the. neighborhood of a hundred thousand,, is small, but the variety of races and the sub-sequent mixtures of these is almost incredible. This unusual and colorful population readily falls into three groups, (1) those of the bush, (2) those of the planta-tion, and (3) those of the towns. The bush people con-sist of two groups, the Indians and the bush-negro. The Indians, of whom there are but a little over two thousand, keep entirely to themselves, lead a. very primi-tive existence, and rarely come into the towns. The bush-negro gives the touch of human life to the thick jungles. A descendant of escaped slaves of by-gone days, he is at present a proud, stalwart, happy, clean, moral, carefree and ease-loving individual-a distinct personality. Work is not his pal, although he does cheerfully that which is necessary for his existence. He will transport passengers and goods in his dugout and guide the stranger through the jungles, but that is all. The plantations employ some negroes, but more Javanese, imported as contract labor from the East Indies. Paramaribo, the only large town in Dutch Guiana, is the center of all government, commercial, and social activities. It has a motley population of 40,000, in-cluding negroes native to Dutch Guiana and many from all parts of the West Indies. In addition, there are Javanese, Chinese, and some Hindus. The two latter groups control most of the retail trade. The Dutch and a few scattered Germans are the only Europeans. All of the higher government offices are held by the Dutch, the others are filled from the ranks of the natives. Marriages and intermarriages among these various groups have produced types that defy almost any effort to decipher an individual's forbears. Dutch is the official language, but among the people a jargon known as Tahki-Tahki holds sway. English is taught in the schools, consequently many speak it, and many more can make themselves understood in that language.

Jan 2, 1928

By G. C. Gardner

A new correlation of blast penetration in model and operating blast furnaces is presented which overcomes objections to prevzous correlations. It is shown that the tuyere diameter is only of small importance in determining the magnitude of the penetration. The data from operating furnaces correlate satisfactorily with the data from models whose cross section is a segment of a circle. A bubble-shaped gas space called a "raceway" normally exists in front of the nose of a blastfurnace tuyere. The periphery or "wall" of the raceway is well defined and is composed of densely packed charge particles. The blast. issuing as a jet from the tuyere, entrains gas from the raceway, in the usual manner of a jet, as it flows to the opposing wall. There it divides and flows back around the walls to be entrained once more. The blast must, of course, eventually leave the raceway and pass up through the charge column, but before it does it may, on the average, make several circuits of the kind described and in doing so it carries with it charge particles, which, because of their fast and circulating motion, suggested the name of "raceway". A fuller description of the raceway may be found elsewhere."' Here it is only necessary to mention, in addition, the measurement which characterizes its size. If an iron bar is pushed through the opened peepsight of a tuyere, .as far as it will go with ease, it will be stopped by the relatively impenetrable opposing raceway wall arid the distance that the bar has been pushed from the tuyere nose is called the "penetration". This measurement may be of practical importance and has; been the subject, with considerable success, of several attempts at correlation with other variables2-5 as stated in the next section. There is still, however, a serious objection to the most recent correlation,3 which is overcome by the correlation proposed in this paper. It will be shown that the objection is not trivial, in that the new correlation will attribute considerably different importance to the variables affecting the magnitude of the penetration. The data to be treated were made available by Wagstaff and are the same as those used in his correlation. They include measurements taken on small models, in which a cold air blast was used and no combustion occurred, and they also include data from operating blast furnaces at Wheeling and Fairless. The information from the blast furnaces will be shown to be in substantial agreement with those models whose plan views are segments of a circle and which are thus like segments of a blast furnace. First, Wagstaff's most recent correlation3 will be discussed and the objection will be given. Then the new correlation will be developed and its implications considered. CORRELATION OF J. B. WAGSTAFF Wagstaff's correlation is where and D — the penetration D,-the internal tuyere diameter V —the blast velocity in the tuyere

Jan 1, 1961

By R. K. Matheson

The Nivloc mine is 9 miles west of Silver Peak, Esmeralda County, Nevada. It has been operated by Desert Silver, Inc., since the summer of 1937. The cyanide mill treats 190 tons of silver-gold ore per day. In March of this year the main operations were about 2000 ft. west of the main shaft, which served as the downcast air passage for the mine. The mechanical blowers in use were sufficient to ventilate the west end at that time, but it was foreseen that as the mining area was extended ventilation would be an increasingly difficult problem, especially during the summer months. It was decided, therefore, to continue a raise from a point 100 ft. above the highest mining level (440-ft. level) to the surface; a distance of 443 ft. at approximately 65°. From this point 40 ft. of drifting would be necessary to connect with a shallow shaft on the surface. (Fig. I.) All the raise was serviced from the 440-ft. level; costs and progress obtained in the initial 100 ft. are included in the following pages. Choice of Raise A raise in this locality would have four objects: (I) to provide an ample supply of fresh air to the working places; (2) to explore the vein from the highest mining level to the surface; (3) to provide waste fill for stopes in the area, and possibly provide a transfer for waste from the surface; (4) to provide an emergency exit in the west end of the mine. Before the mining of an ore shoot in the eastern part of the mine, a similar raise had been driven for purposes of ventilation and exploration. This raise and subsequent mining showed the vein to be noncommercial for a distance of about 250 ft. below the surface. However, in the area under operation, old surface workings showed occasional values. Thus from a development standpoint the raise was an excellent prospect. The raise chosen was a three-compartment, stulled raise having two chutes and a manway. The manway inside timber measured 4 ft. 8 in. by 4 ft. 6 in.; it carried air and water pipes, a ladderway and a slide for a timber skip. The two chutes would be used for drawing off the muck and the cleaning of the bulkhead, as well as for ore and waste passes, which would be especially useful for lateral exploration from the raise. The fairly large cross section of the raise (6 by 15 ft.) would provide a large quantity of waste for backfilling, and more could be obtained from a glory hole on the surface. Time Schedule As speed in completing the raise was the main requisite, three shifts were considered, but two shifts were decided upon when it was found that three would necessitate considerably larger fans than those available for clearing the raise. The main diffi-

Jan 1, 1943

By F. W. Bloecher

S. Mortsell and J. Svenssofi (Royal Institute of Technology, Stockholm, Sweden)—Bloecher states that the apparatus described by him should be a suitable instrument in the mineral dressing laboratories for determining the specific surface of such granular products that are generally investigated. We want to express a somewhat different opinion, as we doubt that the gas adsorption method of surface measurement really is the most suitable method in mineral dressing laboratories. It is well known that the surface area, as measured by this method, does not coincide with the surface area calculated from the size, shape, and number of particles, i.e. the superficial surface of the particles. It can be seen easily from Mr. Bloecher's own determination on ilmenite-leucoxene concentrate that the surface area measured according to this method may be several hundred times larger than the superficial surface. It is supposed that the surface of cracks, fissures, and pores within the particles also is included in the measured surface. However, this is an assumption which scarcely can be proved and certainly has not been proved. We believe that the mineral engineer could better use a method giving the superficial surface of the particles, as in most cases this surface only is of importance to him. Such is certainly the case when calculating the resistance of the fluid flow through granu- lar materials and the capillary forces of such materials, problems which are of great importance in connection with dewatering, filtration, and pelletizing'" etc. The same may also be the case when studying flotation problems, partly since the quantities of the introduced reagents usually are so small that they can only cover the most accessible surface of the particles, and partly since the molecules of at least the collecting reagents are so large that they cannot enter the smallest cracks and pores. As to the surface, which is of consequence in the application of Rittinger's law of crushing, one is inclined to believe, from theoretical reasons and from the work by Gross and Zimmerley, that the total surface here is the most important one. However, it seems as if in that case the superficial surface was as important as the total surface. Other reasons making the gas adsorption method less suitable for use in the mineral dressing laboratories are that the granular materials there very often are too coarse to be measured by this method, that the experimental technique is rather difficult, and that it takes a long time to measure the specific surface of a sample. Besides, the validity of the method must still be considered questionable. It is thus well known that with different gases one usually gets different values of the specific surface of the same sample, which values, even with such an ideal material as metal foils, can differ as much as about 50 pct.'" Still more serious

Jan 1, 1952

By Louis A. Panek

The purpose of this paper is to present a method of designing safe and economical arch structures that are to be constructed of concrete or directly of original earth materials. The experimental data used to illustrate the design method are taken from Bucky and Fen tress,1,2 and the application of the ideas presented by them is extended by obtaining an additional set of curves from which arch structures may be designed. This design method is based upon the following principles of similitude: A scalar model made of the same material as that of the full sized structure will behave similarly to its prototype if the density of the model is increased in the same proportion as its linear scale is decreased. Let R represent the scale or model ratio, namely, the ratio of a prototype linear dimension to the corresponding model linear dimension. The effect of increased density may be obtained by rotating the model in a centrifuge at a speed such that the centrifugal held exerts on the model an accelerative force R times the accelerative force of gravity. Under these conditions; (I) model unit stresses are equal to corresponding prototype unit stresses, (2) the ratio of the prototype deformation at a given point to the model deformation at a similar point is equal to the scale ratio, R, and (3) the model and the prototype will behave similarly within and beyond the elastic limits of strain. According to the above relations, such models can be used to solve the following problems: I. The dimensions of the most economical safe arch structure for a given load. 2. The miaximum safe roof span underground for given geologic conditions. The procedure is described briefly as follows. Tests are made by rotating in a centrifuge a number of models that repiesent a variety of prototype loads and dimensions. From the experimental data, curves are plotted to show the relations between these loads and dimensions. The curves are used to solve directly the problems stated above. Structural design is based upon the working (safe) stress of the material of construction, which is established by applying a factor of safety to the experimentally determined ultimate strength of the material. In conformity with this principle, each model is gradually loaded until failure occurs. A safety factor is then applied to the results. If a safety factor of 4 is used, a given model at failure represents a safe prototype, the linear dimensions of which are equal to one-fourth the model ratio at failure times the corresponding model dimensions. The safe model ratio is computed from the following formula: Safe R =(1/4)4p2n2r/g=p2n2r/g [I] in which: n = centrifuge revolutions per second at model failure; r = radius of rotation of the model in feet; g = 32.2 ft/sec2.

Jan 1, 1949

By B. D. Cullity

A distribution of residual stress after plastic elongation is proposed, in which the bulk of the material is strained in compression and a very small portion in tension, This distribution is shown to account both for the X-ray observations and for the effect of plastic straining on magnetic losses in silicon steel. The regions left in compression are tentatively identified with the interiors of the sub-grains revealed by the electron microscope. THIS paper is concerned with the general problem of the nature of the residual stress which is produced in metals by plastic elongation. A particular stress distribution will be suggested which accounts for the main X-ray diffraction observations which have been made. It will also be shown that certain magnetic measurements made on silicon steel can be understood in terms of this stress distribution. X-RAY OBSERVATIONS Uniform straining of a crystal lattice (macro-strain) causes a diffracted X-ray beam to shift its position, and nonuniform straining (microstrain) causes line broadening. If a bar of metal is plastically elongated in the longitudinal x direction and then unloaded, diffracted X-ray lines are observed to be both shifted and broadened.'-= Although the observed line shift indicates the presence of macro-strain, experiments have shown that no macrostrain, in the usual engineering sense of the term, exists.5"7 For example, removal of a layer from one surface of a stretched and unloaded bar does not produce curvature in the remainder of the bar 5'7 Fig. 1 shows the variation of lattice strain in the x direction, as determined from X-ray line shifts, with the applied stress in the same direction. The curve ABC is the normal stress-strain curve of the material, AB being the elastic portion and BC the plastic. In the elastic region AB, the strain measured by X-rays, called the lattice strain, and the strain measured mechanically are both proportional to the applied stress, and both return to zero when the applied stress is removed. Above the elastic limit the lattice strain would be expected to continue to increase along the line BD. Actually, it increases less rapidly with respect to stress, or even decreases, as shown by BE. If the applied stress is then decreased, the lattice strain decreases along EF, which is parallel to AB. As a result, there is a compressive residual lattice strain FA after unloading. STRESS DISTRIBUTION. Since a stretched and unloaded bar does not contain macrostress, it follows that the observed line shift must be due to a special distribution of mi-crostress. Such a distribution is shown in Fig. 2, which shows the variation of the stress u, across a stretched and unloaded bar. The material is assumed to consist of two kinds of regions: regions C, stressed more or less uniformly in compression, comprising the bulk of the volume of the material and responsible for the observed diffrac-

Jan 1, 1963

By Louis A. Panek

THE purpose of this paper is to present a method of designing safe and economical arch structures that are to be constructed of concrete or directly of original earth materials. The experimental data used to illustrate the design method are taken from Bucky and Fentress,1,2 and the application of the ideas presented by them is extended by obtaining an additional set of curves from which arch structures may be designed. This design method is based upon the following principles of similitude: A scalar model made of the same material as that of the full sized structure will behave similarly to its prototype if the density of the model is increased in the same proportion as its linear scale is decreased. Let R represent the scale or model ratio, namely, the ratio of a prototype linear dimension to the corresponding model linear dimension. The effect of increased density may be obtained by rotating the model in a centrifuge at a speed such that the centrifugal field exerts on the model an accelerative force R times the accelerative force of gravity. Under these conditions; (I) model unit stresses are equal to corresponding prototype unit stresses, (2) the ratio of the prototype deformation at a given point to the model deformation at a similar point is equal to the scale ratio, R, and (3) the model and the prototype will behave similarly within and beyond the elastic limits of strain. According to the above relations, such models can be used to solve the following problems: I. The dimensions of the most economical safe arch structure for a given load. 2. The maximum safe roof span underground for given geologic conditions. The procedure is described briefly as follows. Tests are made by rotating in a centrifuge a number of models that repIesent a variety of prototype loads and dimensions. From the experimental data, curves are plotted to show the relations between these loads and dimensions. The curves are used to solve directly the problems stated above. Structural design is based upon the working (safe) stress of the material of construction, which is established by applying a factor of safety to the experimentally determined ultimate strength of the material. In conformity with this principle, each model is gradually loaded until failure occurs. A safety factor is then applied to the results. If a safety factor of 4 is used, a given model at failure represents a safe prototype, the linear dimensions of which are equal to one-fourth the model ratio at failure times the corresponding model dimensions. The safe model ratio is computed from the following formula: [Safe R = (rl 47r2n2r _ 7r2n2r~r~ \4/ gg in which: n = centrifuge revolutions per second at model failure; r = radius of rotation of the model in feet; g = 32.2 ft/sect.]

Jan 1, 1947

By H. R. Nelson, Martin S. Maier

The diffusion of metals plays an important role in many metallurgical processes,lJ as, for example, the formation of alloys by the annealing of mixed powders, the homogenization of segregated alloys, and the carburization of steel. Other equally important phenomena, such as recrystalli-zation and creep, are based on the movement of like atoms. Although much is known concerning the effects of diffusion in metals, very little is known concerning the details of the physical processes involved. 1,3-5 The data that should give most information concerning the processes involved in diffusion are those on the diffusion of like atoms—self-diffusion. Very little is known concerning the rates of self-diffusion of metals because before the discovery of artificial radioactivity in 1934 there was no means of detecting the movement of the diffusing atoms through the solvent metal except in the elements possessing natural radioactive isotopes—lead, thallium and bismuth. Using the induced radioactivity of the element as an indicator, atoms can now be "tagged" so that the rate of diffusion of the radioactive isotopes of an element (tagged atoms) into the stable isotopes of the same element can be measured. As the radioactive and stable isotopes of an element differ only in nuclear structure, it is quite certain that all isotopes of an element diffuse at the same rate. Two techniques have been used to measure self-diffusion. In one, a thin layer of radioactive metal is deposited on the face of the solvent metal, and the decrease in the surface activity as the radioactive atoms diffuse into the solvent metal is measured. This method has been used to determine the coefficients of self-diffusion of lead6-l0 and bismuth10 using natural radioactive isotopes, and of gold1' and copper12 using artificial radioactive isotopes. In the other technique a thin layer of the radioactive metal is deposited oil one face of the solvent metal, and after diffusion has occurred the specimen is cut in thin sections to obtain the concentration-penetration curve. This method has been used to determine the coefficients of self-diffusion of gold,13 zinc,14-l6 silver," and copper.17 The second method was used for the work described in this paper. Materials The coefficients of self-diffusion were determined in two grades of polycrystalline copper and in single crystals of copper. Commercial OFHC copper was used in the form of disks I in. in diameter by % in. thick. Samples of high-purity copper described by Phillips and Skinnerl9 were supplied through the courtesy of the American Smelting and Refining Co. The high-purity copper rod was used without remelting by cross-rolling to a thickness of 0.075 in., from which disks 7/8 in. in diameter were cut. A good grade of bus-bar copper was used to make the single crystals.

Jan 1, 1942

By A. J. Miles, A. D. Topping

in this paper, the theory of elasticity has been applied to the rock about a deep well, It is assumed that the rock has a modulus of . elasticity and a Poisson's ratio and that the theory of elasticity applies. It is necessary to know or assume the state of stress existing in the rock before it is penetrated by the well drill. The application of this theory indicates that stress concentration of shear, tensions, and compressions about the bore hole are of a high order. This is particularly true when a horizontal compressive stress exists in one direction only in the formation before drilling. If such an initial state of stress exists before drilling, then the rock will have stress concentrations of both tension and compression at the same elevation and of such magnitude that failure of the rock is likely. Accompanying these is a shearing stress of large proportion which is likely to produce spalling of the well walls. Internal pressure applied to the well bore will relieve the extreme compression but not the tension and has little effect upon the shear. Plastic deformation of the rock through a geological time tends to mitigate the stress concentrations. IntroductioN It has long been known that the stresses about holes and re-entrant corners of elastic solids under the influence of loads are different and generally more intense than those imposed upon the body else- where. Stress concentration at a re-entrant corner can be reduced by increasing the radius of the fillet at the corner. Stress concentration at the end of a crack in a plate undergoing either tension or compression Can be relieved appreciably by drilling a hole at the very end of the crack. Also, stresses can be increased in an elastic body in tension or compression by making a hole in it. The stress concentration is greatest at the edge of the hole. Considerable knowledge of the stresses which exist about the bore hole of a deep well may be had by applying our knowledge of engineering mechanics; or more particularly the theory of elasticity. It is, of course, necessary to make the assumption that the rock is elastic and behaves as an elastic solid, that is, that it obeys Hooke's law, has a modulus of elasticity, and a Poisson's ratio. It is also necessary to know, Or assume, the state of stress which exists in the rock prior to the penetration of the drill. With this knowledge it is possible to compute the stresses about the bore hole. The problem can be simplified by considering several simple cases separately and then by applying the principle of superposition to solve the more complex cases which are made up of the simpler ones. Case i—Rock Strata in Compression Penetrated by a Hole This case was first worked out by G. Kirsch* for structural material, but applies equally well to a horizontal rock strata in a state of in in one direction

Jan 1, 1949

By R. W. Gurry

WHEN a steel in the austenitic state, with all its carbon in solution, is maintained, at constant temperature, in contact with a gas that removes the carbon from the surface, yet without otherwise altering the composition of the steel, the change in weight of the specimen, after any period of exposure, is a direct measure of the amount of carbon it has then lost. The purpose of the present paper is to show how, by application of established mathematical reasoning based on the fundamental diffusion law and on certain postulates, this loss of weight yields a complete picture of the distribution of carbon throughout a specimen of regular shape after any period of exposure, with all the accuracy required in any practical occurrence of decarburization. Precisely the same is true for the gain in weight during the reverse process of carburization, or for the gradual gain or loss of any element, so long as the steel comprises only a single phase at temperature, provided always that the experimental conditions conform to the postulates. The change in carbon content from surface to center of the specimen is always continuous, when everything is in solution, even though under the microscope the decarburized zone (or carburized rim) may appear to be sharply defined. The experimental procedure is described and some typical results are presented, which are then interpreted by application of the diffusion law. EXPERIMENTAL PROCEDURE Initially an endeavor was made to follow the gain of weight during the process of carburization in a hydrogen-toluene atmosphere, but it was soon found that the specimen became coated with graphite, which could not be removed without some loss of metal; therefore it was decided to continue carburization until the steel was saturated throughout, to remove all graphite from the surface of the specimen, and then to follow the loss of weight during decarburization in a moist hydrogen atmosphere at the same temperature as that used for carburizing. The carburizing procedure has been described in an earlier paper;1 it suffices to state here that the carburization lasted about 140 hr. at 760°C. (1760°F.) and 44 hr. at 1110°C. (2030°F.), these periods being long enough-on the basis of the calculations presented later-to ensure that the specimen was about 99.9 per cent saturated at each temperature. The specimen was a cylinder about 3.8 cm. long and 0.475 cm. in diameter; the materials were carbonyl iron, a commercial low-metalloid iron, and a steel with 3.6 per cent nickel and the usual proportion of other elements. The fully carburized specimen, after cooling at the top of the furnace tube, was placed in a lathe and sandpapered to remove all graphite; then the diameter was measured with a micrometer caliper, and the initial weighing (corresponding to complete saturation) was made. By means of a

Jan 1, 1942

By R. P. McLaughlin

Discussion of the paper of R. P. McLaughlin, presented at the San Francisco meeting, September, 1915, and printed in Bulletin 'No. 108, December, 1915, pp. 2313 to 2319. M. E. LOMBARDI, Berkeley, Cal.-Mr. Bell, I would ask you to state before the meeting what you said to me the other day, about the final recovery of oil from oil sands that have become inundated with water. A. F. L. BELL, San. Francisco, Cal: In stating my opinion, I give it with certain qualifications. In the Kern field, and possibly in some others, I think we will eventually get all the oil out, whether the water comes in or not, since it is peculiarly constructed, being in the form of a partial dome. In 1909, while I was field manager of the Associated Oil Co., I installed some large air compressors to handle the water. At that time the oil had migrated to the higher and more shallow portions of the field, because the water had inundated the deeper sections. The result was that the shallow districts within the center of the dome were in many, cases producing as much or more oil per well than the deeper wells. And in the inundated sections the water had slowly spread over large areas, so that nothing but water was pumped from them. Under these conditions, we decided to pump the wet. well with air, and installed four 2,000-cu. ft. compressors. The air was compressed and forced into the wet wells at, about 250 lb. pressure. Shortly the inundated district that had produced almost all water and only about 150 bbl. of oil a day increased its production to over 2,000 bbl. of oil per day. It was not-long before, we noticed that the production of the shallow territory was falling off to such an extent that in many cases the wells were abandoned; at the same time the area of the water-inundated district was greatly reduced, showing that when the water pressure was reduced the oil migrated back to the lower depth. The wells which at first had had the air jets put into them finally produced so much oil that we had to take the 'air jets out and return to the use of the beam pump, because it is not commercially feasible to pump, an oil of 14° gravity with air. To handle such oil profitably by air, it must carry with it 50 to 90 per cent. of water. The results have led me to believe that oil monoclines, where there is a gradual rise to a certain point, or in plunging anticlines, even though water gets into oil sands, eventually we get the larger percentage of the oil. However, the man with a small acreage in the deeper sections may find some morning that his oil has gone, another man in some higher portion of the district having gotten it.

Jan 12, 1915

By R. W. Gurry

WHEN a steel in the austenitic state, with all its carbon in solution, is maintained, at constant temperature, in contact with a gas that removes the carbon from the surface, vet without otherwise altering the composition of the steel, the change in weight of the specimen, after any period of exposure, is a direct measure of the amount of carbon it has then lost. The purpose of the present paper is to show how, by application of established mathematical reasoning based on the fundamental diffusion law and on certain postulates, this loss of weight yields a complete picture of the distribution of carbon throughout a specimen of regular shape after any period of exposure, with all the accuracy required in any practical occurrence of decarburization. Precisely the same is true for the gain in weight during the reverse process of carburization, or for the gradual gain or loss of any element, so long as the steel comprises only a single phase at temperature, provided always that the experimental conditions conform to the postulates. The change in carbon content from surface to center of the specimen is always continuous, when everything is in solution, even though under the microscope the decarburized zone (or carburized rim) may appear to be sharply defined. The experimental procedure is described and some typical results are presented, which are then interpreted by application of the diffusion law. EXPERIMENTAL PROCEDURE Initially an endeavor was made to follow the gain of weight during the process of carburization in a hydrogen-toluene atmosphere, but it was soon found that the specimen became coated with graphite, which could not be removed without some loss of metal; therefore it was decided to continue carburization until the steel was saturated throughout, to remove all graphite from the surface of the specimen, and then to follow the loss of weight during decarburization in a moist hydrogen atmosphere at the same temperature as that used for carburizing. The carburizing procedure has been described in an earlier paper;1 it suffices to state here that the carburization lasted about 140 hr. at 760°C. (1760°F.) and 44 hr. at 1110°C. (2030°F.), these periods being long enough-on the basis of the calculations presented later-to ensure that the specimen was about 9.99 per cent saturated at each temperature. The speci¬men was a cylinder about 3.8 cm. long and 0.475 cm. in diameter; the materials were carbonyl iron, a commercial low-metalloid iron, and a steel with 3.6 per cent nickel and the usual proportion of other elements. The fully [carburized] specimen, after cooling at the top of the furnace tube, was placed in a lathe and sandpapered to remove all graphite; then the diameter was measured with a micrometer caliper, and the initial weighing (corresponding to complete saturation) was made. By means of a

Jan 1, 1942

By Martin S. Maier, H. R. Nelson

The diffusion of metals plays an important role in many metallurgical processes,lJ as, for example, the formation of alloys by the annealing of mixed powders, the homogenization of segregated alloys, and the carburization of steel. Other equally important phenomena, such as recrystalli-zation and creep, are based on the movement of like atoms. Although much is known concerning the effects of diffusion in metals, very little is known concerning the details of the physical processes involved. 1,3-5 The data that should give most information concerning the processes involved in diffusion are those on the diffusion of like atoms—self-diffusion. Very little is known concerning the rates of self-diffusion of metals because before the discovery of artificial radioactivity in 1934 there was no means of detecting the movement of the diffusing atoms through the solvent metal except in the elements possessing natural radioactive isotopes—lead, thallium and bismuth. Using the induced radioactivity of the element as an indicator, atoms can now be "tagged" so that the rate of diffusion of the radioactive isotopes of an element (tagged atoms) into the stable isotopes of the same element can be measured. As the radioactive and stable isotopes of an element differ only in nuclear structure, it is quite certain that all isotopes of an element diffuse at the same rate. Two techniques have been used to measure self-diffusion. In one, a thin layer of radioactive metal is deposited on the face of the solvent metal, and the decrease in the surface activity as the radioactive atoms diffuse into the solvent metal is measured. This method has been used to determine the coefficients of self-diffusion of lead6-l0 and bismuth10 using natural radioactive isotopes, and of gold1' and copper12 using artificial radioactive isotopes. In the other technique a thin layer of the radioactive metal is deposited oil one face of the solvent metal, and after diffusion has occurred the specimen is cut in thin sections to obtain the concentration-penetration curve. This method has been used to determine the coefficients of self-diffusion of gold,13 zinc,14-l6 silver," and copper.17 The second method was used for the work described in this paper. Materials The coefficients of self-diffusion were determined in two grades of polycrystalline copper and in single crystals of copper. Commercial OFHC copper was used in the form of disks I in. in diameter by % in. thick. Samples of high-purity copper described by Phillips and Skinnerl9 were supplied through the courtesy of the American Smelting and Refining Co. The high-purity copper rod was used without remelting by cross-rolling to a thickness of 0.075 in., from which disks 7/8 in. in diameter were cut. A good grade of bus-bar copper was used to make the single crystals.

Jan 1, 1942

By O. Jantti, Erkki Laurila, R. T. Hukki

INVESTIGATION of the methods of analyses for magnetite and ilmenite in the Otanmaki iron-titanium ore and respective mill products has resulted in certain improvements in the methods conventionally employed. The outcome of the investigations is summarized as follows: 1—Development of a diamond-drill core analyser for the determination of magnetic susceptibility of drill cores containing magnetic minerals. 2—Development of a magnetite-ilmenite analyser for pulverized products. 3—Development of a potentiometric method of chemical analysis for the determination of iron, vanadium, and titanium in succeeding steps from one test sample. 4—Proposal for a continuous and automatic magnetite assay method. Diamond-Drill Core Analyser The conventional method of assaying diamond-drill cores consists of splitting the cores, pulverizing them and running chemical analyses on representative samples. The physical properties of magnetic minerals, however, offer an entirely different approach to obtain similar final result by the measurement of respective magnetic susceptibilities in a properly designed instrument. This instrument is shown in Fig. 1. Essentially it consists of a magnetic unit, A, and of a meter unit, B. On top of A is a slot where a piece of diamond-drill core C is visible. The magnitude of disturbance in the original magnetic field caused by the magnetic minerals in the drill core can be observed from the microammeter. This reading is a function of the combined susceptibilities of all magnetic grains in the magnetic field. The magnetic unit is built of E-sheets of a shell transformer, as shown in Fig. 2. The primary coil and two identical secondary coils are wound around the central leg and end legs, respectively. The drill core under examination is placed into one of the air gaps as shown in Fig. 2. The number of ampere-turns of primary coil excited by 50 or 60-cycle ac give a maximum induction of about 600 gauss in the air gaps. The secondary coils are connected in series in the opposite direction. Thus their voltages compensate each other if no drill core exists in the air gap or if the susceptibility of the drill core is equal to zero. As soon as the susceptibility becomes greater than zero, the balance of the magnetic circuit is disturbed, resulting in increased voltage in the terminals of coil 1 and in a respective decrease of voltage in the terminals of coil 2. The voltage difference thus produced is a function of the effective susceptibility of the drill core. If the magnetic grains in the drill core are unevenly distributed, the voltage difference varies as the core is rotated around its own axis. The voltage difference is measured with a Bren-tan0 tube voltmeter. If extreme sensitiveness is desired, the established symmetry of the magnetic circuit does not allow reduction of the 0-voltage difference far enough. Therefore a compensatory voltage of a correct phase taken from the terminals

Jan 1, 1952

By R. W. Gurry

When a steel in the austenitic state, with all its carbon in solution, is maintained, at constant temperature, in contact with a gas that removes the carbon from the surface, yet without otherwise altering the composition of the steel, the change in weight of the specimen, after any period of exposure, is a direct measure of the amount of carbon it has then lost. The purpose of the present paper is to show how, by application of established mathematical reasoning based on the fundamental diffusion law and on certain postulates, this loss of weight yields a complete picture of the distribution of carbon throughout a specimen of regular shape after any period of exposure, with all the accuracy required in any practical occurrence of decarburization. Precisely the same is true for the gain in weight during the reverse process of carburization, or for the gradual gain or loss of any element, so long as the steel comprises only a single phase at temperature, provided always that the experimental conditions conform to the postulates. The change in carbon content from surface to center of the specimen is always continuous, when everything is in solution, even though under the microscope the decarburized zone (or carburized rim) may appear to be sharply defined. The experimental procedure is described and some typical results are presented, which are then interpreted by application of the diffusion law. Experimental Procedure Initially an endeavor was made to follow the gain of weight during the process of carburization in a hydrogen-toluene atmosphere, but it was soon found that the specimen became coated with graphite, which could not be removed without some loss of metal; therefore it was decided to continue carburization until the steel was saturated throughout, to remove all graphite from the surface of the specimen, and then to follow the loss of weight during decarburization in a moist hydrogen atmosphere at the same temperature as that used for carburizing. The carburizing procedure has been described in an earlier paper;' it suffices to state here that the carburization lasted about I40 hr. at 760°C. (1760°F.) and 44 hr. at iiio°C. (2030°F.), these periods being long enough-—on the basis of the calculations presented later—to ensure that the specimen was about 99.9 per cent saturated at each temperature. The specimen was a cylinder about 3.8 cm. long and 0.475 cm. in diameter; the materials were carbonyl iron, a commercial low-metalloid iron, and a steel with 3.6 per cent nickel and the usual proportion of other elements. The fully carburized specimen, after cooling at the top of the furnace tube, was placed in a lathe and sandpapered to remove all graphite; then the diameter was measured with a micrometer caliper, and the initial weighing (corresponding to complete saturation) was made. By means of a

Jan 1, 1942

By R. W. Gurry

When a steel in the austenitic state, with all its carbon in solution, is maintained, at constant temperature, in contact with a gas that removes the carbon from the surface, yet without otherwise altering the composition of the steel, the change in weight of the specimen, after any period of exposure, is a direct measure of the amount of carbon it has then lost. The purpose of the present paper is to show how, by application of established mathematical reasoning based on the fundamental diffusion law and on certain postulates, this loss of weight yields a complete picture of the distribution of carbon throughout a specimen of regular shape after any period of exposure, with all the accuracy required in any practical occurrence of decarburization. Precisely the same is true for the gain in weight during the reverse process of carburization, or for the gradual gain or loss of any element, so long as the steel comprises only a single phase at temperature, provided always that the experimental conditions conform to the postulates. The change in carbon content from surface to center of the specimen is always continuous, when everything is in solution, even though under the microscope the decarburized zone (or carburized rim) may appear to be sharply defined. The experimental procedure is described and some typical results are presented, which are then interpreted by application of the diffusion law. Experimental Procedure Initially an endeavor was made to follow the gain of weight during the process of carburization in a hydrogen-toluene atmosphere, but it was soon found that the specimen became coated with graphite, which could not be removed without some loss of metal; therefore it was decided to continue carburization until the steel was saturated throughout, to remove all graphite from the surface of the specimen, and then to follow the loss of weight during decarburization in a moist hydrogen atmosphere at the same temperature as that used for carburizing. The carburizing procedure has been described in an earlier paper;' it suffices to state here that the carburization lasted about I40 hr. at 760°C. (1760°F.) and 44 hr. at iiio°C. (2030°F.), these periods being long enough-—on the basis of the calculations presented later—to ensure that the specimen was about 99.9 per cent saturated at each temperature. The specimen was a cylinder about 3.8 cm. long and 0.475 cm. in diameter; the materials were carbonyl iron, a commercial low-metalloid iron, and a steel with 3.6 per cent nickel and the usual proportion of other elements. The fully carburized specimen, after cooling at the top of the furnace tube, was placed in a lathe and sandpapered to remove all graphite; then the diameter was measured with a micrometer caliper, and the initial weighing (corresponding to complete saturation) was made. By means of a

Jan 1, 1942

By Wallace Thorneycroft

This paper by Mr. Thorneycroft, Past President of the Institution of Mining Engineers (Great Britain), and chairman of its Subsidence Committee, is a valuable contribution to the assemblage of data on ground movement and subsidence from coal mining. It deals with the effects of advancing longwall, a method almost universal in Europe but as yet little used in the United States outside of Northern Illinois and in a few scattered places elsewhere in thin coal beds, as in the Lexington field, Missouri. Canon City field, Colorado, and the Birmingham field, Alabama. Although by advancing lougwall, a maximum recovery of coal is obtained and it has to be employed where there are weak shale roofs, especially in deep mines, it is a more expensive method of operation unless the mining conditions are suitable, which is unusual in the United States. On the other hand, there is a growing use in this country of what is essentially retreating longwall, with an approximately straight break-line as described by Rayburn.1 With this method, in which the pillars are substantial, it was shown that the ground movement effect is practically identical with that of advancing longwall, hence the value to mining engineers of the precise data given by Mr. Thorneycroft. His study is unusual in several respects; it deals with conditions connected with the mining of a coal bed 20 in. thick, about 490 ft. below the surface, over a mined-out area in a 5-ft. bed about 90 ft. below, which had been mined about 12 years earlier. Both mines passed under Plean House, a building of massive cut stone construction built over a century ago, in which Mr. Thorneycroft resided throughout the under-mining period. Slight cracks opened in the walls as the longwall face approached the house in mining the upper seam, and closed after the face passed and subsidence took place. In the northwest wall, the last wall the face passed under, some stone blocks were broken, which was attributed to a strike period when the advance of face came nearly to a standstill. Other than this there was no perceptible permanent damage, as indicated by the photographs taken by the undersigned in 1923 (Figs. 1 and 2). Mr. Thorneycroft used a novel means of studying the successive positions of the corners of the building: he hung plumb bobs from the eaves and traced the paths of the respective plumb bob points on level surfaces. The angles made by the house walls with the plumb bob lines gave measures of the slope of the subsidence wave at the respective points. A slight upward wave was indicated by the surveys ahead of the subsidence, which would appear to be due to a tilting over of the massive sandstone stratum in the roof measures near the surface, the edge of the coal face acting as the fulcrum.

Jan 1, 1931

By Wallace Thorneycroft

FOREWORD This paper by Mr. Thorneycroft, Past President of the Institution of Mining Engineers (Great Britain), and chairman of its Subsidence Committee, is a valuable contribution to the assemblage of data on ground movement and subsidence from coal mining. It deals with the effects of advancing longwall, a method almost universal in Europe but as vet little used in the United States outside of Northern Illinois and in a few scattered places elsewhere, in thin coal beds, as in the Lexington field, Missouri, Canon City field, Colorado, and the Birmingham field, Alabama. Although by advancing- longwall, a maximum recovery of coal is obtained and it has to be employed where there are weak shale roofs, especially in deep mines, it is a more expensive method of operation unless the mining conditions are favorable, which is unusual in the United States. On the other hand, there is a growing use in this country of what is essentially retreating long wall, with an approximately straight break-line as described by Rayburn.1 With this method, in which the pillars are substantial, it was shown that the ground movement effect is practically identical with that of advancing longwall, hence the value to mining engineers of the precise data given by Mr. Thorneycroft. His study is unusual in several respects; it deals with conditions connected with ,the mining of a coal bed 20 in. thick, about 490 ft. below the surface, over a mined-out area in a 5-ft. bed about 90 ft. below, which had been mined about 12 years earlier. Both mines passed under Plean House, a building of massive cut stone construction built over a century ago, in which Mr: Thorneycroft resided throughout the under-mining period. Slight cracks opened in the walls as the longwall face approached the house in mining the upper seam, and closed after the face passed and subsidence took place. In the northwest wall, the last wall the face passed under, some stone blocks were broken, which was attributed to a strike period when the advance of face came nearly to a standstill. Other than this there was no perceptible permanent damage, as indicated by the photographs taken by the undersigned in 1923 (Figs. 1 and 2). Mr. Thorneycroft used a novel means of studying the successive positions of the corners of the building: he hung plumb bobs from the eaves and traced the paths of the respective plumb bob points on level surfaces. The angles made by the house walls with the plumb bob lines gave measures of the slope of the subsidence wave at the respective points. A slight upward wave was indicated by the surveys ahead of the subsidence, which would appear to be due to a tilting over of the massive sandstone stratum in the roof measures near the surface, the edge of the subsidence area acting as the fulcrum.

Jan 1, 1931

By R. L. Fullman

Relationships are derived between average dimensions measured on a polished cross section and the spatial dimensions of particles dispersed as uniform cylinders. The equations are applicable to the measurement of rods or plates without the stipulation that they be extremely thin. QUANTITATIVE analysis of the elements of mi-crostructures has been limited by a lack of knowledge concerning the relationships between observations on plane surfaces and the spatial nature of the structure. This paper is an extension of previous efforts to establish equations suitable for the quantitative investigation of microstructures. The equations derived fill two gaps in the range of particle shapes for which measurements may be required, namely those shapes lying between extremely thin plates and equiaxed particles (approximately spherical), and those intermediate between equiaxed particles and extremely thin rods or needles. In a previous paper' equations were developed relating dimensions measured on a polished cross section to the average spatial dimensions of particles in the form of spheres, extremely thin plates, and extremely thin rods. For extremely thin rods the dimensions measured on a polished cross section were found to depend on the radius only, so that no information concerning the rod lengths or number of rods per unit volume could be obtained from observations on a cross section. In this paper equations are derived relating dimensions measured on a polished cross section to the spatial dimensions of particles in the form of identical cylinders. These equations may be used to evaluate the dimensions of rods, as well as the dimensions of plates, without the stipulation that the plates or rods be extremely thin. The impossibility of evaluating the length of extremely thin rods from planar measurements appears as an increase in the amount of data required for a given precision, as the length-radius ratio increases. Consider a sample containing particles of a phase in the form of cylinders of uniform radius T and height t. If the cylinders are randomly oriented, as in a sufficiently large sample of a fine grained poly-crystalline material, area and lineal analysis may be carried out with parallel cross-sectional planes and lineal traverses. If the cylinders are not randomly oriented, it is necessary to randomize the orientations of the cross-sectional planes and traverse lines. Let a unit cube be cut from the sample, and a cross-section plane be passed through the cube parallel to one of the cube faces. The number of particles N, cut by the cross-sectional plane per unit area is equal to the number of particles N, per unit volume times the probability p1 of a plane intersecting a single randomly positioned and randomly oriented cylinder in the cube. Let J be the maximum cylinder dimension in the direction normal to the cross-section planes, as shown in Fig. 1. Since, of the various possible positions for the cross-sectional plane over the unit length from top to bottom of the cube, only those positions existing over the length J

Jan 1, 1954