Search Documents

By John W. Bell

The aim of the laboratory experiments described in this paper was twofold: 1. To measure as accurately as possible the maximum amount of crushing that can be effected by 1 hp. acting for 24 hr., by the two proposed methods for measuring the amount of crushing, namely: (a) by Rittinger's method; (b) by Stadler's method, based on Kick's law. 2. To show, if possible, that by one of these methods, the amount of crushing produced by 1 hp. in 24 hr. is a constant amount throughout a wide range in the diameter of the piece crushed, and that by the other method, the amount of crushing is a variable quantity over the same range of diameter; or else show that neither of the proposed methods for measuring the amount of crushing indicates a fixed relation between power and crushing and consequently a new method for computing this amount will have to be found to permit the establishment of a law of crushing. Experimental Methods Testing machines, calorimetric methods and commercial types of rock-crushing machines were considered with reference to their value for measuring power used in crushing. The following electrically driven crushing machines, installed and available in the McGill Ore-dressing Laboratory and which had been used in previous rock-crushing tests, on the whole seemed especially suited to the investigation: (1) Comet "A" crusher; (2) 7 by 9-in. Dodge crusher; (3) 10 by 16-in. rolls. In later tests a 3 ½-ft. Huntington mill was used.

Jan 1, 1918

By W. W. Odell

ONCE it was believed that anthracite or coke were the only fuels generally available and suitable for the generation of water gas, particularly so when this gas was made in the generators of standard carburetted water-gas sets such as are commonly used in the manufacture of city gas. During the past 8 years, practice has changed considerably and it is -now common to find gas plants operating without anthracite or coke and using bituminous coal in the generator. This change in fuel has appreciably increased the marketability and sales of certain coal producers' products; accordingly it is of interest to coal operators and others to know the reason for this change, the present trend in practice and the adaptability of various coals for generation of water gas. This paper briefly discusses these conditions and some of the factors relating thereto. BITUMINOUS COAL NOT IDEAL BUT CHEAPER THAN COKE In spite of many changes which have been made in both process and generating apparatus for using bituminous coal as generator fuel more efficiently, in general it is not an ideal fuel for this purpose. It is supplanting coke because it costs less; it is so much cheaper per ton delivered at the gas works that water gas made therefrom costs less per unit-volume than that made from coke, even though the quantity of generator fuel consumed per therm (100,000 B. t. u.) of gas made is greater with coal. This is the reason why so many city gas plants, as well as others, are using bituminous coal instead of coke.

Jan 1, 1927

By Harold Rabling

THE following notes are taken from results obtained on a standard 25-ft. Hancock jig 1 tested during regular operation in the Bonne Terre mill of the St. Joseph Lead Co. The object of the tests was to determine the conditions for most effective 'work, and the nature of products that could be made. The material treated was sized between 9mm. and 2-mm. round-hole screens, and was constant throughout the experiments, Three products were made, concentrates from the first three hutches, middlings from the fourth and fifth, and tailing from the sixth. The middlings were crushed in one pair of Allis-Chalmers Style B 30 by 14-in. rolls, and elevated with the original ore to the 2-mm. screens, oversize of which formed the jig feed. While this arrangement was very simple, it gave rise to undesirable crowding of the circuit. Table 1 shows tonnage and assay values of the jig feed and products at the beginning of the experiments, and brings out the excessively large amount of middlings, comprising 63.3 per cent. of the total material fed to the jig. The rolls, being heavily overloaded, did very little crushing, a screen analysis of the product showing that only 27 per cent., or 128 tons, was crushed fine enough to pass to the tables. The remaining 347 tons returned to the jig with 400 tons of original ore. At times, this overcrowding became much worse, and the tonnage in circuit grew to enormous proportions. It became imperative

Jan 8, 1917

By Harold Rabling

The following notes are taken from results obtained on a standard 25-ft. Hancock jig1 tested during regular operation in the Bonne Terre mill of the St. Joseph Lead Co. The object of the tests was to determine the conditions for most effective work, and the nature of products that could be made. The material treated was sized between 9-mm. and 2-mm. round-hole screens, and was constant throughout the experiments. Three products were made, concentrates from the first three hutches, middlings from the fourth and fifth, and tailing from the sixth. The middlings were crushed in one pair of Allis-Chalmers Style B, 30 by 14-in. rolls, and elevated with the original ore to the 2-mm. screens, oversize of which formed the jig feed. While this arrangement was very simple, it gave rise to undesirable crowding of the circuit. Table 1 shows tonnage and assay values of the jig feed and products at the beginning of the experiments, and brings out the excessively large amount of middlings, comprising 63.3 per cent. of the total material fed to the jig. The rolls, being heavily overloaded, did very little crushing, a screen analysis of the product showing that only 27 per cent., or 128 tons, was crushed fine enough to pass to the tables. The remaining 347 tons returned to the jig with 400 tons of original ore. At .times, this overcrowding became much worse, and the tonnage in circuit grew to enormous proportions. It became imperative

Jan 1, 1918

By D. Walton

The effect of small solute additions of Cu on the activation energies for creep A1 single crystals were determined over the range from 78° to 850° K. Below 240°K and above 800°K activation energies were unchanged. Between 240°K and 400°K strain aging causes the creep rate to become vanishingly small and the flow stress was independent of the temperature. Between 500° and 50°K the activation energy was increased to 41,000 cal per mole. NUMEROUS investigationsL have shown that solid solution alloying invariably increases the flow stress. A typical example is documented in Fig. 1. Such increases in flow stress can arise from two general effects: Alloying can increase the long-range stress fields through which dislocations must move by causing appropriate clustering; modification of dislocation densities, and so forth. Alloying can also modify the short-range stress fields that determine the activation energies for thermally stimulated migration of dislocations, such as occurs in some solute atom pinning processes, and so forth. Actually both long- and short-range stress-field effects can occur simultaneously. It was the purpose of this investigation to evaluate the effect of alloying on the short-range stress fields by determining the effect of alloying on the activation energies for creep. The resulting observations also shed some light on the effect of long-range stress fields on dislocation processes. EXPERIMENTAL TECHNIQUE Single-crystal bars (6 in. long 3/8 in. wide and l/4 in. thick) of dilute solutions of Cu in A1 were seeded, so as to provide extensive single slip, with the pole of the tensile axis located in the standard stereographic projection as shown in Fig. 2. Each crystal was grown under an argon atmosphere in a graphite mold containing a large reservoir of molten alloy for the purpose of obtaining rather uniform composition of Cu along the length of the bar. The nominal compositions of the alloys which were produced from 99.99 pct Cu and high-purity A1 containing 0.001 pct Fe and 0.001 pct Si, are given in Table I. Strains, measured by linear variable transformers attached by quartz rods to a 3 in. gage section were sensitive to and the stresses were ap- plied by direct loading of the specimen. Temperatures were determined by thermocouples attached to the gage section of the specimens. Below 400°K the apparent activation energies for creep, Q, were determined by the effect of small, rapid changes in temperature of about + 10°K from T1 to T2 on the corresponding creep rates ?1 and ?2 during primary creep according to where R is the gas constant.' Above 400°K, Q was evaluated from Eq. [1] for the secondary creep rate using temperature changes of about ±20K. A typical set of results is shown in Fig. 3. RESULTS AND DISCUSSION The previously obtained effect of temperature on the apparent activation energy for creep of single crystals of high-purity A12 is shown by the broken curve of Fig. 4. The datum points on the same graph illustrate the effect of dilute a solid solution alloying with Cu on the activation energy between 600" and 750°K. Each plotted point is the average of not less than five independent test values between 600° and 750°K and not less than three at other tempera-

Jan 1, 1962

YEARS ago, when metal mining was more an adventure than a well-organized business, it was often said (with some color of truth) that unless a mine could stand bad management or mismanagement, it was not much good. Today, the natural excellencies of an ore deposit are merely a challenge to the engineers to exploit it as skillfully, as efficiently, and as profit- ably as possible. A vast amount of competent and imaginative engineering management (which includes economic analysis no less than the technologic arts of mining ore and treating it) has been required to make New Cornelia the smooth-running and successful operation that can be seen at Ajo today. New Cornelia was acquired by the Phelps Dodge Corporation when it absorbed the Calumet and Arizona Mining Company in 1931. Next to Morenci, it doubtless is the corporation's biggest asset. It in no way detracts from the achievement of Phelps Dodge engineers to say that they had an admirable ore deposit to work on. New Cornelia ore, as mined, is relatively low in grade and the deposit is by no means the largest of the Porphyries; but it has these attractive features : (1) It is peculiarly suitable for opencut mining because: (a) The overburden of waste rock lying above the ore deposit was (it has long since disappeared) almost negligible. (b) The deposit is roughly circular in horizontal section, with the horizontal dimensions four to five times as great as the maximum thickness. (Nevertheless, eventually some underground mining may be desirable.)

Jan 1, 1957

Jan 1, 1936

By A. F. Henderson, C. F. Cornell, A. F. Dunyon, D. A. Dahlstrom

IN processing magnetic taconites several steps of crushing, grinding, classification, and magnetic separation are required to produce a 60' pct Fe concentrate. Usually the final concentrate is in a slurry form of 55 to 70 pct solids by weight and is further characterized by a particle size of 60+ pct —325 mesh. In taconite plants already completed or under construction the conventional method used to obtain pellets suitable for blast furnaces is to filter the concentrate slurry, ball the filter cake, and then harden the green pellets in a traveling grate or in a vertical shaft furnace. Large tonnages must be handled in minimum floor space, with reduced maintenance and lowest possible operating cost. Filtration must therefore achieve two objectives: I) high filtration rate—in terms of pounds of dry solids per hour per square foot of filtration area—at a low filtrate solids concentration, and 2) a filter cake moisture that will always permit efficient balling within the normal fluctuations of ' plant operating conditions. If moisture contents are too high, balling may be impossible or result in a mushy ball .without enough green strength for the firing operation. Development work has met these objectives. At low investment and operating costs, filtration rates of 250 to 600 dry lb per hr per ft2 have been attained for concentrates of 60' pct —325 mesh. By combining the disk-type continuous filter with proper agitation, the Agidisc filter produces a homogeneous filter cake, easily discharged from the disk sectors. This application provides the lowest initial cost per square foot of filtration area, minimizes floor space, reduces down time for changing filter media to a negligible factor, and simplifies process control and maintenance for the operator. It is now possible to predict very closely the taconite concentrate filtration rates for a wide variety of conditions—either by experience or with a few simple leaf tests. When balling and pelletizing operations began on a continuous basis, it became apparent that moisture content of the filter cake was a critical factor and must always be controlled by the filter operator below some maximum allowable value. Accordingly the Reserve Mining Co. and the Eimco Corp. joined in a study of the many variables affecting filter cake moisture content. The study was conducted at the Reserve Mining Co. taconite plant at Babbitt, Minn. To process plant tonnage the Babbitt plant utilizes a 6-ft diam Agidisc filter with six disks and an 8-ft diam drum filter with 10-ft face. However, most of the plant filter investigations were made on the Agidisc because of its superior performance. A considerable amount of small-scale test work was first carried out to study more closely the individual variables and ascertain their effect on filter cake moisture content. A general correlation method was then developed which allowed reasonable prediction of moisture content at any operating conditions. This step was made necessary by the fact that in full-scale filtering, operating variables usually interact to produce a combined rather than a singular effect. Studies were then made on the full-scale filters to prove the validity of the correlations developed, and past operating records were examined to ascertain whether or not significant variations could be traced to any single factor or to a combination of factors.

Jan 1, 1958

By Frederic P. Dewey

By the old method of determining platinum in ores and bullion, the silver-alloy first obtained in the regular course of assay is parted in strong sulphuric acid and the residual metal weighed. This is re-alloyed with silver by a second cupellation and parted in nitric acid, the residual metal being again weighed. Any difference shown between the two weighings is assumed to be, and is called, platinum. Sometimes it is SO, and if any considerable amount of platinum be present, there will be a decided difference between the two weighings; but a slight difference is no real evidence whatever of the presence of platinum. On the other hand, the second weight may equal or possibly exceed the first, even when traces of platinum are present. Again, other members of the platinum group may go into solution in nitric acid more or less. If present, these would be called platinum and escape detection. The method does not provide any direct tests whatever as to the presence or absence of platinum. It is often indecisive, and sometimes gives erroneous results. It is, therefore, quite unsatisfactory. Being called upon many times to determine platinum in a wide variety of materials, particularly when present in very small amounts, I have realized the disadvantages and defects of this old method. In an article on the solubility of gold in nitric acid,' I have briefly outlined a method of gathering a little gold out of a solution containing much silver, which furnishes the basis of

Jan 1, 1913

By W. M. Peirce, E. A. Anderson

A FEW metals have been known since the dawn of history. Many have been prepared in quantity only within the present century. Zinc falls into an intermediate category, although there is some evidence of its existence in elemental form in Greece as early as the third or fourth century B.C. Nevertheless, the first important use of zinc antedates the earliest records of civilization. This first use was in the alloy, brass, which is roughly one third zinc. This continues to the present time as one of the major uses of the metal. The first production of brass, in light of present knowledge, was perhaps one of the 'most romantic and fortuitous of technical discoveries. Copper, the major constituent of brass, had been given to man by Nature in the native state; that is, as lumps of fairly pure copper. The art of melting and working it had been learned. But zinc does not occur as the metal in Nature and even if some early metallurgist happened to heat zinc ore with a reducing agent, the reduced metallic zinc passed off as vapor and was lost. Doubtless some early coppersmith, melting copper in accidental contact with zinc ore, reduced the zinc, and the vapor was absorbed by the molten copper. Its color, hardness, and other properties were thus wonder- fully changed. And so the earliest metallic form of zinc, as nearly as we can deduce, existed in the alloy brass. Probably knowledge of the kind of rock with which to heat copper to change it to brass, the best

Jan 1, 1953

By R. F. Decker, J. R. Mihalisin

Phase transformations in a series of relatively pure nickel-titanium-aluminum binary and ternary alloys were studied. The purpose was to clarify age-hardening mechanisms, especially in predominantly titanium-hardened compositions. The micro-structures, as determined on a tirne-temperature-transformation basis by light and electron microscopy and X-ray diffraction, were correlated with room temperature and hot hardnesses. The stability of the ? phase under nonequilibrium conditions was shown to be important in this system. MANY of the alloys designed for service in the temperature range of 1200oF (649°C) to 1800°F (980°C) are austenitic iron-.nickel base alloys hardened by titanium and aluminum. The control of high-temperature properties of these alloys has been advanced by the knowledge of microstructures provided by the equilibrium diagrams of Taylor and Floyd1 and Taylor.2, 3 However, there have been notable instances in the predominantly titanium-hardened alloys where the correlation of properties with microstructure and the equilibrium diagrams has been unsuccessful. Mihalisin and carrol14 have noted in electron microscope studies of commercial alloys of this type that a fee phase precipitates and transforms to 71 (hep Ni3Ti) in stress-rupture tests. Bieber5 found that this fee phase remained in nickel-chromium-titanium alloys even when aluminum was excluded. Age hardening in iron-nickel base alloys has been attributed to 71 by Craver, Aggen, and Dyrkacz,6 ? (fee phase based on Ni3Al) by Beattie and Hagel7 and both phases by Lena,8 Yet, on the basis of the nickel-titanium-aluminum equilibrium diagram of Taylor and Floyd shown in Fig. 1, it has been difficult to understand the presence of ? in low aluminum alloys. Furthermore, the characteristic coarse dispersion of 71 has not been commensurate with the marked aging response. Also there have been anomalous grain boundary transformations in the iron-nickel base alloys which have markedly reduced notched stress rupture properties.9 These have been related to boron content and residual cold work. Recently, Golubtsova and Mashkovich,10 Buckle and Manene11 and Bogaryatskii and Tyapkin12 have noted a transition fee Ni3Ti in titanium-hardened nickel-base alloys. It was apparent from the latter studies that a transition phase was important in the nickel-titanium-aluminum system. The purpose of this study was to identify this phase, measure its transformation characteristics and kinetics, and attempt to correlate these findings with age hardening. It was believed that time-temperature-transformation studies, because of their adaptability to transition phases, would be of most value for this investigation. Accordingly, this method of study was undertaken on relatively pure nickel-titanium-aluminum alloys with controlled variations in boron content and residual cold work. EXPERIMENTAL PROCEDURES Compositions were selected from the nickel-titanium-aluminum isothermal section of Taylor and Floyd.' These compositions are plotted in Fig. 1 and were selected to obtain equal aging potential of approximately 15 pet by volume of the equilibrium precipitate. The materials were vacuum melted from

Jan 1, 1961

HAVING told you of the practices of the arts involving other metals, I wish to tell you also of the practice of that of tin.* Indeed, since this is an easily melted metal, in common use for the utensils which are made for human needs, it is an art known not only to skilled men but also almost to children, wherefore I could proceed with- out telling of it. However, besides the production of plates and bowls and many other dishes for eating from and keeping things in, this embraces two other arts, both of which you should know about because they are useful and beautifid. One of these is the art .of making the letters with which books are printed, and the other is that of making the sheets from which organ pipes are built. The vessels, as perhaps you have seen, are cast one by one in moulds made of white tuff. Several pieces fastened together are then put on the iron axle of a wheel rotated by hand, and [138v] they are turned with a slightly bent tool which has a cutting edge. In this way they are made thin and of a good shape. Then they are polished with a piece of linen cloth and a little powdered tripoli, and thus they are finished. The letters for printing books are made of a composition of three parts of fine tin, an eighth part of black lead, and another eighth part of fused marcasite of antimony.t The desired quantities of these metals are melted

Jan 1, 1942

IN order that you may be expert in every phase of this art of parting, I wish to caution you upon twelve points, all necessary, so that you may know beforehand what may happen to you and what provision you must make to obtain good results. Understand that the gain derived from this art lies in carrying out the work that you wish to do accurately and with profit, which means nothing other than working with certainty and seeing that there is no loss of gold or silver [69v] or substance of anything that is valuable in itself. For, since everything lost is silver or gold, it is valuable however little. it may be, and many little things obviously make one large one; for gain increases gain, and loss, loss. Therefore, you will have as your first task to seek to have the materials that you are to use for making the aqua fortis of as perfect a nature as possible. I tell you this because the saltpeter is often weaker and of worse quality than the alum. For this reason it is necessary to make an effort to have it not only well refined and free from earthiness but also of a nature as powerful as pig [dung] saltpeter, the color of which tends strongly toward yellow. In order to work better, you also refine this again before you put it to use, as I shall tell you in its place. Likewise, see if you can get alum of the red kind that comes from the Levant or from Cartagena, for it costs less and is more powerful. Furthermore; see that you have a good number of well-formed cucurbits of properly purified glass not too large and of, equal content and size. Above all, see that they contain, no indentations, bubbles, folds, or scratches, especially, in the, body or in the bottom, because they break easily and the nature and great force of the acid pierces them by eating away, and makes a little hole. For this reason, cucurbits. are always divided into four grades; one is taken for making aqua fortis, another for drying the waters when they are laden with silver, the third for recovering the acids from the separations, and the fourth are either rejected as useless or dangerous, or are saved for recovering the whites or for some other service. Whoever wishes to practice this art well should remember to keep a large supply of cucurbits, receivers, and alembics. The third is the precaution which one must take in luting them; first, to

Jan 1, 1942

By J. E. Lawrence

Copper-decorated lattice disorders in silicon have been analyzed by electron transmission microscopy, chemical etching, and P-N junction reverse current measurements before and following different gettering treatments. Lattice disorders acted as the nucleating sites for copper precipitation. Copper precipitates in silicon resist chemical etching, absorb electron beams more than silicon, and cause excess reverse currents. Totally gettered disorders will chemically etch rapidly, be transparent to electron beams at extinction contrast, and not contribute to excess reverse currents. Three different gettering procedures were performed to determine the effectiveness of different EXTRANEOUS metal contamination can alter the minority carrier lifetime and resistivity of bulk silicon as well as cause excessive reverse currents in semiconductor devices. Previous workers have reported such activity for gold,1,2 copper,'-* iron,' and rnanganese.2,5 Gettering is a method employed to render these impurities electrically inactive. However, some property of the junction-degrading metals prevents gettering procedures from being wholly reliable. The comparative effectiveness of gettering with different agents has been the topic of earlier publications.2,3,6-11 Two mechanisms for gettering metallic impurities have been proposed: the distribution coefficient mechanism by Thurmond and Logan7 and the compound formation mechanism by Goetzberger and Shockley.2 Both of these models evolved from data gettering mechanisms. First, phosphorus was allowed to enter a protective SiO2 film but remain away from the Si-SiO2 interface to test the effectiveness of the compound formation gettering mechanism. Second, phosphorus was allowed to enter the backside of silicon wafers to test the effectiveness of the lattice strain gettering mechanism. Finally, nickel was diffused into the backside of silicon wafers to test the effectiveness of the distribution coefficient gettering mechanism. In each case, copper precipitates dissolved as a result of the gettering treatment. Tightly packed precipitates will resist gettering more than loosely packed copper precipitates. acquired by the electrical analyses of gettered semiconductor materials. An electrical analysis, however, cannot identify the fate of the gettered metals. The radiotracer technique has recently been applied to determine the redistribution of metals following different gettering treatments.3,9,10 The spatial resolu tion of the radiotracer technique prevented the analysis of metal precipitates which contribute to the excessive reverse currents of P-N junctions. The gettering processes employed by the earlier workers, almost without exception, allowed the gettering agent to enter the silicon lattice. This procedure makes it quite difficult to differentiate between the distribution coefficient and the compound formation mechanisms. The metallographic analysis of gettered copper precipitates discussed in this work supplements the earlier work by obtaining direct evidence of silicon lattice defects decorated with copper before and following different gettering treatments. The effectiveness of a process which restricts the gettering agent, phosphorus, to a dielectric film and thus away from the silicon lattice is discussed in addition to the ef-

Jan 1, 1969

By Carle R. Hayward

The growing importance of metallography has caused a corresponding interest in the improvement of apparatus for preparing specimens of metals and alloys for microscopic examination. The purpose of this paper is to describe an electric heating-furnace, a grinding- and polishing-machine, and a device for mounting specimens, which are used in the metallographical work at the Massachusetts Institute of Technology. These three pieces of apparatus were designed and made in the laboratory of the Institute, and each possesses some original features which may be of interest. Electric Heating-Furnace. The accurate control of heat necessary for metallographic work is best obtained by the electric-resistance furnace. Uufortunately, however, even the platinum-wound furnaces cleteriorate with use and ultimately burn out, while the cheaper resistance-materials are often short-lived. Since the resistance-coil must be replaced from time to time, it is of advantage to be able to make this change with as little trouble as possible, and this is an important feature in the furnace shown in Fig. 1. A is a cylindrical galvanized-iron can, with two handles, B. A porcelain tube, C, passes through the central hole in the cover of the can and rests upon the asbestos disk, D; the bottom of the tube is held in place by the asbestos ring, 3. The space, F, around the tube is filled with powdered magnesia. G is an ordinary assay-crucible, with the cover inverted so as to present a smooth top for supporting a second porcelain tube, H. The latter is wound with " Excello " resistance-tape, 0.014

Jan 1, 1912

By C. W. Spencer, F. G. Hochgraf, R. F. Cheney

Columnar grained specimens of nickel, containing 0.25 pct Si and 0.22 pct Mn, have been exposed to liquid bismuth in the temperature range 670° to 1050°C. Under isothermal conditions the liquid penetration rate is constant. The activation energy for the process is 22,000 cal per g-atom. Small amounts of plastic deformation, introduced into specimens prior to exposure, increase the penetration rate. The slou. step in liquid film penetration is associated with a reaction which occurs at or near the leading edge of the film. Either transfer of nickel across the liquid-solid inteqace or diffusion in the grain boundary ahead of the liquid film is suggested. Aliquid metal, if brought into contact with the surface of an unstrained solid metal, will usually penetrate to some extent into the grain boundaries of the solid. Under isothermal conditions the extent of penetration after any particular time of exposure is related to the relative magnitudes of the interfacial tensions concerned.1,2 If the liquid-solid interfacial tension is less than one-half the grain boundary tension the liquid may penetrate at a relatively rapid rate for an indefinite distance into the grain boundary surfaces. If the liquid-solid interfacial tension is greater than one-half the grain boundary energy, liquid penetration occurs at a continually decreasing rate so that after the liquid has penetrated to a relatively short distance below the surface the process essentially ceases.3 Many observations of a qualitative nature concerning inter crystalline attack by liquid metals have been reported, for example, liquid bismuth upon copper,' liquid mercury upon brass,4 and liquid mercury upon nickel.' Scheil and Schessl determined the rate of penetration of liquid bismuth into the grain boundaries of hot rolled copper rod? At constant temperature, penetration rate was found to be independent of the depth penetrated, the rate in- creasing with increasing temperature corresponding to an activation energy of 20,000 cal per g-atom for the process. It was suggested that at lower temperatures the liquid travelled mainly along lines of three grain intersections while at temperatures greater than 1000°C essentially all grain boundary surfaces were covered. Recently, in a study of copper bicrystals having a common [loo] axis it was found that at 649°C bismuth penetrated 22, 25, and 63 deg boundaries while it did not penetrate 2, 12, 15, and 72 deg boundaries.' Elbaum reported that liquid gallium penetrates aluminum grain boundaries to a depth of 0.6 cm after 11 hr at 27°c.8 Assuming that the rate controlling step in the process was diffusion down the liquid film to the specimen surface, a diffusion coefficient of 10-5 cm2 per sec was calculated. This paper reports a study of the penetration of bismuth-nickel liquids into the grain boundaries of a nickel alloy, containing 0.25 pct Si and 0.22 pct Mn as principal alloying elements. The phase diagram

Jan 1, 1962

By Herbert H. Uhlig

ALTHOUGH the mechanical properties of high-carbon manganese-iron alloys, particularly the Hadfield manganese steels, have been established, the literature discloses discrepancies in the reported properties for the low-carbon alloys. Hadfieldlo summarized the existing data up to 1927, including the investigations of Guillet, Arnold, Strauss, Burgess and Aston, and reported his own values for alloys containing for the most part from 0.06 to 0.15 per cent carbon. He noted: From these comparisons it will be seen that while there is a broad and general agreement as to the character of the alloys at the different manganese percentages, there are disparities in individual cases, especially as regards the degree of toughness or brittleness displayed by an alloy with certain percentages and with particular types of treatment. These differences are not always readily explainable. Ib His data showed that alloys containing 4 to 10 per cent manganese water-quenched from 1000°C. were brittle and hard with maximum brittleness at about 7 per cent. Slight improvement in ductility was obtained with compositions from 10 to 17 per cent manganese. Alloys containing up to 17 per cent manganese slowly cooled rather than quenched from 870°C., he found, were slightly improved in ductility with significant improvement in alloys analyzing 4.1 and 4.8 per cent manganese. Recently Walters, Kramer and Loringa reported mechanical properties for alloys made up from electrolytic manganese and ingot iron and containing from 0.02 to 0.04 per cent carbon. They showed that a tempering treatment at 540°C. (1000°F.) for one hour significantly improved the ductility of compositions ranging from 6.7 to 20.7 per cent manganese. The differences of these compared with Hadfield&apos;s results they considered to originate in this tempering treatment plus the fact that Hadfield&apos;s alloys contained higher carbon and phosphorus (0.06 to 0.15 per cent C; 0.06 to 0.07 per cent P compared with 0.02 to 0.04 per cent C, <0.001 per cent P). In this paper it is shown that of the low- carbon alloys (0.02 to 0.03 per cent) containing from 3.3 to 21.5 per cent manganese moderately or rapidly quenched from 1000°C. some are brittle or ductile, depending on presence or absence of hydrogen. Hydrogen is readily absorbed at high temperatures, and much of it is retained when the alloys are cooled to room temperature. The presence of hydrogen in the range of composition from 10 to 18 per cent manganese is a major factor in determining relative ductility. At 22 per cent manganese, and also at compositions below 10 per cent manganese, factors other than or in addition to hydrogen play a role. These are discussed. Also data are presented to show that hydrogen absorbed by the high-carbon alloys of Hadfield composition is without effect on their mechanical properties.

Jan 1, 1944

By H. W. Schadler

Single crystals of tungsten, which were grown by electron bombardment floating zone refining, were strained 2 pet in tension and bending at 298°, 77°, and 20°K to determine the modes and crystallography of the plastic deformation of these temperatures. At 77° and 20°K plastic deformation occurs by slip on systems of the type (011) [111] and by twinning on (112)-type planes. For crystals strained at 298°K, deformation occurs by slip in [111] directions on either (Oil)- or (112)-type planes. CURRENT models for the initiation of brittle fracture in crystalline materials are based on a knowledge of the details of their plastic deformation.1 Body-centered cubic tungsten exhibits brittle failure at temperatures below about 500 °K yet the modes and crystallography of its plastic deformation have not been studied below 1000°K. In 1924, Goucher2 reported that tungsten single crystals with a total impurity content of less than 0.01 wt pet deformed by slip on the (112) [111] system in the temperature range 1000° to 3000°K. He also observed (001) [010] slip in one crystal which was constrained by the presence of a large fissure. According to Barrett3 tungsten also deforms by twinning on the (112) plane in the [111] direction, but he gives no details on the tendency for twin formation or the temperature range in which twinning occurs. Therefore, as a necessary part of a study of the factors influencing the brittle fracture of tungsten, the modes and crystallography of the plastic deformation were determined at 298°, 77°, and 20°K. Tungsten single crystals, grown using the electron-bombardment zone-refining technique first described by Calverly,4 were deformed to between 1 and 2 pet plastic strain in tension and bending. The surface offset resulting from slip or twinning was observed with an optical microscope to determine the crystallography of the deformation. The existence of twins was proven by the standard polishing and etching technique; and the crystallography of dislocation motion was followed using the dislocation etchant described by Wolff.5 Some observations on the fracture characteristics of tungsten single crystals are included. EXPERIMENTAL TECHNIQUES Growth, Purity, and Perfection of the Crystals-Several years ago Calverly, Davis, and Lever4 dem-onstrated the feasibility of growing tungsten single crystals by the electron-bombardment floating zone-refining technique, and Carlson6 showed that crystals produced in this way had a high degree of purity and were ductile at room temperature. Tungsten single crystals produced by this technique in equipment similar to that described by Carlson were used in this study. The starting material, a powder metallurgy product, was obtained in the form of a swaged 1/8-in. rod from the General Electric Wire Plant in Cleveland, Ohio. The crystals were grown from this material in a vacuum of 10"4 mm of Hg by passing a molten zone 1/8 in. long down the rod twice at a speed of 3 mm per min. At this zone speed

Jan 1, 1961

By J. R. Kyte, C. C. Mattax

Previous workers have developed differential equations to describe oil displacement by water imbibition, but have not explicitly defined the relationship between recovery behavior for a single reservoir matrix block and its size. In the present work, imbibition theory is extended to show that the time required to recover a given fraction of the oil from a matrix block is proportional to the square of the distance between fractures. Using this relationship, recovery behavior for a large reservoir matrix block is predicted from an imbibition test on a small reservoir core sample. The prediction is then extended to analyze recovery behavior for fractured- matrix, water- drive reservoirs in which imbibition is the dominant oil-producing mechanism. Experimental data are presented to support the basic imbibition theory relating matrix block size, fluid viscosity level and permeability to recovery behavior. INTRODUCTION Imbibition has long been recognized as an important factor in recovering oil from water-wet, fractured-matrix reservoirs subjected to water flood or water drive. l-3 Recently, two approaches have been published which might be used to predict imbibition oil-recovery behavior for reservoir-sized matrix blocks. Graham and Richardson4 used a synthetic model to scale a single element of a fractured-matrix reservoir. lair,&apos; on the other hand, used numerical techniques to solve the differential equations describing imbibition in linear and radial systems. This latter method requires auxiliary experimental data in the form of capillary pressure and relative permeability functions. These two approaches, i.e., synthetic models and numerical techniques, have been used to study a variety of reservoir fluid-flow problems. One pur- pose of this work is to present, with experimental verification, a third method for predicting imbibition oil recovery for large reservoir matrix blocks. This method uses scaled imbibition tests on small reservoir core samples to predict field performance. The imbibition tests are easier to perform than the capillary pressure and relative permeability tests required to apply the numerical method. Furthermore, when suitably preserved reservoir-rock samples are used, the properties of the laboratory system are the same as those of the field. This offers an important advantage over the use of synthetic models because there is usually some question as to how accurately reservoir-rock properties can be duplicated in such models. Based on the recovery behavior for a unit matrix block, an analysis is presented to predict oil recovery for a fractured, water-drive reservoir made up of many such unit blocks. In the analysis, it is assumed that the flow resistance and the volume of the reservoir fracture system are negligible compared with that of the porous matrix. These assumptions are generally consistent with observed characteristics of many fractured-matrix reservoirs, and have been employed in previous studies.2,3,6-8 further assumed that the effect of gravity on flow in the matrix blocks is negligible. On first thought, the latter assumption might appear to seriously limit application of the method. However, in a fractured reservoir, the effect of gravity on flow in a matrix block will be restricted by the height of the block. Furthermore, matrix permeabilities are often very low (10 md or less) in such reservoirs. This means that capillary or imbibition forces will be large, thus tending to minimize the relative importance of gravity. For these reasons, imbibition should be the dominant oil-recovery mechanism in many fractured-matrix, water-drive reservoirs. The predictive method presented in this report is applicable to such reservoirs. SCALING IMBIBITION THEORY The "scaling laws" applicable to the operation.

By P. B. Crawford, W. L. Huskey

A study has been made to determine the effects of random vertical fractures existing in a reservoir matrix on the effective matrix properties and producing characteristics of a well. The study indicated that short vertical fractures may distort the isopotentials by 1 percent in the vicinity of the fracture. A correlation was developed which showed the effect of fracture density on the producing capacity of a well. The correlation indicated that fracture shape had little effect on producing capacity, but that total fracture length was closely correlated with the producing capacity. Effective values of reservoir permeability were found to correlate closely with fracture density when measured parallel to the direction of flow. INTRODUCTION In reviewing the petroleum literature, it is found that in most analytical solutions of hydrocarbon reservoir behavior the assumption is made that the rock is uniform throughout although core analysis and well logging surveys indicate that substantially all reservoirs are heterogeneous and some are fractured. Only in recent years have investigators studied the effects of reservoir hetrogeneity. Russell1 discussed the effect of fractures in reservoirs. The literature indicates that fractures greatly affect the exploitation of man reservoirs. This is described by Elkins and Skov2 and Little-field et al. 3 Several types of reservoir heterogeneity have been studied by unsteady-state methods. In 1959, Landrum et al. 4 showed that transient phenomena in hydrocarbon reservoirs could be studied by thermal models. pickering5 studied the flow of heat in stratified linear reservoirs consisting of plates of different metals separated by sheets of rubber which were suddenly exposed to a low-temperature source at one end. Cotman6 studied both laterally and vertically heterogeneous thermal model reservoirs with special reference to cross flow. Another type of reservoir heterogeneity involves the presence of highly permeable, circular lenses around or near the wellbore which was studied by Miesch7 using a thermal and a steady-state model. Many studies have been made of the effects of induced hydraulic fractures on well performance. 8-10 Work regarding vertical fractures that exist initially in the reservoir was reported by Givens, et al.11 Since there appears to be little published work on the importance and influence on reservoir performance of fractures that exist out in the rock matrix, this study was conducted to determine the effects of vertical fractures on pressure distribution, producing capacity and the effective permeability of the reservoir rock. The analogy between the flow of electrical current through a conducting body and the flow of fluid in a porous medium provides a convenient method of studying various aspects of petroleum reservoirs. For electrical radial steady-state flow,12 For steady-state fluid flow, With the above analogy, it is only necessary to establish the ratio of the values of the electrical units in terms of the corresponding flow units to make the analogy quantitative. The ratios are shown below.