Search Documents

By P. R. Sperry, P. A. Beck, J. Towers

Dahl and Pawlek1 found that electrolytic copper develops extremely coarse grains at 1000°C after about 90 pct reduction by rolling. This coarsening occurs only under conditions of penultimate grain size, deformation, and alloying which lead to the "cube" recrystallization texture.l,2,3,5 The peculiar angular shapes and straight grain boundaries of the coarse grains were noted by several investigator.1,4,5 On the other hand, coarsening in Fe-containing aluminum or in AI-Mn alloys8 does not depend on a "cube" (or any well developed) recrystallization texture. It is true that increasing deformation by rolling, and, therefore, an increasingly well developed re-crystallization texture, are associated with decreasing incubation periods of coarsening.6-7-8 Nevertheless, coarsening readily develops in aluminum even after only 30 pct reduction by rolling, where the recrystallization texture is very weak.6,8 Also, coarsening was observed by Jeffries9 many years ago in sintered thoriated tungsten, which presumably has no preferred orientation. In all these cases coarsening is associated with grain growth inhibition by a dispersed second phase.8,9 The annealing temperature has to be suficiently high to overcome the inhibition at a few locations. But if it is too high, growth starts at many points, and the resulting grain size becomes much smaller.9 Normally, the coarse grains are more or less equiaxed, and the boundaries have a typical ragged appearance.6.8 Cook and Macquarie4 demonstrated that, in addition to the texture-dependent coarsening previously found at 1000°C,l electrolytic tough pitch copper may also coarsen at 800°C after 50 pct reduction by cross rolling. The coarse grains formed under such conditions have rounded shapes and ragged boundaries, like those in aluminum. When the annealing temperature is higher, the tendency for their formation decreases. All these observations suggest that the coarsening at 800°C is associated with inhibition by a second phase. Actually, coarsening at 800°C after 50 pct reduction by cross rolling was observed only in tough pitch copper,4 which contains Cu2O particles. On the other hand, the texture-dependent 1000°C coarsening occurs in both tough pitch and oxygen-free copper;4 it does not appear to depend on the presence of a dispersed second phase. However, the interpretation of the 800°C coarsening in Cu after 50 pct rolling as an inhibition-dependent process, similar to the coarsening in A1-Mn alloys, is somewhat weakened by the fact that this coarsening was reported4 to occur only after cross rolling, and not after straight rolling. It was, therefore, decided to re-examine this question. A 1 in. diam electrolytic tough pitch copper rod, No. 2 hard drawn, was annealed for 20 min at 700°C, rolled to 0.5 in., annealed 10 min at 700°C, and straight rolled to 0.064 in. It was then given a penultimate anneal of 20 min at 500°C and it was cut into four sections, which were given final reductions by straight rolling as follows: A 30 pct reduction of area B 50 pct reduction of area C 70 pct reduction of area D 90 pct reduction of area Specimens cut from the four sections were finally annealed at 800°C in an oxidizing atmosphere. Strip A remained fine grained up to 10 hr, but the specimen annealed 12 hr consisted of only 2 large grains. Strip B had a few scattered large (1/2 to 3/4 mm) grains after 1 min, although the balance of the specimen consisted of fine grains of about 0.02 mm. After 5 min there were several 10 to 15 mm grains present, and after 1 hr strip B was completely coarsened. The coarse grains had the same characteristics (see Fig 1) as those obtained by Cook and Macquarie at 800°C after cross rolling. Strip C had several grains of 0.05 to 1 mm after 1 min, but it was still largely fine grained after 12 hr. After 48 hr it consisted entirely of grains of about 0.5 to 4 mm, with an extraordinarily large number of twin bands. Strip D remained com- pletely fine grained after 4 hr at 800°C. These results indicate that, in the deformation range of 30 to 70 pct reduction, the incubation period for coarsening as well as the rate of growth and the final size of the coarse grains decreases with increasing deformation. Similar

Jan 1, 1950

By F. A. Crossley

The additions: o, Sn, Zr, Cb, Mo, and V were studied for their effects on the a/a + Ti3A1 boundary of Ti-A1 base ternary alloys. These additions were chosen because of their importance to commercial titanium alloy practice—oxygen as the major impurity (in terms of influence on mechanical properties) and the others as major alloying additions. The composition and temperature ranges of investigation varied according to the system as shown in Table I. The principal means of investigation was light-microscopy. The basis of the ternary studies was the recently determined titanium-rich end of the Ti-A1 system.' MATERIALS, METHODS, AND PROCEDURES Ingots were prepared in 15 g size for all alloys except those of the Ti-Al-Mo and Ti-A1-V systems which were prepared in 20 g size. These alloys were prepared from high-purity, electrolytically refined titanium obtained from the Bureau of Mines, Boulder City Metallurgical Laboratory, Boulder City, Col., and Chicago Development Corp., Riverdale, Md., for the 15 g and 20 g ingots, respectively. The alloying additions were obtained from commercial sources, and were at least of 99.9 pct purity, except the V-15 wt pct A1 master alloy for vanadium additions. This alloy was of 98.2 pct purity; the major impurities were 0.37 wt pct Fe and 0.38 wt pct Si. The ingots were prepared by multiple melting in a nonconsum-able-electrode arc furnace. The ingots were covered with Markal coating "CR-T22" (Markal Co., Chicago) to protect them from scaling. Markal coating is applied as a slurry of solids in a volatile carrier liquid. The coated article is air dried. Upon heating to the hot working temperature the coating fuses to form a glassy, protective film. The ingots were heated to 1000°C and rolled to 0.16 in. thickness, reheating after each pass. The glassy film was removed by sand blasting. Ten mils were removed from all surfaces by grinding and/or pickling (3 pct HF, 30 pct HN03, balance water). Grinding was always followed by pickling to remove traces of the grindmg medium. Hydrogen removal was accomplished by vacuum annealing at 1000°C for 2 hr to a final pressure of about 0.02 µ Hg. This was followed by annealing in evacuated Vycor bulbs (800°C, 2 hr, ice brine quenched (IBQ)) in order to maximize ductility. The materials were then lightly pickled and cold rolled to 0.12 in. thickness or cracking, whichever occurred first. After rolling, 1/2 by 1/4 in. specimens were cut for the equilibration anneals given in Table 11 and carried out as previously described.1 Final polishing of the specimens for metallographic examination was done by electrolytic means. Specimens were first examined without etching under bright field and polarized light illumination, for this is the only means of detecting the "two-phase" phenomenon.1,2 The term "two-phase syndrome" refers to Ti-A1 al-loys which appear on the bases of their chemical, phys-ical, and mechanical properties, to be two-phase, but are actually single phase and inhomogeneous in composition. Since processing and heat treatment of the alloys of the first group were completed before the two-phase syndrome was discovered, a number of the specimens were found to exhibit this phenomenon—80 pct of those of the Ti-Al-Zr system, 5 pct of those of the Ti-Al-Sn system, 30 pct of those of the Ti-Al-Zr system. The affected specimens were homogenized by heating above the ß transus temperature and quenching. The heat treatment and working schedules to accomplish this are given in Table 111. After rolling, specimens were pickled (3 pctHF, 30 pct HNO3 in water) to remove 1 to 2 mils from the

Jan 1, 1970

By Anthony Tuke

Members of the Society of Mining Engineers may well regard it as rather unusual that a paper on this subject is being presented by someone whose first taste of mining came at the age of 60 or so - someone who last heard of cathodes and anodes in 1938. On the other hand, the title does refer to the financing of a mine and at least I can claim to have spent all my business life since the war until 1981 in the world of banking, though this inevitably means that I have been used to looking at these and other problems through what you might regard as the wrong end of the telescope. The lending of money, and more particularly the provision of finance for commerce and industry, is the central job of a banker and I have been involved in this in various ways over many years, but regrettably only very marginally in the world of mining. As you will see from what follows, the rather specialised form of finance which RTZ required was, to my regret, not provided by the British banking community since the great majority of the finance came from American banks and, to a lesser extent, from German banks. I am most indebted to Mr. Roy Wright for a great deal of the nuts and bolts in this paper. He was one of the three central figures in the growth and expansion of RTZ during the 1950's and 1960's and was right at the centre of the negotiations with bankers, with governments and with many others. The going, as you will see, was far from easy. The essential basis for the financing of each of the RTZ group's mining projects was a firm long-term sales contract for sufficient of the output of the particular mine to produce the cash flow necessary to service the loans. Wherever possible, loans were raised in the same currency as the sales contracts were made. Each sales contract was designed to give the customer the product he wanted with a long-term assurance of supply and at the same time to give the bankers the protection they required. Thus the marketing concept and the financing concept for a particular mine formed one overall plan, and talks with the customers and the bankers were conducted in parallel until a satisfactory marketing/financial package was agreed in principle. Preliminary talks were begun soon after the discovery of a potentially viable orebody. As confidence about a discovery increased, customers were persuaded to enter into contingent sales contracts before the very detailed and costly business of the economic and technical feasibility study of the mine was launched; such a study might cost 10% or more of the estimated total capital cost. At the same time provisional understandings were reached with the bankers. The contingent sales contracts usually gave a period of grace of eighteen months to two years, during which time the decision whether or not to continue with the mine had to be made. This decision rested mainly on the result of the feasibility study and also on coming to a firm agreement with the bankers. If all went well and the mine went ahead, then the contingent sales contracts became firm, but they fell away if the decision was negative. The Company was fortunate that during the whole period when its major mines were being developed, world trade was increasing rapidly and Japan was becoming a dominant industrial power. The Japanese supported many of the mines, including Hamersley Iron Ore, Lornex Copper, Bougainville Copper to name a few, with the basic long-term sales contracts that enabled their development. The Germans supported Palabora and to a lesser extent Bougainville. The mines were financed as individual projects generally with a loan/equity ratio of about 65/35, without legal recourse to RTZ for the loans. RTZ was, however, responsible for providing any capital overruns required to bring the mines into commercial operations. Nevertheless, the RTZ management regarded itself as morally responsible to the lenders because, Palabora aside, the sales contracts had either a fixed selling price (like Hamersley), an indexed price (like the Elliot Lake Uranium Mines and Mary Kathleen), or a floor price (as in Lornex, Bougainville and Rossing). Thus the Company believed that any financial failure would have been due to poor management or technical

Jan 1, 1985

By R. E. Hudrlik, G. W. Preckshot

The diffusion coefficients of silver from solid silver in molten lead were measured to within ± 0.8 pet in a columnar type diffusion cell ower, the temperature range of 326° to 530°C. Fick's law describes the process up to 530°C where the laminar mechanism appareltly breaks down. These is negligible resistance at the interface as shown by mathematical analyses. The diffusion coefficients are found concentration independent. IT would seem that diffusion in liquid metals would be free of such effects as molecular structure, dissociation. polarization. and compound formation. This view was taken by Gorman and preckshot in their study of diffusion of copper from solid copper into molten lead. They reported diffusion coefficients which were independent of the concentration over the range of 478° to 750°C. They found that the Stokes-Einstein equation with constant radius of the diffusing specie represented the diffusion data better than Eyring's rate theory equation and Sheibel's correlation. The radius of diffusion was found to be that of the doubly charged copper. There appeared to be no resistance across the solid-liquid boundary. In the present work the diffusion coefficients for silver in liquid lead were measured over a range of temperatures of 350° to 505°C. The solubility of silver in lead over the range of 303° to 630°C was also obtained. These results are compared with calculated or correlated values or with data in the literature. EXPERIMENTAL Procedure—The experimental equipment techniques and procedures were those reported in detail by Gorman and preckshot9 and will not be repeated here. Measured values of WT, Co, A. L were obtained for various diffusion times and the diffusion coefficient was computed for the case of no resistance at the interface9, 11 by: WT/CoAL = 1- 8/p2 n=1 1/(2n - 1)2 exp[-(2n - 1)2p2 Dt/4L2] [1] or where there was resistance at the interface by: WT = 1- ?n=1 2h2/ap2L [sxp [-Dan2t]/[(h2 + an2) L + h] The roots an are those of the transcendental equation3 tan (an L) = Iz/cun. The diffusion coefficient is that defined by Hartley and Crank.7 The total silver in the lead cylinder and equilibrium slug was determined by a cupellation technique' with proper correction for losses. Analysis of known samples showed that this method is surprisingly accurate. The amount of silver in the lead adhering to the silver cylinder was obtained in the same fashion as shown by Gorman and preckshot.9 The small errors involved in this determination are not critical since the silver in this adhering lead layer is only 3 to 15 pet of the total diffused. Materials—Electrolytic silver containing 99.9+ pet Ag obtained from General Refineries of Minneapolis, Minn. was used for all but runs 7 and 8. For the balance of the runs this silver was reduced with hydrogen at 1100°C and its oxygen content was found to be about 0.017 pet. For the runs. 7 and 8, phosphorous-reduced silver of the same purity was obtained from Handy and Harman Co. of Chicago, Ill. The densities of the phosphorus-reduced silver and the hydrogen-reduced electrolytic silver were 10.484 and 10.487 g per cm3, respectively. These values agree with those reported for pure silver. Silver which was reduced at 900 C had an average density of 9.998 g per cm3, indicating porosity. This silver was used for a number of runs which were not tabulated in Table I. These are indicated by crosses on Fig. 2. The 99.999 pet Pb was obtained from the National Lead Co. Research Laboratory of Brooklyn, New York. DISCUSSION OF RESULTS The diffusion and solubility results are reported in Table I for eleven runs using either phosphorus-reduced electrolytic silver or hydrogen-reduced silver at 1100° C. The solubility data shown in Fig. 1 show the excellent agreement with that reported by Heycock and Neville.8 The data of Friedrichs5 apparently are in error. The experimental solubility data of this work are reported to 0.3 pet. The experimental diffusion coefficients computed from Eq. [1] are reported within 1.2 pet of the mean and are plotted in Fig. 2. These are expressed within +0.8 pet of the experimental values over the entire temperature range by: D= 8.26 x 10 -5 e-1925/RT . [3] There appears to be little difference due to the

Jan 1, 1961

By C. G. Dunn, M. Sharp

IN twinned crystals of the face-centered cubic metals the lattice of one twin is a mirror image of the other in a common twin boundary. When several twins appear within large grain in a sheet specimen, the twin one boundaries form a set of lines at the surface of the specimen which coincide with (111) planes of the large grain. Furthermore, for twins of the same orientation, these lines are parallel. Generally, the presence of identically oriented regions with straight parallel boundaries coinciding with a (111) plane of the surrounding crystal is strong evidence for identifying the island regions as twins of the parent crystal. However, Fig. 1, which shows the macrostructure of a large grain of copper with island regions that satisfy these conditions. is not an illustration of (111) twins. Since the reverse side of the specimen has much the same appearance, it was thought at first that these regions, which appear dark in the macrograph, actually were twins. According to X-ray data, however, these regions are second-order twins of the large crystal. With regard to their formation, these second-order twins formed by secondary recrystallization in a cube texture matrix. Growth occurred in the direction of the arrow (see Fig. 1) as the specimen moved slowly into a gradient temperature furnace as described previously.' Nucleation of the second-order twins occurred, therefore, on the ends facing opposite the arrow. If the origin of the second-order twins were due to repeated twinning, some first-order twin structure should be visible on these ends. This proved to be the case, as very small twins were readily found with the aid of a microscope, and probably could have been seen, in some instances, under ideal lighting conditions without aid of a microscope. Fig. 2 shows a cross-section view taken perpendicular to both the surface and the (111) trace of the parent crystal (visible as a straight boundary in Fig. 1) at the beginning point of growth of a second-order twin and where one first-order twin was relatively thick. In the micrograph, A is the large parent grain; B is the first-order twin of A; and C, which is a first-order twin of B; is a second-order twin of A. Between A and B and between B and C the major straight portions are traces of common (111) twin boundaries. The straight portion of boundary between A and C, however, is not a common crystallographic plane to the two lattices; it is a (111) plane of A and a (115) plane of C. Without considering the mechanism of twinning itself, the origin of the second-order twins may be accounted for in terms of repeated twinning and special growth characteristics. After each nucleation, a selective growth process can be thought of as favoring growth of the first-order twin in local spots only and favoring growth of the second-order twin to an extent comparable with that of the parent grain over relatively large areas in a way similar to that described for twinning in aluminum.' It has already been pointed out that the boundary between the large grain (A) and the second-order twin (C), which is responsible for the straight boundary portions in Fig. 1, involves a (111) plane of A and a (115) plane of C. The same combination of planes is not only possible in first-order twins, but actually appears quite frequently.3 Their prevalence in first-order twins and their presence here in second-order twins, together with the necessary occurrence of a large number of common lattice sites at the boundary, is an indication that this combination produces an "energy cusp"' boundary. (Energy cusp boundaries have been described by Shockley and Read.") The configuration of atoms near a {Ill), (115) boundary in first-order twins is of course different from the configuration near the same type of boundary in second-order twins. References 1 M. Sharp and C. G. Dunn: Secondary Recrystallization Texture in Copper. Journal of Metals (January 1952) Trans. AIME, p. 42. 2W. G. Burgers and W. May: Stimulated Crystals and Twinning in Recrystallized Aluminum. Recueil des travaux chimiques des Pays-Bas (1945) 64, p. 5. aD. Whitwham, M. Mouflard, and P. Lacombe: Discussion of W. C. Ellis and R. G. Treuting, "Atomic Relationships in the Cubic Twinned State." Trans. AIME (1951) 191, p. 1070; Journal of Metals (October 1951). 4 W. Shockley and W. T. Read: Dislocation Models of Crystal Grain Boundaries. Physical Review (1950) 78, p. 275.

Jan 1, 1953

By V. Steve Reed, Ed L. Reed

The in situ leach mining process generates a waste stream that is high in sulfates, total dissolved solids, and radium 226. During the mining phase, the volume of the waste stream is relatively low and consists primarily of the bleed stream. During the restoration phase, larger volumes of waste water are generated. These waste streams require environrnentally sound disposal. The low net evaporation rate in the Coastal Bend area precludes pond evaporation as a feasible disposal alternative. Reverse osmosis is a practical method of reducing the volume of the waste water handled, but the concentrated waste stream from the reverse osmosis unit must be disposed properly. Deep well injection into highly saline reservoirs is considered a sound method of disposing of the liquid waste generated by in situ mining in the Gulf Coast uranium district. Thirteen injection wells have been permitted to serve the disposal needs of the leach mining industry in Texas. Of these 13, 11 have actually been drilled. Seven applications are pending. The injection zones for the permitted wells range from depths of 3050 to 6200 feet. Pressure limitations imposed on these wells range from 500 psi to 1350 psi. The following criteria are used to determine the desirability of a disposal well site: 1. A minimal number of nearby, improperly plugged borings which penetrate the disposal zone; 2. Minimal crustal disturbance; 3. Sufficient salinity of the water contained in the disposal zone; 4. Protection of oil and gas producing zones; and 5. Sand of sufficient permeability and areal extent to handle the desired volume without fracturing the reservoir. 1. Improperly plugged borings: During the early part of the century, oil wells, gas wells and test holes were drilled using cable tool equipment, often with a minimum amount of surface casing. Production casing, when it was set, was often partly removed when the holes were abandoned. Thus, wells drilled prior to 1940 frequently have less than 100 feet of surface casing and either no production casing or the upper part of the production casing removed. Additionally, these holes are often plugged only with mud. The close proximity of these holes to an injection well location are a concern in that they can provide an avenue for injection-depth fluids to migrate up the bore hole and jeopardize shallower fresh water reservoirs. Usually, where there are more than 6 or 8 poorly plugged borings in a 2 1/2 mile radius of the well site, it is preferable to examine deeper zones for disposal well potential. The deeper zones are especially attractive where the borings are not in a cluster, which renders monitoring more difficult. Often, even the deeper disposal zones are penetrated by a few improperly plugged borings. When this condition arises, the potential for leakage through the borings can be addressed in the following ways. a. Demonstration that the static head in the boring is higher than the anticipated increase in bottom hole pressure generated at the boring by the disposal well. A 100 psi differential between these two pressures is recommended. The calculated increased pressure at a boring caused by injection should be refined using annual bottom hole pressure measurements in the disposal well. Figure 1 illustrates an injection pressure map which can be overlain on the oil well map to determine the anticipated increase in pressure expected at each oil, gas or abandoned hole. b. Shallow ground water monitoring. A shallow monitor well is drilled next to the boring and both pressure and quality measurements are made periodically in the shallow well. c. Disposal zone monitoring. Recently there has been a tendency for regulators to require disposal depth monitor wells instead of shallow well monitoring. We consider disposal depth monitoring to be a less effective method of monitoring because it provides only indirect evidence of potential problems. Assumptions have to be made for the unplugged borings, such as mud weight, that are not addressed by the disposal zone monitoring program. There is little improvement with this system to that discussed in "a" above. A shallow zone monitoring program, however, yields direct evidence of a developing problem with an unplugged boring. Leakage by the boring will be detected quickly by an abnormal increase in pressure in the shallow well. Quality monitoring will detect upward migration of poor quality fluids. The pressure data provide an early warning of impending leakage; the quality monitoring will detect actual fluid migration.

Jan 1, 1980

By A. E. Dwight, Paul A. Beck

Known CsCl-type phases in transition metal alloys are shown to be formed preferentially at an electron concentration of approximately 6; a tetragonally distorted version of the CsCl structure is stable at -e/a = 6.5. The relationship of the distortion to electron concentration is verified by experimental data for the quasibinary systems Time ,Os) , Ti(Re,Ir), and Ti(Re,Pt) and the ternary system Ti-Ru-Rh. It was noted some years ago1j2 that the occurrence of CsC1-type ordered structures in equiatomic binary alloys, AB, of transition metals correlates with the location of the component elements in the periodic table. One of the components, A, must be to the left and the other component. B, must be to the right of the chromium group. It was pointed out2 that the CsCl structure does not occur in CrFe. is metastable in VFe,3 and is stable in TiFe up to high temperatures, and that this progressively increasing stability may correspond to an increase in electron transfer between the two components and to a correlated decrease of the magnetic moment associated with the iron atoms. Such a decrease was in fact later demonstrated by Nevitt. 4 The binary CsC1-type phases that are formed by a transition metal with either another transition metal or a lanthanide are listed in Table I. Thirty-three of these phases occur at an average electron concentration (average number of electrons outside of closed shells) of e/a = 6. The five phases that are formed at an electron concentration of e/a = 5.5 appear to be borderline cases. This is suggested for instance by the fact that TiRe is bcc. without any long-range order, while TiTc (at the same electron concentration) has a CsCl ordered structure. Six of the undistorted CsCl phases formed by transition metals with one another have an average electron concentration of e/a = 6. 5. Table I. With the exception of Zr4,Rhs4 and Hf48Rh52. the B component in these phases is a first long-period transition metal, i.e.. iron or cobalt. TiNi. with an electron concentration of e/n = 7, appears to undergo transformation at around room temperature to a more complex ordered struct~re.~ The five other CsCl structures known to occur at e/a = 6.5 have either scandium or a lanthanide metal as the A component; the difference in the elec-tronegativity and in the e/a ratio between the component elements is in these phases very large, A(e/a) - 7, and they may be expected to have a rather strong ionic bond component. Presumably the ionic bond stabilizes the CsCl-type structure at the equiatomic composition and it overrides the electron concentration requirement. When the B element is not from the first long period and the A component is not a rare earth metal. equiatomic (AB) alloys of transition metals with an average electron concentration of e/a = 6.5 in most cases have a crystal structure closely related to the CsC1-type, but slightly distorted tetragonally, Table I. This was first found6 for VRu, NbRu. and TaRu. It was noted6 that the undistorted CsCl structure is also stable in the three systems. but at a lower electron concentration, near e/a = 6. In these systems this requires a deviation from equiatomic stoichiometry. which is otherwise ideally suited for the CsC1-type structure. A similar situation was later found in the Ti-Rh7 and the Ti-Ir systems. 8 The Ti-Rh alloys undergo a second distortion and become orthorhombic when the rhodium content becomes somewhat larger than 50 pct. As mentioned previously. TiRe does not have the CsC1-type ordered structure. In order to investigate how close TiRe is to the borderline of stability of the

Jan 1, 1970

By F. W. Luersson

High carbon, low phosphorus steel can now be made from pig iron containing 0.7 pct P or more, in a commercial sized open hearth furnace. No external heat is required for refining, and steel produced is identical with basic open hearth steel. THE classical Thomas process, which requires liquid pig iron containing over 11/2 pct P, yields a product relatively high in phosphorzs and nitrogen. Unless large quantities of recarburizers are added to the refined steel, the product is low in carbon content. Thus, the Thomas process is not well suited to American raw materials, and its product is often not acceptable according to the standards set by American industry. The shortcomings of the Thomas process have long been recognized; with the advent of low cost oxygen, studies were initiated to develop pneumatic processes which would conceivably overcome the inherent objections to Thomas steelmaking. Several such modified basic pneumatic processes have been tested experimentally in commercial trials during the past decade. Probably the most widely publicized of these developments has been the oxygen converter process as practiced in Austrial,2 and in Canada.8 n this process, oxygen is substituted for air and introduced above the liquid bath in order to increase the heat available to the metallurgical system. This avoids the need for high phosphorus raw materials and gives a product low in nitrogen. In this process, the phosphorus content in the final product depends, to a considerable extent, on the total amount of that element in the burden. The rotary furnace4 which is now in commercial operation at Domnarvet Steelworks, Sweden, is producing low phosphorus steel from iron containing 1.8 to 2 pct P. One or more early slags are removed and by this means high carbon steels of open hearth quality can be produced. At this stage of development there is some question as to the applicability of the rotary process to large heats. Another development has been the use of oxygen in combination with either steam6 or carbon dioxide yn the basic bessemer, thus avoiding the introduction of nitrogen gas into the refining system and giving a product relatively low in nitrogen content. A double slag process has been used in Europe7 in conjunction with a regular Thomas converter, by means of which it is possible to obtain a phosphorus content in the product approaching that obtained in basic open hearth steels. The Turbo-Hearth process:' which uses the side blowing principle so as to produce low nitrogen steel from normal American pig irons, has been developed experimentally in this country. A process fitted to U. S. raw materials All of the preceding processes, with the exception of the rotary furnace, suffer from the common shortcoming that they cannot be adapted to the production of high-carbon, low-phosphorus steels unless large proportions of recarburizers are added. In 1947, a program was initiated by the Inland Steel Company1' to study an oxygen steelmaking process specifically designed to utilize the full range of phosphorus analyses encountered in American pig irons and to give finai products with phosphorus and nitrogen contents comparable to steel produced by the basic open hearth process. Some of the objectives of the proposed process are: 1) the process should be able to accommodate molten pig iron conta;rl;rrg up to 1 pct P., 2) sne process should be subject to close control of carbon analysis, 3) the process should be capable of producing high carbon

Jan 1, 1958

By A. M. Gaudin

Activation of sphalerite with silver ions results from the exchange of two Ag ions for one Zn ion in the sphalerite lattice. The exchange is proportional to the logarithm of time and proceeds until either of the reactants is consumed. Sodium cyanide prevents the activation of sphalerite by controlling the ratio [Zn++]/[Ag+]' in solution. If sphalerite has already been activated with excess silver, cyanide ions can deactivate sphalerite provided oxygen is present. SOLUBLE: silver salts are effective activators for flotation of sphalerite with potassium ethyl xan-thate as collector.* Since sphalerite takes up large quantities of silver ions rapidly,2-4 investigation of this system appeared fruitful for studying the mechanism of sphalerite activation and deactivation. Experimental data presented here comprise studies of the mechanism and rates of activation and deactivation of sphalerite with silver and cyanide ions. A 20-lb sample of sphalerite from Ottawa County, Oklahoma, was handpicked, crushed, screened, hy-drosized to produce a 65/100-mesh fraction, and purified by electrostatic separation. The sphalerite was then leached with benzene and absolute ethyl alcohol to remove possible organic contaminants and with 0.1 M NaCN to remove any copper that might have been picked up during screening. The benzene had been specially purified by grinding with fine sphalerite for 12 hr to remove trace impurities that might have had a specific affinity for sphalerite. Ii'inally, the mineral was washed, dried in flowing nitrogen, and stored in large ground-glass stoppered weighing bottles under a small nitrogen pressure. The specific surface of two samples from the 65/100-mesh sphalerite stock, measured by krypton gas adsorption," was 420 cm2 per g. Compared with spheres of the same size and specific gravity, the prepared mineral had five times as much surface. Results of a wet chemical analysis and of a spectro-graphic study are given in Table I. Manganese, nickel, tin, indium, thallium, cobalt, and magnesium were not detected spectrographically. All solutions were preparid by -adding reagent-grade chemicals to conductivitv water. which had been previously saturated with the desired gases. Nitrogen and oxygen were purified, but no attempt was made to purify the laboratory air. Radioanalysis was carried out with prepurified Ag110 tracer obtained from the Oak Ridge National Laboratory. The complete decay scheme for Ae is complex. Its primary emissions are 0.09 and 0.55 mev ß rays and 1.48 mev y raysB and its half-life is believed to be 270 days. In preparation of solutions of AgNO3, the amount of Ag110 added was negligible compared to that of the inert silver. Abstraction of silver from solution by sphalerite was measured after agitation of 1.5 g sphalerite with 20 ml of solution in 50-ml pyrex flasks. All sample preparations were carried out in a dry box so that the proper atmosphere could be maintained. After the mineral had been agitated in the silver nitrate solution, an aliquot of the solution and a sample of the wet mineral were transferred to separate counting vials and weighed. The solid sample was dried and re-weighed and the activity in both vials measured by means of a scintillation counter. With scintillation counting it is possible to count y radiation from the solid and liquid, a distinct advantage over the older ß-ray counting techdiques.4 All measurements of silver activity were made using a sodium-iodide well-type scintillation counter. The amount of zinc in solution was measured colorimet-rically by the monocolor method of Sandell.7 Activation Studies Exchange Between Zinc Ions and Silver Ions During Activation: The increase of Zn++ in solution during activation with Ag+ is shown in Table 11. These data indicate a replacement of one Zn++ by two Ag+. Analysis of a separate two-month agitation test indicated that the replacement of Zn++ by Ag+ was, as expected, in the molar ratio of 2.0.

Jan 1, 1958

T.B.King (Depaytment of Metallurgy, Massachusetts Institute of Technology)— A valuable contribution of the authors is in the factual information which they have been able to gather; this type of information is quite difficult to obtain. In many respects, however, it would have been better if they had not subsequently embarked on a discussion of the chemistry of the converter process. It seems inconceivable that the authors do not refer to the papers of Schuhmann and his associates14 which have set the thermodynamic foundation for the whole copper smelting operation. In addition, a very useful review on the physical chemistry of copper smelting by Ruddle 5 appeared as long ago as 1953. An examination of this literature would have convinced the authors that there is no cause to be surprised at a correlation between the magnetic content and the silica content of converter slags, though they rightly point out that one should distinguish between the total magnetite content of the slag and the amount of magnetite which may be considered to be in solution. It is not true that the lowest melting converter slag is that corresponding to the eutectic between ferrous oxide and silica. The simplest slag system which can be considered is a three-component system, since both ferric and ferrous iron are present. As Schuhmann, Powell, and Michal have shown, there are lower melting compositions than this eutectic in the ternary system. The most unfortunate impression given by this paper is that the driving force for chemical reaction is determined by the heat of reaction: of course the entropy change must be taken into account. It would have been more correct to list, in Table VI, values for free energies of formation. Nor can it be said that the data in Table VI represent the "best available data." They do not corregtond with any of the recent, acknowledged sources. F. E. Lathe and L. Hodnett(author's reply)— We are pleased that Dr. King finds the factual information in our uawer of some interest. Dr. King suggests that it would have been better if our analysis and discussion of the data had been omitted, largely because our list of references is so. incomplete. If he will carefully read our introduction, he will see that the questionnaire was sent out in the hope of obtaining data which would throw light on certain questions relating to the use of converter refractories. We did not attempt (nor would the AIME have published!) a complete review of the literature on copper converting, as Dr. King has apparently assumed, nor indeed a complete analysis of the data submitted, but tried only to find a sound basis for the choice of refractories, taking into consideration common variations in converter practice. We hope our paper indicates that, by raising the silica content of the converter slag and operating at a higher temperature, the normal circulating load of magnetite can be greatly reduced, and the whole reverberatory-converter operation improved to a major degree, with resultant important savings. Under such operating conditions, chrome-magnesite brick may be expected to stand up better than those of straight magnesite. Regardless of the choice as between these brick types, however, we find the cost of converter refractories to be so low in comparison with other converter costs as to justify operation under the more severe conditions suggested. Valuable as are the papers by Schuhmann and associates and the book by Ruddle, we make no apology for omitting reference to them, nor for using heats of reaction without mention of entropy changes or free energies of formation. Our primary object was to interest the practicing copper metallurgist, with whose language we may claim to be fairly familiar; we think it would have been unwise to include the highly theoretical phases of the subject which Dr. King suggests. The interest shown in our preliminary paper presented at the New York meeting in 1956, and the trends in practice whtat we have observed since that time, suggest that we did not wholly miss the target. In conclusion, we sincerely hope that Professors King and Schuhmann will independently review the data obtained in our questionnaire and submit a paper giving their own recommendations as to the choice of refractories and the particular converter operating conditions which will result in the lowest overall cost of copper smelting and converting.

Jan 1, 1960

By D. B. McGilvra

ANY mining operation in the Altiplano of the Andean range presents problems and imposes obligations on the mining company not usually met with in the United States. The Aguilar mine is in a sparsely populated, semiarid region, 50 km by road from the Argentine State railroad station of Tres Cruces, Jujuy, Argentina. In starting an operation in such a location it becomes necessary for the mining company to prepare first for the needs of the required force of workmen and their families. This program calls for houses, schools, hotels, water, electric power, laundries, post offices, police

Jan 1, 1947

By Joseph W. Roe, L. L. HEWES

A GRAPHIC solution of Kutter's formula for the flow of water has been worked out By Dr. L. I. Helves in connection with his course in Graphic Computations, given in the Sheffieild Scientific School, Yale University, which may be of interest to those who deal with canals, flumes and streams.

May 1, 1909

By AIME AIME

THE 1944 Chairman of the Institute of Metals Division might be classed as metallurgically ambidextrous ; he is teacher of theory and practice of both nonferrous and ferrous metallurgy, and he is consultant on brass and on highspred steel. Arthur Phillips was born in Troy N.Y. and attended the New Haven Connecticut high school. Upon graduation he entered Yale, receiving a degree in metallurgy in 1913. At that time Professor Mathewson was starting his plan of '"co-operation with the metal industries in metallographic work at the Hammond Laboratory of the Sheffield Scientific School, Yale University, and Arthur Phillip, became the first co-operative graduate student, working in the laboratory of the Bridgeport Brass Co., and receiving his Masters of Science degree in 1915. A little later, when the department of metallurgy at Yale was expanded. Me. Phillips was appointed assistant professor. In 1928 he was made associate professor, and in 1934 he was promoted to full professor.

Jan 1, 1944

By AIME AIME

COUNCIL.* PRESIDENT OF THE COUNCIL. CHARLES KIRCHHOFF NEW YORK, N. Y. (Term expires February, 1912.) VICE-PRESIDENTS OF THE COUNCIL. BENJAMIN B. LAWRENCE NEW YORK, N. Y. JOSEPH W. RICHARDS SOUTH BETHLEHEM, PA. ALBERT SAUVEUR CAMBRIDGE, MASS. (Term expires February, 1912.)

Feb 1, 1912

By George W. Maynard

IN the biographical notice of Thomas F. Witherbee, published in Bulletin No. 32, August, 1909 (p. xxv), it is said that ". he is believed to have been the first manager in America to use the chemical laboratory for the purpose of controlling the regular running of the blast-furnace." Since the year is not given, I cannot decide as. to Mr. Witherbee's priority; but the statement leads me to contribute a bit of history, showing an early departure from the "rule-of-thumb " blast-furnace work to the employment of a chemist. In the autumn of 1868 the department of mining and metallurgy was established at the Rensselaer Polytechnic Institute. In connection with the assay-department a laboratory was equipped for making general commercial analyses. In January, 1869, Alexander L. Holley, the manager and builder of the Bessemer plant at the Rensselaer Works of John A. Griswold & Co., the first plant erected in the United¬States under the Bessemer patents, began to send me samples of ore, pig-iron, and steel for analysis. Early in 1869 I was regularly retained as chemist for the Rensselaer Works; and while the laboratory was not located at the works, for all practical purposes it was the works-laboratory. The increase in work soon made it necessary for me to secure an assistant. I applied to Prof. Charles F. Chandler, who recommended Dr. August Wendel, who had been an assistant of Rammelsberg and had lately landed in New York. Dr. Wendel remained with me until 1873, when I left Troy. My leaving made necessary the erection of a laboratory at the Rensselaer Works, which, was done under my supervision, and Dr. Wendel was employed as chemist. During the period from 1868 to 1873 my work as consulting mining engineer frequently called me to the Lake Champlain

Nov 1, 1909

By J. C. Martin

A theoretical analysis of steam stimulation is presented for single sands. The analysis includes the effect of steam production and most of the effects of heat conduction. The results show the effects of a number of important variables on the performance of an idealized well under steam stimulation. Calculated responses are presented which indicate the effects of steam production, amount of steam injected, water production, formation thickness and formation damage. Results indicate that steam production can cause large reductions in the heat contained in the formation. This effect can be eliminated by drawdown control. Water production reduces the amount of oil produced during stimulation. The optimum amount of steam to inject depends on economic factors as well as the well response. In many cases, the increased temperature resulting from stimulation reduces oil viscosity near the well sufficiently to overcome the effects of formation damage even if the damage is not removed during steam injection. Calculated responses for thin sands are more favorable than anticipated. INTRODUCTION Little has been published on the theory of steam stimulation'-' despite the interest it has created and the wide variation in well responses. The results of the present analysis provide an insight into steam stimulation, and the methods employed provide a foundation for future work. Analyses presented in Refs. 1 and 2 are very limited and apply to gravity drainage conditions. Ref. 3 contains an analysis similar to the one presented here. The idealized models used and the assumptions made in Ref. 3 are different from those used in this paper. The analysis assumes that after steam injection has heated a small portion of the volume within the radius of drainage of a single uniform sand, a shut-in soaking period is allowed before returning the well to production. The effects of gravity, capillarity, transient pressure and water-sensitive sands are neglected. The injection and soaking times are assumed short compared to the stimulated production time. The initial temperature is assumed uniform; thus, the results apply primarily to first-cycle stimulation. The effects of gas production other than steam are neglected, and the water-oil ratio during production is assumed constant. Steam stimulation involves the simultaneous variation of the temperature, pressure and saturations. General mathematical equations for these variations are complicated and can be very difficult to solve. Simplified equations based on idealized models are used to reduce the mathematics sufficiently to allow approximate solutions to be obtained. DISCUSSION INJECTION An idealized model for which heat conduction is neglected is used to represent the behavior of a well during steam injection. The mathematics for this model is presented in Appendix A which also contains an approximate solution for the behavior of the no-conduction model. SOAKING During soaking, the well is shut in. Only the temperature, pressure and saturation distributions at the end of soaking are needed in the analysis. During soaking the heat is considered to be conducted in a uniform medium from an initially uniformly-heated circular cylinder confined to the producing interval. At the end of soaking the saturations and pressures are assumed to correspond to the cold zone. Analysis of the heat flow during soaking is included in the next section. The radius of the heated cylinder is calculated from the following heat balance (for constant quality steam injection). At the end of the soaking period it is assumed that little or no free gas is present near the well, and that the soaking period has been sufficiently long that the steam zone has had time to expand and the steam has condensed. The condition where there is no soaking is considered in the next section. PRODUCTION In this section, an approximate method is presented for solving the equations of heat and fluid flow associated with the production of oil and water during steam stimulation. Where initial pressure drawdowns are sufficient to cause steam to flow into the wellbore, steam production is assumed to occur within a short initial adjustment period (Appendix B). The production practices followed soon after the well is returned to production can have a large influence on the amount of oil produced during the stimulation cycle. Under most conditions, there is a period of time in which some or all of the water in the heated zone is converted into steam and produced. This flashing of hot water is caused by the pressure in the heated zone dropping below

By S. G. Gibbs

A new method for predicting the behavior of sucker-rod pumping systems is presented. The pumping system is described by a flexible mathematical model which is solved by means of partial diflerence equations with the aid of computers. Polished rod and intermediate-depth dynamometer cards can be calculated for various bottom-hole pump conditions. The technique permits simulation of a wide variety of operating conditions, both normal and abnormal. The data generated with the new technique are useful in refining the criteria for design and operation of sucker-rod systms. INTRODUCTION Sucker-rod pumping systems are used in approximately 90 per cent of artificially lifted wells. In view of this wide application, it behooves the industry to have a fundamental understanding of the sucker-rod pumping process. Oddly enough, our understanding has been rather superficial. This is evidenced by the semi-empirical formulas which have been used as the basis for design and operation of sucker-rod installations. Thoueh- we have realized the limitations of our methods for many years, it has not been computationally feasible to use more refined techniques. With the advent and widespread use of digital computers, it is now possible to handle the mathematical problems associated with sucker-rod pumping. This paper summarizes a computer-oriented method which can provide greater insight into the sucker-rod pumping process. It is hoped that this technique, and techniques which may evolve from it, will prove to be the tool needed by industry to obtain the most efficient use of rod pumping equipment. THE MATHEMATICAL MODEL Prediction of sucker-rod system behavior involves the solution of a boundary value problem. Such a problem includes a differential equation and a set of boundary conditions. For the sucker-rod problem, the wave equation is used, together with boundary conditions which describe the initial stress and velocity of the sucker rods, the motion of the polished rod and the operation of the down-hole pump. Of these items, the wave equation, the polished rod motion condition and the down-hole pump conditions are of primary importance. Discussion of the mathematical model centers about these factors. ROD STRING SIMULATION WITH THE WAVE EQUATION The one-dimensional wave equation with viscous damp- is used in the sucker-rod boundary value problem to simulate the behavior of the rod string. This equation describes the longitudinal vibrations in a long slender rod and, hence, is ideal for the sucker-rod application. Its use incorporates into the mathematical model the phenomenon of force wave reflection, which is an important characteristic of real systems. The viscous damping effect postulated in Eq. 1 yields good solutions, even though nonviscous effects such as coulcomb friction and hysteresis loss in the rod material are present. Fortunately, the nonviscous effects are relatively small, so the viscous damping approximation used in the wave equation is adequate. The coefficient v is a dimensionless damping factor which is found in field measurements to vary over fairly narrow limits. For mathematical convenience the gravity term is omitted in Eq. 1. The effect of gravity on rod load and stretch can be treated separately, as will be noted later. Since Eq. 1 is linear, the legitimacy of this procedure is easy to demonstrate. POLlSHED ROD MOTION SIMULATION The motion of the polished rod is determined by the geometry of the surface pumping unit and the torque-speed characteristics of its prime mover. By determining the motion of the polished rod, we formulate an important boundary condition. From trigonometrical considerations it can be shown that the position of the polished rod vs crank angle 0 is given by (see Fig. 1) These equations are obtained from the general solution of the "four-bar" linkage problem and can be used to describe the kinematics of any modern beam pumping unit.' If prime mover speed variations are disregarded, the angular velocity of the crank is constant, and Eq. 2 can be used to predict the position of the polished rod vs time. However, the constant-speed condition leading to constant crank angular velocity is only approached in practice: hence, it is better to make provisions for prime mover speed variations in the mathematical model. The speed at which the prime mover runs is determined by its torque-speed characteristics and the torque imposed upon it. The torque that the prime mover "feels" is the net torque arising from the polished rod load and the opposing torque from the counterbalance effect. The

By Shiou-Chuan Sun

THIS paper presents the design and operation of a frothmeter capable of measuring the frothing characteristics of pine oils and other frothing reagents. The experimental data show that the froth-ability of pine oil is governed by: 1—rate of aeration, 2—time of aeration, 3—height of liquid column, 4—chemical composition of pine oil, 5—pH value of solution, 6—temperature of solution, and 7—concentration of pine oil in solution. The effect of mineral particles on the behavior of froth also was studied, and the results can be found in a separate paper.' The results also show that the relative froth-abilities of pine oils in the frothmeter generally correlate with those in actual flotation, provided that other factors are kept constant. In addition to pine oils, the other well-established flotation frothers were tested, and the results are included. In this paper, compressed air frothing is the frothing process performed by means of purified compressed air, whereas sucked air frothing is the frothing process accomplished by purified air sucked into the glass cylinder by a vacuum system. The term vacuum frothing denotes that froth was formed by degassing of the air-saturated liquid under a closed vacuum system. Apparatus The frothmeter, shown in Fig. 1, is capable of re-producibly measuring the volume and persistence of froth as well as the volume of air bubbles entrapped in the liquid and is capable of being used for compressed air frothing, sucked air frothing, and vacuum frothing. Fig. la shows that for compressed air frothing, the apparatus consists of an airflow regulating system, 1-3; a purifying and drying system, 4-8; a standardized flowmeter to measure the rate of airflow from zero to 500 cc per sec, 9; and a graduated glass cylinder, 13; equipped with an air regulating stopcock, 10; an air chamber, 11; and a fritted glass disk to produce froth, 12. The fritted glass disk, 5 cm in diam and 0.3 cm thick, has an average pore diameter of 85 to 145 microns. The pyrex glass cylinder has a uniform ID of 5.588 cm and an effective height of 63 cm. The inside cross-sectional area of the glass cylinder was calculated to be 24.53 sq cm, or 3.8 sq in. For sucked air frothing, Fig. lb shows that the apparatus for compressed air frothing is used again, with the following modifications: 1—compressed air and its regulating system, 1-3, are eliminated; and 2—a vacuum system, 16, equipped with a vapor trap, 15, and a vacuum manometer, 17, is added. The vacuum system can be either a water aspirator or a laboratory vacuum pump. Any desired rate of airflow can be drawn into the glass cylinder, 13, by adjusting the opening of the air regulating stopcock, 10. The sucked air stream is cleaned by the purifying and drying system, 4-8, before entering the glass cylinder, 13. When this setup is used for vacuum frothing, the air regulating stopcock is closed. The frothmeter has been used for almost 3 years and has proved to give reproducible results, as illustrated in Table I. With a magnifying glass and suitable illumination, the frothmeter also can be used to study the attachment of air bubbles to coarse mineral particles.' Experimental Procedures Except where otherwise stated, the data presented were established by means of the compressed air method. The volume and persistence of froth were recorded respectively at the end of 4 and 6 min of aeration at a constant rate of airflow of 29.3 cc per sec, which is equivalent to 71.6 cc per sq cm per min, or 462.6 cc per sq in. per min. The aqueous solution for each test, containing 1000 cc of distilled water and 19.2 ± 0.5 mg frothing reagent, was adjusted to a pH of 6.9 0.2. The volume of froth is expressed as cubic centimeter per square centimeter and is equivalent to the height of the froth column (the distance between the bottom and the meniscus of the froth). The volume of froth was obtained by multiplying the height of froth by the cross-sectional area of the glass cylinder, 24.53 sq cm. Before each test, the glass cylinder, 13, was cleaned thoroughly with jets of tap water, ethyl alcohol, tap water, cleaning solution, tap water, and finally distilled water. The cylinder with stopcock,

Jan 1, 1953

By P. B. Baxendell, R. Thomas

Work on the calculation of vertical two-phase flow gradients by Cia. Shell de Venezuela has been based mainly on the "energy-loss" method proposed by Poett-mann and Carpenter in 1952. The "energy-loss-factor" correlation proposed by Poettmann and Carpenter was based on relatively low-rate flow data. This correlation proved inapplicable to high-rate flow conditions. In an attempt to establish a satisfactory correlation for high rates, a series of experiments was carried out at rates up to 5,000 BID in Cia. Shell de Venezuela's La Paz field in Venezuela, using tubing strings fitted with electronic surface-recording pressure elements. As a result of these experiments a correlation between energy-loss factor and mass flow rate was established which is believed to be applicable to a wide range of conduit sizes and crude types at high flow rates (e.g., above 900 BID for 27/8-in. OD tubing). It is anticipated that the resulting gradient calculations will have an accuracy of the order of % 5 per cent. At lower flow rates the energy-loss factor cannot be considered as constant for any mass rate of flow, but varies with the free gas in place and the mixture velocity. No satisfactory correlating parameter was obtained. As a practical compromise for low flow rates, a modification of the curve proposed by Poettmann and Carpenter was used. In practice this was found to give gradient accuracies of approxirnately ± 10 per cent clown to flow rates as low as 300 B/D in 27/8-in. tubing. INTRODUCTION Production operations in Cia. Shell de Venezuela's light- and medium-crude fields are principally concerned with high-rate flowing or gas-lift wells. Under these conditions the analysis of well performance, the selection of production strings and the design of gas-lift installations are vitally dependent on an accurate knowledge of the pressure gradients involved in vertical two-phase flow. Initially, attempts were made to establish the gradients empirically as done by Gilbert,' but the results were not reliable due to scarcity of data over a full range of rates and gas-oil ratios. Several methods of calculation based on energy-balance considerations were attempted, but the computations were cumbersome and the results cliscouraging. In 1952 a paper was published by Poettmann and Carpenter' which proposed a new approach. Their method was also based on an energy-balance equation. but it was original in that no attempt was made to evaluate the various components making up the total energy losses. Instead, they proposed a form of analysis which assumed that all the significant energy losses for mutiphase flow could be correlated in a form similar to that of the Fanning equation for frictional 1osses in single-phase flow. They then derived an empirical relationship linking measurable field data with a factor which, when applied to the standard form of the Fanning equation, would enable the energy losses to be determined. The basic method was applied in Venezuela to the problem of annular flow gradients in the La Paz and Mara fields" This involved establishing a new energy-loss-factor correlation to cover high flow rates and, also, some adaptation of the method to permit mechanized calculation using punch-card machines. The final result was 1 set of gradient curves for La Paz and Mara conditions which proved to be surprisingly accurate. With the encouraging results of the annular flow calculations, several attempts were made to obtain a corresponding set of curves for tubing flow. Here, unfortunately, little progress could be made. The original correlation of Poettmann and Carpenter was based on rather 1imited data derived from low-rate observations in 23/8- and 27/8-in. OD tubing. It did not cover the higher range of production rates, and extrapolation proved unsuccessful. A new correlation covering high flow rates was required, but this proved to be extremely difficult to establish since tubing flow pressure measurements at high rates did not exist—due to the difficulty of running pressure bombs against high-velocity flow. The necessity for reliable tubing flow data increased with the development of the new concessions in Lake Maracaibo, where high-rate tubing flow from depths of 10,500 ft became routine. Thus. it was decided to set up a full-scale test to establish a reliable energy-loss factor for tubing flow conditions. A La. Paz field light-oil producer with a potential of approximately 12,000 B/D on annular flow was chosen. To obtain full pressure gradients, a special tubing string was installed which was equipped with electronic surface-recording pressure measuring devices,

By H. C. Casey, M. B. Panish

The roles of the phase diagram and the diffusion and solubility analyses in the selection of sources for the diffusion of zinc into GaAs are discussed. Isothermal sections of- the phase diagram are described at 1000°, 775°, 744°, and 700°. Below 744° the three solid phases, Zn3As2, GaAs, and ZnAsz, may be in equilibrium. In the composition regions where the three solid phases are in equilibrium, the partial pressures of the components do not vary with composition and there is no liquid phase. A diffusion source composition within that region containing 5, 50, and 45 at. pct Ga, As, and Zn was selected to yield reproducible diffusion profiles without surface damage. The 5/50/45 ternary source Provides a significant improvement in the reproducibility and planarity of- the zinc diffusion as compared with other zinc diffusion sources that result from starting with elemental zinc, dilute solutions of zinc in gallium, or various combinations of zinc and arsenic. Several experimental diffision profiles for the 5/50/45 diffusion source have been determined. The profiles are very steep in the region of the junction and produce devices that exhibit step junction behavior. The junction depth varies as the square root of time. Diffusion at 700°C for 4 hr gives a 5-µ junction depth and a surface concentration of 2.4 x1020 cm-3. The surface concentration is enhanced by a factor of three and the junction depth decreased by a factor of ten as compared with diffusion with a starting source 01- elemental zinc. The differences in dijyusion behavior with these two sources are due primarily to the approximately 10' greater equilibrium As4 partial pressure for the 5/50/45 source. ThE most common technique for the preparation of GaAs p-n junctions for diverse applications ranging from switching diodes to injection lasers has been to diffuse zinc into the solid from the vapor phase. The zinc impurity has generally been provided by introducing into a diffusion ampoule elemental zinc, dilute solutions of zinc in gallium, or various combinations of zinc and arsenic. For the purpose of eliminating surface deterioration, zinc has been diffused into GaAs through SiO2 films1 and from zinc-doped SiO2 films.2 During a study of the injection and recombination mechanisms in GaAs junctions, we found that these diffusion techniques either resulted in nonreproducible diffusion profiles and nonplanar junction interfaces or were extremely inconvenient to use. An unpublished report by Allen and pearson3 emphasized that an understanding of diffusion requires relating the diffusion conditions and system compositions to the ternary phase diagram. Recent studies have made it possible to relate the effect of starting source composition on the GaAs sample surface and diffusion profile reproducibility and, in addition, to estimate the expected surface concentration and zinc diffusivity. These studies include the Ga-As-Zn phase diagram,415 the determination of the incorporation reaction for zinc as a substitutional impurity,6 and the analysis of the concentration dependence of the diffusion coefficient.' In this paper the role of the phase diagram in the selection of the starting diffusion source composition is illustrated, and an application of the diffusion and solubility analyses is used to demonstrate their utility in the estimation of surface concentration and diffusivity for various starting source compositions. Consideration of the isothermal sections of the ternary phase diagram shows that below about 744° C a composition region exists where the three solid phases Zn3As2, GaAs, and ZnAs2 are in equilibrium without a liquid phase. In this region the partial pressures of the components do not vary with composition, and thus the surface concentration and diffusivity depend only on temperature. A diffusion source composition within that region containing 5, 50, and 45 at. pct Ga, As, and Zn was selected. As will be described in detail, this diffusion source yields reproducible diffusion profiles without surface damage, planar junction interfaces, and steep profiles that result in devices with step junction behavior. Only closed diffusion systems are considered. Open or flowing vapor systems require the maintaining of constant vapor pressures in the flowing gas and represent control problems outside the scope of this discussion.

Jan 1, 1969