Search Documents

By Attilio Ligasacchi

The production of magnetite in copper smelting has been the subject of wide investigation in the past.1-9 It is accepted that most of the magnetite is produced in the converters and returned to the re-verberatory furnace with the converter slag. Part of this magnetite is circulated back to the converters in the matte. As a result, if proper fluxing practice is not followed, and temperatures in the converting process are not strictly controlled, an undesirable circulating load of magnetite may occur. This might ultimately lead to saturation in the reverberatory bath with consequent settling of the magnetite at the furnace bottom. This paper contains the cumulative results of a series of investigations on the formation and control of magnetite presently being conducted by the metallurgical department of the Chino Mines Division of Kennecott Copper Corp. THE METHOD USED FOR THE DETERMINATION OF MAGNETITE AND OTHER MAGNETIC MATERIALS Conventional wet chemical magnetite analyses are neither reproducible nor theoretically sound. These analyses are sensitive to time, temperature, and heating rate-all factors which tend to decrease their accuracy and reproducibility. To overcome the unreliability of the above method, H. W. Mossman1 developed an inductance bridge magnetite "meter" which completely replaces the wet chemical analysis with the advantage of quicker and reproducible results. Fig. 1 is a schematic diagram of the meter in use in the Chino metallurgical laboratory. When a pulverized specimen with magnetic properties is placed in coil A of the inductance bridge circuit, the permeability of the core is increased, causing a voltage drop across the coil. The unbalance is amplified to operate the balancing motor which adjusts the movable iron core in coils C and D to bring the bridge back to balance. An indicator arm connected to the movable iron core indicates directly the percent, or weight, of the magnetic material in the specimen. Coil B acts as a reference coil and a zero adjustment. The meter is standardized with a set of various amounts of -100 mesh pure magnetite crystals (99.8 pet Fe3O4) thoroughly mixed in a test tube with -100 mesh quartz sand. These mixtures are combined with inert transoptic powder and heated to 140° to 160°C (melting point of the powder) and subsequently cooled to room temperature, solidifying the whole mixture, in this manner eliminating any possible preferential orientation of the magnetite grains. In order to determine the effect of the particle size" of the magnetite and other magnetic materials on the meter determinations, five size fractions were obtained by wet screening of the 99.8 pet pure magnetite and several reverberatory matte and slag samples. The following size fractions were investigated: -80 + 100; -100 + 150; -150 + 180; -200 + 250; -250 + 300 and -100 mesh. The deviation of the meter readings for the various sizes of each sample never exceeded ±0.3 pet. This deviation is considered well within the accuracy needed for common smelter operation procedures. Further investigation proved also that grinding slag and matte samples in our laboratory disc grinder normally produced a powder constituted of 95 pet -200 + 300 mesh. For practical purposes a - 100 mesh size was used as standard for our studies. It was also found that for rapid determinations tamping of the sample in the test tube was sufficient and gave comparable results with those samples solidified with transoptic powder. For the purpose of this study we define the magnetite content in copper smelter slags and matte as the equilibrium amount of crystalline Fe3O4 existent in the specimen at a given reference temperature, excluding any secondary magnetite precipitated in the specimen during nonequilibrium conditions. MAGNETITE DETERMINATION IN SLAGS Since it is impossible to ascertain the status of equilibrium of a conventional slag sample, due to

Jan 1, 1965

By Roy A. Lindgren

Since the year 1643, when the first blast furnace in America for treating iron ore was built at Saugus, Mass., out of mica schist quarried in the neighboring district, the procurement of a suitable refractory for furnace lining has been a problem of concern to the operators of furnaces. The stacks built of mica schist continued to smelt iron ore until about 1836, when, according to F. H. Norton, the first firebrick were produced1, at Queens Run, Pa. Other writers speak of brick having been molded and burned in Massachusetts about the year 1834. In 1841, Andrew Russell began to produce medium refractory plastic clay brick near East Liverpool, Ohio, that were used for lining blast furnaces1. The well-known Kentucky clay-producing district was not opened up until the year 1871, but since then it has produced a large percentage of the linings for iron blast furnaces. While some strides have been made by the refractories industry during the 100 years that have passed since the first firebrick were produced, it has been only during the last two decades that any real progress has been made towards bettering the product, even though the method of production had improved. Perhaps the fault lies with the user of the brick rather than with the producer, for not sooner demanding a superior product. During the past 15 years the tonnage produced per lining has increased from 500,000 gross tons to 1,000,000 gross tons, and now some furnaces are producing 1,600,000 gross tons and better on a single lining. It is true that enlarged capacity of furnaces and improved practice have accounted for some of this increase in tonnage, nevertheless better quality in firebrick must be given credit for its share. However, we are not yet ready to say that we have reached a maximum life of furnace lining. We believe that a better product can be produced and that the refractories industries of America will, through their extensive research depart-

Jan 1, 1937

By F. H. Vitovec, D. W. Hoeppner

Pusk-pull fatigue tests were conducted on copper tricrystals of 99.988 pct purity to ascertain the role of grain boundaries in the initiation and propagation of fatigue cracks. Significant differences in behavior were found for specimens which possessed different transverse-boundary misorientation. In speciwens with low boundary angles cracks initiated within the transverse boundary, while higher angles led to transcrystalline fatigue failure. It is suggested that at low angular misorientation moving dislocations may interact with dislocations of the boundary or dislocations present in adjacent pains on favorably oriented glide planes, thus initiating a fatigue crack. MANY fatigue studies have been concerned with fatigue-crack initiation within grains and the mechanism causing initiation and propagation1-3 Although the initiation of fatigue cracks in or near grain boundaries of pure metals has been observed and reported in the literature, the mechanism of this phenomenon has received little attention.4"11 EXPERIMENTAL PROGRAM Testing Procedure. To investigate the role of grain boundaries regarding initiation and propagation of fatigue cracks, copper tricrystals were tested in push-pull. Axial-stress tests were used to avoid the stress gradients introduced by stressing of some other nature. Copper was selected as the most suitable test material since extensive work has been done on the formation of fatigue-induced slip in copper. Tricrystal specimens were used to provide a grain boundary geometry with one boundary transverse to the principal stress axis and one or more boundaries nearly coincident with the direction of maximum resolved shear stress. The boundary energy and the slip characteristics in the vicinity of the transverse boundary depend on the relative orientations of grains across the boundary. Testing was done at room temperature, at a frequency of 700 cpm, in an atmosphere of either air or argon. It is known that the incidence of grain boundary cracking increases with increasing temperature and decreasing frequency.4 The fatigue machine applied a uniaxial tension-compression load to the specimen by a flat spring which was actuated by an adjustable eccentric. Use was made of an adjustable head and a load cell which had strain gages mounted at 120-deg intervals around its periphery, thus providing a means for eliminating any detectable superimposed bending moment. A clip gage was mounted between the gripping heads to record the cyclic strain amplitude applied to the specimen. Each specimen's hysteresis characteristics were recorded by supplying the load and strain signals to an oscilloscope. Microstructural changes were observed and recorded with an optical microscope which was mounted on the fatigue machine. Immediately prior to insertion in the machine, each specimen was chemically polished. Extreme care was exercised while inserting the specimen in the test machine to avoid either bending the specimen or introducing a mean load. Each specimen was stressed in the tensile direction first and subsequently the load was reversed. Specimen Preparation. Tricrystal fatigue specimens of 99.988 pct Cu were grown from the melt using a modified Bridgman technique. Ingots about 1 in. high were grown to the configuration shown in Fig. 1. Upon sectioning the resultant ingot, several specimens were provided with an identical shape and grain boundary orientation. Except for plane sectioning and final polishing, this method eliminated machining the specimens. The apparatus used for growing the tricrystals is shown in Fig. 2. A spectrographically pure graphite mold, Fig. 3, was inserted in a vycor tube which was mounted on a vertical zone-refining table. Prior to insertion in the tube the mold was assembled as follows. The insert and graphite rods were fixed in position in the mold. The copper was then placed in the mold and the entire mold assembly was positioned in the vycor tube, Fig. 2. At this point the tube was sealed, evacuated, purged with helium gas, and finally placed under a slightly positive helium pressure. The heating coil was then adjusted to melt the copper charge, after which it was raised at a speed of 4 in. per hr. It is also possible to use

Jan 1, 1964

By R. M. Keeney

Before the outbreak of the war in 1914, the only electric-furnace smelting plant operating on a commercial basis west of the Mississippi River was an electric pig-iron plant in California; rare metal ores were not smelted, concentrates of molybdenum, tungsten, uranium, and vanadium ores being shipped East for reduction. Practically no chrome ore was being mined, and only comparatively small amounts of manganese ore. Today there are two electric smelting plants in California, one in Washington, one in lowa, and four in Colorado; two plants are under construction in California, and one in Montana. Ores of manganese, chromium, molybdenum and tungsten are now smelted in the West, and although uranium and vanadium concentrates are still shipped to the East, they will probably be treated in western plants in a short time. The four electric-furnace plants in Colorado are listed in Table 1. Table 1.—Electric Smelting Plants in Colorado Capacity. Company Location KilowattB PRoductS Tungsten Products Co........... Boulder 800 Ferrotungsten, ferromo- lybdenum. Boulder Tungsten Production Co .. Boulder 400 Ferrotungsten. Ferro Alloy Co.................. Utah Junction 1200 Ferrochrome, ferrotung- sten. Iron Mountain Alloy Co......... Utah Junction 3000 Ferromanganese. Power for all of these plants is supplied by the Colorado Power Co. The plant of the Ferro Alloy Co., at Utah Junction, contains one 750-kw., three-phase furnace, to which power is supplied by three 250-lrw. transformers, connected /Y to give 129 volts on the furnace cables, the transformer ratio being 13,200 to 75 volts. The furnace operates with an actual voltage of 120 volts, and produces about 3 tons of ferrochrome per 24 hr. from 40-per cent. Cr2O3 ore. The furnace consists of a steel shell of circular cross-section, 8 ft. (2.43 m.) in diameter by 7 ft. (2.13 m.) deep, lined with magnesite, and having three vertical carbon electrodes 12 in. (30.48 cm.) in diameter. A 450-kw. furnace, of the same size as the 750-kw. furnace, is also operated on chrome ore. This furnace is three-phase in appearance, but operates electrically as a single-phase furnace. There are three vertical

Jan 1, 1920

By Robert Mehl

THE precipitation of the nitride Fe4N from the solid solution of nitrogen in a Fe and of the phosphide Fe3P from the solid solution of phosphorus in a Fe both lead to the formation of structures that may be easily identified as Widmanstätten figures, within the meaning of the term as used in this series of papers. A complete solution has been obtained for the figure formed by Fe4N. The structure of Fe,P, however, is not fully known and a complete solution of this figure thus cannot be obtained; nevertheless, a study of this figure has furnished some new information. PRECIPITATION OF FE4N FROM ALPHA SOLID SOLUTION OF N IN ALPHA FE Constitution of System Iron-nitrogen A knowledge of the constitution of the whole of the Fe-N system is unnecessary for this work, though the recent studies by Eisenhut and Kaupp1 and by Lehrer2 have clarified most of the moot points; it is sufficient for our purpose to state that a Fe of high purity dissolves 0.42 per cent of nitrogen in solid solution at the eutectoid temperature, 591° C., and that this solubility decreases at first rapidly and then more slowly until at room temperature the solid solubility amounts to approximately 0.1 per cent nitrogen.3

Jan 1, 1934

By I. R. Kramer, H. Shen, S. E. Podlaseck

The tensile and creep properties of aluminum in vacuum have been investigated. It was found that the general effect of a vacuum environment was to reduce the rate of work hardening. Results obtained from creep tests showed that lowering the test pressure from 2 x 10-4 to 10-7 torr produced a decrease in the activation energy as much as 500 cal per mole. An attempt is made to explain these effects in terms of how the surface-stress field of a specimen may be affected by a vacuum environment during plastic deformation. In a previous paper,' the results were reported of some of our early work on the effect of vacuum environment on the stress-strain curve of aluminum single crystals. This effort has been extended to the studies of the tensile and creep properties in vacuum as a function of various metallurgical parameters. Some of the recent surface-effect studies2'3 have revealed that a region of high dislocation density, the so-called "debris" layer, is formed near the surface of the specimen as a result of plastic deformation. This layer acts as a surface barrier to oppose dislocation movement. The strength of this barrier depends on the surface condition such as oxidation or surface removal. Experimental results obtained from surface-removal work2, 3 indicate that such a surface barrier, which may be characterized by a stress field, affects the applied stress and the rate of work hardening. In the case of aluminum, this stress field was found to be directly proportional to the strain and the strain rate; however, the slope of this curve is independent of orientation of the crystal and may be altered by using different surface treatment. Since the escape of dislocations may be retarded by the presence of an oxide on the surface of a specimen, it is expected that a change in the rate of oxidation in vacuum will affect the process of dislocation accumulation in the debris layer. Consequently, it also affects the plastic behavior of a metal. This paper reports the changes in some of the creep and tensile properties of aluminum in vacuum and discussions of the observed effects in terms of the concept of the "debris" layer. EXPERIMENTAL PROCEDURE Tensile Tests. A sketch of the tensile apparatus is given in Fig. 1. This apparatus was contained in a bell jar (14 in. diam by 24 in. high) connected to a 10-in. vacuum system. The load was applied to the specimen by means of a gear train which was driven by a magnetic clutch located outside the vacuum chamber. The specimen, held by a pair of square shoulder grips, was attached to the load cell at the top and the drive gear at the bottom. A displacement transducer was placed between the two grips as shown in the figure to measure the elongation. Signals of elongation and load were amplified and recorded on an X-Y recorder. The system was capable of detecting loads as small as 0.025 lb and displacements of 0.0002 in. Shoulder-grip specimens of aluminum single crystals, having a nominal square cross section of 1/8 by 1/8 in. and a 3-1/2-in, gage length, were grown by the Bridgeman techniques. The original purity was 99.997 pet. The orientations of the crystals investigated are shown in Fig. 4(a). The crystals were annealed at 600°C for 2-1/2 hr, electropolished, and placed in the vacuum test chamber without any delay. All the crystals were deformed at a strain rate of 4 x 10-5 sec-1 at 24o ± 3°C. A study was made to find whether a chemically passive surface film would give the same effect as a vacuum environment. Specimens of aluminum 1100 (0.187 in. diam with a grain size of 0.1 mm)

Jan 1, 1965

By E. C. Tveter, L. A. Roe

The authors present a general outline of the theory and development of heavy-liquid application to mineral processing. Patent literature and processes are reviewed with special emphasis on liquid recovery systems which have been employed or proposed. Advantages and disadvantages of the process are discussed together with the recent developments which have revived interest in this old concentration method. The most important single factor in this resurgence of interest is concluded to be the narrowing gulf between chemical engineering and mineral dressing which has opened the field for new concepts of mineral plant design. A partial summary of patents on heavy liquid separations is included. In spite of the fact that heavy-liquid separation with organic liquids has been used in the laboratory for over 50 years, this process has never graduated to large scale commercial use for any extended period. However, a variation of this process, the sink-float or heavy-media separation process has found wide acceptance and is used to separate minerals from diamonds to gravel. Materials used to increase the specific gravity of the pulps used as heavy media include sand, clay, barite, magnetite, galena, hematite, atomized lead and ferrosilicon. Because of the greater ease of recovery, the magnetic materials, magnetite and ferrosilicon, are the preferred media today. In the U. S. alone, heavy-media iron ore plants with a capacity of over 10 million tons of concentrate per year are in existence. Heavy-media separation involving use of solid, inorganic particles suspended in water rapidly found a wide range of commercial use with the introduction of magnetic media. The minerals engineer is experienced in working with suspensions and quickly learned to develop and control such media at a cost compatible to the type of separation desired. The natural superiority of a heavy liquid with uniform chemical and physical properties has never been questioned since its first use in laboratory mineral separatory procedures. It offers the only method for gravity separation of fine particles of relatively close specific gravity. The most important early attempt to make heavy liquid separations commercial were made by the DuPont Co. which began experimental work on Virginia limonite in 1904. The appended "Partial Summary of Patents" compiled by W.L. O'Connell of The Dow Chemical Co. demonstrates the quantity and sequence of this work. Both inorganic and organic parting liquids were investigated and numerous patents were issued on the use and recovery of these liquids. The work culminated in the heavy liquid plant of the Weston Coal Co. at Shenandoah, Pa. Chemical engineering design of this plant was the responsibility of Francis I. and Hubert I. DuPont. Heavy-liquid separation of coal almost achieved commercial status at this plant which actually processed over 20,000 tons of coal. The liquids used were tetrabromethane, pentachloroethane and trichlo-roethylene, with liquid losses ranging from 8.9 to 12.4 oz per ton of cleaned coal. The reasons for failure of this plant are not clear but probably involved toxicity problems as well as other problems in chemical engineering. Excellent economics were reported. A more recent pilot plant was built in 1954-55 by Norris Goodwin for the Inerto Co. to treat hectorite clay. This plant employed carbon tetrachloride in jigs with liquid recovery by evaporation. Although good separation was achieved, incomplete removal of the CCl, from the clay prevented commercial operation. The only present operations known to the authors employing heavy liquids for gravity separations are limited to the use of calcium chloride in certain coal washers and bromochloromethane in a small batch operation (one ton solids per day) for the separation of beryllium metal particles from slag materials. This is not strictly a minerals beneficiation problem but it has demonstrated the feasibility of such separations. PROBLEMS Critics of early attempts to commercialize heavy-liquid separation of minerals summarize the draw-

Jan 1, 1963

HE purpose of this chapter is to present a general outline of the basic open-hearth process for the benefit of students, practicing open-hearth operators, and metallurgists who wish to review the subject. The discussion is limited to a description of the sequence and purpose of the operations, illustrated by data and examples. Subsequent chapters deal with more detailed and fundamental studies of the subject. The function of the basic open-hearth furnace is to convert various types of ferrous material into finished steel of given composition and quality. In comparison with other steelmaking methods, the basic open-hearth process is one of the most versa- tile, both from the standpoint of the raw materials that can be used and the types of steel that can be produced. The process can be considered as a sequence involving melting, refining, and deoxidation. With some types of liquid-metal charges, the melting stage is not present, but refining and deoxidation in varying degrees are fundamental features of the process. GENERAL DISCUSSION OF BASIC OPEN-HEARTH PROCESS The first consideration in open-hearth practice is the type of charge on which the operation is based. This consists of three components-fluxing, oxidizing, and metallic materials. Fluxes. In the basic process the fluxing material is primarily lime, added in the form of either limestone (CaC03) or burnt lime (CaO). This is supplemented by magnesia (MgO) and lime from the furnace bottom and banks. The function of these fluxes is to form a slag with the silica (SiO2) , manganese oxide (MnO), iron oxides (FeO and Fe2O3), and phosphorus oxide (P2O5) that are produced by the oxidation of varying amounts of the silicon, manganese, iron, and phosphorus in the charge during melting and refining. The slag, composed of the basic fluxes and the neutral and acid oxides, floats on top of the steel bath when the charge is melted. The composition of this slag depends on the amount of each slag-forming element present in the charge and on the extent of oxidation of these elements during the course

Jan 1, 1951

By R. H. McLean

Velocity, kinetic energy and shear in crossflow beneath three-cone jet bits may influence cleaning of the bottom of the borehole and the teeth of the bit. Laboratory investigation shows that each of these parameters is a function of the diameter of the borehole and the product of the volume rate of flow and velocity through the nozzles (QVn). Increasing QV, or decreasing the diameter of the borehole increases each parameter. These functions provide means of predicting the magnitude of each parameter and of scaling the cleaning forces. INTRODUCTION In drilling operations using conventional jet-type rock bits, the impinging jels create an important flow mechanism. Called crossflow, this flow mechanism originates in the impact area of the jets, spreads across the bottom of the hole and supplies the principal source of energy to clean the teeth of the bit and most of the bottom.' Besides providing the means of cleaning cuttings from the bit and the hole bottom, crossflow may also have other, less direct, effects on the rate of penetration. The shear stress generated on the bottom by crossflow will inIluence the thickness and permeability of any filter cake of mud solids or crushed material which forms on the bottom.' These factors may affect the rate of penetration. A previous pulblication introduced some fundamental concepts of crossflow.' Crossllow was shown to occur in a thin layer adjacent to the bottom, and to cover the bottom completely. The maximum velocity in the crossflow above any position on the bottom was found to be directly proporlional to the square root of the product of the volume rate of flow and velocity through the nozzles—the jct QV,,—and inversely proportional to the diameter of the borehole. This information indicated that the effectiveness of crossllow in scavenging the bottom can be improved by maximizing the jet QV. The investigation reported herein amplifies the definition of crossflow. Complete velocity profile data above a representative position on the bottom are analyzed. These data better illustrate control of the capacity of crossflow to scavenge the bottom, and also relate shear stress on the bottom to known, controllable parameters. The conclusion reached in the previous publication that maximization of the jet QV. produces the maximunl cleaning beneath current jet bits is unchanged by these new data; rather, it is strcngthened. Data presented here show that the kinetic energy flux above a representative position on the bottom is maximized by maximizing the jet QV. The shear stress on the bottom will also be shown to be maximized in the same manner. Since the functions relating the jet QV. to velocity, shear stress and kinetic energy also involve the diameter of the borehole, means of equating, or scaling, these quantities in different sizes of boreholes will be illustrated. EXPERIMENTAL EQUIPMENT AND TECHNIQUE JET BIT MODEL Data were recorded from the same laboratory model as used in the aforementioned investigation of the flow around a jet bit.' The model consisted of a 43/4-in. three-cone, jet-type rock bit in a smooth, flat-bottomed borehole constructed of lucite. The bit had a shape and nozzle placement closely resembling larger three-cone jet bits commonly used in field operations. Fig. 1 illustrates the impact area on the bottom of a jet from this bit. Details of the orientation of the jets may be found in the previous publication. TECHNIOU'E OF MEASUREMENT Measurements of the crossflow were made by inserting a very small Pitot tube through the bottom of the simulatetl boreholc. Extreme thinness of the layer of crossflow necessitated accurate measurements of the height of the Pitot tube above the bottom to achieve close definition of the velocity profile. A cathetometer, which could be read to the nearest 0.005 cm, was used to make this measurement.

Jan 1, 1966

By C. E. Taylor, L. V. Fegan

With on-stream electrical-resistance measurement as a basis, Bethlehem Steel developed a system for the completely automatic control of moisture in the feed to pelletizing. This system was later successfully adapted to the sintering process, with automatic control of moisture serving to maintain the air permeability required for effective burnthrough. Control of moisture content is an important factor in the agglomeration processes by which iron ores are converted into either pellets or sinter for use in blast furnace burdens. In both pelletizing and sintering the iron ore feed must be mixed with the proper amount of water so that capillary attraction and surface tension forces can be utilized to cause the ore particles to coalesce and form larger agglomerates. The success or failure of the entire pelletizing or sintering process depends on maintaining the proper moisture. Therefore, continuous automatic control of moisture is one of the conditions for the automation and optimum use of both pelletizing and sintering processes. In pelletizing, variation in moisture can cause ball size to change and is a common cause of cycling which results in a fluctuating feed to induration. Excessive moisture can cause lumps to form at the cutter bar and contributes to the formation of buildup on the green pellet screens. Green pellets with excess moisture deform and reduce bed permeability. In sintering, accurate control of moisture is also important because moisture must be within the limits that give the bed the required permeability for good burnthrough. In response to these problems, a moisture meter, based on the principle of electrical resistance measurement, was developed. This moisture meter has been a key factor in the development of systems for providing uniform feed of concentrates in most of Bethlehem's pelletizing and sintering plants. The automatic moisture control system for pelletizing was successfully plant-tested and resulted in the following: (1) Control of moisture in the green pellets to +0.2%. (2) Up to 50% reduction of bentonite consumption. (3) Control of unscreened green pellets so that 85% of the furnace product at Cornwall was within the size range of -5/8 +3/8 in. (4) Improved quality of product with a tumble test of 94 and a "Q" index of 90 as the result of feeding a consistently uniform feed to a shaft furnace. In sintering, where sinter-mix temperature change and variable feed could cause errors in moisture measurement, the system was designed to maintain the correct moisture under virtually all plant operating conditions. It was successfully plant-tested and is now in use in most of Bethlehem's sintering plants. The system achieves: (1) Control of moisture within the desired limits for optimum production. (2) Continuous measurement and control of moisture even with intermittent feeding of mix, so that attention by the operator is seldom required. The two automatic control systems provide smoother startup, minimize upsets and contribute to the consistent control of process variables in pelletizing and sintering. MOISTURE-CONTROL SYSTEM FOR PELLETIZING Fig. 1 shows the general arrangement of the system for automatic control of moisture in pelletizing.1 In this system a bin level controller, using a single probe which regulates filter speed, maintains a uniform low concentrate level in the bin.' The moisture meter, comprised of an analyzer and control components, continuously measures the moisture in the bed and adds water to adjust to the desired level. The meter is placed in an open-loop (feed-forward) control, i.e., water addition occurs downstream from the probes. Filter vacuum must be set so as to produce a filtercake moisture which is below the optimum for pelletizing. Bentonite is added after the

Jan 1, 1969

By I. R. Kramer

The stress associated with the high-dislocation layer at the surface of deformed aluminum crystals was measured by progressively polishing the specimen and determining the change in the initial flow stress when the specimen was reloaded. This surface stress increased linearly with strain in the Stage II and Stage III region. The depth to which the high-dislocation layer extends was found to be 0.0025 in. The influence of the surface layer on Stage I is discussed and it is shown that Stage I ends when, on the secondary slip system, the difference between the applied stress and the opposing stress due to the surface layer is equal to the critical resolved shear stress. In previous papers'-3 it was shown that the surface exerts a pronounced effect on the plastic-flow characteristics of metals. For fcc metals it was possible to increase the extent and decrease the slope of Stages I and II and decrease the work-hardening coefficient of Stage II by removing the surface of the specimen continuously during the deformation process.2,3 More recently,L it was shown that the activation energy for plastic deformation and the activated volume were also influenced by the surface. These changes were related to the existence of a high concentration of dislocations in the region near the surface of the deformed specimen and it was concluded' that the effective stress, T, acting on a dislocation was T = Ta-Ti-Ts [1] where 7, is the applied shear stress, Tj is the internal stress, and TS is the stress associated with the surface region containing the high concentration of dislocations. From Eq. [I] it is seen that the effective stress, 7, acting on a dislocation is a function of 7,. Therefore 7, becomes important in any description of the work-hardened state. In this paper the value of 7, as a function of strain and its variation with depth are reported for aluminum single crystals. In particular, it will be shown that the end of Stage I occurs when the net stress, 7, on the secondary slip system is equal to the critical resolved shear stress. To measure Ts, single crystals of aluminum were deformed in tension to various strains, whereupon the load was removed and a known amount of metal was removed by electrolytic polishing. The differ- ence between the final flow stress before the specimen was unloaded and the initial flow stress upon reloading is designated as ?Tp and is the decrease in the stress due to the removal of the surface layer in whole or in part. The value for Tp is taken equal to -Ts. EXPERIMENT PROCEDURE The details of the tensile apparatus, method of electrolytic polishing, and crystal preparation were the same as those used previously,1"3 The single crystals were 1/8 by 1/8 in. in cross section with a 3-in. gage length. The initial purity of the aluminum was 99.997 pet. The specimens were held in vacuo at temperatures 50°C below their melting points for 16 hr and furnace-cooled. Just prior to the application of the tensile stress the specimens were electrolytically polished to remove 0.004 in. from the thickness. This latter operation was conducted after the specimen was placed in the tensile apparatus. The tensile tests were conducted at a strain rate, i, of 10"5 sec-' and a temperature of 3°C. The aluminum crystals used for the determination of 7, are designated A1-3-12 and A1-116. The orientation of the various crystals used throughout the investigation is shown in Fig. 1. The ?Tp values for specimens A1-3-12 as a function of Ax were determined mainly at four strain levels, i.e., ? = 0.0217, 0.0381, 0.0593, and 0.075, where Ax is one half of the total reduction of the thickness after polishing and ? is the re-

Jan 1, 1965

By R. S. French

THE purpose of this paper is to present data that have been obtained during the past two years concerning the effects of prior cold-work and temperature and time of anneal upon the recrystallization and grain growth of 70-30 cartridge brass. It was desirable to study certain of these phases so that an accurate picture of the principles of the subject could be shown. The work is not intended to provide explicit annealing data, but to provide material examples that may help to clarify analogous problems concerning this alloy. Specifications often set limitations upon the grain size of annealed materials because of the importance of grain size upon subsequent manufacturing operations. It was, therefore, worth while to study the variables that affect the recrystallized grain size of this material. C. H. Mathewson and A. Phillips,1 W. R. Webster,2 and others have presented annealing characteristic curves of this alloy. A recent typical curve presented by R. S. Pratt3 is shown in Fig. I. The tensile properties and grain sizes are shown as functions of the annealing temperature. Such curves serve as a useful guide in determining the physical properties of annealed material, but as they are made generally under carefully controlled laboratory conditions they do not indicate the performance of metal annealed in a mill muffle, where conditions are different and the amounts of metal require necessarily longer heating times and temperatures. Study of the effect of time and temperature upon grain growth and subsequent recrystallization was made with a coil of metal having the following analysis: 70.04 per cent copper; 0.007 lead; 0.007 iron; 0.00 tin; 0.00 silicon; 0.001 nickel; 0.000 phosphorus; balance zinc. The sample coil was obtained from metal that had been rerolled from hot-rolled mill stock, and was received at 0.228-in. gauge, soft, with a grain size of 0.125 mm. This material was then rolled 43.5 per cent to 0.125-in. gauge and annealed to a 0.053-mm. grain size. In this condition, samples were cut and used in all of the experimental work reported in this paper. TIME The first study was of crystal growth at a constant temperature over a moderate length of time. Material was cut from the stock coil at 0.129-in. size and rolled 75 per cent hard to 0.032-in. gauge. From this piece small samples were cut ½ by 2 in., suitable for grain-size determinations. Twelve samples were placed in an electric furnace uniformly across the width, approximately an inch from the floor, A thermocouple was wired to the center sample, so that the approximate metal temperature could be followed. At various

Jan 1, 1944

By N. Mungan

A study was made of the effects of wettability and interfacial tension on the immiscible displacement of a Iiquid by another Iiquid from porous media. The influence of viscosity ratio was also investigated. Porous media used were polytetrafluoroethylene (TFE) cores prepared by compressing TFE powder under different pressures. lt is shown that displacement of a wetting by a nonwetting Iiquid is always less efficient than the displacement of a nonwetting by a wetting fluid, all other things being equal. In the former case, the recovery efficiency can be increased substantially by either reducing the interfacial tension or increasing the viscosity of the displacing fluid. A qualitative discussion is given on the implications of this work to the recovery of crude oil by water flooding. INTRODUCTION The high cost of oil exploration and new recovery schemes makes it imperative that waterflooding be conducted under conditions favoring most efficient oil recovery. To improve oil recovery by waterflooding, it is essential that the role played by interfacial forces in the entrapment of residual oil be studied and understood. Interfacial phenomena in natural rock, connate water and crude oil systems are very complicated because of the complexity of the natural liquids found in petroleum reservoirs, because of our inability to adequately describe the geometrical structure of the porous media and because of a lack of understanding of physical and chemical interactions between the liquids and surface of the pores. The problem becomes further complicated when one tries to elucidate the role of interfacial phenomena in fluid flow. Numerous studies of the displacement of oil by water under different interfacial tension or wet-tability conditions have been made.lm3 These studies have been performed in silica, alundum or sandstone systems using water and paraffin oil and also some surface active material to control the interfacial tension s and the contact angle 0. Unfortunately, the high energies of various interfaces involved favor adsorption and orientation of the surface active material at the interfaces. Also the surface active material concentration at the interfaces exceeds that in the bulk of the liquid phases. Such surface excess may cause the surfactant distribution, the contact angle and the interfacial tension to differ from their measured static equilibrium values and makes interpretation of the displacement experiments difficult. Furthermore, as changes in s also lead to changes in cos 0, the role played individually by one of these parameters in the displacement becomes obscured by the effect of the other. To circumvent these difficulties, a low surface energy solid and true solutions or pure liquids should be used. Use of a low surface energy solid minimizes adsorption and orientation effects at the solid-liquid interfaces. By controlling s and cos 0 through use of selected pairs of pure liquids or true solutions rather than by surfactants, the adsorption effects at liquid-liquid interfaces are eliminated. In the present study TFE cores were used as the porous media. Liquids used were water, sucrose solutions, paraffin oils and benzyl, n-butyl and isobutyl alcohols. The interfacial tension was varied from 40 to 1.1 dynes/cm by suitably choosing the liquid pair. A surface active material was added to the water-oil system only in the case where interfacial tension of 0.5 dyne/ cm was desired. No precise changes of cos 0 were attempted. However, either the displaced or the displacing liquid could be made the one which preferentially wets the TFE surface. Using sucrose solutions and blends of paraffin oils proved to be a convenient way of changing the viscosity ratio between the displaced and displacing liquids. The present investigation examines the effect of interfacial tension, wettability and viscosity ratio on the immiscible liquid-liquid displacement from porous media. Observed data permit a

Jan 1, 1967

By D. J. Rose, W. M. Armstrong, A. C. D. Chaklader

The wetiability of single crystals of MgO by specimens of vacuum-cast iron was studied using the sessile drop technique in vacuo at 1550ºC. Formation of FeO at the liquid-vapor interface caused the contact angle (6) to decrease from 117 to 65 deg during the first minute. After cooling, all specimens possessed a peculiar annular interfacial deposit of Fe,O,. Within the annulus the interface showed no sign of chemical attack. Chemical reaction occurred where the iron was alloyed with Ti, V, Cr, Nb(Cb), Ta, cmd Zr. Vnriation of 6 with alloy concentration was studied. Although vanadium md chromium improved the wettability of MgO by iron, the effect of Zr, Ti, Nb, and Ta was indeterminable because the 8 derived from sessile drop considerations was that of the metal against a restrictive peripheral volume of liquid oxide wetting the substrate. INTEREST in the nature of metal-ceramic interactions has been stimulated by progressive development of metal-ceramic combinations. One of the more valuable methods used for bond investigation is the sessile drop technique. Recent attempts to improve wettability through solute additions to the drop have revealed that the solute may a) react with the oxide creating new compounds at the solid-liquid interface, b) adsorb at the interface in a monolayer formation, or c) distribute throughout the drop uniformly causing no wettability variation. This work, part m in a series1,2 of investigations of interface reactions between metals and ceramics, embraces a study of the Fe-MgO system. MgO possesses a large negative free energy of formation (-173 kcal per g mole 0, at 1550ºC) compared to FeO at this temperature (-73 kcal per g mole 02). Hence the electronegativity difference between the drop and substrate allows selection from an extensive group of metal solutes with intermediate electronegativity differences that would, in the absence of chemical reactions, be expected to adsorb pre- ferentially at the solid-liquid interface. However, the high oxygen pressure of MgO in vacuo creates an oxidation environment which influences the solute behavior. Recent studies of the Fe-MgO systemJ74 have been restricted by chemical reactions that obscured observations. In view of the current interest in liquid-phase sintering of metal-ceramic combinations under partial oxidizing conditions,5 a more comprehensive study of the Fe-MgO system seemed beneficial. EXPERIMENTAL PROCEDURE 1) Specimen Preparation. Optical-grade MgO single crystals and Ferrovac vacuum-cast iron were used in all experiments. A spectrographic analysis of the iron indicated 0.008 C, 0.05 V, 0.005 Mo, 0.01 Ti, 0.002 S, 0.01 Cr, 0.001 Mg, 0.01 Si, 0.003 A1 (wt pct). The MgO crystals were cleaved along (100) planes into plates approximately 20 by 20 by 1 mm. Following annealing in vactto at 1100°C for 3 hr, surface irregularities were removed by a chemical etch in phosphoric acid at 100°C. The iron rod was machined into small cylinders approximately 0.250 in. in diam and length and these were carefully cleaned in organic and acidic solutions to remove surface impurities. All specimens were weighed to the nearest 0.1 mg. 2) Apparatus. The apparatus described in detail by previous authors1,2 was modified for this work. A Polaroid Land Camera was incorporated into the optical system so that drop measurements at ten-second intervals after melting were possible with ultra-high-speed self-develop ing film. 3) Procedure. The iron cylinders were placed upright on the MgO plates and positioned in the molybdenum susceptor. After furnace assembly the system was pumped down to a vacuum of 1 µ and flushed with H2 at 800°C. The system was again evacuated to 5 x 10-5 mm of mercury and the temperature raised to 1550°C within 2 min. The molten drops were photographed at appropriate time intervals. TWO min after melting the iron vapor pressure caused gas discharge, or corona, within the tube. The specimens were then slowly cooled to room temperature, sectioned and examined metallographi-cally. X-ray identification of interfacial reaction products was attempted by the powder diffraction technique. Alloying elements (Ti, V, Cr, Zr, Ta, Nb) were added to the iron in various concentrations up

Jan 1, 1963

C. M. Bishop (Aircraft Div., A. V. Roe Cnnada Ltd., Toronto)—Reference is made to the coherency hardening of ß due to ß' and also to the hardening of ß due to a coherent precipitation of the a phase. What is meant by the term "ß'?" Would not the coherency hardening of ß during cooling be due to precipitation of a sub-microscopic dispersion of a, as is the case during reheating a ß-quenched alloy? I feel that these two phenomena are essentially the same mechanism, i.e., precipitation of a from unstable or supersaturated ß. From the text, it would appear that Tables III and IV are intended to show the effect of the martensite transformation on the properties of Ti-Mn alloys. Referring to Table 111, the ß-quenched alloys would contain a greater proportion of retained ß than would the equilibrated alloy. Could not the increase in properties be attributed to this, along with possible coherency hardening of the ß? In the same way, the structures referred to in Table IV contain ß from which they could derive their hardness. It would appear to be difficult, if not impossible, to assess the role of the martensite transformation from the data contained in these tables. F. C. Holden, H. R. Ogden, and R. I. Jaffee (authors' reply)—In reply to the first question raised by Mr. Bishop, the term "ß"' designates an apparent retained-p structure which is harder than it is after the most rapid cooling from the ß field. It is thought to be the result of a coherent precipitation of a. The mechanism by which hardening takes place is undoubtedly the same as that involved in the early stages of aging of a re-tained-p structure, and we are in agreement with Mr. Bishop on this point. The term "ß'" to designate the anomalously hard retained-ß structure was agreed upon at a symposium in 1953." The authors do not agree that the quantity of retained ß and/or ß' in the ß-quenched specimens of Table III is necessarily greater than that of the 600°C equilibrated specimens. The ß-quenched specimen has a structure which appears to be completely a' formed by the martensite reaction. It is recognized that there could be some retained ß (or ß') between the a plates if the M, temperature is below room temperature. This is unlikely for the 3.4 pct Mn alloy where the M, is known to be at about 640°C. Therefore, the increase in strength cannot be attributed to the presence of a greater quantity of coherency-hardened ß, but, rather, the increase in strength must be attributed to the transformation products formed, e.g., a' (martensiti-cally formed a). Concerning the data in Table IV, we agree that it is difficult to estimate the role of the martensite reaction, since the 6.4 pct Mn alloy was retained ß in the quenched condition.

Jan 1, 1955

By William Rawles

THIS report is the first of a series planned by The Mineral Inquiry, organized by the American Institute of Mining and Metallurgical Engineers to make factual studies of the world's mineral resources in their political and international relations. The efforts of The Inquiry are directed toward filling certain notable gaps in available information, without duplication of many other investigations touching this field which are being conducted by other organizations, public and private. Much of its work is indirect-suggesting topics for study by other individuals and organizations and aiding in their planning and correlation. The present report is designed to show the nationality of the commercial control of mineral resources, in addition to the geographic distribution of production by nation-in other words, the overlapping of political and commercial control. Detailed accuracy cannot be claimed for the figures presented. For some companies division of ownership by nationality has been very difficult to ascertain, as many companies are reluctant to dis-close the extent of their holdings in other corporations, and the impossi-bility of tracing all stock ownership is obvious. Also there are frequent changes in the distribution of stock control. The percentages given rep¬resent only the participation that has been identified definitely, even though it is known in some cases with fair certainty that larger interests exist, as, for instance, American participation in foreign chromite companies. In bringing together these figures we have had the cordial co-operation of the United States Bureau of Mines, the United States Geological Survey, the Mineral Division of the Bureau of Foreign and Domestic Commerce, the Brookings Institution, and of many specialists connected with the mineral industries covered. The report does not cover the entire list of minerals. Coal, although it is, one of the most important minerals, was omitted because not more than one or two per cent of the world's production comes from mines that are owned by companies foreign to the countries where the coal is produced. Others, such as gold, are omitted because sufficient data have not yet been accumulated. It is hoped that some of these may be covered in subsequent reports. It is not our purpose here to discuss the significance of the figures sub-mitted, but only to point out the general nature of the problems on which they have a bearing.

Jan 1, 1933

By F. N. Ward, E. H. Bailey

Camp and sample-site methods useful for determining about 0.5 to 16 ppm of mercury in soils and rocks have been devised to complement the analytical methods already widely used in geochemical prospecting. In the camp-site procedure the sample solution is obtained by digesting the finely powdered sample with hot nine-molar sulfuric acid and bromine, the latter generated in place. The solution is buffered at a pH of 4 and the mercury is extracted into a relatively small volume of an organic solvent. Estimations are made by comparing the color of the mercuric dithizonate extracted from the sample with that extracted from a standard. In the sample-site method the mercury is volatilized as the iodide and the latter is dissolved in a buffer solution from which the mercury dithizonate is extracted as in the camp-site method. The estimation is made by a similar comparison with standards or by comparing the color of the mercuric dithizonate with artificial standards prepared from Orange II, an azo dye. Except for the establishment of standards, the time required by an experienced analyst seldom exceeds 10 min for a camp-site determination and 5 min for a sample-site determination. Camp-site determinations are useful for commodity studies and intensive exploration programs; sample-site determinations are applicable to extensive and reconnaissance-type geochemical prospecting programs. The results obtained on samples taken along traverses above known mercury ore deposits in the California Coast Range and western Nevada demonstrate the usefulness of both methods in geochemical prospecting. The increasing use of the mercury content of soils, rocks, plants, or water in geochemical prospecting emphasizes the need for analytical methods which are sensitive and at the same time require a minimum of readily available equipment. Sensitive methods based on the absorption of ultraviolet light by mercury vapor are common in the literature31 12. Although none of these methods have been used to analyze the large number of samples required in geochemical prospecting, they could be extremely useful except for the expensive instrumentation usually required to make the methods quantitative Also, such methods are perhaps too sensitive for field use, especially around areas of possible contamination. Other methods involve the distillation of mercury from a relatively large sample and subsequent collection of the vapors in nitric acid6. The mercury is then estimated by titration with a suitable reagent such as thiocyanate. These methods are not trace methods because the actual amount of mercury measured is often as great as 50 to 100 milligrams. Also, such methods are unsuitable for geochemical prospecting because of the amount of equipment necessary and the time required. Still other methods involve colorimetry, and the sensitive reaction between dithizone and mercury (11).. is the basis for a number of reports on the determination of mercury in biological materials 5,7. Moreover, by the judicious use of solvent extraction and suitable complexing agents, the dithiaone method can be made almost specific for mercury. However, none of the dithizone methods have yet been applied to the determination of mercury in soils and rocks. In most colorimetric methods he sample solution

Jan 1, 1961

By John M. Tinsley, Calvin D. Saunders, H. K. van Poollen

In the past few years much con-sideration has been given to the evaluation of the effect of hydraulic fracturing on the productivity of wells. Generally, these studies included the evaluation of fracturing materials, fracture extension and formation damage due to the use of various fracturing fluids. Only little consideration has been given to the characteristics, and in particular the flow capacity, of the fracture itself and its effect on well productivity. This paper presents the results of laboratory investigations pointed toward fie evaluation of the efficiency of various fracrures with special emphasis on the flow capacity of these fractures. Data presented in this paper are the results of both an electrical model study and physical testing. Under consideration are (I) effect of overflush, (2) premature production of well after treatment, (3) "tailing-in" with coarse sand near the end of the treatment, (4) effect of propping agent size and concentration, (5) reduction in effective frac-ture permeability caused by formation caused by formation fines, silt and clays, and (6) effect of various fluids on formation strength and competency. The results of this investigation indicate that the flow capacity of a fracture is affected by any or all of the various parameters mentioned above. The authors believe that a better understanding arid utilization of these factors should result in more efficient formation fracturing. INTRODUCTION Hydraulic fracturing has become almost a standard practice of many companies for stimulating production from old and new wells. Although most companies utilize this service, techniques of application vary widely between companies and areas. Probably too often when a well in an area responds favorably to a particular technique all future wells in the same area are treated in a similar manner. Possibly a modification of the technique would result in a further production increase. Variables, of which many are extremely difficult to evaluate from field results, hamper the selection of procedure changes. Attempts are being made by a number of organizations to analyze statistically treating techniques from production data. This is a very worthy and necessary approach but very possibly laboratory investigations may aid in evaluating some of the variables which tend to affect the results of a fracturing treatment. Some of the factors cannot be studied from practical field experience and only laboratory tests can show the possibilities which might exist. One of the factors which appears to be of major concern today is the flow capacity of the created fracture and how it can be changed. Papers on this general subject have been primarily concerned with the size of propping agent and the extent of fracture. Papers have also been written on the possible permeability damage to formations by fracturing fluids. In addition, it might be possible that another type of flow restriction is prevalent. This would be a restriction of flow through a sand-packed frac- ture caused by foreign materials integrating within the propping agent. This paper presents preliminary data obtained in an attempt to evaluate the effect of some factors affecting flow through sand-packed fractures. No attempt is being made to offer a fracturing technique adaptable to all areas and conditions, but to furnish data tending to show the possible effects which might be caused by variations in procedures and materials. PROPPING AGENT PERMEABLLITIES A hydraulically induced fracture containing sand as a propping agent may theoretically be classified as a packed-sand system. The flow of fluids through such packed systems has been the subject of much research. Although there have been numerous methods proposed for the evaluation of such systems, most writers agree on the general properties affecting their flow capacity or permeability. These properties include porosity, particle size, sphericity and the roughness of the particle. In some methods of evaluation . the particle size and sphericity terms are combined to produce an equation which is a function of the surface area of the particles. In this study the permeability of various fracturing sands was both measured and calculated. The apparatus for the permeability measurements consisted of a 52-in. Lucite tube with a 2.5-in. ID. A screen and drain plug were fitted in the bottom of this tube to retain and hold the sand in place while allowing fluid flow. Two pressure taps consisting of thin, highly perforated

By G. M. Yocom

Blowing acid Bessemer converters with oxygen-steam produces steel of below 0.002 pct N2 content. This method of blowing, combined with a dephosphorizing treatment in the steel ladle, results in low-carbon steels of low nitrogen and low phosphorous (under 0.035 pet) contents, which has physical properties equivalent to open-hearth steels of similar analysis. Using a 50-50 mixture of oxygen and steam, the refinitzg rate is increased 25 pct over blowing with natural air, and scrap charge increased from 3 to 10 pet. Bottom life is normal with proper tuyere area and arrangements, fumes are decreased, yields increased, and hydrogen content is normal. THE acid Bessemer plant at the South Works of Wheeling Steel Corp., consists of two 15-ton bottom blown converters with a monthly capacity of 57,000 N.T. The product of the shop is skelp billets for continuous welded pipe and slabs for ordinary drawing and forming quality sheets. Approximately 50 pct of ingot production is regular Bessemer steel of natural Phos content and the remainder is a dephosphorized grade of steel made by a special treatment of the blown metal as it is poured into the steel ladle. The low Phos grade of steel has certain advantages over the higher Phos grade but since both grades were produced by blowing natural air, the N2 content was in the range of 0.015 pct which limited its application. In 1954 it was decided to explore the possibilities of blowing with a steam-oxygen mixture for the production of steel of both low N2 and low Phos contents. The necessary equipment was installed to operate one converter in this manner and early in 1955 an experimental run of 160 heats was made by blowing with a steam-oxygen blast and excluding natural air entirely. During this period the proper operating techniques were established, such as blast pressures, steam-oxygen mixtures, valves and instrumental control equipment, tuyere arrangement in the bottoms, blowing times and production rates, and a thorough study made of the final steel quality. Also during this experimental period the dephosphorizing practice was improved by the use of a tap hole below the lip of the vessel. This provided a clean separation of the acid converter slag and blown metal which made the dephosphorizing treatment more effective. The results of this experimental run dictated further development of this practice and a second run of 720 heats was made in 1957. The quality features and conversion cost results were in line with expectations and accordingly a 400-ton per day oxygen plant is now being installed. The plant is scheduled for completion in September of this year. This will provide sufficient oxygen to operate both vessels on steam-oxygen blast and delete natural air blowing entirely. The steel will then be below 0.002 pct N2 bar content and the dephosphorized grades will be between 0.015 and 0.040 pct Phos. STEAM-OXYGEN BLOWING The steam for the process is fed to the plant at 220 psig pressure through a 6-in. line. The high-purity oxygen is compressed to 200 psig and conducted through an 8-in. line. The oxygen from the main line is valved down to 100 psig and passed through a steam heated heat exchanger. The heat exchanger is regulated to supply oxygen at 300°F to the steam-oxygen mixing station. It is essential that the incoming oxygen be held at this temperature to avoid condensation of the steam with resulting excessive erosion of the clay tuyeres in the vessel bottom. Oxygen is admitted to the mixing chamber by a 6-in. hydraulically operated valve driven by the ratio control regulator on impulse from the flow of steam. Steam is admitted to the steam-oxygen mixture station through a 2 1/2-in. hydraulically driven valve. The ratio control regulator acts to increase or decrease oxygen input as the steam flow increases or decreases with changing positions of the Blower's control lever. The important point to note here is that steam flow always precedes the oxygen flow as a safety measure. The control valves have sufficient capacity to afford protection should blow pipe trouble develop. A 50-50 mixture for these 15-ton heats demands an oxygen flow of 3800 standard cu ft per min along with 317 lb of steam. The Blower's stations is provided with an indicating blast pressure gage, and indicating steam and oxygen flow meters. Signal and warning lights indicate the valve positions and line pressures. A control room at the real of the Blower's pulpit room houses the ratio control and pressure regulators, as well as the various meter bodies. The hand actuated wheels used to change the conditions are mounted on a panel on the front of the meter control house. The recording steam and oxygen meters used for totalizing and accounting purposes are also mounted on this panel.

Jan 1, 1962

By P. H. Wendland

Calculations are presented for the design of a silicon photodiode in which the incident light beam makes multiple passes between the detector surfaces. Total internal reflection is used for this "light-trapping" effect. By this means, the optical path length can be extended to several millimeters, while the electrode separation remains less than 102 cm, as required for nanosecond response time. Data are presented for a Schottky barrier photodiode constructed on a multiply reflecting silicon base wafer. It is shown that the long-wavelength response is considerably extended in such structures without a corresponding sacrifice in high-speed response. The development of efficient and powerful lasers at 1.06 p has stimulated interest in detectors which operate at this wavelength. In typical silicon photodiodes, for detecting 1.06 p radiation, the requirements for high speed and high sensitivity are mutually exclusive. Since the absorption coefficient is only 25 cm-', a lo-'-cm path length is required to absorb 92 pct of the incident 1.06 p radiation. If the electrode separation is greater than 10 cm, however, the carrier transit time will be greater than 1 nsec. This problem can be solved by allowing the incident light beam to make multiple passes between the electrodes. The optical path length can then be extended to several millimeters, as required for complete absorption, while the electrode separation remains less than 10' cm, as required for nanosecond response time. In a typical photodiode geometry, one ohmic contact and one rectifying contact are formed on the two opposite surfaces of a base wafer, and the wafer thickness determines the electrode separation. The objective of the multiple reflection design is to allow all 1.06 p radiation to enter the detector front surface and to form the back detector surface so that no 1.06 radiation can exit. Total internal reflection at the back detector surface is well-suited for light trapping of 1.06 p radiation because the relatively large dielectric constant of silicon leads to a critical angle of 16.5 deg for total internal reflection. LIGHT TRAPPING It is well-known that, as light passes from one medium such as air into another medium such as glass or silicon, the angle of refraction is always less than the angle of incidence. In the limiting case, where the incident rays approach an angle of 90 deg with the normal, the refracted rays approach a fixed angle +, beyond which no refraction is possible: this is called the critical angle. It follows from Snell's law that where = critical angle, n - index of refraction of air, n' - index of refraction of the medium. Applying the principle of reversibility of light rays, all internal angles of incidence greater than +, will produce total internal reflection and "light trapping". The index of refraction of silicon at 1.06 p is 3.5,' and the critical angle is thus 16.5 deg. Fig. 1 shows these relationships for silicon. This very small critical angle in silicon is significant because all incident angles between 16.5 and 90 deg will produce total internal reflection and "light trapping". This effect can be implemented with a "prismlike" geometry, so that incident light can be introduced into the sample without loss and "trapped". PHOTOSIGNALS A precise knowledge of the absorption coefficient at 1.06 in silicon is of critical importance to the design of fast and efficient silicon photodiodes for 1.06 radiation. Dash and newman2 show a value of 25 cm-l, and our measurements have corroborated this value. Assuming that the collection of photoinduced minority carriers is perfect, the quantum efficiency of a photodiode is dependent only on the absorption coefficient. It then follows from Lambert's law that where QE is the quantum efficiency in pct, a is the absorption coefficient, d is the optical path length, and the reflectivity at the surface is assumed to be completely suppressed by an optical interference layer. Fig. 2 gives the maximum quantum efficiency for 1.06p radiation of a silicon photodiode with optical path length d, using Eq. [2]. The ultimate response time of a fully depleted photodiode to an incident light pulse can be considered to be the arrival times of all photoinduced carriers at the contacts, i.e., the minority carriers at the junction interface and the majority carriers at the oppo-

Jan 1, 1968