Search Documents

By D. Muraoka

Introduction Soda ash, anhydrous sodium carbonate, is produced from underground trona deposits occurring in the Green River Basin of southwestern Wyoming. Stauffer Chemical Co. of Wyoming, a jointly owned subsidiary of Stauffer Chemical Co. and Rocky Mountain Energy Co., mines the trona and processes it into high-purity soda ash. Stauffer Wyoming has been in operation since 1962 with a current production capacity of about 1.8 Mt (2 million st) per year of soda ash. The refinery has five parallel operating trains using the monohydrate process which is a series of chemical processing unit operations described below. Screening and Crushing The mined trona is brought to the surface in two counterbalancing ore skips. The ore is initially screened with refinery feed (less than peagravel size material) entering the process. Oversize is conveyed to an outdoor stockpile. Ore is recycled from under the stockpile via cone feeders as the refinery needs require. The ore is conveyed to hammer type crushers and screened. The refinery feed enters the process and the oversize is recycled to the crushers. Calcining The first step in refining trona to soda ash is calcining the trona ore. This is accomplished in concurrent flow, rotary kilns that are direct fired with natural gas. [2 (Na CO NaHCO 2H O) Heat 3Na CO + 5H20 + CO 2 3~3~ 2 ~~2 32] The trona ore is fed to the kiln at the burner end and is intimately mixed with the hot exhaust gases from the burner. Direct contact with the flame is avoided. The ore moves down the kiln by gravity and rotation aided by the kiln's internal lifters. The exhaust gases are cleaned to 99 + % particulate removal and discharged to the atmosphere. Dissolving the Calcined Trona The calcined trona is conveyed from the kiln discharge to a rotary drum dissolver where it is mixed with water and weak solution (unsaturated solution of calcined trona in water). The calcined trona is soluble whereas the impurities, primarily shale, are insoluble. The discharge from the dissolver is a slurry containing saturated solution (about 30% dissolved calcined trona in water), some undissolved calcined trona, insoluble impurities, and organic impurities. The dissolver discharge enters a rake classifier that makes the first solid/liquid separation in the process by removing the heavy undissolved material from the saturated solution. This undissolved material is conveyed to a second rotary drum dissolver and mixed with water to dissolve any remaining calcined trona. The secondary dissolver discharges to a second rake classifier that separates the solids from the liquid. The solids are now primarily insoluble impurities with very little unrecovered calcined trona. These wastes are conveyed to evaporation ponds for disposal. The liquid from the secondary rake is recycled to the primary dissolvers' weak solution system. Returning to the primary rake classifier, the saturated solution which contains suspended insoluble muds is pumped to a thickener. Flocculant is added to aid settling of the suspended muds. The flocculated muds are collected at the bottom of the thickener and pumped to vacuum cloth filters. The adhering saturated solution is separated from the muds that remain on the filter cloth. The recovered solution is returned to the dissolvers' weak solution system. The muds are reslurried and pumped to the evaporation ponds for disposal. Saturated solution, containing some suspended insoluble muds along with the organic impurities, overflows the thickener. Filtering the Saturated Solution The saturated solution overflowing the thickener enters a holding tank where it is pumped to a bank of pressure leaf filters. The suspended muds are captured on the filter leaves while the cleaned saturated solution is pumped to polishing filters. The polishing filters are also leaf pressure type, intended to remove any suspended solids that may pass the primary filters. Organic impurities are removed with activated carbon. The purified solution is now free of insoluble muds and has most of the organic impurities removed. Crystallizing and Drying the Monohydrate The purified saturated solution is pumped to evaporative crystallizers. The solution is circulated from the crystallizer vessel through a heat exchanger. Indirect contact with steam in the heat exchanger causes the saturated solution to boil, thus crystallizing the soda ash in its monohydrate form. Slurry crystal density is maintained in the crystallizer by simulta¬neous feed of fresh saturated solution and drawoff of the slurry.

Jan 1, 1986

By A. D. Trippel

The Cripple Creek Mining district is located in Teller County, Colorado, 30 km west of Colorado Springs and 9 km southwest of Pikes Peak (Fig. 1). Since its discovery in 1891, the district has produced almost 21 million tr oz of gold (Gott et al., 1967), most of which has been from high-grade vein deposits. Globe Hill is located in the northern part of the district (Fig. 2). Production from its summit has yielded over 100,000 tr oz of gold. Early production was concentrated along high-grade veins and fault structures, but emphasis was soon placed on the extensive, low-grade (0.0X oz/t), breccia-hosted mineralization. From 1977 to 1981 Newport Minerals, Inc., a subsidiary of Gold Resources, Inc., mined 680,000 t of low-grade gold ore from the Globe Hill pit (Fig. 3) and production is expected to resume in 1986. These strongly-altered and -oxidized breccias are amenable to bulk-mining methods, and thus similar deposits are currently being sought elsewhere in the district. Numerous publications and reports discuss the breccia-hosted mineralization at Globe Hill. They include Cross and Penrose (1895); Argall (1905, 1908); Lindgren and Ransome (1906); Keener (1962); Peters (1982); Thompson, Trippel, and Dwelley (1985); and Trippel (1985). This preprint is a summary of the recently-completed research documented by Trippel (1985).

Jan 1, 1986

By Keith L. Harris

Buoyed primarily by growing consumer spending, the economy rebounded from a weak second quarter. According to preliminary estimates released by the Department of Commerce, real gross national product (GNP) grew $2 1.9 billion during the third quarter, producing a seasonally adjusted annual growth rate of 2.4%, compared with a rate of 0.6% in the second quarter of 1986 and 4.1% for the third quarter of 1985 (table 1). Third-quarter real personal consumption expenditures and Government purchases of goods and services rose $42 billion and $7 billion, respectively, but those gains were partially offset by the continuing decline of gross private domestic investment, off $16 billion, and the growing trade deficit, up $11 billion. Three quarters of the growth in consumer spending occurred in the durable goods sector as consumers responded to discounted financing offered by automobile manufacturers attempting to lower inventories of 1986 vehicles, Increased purchases of other durable goods, such as furniture and home furnishings, also contributed to the strength of the durable goods sector. The success of the automobile inventory reduction program as well as weak factory output led to a $20 billion decline in inventory growth. Investment for housing and producers' durable equipment expanded, however, and mitigated some of the impact of slower inventory depending on gross private domestic investment, Capital spending for nonresidential structures declined for the third consecutive quarter. The declining capital spending and the growing trade deficit continued to be significant and interrelated problems for the economy. Real capital spending for new plants and equipment during 1986 is expected to decline 2.5% after rising 8.7% in 1985 and 15.8% in 1984 (table 2). It is likely that capital expenditures bottomed out in the second quarter and that they will expand in the third and fourth quarters. The first-half decline was partially attributed to uncertainty arising from the pending tax- overhaul bill. Furthermore, there is little impetus for corporations to invest in new capacity since factories are operating at less than 80% of capacity (table 3) largely because of the continuing growth of imports. The growth rate of the trade deficit dropped sharply in the third quarter to less than half that of the previous quarter. Low oil prices, declining auto- mobile and steel imports, and increased exports of agricultural goods contributed to this decline. The continued low growth of the economy caused the Federal Reserve System to reduce the discount rate for the fourth time this year (in late August), drop ping it to 5,596, the lowest level in about 9 years. Inflation, as measured by the fixed- weighted GNP price Index, continues to remain low, but rose from the revised second-quarter level of 1.7% to 2.596, mostly because of rising food costs.

Jan 1, 1986

By R. G. Burdick

The problems presented by old mine workings affect both present day mining and land development. An automated method of locating these old mines from the surface using electrical resistivity techniques was de-veloped earlier under a Bureau of Mines contract. Subsequent Bureau research has refined the techniques and expanded the area of study to a variety of geologic provinces. During this research, six mining areas in the United States were in-vestigated with the Bureau's automated resistivity method. This report describes the mining areas involved and the results of the resistivity investigations, which showed a high rate of success in detecting old mines. It also describes the field measurement techniques and data analysis procedures.

Jan 1, 1986

By Ronald H. Church

The Bureau of Mines tested & ground penetrating radar (GPR) system in the central Florida phosphate district. to determine the feasibility of utilizing GPR technology for subsurface cavity detection. The test area is located in karst topography where sinkhole development is prevalent. State regulations require subsurface drilling of damsites to identify underground anomalies; however, this method is not totally accurate in locating subsurface cavities that could lead to sinkhole development. An electromagnetic method of identifying anoma-lous subsurface conditions could reduce or eliminate the possibility of dam failure due to sinkhole development. The Bureau devised a GPR system that successfully penetrated over 50 ft of overburden and gave recognizable radar return signals characteristic of a rock-cavity interface.

Jan 1, 1986

By Y. Y. Liu, H. M. Domanus, W. T. Sha

"Casting is fundamental to manufacturing of many types of equipment and products. Although casting is a very old technology that has been in existence for hundreds of years, it remains a highly empirical technology, and production of new castings requires an expensive and time-consuming trial-and-error approach.In recent years, mathematical modeling of casting has received increasing attention; however, a majority of the modeling work has been in the area of heat transfer and solidification. Very little work has been done in modeling fluid flow of the liquid melt. This paper presents a model of fluid flow coupled with heat transfer of a liquid melt for casting processes. The model to be described in this paper is an extension of the COMMIX code and is capable of handling castings with any shape, size, and material. A feature of this model is the ability to track the liquid/gas interface and liquid/solid interface. The flow of liquid melt through the sprue and runners and into the mold cavity is calculated as well as three-dimensional temperature and velocity distributions of the liquid melt throughout the casting process.IntroductionBower et al. (1), in their 1969 review of the evolution of some of the casting techniques, stated that:""The evolution of technology requires alternatives to be available so that choices can be made that lead to changes. The role of solidification science and other disciplines related to casting technology is to develop a better understanding of process and product, so that a wider range of alternatives is possible.Alternatives, whether from science or from invention, must lead to improvements in casting ""utility."" Utility, or economic advantage, must include consideration of total system costs, and final product value. The cost of materials, melt processing and finishing should be optimized with casting costs, since all are part of the production system. It is the favorable relationship between total processing costs and product value that makes a change in casting technology viable, but the best change is not possible unless technology has provided an understanding."""

Jan 1, 1986

Provide sound battery safety guidelines for mine workers who maintain mobile surface mining equipment. Approach Battery-related accidents account for about 5 percent of all nonfatal disabling injuries in coal and metal nonmetal surface mines. These accidents occur despite the fact that appropriate battery safety technology (maintenance-free batteries, cable removers, fault detectors, hot-jumping systems, etc.) exists. Mine operators are not always aware of the safety measures that could be taken because this safety information is widely scattered in numerous unrelated technical reports, vendor brochures, and journal articles. A literature search of these reports and a review of relevant industry standards, handbooks, and references on battery safety, followed by discussions with mining company and MSHA personnel, provided the Bureau of Mines with the background necessary to develop procedures for proper and safe battery usage. These procedures were condensed and written in nontechnical language to serve as guidelines in assuring battery safety during maintenance of mining equipment.

Jan 1, 1986

By S. P. Marsh, M. E. Glicksman

"The effect of capillarity-driven diffusion on the recalescence of mushy zones is studied in pure materials. The solid-liquid mixture is formed by rapid partial solidification of an initially undercooled melt, and the ratio of the two phases is subsequently maintained constant by an externally imposed adiabatic constraint. A novel technique has been developed that permits tracking of the kinetics of the diffusion-controlled evolution process in situ using precision temperature measurements averaged on a length scale larger than that of the microstructure.The mean ""temperature of a mushy zone is slightly below the bulk equilibrium melting point, and is related to a characteristic length scale of the microstructure through an appropriate statistical average of the GibbsThomson equation. Theoretical considerations indicate that the diffusion controlled recalescence will progress as t- l / 3, where t is time. This behavior is experimentally verified for both succinonitrile (SCN) and ethylene carbonate (ETC), which have markedly different morphologies. The kinetic rate constant is observed to increase monotonically with increasing volume fraction of solid in the two-phase mixtures, although a firm theoretical foundation for this result has not been developed.INTRODUCTIONA dispersed multiphase system seeking equilibrium will generally contain variations in the local .free energy density. These density variations arise from such factors as variations in the bulk free energies of the phases and the distribution of interfacial energies. In this study, we examine the kinetics of diffusion-controlled recalescence in 1 solid-liquid mixture of a pure substance at its melting point. This type of two-phase system is particularly convenient to analyze because of the well-characterized physical properties of pure phases (crystals and their melts) and their interfacial energies, and the absence of any solutal effects."

Jan 1, 1986

By H. N. Kalia, J. B. Gresham

Several attempts have been made in the past 60 years to commercialize the vast oil shale deposits of the US. However, the industry has yet to become economically viable. To date, it has not been possible to produce oil from oil shales at a competitive price compared with oil produced by conventional methods. The development work, nevertheless, has been fruitful in furthering the art of thick bed mining and technology of in situ retorting. The in situ method has the potential of recovering oil from thick beds, and the resource recovery potential is much greater than that from underground mining and surface retorting. In situ methods require that the shales/marlstones be fragmented in place to induce adequate porosity and permeability to allow in situ heating of the fragmented shales. This can be accomplished by using explosives. However, it has been found that the existing blasting systems, though very efficient, have some drawbacks. The explosives become insensitive when shot in multiplevertical decks. They can also become insensitive when spaced horizontally at close proximity to other charges. Also, since the blasting is done in a very limited free face, the delay time that can be allowed between charges is also quite limited. Consequently, it is not always possible to routinely predict the performance of the blasting system. Furthermore, since it is not possible to undertake secondary blasting, it is imperative that the blast design function properly. Recent successful retorting by Occidental Petroleum Co. Of Retorts 7 and 8 suggests that some of these problems may have been overcome. This paper addresses some aspects of in situ retorting and problems that have impeded the development of a viable oil shale industry in the US.

Jan 1, 1986

By C. M. Paiencia

The Bureau of Mines studied the potential availability of primary antimony (Sb) from demonstrated resources in 21 mines and deposits in market economy countries (MEC's). Twelve of these properties were evaluated as producers and nine as non- producers. The 21 studied properties contain 499,600 mt Sb as demonstrated resources, of which 304,000 mt is recoverable as antimony metal; another 973,500 mt Sb is contained in inferred resources. Recoverable antimony from demonstrated U.S. deposits is only 10,700 mt, just 3.5 pct of the MEC total. Domestic production of primary antimony from operating mines has averaged only 561 mt/yr over the 10-yr period 1975-84. This represents just over 5 pct of total U.S. consumption of primary antimony, which averaged 11,081 mt/yr during the same period. The United States will thus continue to rely on imported ores andlor concentrates, antimony oxides, and metal to satisfy domestic industrial requirements. At production costs up to $1.15/lb Sb, 110,000 mt of refined antimony is potentially available from 10 of the 21 studied properties. At $2/lb, 294,000 mt Sb is available, and at $3/lb, the total rises to 304,000 mt Sb available from all evaluated producers of refined antimony.

Jan 1, 1986

By C. A. Holley

The generation of coal fines is a growing problem which has serious environmental and resource recovery Impact. Several methods of agglomeration are studied including the production of briquettes for the consumer market. Simplified process flow diagrams, "order of magnitude" plant capital cost and operating costs are presented.

Jan 1, 1986

It is a pleasure to submit to the National Critical Materials Council (Council) some of my thoughts concerning the important tasks which lie before you. It long has been my belief that America's economic health and national security may depend to a substantial degree on whether we have available to us an adequate supply of critical and strategic materials at reasonable cost. To the extent the Council can contribute to the identification and implementation of wise public policies which facilitate the achievement of that goal, we and our children should be grateful. Even before assuming office, President Reagan warned of the danger of undue dependence on imported materials essential to the Nation's defense and economic well-being. After his election, he appointed a transition task force on strategic and critical materials, and he later directed the Cabinet Council on Natural Resources and Environment to prepare .a National Materials and Minerals Program Plan and Report to Congress, which he forwarded to the Congress in April 1982. In 1984, my predecessor, Secretary Clark, established the National Strategic Materials and Minerals Program Advisory Committee, popularly known as the "Mott Committee" because of its distinguished chairman, retired Admiral William C. Mott.

Jan 1, 1986

By Chan. C. L., Zehr. R., M. M. Chen, J. Mazumder

"Two three-dimensional models of a surface tension gradient driven flow during laser surface heating are developed. The first model is based on a perturbation solution. The basic solution corresponds to the stationary axisynwnetric case and perturbation is based on a small scanning velocity. The advantage of seeking a perturbation solution, as it turns out, is that the three-dimensional flow is modeled by two sets of two-dimensional equations which are presumably much more tractable than the original three-dimensional equations. Numerical solutions are obtained and discussed. The second model is a full three-dimensional numerical solution of the Navier-Stokes equations, using a point-by-point partially vectorized iteration scheme. Surface shape is also determined in a self-consistent manner. The effect of the presence of convection on pool geometry, cooling rate and solute redistribution is presented and discussed.IntroductionThe existence of fluid flow and convection and their effects were first reported in the fifties (1,2). The role of convection in controlling pool geometry such as pool shape, undercut and ripples, and defects such as variable penetration, porosity and lack of fusion and weld properties such as weld homogeneity and bead formation is studied by a number of researchers and is reviewed by the authors in a recent paper (3).Recently, surface tension gradient driven flow has been identified as one of the forces responsible for convection within the molten pool (4,5). It is postulated that surface tension gradient driven flow is responsible for the surface ripple formation in the weld pool (6-8). Oreper, et al. (9) developed a two-dimensional convection model. Buoyancy, electromagnetic and surface tension gradient forces were considered. The formulation was based on a specified pool profile. It was found that the surface tension gradient is the dominant force in most cases. However, the solid-liquid interface is not known a priori. In fact, the solid-liquid interface, i.e. the pool shape, is a piece of information that the model should be able to predict. A two-dimensional transient self-consistent (i.e. the solid-liquid interface is solved as part of the solution] model for laser surface melted pool was first developed by Chan, Mazumder, and Chen (3)). This model predicted a non-uniform cooling rate, hence non-uniform microstructure. Depending on the Prandtl number (i.e. material) and surface tension number, the shape of the molten region as well as the cooling rate changes. It was also predicted that the recirculating velocity is of one or two order of magnitude higher than that of the scanning velocity. This model provides the detail of the flow field and heat transfer on a plane perpendicular to the scanning direction. However, the front-to-back motion and its effect on heat transfer also play important roles in the physical process, especially for the fact that the recirculating velocity is so much higher than the scanning velocity. Such a motion cannot be obtained from two-dimensional models."

Jan 1, 1986

By G. A. Smyres

A GaC12--B2 l-eachtng procedure and meEal-recovery method for treating an arsenical Cu-Co sulfide concentrate was developed by the Bureau of Mines as an alternative to smelting, which is no longer considered an environmentally acceptable method for treating arsenic-containing ores. The procedure involved leaching in a CaC12 solution with CaC03 for 96 h at 26 pct solids (concentrate and CaC03), 1150 C, and a pressure of 50 psig (345 kPa) with oxygen. More than 98 pct of the cobalt and 97 pct of the copper were extracted, and 99.98 pct of the arsenic and 99.9 pct of the iron remained in the leached residue. A filtration rate of 2.9 gal/(ft2.h) [118 L/(m2.h)] and a cake formation rate of 36 Ib/(ft2.h) [172 kg/(m2.h)] were obtained by filtering reacted pulp at 1000 C. More than 99 pct of the copper in the pregnant solution was extracted as CuCl2 by golvent extraction with a new copper extractant. Copper was stripped from the loaded organic with a sulfate-chloride solution, which was used as the anolyte for electrowinning of copper in a membrane cell. Copper plate produced was more than 99.99 pct pure and requi.red 0.81 (kW'h)/lb Cu [1.78 (kW·h)/kg Cu]. Cobalt sulfide was precipitated at pH 3 with Ca(HS)2.

Jan 1, 1986

By K. Pagalthivarthi, P. V. Desai

"During solidification of alloys, many inter-related physical processes, such as melt convection, mass diffusion, heat transport, macrosegregation and melt-undercooling occur. Several research workers have studied one or more of these effects separately. To answer important questions on solidification behavior, an overall approach interrelating the processes is necessaryAs a first step toward accomplishing the overall modeling, the freezing of a binary liquid metal is examined. The various stages during freezing of the alloy are critically described to build a continuum model for the entire process. Diffusive mixture theory is used to write the conservation equations for the liquid and the mushy zones. The concept of local thermodynamic equilibrium in the presence of curvature effects is used to restrict the number of degrees of freedom within the mushy zone. Constitutive equations relating flux quantities and affinities, the governing differential equations and boundary conditions are discussed briefly for the freezing process.IntroductionA solidification study encompasses such facets as solifidication time, microstructure, macrosegregation of the solute, shrinkage effects and soundness of the casting. Influence of one or more of the coupled heat and mass transfer phenomena such as local solute redistribution, thermosolutal convection, melt undercooling at dendrite tips and heat dissipation to the ambient have all been previously examined [1-3J. Several studies have examined the individual facets of alloy solidification as isolated processses. In fact, the transport processes occur simultaneously and a comprehensive model should include their inter-relationships. In particular, some of the existing studies of the mushy zone [4-6J do not emphasize thermodynamic completeness. This work represents an attempt to develop a thermodynamically consistent model of the entire solidification process on physical grounds."

Jan 1, 1986

Spiral separators have been used to treat heavy mineralsands in Australia for some 35 years. During the last five years the logical extension of this technology has been torecover cassiterite from alluvial deposits. Plants incorporating spirals are already operating in Indonesian andAustralian tin mines. A more fundamental understanding of the design of spiral separators has recently led to their use in other mineral treatment plants.In particular, the treatment of alluvial gold and fine coal has been most successful. The new generation of spirals that has been developed to treat these minerals is more easily operated than other competing processes and is therefore ideal for new mine situations.

Jan 1, 1986

By J. S. Roti

Calcium-based scale precipitation is a common problem in mill reclaim waters. An accumulation of these adherent deposits can result in a rapid reduction in water flow, loss of pump efficiency and eventual interruption in ore production. Heap leach applications can also experience the added problems of plugged sprinkler heads, blinded leach pads and fouled carbon columns when scale control is not maintained. The primary cause of scaling in precious metal recovery is the relative insolubility of calcium carbonate. The presence of the calcium ion and bicarbonate alkalinity is common in almost every water source. If the temperature or pH of the water is allowed to increase, the bicarbonate breaks down to carbonate and combines with calcium by the following reaction: Ca +2 +2HCO3 ? CaCO3 + H2 + CO2

Jan 1, 1986

By Michael Robert Hodson

I. Introduction The free market cash price of aluminium in the third quarter of 1985 is approximately $1000/tonne, well below the short as well as the long run marginal costs of production. The current disparity between marginal costs and price, as well as the pattern of prices since their high level in 1980, has led many economists to question whether prices are governed by the 'fundamentals'. Prices, some would argue, are the result of speculation; and speculation is irrational, little influenced by the fundamentals of the industry. The object of this paper is first to examine the relation between the fundamentals and price. A number of simple models are formulated and some empirical evidence is given to show that prices are not very different from costs. There are good reasons why costs and prices have fallen since 1980. Second, the connection is drawn between the fundamentals and speculation. To do this the concept of 'rational expectations' is defined and applied to a model of the aluminium industry. Then speculation is introduced into the model and the characteristics of the solution are examined. The object is to discover the new factors that determine prices and the basic change in the nature of the solution that is caused by speculation. 2. Prices and the Simple Supply-Demand Model Consider a simple model of the aluminium industry. Demand, D, is determined by economic activity, A, given exogenously, and inversely by the price, P. Supply, Q, is determined by the inverse of the rnarginal cost curve, and so varies inversely with exogenous factor costs, C, and directly with price. Price is determined by the equality of supply and demand. Given the equilibrium level of demand, price in this model is equal to the rnarginal cost of producing the required level of demand, P = MC (Dl. Algebraically, [ ] How well does this simple model explain the change in prices during the last five years? Table I shows the LME cash price and the estimated marginal cost of producing consumption during the second quarters of 1980 through 1985. The figures in the table show that price and the marginal cost of supplying demand both fell between 1980 and 1982, rose in 1983 and by 1985 had fallen again. Only in 1984 is the direction of change different. And during the 1980 to 1985 period, marginal costs fell by approximately $560/tonne while the LME cash price fell by $678/tonne. There is a correspondence between the fundamentals of cost and price over most of the individual years and the period as a whole.

Jan 1, 1986

By R. E. Sundell, S. M. Correa

"A mathematical model is developed for the fluid flow in tungsten-inert gas weld pools. The fluid is taken to be incompressible, viscous, electrically conducting, and confined within a pool of specified shape. The flow is driven by surface-tension gradient (thermocapillary), electromagnetic, and buoyancy forces generally in this order of importance. Analysis indicates that the electromagnetic field may be calculated a priori and so decoupled from the nonlinear fluid mechanics and heat-transfer problem. The resulting set of equations is solved numerically by a line relaxation process. Computational errors caused by truncation in finite-difference schemes and by inadequate grid distributions are considered. The model is used to study several cases of differing heat and current input to the pool with and without surfactant-induced thermocapillary forces. The principal conclusions are that the thermocapillary flow, when present, is dominant with surface velocities on the order of 1 m/s. The heat transfer, as deduced from isotherms within the pool, is crucially dependent on the flow. Also, flow velocities and weld-pool surface temperatures were measuredand agree well with the model.2. IntroductionThe requirements of improved control and of automation in the TIG (tungsten-inert gas) welding process have led to the development of mathematical models for the heat transfer to the workpiece. It is now widely recognized that the fluid flow in weld pools sufficiently distorts the conduction heat transfer pattern to warrant inclusion in a process model. The experiments of Woods and Milner (1), among others, demonstrated that TIG weld pools could be driven by the electromagnetic J x B body forces present, J being the current distribution and B the induced magnetic field. Atthey (2) calculated that in typical pools the buoyancy force was about an order of magnitude less important - despite the large temperature gradients present - and neglected it in a computational study of the magnetohydrodynamic (MHD) flow. This work also reviewed many earlier studies of MHD flow caused by a source of current at a fluid surface, including analytical studies by Shercliffe (3), Sozou (4), and Sozou and Pickering (5), and a numerical study by Andrews and Craine (6). These works establish the importance of the electromagnetic force in the weld-pool flow."

Jan 1, 1986

By Robert H. Peters

Several significant problems associated with employee turnover among underground miners are described in this Bureau of Mines report. Although a vast amount of research has been conducted on turnover among employees in nonmining industries, there has never been an attempt to empirically investigate the causes of turnover among underground miners. This report presents the results of empirical research on employee turnover in other industries and describes and critiques the research methods that have been used to investigate turnover. Using the findings on employee turnover in other industries and known data about miners and the mining industry, a conceptual model of the factors that cause turnover among miners was generated. This model, which is expected to provide a basis for later empirical studies, is presented. Several recommendations for controlling miner turnover are also presented.

Jan 1, 1986