METALLURGICAL AND MATERIALS TRANSACTIONS B
CONTENTS and ABSTRACTS
Volume 29B, No. 6, December 1998

This Month Featuring: Hydrometallurgy; Pyrometallurgy; Electrometallurgy; Transport Phenomena; Physical Chemistry; Solidification; Solid State Reaction; Materials Processing; Surface Treatment; Mathematical Modeling.

HYDROMETALLURGY

Beneficiation of West Sibaiya Phosphate Ores by Flotation in Alkaline Media [pp. 1149-1156]
R.M. AWADALLAH, A.E. MOHAMED, N.T. El HAZEK, and M.Y. HASSAN
Four big head phosphate samples (about 200 kg) were collected from West Sibaiya mines medium grade (sample 1, southern mine), high grade (sample 2, the mine in the cultivated area), low grade (sample 3, oversize), and tailings (sample 4) phosphate ores. The representative samples were analyzed chemically for P2O5, Fe2O3,Al2O3, InR, CaO, MgO, TiO, MnO, SO2-4, F-, Cl- loss on ignition (LOI), and U before and after screen analysis (-10.0 +9.50, -9.50 +6.70, -6.70 +2.0, -2.0 +1.0, -1.0 +0.85, 0.85 +0.710, -0.710 +0.500, -0.500 +0.350, -0.350 +0.250, -0.250 +0.180, -0.180 +0.125, -0.125 +0.075, and -0.075-mm-size fractions). The different grades of phosphate ores (organogenic-granular phosphorite ores with carbonates, calcite, calcedony, etc.) were beneficiated using local crude rice bran oil and imported oleic acid (individuals) as collectors, NaOH or KOH as the pH adjustor, and Na2SiO3 as the silica depressant. The results indicate that -0.355, +0.180 mm grain size, pH = 9.6, 20 minutes conditioning time, 0.7 kg Na2SiO3 (depressant)dose/ton feed ore, 4.25 kg rice bran oil dose/ton feed of phosphate ore or 3.4 kg oleic acid (collector) dose/ton feed of phosphate ore, and cleaning (reflotation) of phosphate concentrates are the optimum conditions for beneficiation of Sibaiya phosphate ores. On cleaning (reflotation) the phosphate ore feeds of samples 1 (medium, P2O5 = 28.67 pct), 2 (high, P2O5 = 31.70 pct), 3 (oversize, P2O5 = 25.30 pct), and 4 (tailings, P2O5 = 22.61 pct)using rice bran oil as a collector, the feeds beneficiate into concentrates of P2O5 pct equal to 31.87, 34.70, 26.16, and 31.90 pct (recoveries equal 49.41, 61.41, 52.01, and 29.03 pct P2O5) for samples 1, 2, 3, and 4, respectively. On using oleic acid as a collector, P2O5 pct in phosphate feeds of samples 1, 2, 3, and 4 increases (upgrades) into 31.28, 33.95, 26.82, and 30.70 pct in concentrates (recoveries equal 61.74, 53.60, 53.31, and 36.30 P2O5). Comparing the results obtained on using rice bran oil and oleic acid as collectors from the economic point of view, the costs of producing 1 ton phosphate concentrate by rice bran oil are lower (~LE 70) than those on using oleic acid (~LE 100) as a flotation collector.

The Acid-Base Behavior of Zinc Sulfate Electrolytes: The Temperature Effect [pp. 1157-1166]
WEIDONG WANG and DAVID B. DREISINGER
The pH of both synthetic zinc sulfate solutions of various compositions and commercial zinc plant electrolytes was measured over a range of temperatures. A model for the solution thermodynamics has been developed to predict the solution speciation, temperature, and concentration effects on the pH. It was found from both the thermodynamic predictions and the pH measurements that the pH of zinc sulfate electrolytes, in the absence of free acid, drops with increasing temperature. The pH temperature behavior was largely dominated by zinc hydrolysis. The pH of zinc sulfate electrolytes with small amounts of free acid both increased and then decreased in the temperature range of interest. This was explained by taking into account the additional effects of bisulfate/sulfate equilibrium and/or ZnSO4 ion pairing on the overall pH behavior. Based on the correlation between the model and pH measurements, it is evident that the dinuclear species Zn2(OH)3+ exists at a much higher concentration than Zn(OH)+ ions and dominates the pH-temperature behavior of the solution. Speciation and the acid/base composition of a ZnSO4 solution, against pH at 100°C, were also predicted. The pH-temperature behavior of zinc plant electrolytes from Kidd Creek (Falconbridge Limited, Timmins, Canada) and CEZinc (Noranda Limited, Valleyfield, Canada) was measured by saturating the electrolytes with ZnO at 100°C and then allowing the solutions to cool. The pH first increased slightly and then dropped from a maximum pH of 4.2. By including species involving Al3+, Mg2+, Mn2+, and Na+ in the zinc plant electrolytes in the solution model calculation, model predictions of the pH-temperature were again correlated with the pH-temperature measurements on zinc plant electrolytes.

PYROMETALLURGY

Titanium Powder Production by TiCl4 Gas Injection into Magnesium through Molten Salts [pp. 1167-1174]
TETSUSHI N. DEURA, MASAHIRO WAKINO, TOMOYA MATSUNAGA, RYOSUKE O. SUZUKI, and KATSUTOSHI ONO
A process to produce titanium powder continuously is proposed and its applicability is examined experimentally. The method is based on the chemical reaction in the conventional Kroll reduction process; however, TiCl4 gas is injected into molten salt on which a molten magnesium layer is floated as the reductant. Bubbles of gaseous TiCl4 can be reacted at the lower surface of the liquid Mg layer, while TiCl4 gas reacts on the upper surface in the Kroll process. The fine Ti particles produced in this study were well separated from magnesium and could be recovered from the bottom of the molten salts. The particles were small and fine enough for use in powder metallurgy, while congregated lumps of about 20 µm in size are obtained by the Kroll process. The composition of molten salts and an operation temperature above 1073 K did not affect the morphology of the Ti particles, if suitable material for the reaction vessel was chosen.

A Study on Carbothermic Reduction of Ilmenite Ore in a Plasma Reactor [pp. 1175-1180]
R.K. GALGALI, H.S. RAY, and A.K. CHAKRABARTI
Ilmenite has been smelted in a plasma reactor to study the effect of various process parameters such as the amount of reductant, flux, and composition of plasmagen gas on the degree of reaction to recover titanium and iron. Experimental conditions suitable for preparing a TiC-rich Fe-TiC master alloy and TiC-reinforced cast-iron composites have been determined. Thermodynamic data on various possible reactions are summarized and analyzed to explain the experimental observations.

Thermodynamic Model for MnO-Containing Slags and Gas Slag-Metal Equilibrium in Ferromanganese Smelting [pp. 1181-1191]
HONGJIE LI, ARTHUR E. MORRIS, and DAVID G.C. ROBERTSON
The thermodynamics of MnO-containing slags was reviewed, and the properties of such slags were correlated by application of the cell model. This model, together with a previously developed unified interaction parameter model for the metal phase, was used to simulate the gas-slag-metal conditions in the hearth of a ferromanganese submerged arc furnace. The calculated slag and metal compositions and distribution of Mn, Si, and K between slag, metal, and fume were compared with plant data and satisfactory agreement was obtained. A steady recirculation of potassium inside the furnace was also predicted. For a potassium percentage in the raw materials of 0.63 pct, the amount of the recirculating potassium was about 160 kg per metric ton of ferromanganese. Due to the recirculation, the average potassium content in the solid mix coming down to the hearth could be as high as 7 pct.

ELECTROMETALLURGY

Measurement of pH in the Vicinity of a Cathode during the Chloride Electrowinning of Nickel [pp. 1193-1198]
KUNIAKI MURASE, TAKESHI HONDA, TETSUJI HIRATO, and YASUHIRO AWAKURA
An antimony microelectrode was prepared by quenching a molten Sb-Sb2O3 mixture (2 pct Sb2O3). The local pH in the vicinity of a nickel-plated copper cathode was directly measured using the microelectrode during the chloride electrowinning of nickel for a MCLE (matte chlorine leach electrowinning) process, where nickel metal is electrodeposited with a high current efficiency, 94 to 97 pct, from low-pH baths. The local pH at 328 K was increased by proton consumption during the electrolysis of aqueous electrolytes containing NiCl2 (1.20 mol dm-3) and NaCl (0.43 mol dm-3) with the same concentrations as employed for the MCLE process. The difference in pH between the cathode surface and bulk solution increased with increasing cathodic current density. Nickel deposits with a metallic luster were obtained when the difference was not more than 1.2 pH units. The current efficiency was a maximum for electrolysis with a current density of 265 A m-2 and bulk pH of 1.0 to 1.5; these optimal conditions coincided with those reported for the MCLE process: temperature 328 to 333 K, bulk pH 1.1 to 1.5, and current density 230 to 260 A m-2. Electrolytes with lower NiCl2 and NaCl concentrations resulted in a drop in current efficiency.

TRANSPORT PHENOMENA

Gas Flow Analysis in Melting Furnaces [pp. 1199-1207]
L.I. KISS, R.T. BUI, A. CHARETTE, and T. BOURGEOIS
The flow structure inside round furnaces with various numbers of burners, burner arrangements, and exit conditions has been studied experimentally with the purpose of improving the flow conditions and the resulting heat transfer. Small-scale transparent models were built according to the laws of geometric and dynamic similarity. Various visualization and experimental techniques were applied. The flow pattern in the near-surface regions was visualized by the fluorescent minituft and "popcorn" techniques; the flow structure in the bulk was analyzed by smoke injection and laser sheet illumination. For the study of the transient effects, high-speed video photography was applied. The effects of the various flow patterns, like axisymmetric and rotational flow, on the magnitude and uniformity of the residence time, as well as on the formation of stagnation zones, were discussed. Conclusions were drawn and have since been applied for the improvement of furnace performance.

Effects of Pore Diameter, Bath Surface Pressure, and Nozzle Diameter on the Bubble Formation from a Porous Nozzle [pp. 1209-1218]
MANABU IGUCHI, MASUO KAJI, and ZEN-ICHIRO MORITA
The effects of the pore diameter, bath surface pressure, and nozzle diameter on the bubble formation from a porous bottom nozzle placed in a water bath and on the behavior of rising bubbles were investigated with still and high-speed video cameras and a two-needle electroresistivity probe. Three types of bubble dispersion patterns were observed with respect to gas flow rate, and they were named the low, medium, and high gas flow rate regimes. The transition boundaries between these gas flow rate regimes were expressed in terms of the superficial velocity at the nozzle exit, i.e., the volumetric gas flow rate per unit nozzle surface area. These transition boundaries were dependent on the pore diameter but hardly dependent on the bath surface pressure and the porous nozzle diameter. The characteristics of rising bubbles in each gas flow rate regime were investigated as functions of the three parameters.

Effect of Cross-Flow on the Frequency of Bubble Formation from a Single-Hole Nozzle [pp. 1219-1225]
MANABU IGUCHI, YUKIO TERAUCHI, and SHIN-ICHIRO YOKOYA
Bubble formation from a single-hole nozzle placed vertically upward in a rotating water bath was investigated using a high-speed video camera. Air was used as the working gas. The measured values of the frequency of bubble formation, fB, were compared with those observed in a stationary bath, fB0. The velocity of cross-flow, v, affected the bubble formation significantly when it exceeded a critical value, vc. The ratio of fB to fB0 was unity for v vc, but it changed in a complex manner for v > vc. In the latter case, when the air flow rate Q was relatively low, fB/fB0 became larger than unity irrespective of Q, and an empirical correlation of fB/fB0 was proposed as a function of v and the inner diameter of the nozzle, dni. As the gas flow rate increased, fB/fB0 decreased monotonically and became smaller than unity, and an empirical correlation of fB/fB0 was derived as a function of Q, v, and dni. These empirical correlations could approximate the measured values of fB/fB0 within a scatter of -15 to +20 pct.

Evaluation of Velocity-Dependent In Situ Leaching Processes: Single-Porosity Model [pp. 1227-1234]
J. LIU and B.H. BRADY
A methodology to characterize processes of in situ leaching is developed to study various parameters that may affect the recovery of a valuable mineral and to assure a successful application of the in situ leaching technology. A leaching kinetics model is first derived based on the concept of representative elementary volume (REV) in porous ore deposits. Every parameter in the model is clearly defined and may be easy to obtain in practice. Then the governing equations are obtained for transport of both the reagent and the dissolved mineral in porous ore deposits. These equations are solved by a fully three-dimensional computer simulator. Effects on the mineral recovery of parameters such as lumped rate constant, ore porosity, injection flux, and macrodispersivities are investigated through the three-dimensional case with one injection well. It is found that all of these four parameters, interconnected through the ore porosity, significantly affect the effectiveness of the in situ leaching. Furthermore, the flow velocity of leach solution within an ore deposit may be the key parameter to the design of a real leaching mine because all of the other three parameters are site specific.

PHYSICAL CHEMISTRY

Kinetics Studies on the Dissolution of Nitrogen in the CaO Al2O3-SiO2 and CaO-Al2O3-TiOx Melts [pp. 1235-1240]
KOICHI IUCHI, KAZUKI MORITA, and NOBUO SANO
The rate of nitrogen dissolution in CaO-Al2O3-SiO2 and CaO-Al2O3-TiOx melts was measured by 14N-15N isotope exchange reaction. The rate constant for the CaO-Al2O3-SiO2 melts at the ratio of mass pct CaO/mass pct Al2O3 = 1 increases as SiO2 content increases, whereas the rate constant for the same melts at the ratio of mass pct CaO/mass pct SiO2 = 1 increases as Al2O3 content increases. The rate constant for the CaO-Al2O3-TiOx melts at the ratio of mass pct CaO/mass pct Al2O3 = 1 decreases as the TiOx content increases. The activation energies of nitrogen dissolution in CaO Al2O3-SiO2 melts are about 1.5 to 3 times larger than that of molten pure iron. Moreover, the rate constant of nitrogen dissolution is independent of the ratio of Ti3+/Ti4+.

Activities in the Spinel Solid Solution FeXMg1-XAl2O4 [pp. 1241-1248]
K.T. JACOB and RASHMI PATIL
Activities in the spinel solid solution FeXMg1-XAl2O4 saturated with -Al2O3 have been measured for the compositional range 0 < X < 1 between 1100 and 1350 K using a bielectrolyte solid-state galvanic cell, which may be represented as Pt, Fe + FeXMg1-XAl2O4 + -Al2O3//(Y2O3)ThO2/[fj (CaO)ZrO2//Fe + FeAl2O4 + -Al2O3, Pt Activities of ferrous and magnesium aluminates exhibit small negative deviations from Raoult's law. The excess free energy of mixing of the solid solution is a symmetric function of composition and is independent of temperature: GE = -1990 X(1-X) J/mol. Theoretical analysis of cation distribution in spinel solid solution also suggests mild negative deviations from ideality. The lattice parameter varies linearly with composition in samples quenched from 1300 K. Phase relations in the FeO-MgO-Al2O3 system at 1300 K are deduced from the results of this study and auxiliary thermodynamic data from the literature. The calculation demonstrates the influence of intracrystalline ion exchange equilibrium between nonequivalent crystallographic sites in the spinel structure on intercrystalline ion exchange equilibrium between the monoxide and spinel solid solutions (tie-lines). The composition dependence of oxygen partial pressure at 1300 K is evaluated for three-phase equilibria involving the solid solutions Fe + FeXMg1-XAl2O4 + -Al2O3 and Fe + FeYMg1-YO + FeXMg1-XAl2O4. Dependence of X, denoting the composition of the spinel solid solution, on parameter Y, characterizing the composition of the monoxide solid solution with rock salt structure, in phase fields involving the two solid solutions is elucidated. The tie-lines are slightly skewed toward the MgAl2O4 corner.

Communication: Solubility of Bismuth in -Iron [pp. 1371-1372]
LEWEN CHANG AND MARK E. SCHLESINGER

Communication: On the Calculation of Ionic Equilibria Using the Gibbs Energy Minimization Method [pp. 1372-1374]
PETR VONKA AND JINDRICH LEITNER

SOLIDIFICATION

Simulation of Microporosity Formation in Modified and Unmodified A356 Alloy Castings [pp. 1249-1260]
JULIE HUANG, T. MORI, and JAMES G. CONLEY
In order to comprehensively model both the performance and inspectability of early design stage safety critical aluminum castings, the size, shape, and location of defects such as pores should be determined by simulation. In this article, a two-dimensional (2-D) model to predict grain size, pore size, pore morphology, and location is presented. The proposed model couples hydrogen gas evolution and microshrinkage pore formation mechanisms with a grain growth simulation model. The nucleation and growth of grains are modeled with a probabilistic method that uses the information from a macroscale heat transfer simulation to determine the rules of transition for grain evolution. Microshrinkage pores and the combination of microshrinkage and gas pores are addressed. The proposed model and postprocessing can provide direct simulated views of the microstructure of the solidifying casting. In the present work, the effect of Sr modifier and hydrogen content on pore size and morphology for equiaxed aluminum alloy A356 is modeled. The simulation results correlate well with the experimental observation of cast structures and other published data. In addition, Sievert's law and the conditions for spontaneous growth of a gas pore are derived from first principles in the Appendix.

Fractal Analysis of the Surface Cracks on Continuously Cast Steel Slabs [pp. 1261-1267]
HIROSHI KAMETANI
Data concerning the length of longitudinal cracks on the surface of continuously cast steel slabs were collected from two plants. The data were analyzed to find the relation between crack length and crack frequency. The analysis revealed the following:

  1. After normalization to eliminate the effect of different casting conditions, the fractal relation characterizing the normalized cumulative frequency distribution (N (m-2)) and the crack length (L (mm)) of the primary surface cracks could be represented by the equation
N = Nc = kc L-1.5
where Nc is the cumulative frequency before normalization, is the normalizing coefficient, and kc is a constant.
  1. The values for varied over a wide range, but remained constant throughout a heat and were the same for both the upper and lower faces of the slab.
  2. It was found that in some instances, when L exceeded a critical value (Lc), the value of L became times longer than the length predicted by the previous distribution. This increase in L was ascribed to secondary growth of the cracks. This occurred more frequently on the lower, rather than on the upper, face of the slab. The product of Lc and was approximately constant. The formation of the surface cracks is discussed in view of the fractal phenomena.

Solidification Behavior and Microstructural Evolution during Laser Beam-Material Interaction [pp. 1269-1279]
P.S. MOHANTY and J. MAZUMDER
A laser-assisted visualization technique has been used to monitor the solidification behavior at the tail of a molten pool created by scanning high energy density laser beam. A high speed digital camera with spatial resolution of 64 X 64 pixels and temporal resolution of 40,500 frames/s has been employed along with a novel concept of illuminating the interaction zone by a secondary visible probe laser. This technique enabled in situ monitoring of the solid/liquid interface due to the characteristic difference in the reflectivity between solid and liquid surfaces. It is observed that the solidification behavior is unstable and is highly influenced by the instabilities in the flow, which develop from the complex laser-material interaction process. Quite often the growth front remelted back due to the fluctuating thermal field driven by flow instability. The fluctuations in the growth front and the fluctuations in the laser-material interaction process have been monitored simultaneously, however, no correlation is apparent. The influence of flow instability on the resulting microstructure has been analyzed.

The Use of Particle Image Velocimetry in the Physical Modeling of Flow in Electromagnetic or Direct-Chill Casting of Aluminum: Part I. Development of the Physical Model [pp. 1281-1288]
DONG XU, W. KINZY JONES, Jr., and J. W. EVANS
This article describes the development of a physical (water) model for examining flows in the liquid metal pool, at the head of an electromagnetic (EM) or direct-chill (DC) caster for aluminum, that are driven by the inflow of liquid metal. Also described is a particle image velocimetry (PIV) system developed for measuring the flows; the system dispenses with the laser usually required for PIV. Flows in the model were found to be unsteady, but ensemble averaging yielded flow patterns that conformed to expectations and velocities that can be compared to the time-averaged velocities (TAVs) of mathematical models. Some representative results are presented (with more extensive results to be included in a subsequent Part II of this article) showing that the design of the metal delivery system has a great effect on the flows.

The Use of Particle Image Velocimetry in the Physical Modeling of Flow in Electromagnetic or Direct-Chill Casting of Aluminum Part II. Results of the Physical Model, Including Bag Geometry, Blockage, and Nozzle Placement [pp. 1289-1295]
DONG XU, W. KINZY JONES, Jr., and J.W. EVANS
This second part of a two-part article describes the application of the physical model and particle image velocimetry (PIV) system, treated in Part I, to studying the flows in the pool at the top of the casters used for semicontinuous casting of aluminum. The effects of various designs of "bags" that are used in the industry to modify the flow were determined, as were the effects of blocking the outflow "windows" of the bag and the effects of varying the submergence of the nozzle delivering metal into the bag. The flows were found to be strongly dependent on bag design; for example, on the length of the bag and the presence of windows in the bottom of the bag. Flows in the vicinity of the (simulated) solidification front were greatly changed by such apparently small changes in design as lengthening the bag by 50 mm. A qualitative explanation is offered for some of these effects, and the article ends with a discussion of what might be the best flow from the viewpoint of caster operability or ingot structure.

SOLID STATE REACTION

Carbochlorination Kinetics of Titanium Dioxide with Carbon and Carbon Monoxide as Reductant [pp. 1297-1307]
FENGLIN YANG and VLADIMIR HLAVACEK
Kinetic study of the chlorination of titanium dioxide (rutile and anatase) was carried out in a fixed bed reactor at temperature ranging from 800°C to 1000°C and normal pressure. In our experiment, titanium dioxide powder and gaseous chlorine with carbon or carbon monoxide as reductant were used. The products of the reaction are all in gaseous phase under the temperatures and pressure studied. With CO as reductant, reaction is of noncatalytic gas-solid nature and experimental data fit the shrinking core model. When using C as reductant, solid-solid reaction is involved. Reactivity is highly enhanced by solid carbon and it is concluded that an activated C-TiO2-Cl complex contributes to the enhanced reactivity. A reaction model based on phase boundary control applies to the experimental data. Thermodynamic analysis supports the experimental observation.

MATERIALS PROCESSING

Desorption Kinetics of Carbon and Oxygen in Liquid Niobium [pp. 1309-1314]
HYUN GYOON PARK and REZA ABBASCHIAN
Desorption kinetics of carbon and oxygen in liquid niobium were investigated, in the temperature range from 2700 to 3000 K, in a CO/Ar stream using an electromagnetic levitation technique. It was found that desorption of both species occurs via CO evolution, especially when the starting metal has an excess of carbon. Desorption of carbon and oxygen in liquid niobium are second-order processes with a first-order dependence on C and O concentrations, indicating that the overall reaction rate is controlled by the substep of recombination of C and O. The rate equations for C and O desorption are found to be (dO/dt)des = (dC/dt)des = 3.80 · 10-4 (A/V) exp (-47,500/T) (Ce Oe - CO), where C and O are the carbon and oxygen concentrations in atomic percent, respectively, A is the surface area of the sample in square centimeters, V is the volume in cubed centimeters, and the subscript e denotes the equilibrium concentration.

SURFACE TREATMENT

Effect of Oxidation Treatment and Surface Filming on Hydrogen Degassing from TiH2 [pp. 1315-1319]
FUSHENG HAN, ZHENGANG ZHU, JUNCHANG GAO, and WENHAI SONG
An experiment was carried out to determine the effects of oxidation treatment and surface filming on hydrogen degassing from TiH2 during heating and at fixed temperatures. It was found that surface filming has a stronger effect than oxidation, through which the temperature corresponding to the degassing peak of TiH2 moved from 638°C to more than 680°C and the degassing temperature range increased from 125°C to 150°C in the heating process. From the degassing measurements at fixed temperatures and the physics model, the diffusion coefficients of hydrogen were determined to be 0.3 to 3.0 times less than that of the original TiH2.

MATHEMATICAL MODELING

Numerical Investigation of the Interface in a Continuous Steel Casting Mold Water Model [pp. 1321-1327]
A. THEODORAKAKOS and G. BERGELES
In this study, the steady-state Navier-Stokes equations are solved on a curvilinear nonorthogonal grid, following the finite volume approximation, with a pressure prediction-correction method, for the case of a flow in a model steel casting mold. The steel flow is simulated by water flow and the slag layer by an oil film, following conditions of previous experimental studies. The simulation aims at the understanding of the free wave and the interface surface wave behavior and the mechanism that leads to the breakup of the steel-slag interface, and thus induction of impurities inside the final steel product. Boundary conditions are set on the free and the interface surfaces, and an adaptive grid mechanism is used in order to update the grid's shape so as to follow the wave formation. Several cases have been considered with the inlet velocity parameter, and results concerning the velocity field and the generated waves are reported. It is shown that a critical casting speed exists that leads to wave instability, which may be associated with emulsification phenomena.

Evaluation of Nickel Flash Smelting through Piloting and Simulation [pp. 1329-1343]
S.R. VARNAS, N. KEMORI, and P.T.L. KOH
An extensive study of the nickel flash smelting process has been undertaken. It is aimed at the optimization of the burner design to improve the smelting performance and to increase the throughput of the rebuilt furnace. A design-based mathematical model was developed to simulate the operation of the four burners and the reaction shaft of the flash furnace at Western Mining Corporation Ltd.'s Kalgoorlie Nickel Smelter. A modified single burner version of the model was validated against data obtained from the pilot plant at the Pyrometallurgical Research Centre (PRC) of the Sumitomo Metal Mining Co.'s Toyo Smelter. The approach taken involved experimental measurements of key process parameters in the pilot plant and detailed numerical simulation of the fluid flow, heat transfer, and combustion in the entire burner-shaft complex. Several burner designs have been tested experimentally at the pilot plant and theoretically through computer simulation. The main outcome of the study was the development of an experimentally validated mathematical model of the flash smelter providing a new powerful design tool. The insight gained about the process from the application of this tool led to the design of a more efficient nickel flash smelting process.

Mathematical Simulation on Coupled Flow, Heat, and Solute Transport in Slab Continuous Casting Process [pp. 1345-1356]
HONGLIANG YANG, LIANGANG ZHAO, XINGZHONG ZHANG, KAIWEN DENG, WENCAI LI, and YONG GAN
A three-dimensional comprehensively coupled model has been developed to describe the transport phenomena, including fluid flow, heat transfer, solidification, and solute redistribution in the continuous casting process. The continuous casting process is considered as a solidification process in a multicomponent solid-liquid phase system. The porous media theory is used to model the blockage of fluid flow by columnar dendrites in the mushy zone. The relation between flow pattern and the shape of the solid shell is demonstrated. Double diffusive convection caused by thermal and concentration gradients is considered. The change in the liquidus temperature with liquid concentration is also considered. The formation mechanism of macrosegregation is investigated. Calculated solid shell thickness and temperature distribution in liquid core are compared with the measured quantities for validating the model.

Liquid Flow on a Rotating Disk Prior to Centrifugal Atomization and Spray Deposition [pp. 1357-1369]
Y.Y. ZHAO, M.H. JACOBS, and A.L. DOWSON
Video observations of the flow patterns that develop on a rotating disk during centrifugal atomization and spray deposition, and subsequent metallographic studies conducted on solid skulls removed from the disk after processing, have indicated a circular discontinuity or hydraulic jump, which is manifested by a rapid increase in the thickness of the liquid metal and by a corresponding decrease in the radial velocity. A mathematical model has been developed that is capable of predicting both the occurrence and location of the jump, and the associated changes in the thickness profile and in the radial and tangential velocities of the liquid metal. Good correlations have been observed between model predictions and the flow patterns observed on the skull after atomization, and the effects of changes in material and operational parameters such as kinematic viscosity, volume flow rate, metallostatic head, and disk rotation speed have been quantified. Liquid metal flow is controlled primarily by the volume flow rate and by the metallostatic head prior to the hydraulic jump and by the centrifugal forces after the jump. The implications of these observations in terms of the atomization process are discussed.

Table of Contents and Abstracts, Metallurgical and Materials Transactions A, December 1998 [pp. 1371-1372]

Combined Index to Volumes 29A and 29B [follows page 1372]

Direct questions about this or any other Metallurgical and Materials Transactions page to mettrans@andrew.cmu.edu.

Search TMS Document Center Tables of Contents Subscriptions Met. and Mat. Trans. TMS OnLine