| METALLURGICAL AND MATERIALS TRANSACTIONS B | |
|---|---|
|
|
|
Volume 26B, No. 4, August 1995 This Month Featuring: Hydrometallurgy, Pyrometallurgy, Transport Phenomena, Physical Chemistry, Solidification, Solid State Reactions, and Mathematical Modeling. View August 1995 Table of Contents.
|
|
where N represents moles of pyrite S is the surface area of the solid particles k" and k"' are constants b is a stoichiometric factor pO2 is the oxygen partial pressure and [OH¯] is the hydroxyl ion concentration. The corresponding fractional conversion (X) vs time behavior follows the shrinking particle model for chemical reaction control:
The rate increases with the reciprocal of particle size and has an activation energy of 55.6 kJ/mol (13.6 kcal/mol). The relationship between reaction rate and oxygen partial pressure resembles a Langmuir-type equation and thus suggests that the reaction involves adsorption or desorption of oxygen at the interface. The square-root rate law may be due to the adsorption of a dissociated oxygen molecule. The observed apparent reaction order with respect to the hydroxyl ion concentration is a result of a complex combination of processes involving the oxidation and nydrolysis of iron oxidation and hydrolysis of sulfur and the oxygen reduction.
Simulation of In Situ Uraninite Leaching-Part II: The Effects of Ore
Grade and Deposit Porosity
KNONA C. LIDDELL and RENATO G. BAUTISTA
A combined partial equilibrium-mixing cell model has been used to investigate the effects of fluid flow, mineral content, porosity, and lixiviant concentrations on in situ leaching of uraninite. The model couples the rate processes of reactive transport (uraninite and calcite dissolution kinetics and leach solution flow) with solution phase equilibria (acid-base and complexation equilibria). Solution circulation and porosity changes have been explicitly treated in the following way: reacted solution was assumed to be pumped from the system at a constant rate and replaced by fresh lixiviant; the additional void volume resulting from CaCO3 or UO2 dissolution was immediately filled with lixiviant. A solution volume of 1 cm3 was taken for the base, and it was assumed that on each 1200 second increment, loaded solution was removed at the rate of 1.67 X 10-5 cm s-1, equivalent to removal of 2.0 pct of the base volume. The lixiviant considered was NH4HCO3-(NH4)2CO3-H2O2 with reference case concentrations of 1.0 X 10-4, 1.0 X 10-4, and 2.2 x 10-5 mol cm-1. The parameters that were varied in this investigation were the mass fractions of UO2 (0.000 to 0.015) and CaCO3 (0.00 to 0.40) and the initial porosity of the deposit (0.20 and 0.30). Major factors found to affect the uranium content of the solution were UO2 content and initial porosity. Higher UO2 grades were associated with higher U(VI) concentrations, and these were maintained for much longer periods; the consumption of the peroxide oxidant was under mass transfer control. As the leaching reaction slowed, solution replacement began to control the component concentrations, causing decreasing U(VI) concentrations. Higher porosity caused reduced maximum U concentrations and a faster decline. The calcite content had a slight effect on the rate of U leaching; this occurred because high CaCO3 mass fractions led to increased HCO3 concentrations. Early in the leaching process, a lower initial porosity or a higher calcite content led to a higher (less negative) value of the CaCO3 saturation index; however, for the conditions simulated, the solution did not actually become saturated. Also, decreases in the saturation index occurred sooner for higher initial porosities or lower calcite grades. The final porosity was effectively determined by the initial calcite content; dissolution of calcite continued until it had completely reacted, and the uraninite content was too low for it to contribute significantly. Changes in concentrations of the various solution species occurred more rapidly if the ore was more porous, but there were no other sigmficant differences attributable to initial porosity. The H+ concentration was virtually constant throughout leaching if the ore did not contain any calcite; with high calcite contents (40 pct), it remained constant for an extended period following an initial sharp decrease. Changes in the OH¯, NH4+, and NH3 concentrations could be readily predicted from those of H+, and changes in the Ca species concentrations were closely,rplated to those of the Ca and CO3 components. Total U and total H2O2 concentrations behaved oppositely (as required by the reaction stoichiometry), but changes in the concentrations of the minor U(VI) and peroxo species were more complicated. The concentrations of the CO32-- and HCO3¯ species could not readily be predicted from the reaction kinetics, and variations in their concentrations did not reliably indicate pH.
Simulation of In Situ Uraninite Leaching-Part III: The Effects of Solution Concentration
KNONA C. LIDDELL and RENATO G. BAUTISTA
The effects of variations in the concentrations of leaching reagents have been simulated for in situ leaching of UO2 by H2O-(NH4)2CO3-NH4HCO3. The model used in the simulations incorporates rate laws for the mineral reactions, equilibrium reactions among the solution species, and a mixing cell representation of solution flow. Of the component concentrations, the major factor affecting the rate of uraninite dissolution is the oxidant concentration. High peroxide concentrations lead to more rapid reaction with an early maximum in the U(VI) concentration. If lower oxidant concentrations are used, the reaction is under mixed kinetic and mass transfer control and the U(VI) concentration is lower but approximately constant for an extended period. Because they increase the concentration of the HCO3¯ anion, high ammonium carbonate and ammonium bicarbonate concentrations also result in some enhancement in the rate of U leaching; the reaction is known to be half-order in both HCO3¯ and H2O2. A 10:1 ratio of (NH4)2CO3 to NH4HCO3 concentrations was found to result in a nearly constant pH during most of the leaching process. Calcite-containing gangue causes an immediate pH increase from about 8.9 to 9.4. The rate of the calcite reaction, calcite saturation index, and porosity are all independent of the lixiviant concentrations. Detailed calculations of solution speciation are necessary to predict the concentrations of individual species from those of components.
Kinetics of Chlorination and Carbochlorination of Vanadium Peroxide
I. GABALLAH, M. DJONA, and E. ALLAIN
Kinetics of chlorination of V205 with Cl2-air, Cl2-N2, and Cl2-CO-N2 gas mixtures have been studied by nonisothermal and isothermal thermogravimetric measurements. In the temperature range of 500° to 570°, the chlorination of V2O5 with Cl2-N2 gas mixture is characterized by an apparent activation energy of about 235 kJ/mole. This could be attributed to chemical reaction. Between 570° and 650°, the apparent activation energy is equal to 77 kJ/mole, indicating that the overall reaction rate is controlled by chemical reaction and pore diffusion. The reaction order with respect to chlorine is 0.78. The apparent activation energy of the carbochlorination of V2O5 by Cl2-CO-N2 gas mixture is about 100 kJ/mole in the temperature range of 400° to 620°. In this case, the chemical reaction is the limiting step. At temperatures higher than 620°, an anomaly is observed in the Arrhenius plot, probably due to thermal decomposition of COCI2 formed in situ and/or transformation of the vanadium oxide physical state. The maximum reaction rate is obtained by using a Cl2-CO-N2 gas mixture having a Cl2/CO volume ratio equal to about 1.
Mathematical Model of Chalcocite Particle Combustion
A.A. SHOOK, G.G. RICHARDS, and J.K. BRIMACOMBE
A mathematical model has been developed to simulate the combustion of a single chalcocite particle (particle diameter between 10 and 100 microns) in air, oxygen, and oxygen-SO2 mixtures. Neglecting temperature and composition gradients within the particle, the model computes the thermal and compositional changes of the particle as a function of time. Five chemical reactions were considered to describe the chemical interaction between the gas and particle and within the particle. Copper vaporization below the boiling point was calculated by a balance between the vaporization rate (characterized by the Langmuir-Knudsen equation) and mass transfer from the surface of the particle. The model calculations were verified by comparison with published mass loss data obtained in a stagnant gas reactor. The model revealed that for particles combusting in oxygen, copper vaporization below the boiling point does not limit particle temperature, and particles of all sizes between 20 and 100 microns can readily reach the boiling point of copper (2836 K). Thus, the particle explosions observed in an earlier study are likely due to copper boiling within a combusting particle. Calculations of particle combustion in air showed that only small (<20-micron diameter) particles would be capable of exploding, which agreed geneMlly with earlier observations. Consequently, in a commercial flash-converting furnace, careful control of particle size and oxygen atmosphere must be maintained to minimize particle explosions and concomitant dust formation.
model was adopted to incorporate turbulence effects on transport processes in the system. A control-volume-based finite-difference procedure was employed to solve the conservation equations associated with appropriate boundary conditions. Because of high nonlinearity in the system of equations, a number of techniques were used to accelerate the convergence process. The effects of the parameters such as casting speed, steel grade, nozzle configuration on flow pattern, solidification profile, and carbon segregation were investigated. From the computed flow pattern, the trajectory of inclusion particles, as well as the density distribution of the particles, was calculated. Some of the computed results were compared with available experimental measurements, and reasonable agreements were obtained.
Natural Convection in the Hale-Shaw Cell of Horizontal Bridgman Solidification
YILI LU, JIAN LIU, and YAOHE ZHOU
The numerical simulation of natural convection in the Hale-Shaw cell during horizontal Bridgman solidfication reveals that the convection presents even for the very thin cell. The effects of the horizontal temperature gradient, G, thickness of the cell, H, temperature difference between the top and bottom of the cell, and other parameters have been studied. These findings have been confirmed by experiments through direct observation and measurement of convection in the cell containing succinonitrile transparent model alloy.
Numerical Study of Steady Turbulent Flow through Bifurcated Nozzles in Continuous Casting
FADY M. NAJJAR, BRIAN G. THOMAS, and DONALD E. HERSHEY
Bifurcated nozzles are used in continuous casting of molten steel, where they influence the quality of the cast steel slabs. The present study performs two-dimensional (2-D) and three-dimensional (3-D) simulations of steady turbulent (K-) flow in bifurcated nozzles, using a finite-element (FIDAP) model, which has been verified previously with water model experiments. The effects of no design and casting process operating variables on the jet characteristics exiting the nozzle are investigated. The nozzle design parameters studied include the shape, angle, height, width, and thicknessof the ports and the bottom geometry. The process operating practices include inlet velocity profile and angle as well as port curvature caused by erosion or inclusion buildup. Results show that the jet angle is controlled mainly by the port angle but is steeper with larger port area and thinner walls. The degree of swirl is increased by larger or rounder ports. The effective port area, where there is no recirculation, is increased by smaller or curved ports. Flow asymmetry is more severe with skewed or angled inlet conditions or unequal port sizes. Turbulence levels in the jet are higher with higher casting speed and smaller ports.
Communications: Study of Slopping and Splashing in a Cylindrical Bath with Top-Submerged Injection Chart
J.-L. LIOW, W.H.R. DICKINSON, M.J. ALLAN, and N.B. GRAY
mixH°m/ = - 59.6 ± 2.5 kJ · mol-1 at xPd = 0.59 and independent of temperature within the experimental error. The integral molar enthalpy of mixing is given bymixH°m/{kJ·mol-1}= x(1 - x)·(-126.94 - 92.653x - 83.231x2 - 734.49x3 + 949.07x4), where x = xpd. The limiting partial molar enthalpy of palladium in indium was calculated as hm(Pd liquid in 
liquid In) = - 127 ± 5 kJ · mol-1. The results are discussed and compared with the enthalpies of formation of solid alloys. The anomalous behavior of the partial enthalpy of Pd is assumed to be due to the charge transfer of, at most, two electrons of In to Pd.
Potentiometric Determination of the Gibbs Energies of Formation of SrZrO3 and BaZrO3
K.T. JACOB and Y. WASEDA
The Gibbs free energies of formation of strontium and barium zirconates have been determined in the temperature range 960 to 1210 K using electrochemical cells incorporating the respective alkaline-earth fluoride single crystals as solid electrolytes. Pure strontium and barium monoxides were used in the reference electrodes. During measurements on barium zirconate, the oxygen partial pressure in the gas phase over the electrodes was maintained at a low value of 18.7 Pa to minimize the solubility of barium peroxide in the monoxide phase. Strontium zirconate was found to undergo a phase transition from orthorhombic perovskite (o) with space group Cmcm; D2h17 to tetragonal perovskite (t) having the space group 14/mcm;D4h18 at 1123 (± 10) K. Barium zirconate does not appear to undergo a phase transition in the temperature range of measurement. It has the cubic perovskite (c) structure. The standard free energies of formation of the zirconates from their component binary oxides AO (A = Sr, Ba) with rock salt (rs) and ZrO2 with monoclinic (m) structures can be expressed by the following relations:
SrZrO3 (o)
G° = - 74,880 - 14.2T(±200) J mole-1
SrO (rs) + ZrO2 (m) SrZrO3 (t)
G° = -73,645 - 15.3T (±200) J mol-1
BaO (rs) + ZrO2 (m) BaZrO4 (c)
G° = - 127,760 - 1.79T (± 250) J mole-1
The results of this study are in reasonable agreement with calorimetric measurements reported in the literature. Systematic trends in the stability of alkaline-earth zirconates having the stoichiometry AZrO3 are discussed.
Activities of Phosphorous in Liquid Ni + P Alloys Saturated with Solid Nickel
R. KAWABATA, E. ICHISE, and M. IWASE
By using an electrochemical technique incorporating magnesia-stabilized zirconia electrolyte, the activities of phosphorous in liquid {Ni + P} alloys saturated with solid nickel were determined at temperatures between 1477 and 1663 K. The activities of phosphorus referred to gaseous diatomic phosphorous at 1 atm pressure, ap, could be expressed by analytical fommulas:
at temperatures between 1477 and 1612 K, and
at temperatures between 1612 and 1663 K.
The Activity of Calcium in Calcium-Metal-Fluoride Fluxes
YUICHIRO OCHIFUJI, FUMITAKA TSUKIHASHI, and NOBUO SANO
The standard Gibbs energy of reaction Ca (1) + O (mass pct, in Zr) = CaO
(s) has been determined as follows by equilibrating molten calcium with solid
zirconium in a CaO crucible:
G° = -64,300(±700) + 19.8(±3.5)T J/mol (1373 to 1623 K)The activities of calcium in the CaOsatd-Ca-MF2 (M: Ca, Ba, Mg) and CaOsatd-Ca-NaF systems were measured as a function of calcium composition at high calcium contents at 1473 K on the basis of the standard Gibbs energy. The activities of calcium increase in the order of CaF2, BaF2, and MgF2 at the same calcium fraction of these fluxes. The observed activities are compared with those estimated by using the Temkin model for ionic solutions. Furthermore, the possibility of the removal of tramp elements such as tin, arsenic, antimony, bismuth, and lead from carbon-saturated iron by using calcium-metal-fluoride fluxes is discussed.
Determination of the Chemical Diffusion of Oxygen in Liquid Iron Oxide at 1615 °C
Y. SAYADYAGHOUBI, S. SUN, and S. JAHANSHAHI
The chemical diffusion of oxygen in liquid iron oxide has been studied by the oxidation of a melt in a long capillary at 1615°. When pure oxygen was used as the oxidizing agent, the surface composition of the slag was found to be in close agreement with the expected gas-slag equilibrium, suggesting that diffusion is the controlling step. This was not the case when air, 5 pct oxygen in argon or pure CO2 was used to oxidize the slag. The deviation of the surface composition from the expected equilibrium was in accordance with a mechanism of mixed control by both the gas-slag reaction and diffusion in the bulk. The average value of the chemical diffusivity of oxygen (or iron) in liquid iron oxide with Fe2+/FeT between 0.25 and 0.77 was established to be 3(±1) x 10-7 m2/s. This value is one to two orders of magnitude higher than those from earlier studies. There seems to be a reasonable correlation between the chemical and the ionic self-diffusivities through the Darken equation. A quantitative analysis in this respect and on the role of electron hole migration depends on the availability of data on the ionic conductivity and the tracer diffusivities.
Effect of the Bubble Size and Chemical Reactions on Slag Foaming
Y. ZHANG and R.J. FRUEHAN
Slac foams have been investigated with smaller bubbles than those used in the previous studies.[5,6,7] The bubbles were generated by argon gas injection with the nozzle of multiple small orifices and by the slag/metal interfacial reaction of FeO in the slag with carbon in the liquid iron. The foam stability in terms of the foam index for a bath-smelting type of slag (CaO-SiO2-AI203-FeO) was determined for different bubble sizes. The average diameter of bubbles in the foam was measured by an X-ray video technique. When the foam was generated by the slag/metal interfacial reaction at 1450°C, it was found that the average bubble diameter varied from less than 1 to more than 5 mm as a function of the sulfur activity in the carbon-saturated liquid iron. The foam index was found to be inversely proportional to the average bubble diameter. A general correlation is obtained by dimensional analysis in order to predict the foam index from the physical properties of the liquid slag and the average size of the gas bubbles in the foam.
Effect of Carbonaceous Particles on Slag Foaming
Y. ZHANG and R.J. FRUEHAN
Use of carbonaceous particles such as coke or coal char in controlling slag foaming is of great practical significance for bath-smelting and other steelmaking processes. The foamability of the liquid slag in terms of the foam index has been determined with the presence of different amounts of coke and coal char particles. Different sized and shaped particles were used in the experiments. It was found that the foam index decreased significantly as the ratio of the total cross-sectional area of the particles to the liquid slag surface area increased. When the foam was generated by argon gas injection through an alumina nozzle (i.d. = 1.S mm), a liquid slag, CaO-SiO2-CaF2-(AI2O3), depending on the alumina content, could have an initial foam index of about 2 to 4 seconds at 1500° without any carbonaceous particles. When the slag surface was covered only 15 ~ 20 pct with either coke or coal char particles, the foam was totally suppressed regardless of the initial foam index. In order to understand the mechanism of the antifoam effect of the carbonaceous particles, interactions of a coke sphere, an iron ore pellet, an alumina tube, and a coal char particle with the liquid slag foam were examined by X-ray observation. It was concluded that the antifoam effect of coke or coal char particles is primarily contributed by the nonwetting nature of the carbonaceous materials with the liquid slag. Possible mechanisms of carbonaceous particles rupturing a slag film could be ( 1 ) the rapid thinning of the liquid slag film driven by a difference between the instantaneous contact angle and the equilibrium contact angle or (2) the "dewetting" of the liquid slag from the interface when the film is "bridged" by the particle.
Effect of Phosphorus on Solidification Process and Segregation of Directionally Solidified IN738 Superalloy
H.Q. ZHU, Z.Q. HU, Y.X. ZHU, S.R. GUO, H.R. GUAN, C.X. SHI, M. MORINAGA, and Y. MURATA
The effect of phosphorus on the solidification and the solute segregation of a directionally solidified IN738 Ni-based superalloy was investigated experimentally employing a method of partially directional solidification and subsequent quick quenching. It was found that the phosphorus addition widened the solidus-liquidus temperature interval of the alloy. Both the P content and the solidification rate affected the morphology of the solid-liquid interface and the precipitation of phosphides in the alloy. Phosphorus segregated largely in the intercellular/interdendritic regions and promoted the segregation of other alloying elements in the solidified structure.
phase. The densities of these precipitates controlled dislocation movement and regulated subgrain sizes. Prolonged heating resulted in peritectoid reactions to Al6Mn or Al12Mn. Recrystallization, which is associated with the formation of Al12Mn, is advanced by increasing the silicon content; the nucleation and growth of Al12Mn occurs only at the expense of other phases that stabilize the subgrain network.
Optimal Riser Design for Metal Castings
T.E. MORTHLAND, P.E. BYRNE, D.A. TORTORELLI, and J.A. DANTZIG
The optimal design of a casting rigging system is considered. The casting geometry is systematically modified to minimize the gate and riser volume, while simultaneously ensuring that no porosity appears in the product. In this approach, we combine finite-element analysis of the solidification heat-transfer process with design sensitivity analysis and numerical optimization to systematically improve the casting design. Methods are presented for performing the sensitivity analysis, including the sensitivity of important solidification parameters such as freezing time, temperature gradient, and cooling rate. We also present methods for performing Newton-Raphson iteration for solidification models that use the boundary-curvature method to represent the sand mold. Finally, the methods are applied to design risers for an L-shaped steel plate to control microporosity and for a steel hammer to control macroporosity. It is demonstrated that the size of a conventionally designed riser can be reduced by a significant amount while retaining the quality of the cast product.
| Search | TMS Document Center | Tables of Contents | Subscriptions | Met. and Mat. Trans. | TMS OnLine |
|---|