Search Documents

By Fabio J. Esper, Guillermo R. Martín-Cortés, Francisco R. Valenzuela-Díaz, Cassio P. Carvalho, Giorgio Grotti, Reginaldo B. Santos

"Rejected sands from smelting metallurgical process because contain lead and others heavy metals are a problem to dispose as waste in all the countries. A procedure that is being proposed for it is to incorporate it on common clay bodies in the fabrication of structural ceramic products as bricks, blocks and roof tiles. The main goal of this paper is to present the results obtained with mixtures of sands from smelting metallurgical process with common clay from Jundiaí, São Paulo, Brazil. Preliminary incorporation tests (1,0% in weight) were made with hand formed spherical samples. With the higher compression strength of the spheres after drying and firing (950ºC) experiments on bricks production were made for quality technological test. Seven hundred hallow blocks containing 1,0% waste were produced and presented ceramic properties similar to those without sand incorporation. Introduction In last the fifteen years the LMPSol - Raw Materials & Non-metallic Solids Laboratory – Metallurgic Eng. & Materials Depto. – Polytechnic School - University of São Paulo comes studying the possibilities of incorporate industrials waste materials to the production of bricks and other products of construction aiming at to remove pollution products of the environments and to user the properties of the same ones in the structural ceramics products. Examples are the study of the influence of rejected steel furnace dusts in structural ceramic [1], another example is the main subject of the present paper: use of rejected sand from smelting metallurgical process in the production of bricks an other products of structural ceramics. Antecedents More or less one year ago, specialist and members from the direction board of METALÚRGICA GROFE LTD. Company contacted researchers and professors of the LMPSol – PMT – EPUSP seeking for a solution to their concern with the disposal of rejected sand from smelting metallurgical process."

Jan 1, 2008

By Anette Mueller, Thomas Schnellert, Sérgio C. Angulo

"The use of automatically sorting technique, to separate gypsum components from recycled material of processed construction and demolition waste (C&DW), is motivated by environmental and economical aspects. Only if the gypsum can be separated the aggregates can fulfill the environmental requirements and also the structural characteristics according to the standards. Furthermore, the separation of red brick can also improve the quality of the aggregates.This paper aims at analyzing the separation of gypsum components and red brick by using the color sorting technique. The impacts on the composition and the envelope density of recycled aggregates are described. Depending on the gypsum content the application of two processing stages (rougher-cleaner) can be necessary to produce aggregates with sufficiently low gypsum content. Additionally, the separation of red brick and concrete is successful. So, the color sorting may be a promising technique to upgrade the quality of mixed C&DW aggregates.IntroductionConstruction and demolition waste (C&DW) is a mixture of different materials. Such mixtures can be used only in low grade applications. If the material is used as certified material for road construction or as aggregate in concrete the concrete particles must be separated from the other building materials. This paper aims at analyzing the separation of gypsum components and red brick by optical sorting. The needed stages of sorting for getting a product quality according to the German Standards are deduced.Gypsum is a building material that is increasingly used for interior works in Germany like it is shown in figure 1 [1]. At present the content of gypsum in C&DW is estimated at about 1 %. Assessments on the basis of material streams of gypsum and concrete and their life spans show that higher gypsum contents of about 3 % must be expected in future [2]. Also in other countries the use of gypsum has grown considerably in the last decades. From the characterization of C&DW available in some cities in Brazil, 4 to 15 % of total amount of C&DW that may arrive at the recycling plant to be processed is excessively contaminated by gypsum. In one of these cities, the average content of gypsum in the C&DW before processing was as high as 2.1 % [3]."

Jan 1, 2008

By Atsushi Shibayama

The aim of this research was to develop a copper recovery process from mine tailings (0.34%Cu) using flotation followed by high-pressure oxidative leaching (HPOL) and solvent extraction. The results of HPOL using the concentrate of mine tailings obtained by flotation under the optimal conditions of the previous study shown that an efficient copper dissolution of 94.4% was achieved in an H2O media, while the copper concentration of PLS reached to be 2.9 g/L. The solvent extraction of PLS obtained from the optimal HPOL showed that 91.3% copper was recovered in stripped solution under the determined optimum conditions, in which the copper concentration achieved to be 44.8 g/L. Finally, a proposed copper recovery process from the concentrate of mine tailings was developed by combination of HPOL and solvent extraction, while a total copper recovery of 86% was achieved.

The viscosity of high magnesium content FeO-Fe3O4-SiO2-CaO-MgO slags at nickel flash smelting conditions was investigated using a commercial rotational viscometer. The studied parameters were compositions of slag and temperature. The viscosity of these slags was measured as a function of CaO, MgO, SiO2, FeO. Fe3O4 and Fe/SiO2 in temperature range from 1473K to 1723K. These slags contained 5.81 to 10.59 pct MgO, 0 to 10.37 pct CaO, 28.21 to 36.21 pct SiO2, 39.40 to 55.42 pct Fe and 1.1 to 1.5 Fe/SiO2.

Jan 1, 1999

By Musa Karakus

Ceramic build-ups are synthetic mineral deposits formed in critical locations in or near the throats of inductor units in channel induction furnaces, the preferred melting and holding furnaces used in iron foundries. The build-ups are undesirable because they slow down melt circulation, increase power consumption and cost. and decrease the quality of the metal. This study involved the determination of the mineralogy of build-up samples and the quantitative measurement of phase abundance in those build-ups by cathodoluminescence microscopy. These undesirable deposits usually are hard and dense and contain variable amounts of entrained metallic iron. The non-metallic phases dominantly consist of oxides and silicates of several refractory elements, AI, Mg. Ca, Si, and Fe. Minor amounts of the sulfides of Ca. Mg, Fe and REE elements, especially Ce and La. and the oxides and silicates of Ce and La elements are locally present in the build-ups. The build-up samples were found to be ideally suited for study by cathodoluminescence, which has provided an easy means for rapid re-cognition of the phases present in the build-ups. The phase compositions were determined by electron microprobe analysis. The phases identified and their characteristic cathodoluminescence shown in the ceramic build-up samples were: spinel (bright green, blue-green, and red). forsterite (red and purple), periclase (dark reddish-brown, blue. and orange), corundum (red, blue, orange, and white), anorthite (yellow, yellow-green, blue, and blue-purple), hibonite or calcium hexaluminate (dark green), mullite (dull red-brown), gehlenite (deep blue), silica(dark brown, pink, purple, red. and blue), monticellite (light violet blue and light pale yellow-orange), silicon carbide (orange), oldhamite (bright yellow and orange). Ce-La silicate phases (purple), and glass (orange, dull red, and purple). Non-luminescent phases were: hercynite, Fe-oxides, Ce-La-oxides, Fe-sulfides, and metallic iron.

Jan 1, 1989

By E. V. Redin, B. A. Dubrovsky, E. V. Knyazev, A. I. Starikov

"Ti-V magnetite is one of the most abundant composite ore resources in the world. This ore cannot be fully processed via blast furnace technology because of high content of Ti02. According to the investigations carried out in the USA and Japan ITmk3 RHF (Rotary Hearth Furnace) technology is a breakthrough in Ironmaking. Reduction, melting and slag removal can be achieved in just 10 min. The main objective of the investigation is to establish optimum operation conditions for the production of iron nuggets from Ti-V magnetite ore via the ITmk3 by means of the lab scale testing. Samples of ore from the Maly Kuybas deposit (Southern Ural), carbon reductant and fluxes were chemically analyzed. Green pellets were processed via a lab chamber furnace to simulate RHF conditions. The nuggets and slag were chemically analyzed. This preliminary test work provides valuable information which may be used for large-scale testing in a commercially sized RHF.IntroductionThe ITmk3 process upon which this paper focuses was developed by Kobe Steel in its research facilities [l]. After pilot testing at Kobe Steel's Kakogava Works in 1999-2004 the pilot demonstration plant was built and operations were successfully carried out by Mesabi Nugget joint venture in 2002-2004. Steel Dynamics put into operation the commercial plant at the Mesabi Nugget site in Hoyt Lakes (Minnesota) in 2010. Results were consistent with nugget characteristics confirming ITmk3' s flexibility with respect to raw material inputs. At the same time, no research in the field of nugget from the Ti-V magnetite bearing ore with high content of Ti02 and Alz03 has been done. It is well known, if a lot of Ti02 and Alz03 enter the blast furnace, slag becomes viscous. The viscous slag is obstacle for stable operation of the blast furnace. Besides, high content of Alz03 raises the melting temperature of gangue minerals disturbing the coalescence of reduced iron to the nuggets."

Jan 1, 2013

By N. B. Gray, A. E. Guerra, D. E. Morton

"The formation of splash in pyrometallurgical operations that involve the injection of gases into a liquid has been the subject of a number of studies over the last few decades. Given the complexity of the phenomena, there have been some advances in understanding the microscale formation of splash but the ability to accurately predict the amount and character of the splash formed is still in its infancy. First, the macroscale mechanisms for splash formation is very varied. Second, the timescales for splash formation are very small, usually less than 100 milliseconds, making experiments elucidating the formation of splash difficult and costly. Work at the GKW CRC has shown that the microscale mechanism of splash formation is primarily is due to the Kelvin-Helmholtz instability. Substantial progress has been made in understanding the behaviour of large bubbles in effecting splash formation and a simple model has been developed to estimate the behaviour of splash formed from a top submerged lance. Direct numerical computation of splash formation using the GKMAC codes developed at the Centre has been successful in simulating the impact of liquid drops onto a liquid bath.IntroductionInjection of gases into melts is used in the metallurgical industry to promote mixing to ensure acceptable heat and mass transfer rates, conveying solids for alloying, provide efficient contact for gas-melt reactions, and as a reactant. High gas injection rates bring about the formation of splash and slop. Extensive splash formation leads to loss of valuable product, creation of safety hazards, formation of accretion on refractory lining and exhaust ducts, and increased load on dust removal equipment. Many operators and researchers have claimed that splash formation is the limiting factor in many gas injected processes. The problems listed above lead to loss of production and increased maintenance, adding to the cost of processing base metals. Over the last five years, the GKW CRC has been studying the formation of splash from a number of sources. They include the impingment of drops on liquids[ l·2.31, slopping of bath liquidI4.51, breakup of gas bubbles at the liquid-gas interfacel61, shearing of liquids by gases[7·81, and impact of packets of fluids on solidsl81. Theoretically, the extreme deformation of fluid shape and the influence of surface tension makes the simulation of flow resulting from only a drop impacting on a liquid surface a very challenging computational fluid dynamics problem. Recently, we have been successful in simulating the formation of the Rayleigh jet resulting from a drop impact showing the effect of drop Froude number on the shape of the Rayleigh jet."

Jan 1, 1996

By Shawn A. Cefalu, Matthew J. M. Krane, Kent J. VanEvery

"Segregation during electroslag remelting (ESR) of a nickel-chromium-molybdenum alloy (Ni-8 wt% Cr-25 wt% Mo) has been modeled. Features of the model include species transport and AC electromagnetic effects coupled with fluid flow, heat transfer, and solidification. The effect of varying ingot diameter and melt rate, on macrosegregation patterns, local solidification time, and liquid pool profiles obtained from this model are demonstrated. The results show that the macrosegregation in the ingots correlates to the filling velocity.IntroductionElectroslag remelting (ESR) is a secondary melting process used to refine the chemical composition and the microstructure of metals. Macrosegregation of solute elements causes defects in ESR ingots, including freckles and the maldistribution of phases. These defects are not correctable through heat treatment due to the large length scales over which they occur. However, the frequency and distribution of these defects may be minimized, or even prevented, through process control based on an understanding of the conditions that lead to their formation.In the past, studies have been conducted on the effect of the process parameters on the shape of the melt pool and the thickness of the mushy zone [1-7]. In these studies, the shape of the liquid pool is determined by consideration of the heat and fluid flow induced by electromagnetic (Lorentz) force and thermal buoyancy. The transport of solute and the resultant solutal buoyancy effects were not considered in these macroscopic models. Defect formation is tied to formation of secondary phases (e.g., freckling and white spots in nickel superalloys) [5,8]. In cases where secondary phase formation is calculated (e.g., [6]), phase fractions are determined using a micro-scale model given the local solidification time (LST) and local thermal history from the macroscale model.Using the local thermal history from the macroscopic transport model to predict microstructure increases the importance of the accuracy of those results. This investigation models the macrosegregation of molybdenum during the ESR of a Ni-Cr-Mo alloy. The effect of macrosegregation on pool shape and depth is demonstrated, along with its effect on the local solidification time. By comparing the relative effects of the melt rate and ingot radius, a better understanding of the factors that affect macrosegregation, and ultimately secondary phase formation may be obtained."

Jan 1, 2004

By I. Gaballah

The chalcopyrite ore's anodic and cathodic electrodissolution are studied using a Porous Pulsed Percolated Electrode (3PE) reactor. Stationary solid electrode experiments are also performed to explore the mechanism of electrochemical dissolution of chalcopyrite. Experiments with a low - grade natural ore have been carried out at 25 "C in solutions containing 1 M NaCl and 1 M HCl. Galvanostatic and potentiostatic conditions are employed. The influences of current and potential on selectivity of the chalcopyrite ore dissolution are assessed. Anodic oxidation of this ore leads to a preferential dissolution of iron bearing minerals, whereas cathodic reduction is more selective with respect to chalcopyrite. However, in both cases the electrochemical dissolution is incomplete and the current efficiency is low. Design of the actual reactor has to be modified to increase the bed conductivity and to allow continuous experimentations.

Jan 1, 1994

By Fumitaka Tsukihashi

Phase diagrams of the CaO-SiO2-FeOx-MgO, Al2O3 systems are necessary for the analysis of copper smelting reactions and sintering process of iron ore. The effect of MgO and Al2O3 addition on the liquidus of the phase diagram of the CaO-SiO2-FeOx system at low oxygen partial pressure was observed. The liquid phase area changed by the addition of MgO and Al2O3 with a changing oxygen partial pressure from 10-8 to 10-3 atm (10-1 to 102 Pa) at 1573K. The effect of Fe3+/Fe2+ ratio on the melting mechanism is discussed.

Jan 1, 2003

By Qingyu Deng, Shiwei Ma, Wang Hua, Guibao Qiu

"The flow properties of the blast furnace slag bearing high TiO2 at high temperature like the viscocity may change a lot due to the formation of the carbonitride inside, causing serious problems on the separation between the metal and the slag. Therefore, the critical condition of the formation of Ti(C,N) in blast furnace is vital importance for the ironmaking process with iron ore bearing titania. In this study, the formation of Ti(C,N) in blast furnace and the influences of various factors on the Ti(C,N) formation are investigated by FactSage. It is found that the mass of Ti(C,N) would reach the maximum at 1480 oC, and the product is mainly TiN below 1620 oC, while it is mainly TiC above 1620oC. The order of the factors by the importance from strong to weak was: temperature, the content of TiO2 in slag, basicity of slag, the content of MgO and Al2O3.IntroductionThe titanium dioxide (TiO2) in iron ore bearing titania will be partly reduced by carbon to titanium carbonitride (Ti(C,N)) in blast furnace, causing serious problems on the separation between the metal and the slag because of the change of flow property. Therefore, the critical condition of the formation of Ti(C,N) in blast furnace is vital importance for the ironmaking process, especially for the smelting of iron ore bearing titania. Extensive experimental data in regard to the formation of Ti(C,N) has been reported in numerous papers[1-6]. However, only some discrete data can be obtained by experiment, and it is difficult to detect the mass of Ti(C,N) in slag. In addition, the experiment is difficult to be proceeded when the temperature is above 1600oC. Therefore, it is very difficult to make clear the critical condition of the formation of Ti(C,N) by experiment."

Jan 1, 2012

By Metehan Erdogan, Orhan Gökçe Göksu, Ishak Karakaya

"An experiment (mixture) design was created to optimize the process parameters of the electrochemical reduction of CaW04 to W in molten CaCl2-NaCl eutectic mixture. Temperature, applied voltage and the length of Kanthal wire winding of the CaW04 pellets were selected as the process parameters that were allowed to vary between the predetermined minimum and maximum values. The rates of the electrochemical reductions were interpreted from the variations of current and total charge vs. time graphs under different conditions. The analysis pointed out that 640°C and 2.81 V resulted in the fastest reduction. Itwas seen that the effect of Kanthal wire winding on the output current was less pronounced when compared to the other two parameters.IntroductionFFC Cambridge Process [1] has been applied to a number of metal oxides in laboratory experiments to produce metals. In nearly all of these studies, the electrolysis temperature was around 900°C, the electrolyte was pure CaCl2 or CaCl2 containing salt solutions and the anode was graphite. However, it was recently shown that the electrochemical reduction process of metal oxides in CaCl2 containing electrolytes at the mentioned high temperatures was not suitable for tungsten production [2], It was reported that as soon as WO3 contacts CaCl2, the following reaction occurs:"

Jan 1, 2012

By Florian Aurelian, Mihai Popescu

"It will be tested four types of “geochemical barriers”. All those mixtures were contacted with mine waters containing uranium, radium and heavy metals. The loading capacity was determined. The optimum mixture was determined and its efficiency was tested within an uranium areal, where the activity was stopped. It was determined the possibility to recover the radioactive elements and the heavy metals from the investigated barriers.IntroductionThe paper presents the results concerning the performances of some “geochemical barriers” used for the decontamination of mine waters from uranium mining sites, which were tested in lab and in situ.ExperimentalThe research activity was carried out in lab and on the site.In the lab a module was constructed in order to follow up the uptake capacity of the “permeable reactive geochemical barriers” used to remove the radiolelements from the contaminated waters originated from the uranium mining industry. The barriers, which were used, are made of cheap natural materials, which can be easily got (wood chips, peat coal, sawdust, iron span, steel and zeolitic tuff). The following types of barriers were tested:•, 1st barrier: peat coal, sawdust, zeolite on weight ratio of 1/1/1;•, 2nd barrier: bentonite, sand, wood chips, peat coal on weight ratio of 1/1/1/1;•, 3rd barrier: iron span, steel, sawdust, peat coal on weight ratio of 2/1/1;•, 4th barrier: iron span, steel, sand, firry wood chips on weight ratio of 1/1/1.Lab researches concerning the uranium uptake capacity of the above described “reactive barriers” pointed out that the 3rd mixture (iron span, steel, sawdust, peat coal on a ratio of 2/1/1) didn’t reach the maximum uranium uptake capacity not even after eleven contact stages (the uptake capacity being of 70.6 mg U/ kg of mixture) and the mixture of peat coal, sawdust and zeolite has an uranium uptake capacity of 26.6 mg U/kg of mixture and becomes saturated after four contact stages.Each one of the components involved in physical-chemical processes of sorption, reduction and precipitation, will determine the diminution of uranium and other metals contents upstream the permeable reactive barrier."

Jan 1, 2008

By Li Ma, Shaoguo Chen, Gele Qing, Qihang Liu, Keng Wu

Olivine contains higher MgO element, which can be added into blast furnace burden to improve slag stability and fluidity. Effect of calcined olivine and untreated olivine on low silicon pellet production process and quality was studied in this paper. The results shows, calcined olivine has no effect on bursting temperature of green pellet, however the bursting temperature of green pellet will decrease with increase of untreated olivine ratio. Untreated olivine has slight effect on green pellet strength than calcined olivine while the ratio is higher than 2%. The compressive strength of low silicon pellet will increase first and decrease afterwards with increases of calcined olivine ratio under firing temperature of 1250°C, whereas, the changing sequence is reverse fired under 1280°C. The result of untreated olivine pellet compressive strength is same as calcined olivine pellet. However, the compressive strength and metallurgical property of calcined olivine pellet is slightly higher than that of untreated olivine pellet under same condition.

Jan 1, 2015

By Raj P. Singh

Two design-of-experiments studies were carried out to investigate the leaching of technical grade molybdenum tri-oxide (TGM) in the presence of hydrochloric acid and the mother liquor of ammonium di-molybdate crystallization used as an ammonia source. Results indicated that leaching of TGM in HCl is possible only when used at about 1N concentration. In high concentrations of HCl, technical grade MoO3 had very high solubility (7.5 to 40 g/L Mo). The most suitable conditions for HCl leaching of TGM were obtained at a slightly elevated temperature of 80°C. Using about 1N HCl at 80°C, potassium in the leached solids (TGM) decreased to about 110 ppm, a value typically obtained in TGM leached in HNO3 - NH4NO3media (3.5 - 4.0 moles /L HNO3 + 0.5 ? 1.0 moles/L NH4NO3). Mo losses under these conditions were 0.8-0.9 g/L Mo. The HCl-leach method, presented here-in, has a significant advantage in chemical cost and environmental benefits over HNO3 ? NH4NO3 purification method.

Jan 1, 2006

By Carla Lupi

In this work, nickel-cobalt electrodeposition under galvanostatic conditions was investigated. Ni-Co alloy was electrowon from an acid sulphate bath containing 40 g/L Ni and2-8 g/L Co.. The effect of electrolyte composition, current density, temperature, solution pH and presence of buffers such as H3BO3 and (NH4)2SO4 on Co content of the binary alloy has been investigated. In particular, current densities ranging from 180 to 400 A/m2 and temperatures ranging from 27 to 60°C were tested. The composition of Ni-Co alloys may be diversified through the selection of process parameters. The experiments show that the highest cobalt content of the alloy is reached by using a buffer free electrolyte without solution pH control. Though, in these conditions, the most homogeneous, fine and regular deposit morphology is obtained, energy requirements are the highest in tests lasting just two hours.

Jan 1, 2003

By Ragnhild E. Aune, Mark William Kennedy, Shahid Akhtar, Jon Arne Bakken

"To determine the heating rate of billets using ‘short coils’, an appropriate correction factor must be applied to the theoretical relationship. In 1945, Vaughan and Williamson published a semi-empirically modified Nagaoka coefficient applicable for moderate frequency induction heating processes (10 kHz). Recently it was demonstrated that the method of Vaughan and Williamson gives <10% error in the estimated power when heating aluminum billets at 50 Hz.In the present study, experiments have been conducted on aluminum billets in order to verify an empirical frequency corrected ‘short coil’ equation. Measurements of electrical conductivity (<± 0.5%), current (± 1%), heat (± 1-3%), and magnetic flux density (± 1-2%) have been performed. The results are compared with 1D analytical calculations, and 2D axial symmetric FEM modeling using COMSOL 4.2®. The frequency corrected equation has proven to provide accurate predictions of power (<4% error) within the frequency range 50 Hz to 500 kHz.IntroductionInduction heating is commonly applied to the re-heating of billets before forging or extrusion processes. The resistive heating produced by eddy currents in the solid or semi-solid work piece during this procedure, is driven by the time varying magnetic flux density in the air-gap between the work piece and the coil. The flux in the air-gap is created by the current flowing in the induction coil, i.e. the magneto-motive force. In Figure 1 it can be seen that the currents present in the work piece are concentrated in the outer ‘shell’, or the first electromagnetic penetration depth (dw), and flow in a direction which opposes the magnetic field produced by the induction coil."

Jan 1, 2012

By Florian Kongoli

The behavior of cobalt as a by-product in the NilCu smelting and converting processes effects directly the quality of the final products as well as the revenues gained from the overall process. An effective way to evaluate and predict this important behavior is thermodynamic modeling which reduces the time and the cost of experimental investigation. In this work a thermodynamic model of cobalt-containing matte phase in NilCu smelting and converting processes is presented. The model predicts a priori the thermodynamic properties and phase relations of the ternary and multicomponent matte based solely upon the optimized parameters of the binary systems. This model can be used with other models such as those of slag, metal, gas etc. to predict various aspects of cobalt behavior such as matte- slag, metal-slag distribution, etc.

Jan 1, 1999

By R. H. Lien

The Bureau of Mines, U. S. Department of the Interior, is investigating chemical and biological decontamination of complex wastewaters such as tailings pond water containing 280 ppm CN and 5 ppm Se as well as significant concentrations of arsenic, copper, iron, silver, and zinc. The most effective chemical procedure involves cyanide oxidation using hydrogen peroxide or sodium hypochlorite followed by selenium reduction using ferrous hydroxide. The effluent contains <l ppm CN and 20 ppb Se; concentrations of other major contaminants meet EPA drinking water standards. Preliminary research indicates that biological cyanide oxidation is possible even in highly alkaline solutions (pH 10.5) containing high cyanide concentrations; indigenous bacteria destroyed 85 pet of the cyanide. Exploratory research shows that other bacteria removed up to 79 pet of the selenium from chemically oxidized, low-cyanide water. These promising results may lead to a final processing scheme that involves a combination of chemical and biological techniques.

Jan 1, 1990

By J. P. Bellot

During the non-ferrous metal processing, the melting step under vacuum or under rarefied gas leads to volatile metal losses. It is the case of the metallurgical processing of the nickel-based, titanium-based or zirconium-based alloys where the evaporation losses don&apos;t allow a sufficient control of the alloy chemical composition. A theoretical approach has been carried out involving both a continuum description in the liquid phase and a molecular approach in the rarefied gas phase. The finite volume method is used in order to solve the coupled transport equationsin the liquid, whereas the Boltzmann equation is modelled by using a Monte Carlo method. At the liquid metal and metallic vapor interface, the volatilization and recondensation fluxes link up the two calculations. Simulation results are given for the melting and refining of reactive metals such as titanium alloys where the control of the chemical composition remains a major difficulty. The validity of the kinetic Langmuir&apos;s law is discussed for different operating conditions.

Jan 1, 1996