Search Documents

By C. J. Martin, M. P. Hay

"A methodology was developed in the mid-1980’s to predict plant performance from standard laboratory flotation tests. The technique is based on a simple, empirical kinetics model, and is tailored for use by the practical metallurgist.To date the performance of more than 20 flotation plants has been predicted, including copper, lead, zinc, phosphate, pyrite, graphite, cassiterite, platinum, nickel and various slags. The method has successfully been used to predict and size four flotation plants directly from bench-scale laboratory flotation testwork. All have since been commissioned, and have performed as predicted.The technique has now received broad acceptance within the South African mining industry, for example in the last year alone it has been used to drive flowsheet development for at least seven new plants. It is also being used to assist in the optimisation of existing operating plants.The methodology is described in this paper and verification of the methodology is presented using case studies of porphyry, polymetallic and PGM operations.INTRODUCTIONThere is considerable interest these days in the use of modelling to simulate flotation circuits, for process design and plant optimisation purposes. Although these models are proving to be useful tools for plant design and optimisation, their complexity can make them cumbersome – expensive to use and difficult for metallurgists to visualise and mentally absorb, without a background in each specific technique. In “SUPASIM” no attempt is made to measure flotation fundamentals, rather to use empirical relationships between laboratory flotation kinetics and plant performance – relationships that hold up well to validation exercises as will be demonstrated later in this paper. Although not designed to be as comprehensive as other models, “SUPASIM” is easily understood by the operating metallurgist and can be quickly and easily applied."

Jan 1, 2004

Ferrochrome is the major chromium source for stainless steel production, and submerged-arc furnaces (SAF) are most commonly utilised to smelt chromite ores into ferrochrome by using suitable carbonaceous reductants such as coke. Because of the complexity of feed structure and electrical-thermal-chemical interactions, there exist large gradients and wide distributions of temperature, mineralogy and other process variables in the furnace. This leads to various zones in the furnace, and different reduction mechanisms. Control of the furnace is thus often difficult due to the complicated physical and chemical situations in the furnace. The temperature profile in the submerged arc furnace has a great influence on the reduction rate and the production efficiency. Therefore, the energy generation through electrical arcs submerged in the solid feed and its distribution within the packed bed play an important role. In the present paper, an overall process model is illustrated to simulate inter-linked transport processes of fluid flow, heat transfer and reduction kinetics within the submerged arc furnace. The application of reaction kinetic data to simulate the chromite pellet/lumpy ore reduction with CO gas is presented and further adjusted to provide a submodel for a flow and heat transfer simulation for the packed bed by using computational fluid dynamics (CFD) technique. The comprehensive process model is aimed to simulate materials and gas flow, temperature and energy distribution throughout the furnace system. A preliminary model for gas flow and temperature distribution for the upper part of a 20 MW furnace (the packed bed) for high carbon ferrochrome production has been developed under simplified conditions. The integration of the reaction kinetic model into process simulation, and the future vision for a more comprehensive process modelling are discussed. It is expected that better understanding of the furnace insights through process modelling will help process optimisation and furnace control.

Jan 1, 2004

A model has been developed for drum scrubbing which allows selection of geometry and power input. Specific aspects considered are: Calculation of total static and operating volumes. Total residence time based on total operating volume. Solids residence time based on solids operating volume. Selection of the aspect ratio. Alteration of operating fill to allow for smaller or larger feed openings. Selection of proportion of critical speed. Calculation of shell and motor power requirements. The model uses either solids or total residence time to calculate the required operating volume. The solids residence time is based on the proportion of coarse solids in the feed. Material finer than the reject size is considered as part of the slurry. The solids operating volume is the static volume plus the surcharge on the discharge weir required for flow of the solids from the unit. The total operating volume is the static volume plus the surcharge on the discharge weir for the total flow to exit the scrubber. The burden density is calculated from the coarse solids bulk density and the slurry density that includes all the fine solids. Power is calculated using the Liddell and Moys correlations for burden position and shape, and the operating rotational speed. Results from the model (especially power) have been verified from industrial installations of significant size. Predicted and observed power draws are within two per cent. A check against the JKMRC mill model shows power draw prediction agreement to within four per cent.

Jan 1, 2004

By Sri C. Sawmliana, P. Pall Roy

Damage to underground roof and pillars in blasting gallery (BG) panels has been an increasingly important area of study in recent years due to the invention of new explosives and cord systems which has led to the modification and alteration of the design aspects developed by Charbonnage de France (CdF). According to extensive mine experiments conducted by Indian and foreign experts, it has been established that the induced peak particle velocity (PPV) is predominantly related to the emanated seismic energy of a source and the distance from the source. With that concept in mind, nearly 160 experimental blasts were carried out in four BG panels in India under varying geomining conditions. Based on the results obtained thus far, the authors have established predictor equations correlating explosive energy (in Joules) and distance from the shot to that of induced peak particle velocity received at the sensor point. Good correlation was obtained between the actual and the predicted values. It is realized that this equation may be used as an efficient tool to determine the size limit of blasts in BG panels so that the induced particle velocity does not create damage in underground workings as per the established Blast Damage Index (BDI) and other available safety norms.

Jan 1, 2004

By Y. Mao, D. Xie

The viscosity of TiO2-containing blast furnace slags may increase rapidly (commonly referred to as thickening) under strong reducing conditions in the blast furnace. To elucidate the mechanisms for slag thickening, measurements have been made of changes in the composition and rheological properties of TiO2-CaO-SiO2-Al2O3-MgO slags during the reduction process. The slag composition covered the range of 17–33 wt% TiO2, 9–19% Al2O3, 8–16% MgO and theCaO/SiO2from 0.8 to 1.4. The viscosities were measured using the rotating cylinder method at temperatures up to 1550ºC and the fusibility temperatures were determined from ‘breaks’ in the viscosity-temperature curves. The reduction of the slag was carried out in graphite crucibles at1500ºC for 2–6 hours The rheological properties of the slags during reduction were predominantly influenced by the slag basicity (CaO/SiO2), and the effect of other components depended largely on basic or acidic behaviour of the oxide. The reduction of TiO2in the slag by carbon was initially fast to formTi2O3, followed by further reduction to TiO and TiC. Correspondingly, the viscosity showed initially a slight decrease, followed by a gradual and then rapid increase. The thickening was found to be accelerated by an increase in TiO2content and a decrease in the CaO/SiO2in the slag. In addition to the existence of 1~3 wt% TiC, the thickened slags also had a significant amount of TiO (3~7 wt%), suggesting that TiO may play an important role in the ‘thickening phenomena’ ,possibly by forming solid solutions with TiC to significantly increase the number of solid particles in the heterogeneous slag. Keywords: TiO2slags, viscosity

Jan 1, 2004

By R. J. Fruehan

Slags have many useful functions in iron and steel making. In particular, slags are used to remove impurities such as sulphur, phosphorus and inclusions from the metal, and a large amount of research has been done in this area. Slags also insulate the metal thermally and form reactions with the atmosphere such as reoxidation and nitrogen pick-up. However, slags serve other more unique functions, several of which are reviewed in this paper. The role of slags in new iron smelting processes with regards to reduction of ore and gaseous desulphurization are discussed. In iron smelting processes, much of the reduction occurs by the reaction of carbon in the char and the metal suspended in the slag with iron oxide dissolved in slags. The slags also react with the gas resulting in sulphur removal from the process. The mechanisms of these reactions and pilot plant data for these processes are given in this paper. Slag formation within ferrous burden materials affects the cohesive zone and the performance of the blast furnace. Recent theoretical and experimental findings on this topic are presented. In the electric arc furnace, foamed slags are used to allow for higher power operation and productivity. Of particular interest, foaming in stainless steel production is more difficult than for normal steels. The foam index, or the ability of the slag to foam, for stainless steelmaking is similar to that for carbon steelmaking slags. The poor foaming is due to the lack of gas generation for the foam. Methods to increase foaming in stainless steelmaking are discussed.

Jan 1, 2004

By P. Dennis, T. Pieters, A. De Kievit, S. Ganguly

TEMCO has been producing manganese ferroalloys in submerged arc electric furnaces for over 40 years. The furnace linings all along have been of conventional insulation type until 2001 when a freeze lining of UCAR concept was installed for a 20 MW Furnace. Continuous improvement of furnace lining performance has been a key component in improving the business competitiveness of TEMCO. The expected minimum campaign life of the freeze lining is fifteen years as compared to the ten years of the conventional linings. The success of the freeze lining is very critical to the quanta and direction of heat flow through the lining that control the freeze protection. Under freeze protection can lead to reduced lining life and over freeze protection, on the other hand, can lead to operating difficulties and loss of smelting efficiency. It was, therefore, necessary to develop a computer based system to monitor and control the freeze protection. The computer based monitoring and control system uses the basic principles of heat transfer, activeX object and heat flux from dual thermocouples and that is reviewed.

Jan 1, 2004

By L. A. Cramer

South African platinum mines produce the world?s majority of platinum metal that is used in autocatalysts and jewellery. As the platinum market is forecast to grow at healthy rates into the future, the South African producers, Anglo Platinum, Impala, Lonmin and Northam, are planning to expand in the next decade. Furthermore, a number of smaller producers who are accessing other mineral rights are entering the buoyant market. The South African platinum mines have traditionally exploited the Merensky Reef of the Bushveld Igneous Complex (BIC); however, the mining of this reef is reaching depths that require significant vertical shaft systems and costly refrigeration for access to the orebodies at deeper levels. In the past decade all of the producers have begun to mine the less attractive UG2 reef because of its cheaper mining costs. This UG2 reef contains significant quantities of chromite with typically a chrome-to-iron (Cr/Fe) ratio of 1.35. UG2 ores can contain between 10 and 25% Cr2O3dependingupon the reef width and the selected mining method. It is relatively cheap and easy to extract a portion of the chromite from a platinum concentrator tailings stream at a grade of between 40?42% Cr2O3. However, it is only in recent years that some ferrochrome producers have begun to source some of their chromite from the tailings of platinum producers. South Africa is also the world?s largest single producer of ferrochrome and supplies much of the world?s stainless steel producers and thus the growing market for their product. As the traditional sources of high Cr/Fe ratio ores have been depleted at the shallow economic depths and thus have become more expensive sources of chromite, technical developments have improved the ability to use fine chromite economically to make acceptable grades of ferrochrome. Thus UG2 fines are now becoming possible cheap sources of acceptable feed to the ferrochrome industry. The abundance of feedstock for ferrochrome production and the lessening dependence on mined material will impact upon the structure of the ferrochrome business in South Africa. With continued platinum mining from UG2 reefs, South Africa is set to remain highly competitive in the world?s ferrochrome/stainless steel industry.

Jan 1, 2004

By M. Sharafodin, A. A. Hassanipak

Abstract - The essential aims of additional borehole drilling are to improve the reliability of grade and tonnage estimates in each reserve class and to increase ore tonnages. The "GET" function presented in this paper considers strategies for achieving both of these goals simultaneously, and therefore is advantageous for selecting sites for additional boreholes. The "GET" function is either a linear or a non-linear product of three variables G, E, and T: f(G,E,T,) = G??E??T??where the values of any or all of the exponents ?, ?, and ??may differ from unity at the discretion of the user. G and E are the average estimated block grade and the average estimation error for ore blocks in one vertical column, and T is the compounded ore thickness within the column. To illustrate its utility, the GET function has been used for determination of the most advantageous sites for additional drilling in the Shah-Kuh Pb-Zn deposit in west central Iran.

Jan 1, 2004

By J. Madias, E. Benavidez, E. Brandaleze, L. Castella

Knowledge of the physical and chemical properties as well as the behaviour of the mould fluxes during melting, important in order to improve continuous casting operations. Lubrication and control of heat transfer in the mould are the more relevant functions of these materials. Values of physical properties like viscosity and surface tension at melting range temperatures are required, as much as the thermal behaviour under process conditions. The heat extraction from the liquid steel has an important effect in the final quality of products. Mould fluxes are constituted by different oxides, and may form glasses and crystal phases during cooling. The solidification rate and the chemical composition determine the proportion, type and crystalline morphology, which are responsible of the thermal transfer processes. In this paper, viscosity values are reported in the melting range temperatures calculated by means of the Riboud model and compared with fluidity results obtained by the inclined plane technique. Also the linear expansion of the mould fluxes at high temperatures was determined by dilatometry. The crystallization in two different fluxes was determined by DTA tests and correlated with the basicity index. Crystal phases were evaluated by scanning electronic microscopy on quenched samples at 1300°C. This information is relevant to increase the knowledge on the peritectic steel grades casting process in order to improve its quality. Keywords: mould fluxes, heat transfer, crystal phases, peritectic steels, thermal properties.

Jan 1, 2004

By J. Hop, A. Støre, T. Foosnæs

The carbon cathode for aluminium electrolysis is also part of the container of the process. In order to understand the behaviour and possible failure of the cathode, detailed material properties under operating conditions are needed. The paper reports experimental studies of bath and sodium in filtration, graphitization with time, loss of strength due to sodium, sodium expansion as function of pressure, as well as creep due to the developed stresses. As an example, sodium and thermal expansion data are used to calculate strain-stress models of realistic cell designs, and possible design weakness can be pointed out. Keywords: carbon, cathodes, sodium expansion, strength, graphitization, creep

Jan 1, 2004

With the continued implementation of stricter limitations on the discharge of excess process water from all existing and future mining and metallurgical operations, the design and operation of the entire process water circuit must form an integral part of any total operational package. It is far better to æanticipate and circumvent the problemÆ rather than be forced into an expensive remediation program after the event. By far the greatest problem relates to the treatment and disposal of what is generically known as acid drainage. In simple terms, an acidic metal sulfate solution is generated by a series of chemical and biological reactions as soon as most sulfide minerals are exposed to air and water. Treatment of this solution by one or more of a wide range of chemical and physical options can involve substantial capital and operating costs, although there are recent processing options that yield saleable by-products, thereby reducing the financial liability of what is effectively a waste stream. This presentation provides an overview of the Australian situation compared with the rest of the world and outlines the basic features of a number of the more advanced passive and active acid drainage options.

Jan 1, 2004

This paper examines the integration of risk management into the design philosophy of a project to produce a æzero baseÆ design, which targets a minimum capital cost without adversely impacting upon the operating costs of the resulting plant. The concept of zero base design is similar to zero base budgeting and the desired outcome is to improve the economy of the design. Design requirements are questioned in an attempt to distinguish preference, prejudice and myth from necessity. Where doubt remains a risk management approach is adopted. The key to risk management is not risk elimination, but risk mitigation. Risks can be managed by designing them out, but often this involves considerable capital cost penalties and it is more attractive to accept the risk and design an operating strategy to deal with it. The paper discusses this with some examples of real world risk mitigation strategies that resulted in substantial capital cost savings. The paper concludes that as much as a quarter of project capital costs can be eliminated by a combination of zero base design and carefully crafted risk mitigation strategies.

Jan 1, 2004

By Y. I. Rabinovich, S. C. Brown

In the Florida phosphate industry, fatty acid collectors are used to separate phosphate ore from silica during the rougher flotation stage. However, the presence of multivalent ions (in particular Ca++) in the processing water has been reported to result in the activation of silica with respect to oleate adsorption, rendering it floatable. Hence, adequate separation may not be achieved during the rougher flotation stage. In this study, the direct measurement of surface inter-action forces via atomic force microscopy was used to delineate possible mechanisms for calcium activation under phosphate flotation conditions at oleate/calcium concentrations both above and below the calcium dioleate precipitation limit.

Jan 1, 2004

By J. A. Kromhout, J. Trouw, A. A. Kamperman, M. Kick

Mould powders play an important role in the stability of the continuous casting process. The main functions of mould powders are to provide sufficient lubrication and to control the horizontal heat transfer. Because process control of thin slab casting is more stringent, it is generally assumed that mould powder properties become more critical as well. In 2000, the Direct Sheet Plant at Corus IJmuiden (DSP) was commissioned, incorporating a thin slab caster designed for casting speeds up to 6 m/min. During commissioning, several operational problems with mould powders were encountered i.e. insufficient liquid pool depth and the formation of rims/lumps. To investigate the operational mould powder problems, some of the mould powders were characterized using physical and mineralogical methods. The results were linked to operational experiences. It was concluded that for thin slab casting, a stable slag formation is obtained with a proper choice of free carbon and an even distribution of the carbon particles within the granules. The physical properties of mould powders are not as critical as expected when the additional mineralogical requirements are taken into account. Keywords: mould powder, characterization methods, thin slab casting.

Jan 1, 2004

By Eugene Nel, Johan Theron, Chris Martin, Hans Raabe

"Impala Platinum Limited is the world’s second largest platinum producer, producing more than 1.7 million ounces of the precious metal last year, much of this from the UG-2 reef located in the Bushveld Complex, South Africa. As part of Impala’s ongoing expansion programme, tonnage through their principal UG-2 processing facility, near Rustenburg, was increased 40% in 2001.To accommodate the increased throughput, the role of the existing primary mills was converted from conventional semi-autogenous grinding units, to attrition and selective grinding units, operating in fully autogenous mode. The finer chromite and coarser silicate components in the mill discharge are then separated by screening and processed in dedicated circuits downstream. This way, the need to spend considerable capital to increase primary grinding capacity was averted.In the last year, considerable work has been conducted to evaluate and optimise the two downstream circuits. This has involved an optimisation programme of sampling and diagnostic flotation testing, supported with QemSCAN analysis. This paper will describe the operation of the new circuit. It will also describe some of the more novel aspects and interesting findings from the optimisation programme.INTRODUCTIONThe South African platinum industry is the world’s largest producer of platinum and the world’s second largest producer of palladium. Located in a wide area north of Johannesburg, the industry mines ores from the Bushveld Complex, one of the world’s great mineral treasure chests, containing over 70 percent of the world’s platinum, as well as extensive reserves of the other PGE (platinum group elements), chrome, manganese and vanadium. The industry employs about 180,000 people in, predominantly, underground mines, which together form an almost unbroken circle of mining operations 200 km in diameter."

Jan 1, 2004

By D Craw, C Smith, C Rufaut

Oceana Gold (NZ) LtdÆs, Macraes Gold Operation currently has two tailings impoundments, the larger of which covers approximately 75 ha. When an impoundment is temporarily out of use, the surface dries and dust can be blown from the surface to the surrounding area. At present water spraying is used to stabilise the surface, but this is not ideal as Macraes mine can experience water shortages during the summer. This paper examines an alternative method of surface stabilisation using vegetation as a ground cover. The tailings have a low natural fertility, so it is important to determine what will actually grow in the tailings, and whether amendments are necessary. Also important is the effect of the amendments on arsenic mobility in the tailings as the schist rock at Macraes has naturally elevated levels of arsenic. Three plants species, rye corn, barley, and lupin, were grown in metre-deep lysimeters at Macraes mine site. The plants were grown in unaltered tailings, and tailings with one of two amendments added: Bioboost (an organic fertiliser containing biosolids), and superphosphate fertiliser. The leachate from each lysimeter was collected and analysed monthly to determine the extent of arsenic mobility. In addition, laboratory experiments were set up to determine arsenic mobility in tailings with amendments at the centimetre scale.   All plants grew best with the fertiliser amendment, with rye corn growing the best overall. Plants growing in unaltered tailings either barely survived (rye corn, barley) or died (lupin). Plants in the Bioboost amended tailings grew only slightly better than in the unamended tailings. Laboratory experiments indicate that the addition of phosphate fertiliser greatly increases arsenic leaching at the centimetre scale when compared to unamended tailings. The addition of Bioboost only marginally increases arsenic leaching, compared to unamended tailings. This trend has not been repeated in the lysimeter trials on the metre scale over four months. Current results suggest that of the plant types trialed and amendments used, rye corn with superphosphate is the most practical short-term tailings stabilisation method at Macraes.

Jan 1, 2004

By R. Dahlke, J. A. Finch

A campaign was undertaken during the greater part of 2001 at Noranda’s Brunswick mine concentrator to investigate the role of gas (air) superficial rate (Jg) distribution to the cells in the final Zn cleaning stage (bank of seven cells), i.e. the Jg profile. Three profiles were tested: “balanced” (each cell with same Jg), “increasing” (from cell 1 to 7 proportionally), and “decreasing,” along with an “as found” profile in a parallel control bank. The experimental design allowed for two confounding factors, changes in ore type and circulating load. The increasing profile consistently gave the best metallurgy (down-the-bank grade / recovery). The reason for improved performance was the response of the first cells in the bank where the increasing profile gave the highest selectivity of sphalerite against nonsulphide gangue (NSG). The other profiles gave lower selectivity and higher zinc recovery in the first cells, which combined to reduce overall bank performance. The higher selectivity with the Increasing profile is attributed to the low gas rate in the first cells reducing water recovery and hence limiting entrainment of NSG. The increasing profile was subsequently configured for all four stages in the zinc cleaner circuit.

Jan 1, 2004

By P. Wells

"In the past decade, there have been many developments in processing of nickel/copper sulfide ores. Improved pentlandite/pyrrhotite flotation selectivity with the use of new reagents, fine regrinding and Redox measurement (control) has led to increased pyrrhotite rejection and thus higher concentrate grades. Flash flotation has increased overall nickel and associated precious group metal recoveries. Recovery of both fine and coarse particles has been improved through an understanding of flotation cell hydrodynamics. The separation of copper from nickel in such ores has seen a similar level of development. From the late 80s to very recently, the use of flotation columns, the use of pyrrhotite depressing reagents and the improved understanding of the complex mineralogy of chalcopyrite, pentlandite and pyrrhotite have considerably increased copper / nickel separation efficiency. Copper concentrate with more than 30% copper and less than 0.5% nickel grade is now obtainable at 80-85% copper recovery from mill feed compared to 70-75% copper recovery with 30% copper and 1% nickel grade only a few years ago. Improved copper / nickel separation in milling operations has the potential to substantially improve the processing of the resulting concentrates and thus the overall return from the extraction of nickel and copper from sulfide ores.INTRODUCTIONNickel metal by itself or in alloys with other metals and materials makes a significant contribution to modern society. The relatively high price that nickel commands compared to other base metals is a reflection of the fact that, although abundant in the universe, nickel is sparingly distributed in the earth’s crust. There are two types of nickel deposits that are currently being exploited: complex sulfide ores and lateritic oxide ores. Lateritic contain no copper but in sulfide deposits, nickel– bearing minerals, principally pentlandite (Pn, [Ni,Fe]8S9), some millerite (NiS) and violarite ([Ni,Fe]S) are often found in association with copper-bearing minerals, mainly chalcopyrite (Cp, CuFeS2) and cubanite (CuFe2S3). The copper content of such ores can vary considerably from as high as 5-8% (in some Sudbury basin ores) to as low as 0.1% (in ores mined by Western Mining at Mount Keith, Western Australia and by Inco in Thompson, Manitoba). The valuable minerals Pn and Cp are almost invariably found associated with pyrrhotite (Po, Fe7S8) which along with nonsulphide rock has to be rejected to tails in the primary milling stage. The bulk concentrate from the primary stage is then treated to depress Pn and float Cp. A complete Cu/Ni separation at the mill would be highly desirable but this is not possible because of locking between a small fraction of the valuable minerals. As a result, Cu/Ni separation has to be carried out both during milling and during refining of the matte following smelting. At Inco, Jinchuan and in Norilsk, a chalcocite / heazlewoodite separation is made by flotation following slow cooling of the matte. In the present Sudbury Inco operations, Cu/Ni separation has not been conducted at the mill since 1991 and all Cu/Ni separation takes place at the matte processing plant. Because of the rising ratio of Cu to Ni in the ore, this strategy is under review and this paper describes some recent studies of circuits in preparation for possible installation of a copper / nickel separation circuit at Inco’s Clarabelle Mill in the Sudbury area."

Jan 1, 2004

In ilmenite smelting, the FeO content of the feed material (nominally FeO.TiO2) is decreased by reduction with carbon. In a parallel reaction, a significant amount of the TiO2 is reduced toTi2O3. One of the striking features of these reactions is a consistent relationship between the extents of FeO and TiO2 reduction; this relationship is independent of furnace size. It was recently proposed that this relationship follows from the solidification behaviour of the slag, where the final slag composition corresponds to the minimum melting point (?eutectic groove?) in the FeO-TiO2-Ti2O3 phase system; the hypothesis is that solidification equilibrium with the freeze lining stabilizes the slag composition. If this is the case, the system of slag and freeze lining must be capable of correcting two types of disturbances: thermal and compositional. Thermal disturbances result from incorrect power input, for the feed rates of ilmenite and reductant. For example, excessive power input will tend to overheat the slag, causing the freeze lining to melt partially, establishing a new steady-state heat loss from the furnace. Compositional disturbances result from an incorrect reductant input for the given power and ilmenite input. In this work, the dynamic response of the slag and freeze lining to different inputs (flow rates of ilmenite and reductant, and power input) is modelled. The model is based on solidification equilibria, and quantified heat loss through the freeze lining. The conclusion is that the solidification equilibrium cannot constrain the composition to the eutectic groove for typical operating conditions. Keywords: ilmenite smelting, freeze lining, solidification, high-titanium slag

Jan 1, 2004