A fully 3-dimensional model for an inductively-stirred ladle is presented. Since the fluid flow during induction stirring is not symmetric with respect to the centre of the ladle, this model gives a more accurate description of the real stirring conditions than more traditional models that use cylindrical or polar co-ordinates. Typical output from the model are steel velocities, turbulent kinetic energy and dissipation of turbulent kinetic energy. The calculated steel velocities agree fairly well with experimental data. © Munksgaard, 1997.

In 1982, SSAB, Luleå Works, converted the blast furnace ferrous burden from an acid to an olivine pellet with a high Fe content. This change made it possible to gradually reduce the slag volume to 150 kg per metric ton of hot metal with huge positive effects on the blast furnace operation. SSAB, Oxelösund Works, can present a similar development, initially by increasing the Fe content in sinter and later by converting to a 100% olivine pellet operation in 1995. In addition to a description of the historical development of the properties of SSAB blast furnace slags, this paper describes the technical means and issues as well as the environmental constraints governing future developments of slag volume and chemistry, including effects of changes in burden quality and possible tuyere injection of slag formers. © Blackwell Munksgaard, 2002.

Within the framework of a joint Swedish-Soviet co-operative project, refining of impurities in steel has been studied. The work was split in two parts: 1) refining of copper from steel melts; and 2) phosphorus removal under reducing conditions with subsequent slag destruction. Refining of copper was carried out in the laboratory and in one pilot trial. The laboratory tests showed notable copper removal with sodium-based reagents, but the consumption of reagents is very high in addition to the necessity for an extra desulphurizing operation. A few lab-tests with soda-based (Na2CO3) reagents show promising results which could lead to a more economical procedure. Phosphorus removal under reducing conditions followed by destruction of the slag has been performed for 100 kg and in a pilot study. In the pilot study a method has been developed with controlled destruction of the slag in the ladle before removal of the phosphorus-containing slag. Including the destruction, the total dephosphorization reached 35%.

The ongoing struggle to reduce the production costs in steelmaking is one of the important driving forces for R&D engineers. By combining new technical solutions from different areas with new ideas for steelmaking, it is possible to find new cost effective processes for steel production. During the last decade, great efforts have been made worldwide to develop new techniques for near net shape casting of steel. Most of these projects have dealt with production of steel strip. This paper describes the work carried out at MEFOS to develop the DSC (Direct Strip Casting) process. The process is based on the horizontal single belt principle. Based upon the experience to date, we can conclude that belt casting is a technology suitable for high productivity demands, e.g., for mass production of low carbon steels. © Munksgaard, 1997.

SSAB Lulea, SSAB Oxelosund, MEFOS and Lulea University of Technology have cooperated in developing theoretical models for prediction of thermal stratification and drainage in steel ladles. Predicted results have been verified by measurements done in production ladles with heat weighs of 105 and 200 tonne as well as in a 7-tonne pilot plant ladle. The thermal stratification was measured by thermocouples at different depths. The drainage flow was studied by means of tracer experiments. Numerical simulation models of the steel flow were developed for 7, 105 and 200 tonne steel ladles. The agreement between predictions and measurements was found to be good. The conclusion is that CFD simulation gives a good prediction of stratification and drainage flows in production ladles.

The combined movement of steel, slag and gas in a ladle has been studied with the aid of a multiphase CFD model and measurements In production scale. Comparison of model, data and measurements In a production ladle has been carried out for three cases: The 'open eye' created during gas stirring, the mixing of top slag during gas stirring In the CAS-OB and finally for studying sculling phenomena. It was found that the simulated 'open eye' appears as a central slag-free zone with a surrounding 'splash zone' characterized by the fact that steel is found on top of the slag. The simulated 'open eye' corresponds to the 'splash zone'. The predicted gas stirring in CAS-OB gives an efficient stirring of the top slag also. This was confirmed by noting that slag samples taken from the ladle seem to be representative of the mean slag composition. Finally, streamlines during stirring have been compared with sculling and wear observed In the production ladle. Good agreement was obtained. © Blackwell Publishing, 2003.

The meltdown behaviour of pellets reduced under different conditions in a reduction under load furnace has been examined. The pellet type, temperature increase rate, the gas flow and the applied load over the crucible were varied between trials. Samples from 12 trials interrupted just before meltdown were studied under an optical microscope and it was found that all iron oxides in the samples were reduced to iron. After the etching of the samples the carbon content was estimated. The carbon content differs within the samples as well as between the samples. A correlation exists between the average carbon content in a sample and the meltdown temperature. Some differences in the iron structure and the carbon content between the crucible wall and the crucible centre are discussed. Because no big difference in carbon content was found between the crucible wall and the crucible centre most of the carbon in the samples probably dissolves in the metal by a gas reaction where CO gas forms carbon and CO2.

Gas stirring of steel in the ladle is widely used in metallurgical industries to homogenise chemical composition and temperature, enhance reaction rates, etc. This study focused on the homogenisation, i.e., the mixing of alloy additions and steel during gas stirring. More specifically, the influence of porous plug location on alloy mixing into steel was studied theoretically. Flow fields were calculated and presented for simultaneous stirring with 2 porous plugs in 4 different arrangements. The alloy mixing into steel in the ladle with these different arrangements was studied with 2 different alloy addition methods. In the 1st method, the alloy was added in a batch at the beginning of the stirring. In the 2nd method, the alloy was added continuously throughout the stirring. When comparing the calculated results for the 4 different pairs of porous plug placements, it was found that centric stirring and centric addition rendered the best average mixing and also the least likelihood for "dead rooms" to develop. However, the combination of centric stirring and addition was found to be extremely sensitive to small variations in alloying position. It was therefore concluded that for steelmaking, the best of the studied alternatives is using 2 porous plugs placed such that their respective radii form an angle of 60°. Verification of calculated steel velocities was done with experimental data taken from a previous publication.

A research program with the objective of studying the separation of non-metallic inclusions is currently being undertaken by the Swedish Steel Producers' Association. The aim is to develop a pc-based model that can easily be used by the participating steel companies. Important inputs to the model are, among other parameters, data for steel velocities and dissipation of turbulent kinetic energy.

Within the metallurgical industry as well as the underground mine industry, there is a need for on-line level detection in the different production processes and their material feeding systems. Microwave technology is a versatile, non-invasive measurement technique, which has a number of advantages compared with penetrating probes, laser and radar technology, and echo sound technique. In its extension, it could also be used for qualitative and quantitative analysis of waste gases due to the presence of specific molecules. This contribution describes the introduction of microwave technology in different metallurgical as well as underground mine production processes with the emphasis on EAF's.

Asymmetric rolling conditions often cause front-end bending in the plate mills. In this work, experiments were performed to study how different parameters influence front-end bending in plate rolling. The parameters studied were: the influence of different peripheral speeds on the top and bottom work rolls, different reductions, a thermal gradient across the thickness of the plate and the position of the roller table in relation to the neutral axis of the roll gap. The results were evaluated statistically and a good model was obtained. Finite-element simulations were made to simulate ski-end behaviour for the different plates. A comparison with the experimental results was made.

In order to increase the understanding of continuous casting, the expert committee in casting and solidification started a project called 'Mobile Measuring Unit for Continuous Casting'. The committee works within the Swedish Steel Producers Association (Jernkontoret) and has participating companies from the Nordic countries. This paper describes results from measuring campaigns performed at industrial slab casters. The topic is mould function and mechanical behaviour of slab casters.

At MEFOS, various metal working problems have been analysed using the finite element method. The general nature of FEM also makes it possible to analyse the rolling of complex section profiles. The aim of this project was to investigate the precision of the FEM calculations, including the rotating boundary calculations developed at MEFOS, and the possibility of using FEM as a tool in computer-aided roll-pass design.

Mathematical models of inclusion behavior in stirred ladles are useful both for increasing our fundamental understanding of the growth and removal of inclusions as well as for future use in process control. This study reports on some efforts to use both static and dynamic modeling to better understand inclusion behavior in gas-stirred ladles. A computational-fluid-dynamics mathematical model of a gas-stirred ladle was developed earlier. In the investigation covered in this report, instantaneous fluid-flow results from the model were used in combination with inclusion growth and removal theories in order to study the importance of bubbles on inclusion flotation. The study results proved to be highly dependent on the theory used to describe bubble flotation. The model of the gas-stirred ladle was also used together with the inclusion theories to study the transient behavior of inclusions during growth and removal. The dynamic simulation results indicated that inclusion concentration gradients exist. The most important research task in the near future is to verify static and dynamic modeling results of inclusion behavior during stirring with experimental data. Here, the authors feel that carefully performed plant trials could provide useful information. © Blackwell Munksgaard, 2002.

Since the early 1990s, mathematical modelling has supported metallurgical research at MEFOS. In addition to thermodynamic calculations and trials in various scales, the commercial software PHOENICS has been used to describe the transport of mass and heat. Steady-state models, as well as transient models, have been developed for single-, 2- and 3-phase systems. In some applications, chemical reactions between the phases are also incorporated. Modelling activities have focused on the process route of steelmaking, including melting, secondary metallurgy and continuous casting. In this paper, highlights from some of these modelling activities are presented, including verification measurements. Process improvements are suggested, as well as some implications for treatment strategies and practices. Finally, future work and planned research are discussed.

Some aspects of inclusion behaviour in the tundish have been investigated both theoretically and experimentally. Good agreement was obtained between measured and predicted temperature and flow fields for 1- to 6-strand continuous-casting tundishes. In this study the flow field was redesigned with weirs, resulting in the addition of a vertical component to Stoke's equation. The results indicate an increase in the velocity that cause a rise of inclusions (smaller inclusions (<20 μm)). Consideration of slag, flux and refractory in the model has also made it possible to simulate the mixing of steel and slag. Special sampling techniques were used to gather information. Samples were analysed using ultrasonic testing, LOM, SEM and Atomic Force Microscopy (AFM). The analysis results were used to verify the predictions regarding steel/slag mixing and understanding of physical conditions at the interfaces. As a result, the casting praxis was improved (cleaner steel) and the products were of higher quality. © Blackwell Munksgaard, 2003.

The purpose of this study has been to theoretically evaluate which inclusion growth mechanisms are important in an inductively stirred ladle. This has been done using data from a computational fluid dynamics model of a real ladle. The data was utilized as the input to the different collision equations. It was concluded that diffusion of oxygen and deoxidant to the inclusion surface, diffusion coalescence and Brownian motion collisions contribute very little to growth of inclusions during the stirring period in the ladle. This was in accordance with earlier findings in the literature. This study also showed that laminar shear collisions could be excluded from a growth model. The major growth mechanism is turbulent collisions. It was also shown that a 700-A stirring current leads to more growth due to turbulent collisions than a 500-A stirring current. The importance of a correct value of the collision efficiency coefficient for turbulent collisions was pointed out. If a large difference in inclusion size exists then collisions due to difference in buoyancy (Stokes collisions) also need to be considered in a growth model. Finally, it is concluded that the variation in fluid flow and turbulent properties in different parts of the ladle should be incorporated into a growth model, since it can affect average turbulent collision volumes by 25-30%.

The main objective of cold rolling is to reduce the thickness of hot rolled strips to the desired final thickness. The process control must reach the target for several other factors too, as regards geometrical, mechanical, chemical and surface properties. During recent decades, process control has taken advantage of new measurement devices, new control actuators and the rapid development of information technology. Automatic process control can improve the quality beyond what is possible by manual control. Today, rolling of flat products needs advanced process control to increase the productivity and to reduce the variations in the final properties. A typical cold rolling stand performs one step in a chain of processes in the cold rolling mill, which can include pickling, rolling, annealing, temper rolling and downstream processes. All these processes contribute to the final properties of the strips. When the main process or the main objective is well controlled, it is important to continue with the other processes. In continuous annealing furnaces, the temperature controls the mechanical properties, but temperature differences and bending around rollers change the flatness. Temper rolling needs the same flatness control as other cold rolling processes. Cooling and lubrication can affect several properties of the strips.

Different empirical and hybrid modelling techniques were evaluated in order to improve the modelling compared with the present situation in the plate rolling mill at Rautaruukki, Raahe. Logged data from the production were used to develop new models and to compare them with each other and with the present model. The data describe the chemical composition and various process parameters of the rolling process, including thermal, geometric and kinematic parameters. Various neural networks and multivariate models were evaluated in order to take non-linear effects into account. Hybrid methods use a combination of physically based models and empirical models. The simple physically based formula, force = area x yield strength factor, proved useful. A more advanced model solving von Karman's differential equation was evaluated as a stand-alone model and as part of a hybrid model. The measured force of the previous rolling pass was tested as an additional input variable. Various aspects of the different modelling techniques are discussed and compared. Simple but improved models are described as suitable candidates to replace the present on-line model. © Blackwell Munksgaard, 2002.

Extensive researches on slag reduction for recovery of valuable metals and oxide materials from metallurgical slags and other wastes using a DC furnace with a hollow electrode have been conducted in 2 major EU projects with close cooperation between MEFOS, FEhS and CRM. Steel slags and other residues were introduced into the reactor through the hollow electrode to the hot plasma. The materials were melted, reduced and mixed. The final products were a metal product, a slag product with targeted chemical composition and a dust fraction with a high content of ZnO. Different steel-making slags and residues including BOF slag with low and high V-content as well as EAF and AOD slags from stainless steel production, EAF dust, oily millscale, hydroxide sludge, BOF and BF dust have been treated. The slag products include a metallurgical powder for desulfurization of steel, hydraulic binder and slag stones for construction applications. The metals obtained are rich in Fe, Mn, V and Cr depending on the treated slag and residues. The environmental compatibility and mechanical properties of the slags have been improved after slag reduction. © Blackwell Munksgaard, 2003.

The phase relations in the Ti-Ca-Si-O system were studied between 1300 and 1600°C. In the Ti-Si-O sub-system at 1300°C Ti2O3 coexists with Ti5Si3 and SiO2. Also, the silicides Ti5Si4, TiSi and TiSi2 as well as metallic silicon coexist with SiO2. At 1600°C Ti5Si3 can coexist with TiO + Ti2O3, with Ti2O3 + SiO2 and with SiO2 + Ti5Si4. Liquid Si-Ti alloy with up to 40 atom% Ti coexists with SiO2. In the CaO-Ti2O3-SiO2 sub-system there is a ternary phase Ca3Ti2Si3O12 with garnet structure and a = 12.165±0.001 A as well as the binary compounds Ca2Ti2O5 and Ca8Ti6O17. The two last compounds were found to be members of the solid solution series CaTiO1+x and C4Ti3O4+3x where the former has the perovskite structure. The metallic phase which can be in equilibrium with all three-phase combinations in the CaO-Ti2O3-SiO2 system was found to be the compound Ti5Si3 (m.p. = 2125°C) with only small variations in its silicon content. In order to obtain liquid metal and slag by silicothermic reduction of titanium oxide at 1600°C an excess of silicon must be used to give a silicon-titanium alloy with 40 atom% (52 wt%) Ti, or less.