The fluid mechanics of near-net-shape or thin-strip casting
Steel is traditionally cast in ingots which are massive pieces of solid steel. The ingots are then turned into steel plate or other forms by rolling and other processes. This method of production has several drawbacks. The casting has to be carried out in batches and the ingots are clumsy to handle. Continuous processes for the casting of steel have been in use for a few decades. They yield a material that is easier to handle in its subsequent processing. This results in fewer rolling stages and an increased yield, which in turn leads to better economy and lower consumption of energy. Near-net-shape or thin-strip casting is a continuous casting process that has been developed during recent years and is already being used for the production of sheet, plate, pipes and rods. This process is presently going through considerable further development. Compared to earlier continuous processes, near-net-shape or thin-strip casting offers advantages such as: a short processing chain, which results in low costs for investment and a short through-put time. Near-net-shape or thin-strip casting results in a significantly higher yield compared to casting of ingots or earlier forms of continuous casting. The higher rates of cooling and solidification affect the material's segregation, which makes it possible to process materials with totally new properties. Experimental and theoretical investigations of near-net-shape or thin-strip casting will be carried out in collaboration with ABB Industrial Systems. This company will finance an Associate Professorship (at 20% of full time) at the Department of Materials Processing, KTH.

Near-net-shape or thin-strip casting can be carried out in many different ways. An example in which fluid mechanical phenomena play a significant role is the Twin roll casting method for casting sheet. In this method the molten material or melt is supplied continuously to a reservoir or tank. Solidification takes place as the melt flows out through a narrow gap between two rapidly rotating rollers in the bottom of the tank. After passing between the rollers the plate is cooled with air or water. The structure in the plates, which determines among other things the properties of the material, as well as its final surface finish depend to a great extent on the flow of the melt in the entrance to the gap between the rollers. The flow between dendrites in the solidifying shell during the process of solidification also has an effect on the structure. Experiments on models of this type of casting, as well as on other types, are being carried out at the Department of Materials Processing, KTH. Theoretical calculations of the transport of heat and mass in the turbulent flow field in the melt as well as of the phase change in the process of solidification will be started at the Centre. The L.E.S. method will be used in this case as well. The goal of the theoretical investigations is to achieve the ability to predict the properties of the plate and in this way make optimization of the process possible.

Manufacture of 'in-situ' composites
The technology of metallic composites has attracted a great deal of interest within both Natural Sciences and Engineering Sciences for some time. There are a large number of methods for the manufacture of metallic composites. Since the superior properties of these composites can be considered to be well-known, it will suffice just to point to the possibility of tailoring the material to have the pre-determined properties of resistance to wear, thermal stability, thermal expansion and so on.

One promising method for the manufacture of metallic composites consists of allowing a molten alloy of, for example, iron and titanium, to come into contact with solid graphite. Diffusive transport of carbon into the molten alloy leads to a reaction between the carbon and the titanium which results in a precipitation of small, almost cubic, crystals of solid titanium carbide. The variations in the mass density of the melt, which the reaction causes, start a convective motion in the melt. Since diffusion is a relatively weak transport mechanism, the rate of production of the crystals will be determined by the convective transport of the carbon and the titanium. The ability to control the particles' volume fraction, the distribution in the matrix, the rate of the process and its yield thus requires quantification of the convective and diffusive transport in the melt. Experiments are in progress at the Department of Materials Processing, KTH. These experiments will be complemented within the Centre by mathematical modelling and numerical calculations with the aim of developing an industrially useful tool for process design.

Hardening of steel
In the mechanical engineering industry, heat treatment is a central process in the manufacture of high-performance components such as bearings, gears and sledge-hammers. These products obtain their mechanical properties as a result of the phase changes which take place during cooling (quenching or hardening) after heat treatment. The most commonly used cooling agents are oils and suspensions in water. In the sense used here, cooling includes conversion of water to steam, i.e. boiling which is an instance of multi-phase flow, and turbulent convection. The use of a gas under high pressure is becoming more and more common in cooling, primarily because of its advantages for the environment. Trials with fluidized beds have yielded promising results. The activities of the Centre within the hardening of steel will be carried out in collaboration with the Swedish Institute for Metals Research at which theoretical and experimental research is being carried out under the supervision of the Technical Council of the Swedish Mechanical Engineering Industry.

One of the problems in hardening that is the most difficult to cope with and most costly consists of an uncontrolled deformation of the component due to thermal stresses. Consequently it is essential to be able to control the temperature distribution during hardening. In computations of the hardening process, an estimate has to be made of the transport of heat between the component and the cooling agent, which involves fluid mechanics, as was mentioned above.

Mathematical models for numerical simulations of the way in which the variation of the temperature in time and the phase changes in the steel affect the properties of the components, such as residual stresses, hardness and shape, have been successively developed during recent years at the Swedish Institute for Metals Research as well as at other Swedish institutions. The investigations have mainly concerned transmission components (gears), bearings and tools for forging and plastic injection moulding. A serious limitation of these models is that the transport of heat from the component to the cooling fluid is only described by empirical relations. This means that the usefulness of the simulations has so far been limited to idealized situations which are not in general typical of industrial production. At the Centre, realistic computational methods will be developed for the heat transfer between component and cooling fluid. These methods will complement the existing simulation models for the transport of heat, the phase changes and the mechanical response within the component. The resulting computational model will constitute both a unique and a powerful tool for controlling hardening processes.