Sami Vapalahti

The effect of Fluid Flow on Heat Transfer and Shell Growth in Continuous Casting of Copper

Molten metal is cooled in a continuous casting mould forming initially a thin shell that grows thicker. The main phenomena in the mould are: fluid flow, heat transfer and solidification. A lot of mathematical models have been developed to simulate these phenomenons in continuous casting machines but most of the models developed are not calculating the fluid flow at all. In these models, it is assumed that the strand (solid and liquid) is withdrawn through the machine with a constant velocity field (= casting speed) and the convective heat transfer generated by the fluid flow is taken into account by using an effective thermal conductivity method. Also at the Helsinki University of Technology, these kinds of heat transfer models have been developed (TEMPSIMU for steels and CTEMP3D for coppers). The flow in the mould is three-dimensional and turbulent. Coupled models calculate the fluid flow, heat transfer and solidification simultaneously. The fluid flow is affected by many things: inlet flow rate, design of the inlet nozzle (SEN), immersion depth of the SEN, movement of the solid shell, natural convection, solidification shrinkage, etc. and the fully coupled, turbulent fluid flow and heat transfer models are generally subjected to convergence difficulties and they need a lot of computing time. Due to these reasons, these kinds of models are not so much used in industry so far. In the present study, a commercial FLOW-3D package is used to make coupled simulations of heat transfer, turbulent fluid flow and solidification in a copper continuous casting machine. The effect of thermophysical material data are also studied and presented. The material data are calculated by a model developed at the Helsinki University of Technology, called CASBOA.

Experimental Work on Possibilities to Predict Casting Defects in LPDC Brass Castings

In water tap production 0.5 mm of material needs to be ground off from the surface of LPDC (Low Pressure Die Casting) brass castings in order to remove the defects deteriorating the quality of later applied coating. In order to minimize the amount of removed material the causes of these defects need to be discovered and properly connected with the process parameter window. At Oras Oy foundry in Finland, nearly 100 castings were produced under actual process conditions. To monitor the process seven thermocouples were inserted into the die. Thermal camera was also used for monitoring the die conditions during the open time of the die. Castings were divided into sets of ten pieces for statistical reasons. A few key process parameters were selected based on the basis of earlier knowledge and they were systematically varied during casting experiments. Each cast piece was marked and later analysed in order to find the dependencies between detected defects and process parameters. Computer simulations of the process were conducted to study the possibility to use numerical simulations for defect prediction. It was found that shrinkage defects could be reasonably well predicted and the influence of the process parameters on their formation was also apparent. The predictability of surface defects, however, was poor and only indirect conclusions could be made. Observations were made using as cast, ground and polished and cut surfaces from certain sections of the castings. It is very difficult to make any conclusions on surface defect formation based on parameter variation. One reason probably is the too narrow process window, but several promising ideas on the influence of e.g. mould shape, temperature and composition of the graphite coating on the defect formation was discovered.