Physical Modelling of Cu-Sn Casting

In the casting of real metals the opacity of the melt and the high temperatures involved make it difficult to measure, or even observe, the flow patterns resulting from different metal delivery devices. Consequently, mathematical and physical modeling have been extensively used to determine the flows in casters, particularly those of steel casters. While mathematical models have been widely accepted, their performance has frequently been judged by their ability to predict flows measured in physical (water) models. In the case of steel casting, physical modeling of the appearing flow fields in the liquid melt is often performed by using water models and particle image velocimetry or similar flow field measurement techniques. The most interesting region for this type of investigation is the region near the nozzle and industrial research groups test new nozzle designs with PIV measurements near the submerged entry nozzle. The physical models referred to generally either ignore the formation of the solidification front or just take it into account by changing from a purely cylindrical geometry to a conical geometry. For steel casting this is a good approximation because the final solidification point, i.e. the point where the casting product is fully solidified is several meters away from the submerged entry. There are, however, other metals and alloys where the solidification front is much closer to the entry nozzle and has therefore to be taken into account in both numerical and experimental modeling of the casting process. A recent work at the Department of Materials Science and Engineering, University of California, Berkeley has shown that in the case of Al direct chill casting the distance between entry of the liquid metal and the solidification front is in the order of 0.4 m. Recent calculations and cooperation with industrial partners of the applicants home institute revealed that in the case of Cu-Sn alloys this distance is in the same order of magnitude. Therefore, if a physical model of the flow fields appearing in Cu-Sn casting processes is to be realistic, it has to imply the distance and shape of the solidification front. For the physical modeling of an Al casting, the group in Berkeley simplified the parabolic shape of the solidification front by introducing a pyramid covered with porous material. This fixed shape was based on calculations for a fixed geometry and casting speed. The model showed a good agreement with a numerical model and provided the ability to test the influence of different nozzle geometries on the flow field, however, it could not cover the influence of different shapes of the solidification front. The aim of the proposed project is to adapt the existing model and measurement setup in Berkeley to the modeling of Cu-Sn casting and to improve it in order to investigate the influence of the position and shape of the solidification front on the flow field. The final model should give the experimentalist the opportunity to change the shape of the solidification front to the one predicted by a numerical model and compare the resulting flow field with the numerical prediction. If new numerical simulations predict a different shape of the solidification front due to a change of the casting speed or an improvement in the numerical model, the flexible setup will provide means to check the new results with the corresponding velocity field measurements in water. Apart from the ability to test different numerical predictions a systematic study of the influence of the shape of the solidification front on the flow of the liquid melt will improve the understanding of the entire process.