AbstractSolidification modelling appears to be well developed with regard to simulating macroscopic aspects of the freezing of castings. Therefore, it is most valuable for the identification and solution of feeding problems. One unsolved problem in solidification simulation is that of due allowance having to be made for the heat transfer at the casting/mould interface, where the gap which develops can vary widely over a casting. Any good solidification model will have to take account of this important effect. A further unsolved problem is the redistribution of heat within the solidifying casting by convection in the liquid. Where this predominates, the redistribution of heat affects the location of defects and the effective feeding of castings. Despite excellent progress, it is evident that considerable problems remain with regard to castings. Usually these are filling problems, leading to the incorporation of the surface oxide into the bulk metal by surface turbulence. Computer modelling of this area has only just started. The work being carried out at The University of Birmingham to define good filling regimes is described. As regards simultaneous flow and solidification, modelling is at a relatively primitive stage. The fluidity of pure metals and complex /alloys will not be easy to differentiate accurately. To simulate the phenomenon of remelting in aflowing stream, which the at/hor has termed‘continuous fluidity’, requires accurate modelling. It describes important effects such as the situation in running systems or the gating of a heavy casting through a thin section, in which, despite some early freezing, remelting of the solid allows the flow path to remain open indefinitely. Other aspects of casting, namely, stress, distortion, and the creation of defects, remain on the fringe of current modelling capability and will require new data.MST/1481