Research will develop numerical simulation-based shrinkage prediction criteria for the optimization of casting design of commercial ductile and gray iron castings. The initial focus will be on ductile iron. Shrinkage prediction for this important commercial alloy is complicated not only by the complexity of solidification behavior, but also by its extreme sensitivity to metallurgical processing conditions. Thus, an essential requirement for the development of practically useful design software is the development of a kinetic model of cast iron solidification which realistically accounts for the effects of metallurgical processing on mode and sequence of the solidification. A combined micro-modeling/thermal analysis approach to solificaction modeling will be developed, along with an algorithm for applying the solidifaction model to the estimation of volume changes during freezing on the basis of conservation of energy and mass, and thermodynamic considerations. The ultimate goal of the research is the development of an integrated computer-based methodology for process control, optimization of casting design, and diagnosis of casting defects which takes into account the interdependent effects of metallurgical variables, molding factors, and geometry on casting quality. Castings are designed in industry by a process that involves much trial and error, since experience is the only available tool. This project involves the development of a design tool which incorporates current understanding of the metallurgy of the cast metal and additional insights that result from the research project. Therefore, parts can be designed more precisely, reducing the number of failed attempts.