Parallel computation technique for mold filling simulation of die casting process by optimizing the computational parameters

摘要

Mold filling simulation for the die casting process is a very important aspect of CAE analysis to predict the molten metal flow pattern and to optimize the die design for the process. One of the effective means to improve the computational efficiency and scale is the implementation of parallel computation technique. Based on analysis of the mathematical model and the computational algorithm for the mold filling simulation, some computational parameters that influence the efficiency in the SOLA-VOF algorithm were summarized by numerical experiments. A parallel search algorithm was developed and implemented on a cluster computation platform to optimize the efficiency parameters by searching the most efficient parameters dynamically. The verifications on the mold filling simulation of practical die casting components show that the optimization of efficiency parameters can greatly improve the computational efficiency to some extent while maintaining the computational accuracy.