Microstructure simulation of aluminum alloy using parallel computing technique

Abstract

The formation and evolution of the microstructure of casting are important research areas in the field of material science and engineering. The solidified microstructure of aluminum alloy was simulated by combining the CA (Cellular Automaton) model with macro heat transfer. A modified CA (MCA) model, which uses a more similar shape to the actual dendrite to describe the growth grain, was proposed and studied. Because of the huge computational capacity to simulate the microstructure of casting, a relevant parallel computing technique based on the serial arithmetic was developed, which can greatly improve the computing scale and efficiency and can also ensure the computing accuracy as well. The simulation results are compared with the experimental results and agreed quite well.