Novel Cellular-Automata (CA) based modelling and optimization for recrystallization and grain growth in materials - theory and applications

The mechanical properties of alloys depend on their microstructure, or grain sizes, which are obtained during processes of the grain growth and recrystallization. However, experiments do not allow to observe all details of the process and there is no theoretical model to predict these. This is why computer modelling allows for the prediction of microstructure. Cellular Automata (CA) is known to be used to for the simulation of the grain growth. However, (CA)-based models usually work only on the specific range of parameter, because there is no linear relationship exists between cellular automata and the physical parameters.

This thesis describes a newly developed space and time realistic Cellular Automata technique for the solving of various differential equations. It has been shown that the model is capable of solving the associated differential equations, which describe grain boundary movement, but with the errors that are related to the square grid. The newly suggested type of neighbourhoods allowed to eliminate the effects of the grid and obtain better prediction results.