Investigation on dynamic recrystallization using a modified cellular automaton

To predict and to control the microstructural evolution during dynamic recrystallization (DRX), a modified cellular automaton (CA) model based on mathematical statistics theory and physical metallurgical principles is developed. Initial microstructure and thermo-mechanical parameters are used as input data to the CA model. Dislocation density is used as a crucial internal state variable to link microstructural evolution with macroscopic flow stress. The latter two are output data, which can be compared with experimental one. In order to exhibit the effect of deformation stored energy on DRX, both the nucleation rate and the growth velocity of each recrystallizing grain (R-grain) are calculated from the dislocation density. The growth kinetics of R-grain is calculated from the metallurgical principles, and the nucleation kinetics is evaluated from a statistically based dislocation-related nucleation model. Model parameters are identified by a flow stress-based inverse analysis method, and then their variations with thermo-mechanical parameters (strain rate and temperature) are estimated and integrated into the CA model. The good agreement between the simulations and the experiments demonstrates the availability and predictability of the modified CA model.