Simulation Of Dynamic Recrystallization Process Of Metallic Materials By Cellular Automata Method

CA (Cellular Automata) method has been more and more mature in modeling the non-liner problems and complex dynamic system. It has been applied in many science fields. In materials science, CA is particularly outstanding in simulating the microstructure and its development.The 2-D and 3-D models of dynamic recrystallization (DRX), which combined with the fundamental theory of metallic using CA method, were built up in this paper. The model coupled the heat process parameters at the macroscopic scale with the dislocation development at the mesoscopic scale. The dynamic recovery, nucleation rate, strain rate, and dislocation density were in considered. The nucleation and growth of dynamic recrystallization grain (R-grain) could be monitored. The development of dislocation density and fraction of DRX could be calculated. This model could also display the shape, distribution and size of R-grain. The simulation results described the kinetic rule of DRX well. Main conclusions are as follows:1. The 2-D model succeeds in simulating the growth process of equiaxed grain. The original microstructure obtained with this model is similar with typical microstructure after anneal, and the microstructure of DRX obtained accords with typical DRX characteristic.2. Strain and strain rate typically influence the microstructure of DRX. At large strain, the R-grain is finer when strain rate is rising. While at small strain, with the increase of strain rate, the transition of DRX was more inadequate, sometimes, no DRX transition.3. The mean R-grain size R_d is independent of original grain size. It has been provedtheoretically that R_d is determined by Zener-Hollomon parameter uniquely in this paper. When T was decided, R_d was decreased with the rising of ε.4. A 3-D model is successfully developed on the foundation of 2-D model. The 3-D model can describe the Avrami equation well also, and its transition speed of DRX is faster than 2-D model.5. The influence of different nuclei type on simulation with 3-D model has been discussed. The results indicate that the model of nucleation at boundary with nucleation rate determined by strain rate and particle simulation nucleation (PSN) model are suitable for simulating DRX. And what's more, the PSN model gains finer grain.