Mean field and full field simulations of dynamic recrystallization of austenitic 304L steel

During hot deformation of austenitic 304L steel several physical mechanisms take place at different scales which may lead to the full regeneration of the microstructure. The level of plastic deformation, recovery and recrystallization actually defines the final microstructure. Therefore, in order to control the different microstructure features (e.g. grain size, recrystallized fraction and dislocation density), it is of prime importance to be able to simulate those mechanisms and to link them to the thermomechanical conditions. In the present work, we propose a full field model with explicit description of the microstructure and a mean field model with implicit description of the microstructure. In both models, plasticity, recovery and nucleation are modeled with phenomenological laws. As for grain boundary migration, it is driven by stored energy gradients and capillarity effects. Hence, the migration velocity of each grain boundary is affected by the neighboring grains. Since microstructure is explicitly described in full field simulations, the neighborhood can be directly determined and grain boundary curvature can be locally calculated. However, in mean field simulations, the microstructure is statistically described. The novelty of the present model is that neighborhood of a given grain is implicitly described through statistical topological laws. The comparison of the results of both models validates the proposed statistical neighborhood description. Simulation results are also compared to experimental ones.