Interactions Between Dynamic Recrystallization and Dynamic Precipitation During Hot Forming of Magnesium Alloy: Multilevel Cellular Automaton Modeling

Yu-Ying He,Li-Ping Lei,Gang Fang
DOI: https://doi.org/10.1016/j.jma.2024.06.018
IF: 11.813
2024-01-01
Journal of Magnesium and Alloys
Abstract:Accurately tailoring microstructures, especially grain size, during thermomechanical processing is crucial for achieving the desired strength-ductility synergy of wrought magnesium alloys. This study establishes a multilevel cellular automaton (CA) model to predict the microstructure evolution of wrought magnesium alloys undergoing both dynamic recrystallization (DRX) and dynamic precipitation (DP), surpassing the capabilities of traditional DRX models. Multiple physical metallurgical mechanisms, including variations in dislocation with work hardening (WH) and dynamic recovery (DRV), DRX, DP, and solute diffusion, are integrated and interconnected by their mutual effects. To facilitate the CA modeling, a novel local pinning model is proposed to reflect the uneven retardation of a precipitate to grain boundary migration and the virtual intersections of precipitates and grain boundaries based on their distribution, and its rationality is verified by simulations for grain coarsening. Considering the substantial difference in grain size and precipitate size, a multilevel cellular space is constructed, with a coarse parent cellular space for DRX and a sub-cellular space discretized from parent cells for DP, to balance computational efficiency and accuracy. The simulation successfully captures the microstructure evolution with multiscale characteristics, specifically the refinement of grains from hundreds of micros to a few micros through DRX, aided by dynamically precipitated second-phase particles in the submicron (hundreds of nanometers) range. The high degree of agreement between simulated and experimental results in terms of kinetics for microstructure evolution and microstructure after deformation at various temperatures and strain rates attests to the sound rationality and strong predictive capability of the established multilevel CA model. A comparison between the simulated results of the traditional CA model exclusively for DRX and those obtained from the multilevel CA model that incorporates both DRX and DP highlights the necessity of considering the interaction between these two phenomena for accurate grain size prediction.
What problem does this paper attempt to address?