Exploring plastic deformation behavior and underlying mechanisms of magnesium alloy: Progress in single length-scale approaches and frontiers in cross-scale linkages

The plastic forming of magnesium (Mg) alloys, characterized by their hexagonal close-packed (HCP) crystal structure, involves complex deformation mechanisms spanning from atomic to macroscopic scales. High-accuracy simulation technologies are pivotal for understanding these mechanisms and optimizing forming processes. However, the inherent multiscale nature of deformation poses significant challenges in seamlessly integrating models across different length scales. This review proposes a “multiscale modeling pyramid” framework, which systematically links atomic-scale methods (density functional theory, molecular dynamics), mesoscale approaches (phase-field modeling, crystal plasticity finite element method), and macroscopic simulations (finite element method, constitutive models). The capabilities and limitations of each scale-specific method are critically evaluated, with a particular focus on their roles within the pyramid: providing physical parameters for upper-scale models while being constrained by lower-scale mechanisms. Furthermore, recent advances in multiscale coupling strategies are highlighted, encompassing both mechanism-driven direct coupling and data-driven machine learning (ML) approaches. ML techniques, such as deep learning potentials and neural networks, are revolutionizing this field by enabling accurate and efficient cross-scale predictions. Finally, potential research directions are outlined, aimed at enhancing computational efficiency, predictive accuracy, and the integration of multiscale simulations for the rational design of Mg alloy forming processes.