Solid-state sintering is a crucial thermal post-processing step in metal extrusion additive manufacturing(MExAM),influencing the microstructural evolution,densification,and final properties of fabricated components.Ho...Solid-state sintering is a crucial thermal post-processing step in metal extrusion additive manufacturing(MExAM),influencing the microstructural evolution,densification,and final properties of fabricated components.However,accurately simulating this process remains challenging due to its inherently multiscale and multiphysics nature.This comprehensive and critical review examines the main computational approaches developed to model solidstate sintering in the MExAM context,ranging from nano-to macrostructure scales,including molecular dynamics,kinetic Monte Carlo,discrete element methods,phase-field models,and continuum-based methods.For each,we detail the underlying mathematical formulations,numerical strategies,and implementation environments.Their capabilities and limitations are evaluated in terms of scale resolution,physical accuracy,and computational demands.Particular attention is given to challenges such as the coupling of thermal,mechanical,and diffusive phenomena,as well as the difficulty of bridging disparate spatial and temporal scales.In response to these limitations,emerging trends such as Physics-informed machine learning(PIML)offer promising avenues to enhance predictive accuracy,improve computational efficiency,and streamline simulation workflows.By integrating conventional modeling techniques with data-driven approaches,the field is moving toward faster and more reliable predictions of sintering behavior,advancing the goals of the MExAM initiative.The insights presented aim to guide future research focused on optimizing sintering processes for broader industrial applications and improved material performance.展开更多
基金the financial support provided by Total Energies S.E.,contract No.FR00055666 and the French Em-bassy in Angola.
文摘Solid-state sintering is a crucial thermal post-processing step in metal extrusion additive manufacturing(MExAM),influencing the microstructural evolution,densification,and final properties of fabricated components.However,accurately simulating this process remains challenging due to its inherently multiscale and multiphysics nature.This comprehensive and critical review examines the main computational approaches developed to model solidstate sintering in the MExAM context,ranging from nano-to macrostructure scales,including molecular dynamics,kinetic Monte Carlo,discrete element methods,phase-field models,and continuum-based methods.For each,we detail the underlying mathematical formulations,numerical strategies,and implementation environments.Their capabilities and limitations are evaluated in terms of scale resolution,physical accuracy,and computational demands.Particular attention is given to challenges such as the coupling of thermal,mechanical,and diffusive phenomena,as well as the difficulty of bridging disparate spatial and temporal scales.In response to these limitations,emerging trends such as Physics-informed machine learning(PIML)offer promising avenues to enhance predictive accuracy,improve computational efficiency,and streamline simulation workflows.By integrating conventional modeling techniques with data-driven approaches,the field is moving toward faster and more reliable predictions of sintering behavior,advancing the goals of the MExAM initiative.The insights presented aim to guide future research focused on optimizing sintering processes for broader industrial applications and improved material performance.
