机器学习赋能增材制造难熔高熵合金设计OA

Refractory high-entropy alloys(RHEAs),owing to their exceptional high-temperature properties,show great potential for use in extreme environments such as aerospace systems.However,the inherent complexity of their multicomponent systems poses significant challenges to the efficiency of conventional alloy design methods.Although additive manufacturing(AM)overcomes some limitations of traditional fabrication routes,its intrinsic non-equilibrium solidification behavior and complex process-parameter space introduce new challenges in microstructural control and property optimization.In this context,machine learning(ML)offers a data-driven paradigm for uncovering the intricate composition-structure-property relationships in RHEAs and provides a powerful tool for intelligent materials design.This review systematically summarizes the latest advances in the application of ML to RHEAs,with particular focus on key issues in AM scenarios.On the one hand,the roles of ML in phase prediction,strength and ductility optimization,and hardness design are discussed.On the other hand,ML-enabled strategies for process-parameter optimization,defect mitigation,and microstructural tailoring in AM are thoroughly analyzed.Despite notable progress,several challenges remain,including the scarcity of AM-specific datasets,the limited generalization capability of process-property models,and the insufficient integration of physical mechanisms.Therefore,establishing standardized databases oriented toward additive manufacturing,developing machine learning models integrated with physical constraints,and constructing an integrated process-microstructure-property optimization framework are key directions for promoting the transition of refractory high-entropy alloys in the field of additive manufacturing from"printable"to"designable".