文献类型：期刊文献

英文题名：Artificial intelligence for advanced alloy: multiscale simulation, intelligent manufacturing, high-throughput characterization, and service performance intelligent assessment

作者：Wu, Wenwang[1];Xu, Kang[1,2];Yang, Xianli[1];Zhao, Qiutong[3];Mao, Shilin[4,5];Sun, Dongke[4,5];Wei, Tianyu[6];Zhang, Xu[6];Hou, Xiang[7];Zhang, Liang[7];Deng, Dayang[1,8];Yang, Ping[9];Zhao, Pengyang[9];Wen, Cheng[10];Wu, Liying[1];Liu, Ruxue[1];Zhang, Rui[1];Hu, Zhiqiang[1];Sun, Zhengming[2]

机构：[1]Suzhou Lab, 388,Ruoshui St,Suzhou Ind Pk, Suzhou 215123, Peoples R China;[2]Southeast Univ, Sch Mat Sci & Engn, Nanjing 211189, Peoples R China;[3]Dalian Univ Technol, Sch Phys, Key Lab Mat Modificat Laser Ion & Electron Beams, Minist Educ, Dalian 116024, Peoples R China;[4]Southeast Univ, Sch Mech Engn, Nanjing 211189, Peoples R China;[5]Key Lab Struct & Thermal Protect High Speed Aircra, Nanjing 211189, Peoples R China;[6]Southwest Jiaotong Univ, Sch Mech & Aerosp Engn, Appl Mech & Struct Safety Key Lab Sichuan Prov, Chengdu 610031, Peoples R China;[7]Chongqing Univ, Coll Mat Sci & Engn, Int Joint Lab Light Alloys MOE, Chongqing 400044, Peoples R China;[8]Harbin Inst Technol, Ctr Composite Mat & Struct, Natl Key Lab Sci & Technol Adv Composites Special, Harbin 150080, Peoples R China;[9]Shanghai Jiao Tong Univ, Sch Naval Architecture Ocean & Civil Engn, Dept Engn Mech, 800 Dongchuan Rd, Shanghai 200240, Peoples R China;[10]Guangdong Ocean Univ, 1 Haida Rd Mazhang Dist, Zhanjiang City 524088, Guangdong Provi, Peoples R China

年份：2027

卷号：165

外文期刊名：PROGRESS IN MATERIALS SCIENCE

收录：SCI-EXPANDED(收录号：WOS:001890320200001)、、WOS

基金：This work has been financed by the Advanced Materials-National Science and Technology Major Project (No. 2025ZD0619200) and the National Natural Science Foundation of China (No. 12472347) . This work was supported by the National Natural Science Foun-dation of China (52130107) .

语种：英文

外文关键词：Artificial intelligence; Alloy development; Multi-scale simulation; Intelligent manufacturing; High-throughput characterization; Service-life prediction

外文摘要：Artificial intelligence is reshaping alloy research and development by extending its role from conventional data-driven property prediction to a broader set of functions embedded in physical models, manufacturing workflows, characterization pipelines, and service assessment frameworks. This review summarizes recent progress in AI-enabled alloy development across four interconnected stages: multiscale computational simulation, intelligent manufacturing, highthroughput characterization, and service-life prediction. In computational simulation, machine learning interatomic potentials, CALPHAD-assisted models, discrete dislocation dynamics, phasefield modeling, crystal plasticity finite element methods, and ICME are discussed with emphasis on simulation acceleration, closure-relation learning, parameter inversion, and multi-scale coupling. In intelligent manufacturing, applications of AI in casting, forging, additive manufacturing, rolling, heat treatment, and metallurgical refining are reviewed, highlighting multi-source sensing, physical-field reconstruction, defect detection, process optimization, and digital-twin-based control. High-throughput characterization based on SEM, EBSD, TEM/STEM, and automated microstructure quantification is further considered as a critical data-feedback layer that links processing history, microstructural descriptors, and model calibration. For service assessment, AI-assisted prediction of corrosion, fatigue, creep, and coupled damage is summa-rized to clarify how composition, microstructure, defects, residual stress, environment, and loading history can be integrated for lifetime prediction and reliability evaluation. Finally, key challenges are discussed, including data scarcity and heterogeneity, insufficient physical consis-tency, limited extrapolation reliability, weak interoperability among models and scales, and the lack of uncertainty-aware decision-making. Future perspectives are proposed toward physics-informed learning, standardized databases, autonomous laboratories, digital twins, and closed-loop alloy development systems.