Lijun Zhang
State Key Laboratory of Powder Metallurgy, Central South University, Changsha, 410083, China 

ABSTRACT: Efficient development of high-performance alloys is the key to driving industrial progress. However, with the traditional "trial-and-error" method that is quite time- and labor-consuming, it is difficult to realize the accurate design of alloy compositions and processes, and it is thus impossible to achieve efficient development of high-performance alloys. Thermodynamics and kinetics are the two theoretical pillars of materials, and an in-depth understanding of the thermodynamics and kinetics mechanisms of the microstructural evolution during alloy preparations and services is expected to realize the theoretical design of alloys. Based on the established alloy thermodynamic/kinetic databases and developed software, the quantitative relationship of the "composition-process-microstructure" in target alloy can be constructed by combining key experiments. A further integration of machine learning/active learning techniques and key experiments, the quantitative relationship of "microstructure-property" can be obtained. Furthermore, on the basis of the established quantitative relationship "composition-process-microstructure-property" together with the multi-criteria decision-making strategy, a high-throughput design platform for high-performance alloys can be developed. Finally, successful applications of the platform in the design of e.g., advanced Al alloys, Ni-based single-crystal superalloys and high-entropy superalloys are demonstrated.