Development of an automated evaluation system for the process-structure-property relationship of structural metallic materials towards autonomous exploration

Automated autonomous experiment systems in material science are desired for exploring numerous conditions but barely any examples for solid bulk metals are reported. In this study, an automated experimental system for structural metallic materials is designed to evaluate the process-structure-property relationship of structural metallic materials. Fully automated heat treatments, X-ray diffraction (XRD) measurements and impact testing on bulk metallic samples with a size of 40 × 4 × 1.5 mm3 are realized by integrating a robot arm. The experimental process, excluding heat treatment, can be finished within 800 s for 1 sample. Automated high-throughput experiments on the one- and two-step tempering of a JIS SCM435H alloy are performed to demonstrate the capability and efficiency of the system, and the accuracy of the resulting property data is validated by manually measured hardness. It is also found that the diffraction peak width is positively correlated with strength, indicating the capability of the system to achieve structure–property relationships. An AI optimization tool built in the NIMS Orchestration System (NIMO) is implemented, thereby providing a basis for future closed-loop autonomous optimization.