[S3] S3 Materials DX for the research and development of structural materials for extreme environment(IV)

2:25 PM - 2:40 PM JST(5:25 AM - 5:40 AM UTC)

[S3.12] Wide-range Microstructure Generation of MoSiBTiC Alloys Using Deep Learning Models

○Shunto Obana¹, Chihana Kudo¹, Shugo Tomioka¹, Takahiro Kaneko², Kyosuke Yoshimi² (1. Tohoku Univ. Eng. (Graduate student), 2. Tohoku Univ. Eng.)

Keywords:

MoSiBTiC Alloy,Deep Learning,Microstructure Generation,Materials Informatics,Transformer

This study investigates a method for generating wide-area, high-resolution microstructures of multiphase MoSiBTiC alloys using VQGAN and Transformer. The generated microstructures were quantitatively compared with real microstructures to evaluate their fidelity.