[11p-N107-1~16] 23.1 Joint Session N "Informatics"

[11p-N107-10] Optimization of flow distribution in SiC solution growth using topology and machine learning

〇(M2)Masaru Isono¹, Tomoo Yokoyama², Kentaro Kutsukake^3,4, Shunta Harada^1,4, Miho Tagawa^1,4, Toru Ujihara^1,4 (1.Grad. School of Eng. Nagoya Univ., 2.Dep. of Eng. Gifu Univ., 3.AIP RIKEN, 4.IMaSS Nagoya Univ.)

Keywords:optimization, topology, SiC solution growth

In the SiC solution growth method, it is important to control the flow distribution of the solution in the crucible, which affects the crystal growth rate and morphology. However, it is difficult to reflect the distribution such as flow distribution in the objective function. In this study, we focused on the stagnation point where the flow velocity becomes zero in the flow distribution and searched for the process parameters that can reproduce the flow distribution where the high-quality 3-inch crystal was obtained in 6-inch crystal growth.