[SY-F5] Symposium F-5

[SY-F5] Solidification analysis by non-equilibrium phase field model using thermodynamics data estimated by machine learning

^○Sukeharu Nomoto¹, Hiroshi Wakameda¹, Masahito Segawa¹, Toshiyuki Koyama², Akinori Yamanaka³ (1.ITOCHU techno-Solutions Coorporation, Japan, 2.Materials Design Innovation Engineering, Nagoya Univ., Japan, 3.Department of Mechnical Systems Engneering, Tokyo University of Agriculture and Technology, Japan)

A multi-phase field (MPF) model coupled with finite interface dissipation proposed by Steinbach et al. is applied to simulate the dendritic solidification in Fe-Cr-Ni-Mo-C steel. Thermodynamic calculation using CALPHAD database is replaced by machine learning prediction procedure in this MPF program. Solidification calculations are performed in a case of quinary system: Fe-16Cr-2Mo-10Ni-0.08C. We confirm that the microstructure evaluated by using machine learning parameter is good agreement with one directly coupled with CALPHAD database. Furthermore, this calculation is also confirmed to be approximately five times faster than the direct CALPHAD calculation method.