About this Abstract |
Meeting |
7th World Congress on Integrated Computational Materials Engineering (ICME 2023)
|
Symposium
|
ICME 2023
|
Presentation Title |
Multi-phase-field simulation of rapid solidification in SUS316L stainless steel using aritificial neural network-based thermodynamic calculation |
Author(s) |
Akinori Yamanaka, Masahito Segawa, Shoichiro Nakamura |
On-Site Speaker (Planned) |
Akinori Yamanaka |
Abstract Scope |
Additive manufacturing is a powerful method to produce an industrial part with flexible shape and superior mechanical properties. In the additive manufacturing process, a laser heating causes a melting of material, followed by a rapid cooling that results in formation of unique solidification microstructures. The multi-phase-field method is a powerful numerical simulation method to quantitatively predict the microstructure evolution during the rapid solidification in the additive manufacturing process. In this study, the rapid solidification in SUS316L stainless steel during the additive manufacturing is simulated using the non-equilibrium multi-phase-field model that is able to simulate microstructure evolutions under a strong non-equilibrium condition. The thermodynamic calculations for the multi-component SUS316L stainless steel in the multi-phase-field simulation were accelerated using aritificial neural networks. In this presentation, we present a computational framework to train the aritificial neural networks with the thermodynamic database and to implement the trained neural network into the multi-phase-field simulation. |
Proceedings Inclusion? |
Planned: Other (describe below) |