About this Abstract |
Meeting |
2023 TMS Annual Meeting & Exhibition
|
Symposium
|
Transmutation Effects in Fusion Reactor Materials: Critical Challenges & Path Forward
|
Presentation Title |
Machine Learning Generation of Trajectories for Accurate Modeling Plasma Material Interactions |
Author(s) |
Osetsky Yury, German Samolyuk, Eva Zarkadoula, Markus Eisenbach, Cornwall Lau, Juergen Rapp |
On-Site Speaker (Planned) |
Osetsky Yury |
Abstract Scope |
The Materials Plasma Exposure eXperiment (MPEX) is a device for studying fusion reactor-relevant plasma-materials interactions (PMI). A corresponding MPEX digital twin aims to model the main processes of MPEX, including the PMI’s that ultimately define the degradation of the target material. Traditionally, PMI are considered within the binary collision approximation (BCA) implemented in various codes such as SRIM. These codes are not able to model the realistic trajectories of ions in materials depending on crystallography, temperature, and microstructure. To address this limitation, we developed a machine learning approach to model ion trajectories using classical and ab initio molecular dynamics (MD). First, we demonstrated that the full MD modeling results in different trajectories and penetration depths. Second, we have generated a database that can be used to reproduce PMI for MPEX-relevant ions and their energy and velocity spectra. The developed approach is applicable to different target and plasma compositions. |
Proceedings Inclusion? |
Planned: |
Keywords |
Modeling and Simulation, Nuclear Materials, |