About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
Chemistry and Physics of Interfaces
|
| Presentation Title |
Tritium Transport across Solid-Liquid Interfaces in Nuclear Fusion Breeder Blankets using Machine-Learning-Based Atomistic Simulations |
| Author(s) |
Prashanth Srinivasan, Krishna Chaitanya Pitike, Wahyu Setyawan, Matthew R Ryder, Mark R Gilbert, Duc Nguyen-Manh |
| On-Site Speaker (Planned) |
Prashanth Srinivasan |
| Abstract Scope |
A commercially viable nuclear fusion reactor requires an efficient tritium breeder blanket, which contains different interfaces including the one between the breeder and the surrounding structural material. Modelling necessitates physical understanding and accurate measurement of tritium atoms in the breeder and across the interface. Here, we develop machine-learning interatomic potentials(MLIPs) to gain an atomistic-level understanding. Specifically, we train an atomic cluster expansion(ACE) and a neuroevolution potential(NEP), for analyzing tritium in a liquid lithium breeder–irradiated solid vanadium structural material interface. Our results indicate large negative segregation energies for tritium at the interface, a significant barrier to travel into the structural material and decrease in the free energy with increasing vacancies. Large-scale molecular dynamics are performed to investigate Gibbs free energies, obtain tritium diffusivity as a function of temperature along and across the interface, and calculate steady-state concentrations. The methodology provides insights into tritium transport modelling and extendable to other interfaces. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Modeling and Simulation, Nuclear Materials |