About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
Accelerated Qualification of Nuclear Materials Integrating Experiments, Modeling, and Theories
|
Presentation Title |
Physics-informed Smart Scaling for Accelerated Fuel Testing |
Author(s) |
Anant Raj, Chalie Owen, Hany Abdel-Khalik, Khafizov Marat, Colby Jensen, Aysenur Toptan, Jason Hales |
On-Site Speaker (Planned) |
Chalie Owen |
Abstract Scope |
Scaled experiments like MiniFuel and fission accelerated steady-state tests (FAST) will play a key role in accelerating the fuel development cycle. However, scaling of the test parameters, like the higher fission rate in FAST experiments, constrains the regulators to adopt conservatively bound scaling distortion uncertainties. The present work alleviates this issue through a concerted use of high-fidelity simulations and machine learning (ML). Using BISON for simulating the fuel response, discriminative ML is employed to extract the most informative pseudo response between the scaled (FAST) state and the application state. The approach is employed for FAST test capsules as a test case for computing the scaling of biases in the peak centerline temperature and fission gas release between the scaled (FAST) state and the application state. |
Proceedings Inclusion? |
Planned: |
Keywords |
Modeling and Simulation, Machine Learning, Nuclear Materials |