| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Qualification of Nuclear Materials Integrating Experiments, Modeling, and Theories
|
| Presentation Title |
Uncertainty Quantification and Bayesian Calibration Applied to Mechanistic Models of Nuclear Fuel Performance |
| Author(s) |
David Andersson |
| On-Site Speaker (Planned) |
David Andersson |
| Abstract Scope |
Fission gas evolution, swelling and creep are important for the performance of nuclear fuels. They are controlled by atomic scale diffusion processes that couple to the chemistry of the fuel, temperature, and irradiation conditions. Mechanistic models have been developed for fission gas evolution, swelling and creep in UO2 and UN by investigation of the underlying atomic scale defects and diffusion processes using atomic-scale calculations, followed by cluster dynamics simulations and, finally, gas evolution, swelling and creep rates are captured by use of the controlling rates predicted by cluster dynamics simulations in model equations governing the macro-scale response of the fuel pellet. To reduce uncertainties of the mechanistic models, a Bayesian framework was developed to calibrate the model to experimental data for UO2. The results indicate a good agreement between the calibrated diffusivity, non-stoichiometry and creep predictions and the given data. Similar techniques were applied to swelling in UN. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Modeling and Simulation, Nuclear Materials, |