About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
Computational Thermodynamics and Kinetics
|
| Presentation Title |
Evaluating the Thermodynamic Behavior of CsF Through Machine Learning-Accelerated Nanoscale Simulations |
| Author(s) |
Pandu Wisesa, Juliano Schorne-Pinto, Theodore M. Besmann |
| On-Site Speaker (Planned) |
Pandu Wisesa |
| Abstract Scope |
As cesium is both a major fission product in molten salt reactors, present as CsF, and its volatility and radiotoxicity issue, representing its properties and thermodynamic behavior is important. The current understanding of this system suggests significant anharmonic behavior in the solid at roughly 400 K, an unexpectedly low temperature. This anomalous behavior thus warranted work to re-evaluate the results. To address the issue a combination of using density functional theory calculations and machine learning interatomic potential were utilized to perform simulations at nanoscale levels to capture the thermodynamic properties. Such system size and simulation time allow for a sufficient representation of the system in the solid and liquid state as reflected in the minimal deviations of the properties. The results indicated that while CsF exhibits anharmonicity, it is to a lesser degree and at higher temperatures than previously reported. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, Nuclear Materials, Machine Learning |