| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Ceramics and Ceramic-based Composites for Nuclear Fission Applications
|
| Presentation Title |
Correlating Atomic Structure with Elastic Properties in Non-textured Pyrocarbon |
| Author(s) |
Raphaelle David, Yongfeng Zhang |
| On-Site Speaker (Planned) |
Raphaelle David |
| Abstract Scope |
Pyrocarbon has attracted intensive research efforts because of its wide applications from the medical field to the nuclear industry. Among those, the TRISO particle fuel, which is considered as a leading fuel candidate for various reactors, contains two high-density, textured pyrocarbon layers, IPyC and OPyC, and a low-density, non-textured pyrocarbon buffer layer. While previous efforts have mostly focused on textured pyrocarbon, revealing the atomic structure of non-textured pyrocarbon and correlating it with material properties has remained a challenge. Aided by molecular dynamics simulations, we identified a list of microstructural descriptors, including a newly proposed atomic orientation parameter for extracting the crystallite size for non-textured pyrocarbon. A machine learning model was trained to predict the elastic moduli based on the list of descriptors with excellent accuracy. Furthermore, these descriptors were used to characterize the pyrocarbon deformation mechanisms as well as its behavior under irradiation. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Nuclear Materials, Mechanical Properties, Machine Learning |