| About this Abstract |
| Meeting |
2022 TMS Annual Meeting & Exhibition
|
| Symposium
|
Grain Boundaries and Interfaces: Metastability, Disorder, and Non-Equilibrium Behavior
|
| Presentation Title |
5D Grain Boundary Energy Landscapes, Paths and Correlations from Bayesian Inference |
| Author(s) |
Sterling G. Baird, Eric R Homer, David T Fullwood, Oliver Johnson |
| On-Site Speaker (Planned) |
Sterling G. Baird |
| Abstract Scope |
Leveraging our recently developed Voronoi fundamental zone (VFZ) framework, we use Bayesian inference strategies to infer 5D structure-property models for GB energy. We discuss how distributions of metastable GB states can be easily incorporated into this method both to inform GB structure-property model development, and to incorporate the effects of metastability into property predictions and subsequent mesoscale simulations. By analyzing the resulting models, we quantify the size and shape of FZs for cubic materials, and test and give context to commonly assumed correlation lengths for GBs. We demonstrate the way in which correlation length depends on crystallographic character and describe the implications for computational modeling. Finally, we identify important paths through the resulting GB energy landscapes that may influence microstructural evolution in ways that have not previously been investigated. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Other, Other, Machine Learning |