About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Bridging Scales: Deformation and Damage Mechanisms in Microstructurally and Compositionally Complex Metallic Alloys
|
| Presentation Title |
Bridging Atomistic and Mesoscale Models for Predicting Materials Response Under Radiation |
| Author(s) |
Khanh Dang, Darshan Bamney, Minh Nhat Vu |
| On-Site Speaker (Planned) |
Khanh Dang |
| Abstract Scope |
The interactions between irradiation-induced defects and intrinsic microstructural features (dislocations, twins, secondary phases, grain boundaries) across various length and time scales dictate the thermomechanical performance of materials in irradiation environments. However, the current multi-scale modeling approaches for radiation damage typically neglect cascade-microstructure interactions, limiting our ability to predict radiation-induced destabilization of sinks and forecast materials’ failure. We propose a new modeling framework, the Realistic Atomistic-to-Mesoscale Microstructure Mapping Package (RAMMMP). RAMMMP integrates realistic dislocation networks, generated by spatially resolved cluster dynamics-discrete dislocation dynamics (CD4) simulations, and aggregates them using advanced machine learning (ML) techniques. This ML module learns defect signatures (e.g., Nye tensor, plastic distortion, stress field) to map continuum-to-atomistic structures. Cascade simulations are then performed on atomized networks via molecular dynamics, and the resulting atomistic microstructures are upscaled to CD4. Long-term microstructure evolution is simulated via CD4, and the process is repeated, enabling accurate modeling of radiation-induced damage and failure. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Computational Materials Science & Engineering, Nuclear Materials |