About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Ceramics and Ceramic-Based Composites for Nuclear Applications III
|
| Presentation Title |
AI/ML-assisted Design of Phosphate Nuclear Waste Forms |
| Author(s) |
James Edward Saal, Kyle Miller, Vinay Hegde, Sarah Allec, Miroslava Peterson, Thiruvillamalai Mahadevan, Thanh Nguyen, Jayani Kalahe, Jared Oshiro, Robert Seffens, Ethan Nickerson, Xiaonan Lu, Jincheng Du, Brian Riley, John Vienna |
| On-Site Speaker (Planned) |
James Edward Saal |
| Abstract Scope |
Current disposition pathways for salt wastes from molten salt reactors or used nuclear fuel reprocessing produce waste forms with relatively low halide loading, large volumes, poor stability, and poor durability. Recent work has shown the potential for dehalogenation of the waste salt (recycling half the waste mass) and immobilization in iron phosphate glasses to increase waste loading and decrease volume. While these glass waste forms have promise, alternative fully-optimized ceramic or glass-ceramic waste forms may be adopted, with improved cation loading, thermal stability, and mechanical durability by designing waste forms with dehalogenation in mind. Using machine learning, artificial intelligence, and physics-based simulation methods, we have developed novel ceramic phosphate waste forms for dehalogenation and more secure immobilization of salt waste. We will summarize progress in building waste-related materials property databases, machine learning property models for glass formability and durability, and using AI to identify promising phosphate waste forms. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, ICME, Nuclear Materials |