The Research Group
- ARPA-E funded project
- PNNL, U. of North Texas, Citrine Informatics
Award# DE-AR0001613
ONWARDS, PM Bob Ledoux
The Problem
- Current nuclear waste storage is too expensive
- Waste forms are space-inefficient and prone to hazardous leaching, mechanical failure, and thermal degradation
- Design novel waste form with improvements in five performance metrics
The Process
- Built database of nuclear waste form materials properties for model waste stream
- Developed ML models to predict strength, chemical durability, phase stability, vitrification
- Optimized nuclear waste glass for waste loading, strength, chemical durability, thermal stability
The Outcome
- >2x waste loading
- 20x reduction in waste volume
- >2x mechanical strength
- Multi-objective design workflow
Hegde et al. Digit. Discov. (2024); Allec et al. J. Am. Ceram. Soc. (2024); Du et al. Trans. Am. Nucl. Soc. (2024); Kalahe et al. J. Nucl. Mat. (2024)
