Title:AI-driven Uncertainty Quantification & Multi-Physics Approach to Evaluate Cladding Materials in a Microreactor

Abstract:The pursuit of enhanced nuclear safety has spurred the development of accident-tolerant cladding (ATC) materials for light water reactors (LWRs). This study investigates the potential of repurposing these ATCs in advanced reactor designs, aiming to expedite material development and reduce costs. The research employs a multi-physics approach, encompassing neutronics, heat transfer, thermodynamics, and structural mechanics, to evaluate four candidate materials (Haynes 230, Zircaloy-4, FeCrAl, and SiC-SiC) within the context of a high-temperature, sodium-cooled microreactor, exemplified by the Kilopower design. While neutronic simulations revealed negligible power profile variations among the materials, finite element analyses highlighted the superior thermal stability of SiC-SiC and the favorable stress resistance of Haynes 230. The high-temperature environment significantly impacted material performance, particularly for Zircaloy-4 and FeCrAl, while SiC-SiC's inherent properties limited its ability to withstand stress loads. Additionally, AI-driven uncertainty quantification and sensitivity analysis were conducted to assess the influence of material property variations on maximum hoop stress. The findings underscore the need for further research into high-temperature material properties to facilitate broader applicability of existing materials to advanced reactors. Haynes 230 is identified as the most promising candidate based on the evaluated criteria.