Multiphysics continuous shape optimization of LWR nuclear reactor components under uncertainty

We present the capabilities of the continuous shape optimization module within the MOOSE Multiphysics simulation framework and showcase its relevance for nuclear engineering applications. We demonstrate the module's capabilities through an optimization study of annular fuel pin-cell shape, incorporating coupled neutronics and thermal-hydraulics simulations. The optimization process reveals significant improvements in both neutron multiplication and thermal performance. Our analysis explores the impact of various objective functions and physics feedback mechanisms, highlighting the importance of Multiphysics modeling in reactor design optimization. When including both spatial and thermal feedbacks in the pin-cell two-dimensional model, we achieve improvements of 1500 pcm in keff and around 50K reduction in maximum temperature. Furthermore, we introduce a novel DAKOTA-MOOSE interface for uncertainty quantification that provides a robust shape optimization capability under uncertainty. Both thermal conductivities and linear power uncertainties were propagated through the shape optimization process. This approach not only enhances design resilience but also quantifies the uncertainty in the expected economic benefits. By applying uncertainty quantification to the economic value derived from fuel area reduction, we can provide a more robust assessment of potential savings. Our optimized designs consistently achieve a 16-22% reduction in fuel area while maintaining or improving performance, with quantifiable confidence in the associated economic outcomes. The optimization module that we demonstrated in MOOSE, coupled with uncertainty quantification, provides a robust framework for exploring innovative fuel designs that facilitate the next generation of more efficient and safer nuclear reactors.