High-Fidelity Multiphysics Modeling of the SNAP 10/A Space Reactor Using Cardinal

This work presents the development of a coupled neutronic-thermal hydraulics model using the high-fidelity multiphysics code Cardinal for the Space Nuclear Auxiliary Power (SNAP) 10/A space reactor. The first step was to perform a verification and validation process for the neutronic model developed in OpenMC. Results were compared against the Monte Carlo N-Particle (MCNP5) code and experimental measurements. Comparison with computational benchmarks showed an average absolute difference of 292 ± 71 pcm, or 345 ± 800 pcm when comparing with experimental results. The primary source of difference is the use of different cross-section libraries. MCNP5 simulations used a combination of the cross-section libraries ENDF/B-VI and ENDF/B-V for minor isotopes. At the same time, OpenMC simulations utilized ENDF/B-VIII.0. Differences with experimental results can be explained by the approximations made in computational models by reference data. Results showed good agreement, concluding that OpenMC is suitable for performing neutronic simulations for the SNAP 10/A space reactor. Then, a coupled model of the SNAP 10/A reactor was developed in Cardinal, solving neutron transport equations in OpenMC, thermal hydraulics in NekRS, and heat conduction in solid components in MOOSE (Multiphysics Object-Oriented Simulation Environment). Results for the heat source, temperatures, and velocities showed reasonable behavior. Temperature results showed good agreement with reference calculations, obtained with codes HEAP and TRANCORE. Differences can be explained considering that the top and bottom boundary conditions in the reference model were not adiabatic, while in the Cardinal model, these surfaces had adiabatic boundary conditions. Additionally, the reference model considered beryllium internal reflectors when solving the heat conduction equations. These reflectors also have nonadiabatic boundary conditions and therefore there was another heat loss at those surfaces as well. The Cardinal model did not consider the internal beryllium reflectors when solving heat conduction; these were only considered in the neutronic simulations. The average differences are +3.1 K for the fuel centerline temperature and +0.8 K for the coolant temperature.