A Cross-code Verification Study for Conjugate Heat Transfer in Gas-cooled Pebble Bed Reactors

Test and small-scale commercial pebble bed reactors have been established in the US and various other regions globally. Traditionally, heat transfer studies have primarily focused on system-level and porous media methodologies. However, adopting a Computational Fluid Dynamics (CFD) approach for Pebble Bed Reactors (PBRs) has posed challenges, primarily due to computational constraints. Nonetheless, in recent decades, significant strides have been made owing to advancements in algorithms and computational power. A plethora of CFD codes populate the market. This study undertakes a cross-code verification analysis centered on a randomly packed bed configuration within a cylindrical vessel. All simulations are conducted under Conjugate Heat Transfer (CHT) conditions, entailing the meshing and coupling of solid and fluid domains. Specifically, four codes are employed: OpenFOAM, STAR-CCM+, Cardinal, and NekRS. Notably, simulations employing Cardinal continue to leverage NekRS for fluid domain calculations while using MOOSE for solid domain computations, with information only transferred at the interface. For NekRS, we adopted the over-set mesh approach to couple solid and fluid meshes. We also performed simulations to compare with analytical solutions to justify this method. Two Prandtl numbers are simulated: one represents gas flow and the other represents salt flow, as gas and salt are commonly used coolants for pebble bed cores. Furthermore, the Reynolds number, based on inlet parameters, spans a range from 20 to 10,000, effectively encompassing a broad spectrum of operating conditions. Based on our observations, using the over-set mesh approach in NekRS, we can correctly calculate the temperature field while maintaining computational speed and high scalability, which are compromised through interface mapping when using Cardinal.