Summary on the results of two computational fluid dynamic benchmarks of tube and different channel geometries

Abstract

Two computational fluid dynamic (CFD) benchmarks have been performed to assess the prediction accuracy and sensitivity of CFD codes for heat transfer in different geometries. The first benchmark focused on heat transfer to water in a tube (first benchmark), while the second benchmark covered heat transfer to water in two different channel geometries (second benchmark) at supercritical pressures. In the first round with the experimental data unknown to the participants (i.e., blind calculations), CFD calculations were conducted with initial boundary conditions and simpler CFD models. After assessment against measurements, the calculations were repeated with the refined boundary conditions and material properties in the follow-up calculation phase. Overall, the CFD codes seem to be able to capture the general trend of heat transfer in the tube and the annular channel but further improvements are required in order to enhance the prediction accuracy. Finally, sensitivity analyses on the numerical mesh and the boundary conditions were performed. It was found that the prediction accuracy has not been improved with the introduction of finer meshes and the effect of mass flux on the result is the strongest among various investigated boundary conditions.