The development of subchannel models for fuel assemblies and reactor cores requires accurate information on flow distribution, wall friction and loss coefficients in order to accurately predict the pressure, temperature and flow distribution on a subchannel level. This paper discusses the use of Computational Fluid Dynamics (CFD) simulations as a practical tool for characterising inlet velocity boundary conditions, an approximation of wall friction factor and spacer grid pressure loss coefficient, which are of fundamental importance to correctly generate a consistent subchannel model of a given assembly system. The geometry of the simplified PWR assembly presented here is based on the NUPEC PWR subchannel and bundle tests. Comparison of the derived friction factors and grid pressure loss coefficient with published and recommended values are reported. Discrepancies are also explained using additional calculations. A comparison of the overall system pressure drop, by comparing_numerical and analytical solutions, and the local axial pressure distribution at subchannel level are presented. To make a one-to-one comparison between CFD and subchannel solutions, volume-averaging is applied to the CFD results according to the chosen subchannel nodalization.
The obtained results show a perfect agreement between the two codes. This outcome reflects the correct approach employed to build two consistent numerical models by properly carrying important information from the high-resolution models (CFD) to the low-resolution models (subchannel code). Furthermore, it has been found that the large discrepancies recorded in the CFD prediction of the grid pressure loss coefficients suggested in the benchmark specifications are mainly because the suggested benchmark values do not take into account the presence of the bounding channel that was present in the experimental facility.
