Predictions in CFD simulations of wire-wrapped SFR fuel assemblies

In response to the goals outlined by the U.S. Department of Energy's Advanced Fuel Cycle Initiative, an effort has been initiated to create an integrated multi-physics, multi-resolution thermal-hydraulic simulation tool package for the evaluation of nuclear power plant design and safety. As part of this effort, a variety of thermal-hydraulic analysis methods are being evaluated for application to the prediction of heat transfer and fluid dynamics in the wire-wrapped fuel-rod assemblies found in a sodium-cooled fast reactor core. The work described herein is an initial assessment of the capabilities of the general-purpose commercial RANS-based computational fluid dynamics (CFD) code Star-CD for prediction of fluid dynamic characteristics in a wire-wrapped fast reactor fuel assembly. Wire-wrapped fuel rod assemblies containing 7, 19 or 37 pins based on the dimensions of fuel elements in the conceptual design of the Advanced Burner Test Reactor (ABTR) [1] were simulated for different mesh densities and distortions, and domain configurations. Expanding on previously completed benchmarking comparisons with high-order numerical methods, specifically large eddy simulation-based NEK5000, an initial dynamic pressure loss validation effort has been completed using the pressure loss correlations developed by Cheng and Todreas [2]. This commonly used correlation was numerically developed and fitted to available experimental data. The agreement between the CFD predictions and the correlation improves as the pin number increases. The 7-pin bundle simulations predict dimensionless pressure loss coefficients that are 60% higher than the predictions of the correlation. The large deviation is not unexpected since the fundamental flow in this case is dominated by the swirl of the coolant in the edge channels, and this small assembly is outside the specified range of the correlation. The 19-and 37-pin simulations predict dimensionless pressure loss coefficients that are within the correlation's 14% error. The 37-pin simulation is within 9% of the correlation value.