Scale-resolving simulations of the flow in a nuclear fuel bundle with a channel spacer grid using partially averaged Navier–Stokes and large-eddy simulation

Abstract

The objective of this study is to evaluate the capabilities of the Partially Averaged Navier–Stokes (PANS) method in simulations of the flow through a rod bundle representative of a nuclear fuel assembly with a channel-type spacer grid. In the PANS bridging turbulence model, filters can be applied to obtain any level of resolution from Reynolds Averaged Navier–Stokes (RANS) to Direct Numerical Simulation (DNS). The closure model is sensitive to the length-scale cutoff by means of unresolved to total kinetic energy ratio (fk) and unresolved to total dissipation ratio (fε). Simulations are conducted to study the effect of the cutoff of length scale on the results obtained for flow statistics at selected locations in the domain. The results obtained with different PANS filters are compared with Unsteady RANS (URANS), Large Eddy Simulation (LES), and experiments. The mean and fluctuating flow components are computed at a representative plane located at the inter-channel between rods. Other flow quantities analyzed include the pressure drop upstream and downstream of the spacer grid, the turbulent kinetic energy, and the unresolved eddy viscosity. Additionally, we use the Lumley triangle to study turbulence anisotropy and to compare the nature of the energy content captured with LES and one of the PANS models. It is shown that the PANS model with fk=0.4, which resolves 60% of the turbulent kinetic energy content, captures the most relevant flow physics and is a suitable modeling approach for this application.