Numerical simulations for determination of minimum representative bundle size in wire wrapped tube bundles

When constructing an experiment or simulation for flow through a wire-wrapped fuel rod bundle, scientists may utilize a smaller bundle than design in order to minimize material or computational costs. Small bundles may not capture the required flow physics experienced by larger bundles. This paper compares the flow fields through wire-wrapped rod bundles of 19, 37, 61, and 91 pins and investigates the ability of each to capture the relevant physics of a larger bundle. For model verification, the SST k-ω and elliptic blending k-∊ RANS models were compared against an LES simulation of the 19 pin domain, finding both RANS turbulent models capable for the given geometry. The central subchannel transverse velocity and inter-channel mass exchange for each bundle was compared, revealing a strong dependence of inter-channel mixing on the bundle size. Furthermore, analysis of the inter-channel mass exchange for subchannels at a varying distance from the surrounding shroud revealed than the mass exchange as a function of height is strongly tied to distance from the wall, with a slight adjustment in magnitude for the bundle size. The asymmetrical aspect of the wall effect was observed, revealing that the outside 2 or 3 rings of subchannels experience a large deviation from the characteristic subchannel behavior, but the central subchannels of the 37, 61, and 91 pin bundles are largely isolated from the asymmetric wall effect. Based on these findings the authors recommend a 37 pin bundle as the minimum surrogate for a larger bundle, and 61 pins as the preferred bundle size.