Sodium fast reactor designs often implement a hexagonal array of fuel rods with wire-wrappers to encourage the exchange of coolant between subchannels. The ability to accurately predict inter-subchannel mixing can be used as a metric for turbulence model performance in capturing wire-wrapped fuel rod bundle flow behavior. In this study inter-subchannel mixing predictions by Large Eddy Simulation (LES) and Reynolds Averaged Navier Stokes (RANS) models are compared. The results indicate that the lower order RANS approach is capable of predicting intersubchannel mixing inside a 19 rod bundle with acceptable accuracy. The RANS model was extended to 37, 61, and 91 rod bundles to observe the effects of bundle size on inter-subchannel exchange for the center-most subchannels. Transverse velocity magnitude and mass exchange were observed to increase with larger bundle sizes. Inter-subchannel mixing is observed to be a strong function of bundle size for bundles up to 91 rods. The results indicate that the inner subchannels of larger bundles may converge upon a characteristic flow pattern. The 91 rod bundle is not large enough to isolate the inner subchannels from shroud effect, and larger bundles will need to be investigated.