Evaluating the Low-Reynolds-Number Models for Predicting Axial Flow in a Tightly-Spaced Square-Pitch Rod Bundle

Single-phase turbulent flow through a subchannel of a tightly spaced rod bundle with a pitch-to-diameter ratio p/d = 1.107 is investigated numerically at the Reynolds number Re = 48 400. Two low-Reynolds-number (LRN) Reynolds-averaged Navier–Stokes (RANS) models were used, namely, the $$k{-} \omega ~{\text{SST}}$$ model and the $${\text{BL}}{-} \overline {{{\upsilon }^{2}}} {\text{/}}k$$ model, which is an elliptic-blending-based $$\overline {{{\upsilon }^{2}}} {-} f$$ model implemented in the in-house EDF code−saturne®. The numerical predictions are presented in terms of the mean velocity distribution, the turbulent kinetic energy, and variation in the wall shear stress. These predictions are then compared with experimental results given by Hooper in 1980. The results indicate that both LRN models are capable of predicting specific flow features. Furthermore, the low-Re near-wall treatment successfully reproduces the correct qualitative behavior of the wall shear stress along the rod surface, in agreement with experimental results. However, the $$k{-} \omega ~{\text{SST}}$$ model demonstrated superior overall performance as compared to the $${\text{BL}}{-} \overline {{{\upsilon }^{2}}} {\text{/}}k$$ model, showing better accuracy in predicting the flow characteristics in this complex configuration.