An OpenFOAM-based evaluation of various wall treatment strategies in simulating separated flows past bluff bodies with heat transfer

The effect of wall treatment on the performance of k-ε model in incompressible, turbulent, separated flows with and without heat transfer has been evaluated in this study. We have simulated two benchmark cases: (i) flow past a circular cylinder at Re = 3900, and (ii) flow past a heated square cylinder at Re = 21 400 using the open source CFD package: OpenFOAM. We have compared three variants of the k-ε model, namely, Launder-Sharma k-ε model (Yap corrected) (LSKEY), Lam–Bremhorst k-ε model (Yap corrected) (LBKEY) and two-layer k-ε model (TLKE) along with the available experimental and direct numerical simulation (DNS) data. Comparisons are made in terms of the models’ capability to predict the mean flow variables, surface integral quantities, and heat transfer characteristics at different wake locations. On the basis of the presented study, we conclude that LSKEY performs better than the other models in predicting the wake and surface flow and heat transfer parameters. Further our comparisons show that, while LSKEY and LBKEY require comparable clock time per flow-through cycle, the computational time needed by TLKE is almost twice as compared to LBKEY or LSKEY. These results call for more attention from the CFD community onto the LSKEY model, in particular, so that it can be incorporated in various other flow fields, especially the scale resolving methodologies like the partially-averaged Navier-Stokes (PANS) equations, wherein a superior wall treatment along with a shorter computational time could be of immense advantage. In authors’ opinion, these benefits of the LSKEY model have largely been overlooked, perhaps because of a biased preference to the TLKE model, which enjoys the default presence in popular commercial CFD packages.