A CFD Evaluation of Multiple RANS Turbulence Models for Prediction of Boundary Layer Flows on a Turbine Vane

Abstract

There is a growing trend toward the use of conjugate CFD for use in prediction of turbine cooling performance. While many studies have evaluated the performance of RANS simulations relative to experimental measurements of the momentum boundary layer, no studies have evaluated their performance in prediction of the accompanying thermal boundary layer. This is largely due to the fact that, until recently, no appropriate experimental data existed to validate these models. This study compares several popular RANS models — including the realizable k-ε and k-ω SST models — with a four equation k-ω model (“Transition SST”) and experimental measurements at selected positions on the pressure and suction sides of a model C3X vane. Comparisons were made using mean velocity and temperature in the boundary layer without film cooling under conditions of high and low mainstream turbulence. The best performing model was evaluated using modification of the turbulent Prandtl number to attempt to better match the data for the high turbulence case. Overall, the models did not perform well for the low turbulence case; they greatly over-predicted the thermal boundary layer thickness. For the high turbulence case, their performance was better. The Transition SST model performed the best with an average thermal boundary layer thickness within 15% of the experimentally measured values. Prandtl number variation proved to be an inadequate means of improving the thermal boundary layer predictions.