Simulations and validation of an axial-flow pit turbine

Abstract

The geometric design of an axial-flow pit-turbine is essential as flow separation will decrease turbine efficiency. In this thesis, steady Reynolds Averaged Navier Stokes (RANS) simulations were conducted with a two-equation closure model and validated with experimental results. The research focuses on evaluating the performance of a k-omega Shear Stress Transport (SST) turbulence model with wall functions in separated flows. RANS simulations were carried out in OpenFOAM using an unstructured grid for both lab and full-scale models of the pit-turbine. By testing a lab-scale model, Particle Image Velocimetry (PIV) was performed to provide insights into the flow behaviour in regions of separation. It was observed that the pressure results from the lab and full-scale RANS models were in agreement with experimental data except in regions of an adverse pressure gradient. The k-omega SST model is sensitive to wall-functions and therefore flow properties near the wall are incorrectly calculated in regions of flow separation. At sharp streamline curvatures, the pressure drop is overpredicted for both the lab and full-scale models. However, at very large Reynolds numbers for the full-scale turbine, the turbulence model underpredicts and delays flow separation at a larger radius of curvature when compared to experiments.