Investigation of CFD Prediction Capabilities for Low Reynolds Turbine Aerodynamics

The objective of this work is to study the performance of low pressure turbines operating at low Reynolds numbers by extensive experiments and to validate numerical simulation results with the experimental data. Particular attention is payed to the prediction capabilities of current numerical turbulence and transition models in order to be able to benchmark the performance of future turbine airfoil profiles and to optimise their aero design. The LPT-Cascade under consideration has been investigated at the High Speed Cascade Wind Tunnel of the Institute of Jet Propulsion to gather information about the performance of turbine airfoils under low Reynolds operating conditions. The experiments were executed in the range of Re = 40^′000 to 400^′000 with steady state inflow conditions at different Mach number levels. The main focus of the investigation thereby was on the range of Re = 40^′000 to 70^′000. The high speed cascade wind tunnel of the University of Federal Armed Forces Munich allows for an independent Reynolds and Mach number variation such that an extensive database can be generated for realistic engine operation conditions. One major test objective was related to flow separation phenomena on the suction surface and its influence on the performance of the turbine profile. For this purpose both the loss behaviour and the pressure distribution on suction and pressure surface of the blade were measured and analysed. In addition to the experiments numerical flow simulations were conducted for the same turbine profile. In order to achieve more information on the influence of different turbulence and transition models on the flow separation, transition, and reattachment behaviour, two different CFD codes were used for comparison purposes. On the one hand the CFD code TRACE, which is developed by the German Aerospace Center (DLR) and MTU Aero Engines and on the other hand the general purpose code ANSYS CFX were applied. The aim is to assess the prediction capabilities of the different codes especially in the low Reynolds number range.