WALL RESOLVED LARGE-EDDY SIMULATIONS OF THE SPLEEN LOW-PRESSURE BLADE CASCADE

The field of aircraft propulsion is currently facing major energy and environmental challenges, requiring faster improvements in engine performance and therefore a better understanding of turbine flows. This has led to the development of new ultra-high bypass ratio engine technologies, in which the fan driven by the low-pressure turbine through a gearbox is responsible for about 80% of the thrust. As the various experimental investigations carried out are limited by the complex nature of the flows and geometries involved, it becomes important to build numerical methods that can correctly predict such flows. However, accuracy in numerical predictions remains a real challenge.
Low pressure turbine blades in geared turbofan engines operate at transonic exit Mach numbers (between 0.6 and 0.9), and experience fairly low Reynolds number flow regimes, allowing to perform wall resolved large eddy simulation of cascade flows, and avoiding wall modelling issues in presence of adverse pressure gradients. This work focuses on the Wall Resolved Large-Eddy Simulations (WRLES) of flows in High-Speed Low-Pressure
Turbine (HS-LPT) transonic cascades. The test case investigated is the ”SPLEEN” cascade, a next-generation HS-LPT designed at the von Karman institute for fluid dynamics, in collaboration with Safran aircraft engines within the ”Secondary and Leakage Flow Effects in High-Speed Low-Pressure Turbines” EU project. These WRLES are performed with the Explicit Compressible Solver (ECS) of the massively parallel code YALES2, using a high-order Finite Volume Method (FVM). The first simulation campaign presented here concerns the
SPLEEN case for the following isentropic exit quantities M2s = 0.7 and Re2s = 70k; the aim being to access the solver for transonic flows, in terms of prediction of the suction-side separation bubble, the separation-induced laminar-turbulent transition and the wakes. The results obtained, compared with high-fidelity experimental data, show fairly good agreement, further studies are in progress to improve wake prediction.