RANS Prediction of Losses and Transition Onset in a High-Speed Low-Pressure Turbine Cascade

Abstract

Current trends in aero-engine design are oriented at designing high-lift low-pressure turbine blades to reduce engine weight and dimensions. Therefore, the validation of numerical methods able to correctly capture the boundary layer transition at cruise conditions with a steady inflow for high-speed blades is of great relevance for turbine designers. The present paper details numerical simulations of a novel open-access high-speed low-pressure turbine test case that are performed using RANS-based transition models. The test case is the SPLEEN C1 cascade, tested in transonic conditions at the von Karman Institute for Fluid Dynamics. Both physics-based and correlation-based transition models are employed to predict blade loading, boundary layer characteristics, and wake development. 2D simulations are run for a wide range of operating conditions ranging from low to high transonic Mach numbers (0.7–0.95) and from low to moderate Reynolds numbers (70,000–120,000). The γ-Re˜θt transition model shows a good performance over the whole range of simulated operating conditions, thus demonstrating a good capability in both reproducing blade loading and average losses, although the wake’s width is underestimated. This leads to an overestimation of the total pressure deficit in the center of the wake which can exceed experimental measurements by more than 50%. On the other hand, the k-ν2-ω model achieves satisfactory results at Ma6,is = 0.95, where the boundary layer state is affected by the presence of a weak shock impinging on the blade suction side which thickens the boundary layer, leading to a predicted shape factor equal to five, downstream of the shock. However, at low and moderate Mach numbers, the k-ν2-ω model predicts long or open separation bubbles contrary to the experimental findings, thus indicating insufficient turbulence production downstream of the boundary layer separation. The slow boundary layer transition in the aft region of the suction side that is exhibited by the k-ν2-ω model also affects the prediction of the outlet flow, featuring large peaks of a total pressure deficit if compared to both the experimental measurements and the γ-Re˜θt predictions. For the k-ν2-ω model, the maximum overestimation of the total pressure deficit is approximately 60%.

Keywords

• low-pressure turbine
• high-speed low-reynolds number flows
• computational fluid dynamics
• transition modeling