Utilizing Optimization Tools for Profile Contouring to Reduce Endwall Losses

Low-pressure turbines (LPTs) play a crucial role in fuel efficiency and thrust generation of aero-engines. Traditional LPT designs, however, involve multiple stages and numerous blades, resulting in increased weight and manufacturing costs. The challenge of modern day researchers is to reduce the weight and cost but maintain the high efficiency of LPTs. To address this, one approach is to increase the aerodynamic loading of individual blades, reducing the blade count. However, this can lead to increased secondary flow losses, particularly in the endwall regions, where flow separation is more prominent. This research focuses on optimizing the blade profile at the junction with the endwall to reduce these losses. A binary evolutionary genetic algorithm is used to design an improved contoured blade geometry that minimizes total pressure loss across the blade row. By modifying user-defined parameter bounds, two optimization routines produced a high and low stagger angle optimized profile. The optimized profiles were experimentally validated in a low-speed linear-cascade wind tunnel environment, and comparisons between the computational predictions and experimental measurements are presented. Experimentally, the total passage loss was reduced by 15.2% and 20.2% for each optimization routine. The observed reduction in passage loss confirms the performance gains and demonstrates the effectiveness of the optimization framework in improving the high-lift LPT blade design.