Multi-objective optimization of aerodynamic and erosion resistance performances of a high-pressure turbine

Abstract

Focusing on a high-pressure turbine, this study conducts a multi-objective optimization aimed to improve its aerodynamic and erosion resistance performances. With a self-adaptive updated Kriging model first applied, the Latin hypercube sampling method and NSGA-II algorithm are then used to find trade-off solutions. The sensitivity analysis shows that the efficiency is mainly influenced by the flow angles near trailing edges of three selected blade spans, among which the flow angle near the trailing edge of middle span is the most significant, followed by the root span and tip span, while the dominant design parameters affecting the erosion are the flow angles near trailing edges of middle span and tip span. After optimization, the erosion around the middle span on pressure side and middle chord on suction side is significantly reduced due to the decrease of impact velocity and impact frequency, and the increase of impact angle. Besides, the low-velocity fluid regions near the pressure side of middle span and the trialing edge of tip are greatly reduced, which can relieve the blockage of flow passage. At all axial sections considered, the optimized blade is shown to significantly reduce losses near the tip and suction side, especially at 80% axial section.