Abstract
The 3D aerodynamic design optimization has been applied in the generation of modern turbine blade profile. However, the traditional design method paid little attention to the decrease of heat transfer coefficients on the blade external surface. In the present work, a typical high load turbine vane, VKI LS89 cascade, was optimized with the decrease of aerodynamic loss and heat load chosen as the optimization objective functions. Numerical simulation methods were validated by the experiment data, and simulations results agreed well with the measured values. Both 2D profiles and stagger curves of the vane were parameterized by no-uniform B-Spline. There were totally seven movable control points for the 2D profiles, and four movable control points for the corresponding stagger curves. And the locations of the B-Spline control points and stagger angles were taken as the design variables. Multi-objective genetic algorithm coupled with surrogate model was adopted to acquire the optimal cases with better aero-thermal performance. The profiles of the vane were firstly optimized in a linear cascade model, and then the stagger curves and sections stagger angle were modified for better overall performance. Mass flow rate of the mainstream and exit flow angle at outlet were constrained by the comprehensive objective functions during the 3D optimization process. The results showed that profiles with high aerodynamic efficiency and low heat load can be obtained by the 2D profiles optimization design. Additionally, the heat load could be decreased by the 3D optimization design. Furthermore, the effects of optimization on the heat load distributions of the endwall were studied, and it can be observed that the 3D optimization obviously modified the heat transfer patterns of the endwall.