Blade geometry optimization for axial flow compressor

Blade geometry design is crucial for turbomachinery performance and stability. A blade optimization method is developed and presented in this paper. The method consists of several elements: geometry parameterization, numerical simulation of flow and performance, optimization algorithm. Sweep and lean are defined by centroid coordinates of sectional blades while the blade geometry is represented by B-Splines approach, through which deformation of blade shape in terms of sweep and lean can then be readily realized. Flow simulation is carried out with a commercial CFD package-NUMECA. In addition, response surface is incorporated to approximate the objective function during the optimization process for the purpose of time-saving, and two different models (polynomial model and basic-function model) are applied respectively to construct the response surface and their differences highlighted. Genetic Algorithms is incorporated to achieve a global optimal geometry of blade. For case study, the developed method was used to optimize the transonic NASA Rotor67 blading. It is demonstrated that an optimal tuning of blade sweep and lean results in evident improvement of both isentropic efficiency and pressure ratio at design and off-design flow.