基于本征正交分解的串列叶栅多目标气动优化研究

Multi-objective Aerodynamic Optimization of Tandem Cascade Based on Proper Orthogonal Decomposition

为提高压气机串列叶栅复杂气动构型优化设计的效率,本文基于本征正交分解(POD)理论,建立了集几何外形参数化、样本空间降维、气动性能求解、降维模型构建和遗传算法寻优于一体的串列叶栅高效优化设计系统,针对叶型型面和叶片相对位置关系参数,开展了多目标优化设计工作。在7°攻角工况下,优化后的串列叶栅的静压升提高了0.53%,总压损失系数下降了9.25%,其他攻角条件下叶栅性能同样也得到了改善。与传统CFD方法相比,极大提高了优化效率,与基于Kriging代理模型相比,本文发展的优化系统由于缩小了设计变量空间,提高了优化的迭代效率,收敛耗时仅为Kriging代理模型的0.0796%。此外,基于POD方法的优化设计系统具有更高的建模精度,使串列叶栅获得了更高的静压升和更低的总压损失系数。优化后的串列叶栅节距系数增大、弯角比减小,减小了掺混损失,抑制了流动分离,改善了全攻角范围内的气动性能。

In order to improve the design efficiency of compressor tandem cascade with complex aerodynamic configuration,an efficient optimization design system for tandem cascade is established based on POD theory,which integrates geometric shape parameterization,dimension reduction of sample space,aerodynamic performance solution,dimension reduction model construction and genetic algorithm.Multi-objective optimization design is carried out by considering the blade profile and relative position parameters.Under the condition of 7 incident angle,the static pressure rise of the optimized cascade is increased by 0.53%and the total pressure loss coefficient is reduced by 9.25%.Cascade performances is also improved at other incident angles Compared with Kriging surrogate model,the optimization design system based on POD method improves the iteration efficiency due to reducing the space of design variables.Optimization time of the system is only 0.0796%of that of Kriging surrogate model.In addition,the optimization design system based on POD method has higher modeling accuracy,which makes the tandem cascade obtain higher static pressure rise and lower total pressure loss coefficient,After optimization,the percent pitch is increased and camber ratio is de-creased,which reduces the mixing loss,suppresses the flow separation,and improves the aerodynamic performance within the range of full incident angle.