叶轮机械叶栅多目标多学科设计优化算法  

MULTI-OBJECTIVE AND MULTIDISCIPLINARY OPTIMIZATION DESIGN METHOD OF TURBOMACHINERY BLADE

作　　者：罗常[1] 宋立明[1] 李军[1] 丰镇平[1] 

机构地区：[1]西安交通大学叶轮机械研究所,陕西西安710049

出　　处：《工程热物理学报》2009年第10期1651-1654,共4页Journal of Engineering Thermophysics

基　　金：国家自然科学基金资助项目(No.50776065)

摘　　要：提出了基于多目标寻优的叶轮机械叶栅多学科设计优化算法,方法包括:采用并行多目标差分进化算法作为优化求解器来搜寻叶栅多学科设计优化问题的Pareto解集,采用非均匀B样条方法对叶片型面进行参数化处理,通过求解Reynolds-Avergaed Navier-Stokes方程评估叶片的气动性能,耦合气动计算得到的叶片表面压力,应用有限元分析方法预测叶片的强度性能.为证明本文方法的实用性,选择叶片的等熵效率和叶片应力为目标函数,完成了NASA Rotor37转子叶栅的多学科设计优化,结果表明本文提出的多学科设计优化算法具有良好的优化性能。Automated multi-objective and multidisciplinary optimization and design of turbomachinery blades are proposed in this paper. In this method, an algorithm named Multi-objective Differential Evolution （MDE） is used as a solver to find the Pareto solution sets of multidisciplinary design problem. The Non-uniform B-Spline approach is adopted to parameterize the turbomachinery blade profiles. The aerodynamic performances of design blade candidates are predicted with a three-dimensional Reynolds-Averaged Navier-Stokes （RANS） solution. The blade stresses are evaluated with means of a finite element analysis coupled with the surface pressure of blades obtained with CFD calculation. To validate the method in the multi-objective and multidisciplinary optimization, the NASA rotor 37 is optimized for the maximization of the isentropic efficiency and the minimization of the maximum stresses. The results demonstrate that the presented multi-objective and multidisciplinary optimization methodology has a good optimization performance.

关 键 词：叶轮机械叶栅 Paxeto解集 多学科设计优化 并行多目标差分进化算法 

分 类 号：O354[理学—流体力学]