摘要
参数化方法是翼型设计的重要方法。在翼型设计阶段,参数化模型中的变量通常认为在一定区间内变化,并进行离散化。参数化模型的变量较多时,设计样本空间大,计算量大。本文提出一种翼型的随机参数化设计方法,采用类/形函数变换方法参数化翼型,参数化模型的变量为服从一定概率分布的随机变量,基于不确定性量化方法与CFD模拟获得翼型气动性能的统计特性。将此方法与OpenFOAM耦合,对S809翼型进行了随机参数化建模,比较了蒙特卡洛、拉丁超立方和基于稀疏网格的多项式混沌法与CFD结合的精度和效率。结果表明基于稀疏网格的多项式混沌法收敛速度快和精度高,可以大幅减小高维设计变量下的计算量;通过敏感性分析寻找了影响翼型气动性能的主要设计变量。
Parameterization method is an important method of airfoil design.In the phase of airfoil design,the variables in the parametric model are usually considered to change in a certain range and discretized.When there are many variables in the parametric model,the design sample space and cost of calculation are large.In this paper,a stochastic parametric design method of airfoil is presented.The airfoil is parameterized by class/shape function transformation method,and the variables of the parameterized model are random variables with a certain probability distribution.The statistical characteristics of aerodynamic performance of airfoils are obtained based on uncertainty quantization method and CFD simulation.Coupling this method with OpenFOAM,the random parametric modeling of S809 airfoil is carried out,and the accuracy and efficiency of Monte Carlo,Latin hypercube and polynomial chaos method based on sparse grid combined with CFD are compared.The results show that the polynomial chaos method based on sparse grid has fast convergence speed and high accuracy,and can greatly reduce the calculation cost under high-dimensional design variables,and the main design variables that affect the aerodynamic performance of airfoils are found through sensitivity analysis.
作者
于佳鑫
陈江涛
王晓东
白雪峰
康顺
YU Jia-Xin;CHEN Jiang-Tao;WANG Xiao-Dong;BAI Xue-Feng;KANG Shun(Key Laboratory of Power Station Energy Transfer Conversion and System,Ministry of Education,North China Electric Power University,Beijing 102206,China;China Aerodynamics Research and Development Center,Mianyang 621000,China)
出处
《工程热物理学报》
EI
CAS
CSCD
北大核心
2021年第5期1184-1192,共9页
Journal of Engineering Thermophysics
基金
国家数值风洞工程项目课题(No.NNW2018-ZT7B14)
国家自然科学基金(No.51876063)。
