基于参数化降阶模型的非线性气动弹性高效分析  

Effective Analysis of Nonlinear Aeroelasticity Based on Parameter Adaptive Reduced Order Model

作　　者：陆召严 肖天航[2] 常亮 邓双厚[2] 付碧红 高海云 LU Zhaoyan;XIAO Tianhang;CHANG Liang;DENG Shuanghou;FU Bihong;GAO Haiyun(Innovation Academy for Microsatellites,Chinese Academy of Sciences,Shanghai 201304,China;School of Aerospace Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing,Jiangsu 210016,China)

机构地区：[1]中国科学院微小卫星创新研究院,上海201304 [2]南京航空航天大学航空学院,江苏南京210016

出　　处：《北京理工大学学报》2024年第1期28-38,共11页Transactions of Beijing Institute of Technology

基　　金：国家自然科学基金青年科学基金资助项目(12202462)。

摘　　要：针对非线性气动弹性分析时需要在时域求解多个流场参数条件下结构运动方程而造成的计算消耗过大的问题,提出了一种适用于参数变化时非定常流场高效计算的参数化降阶模型,不仅可以应用于计算机翼等结构的总体气动力还可以得到每个时刻的结构表面的流场数据分布,并成功应用于典型机翼的跨声速颤振边界的计算,大大地提高了计算效率.结果显示,单流场条件时降阶模型的计算速度比直接使用时域分析方法提高了3倍;在计算多流场参数条件下,参数化降阶模型相比于使用单流场降阶模型计算速度提高了3.5倍,相较于时域分析方法提高了10倍.A parameter adaptive reduced order model was developed to calculate the aerodynamics of aircraft structures and to be applied to the transonic flutter boundary prediction.To get the mode information adapted to the flow condition variation,an interpolation based on Grassmann manifold was employed.For the calculation of the mode coefficients during the variation of the flow condition,a local linear neuro-fuzzy model was similarly employed with adding an extra input of the varying parameter.Based on this,the calculation of aerodynamic for a 3D LANN wing was arranged to validate the developed method and then to be applied in transonic flutter boundary prediction successfully.The results show that the computational efficient of the present reduced order can increase 3 times,compared with the time domain analysis method,and its calculation speed can improve 3.5 times,while the parameter adaptive reduced order model can increase 10 times when investigating of a multi flow conditions aeroelastic problem.

关 键 词：非定常计算流体力学 参数化降阶模型 局部线性神经模糊模型 非线性气动弹性 Grassmann流形插值 

分 类 号：V474.2[航空宇航科学与技术—飞行器设计]