论文部分内容阅读
采用特征正交分解(POD,Proper Orthogonal Decomposition)方法,建立了基于状态空间的非定常气动力模型,耦合结构方程,建立了降阶的气动弹性系统,开展了气动弹性虚拟仿真研究,与CFD/CSD全耦合相比,效率提高了2～3个数量级。具体过程是:首先获取全阶系统的频域快照构成POD核,通过对POD核进行奇异值分解提取POD模态(或POD基),对低能量模态截断形成降阶子空间,并将其映射到全阶系统,从而形成基于状态空间的降阶的非定常气动力模型。对气动弹性标模AGARD445.6进行算例验证,证明了降阶方法正确,可以提供高效、高精度的气动弹性分析。
A Pseudo-orthogonal Decomposition (POD) method was used to establish the unsteady aerodynamic model based on state space and the coupled structural equation. A reduced-order aeroelastic system was established and aeroelastic virtual simulation was carried out. Compared with CFD / CSD compared to full coupling, the efficiency increased by 2 to 3 orders of magnitude. The specific process is as follows: Firstly, the POD kernel is obtained by taking the frequency-domain snapshots of the whole-order system and the POD kernel is extracted by singular value decomposition of the POD kernel to form a reduced-order subspace by cutting off the low-energy mode, Mapped to the whole order system, thus forming a reduced order unsteady aerodynamic model based on state space. The numerical simulation of the aerodynamic elastic modulus AGARD445.6 shows that the method of reduced order is correct, which can provide highly efficient and highly accurate aerodynamic analysis.