提出一种通过计算流体动力学(Computational Fluid Dynamics,CFD)建立离散时间气动模型并识别颤振导数的方法,该方法通过强迫模型按一种3211多阶跃方式运动,由CFD数值模拟得到作用在模型上的气动力。由于3211多阶跃包含了丰富的频率成份,能激励起气动力系统在这一连续频域范围内的广泛响应,因而可基于3211输入气动力输出得到定义在某一频率范围的离散时间气动模型。该气动模型能对对应不同折算风速的简谐位移输入快速仿真得到气动力输出,通过这些输入-输出就可快速识别任意折算下的颤振导数,无需对不同折算风速重复进行CFD计算。应用本文方法识别了薄平板的颤振导数,并与薄平板风洞试验结果进行了比较,颤振导数的一致性证明了本文方法的有效性。
Flutter derivatives show great importance on analysis Of bridge wind-induced vibration. They can be extracted through wind tunnel tests and CFD simulation. In this paper, a novel approach based on the CFD discrete time aerodynamic model is presented to identify flutter derivatives. This method employs the 3211 multi-step as input of velocity to excite CFD model, which can provide aerodynamic forces acting on the body through CFD simulation. An aerodynamic model in discrete time domain, defined on frequency range of interesting, can be developed based on the input-output relationship. Then the obtained aerodynamic model can be used to simulate aerodynamic response for any harmonic input with suitable frequency. Finally, flutter derivatives can be identified based on input and output time histories. The present method can extract flutter derivatives under arbitrary reduced wind speed, requiring respectively only once round of CFD simulation for heaving or pitching motion. Identification of flutter derivatives of a thin plate is carried out by using the present method, reasonable agree- ments between the identified flutter derivatives and those from Wind tunnel tests validate the efficiency of the present approach.