基于流态特征的大跨度桥梁颤振机理研究

Good flutter stability is the key to the design of super long span bridges. However, the mechanism of the flow pattern around the bridge deck on flutter is still not very clear, which will hinder the development of our country to the larger span bridges. This project is aimed at the research on the basic frontier scientific problem in the bridge wind engineering. Firstly, based on the results of wind tunnel testing to bridge flutter, the influences of the wake patterns, the leading separation flow patterns and the change of the flow state on the flutter stability of a long-span bridge will be studied, and the linear and nonlinear mapping relationships between the characteristic flow mode and the structural vibration will be established. According to the above analysis, the influence mechanism of the flow pattern on the whole process of flutter will be revealed. Secondly, based on the results of 3-D numerical simulation of the flutter of large span bridge, the transfer relationships among the energy distribution of characteristic flow, the work distribution of the surface pressures and the vibration energy f bridge structure will be established, and the flutter mechanism based on the flow pattern characteristics will be revealed form the point of energy transfer. Thirdly, based on the results of dynamic mode decomposition (DMD) on the flow around the bridge deck, the vortex dynamics models of characteristic modes of wake, leading separation flow will be established. And the linear and nonlinear unsteady aerodynamic models considering the effects of the flow pattern characteristics of wake and the leading separation flow will be deduced. Finally, the local flow control methods to bridge flutter will be studied, and the flow control theory of flutter based on the flow pattern will be established.

良好的颤振稳定性是超大跨度桥梁设计的关键，然而目前桥梁绕流场流态对颤振的作用机理仍不够明确，这必将阻碍我国向更大跨径桥梁发展。本项目针对这一桥梁风工程中的基础前沿科学问题展开研究。首先，基于桥梁颤振风洞试验结果，研究钝体尾流及流态变化、前缘分离流动及流态变化对大跨度桥梁颤振稳定性的影响规律，建立特征流动模态与结构振动的线性和非线性映射关系，揭示绕流场流态对桥梁颤振全过程的作用机理。其次，基于桥梁颤振三维CFD数值模拟结果，建立绕流场特征流动能量分布、结构表面压力做功及结构振动能量三者间的传递关系，从能量输运角度揭示绕流场流态对结构颤振的作用机理。第三，基于桥梁颤振全过程绕流场动态模态分解结果，建立尾流、前缘分离流动绕流场特征模态的涡动力学模型，推导考虑尾流流态和前沿分离流流态特征的线性和非线性气动力模型。最后，研究大跨度桥梁颤振局部流态控制方法，建立基于流态特征的大跨度桥梁颤振控制理论。

良好的颤振稳定性是超大跨度桥梁设计的关键。大跨度桥梁的颤振问题是典型的流固耦合问题，然而目前桥梁绕流场流态对颤振的作用机理仍不够明确，本项目针对这一桥梁风工程中的基础前沿科学问题展开研究。首先基于泰勒冻结假设启发提出绕流场流速时空关联非线性函数，提出了一种时空深度神经网络，通过少量高频离散风速监测点和低频PIV重构出高时间分辨率绕流场。其次、研究了钝体刚性平板的颤振响应特性，通过外激励荷载得到了亚临界 Hopf 分岔中的不稳定界限换幅值，给出了完整的分岔图和相应的分岔方程。基于流场显示技术揭示了颤振发生过程中的流场演化规律，研究发现在颤振发展早期，前缘涡随振动振幅增大而增大并在传播过程中分裂成较小的漩涡。在充分发展的颤振阶段，前缘涡急剧扩展并向后缘传播，最终随着振动进入尾流区。通过绕流场涡量相关性分析，发现随着来流风速增加，前缘流动分离诱发的涡量分布向下游发展更快，出现相位提前现象，进而导致气动力与振动之间的相位差增大，使得流场向模型输送了更多的能量，最终激发了颤振行为。采用基于流态的钝体绕流气动力重构方法，首次建立了矩形钝体绕流场不同区域关键流态、气动力和颤振响应三者间的定量关系，结果表明边界层和剪切层区域对颤振气动力的贡献为83%-94%，进一步证明了前缘分离流动对矩形结构颤振的控制作用。第三、提出了嵌入动力学控制方程的颤振响应和气动力计算深度神经网络，实现了不同气动外形桥梁非线颤振响应和气动力的长时准确预测。此外，提出桥梁颤振气动力方程机器学习识别方法，获得不同气动外形下气动方程随风速的变化规律。最后提出了一种多孔材料覆盖层控制流动分离的自激振动流动控制方法。.在本项目的资助下，发表SCI论文20篇；参加国际会议3人次，国内会议4人次，其中在国内外会议邀请报告4次；获省部级科技奖励3项； 1人入选2021年黑龙江省优青，1人入选2020年中国博士后科学基金资助者选介（全国仅100人）；培养研究生13名（已毕业6名）。