基于高级统计能量分析的周期加筋板振动特性研究

Study of vibration propagation in periodic rib-stiffened plates using advanced statistical energy analysis

统计能量分析(statistical energy analysis,SEA)是复杂耦合系统中、高频动力学特性计算的有力工具.本文以波传播理论和SEA的基本原理为基础,研究周期加筋板中弯曲波传播特性.分析了周期结构的频率带隙特性和加强筋对板上弯曲波的滤波特性对SEA计算结果的影响规律,发现经典SEA由于忽视了加筋板中物理上不相邻子系统间存在的能量隧穿效应,而导致响应预测结果产生最高近40 d B的误差.为了解决这一问题,本文应用高级统计能量分析(advanced statistical energy analysis,ASEA)方法,考虑能量在不相邻子系统间的传递、转移和转化的物理过程,从而大幅提高子系统响应的预测精度,将误差在大部分频段降低至小于5 d B.设计了模拟简支边界条件的加筋板振动测试实验装置,实验测试结果与有限元结果符合较好,对理论模型进行了验证.

Statistical energy analysis （SEA） is widely used in predicting dynamic response of complex coupled systems. This paper studies the bending wave propagation in periodic rib-stiffened plates in the framework of SEA. Effect of frequency band gap property of the rib-stiffened plate and wave filtering characteristics of the stiffened ribs on the prediction results of SEA is analyzed by using the wave approach and Bloch theory. It is found that due to the fact that classic SEA ignores an energy“tunneling mechanism”between subsystems that are not physically connected, large error up to almost 40 dB is generated in the subsystems of the plate compared with the results calculated from the finite element method. This tunneling mechanism mainly results from the wave filtering effects caused by the periodic arrangement of the ribs and it plays a significant role on the subsystem response at high frequencies. However, this is not incorporated in the modelling of classic SEA thus large errors can occur. To solve this problem, an advanced statistical energy analysis （ASEA） is used to consider the transition, transmission and transport of energy between unconnected subsystems. ASEA divides the energy of each subsystem into two parts： available energy which is the modal energy that could transmit into connected subsystems, and unavailable energy that dissipates within the subsystem;therefore the energy cannot propagate further away. Then the ray tracing algorithm is used to track the power flow across subsystems. By using ASEA, the accuracy of the prediction results can be greatly improved so that the error is reduced to less than 5 dB in most frequency bands. An experimental set-up is also designed to support the plate by simulating the simply-supported boundary conditions along the edges. The test results agree well with the finite element method, and it is su？cient to validate the theoretical models.