ABSTRACT

Considering the frequency dispersion and the scale sensitivity in wave propagation, the central frequency and bandwidth should be carefully selected in Lamb wave-based structural health monitoring. These crucial parameters are hardly determined without the prior knowledge. Therefore, massive trials should be conducted before the formal test. A Fourier spectral Frequency Response Function (FRF) method is proposed to address this problem. As the FRF contains wide-band information on the relationship between excitation and output, the Lamb wave responses in a wide frequency range can be estimated by using an arbitrary virtual input within the frame of single FRF test, being free from the massive experimental works. The theoretical FRF can be obtained from the unit impulse response of a linear system, but this path is impracticable due to the difficulty of exciting the ideal impulse. Considering the integral relationship and Fourier differential properties between step excitation and the ideal impulse, a robust approximation of wave signal by FRF is developed. Synthetic data tests and experiments both verified that the present method can accurately extract the response in isotropic and anisotropic materials for the specifically designed excitation signal in a wide parameter range.

摘要

考虑到波传播中的频率色散和尺度敏感性，在基于兰姆波的结构健康监测中应谨慎选择中心频率和带宽。如果没有先验知识，这些关键参数很难确定。因此，在正式测试之前应进行大规模试验。提出了一种傅里叶谱频率响应函数（FRF）方法来解决这个问题。由于 FRF 包含关于激励和输出之间关系的宽带信息，因此可以在单个 FRF 测试框架内使用任意虚拟输入来估计宽频率范围内的 Lamb 波响应，而无需大量的实验工作。理论频响函数可以从线性系统的单位脉冲响应中获得，但由于难以激发理想冲动，这条路是行不通的。考虑阶跃激励与理想脉冲之间的积分关系和傅里叶微分特性，提出了一种稳健的频响函数逼近波信号。综合数据测试和实验都验证了本方法可以准确提取各向同性和各向异性材料在宽参数范围内对专门设计的激励信号的响应。