铝板中激光Lamb波信号的模态分析与缺陷检测研究

激光激励的Lamb波信号具有较宽的频带,且包含多个模态信息。本文采用二维傅里叶变换和时频分析等信号分析技术用于检测信号中的模态成分及缺陷信息识别。首先,对200组激光Lamb波信号进行二维傅里叶变换,得到信号的频率-波数图,可识别出激光Lamb波信号中的低阶A0、S0和高阶模态,并且A0模态能量高,可用于缺陷检测。随后对有、无缺陷状态下Lamb波信号进行连续小波变换,从时频图中识别出缺陷信号的频率成分,进一步提取特定频率下的小波系数幅值信号,实现了缺陷信息的识别。结果表明,二维傅里叶变换能较好地识别激光Lamb波的模态成分,而提取出的连续小波变换系数图,能准确实现缺陷定位。

On the Mode Identification and Defect Detection of Laser-induced Lamb Waves Signal in an Aluminum Plate

Laser-excited Lamb wave signals have wider frequency band, and contain multiple modes. The modal composition and defect information identification in signals were detected by signal analysis techniques such as two-dimensional Fourier transform and time-frequency analysis. Firstly, two-dimensional Fourier transform was employed to analyze 200 groups of laser-induced Lamb wave, based on which, the frequency-wave number graph was obtained to identify the low order modes A0, S0 and high order modes in Lamb wave signals. Among them, mode A0 has high energy and can be used for defect detection. Subsequently, continuous wavelet transform was used to analyze the Lamb wave signals with or without defects. The frequency component of defect signals can be identified from the wavelet coefficients spectra. Furthermore, the wavelet coefficient amplitude in signals can recognize the defect which is under particular frequency component extracted from the wavelet coefficients spectra. Results show that two-dimensional Fourier transform can identify the mode composition of laser-induced Lamb wave, and the extracted continuous wavelet transform coefficient signals can locate defects accurately.