The operation modal analysis of the structure crack fault diagnosis based on pseudo-successive data

Abstract

In order to monitor the crack propagation of the structure in the working state for a long time, an operation modal analysis method based on pseudo-successive data is proposed. The vibration response signals of the cantilever beam under white noise excitation are collected and the modal parameters are extracted by the time-frequency operation modal analysis method based on the complex Morlet wavelet. In comparison with the experimental modal analysis results of hammering method, it is revealed that the error of the time-frequency operation modal analysis method is less than 10 %. By setting cracks of different lengths on the cantilever beam, the vibration response signals are extracted, and the modal parameters are extracted by the operation modal analysis method separately. By comparing those modal parameters above, it is found that the natural frequencies of the second, the fourth and the sixth orders decrease with the increase of the crack depth, and the changes of natural frequencies show the monotonicity. So, it can be used as an index for quantitative identification of crack damage. The pseudo continuous data monitoring signals of crack propagation can be constructed by means of “first discrete, then continuous”. The modal parameters changes of the whole crack propagation can be observed in one time plane by means of the operation modal analysis method. Therefore, the effective monitoring and diagnosis of the structure can be completed in case of excessive data of long-time vibration monitoring signals.