Study On Wavelet-Based Damage Detection In Pipes Using Ultrasonic Longitudinal Guided Wave

A study on defects detection of the hollow and liquid-filled pipes was carried out by comparison of their detection signals and dispersion curves. A defect identification method in pipes using ultrasonic longitudinal guided-wave is studied. The numerical data of damage detection signals of the pipe based on finite element method are derived by LS-DYNA, involving the reduced stiffness and thickness of elements on the defect location. Furthermore, wavelet transform is proposed to identify the imperceptibility defect. By means of reflection signal peak intensity, the sensitivity of the proposed approach versus the change of local stiffness in pipe wall is evaluated. The effect of noise on the identified results is also discussed. Defect identification in the low signal-noise ratio is realized based on the algorithm of wavelet decomposition and reconstruction. The specific disturbance mode in liquid-filled pipes, as well as its frequency dependence, is analyzed. The results show that the disturbance mode can be adjusted and restrained remarkably by choosing some appropriate exciting signs. Furthermore, the difference of the defect reflecting signals with variable damage levels in the hollow and liquid-filled pipes indicates that there is a nonlinear relationship between the extent of the energy dissipation and the damage extent in the liquid-filled pipes. The present numerical analysis proves a good agreement with the corresponding experiment results.