Structural damage localization using Lamb Wave signals filtered with Gabor functions

Over the past two decades, Structural Health Monitoring (SHM) has gained significant interests as an alternative to the conventional nondestructive evaluation (NDE) techniques for thin-walled structures. Among several different methods investigated in SHM, Lamb Wave-based techniques have many advantages. For the excitation and detection of the Lamb Waves in thin-walled structures, piezoceramic patch transducers are typically used. By exciting Lamb Waves and measuring reflected signals from various places in the structure, one can determine the location and severity of the damage. When the excitation voltage is low or the measurement signal is contaminated with noise, signal processing is usually applied to increase the Signal-to-Noise Ratio (SNR) for measurement signals using Lamb Wave.;In this research, a Gabor filtering technique was applied to a SHM damage detection method to improve detecting and locating structural damage in a plate. In the experimental setup, four circular piezoceramic transducers that were bonded to a 0.3721 m2 square area and 1.016 mm thick aluminum plate were used to send a tone burst signal and record the response signals. In order to simulate structural damage, three holes of different sizes were punctured in different locations. The transducers were measured using an oscilloscope, the SNRs of the raw measurement data turned out to be quite low. This was due not only to the fact that the spacing between the transducers was relatively large at 0.30 meters but also that the signals were measured using a low resolution oscilloscope. In order to improve the SNR, a series of Gabor filters at different center frequencies were applied and the time-delayed sum was calculated to localize the holes. The results show that this approach produced good damage detection and localization when the SNR of the measurement signal is low.