In this work, the most promising ultrasonic signal processing methods—discrete wavelet transform, variational mode decomposition and Hilbert transform—are applied for the analysis of disbond-type defects in the segment of wind turbine blade. Two disbond-type artificial defects having diameters of 81 and 25 mm were located on the main spar of wind turbine blade. The low-frequency ultrasonic system developed by Ultrasound Research Institute of the Kaunas University of Technology was used for the experimental investigation of wind turbine blade using guided waves and only one side of the blade segment was accessed. Two contact type ultrasonic transducers separated by 50 mm distance and fixed on a movable mechanical panel were used as a transmitter–receiver pair during the experiment for the ultrasonic signals recording up to the scanning distance of 250 mm with the scanning step of 1 mm. Both types of defects were marginally detected from the conventional experimental B-scan and therefore appropriate signal processing techniques were used to improve the accuracy of the analysis of defects. The discrete wavelet transform was combined with the amplitude detection method for estimating the size and location of defects. Finally, the variational mode decomposition is combined with the Hilbert transform to compare the instantaneous frequencies and amplitudes of the defect-free and defective signals as well as for the measurement of time-delays between the defect-free and defective signals.
Dirt and mud detection and diagnosis on a wind turbine blade employing guided waves and supervised learning classifiers
