Image reconstruction and characterisation of defects in a carbon fibre/epoxy composite monitored with guided waves

Abstract

In this extensive experimental study, the structural health monitoring (SHM) of a large 1 m × 1 m carbon fibre reinforced polymer laminate was performed using guided waves. Multiple hole- and crack-type defects were induced in the structure and the guided wave signals were collected using a circular network of piezoelectric disc transducers permanently bonded to the structure. Images representing each damage state of the composite plate were obtained by applying the (total focusing method, TFM with full matrix capture) on guided wave signals. Several variations in the image reconstruction algorithm were investigated by using two different baselines—pristine signals or signals collected during the last damage state-, processing only the positive amplitude by implementing a lower limit threshold and normalising the signals. The algorithms’ ability to detect and localise multiple defects, inside and outside the array, including holes as small as 2 mm in diameter, was evaluated. TFM images, involving the use of a lower threshold limit in the signal processing, resulted in more accurate detection. In addition, images obtained using both types of baseline provided complementary information leading to increased confidence in the system. All defects in the plate were detected and located even in the presence of multiple defects. Also, regions corresponding to crack- and hole-type of defects in the resulting images were identified, quantified and differentiated using geometric circularity and eccentricity shape factors. This ability of accurately identifying multiple defects and reducing location error are important contributions in the effort of establishing a reliable SHM system for multi-layered composite structures.