Impact response analysis and Physics-Informed damage classification of sandwich composites using electrical resistance-based self-sensing

Abstract

Sandwich composites face maintenance challenges from low-velocity impacts due to their intricate geometry. We propose a method that uses the self-sensing property of carbon fiber reinforced plastic skins to determine the health of an entire sandwich structure in real time. To reflect the hierarchical nature of impact-induced damage progression and the corresponding increase in signal response, a superposition method was proposed. This method superposes binary classifications to construct a damage index map, addressing the low sensitivity of these specimens and small number of samples. This algorithm achieved an accuracy of 81.36%, which is significantly better than the 74.58-79.69% obtained using the conventional algorithm. Moreover, the color layer yields complementary information on the classification results, instilling confidence in these results. This algorithm has potential applications in hierarchical class relationships. The findings show that self-sensing can be useful in monitoring various complex structures with blind regions.