Cluster analysis of acoustic emission signals for tensile damage characterization of quasi-static indented carbon/glass fiber-reinforced hybrid laminate composites

Abstract

Applying acoustic emission method to understand the damage and failure mechanism of hybrid composites is challenge for industry. For this study, nine types of carbon/glass fiber-reinforced hybrid laminate composite specimens with different structure and indentation were subjected to tensile experiments. The AE signals collected during the tensile process were post-processed using cluster analysis based on Fuzzy C-Means algorithm. It was found that the AE signals can be divided into three types which correspond to three damage modes, and the AE peak frequency characteristics of each damage mode were found for different specimens. Micro-computed tomography imaging was obtained for specimens with indention before loading and after failure. There is seen to be a good correlation between the damage seen with the Micro-CT imaging, the mechanisms identified and the data cluster analyzed by using AE signals. Results prove that AE signals are reliable and can be used for composite structure health monitoring.