Multi-parameters approach for damage propagation analysis in Fiber Reinforced Polymer Composites

Abstract

Assessing damage evolution in Carbon Fiber Reinforced Polymer (CFRP) composites is an intricate task due to the complex mechanical response of the composites. The acoustic emission technique (AE) is a non-destructive technique based on the recording of sound waves generated inside the material as a consequence of the presence of active defects. Proper analysis of the recorded waves can be used for monitoring damage evolution in many materials including composites. The acoustic track associated with the entire loading history of the sample or the structures is usually followed by using some descriptors such as the amplitude of the sound waves, the number of counts and so on. In this paper the acoustic emission in CFRP single lap shear joints will be monitored by using a multiparameter approach based on the contemporary analysis of multiple features such as ASL, Initiation Frequency, Reverberation Frequency and so on to understand if a proper combination of them can be adopted for a more robust description of damage propagation in CFRP structures. For selecting the best features, Principal Component Analysis (PCA) is used. The selected features are classified into different clusters using Fuzzy c-means (FCM) data clustering for analyzing the damage modes.