Damage detection of a hybrid composite laminate aluminum/glass under quasi-static and fatigue loadings by acoustic emission technique

This paper deals with the characterization by Acoustic Emission technique of damages occurring in a hybrid laminate aluminium/glass during quasi-static and fatigue tests. Indeed, hybrid laminates materials metal/composites are more and more considered in structure, automotive and aerospace designs because of their good mechanical performances and lightness. To understand their damages characteristics, several types of laminates (fiber orientations, number of folds, presence or not of an aluminium sheet) have been tested. The acoustic emission analysis has been realized using statistical multi-parameters methods of data clustering: combination of Principal Components Analysis (PCA) and k-means methods in unsupervised analysis and Classification and Regression Trees (CART) in supervised analysis. Using these methods, it was possible to identify damages occurring during both quasi-static and fatigue tests. Acoustic emission parameters such as counts to peak (PCNTS), amplitude, duration, counts and frequency come out as the most relevant to classify the damage mechanisms; and with the energy parameter, friction mechanisms that can occur during fatigue tests have been sorted