Identification of wear mechanisms of glass/polyester composites by means of acoustic emission

The acoustic emission (AE) technique was used for condition monitoring of pultruded glass/polyester composites subjected in abrasive wear. Three wear mechanisms were recognized by means of wavelet and cluster analyses of the AE data. We were able to associate these mechanisms with fiber breakage, debonding and a hybrid type based on the results of pattern recognition, previously performed for the signals recorded during tensile tests. For the latter tests, the temporal order of appearance of the different damage mechanisms allows to draw conclusions more easily about the correlation with the AE signals. A number of AE features were selected for classification using parameter-less self-organized mapping (PLSOM), which is a type of neural network that is not bound to the naturally subjective learning rate, neighborhood function and their annealing with the training progress.