Abstract
Characterization of acoustic emission (AE) signals in loaded materials can reveal structural damage and consequently provide early warnings about product failures. Therefore, extraction of the most informative features from AE signals is an important part of the characterization process. This study considers the characterization of AE signals obtained from bending experiments for carbon fiber epoxy (CFE) and glass fiber epoxy (GFE) composites. The research is focused on the recognition of material structure (CFE or GFE) based on the analysis of AE signals. We propose the extraction of deep features using a convolutional autoencoder (CAE). The deep features are compared with extracted standard AE features. Then, the different feature sets are analyzed through decision trees and discriminant analysis, combined with feature selection, to estimate the predictive potential of various feature sets. Results show that the application of deep features increases recognition accuracy. By using only standard AE-based features, a classification accuracy of around 80% is obtained, and adding deep features improves the classification accuracy to above 90%. Consequently, the application of deep feature extraction is encouraged for the characterization of loaded CFE composites.
Highlights
Composite materials are increasingly used in lightweight transportation systems and civil engineering due to increasing weight constraints and installation costs
When the carbon fiber epoxy (CFE) sample breaks, the highest amplitudes and energies of acoustic emission (AE) signals are measured, while in the case of glass fiber epoxy (GFE) samples, after sample breaking the values of the amplitude and energy remain high
The results indicate the good potential of introducing learning m the need for automated AE-based feature extraction, where the informative features are odsautomatically into the AE-based characterization of loaded materials
Summary
Composite materials are increasingly used in lightweight transportation systems and civil engineering due to increasing weight constraints and installation costs. The composite constituents (fiber, matrix, and interfacial bonding), and the applied mechanical load (stress level of loading sequence, stress ratio) influence the damage evolution [1]. This aggravates the development of damage-tolerant design procedures like those used for metallic materials. In polymer composites, this is mostly related to the development of experimental techniques that can monitor material behavior and offer real-time information about damage evolution [2]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
