Tensile damage mechanisms of carbon fiber composites at high temperature by acoustic emission and fully connected neural network

Abstract In this paper, the tensile damage mechanism of carbon fiber composites at high temperatures is analyzed. The acoustic emission technique was employed to monitor the tensile process of specimens. The acoustic emission signals at high and room temperatures were classified based on k -means and the wavelet packet energy spectrum. The results show that the damage mechanisms at high temperatures and room temperature differ. At high temperatures, there is more stress release, the material instability appears earlier, and redistribution occurs in the specimen. The damage mechanisms include matrix cracking, fiber/matrix debonding, and fiber breakage. For damage mechanism identification, the acoustic emission characteristics were used under room temperature and high-temperature conditions in the fully connected neural network, with an accuracy rate of 97.5%. The results indicate that the network is suited for both high temperatures and room temperature and can better identify various damage mechanisms.