An Artificial Intelligence-Based Scheme for Structural Health Monitoring in CFRE Laminated Composite Plates under Spectrum Fatigue Loading

In the fabrication and monitoring of parts in composite structures, which are being used more and more in a variety of engineering applications, the prediction and fatigue failure detection in composite materials is a difficult problem. This difficulty arises from several factors, such as the lack of a comprehensive investigation of the fatigue failure phenomena, the lack of a well-defined fatigue damage theory used for fatigue damage prediction, and the inhomogeneity of composites because of their multiple internal borders. This study investigates the fatigue behavior of carbon fiber reinforced with epoxy (CFRE) laminated composite plates under spectrum loading utilizing a unique Deep Learning Network consisting of a convolutional neural network (CNN). The method includes establishing Finite Element Model (FEM) in a plate model under a spectrum fatigue loading. Then, a CNN is trained for fatigue behavior prediction. The training phase produces promising results, showing the model’s performance with 94.21% accuracy, 92.63% regression, and 91.55% F-score. To evaluate the model’s reliability, a comparison is made between fatigue data from the CNN and the FEM. It was found that the error band for this comparison is less than 0.3878 MPa, affirming the accuracy and reliability of the proposed technique. The proposed method results converge with available experimental results in the literature, thus, the study suggests the broad applicability of this method to other different composite structures.