Data-driven stress and strain curves of the unidirectional composites by deep neural networks with principal component analysis and selective-data augmentation. (1st June 2023)

• Record Type: Journal Article
• Authors: Kim, Do-Won; Go, Myeong-Seok; Lim, Jae Hyuk; Lee, Seungchul
• Is Part Of: Composite structures. Volume 313(2023)
• Publication Date: 2023-06-01
• Subjects: Unidirectional composites -- Deep neural network -- Stress and strain curve -- Ultimate tensile strength -- Toughness -- Selective-data augmentation; Composite construction -- Periodicals; Composites -- Périodiques; 624.18
• DOI: 10.1016/j.compstruct.2023.116902
• Languages: English
• ISSNs: 0263-8223

Abstract

Highlights: A data-driven prediction of the stress–strain (S-S) curves of unidirectional composites was conducted with DNN. PCA expresses the S-S curves in a reduced domain without losing data information but eliminating the undesired oscillation. The proposed DNN model accurately predicts the S-S curves for normal samples, but performs poorly on rare samples. A selective-data augmentation was proposed to improve the prediction accuracy for rare samples with lower toughness values. Abstract: This study proposes a data-driven approach using a deep neural network (DNN) to efficiently predict the stress–strain (S-S) curves for unidirectional composites. Firstly, a representative volume element (RVE), including arbitrary fiber distribution of circular fibers, was generated. Then, the S-S curves for each RVE were obtained from finite element simulation considering the interfacial debonding phenomenon. Next, input and output features were chosen in terms of the center positions of fibers and S-S curves, respectively, for DNN training. In addition, to learn the S-S curves efficiently in a lower-dimensional space, principal component analysis (PCA) was employed. Subsequently, the data-driven model combining PCA and DNN was developed; this quickly and accurately predicted the S-S curves with a relative error for the toughness of about 2 %. Furthermore, selective-data augmentation is proposed to improve the prediction accuracy in insufficient and nongeneralized datasets, leading to the higher prediction accuracy of ultimate tensile strength and toughness.