A highly efficient self-consistent clustering analysis method with field refinement capability for the mesoscale damage behavior of 3D woven composites

• Published in: Composites science and technology 2024-06, Vol.252, p.110609, Article 110609
• Main Authors: Wu, Siyang, Guo, Licheng, Li, Zhixing, Zheng, Tao, Huang, Jinzhao, Han, Xiaojian, Jia, Fenghao, Man, Shihan
• Format: Article
• Language: English
• Subjects: A. Polymer-matrix composites (PMCs); B. Mechanical properties; C. Computational mechanics; C. Damage mechanics
• ISSN: 0266-3538; 1879-1050
• DOI: 10.1016/j.compscitech.2024.110609

To effectively balance the accuracy and efficiency in solving high-dimensional damage problems, a self-consistent clustering analysis framework with field refinement capability (RESCA) incorporating a mesoscale damage model, is developed to investigate the mesoscale failure behavior of 3D woven composites (3DWCs). The RESCA method includes three stages: offline stage, online stage and field refinement stage integrating damage information. In the third stage, a damage-related field refinement framework is proposed to achieve cluster-based field dehomogenization and efficiently reconstruct the voxel-based field information. The results indicate that the RESCA method can accurately predict the local stress concentration, the voxel-based damage field distribution and the damage accumulation process, which are not available with the traditional SCA method. Importantly, the RESCA method can improve the computational efficiency by 25∼55 times compared to the finite element analysis (FEA) method. The RESCA method has double advantages in the efficiency and accuracy for the damage analysis of 3DWCs. [Display omitted]