Generation of unidirectional fiber random distribution structures based on an artificial fish swarm algorithm with random deletion after fiber filling

Abstract

This study proposes a random distribution method for generating high-content volume fractions of fibers in their cross-sectional area. This approach is referred to as the artificial fish swarm algorithm (AFSA) with random deletion after fiber filling (RDAFF_AFSA). Initial fibers were first generated using AFSA, followed by fiber filling the matrix-rich region with a hard-core model. The desired representative volume element (RVE) was ultimately obtained by random deletion. The proposed method can generate RVEs with high-content volume fractions (up to 67%) of fibers. The generated RVEs were statistically analyzed and compared with the completely spatial random (CSR) pattern and experiment data. The results showed that RDAFF_AFSA exhibited a high degree of consistency with the CSR and experimental data. The elastic constants of the carbon fiber-reinforced plastic composites were predicted by finite element analysis. The predicted results are very reasonable compared to the experimental results. The proposed method can provide a highly valuable alternative for micromechanical and multiscale analyses of unidirectional fiber-reinforced composites.