High-fidelity architecture modeling and compressive strength prediction of 3D woven composite material

The compressive strength of 3D woven composites (3DWC) is significantly lower than the tensile strength, because of the misalignment of the weaving yarn. Therefore, high-fidelity weaving architecture of 3DWC is crucial for the prediction of the compressive strength. In this paper, an optimal modeling strategy is proposed to obtain high-fidelity architecture of the 3DWC using a digital element approach and ABAQUS/Explicit. The virtual architecture obtained from the present method replicates the actual textures. Subsequently, a meso-scale finite element model based on the high-fidelity architecture is created to predict the compressive strength and damage process. To remedy the mesh dependence, the corrected crack band model for logarithmic strain is used to model the softening behaviors of the yarn and matrix. The predicted compressive strengths for 2.4 mm, 3.4 mm, and 5.4 mm thick 3D woven composites in the warp and weft directions are in good agreement with the test results.