Modeling of 3D Woven Composites with Realistic Geometry for Accurate Prediction of Kinking under Compressive Loads

The accuracy of modeling dictates the reliability of predictive analysis for 3D woven composites (3DWC). Past research is based on analysis of ideal geometry with specified nonlinearity. Co-relating the actual non-linearity arising from imperfection in the specimen with that of the model becomes an iterative process. In this study, digital element approach is implemented for creating the composite geometry. This technique simulates the individual fibers and the interactions between them, thus allowing the user to create a realistic RUC with random nonlinear geometry induced during manufacturing stage of 3DWCs. Thus response and strength analysis captures the unique weaving signature and provides better predictive models without the necessity to run iterative analysis. This approach can be extended to predict tensile and compressive failure of 3DWC with more accuracy without having to seed imperfections for failure initiation.