Image-based modeling of moisture-induced swelling and stress in 2D textile composite materials using a global-local approach

The present paper focuses on the employment of a global-local approach for the simulation of moisture-induced swelling and stress in a 2D textile carbon fiber/epoxy matrix composite material. The approach is based on the development of a robust image-based finite element model built starting from X-ray microtomography scans of a repetitive unit cell of the textile material: simulations aim at investigating both the global (sample scale) level and the local micromechanical (interfaces between matrix and fibrous tows, matrix inside a tow) level behaviour. Three models at the different cited scales are first built (ply level model, tow level and fiber level model): then a non-concurrent (node-based) approach is used for coupling the models at different scales. The results of numerical simulation of moisture-induced swelling and stress are discussed, with emphasis on the identification of the most critical zones at which maximal stress values are calculated, at different scales.