Multi-scale experimental and computational investigation of matrix cracking evolution in carbon fiber-reinforced composites in the absence and presence of voids

Manufacturing defects such as voids can influence matrix cracking, which is one of the main forms of damage in fiber-reinforced laminates. Characterization and investigation of matrix cracks is not straightforward, neither experimentally nor computationally. In the current study, we develop a methodology for in-situ analysis of matrix cracking evolution in the absence and presence of voids. The cracks are detected using digital image correlation at three different scales: macro, meso, and micro. Additionally, a combined two-scale computational approach is established for prediction of the effect of voids on matrix cracking. Local properties of the material are calculated using computational micromechanics and then serve as input for the meso-scale model based on extended finite element method. The computational approach predicts the crack density evolution in function of the applied deformation. The results of both experimental and computational studies reveal that voids initiate matrix cracks, causing earlier damage in off-axis plies. They also result in higher crack density, for a given applied strain.