Micromechanical computational analysis of debonding and kinking in microscale composites

Composites are widely utilized across various industries due to their exceptional properties, allowing design flexibility and complexity. There are different scales of composites numerical study, and this study aims to understand the damage mechanisms observed in microscale composites undergoing transverse compressive load. This understanding is achieved through prediction via simulation and micromechanical computational analyses. Based on previously conducted experimental studies, Zumaquero et. al. studied different stages of damage progression through microscopical inspection of the tested coupons[1]. The compressive failure behaviour of composites in the transverse direction was observed, revealing the preferential debonding angle is between 70 to 80°. Subsequently, the growth of the interface debonding failure becomes stable, and the kinking angle towards the matrix was found to be between 50 to 60°, consistent with the numerical predictions by Correa et al. using the Boundary Element Method (BEM) [2]. These findings motivate the current study, where a UD RVE model with a periodic boundary condition (PBC) is developed using the random sequential algorithm (RSA) and the angles of debonding and kinking failure are observed. To predict the onset of matrix cracking and the crack propagation, the extended FEM modelling approach [3] is used and the LaRC05 failure criterion is implemented through a compiled UMAT. The Drucker-Prager model served as the constitutive law for matrix yielding behaviour, while cohesive elements were assigned to predict the debonding failure of the fibre-matrix interface. In addition, the Phase-Field Fracture (PF) method is also used to predict matrix cracking behaviour and the results from both LaRC05 and PF are compared to investigate the efficiencies of both methods. The study concludes that the initial direction of failure predicted agrees with that microscopically observed in experiments. This research aims to contribute to the development of computational tools leading eventually to more resilient and precisely engineered composites for diverse applications.