Finite Element Methods (FEMs) are widely employed for damage simulation in composites, with the Cohesive Zone Model (CZM) being particularly favored for its robustness and scalability. Despite the CZM's mesh-dependent nature, the Discontinuous Galerkin Cohesive Zone Model (DG/CZM) formulation allows for relatively arbitrary crack propagation thanks to highly refined meshes and its strong suitability for parallel implementation, which is crucial for fracture simulation in composites.

However, a challenge with the CZM in modeling damage in composites lies in determining which damage mechanism occurs at a given numerical interface. To date, fracture simulation in composite laminates has mainly relied on methods designed for isotropic materials, which come with notable modeling constraints such as limiting the arbitrariness of damage evolution and requiring experimental data on fracture propagation.

In this thesis, a novel approach is proposed to ensure arbitrary damage development solely based on the structural stress state using the DG/CZM approach.