Three-dimensional semi-analytical investigation of interlaminar stresses in composite laminates

Delamination is the most common failure mechanism in both monolithic and hybrid metal-composite laminates. Understanding its evolution is crucial to predict the failure behaviour of these classes of materials. Analytical and experimental investigations of how such laminates respond in the vicinity of a free edge, served as the foundation for the study of delamination phenomena in structural composite laminates. High interlaminar stress gradients arise near these free edges due to material discontinuities. These high stresses may eventually result in premature failure. Classical laminate theory (CLT) is not adequate to predict such failures due to its two-dimensional nature and assessments of the out-of-plane stress distributions are not possible. Consequently, much attention has been focused on the characterisation of the composite laminates’ interfaces. Several approaches have been proposed to calculate this free-edge stress field. Some of them make use of analytical techniques, while others make use of numerical methods. Although much research has been conducted in this area, a more general approach that could be used at dissimilar media interfaces is still required. In this study, we make use of the finite element method to compute the full stress tensor near the free edges which is then utilised in the development of a new three-dimensional semi-analytical method to calculate interlaminar stresses at the interfaces for any given material system and geometry. An expression is developed to study the dependence of the interfacial stresses on elastic and geometrical parameters. Symmetric cross-ply, and angle-ply laminates subjected to uniaxial loading are used as test cases to demonstrate the accuracy of the developed approach. A bi-material system is also considered, and the non-dimensional stress function values are obtained. A thin resin rich transition layer is introduced at the interface where failure assessment is required. Results are compared to the predictions of other analyses found in the literature. The proposed method is found to be simpler and efficient.