Interface fracture in adhesively bonded shell structures

Two methods for the prediction of crack propagation through the interface of adhesively bonded shells are discussed. One is based on a fracture mechanics approach; the other is based on a cohesive zone approach. Attention is focussed on predicting the shape of the crack front and the critical stress required to propagate the crack under quasi-static conditions. The fracture mechanical model is theoretically sound and it is accurate and numerically stable. The cohesive zone model has some advantages over the fracture mechanics based model. It is easier to generalise the cohesive zone model to take into account effects such as plastic deformation in the adhering shells, and to take into account effects of large local curvatures of the interface crack front. The comparison shows a convergence of the results based on the cohesive zone model towards the results based on a fracture mechanics approach in the limit where the size of the cohesive zone becomes smaller than other relevant geometrical lengths for the problem. However, convergence issues and numerical stability must be addressed.