A fracture toughness identification method for the debonding test of DCB specimens accounting for three-dimensional effects

The fracture toughness of adhesive joints and laminated composites subjected to opening load is commonly evaluated using Double Cantilever Beam (DCB) specimens according to ISO or ASTM standards. The formulations of these standards are based on two-dimensional (2D) elastic beam models, although a three-dimensional (3D) effect, related to the curvature of the crack front in DCB specimens and affecting the measured fracture toughness, is present and reported in literature. In this paper, a novel test set up is proposed to evaluate the energy release rate by revisiting a recently published decohesion model that implicitly considers the presence of 3D effects, without measuring the crack propagation length. The elastic behaviour of adherends in terms of Young's modulus and only load and macroscopic angular displacements response are necessary. Using the finite element code Ansys APDL, several decohesion test simulations of 3D DCB specimen models were performed. The fracture surfaces were modelled using cohesive elements and the adherends as isotropic material. The results obtained from the simulations were used to evaluate the accuracy of the fracture toughness obtained both through the active standards and via the novel approach, by comparison with that imposed in the cohesive law. The results showed that the standardized data reduction schemes were affected by systematic errors up to 30%, whilst the proposed methodology gave a negligible error in the fracture toughness evaluation compared to that imposed in the cohesive law.