Interlaminar fracture behaviour of sandwich structures

The experimental determination of mode I interlaminar fracture toughness (G(IC)) values within the composite material core of sandwich structures, by using double cantilever beam (DCB) specimens, is strongly affected by-amongst other factors-the strain rate applied to the specimen. On the other hand, phenomena such as fibre bridging and crack-tip splitting are closely related to the applied strain rate, so that experimentally determined G(IC) values are expected to be rate-dependent. Another difficulty for G(IC) evaluation is the unknown difference between the real and measured crack length. Thus, although a number of different data reduction methods have been developed in the literature for the determination of G(IC) values, a great discrepancy between predicted G(IC) values has emerged. In the present paper, sandwich structures consisting of two aluminium skins bonded together with an epoxy-based adhesive reinforced with woven E-glass fibres were tested in the form of DCB specimens for the determination of their G(IC) values. Due to the non-linear and inelastic behaviour of the system, both the classical simple beam theory as well as the usual 5% compliance approach failed to predict G(IC). Thus, a new approach based mainly on experimental observation is suggested here for the determination of G(IC) The results are in good agreement with those derived by applying a model developed by Bishop and Drucker, further developed by Devitt ed al. and recently completed and presented in its final form by Williams and Freeman. Moreover, a strong effect of the applied strain rate on the G(IC) values was observed.