Fracture mechanics of plate debonding: experimental validation

Premature plate debonding hampers the efficient use of externally bonded FRP plates for flexural strengthening of concrete beams. Existing research mostly concentrates on finite element (FE) modelling of the concrete–FRP interface but such analyses are of dubious validity because they require far more details than will ever be available for the interface. A fracture-mechanics-based plate debonding model has been developed by the authors; since detailed stress analysis of concrete is unattainable the model is based on the global energy balance of the system. Flaws will inevitability be present in the vicinity of the interface; the model investigates the energy balance when such a flaw propagates. The energy released when the crack extends (GR) is compared with the interface fracture energy required to create the new surfaces GF: If GR > GF the crack will extend causing debonding.Determination of both GR and GF associated with crack extension is not trivial because of the unknowable microstructure of concrete. The early work of the present study developed methods to find both parameters to accuracies sufficient for practical purposes. A modified version of Branson’s model, which takes account of the effects caused by the axial force in the FRP, has been developed for the moment–curvature and subsequent GR analyses, while GF has been determined according to the actual fracture mechanism that takes place in the interface.This paper presents comparisons with a variety of plate debonding test data (including steel plate bonded beams) reported in the literature and shows that the present model can correctly determine both the failure load and the debonding mode. Only simply-supported beams, without additional plate end anchors, under short-term monotonic loads are considered here, but the model could be extendedto analyse more complex practical problems.