A bond strength mechanical model for FRP NSM reinforcements: theoretical developments and their calibration for design guidance

The tension capacity of fibre reinforced polymer (FRP) bars or strips mounted in grooves realised on the surface of existing concrete, masonry or timber structural elements, usually indicated as near-surface mounted (NSM) reinforcement, has been examined in recent years by means of several experimental tests. These tests include both bond and bending tests, and a large amount of numerical and theoretical studies have also been developed, but no mechanically- based model for predicting the maximum tensile force in FRP NSM reinforcements is available in design guidelines nowadays. Since the experimental tests evidenced that, also for NSM systems, the detachment of the reinforcement (debonding failure) may cause the premature collapse of the strengthened element, the study of the bond behaviour at the substrate-NSM reinforcement interface is fundamental. The paper aims to provide a prediction of the maximum tensile force that FRP NSM systems can sustain in case of debonding failure, using an interface constitutive law that includes and generalises the bond strength model commonly used in the literature and in design guidelines for externally bonded FRP reinforcements. A simplified rigid-linear softening relationship followed by a constant frictional resistance branch is, in fact, assumed for the interface adhesion law of FRP NSM systems applied to both concrete and masonry elements. Using this interface bond model and solving the boundary values problem of an FRP reinforcement adhered to the support in which it is inserted, a very simple closed-form analytical prediction is derived for the calculation of the ‘effective length’ and of the corresponding debonding force. In particular, the effective length is divided into two terms, one related to the rigid-linear softening branch of the bond law and the other to the frictional one; therefore, also the debonding force is given by two terms depending on the two branches of the bond law. The reliability of the proposed formulation for the debonding force, in particular for the term related to the rigid-linear softening branch, was confirmed thanks to the comparisons with a wide set of experimental data of tests on NSM systems bonded to concrete and masonry specimens, available in the international literature and collected by the Authors in a database. Finally, statistical analyses were carried out to assess the main coefficients of the proposed formulation, which allow defining design values for the debonding force in FRP NSM systems according to the ‘Design assisted by testing’ scheme suggested in Annex D of Eurocode EN 1990. The proposed formulation can be used for end debonding verifications in both concrete and masonry structures. Thus, it has significant practical implications for the design of this type of NSM strengthening intervention.