Development of a virtual testing methodology for FRP-timber hybrid structures

The civil engineering sector has seen growing interest in timber structures as a means to reduce the lifecycle environmental impact of low-rise and mid-rise buildings. In this context, glued-in rods (GiR) have emerged as efficient solutions for producing versatile timber connectors which meet strength and stiffness requirements with additional ease of assembly. Among the available rod materials for GiRs, fibre-reinforced polymer (FRP) composites offer significant advantages over metallic bars such as improved corrosion resistance, with the main disadvantage being their brittle failure behaviour in the absence of metal ductility. To improve the damage tolerance of such joints, the brittle behaviour of the FRP must be compensated by promoting pseudo-ductility in the surrounding wood and/or adhesive. This makes the design of resilient FRP GiR joints considerably more complex than a metallic equivalent. Current numerical models of GiRs do not include wood failure mechanisms and hence are unsuitable for use in a virtual testing approach. The aims of the present work are to (i) develop a continuum damage mechanics (CDM) model including cohesive zone modelling (CZM) capable of reproducing the various damage mechanisms in spruce wood, (ii) characterise and reproduce the internal meso-scale structure of the wood in representative timber elements, and (iii) verify and validate simulations against physical tests on FRP GiRs. To this end, individual growth rings in each manufactured specimen were characterised through image analysis and modelled as conformal layers of orthotropic properties. The different damage mechanisms for each layer were calibrated using standard test methods, namely tensile tests and block shear tests, as well as material data from the literature. The model was then validated against destructive FRP GiR experiments in the ‘pull-pull’ configuration. The consideration of individual growth rings enabled a significant degree of material variability to be reproduced. The simulations are shown to reproduce calibration tests accurately, and encouraging results are achieved in predicting the progressive failure of a standard GiR joint design.