Modelling mixed-mode crack growth in bio-adhesive wood bonds via cohesive laws

An essential aspect of developing novel bio-adhesives is ensuring sufficient bond strength. Currently, bond strength is typically assessed using stress-based criteria derived from lapshear tests, such as those specified in EN 302-1. However, this method provides limited insight into bond performance and cannot reliably predict behaviour in other joint configurations or structural components. A more comprehensive approach involves fracturemechanical characterisation of the wood-adhesive bond line through measurement of its cohesive law, commonly referred to as the traction–separation (σ–δ) law. This allows for detailed analysis of both strength and fracture energy release rates under various loading modes. The most relevant modes of loading for wood products are mode I (opening/tension), mode II (sliding/shear), and mixed-mode loading. These parameters can be implemented in Finite Element (FE) models using Cohesive Zone Modelling (CZM) to simulate crack growth, enabling strength predictions across a range of wood products, such as bed slats in furniture applications or glulam finger joints in structural applications. This modelling approach, requiring a mixed-mode I–II cohesive zone model, was successfully applied in the 1990s to conventional adhesives such as resorcinol–phenol and polyvinyl acetate, and it is now highly relevant for guiding the development of wood-based bio-adhesives. In parallel research, experimental cohesive laws have been derived from Double Cantilever Beam (DCB) specimens subjected to pure bending moments under varying mode I/II conditions. In the present study, these experimentally obtained parameters are implemented in a COMSOLbased FE model to simulate crack growth. The predicted crack propagation and fracture resistance are then compared with experimental results to validate the model.