MEASURING COHESIVE LAW OF WOOD ADHESIVE BONDS FOR ENGINEERING OF TIMBER JOINTS AND PRODUCTS

Accurate prediction of crack growth and strength in glued timber joints and components can be achieved by using cohesive laws in constitutive models. These laws represent the traction-separation relationship that measures the stress transmitted between crack faces in a cohesive zone, and can be obtained directly from experiments. However, studies on measuring the cohesive laws of wood adhesive bonds are scarce, not standardised, and not commonly used in practice. This work aims to implement the "J-integral approach" for determining the mode I cohesive law of wood adhesive bonds directly from experiments. The investigation used Double Cantilever Beam (DCB) specimens subjected to pure bending oments, with lignin-adhesive, PVAc-adhesive, and beech wood as the test materials. For lignin-adhesive bonded wood, fracture resistance values were notably low (20-60 J/m2), accompanied by small peak stress values (0.8 MPa) and end openings (0.03 mm). Fracture propagation exhibited some instability after crack initiation. The cohesive laws for PVAc-adhesive bonded wood were more consistent, with peak stress values ranging from 1-3 MPa, end openings between 0.1-0.3 mm, and fracture resistance reaching up to 1000 J/m2. PVAc specimens with release film side grooves proved to be most suitable for characterising the bond line fracture behavior of this ductile, strong adhesive. For beech wood, specimens with side grooves exhibited more stable fracture propagation and more reasonable values for peak stress (5 MPa) and fracture energy release rates (360 J/m2). In conclusion, the J-integral approach demonstrates its potential as a valuable method for determining cohesive laws in wood and wood adhesive bonds. However, further development is necessary to improve specimen geometry, preparation, and loading procedures to enhance the reliability of fracture characterisation for all three materials investigated.