Cohesive zone model and quasibrittle failure of wood: A new light on the adapted specimen geometries for fracture tests

•An accurate CZ modeling needs the experimental steady state propagation.•The steady state propagation requires a free development/propagation of the FPZ.•The boundary conditions can induce a FPZ confinement.•A change in the FPZ geometry modifies the resistance to crack growth.•Modification of the FPZ geometry requires a change in cohesive fracture energy.

The present study focuses on the cohesive zone modeling of quasibrittle failure of wood. The outline of a general estimation procedure of the cohesive zone parameters based on the one-to-one correspondence which exists between the R-curve and the softening function is presented for any kind of specimen geometry and for the bilinear approximation of the softening function. The connection between the R-curve characteristics and the cohesive parameters provides a new light on the effect of boundary conditions issued from specimen geometries used for wood fracture tests. More particularly, the steady state propagation of the main crack with its fracture process zone is required to estimate relevant fracture properties in wood. Due to the consequent size of the fracture process zone (FPZ) in wood, the steady state propagation requires specimen geometries exhibiting a sufficiently long ligament length to avoid confinement of FPZ during its development and its propagation. Finally, the variability of the cohesive parameters as well as their dependences with respect to the specimen geometry are discussed in the case of Norway Spruce.