Conference paper Open Access
Bond-line creates an interface between two glued surfaces and, therefore, its brings additional complexity into the mechanical behavior of glued component, especially around the bond-line region because an adhesive has a very different response to a mechanical stress than wood. From this perspective, the bond-line influences the total mechanical response of glued components by both its cohesive and adhesive behavior at wood-adhesive boundary. For timber constructions, there are many various adhesives one can use and each of them has a different mechanical characteristics, advantages and disadvantages.
The goal of this work was to create a numerical finite element models applicable for analysis of fracture problems in mode II. The models were developed for the adhesive that are often applied in timber constructions and wooden materials. The FE models include 2D geometry of the bond line and cohesive law fitted on the outputs of the experimental measurement. The experimental data for the developing numerical models were obtained using 3point end-notched flexure (3ENF) tests with the compliance-based beam method (CBBM) coupled with the digital image correlation to be able to obtain displacement slip needed for the development of the FE models. Furthermore, within the FE analysis, wood was modeled as orthotropic material including both elastic and plastic regions of deformation. The FE models were developed in Ansys computational system. The specific objectives of the work were following: 1) to create the cohesive zone models based on experimental data; 2) to develop parametric 2D and 3D model of the bond-line reflecting experimental data; 3) to analyze the influence of friction coefficient on resulting force-displacement outputs; and 4) to analyze plastic imprint into the specimens for Norway spruce and an influence of the fiber angle inclination.
Implementation of the cohesive law models of wood-adhesive system into the FE analysis was successful. The FE analysis provided the force-deflection response that was validated by the experiment work. The FE model showed that influence of the friction on the simulated force may be up to 5% of the maximal force which is not negligible effect. The imprint of the load head into specimen is substantial if span-to-height ratio is below 17. The influence of the fiber angle with respect to a longitudinal axis is rather high, i.e. angle of 14° means 30% reduction of maximal force.