Specific fracture energy at glue joints in European beech wood

Abstract The specific fracture energy G at glue joints in hardwood was investigated by using a novel approach of testing and data evaluation. European beech wood was bonded with a specific one-component polyurethane (PUR) adhesive and a specific phenol resorcinol formaldehyde resin (PRF) to produce double cantilever beam (DCB) specimens. These specimens were stored at different relative humidities (50%, 65% and 95%) prior to testing. Additionally, DCB specimens manufactured out of differently aged glulam elements were also evaluated. Special care was given to a constant crack propagation rate during examination, and only the middle part of the specimens was evaluated to avoid influences of the discontinuities at crack initiation and final rupture. The PUR series were tested at three different crack propagation rates to check for a potential influence on G. The PRF specimens failed mostly in the wood, with an average wood failure of 90%. Thus no difference was notable between the PRF and solid wood samples. Their G was around 0.85 N/mm for the regular and dry climates and 1.35 N/mm for the wet climate. All aged series had a 10–15% reduced G, but with lack of significance. The PUR specimens failed in adhesion, independent from climate, ageing or testing speed, and had an average G of 0.24 N/mm. The novel approach, i.e. with constant crack propagation rate and evaluating the data only partially, worked well and can be recommended for further research.