A new method for estimating wood failure percentage in adhesive-bonded shear specimens

The traditional visual estimation of wood failure percent (WFP) in shear-failed adhesive-bonded joints is often marred by inconsistencies and produces results with large variations between examiners. This method of determining WFP has also been criticized as inaccurate and subjective. Therefore, an alternative method with improved accuracy and more consistent WFP estimations is desired. This study investigated the feasibility of Fiji-ImageJ software-assisted (F-IMJ-A) estimation of WFP in sheared block specimens of three wood species namely Eucalyptus grandis, Fagus sylvatica, and Pinus sylvestri bonded with three different adhesive systems viz. melamine-urea-formaldehyde, phenol-resorcinol-formaldehyde, and polyurethane. The consistency of the F-IMJ-A method was evaluated between three estimators and against the traditional ASTM D5266:2005 practice. The results obtained showed that the F-IMJ-A method improved, in most cases, the consistency of WFP estimations between the estimators. 5–33%, 1–36%, and 5–64% reduced WFP variations between estimators were recorded in E. grandis, F. sylvatica, and P. sylvestri-bonded joints respectively, across the different adhesive systems and wood failure modes. The F-IMJ-A procedures downscaled, by one-third, the 100% human-error-prone ASTM D5266 procedures as could be applicable with partially trained estimators. However, further improvement in the F-IMJ-A estimation is still desirable considering the range of variations between the estimators despite the recorded improvements. This study has demonstrated the suitability of the F-IMJ-A method in estimating WFP in shear-failed specimens of different wood species bonded with different adhesive systems.