Identification of Shear Sample Test Geometry for Bulk Adhesive Characterization

Abstract

High-fidelity finite element models of adhesively bonded structures require bulk adhesive mechanical properties describing stress-strain response, dependence on deformation rate, and ultimate strength of the adhesive material. In particular, the shear properties of the adhesive are necessary to accurately represent Mode II and mixed-mode loading response, but are not directly available from typical characterization tests such as the lap shear test. To address this challenge, five proposed shear specimen geometries from the literature were machined from flat plaques of cured two-part toughened structural epoxy and assessed by comparing the stress-strain response for quasi-static loading. The stress was calculated from the measured force and ligament area while strain was measured using Digital Image Correlation. There was generally good correspondence between the stress-strain responses for all specimens, although the strain at failure varied between the different sample geometries. In all cases, failure initiated at locations of high hydrostatic stress, as noted by others when testing polymers under shear loading or areas of high stress concentration. The sample with the smallest gauge length provided the highest strain measurement prior to failure and thus the most complete stress-strain response. However, in view of the failure initiation outside of the ligament, the measured shear strains at failure should be treated as a lower bound. Additionally, the small gauge length is conducive to future high deformation rate testing, since lower test velocities would be needed for high loading rates to be applied to the sample.