Investigation of the influence of the loading arrangement on the measured mode-I fracture resistance

This paper investigates the influence of the loading arrangement of double cantilever beam (DCB) test on the computed fracture resistance parameters of a structural adhesive. Aluminium joints were tested on three conventional loading arrangements from the BS ISO 25217:2009. We have found that the compliance of different loading arrangements can lead to significant differences in the fracture resistance computed using the cross-head displacement of the tensile-testing machine, even after considering the take-up of play and machine-compliance effects. In contrast, using the digital image correlation to measure the load-line displacement results in fracture resistance that is essentially independent of the loading arrangement. This finding was confirmed using three different data-reduction schemes, which yielded almost identical values of fracture resistance (less than 1% difference). Furthermore, direct identification of the traction-separation law at the crack tip was performed using Savitzky-Golay filter to compute the derivative of the J integral with regard to normal separation. Again, no significant influence of the loading arrangement was found since identified values of the separation at the peak traction and maximum separation are in excellent agreement. The identified maximum separation can be used as a threshold value of the relative deflection of the DCB arms to automatically determine the crack-tip position.