Assessment of full-field J-integrals through digital image correlation using stress estimation based on measured strains

Application of digital image correlation (DIC) for full-field J-integral measurement with single-edge bending type SE(B) specimens, varying crack length-to-width ratios (a0/W), and a ductile material was studied numerically and experimentally. The specimens were first modelled in FEM and the displacement data was treated as measurement data to validate the method. Experimental studies were then conducted to extract full-field J-integrals by DIC. The study shows comparable or slightly higher J-values obtained with the surface DIC measurement as with the ASTM E1820 standard. However, there are certain considerations regarding convergence of the full-field J-integrals when large scale yielding and large rotations are involved.
The small strain theory behind J-integral imposes a limitation on the full-field method once the rotations are large enough to impose an error in the measured gradients. The rotations cause an inherent negative error in the measured gradients, which in turn imposes a negative error in the measured J. This is especially evident when the contour or domain for J-integral is taken far-field and contains areas that exhibit close to zero strains. A simple remedy is to determine the J-integral relatively close to the crack tip avoiding large rotations, while still also avoiding the crack tip to avoid strain localizations.