Digital image correlation (DIC) is an efficient nondestructive technique for measuring surface displacement in engineering. However, standard DIC is restricted to continuous deformation, and the existing discontinuous DIC (DDIC) techniques are only able to measure small-scale cracks. In this report, a novel subset restore model and a corresponding subset size adaptation algorithm are presented to overcome this limitation for crack-state and displacement field reconstruction for large-scale cracks. The technique introduces a new subset restore method for splicing the segmented subset by tracing the motion trajectory caused by pure discontinuities. The proposed model facilitates the calculation of the rotation angle and the pivot of the subset movement. The subset size adaptation algorithm is designed based on an evaluation of the intensity gradient and correlation coefficient to allow the model to achieve high accuracy. Validation of the approach was performed using two typical crack models, by deforming a numerically synthesized Gaussian speckle image according to the deformation data from finite element analysis (FEA) results and photographing a laboratory tensile test with a high-speed CCD camera, respectively. The results validate the efficacy and high accuracy of the proposed approach compared to standard DIC in the reconstruction of the displacement fields in both continuous and discontinuous regions. The accuracy of resultant displacement reconstruction achieves approximately 0.015 pixels and 0.05 pixels in continuous region and crack vicinity, respectively.
High Accurate Crack Reconstruction Based on an Improved Discontinuous Digital Image Correlation: Subset Restore and Adaptation Method
