Measuring discontinuous displacement fields in cracked specimens using digital image correlation with mesh adaptation and crack-path optimization

Abstract

This paper evaluates mesh adaptation techniques in two-dimensional digital image correlation (2D-DIC) analyses of specimens with large deformations and a single propagating crack. The “finite element” formulation of DIC is chosen as basis for this study. The focus has been on the challenges caused by high-gradient and discontinuous displacement fields in the region of a propagating crack, and the aim has been to improve both the robustness and the accuracy of the correlation in such regions. Mesh adaptation based on node splitting and a novel overlapping-mesh technique have been implemented in a DIC code and used to capture the discontinuous displacement fields of cracked specimens. In addition, a procedure for optimization of the location of the crack path is presented. Synthetic image series with known displacement fields, generated from finite element simulations, have been analyzed to evaluate the performance of the presented techniques. Additionally, an experimental image series of a modified Arcan test has been analyzed using the proposed mesh adaptation and crack-path optimization. The paper contains a detailed description of the proposed techniques and results from the evaluation. It is found that both the node-splitting and overlapping-mesh techniques can be applied to successfully capture the discontinuous displacement field of a propagating crack. In the latter technique, the crack path is described down to pixel level. The crack-path optimization is capable of locating the crack path with sub-pixel accuracy, reducing correlation residuals and thus increasing the robustness of the DIC analysis. In addition, a certain filtering of pixels based on high correlation residual is found to increase the robustness of the correlation in areas affected by a propagating crack.