Two-dimensional meshfree-based digital image correlation for strain-displacement measurement

Abstract

At present, there are various image registration algorithms under active development for digital image correlation techniques. These techniques include the meshfree-based (MFree-based) digital image correlation. However, there exists very little study in advancing the method for better computational efficiency in engineering measurement applications. Also, a comparative investigation against the more well-known finite element-based digital image correlation is absent. This paper presents the development of an efficient two-dimensional (2D) MFree-based digital image correlation technique to obtain the strain and displacement of a planar object. In the present method, the speckle pattern in the form of circular adhesive markers is first detected and tracked by employing the Circular-Hough-Transform and Kanade-Lucas-Tomasi algorithms. The element-free Galerkin approach that specifically employs the moving least-square approximation has been adopted as the meshfree method for the strain and displacement determination. A series of tensile tests on pure epoxy samples has then been carried out to verify the applicability of the proposed 2D MFree-based digital image correlation technique, from which a good agreement of the strain and displacement description is found when compared with that obtained by the conventional strain gauge method. Furthermore, the full-field deformations computed from the ABAQUS model have been well captured by the proposed technique. This concludes that the proposed 2D MFree-based digital image correlation technique is greatly applicable and is recommended for use in determining the planar material engineering properties.