| Summary: | The determination of three components of displacements at material surfaces is possible using surface topography information of undeformed (reference) and deformed states. The height digital image correlation (hDIC) technique was developed and demonstrated to achieve micro-level in-plane resolution and nanoscale out-of-plane precision. However, in the original formulation hDIC and other topography-based correlation techniques perform well in the determination of continuous displacements. In the present study of material deformation up to cracking and filan failure, the ability to identify discontinuous triaxial displacements at emerging discontinuities is important. For this purpose, a new method reported herein was developed based on the hDIC technique. The hDIC solution procedure comprises two stages, namely, integer-pixel level correlation and sub-pixel level correlation. In order to predict the displacement and height changes in discontinuous regions, a smoothing stage was inserted between the two main stages. The proposed method determines accurately the discontinuous edges, and the out-of-plane displacements become sharply resolved without any further intervention in the algorithm function. High computational demand required to determine discontinuous displacements using high density topography data was tackled by employing the graphics processing unit (GPU) parallel computing capability with the paging approach. The hDIC technique with GPU parallel computing implementation was applied for the identification of discontinuous edges in an aluminium alloy dog bone test specimen subjected to tensile testing up to failure.
|
|---|