Wednesday  |  Conference Center B  |  04:20 PM–04:40 PM

#13345, Path-Integrated X-ray Images for Multi-Surface Digital Image Correlation

Digital Image Correlation (DIC) based on X-ray (opposed to optical) images provides a pathway to characterize heterogeneous deformation of specimens that are visibly occluded. However, because X-ray images are path-integrated, conservation of intensity and the principle of optical flow can be violated if any component other than the patterned surface of interest moves during a test. This work presents a novel DIC algorithm that modifies the root equation for conservation of intensity to separate motion of two independent surfaces from a single series of path-integrated images of a deforming specimen. Specifically, the cost function for the correlation is rewritten. The deformed, path-integrated image intensity becomes the multiplicative sum of two undeformed reference images of interest, independently warped with an affine shape function.

We first present the algorithm and new image processing steps, and then demonstrate the ability of the algorithm to separate independent motions using synthetic images undergoing rigid translation, rigid rotation, and uniform strains. Subpixel (<0.05 px in most cases) measurement resolution and accuracy are obtained for noise levels commiserate with experimental images. We further demonstrate the efficacy of the algorithm using real, experimental images to separate two, independent rigid translations of an aluminum plate patterned with tantalum, again with subpixel accuracy and precision. Finally, we investigate using either experimental or synthetic reference images of the two undeformed patterns (in conjunction with the experimental, path-integrated image series of the deforming specimen), and discuss challenges associated with manufacturing defects in the patterns.

This work provides the foundation for using X-ray DIC for complicated test environments, where optical DIC is not feasible. While X-ray DIC has been used in the past for a single moving/deforming surface, the innovative algorithm presented here extends the applicability of X-ray DIC to more challenging conditions where multiple surfaces/components are moving independently.

Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security administration under contrast DE-NA0003525.

Elizabeth Jones   Sandia National Laboratories
Samuel Fayad   Sandia National Laboratories
Dayna Obenauf   Sandia National Laboratories
Benjamin Halls   Sandia National Laboratories
Caroline Winters   Sandia National Laboratories