A comparative study of techniques of distant reconstruction of displacement and strain fields using the DISTRESS simulator

Abstract

Reconstruction and monitoring of displacement and strain fields is an important problem in engineering. We analyze the remote and non-obtrusive method of Digital Image Correlation (DIC) in 2D based on photogrammetry. The method involves covering the photographed surface with a pattern of speckles and comparing the images taken before and after the deformation. The analysis is based on a specially developed Digital Image Synthesizer To Reconstruct Strain in Solids (DISTRESS) Simulator to generate synthetic images of displacement and stress fields in two dimensions in order to investigate the intrinsic accuracy of the existing variants of DIC. We investigated the Basic DIC and a commercial software VIC 2d, both based on displacement field reconstruction with post processing strain determination based on numerical differentiation. We also investigated what we call the Extended DIC where the strain field is determined independently of the displacement field. While the Basic DIC is faster, the Extended DIC delivers the best accuracy. The speckle pattern is found to be playing a critical role in achieving high accuracy for DIC. Increase in the subset size for DIC does not significantly improves the accuracy, while the smallest subset size depends on the speckle pattern and speckle size. Increase in the overall image size provides more details but does not play significant role in improving the accuracy, while significantly increasing the computation cost. We observed that it is not reliable to measure very small strains using grayscale images in DIC. Thus, we propose Color DIC using color images and found that it improves the accuracy in measuring small strains.