Precise Strain Measurement using Digital Image Correlation

Abstract

Digital Image Correlation (DIC) is a reliable and inexpensive newlinetool for in-plane full field displacement and strain measurement on the newlinesurface of a structure. In this thesis, the technology needed for precise newlinestrain measurement based on DIC is developed. The low-cost DIC setup newlineconsists of software module on which the DIC algorithm runs and a highresolution newlinewebcam to capture images. Mainly, Subpixel displacement is newlinemeasured in two steps: first, the displacement to the nearest integer pixel newlineis measured using the DIC technique and second, the displacement is newlineredefined to the nearest pixel using a subpixel DIC algorithm. Full-field newlinedisplacements are measured with sub pixel resolution using a sub pixel newlineDIC algorithm. This thesis presents a careful study of three commonly newlineused sub pixel DIC algorithms to determine which is most appropriate for newlinestrain measurement. The three most commonly used subpixel DIC newlinealgorithms are implemented in MATLAB for careful comparison. These newlinemethods are: Iterative spatial gradient method, Modified surface fitting newlinemethod and Ingrained fine search algorithm. These three algorithms are newlineverified and compared in terms of accuracy and computation time, using newlinesimulated images corresponding to various kinds of deformations. First newlinemodule focuses on measuring the strain based on displacement using newlineiterative spatial gradient algorithm and analyze the error for different newlineorder polynomial based on the displacement also calculate the searching newlinetime. Second module focuses on measuring the actual displacement and newlinecalculating the error based on the displacement using Modified Surface newlineFitting Algorithm (M-SFA). To obtain high-precision strain using initial newlinevalue of displacement and calculating the coefficients of polynomial newlineix newlinefitting function using M-SFA. Final module focuses on measuring the newlinestrain value with the actual displacement value with reduced number of newlineiterations. To measure the accuracy of the strain with reduced complexity, newlineingrained fine search algorithm is implemented. Computation time is newline