Investigation of the applicability of digital image correlation method for measuring the deformation of construction materials under load

Engineered Cementitious Composites (ECC) are construction materials exhibiting multiple cracking and pseudo strain hardening behavior and therefore, enhanced ductility when compared to conventional fiber reinforced composites. Digital image correlation method (DIC) is a recently developed non-contact measurement method based on math and image processing. Various parameters such as speckle density, illumination condition and analysis parameters (subset size, step size etc.) affect the measurement calculations. In this thesis, the effect of DIC parameters on the tensile strain measurements of ECC was investigated. For this purpose, ECC specimens were prepared by using a cement based matrix and 2 percent of high tenacity polypropylene (HTPP) fiber. Tensile tests were applied to specimens by taking the speckle density (coarse/fine), the effect of illumination (controlled/uncontrolled), and the resolution (high/low) parameters into account. Strain values of specimens were calculated both mechanically and digitally by using 4 extensometers placed at each corner of specimen and by DIC analysis, respectively. In DIC calculations, the effect of subset and step size combination to the strain measurement was also investigated. Crack properties of specimens providing the lowest strain error rate were also investigated both visually and digitally. The specimens exhibited average tensile strength of 2.75 MPa and average strain of 3.38 percen. In different step size of DIC; (0.5*subset)*(1/4), (0.5*subset)*(2/4), (0.5*subset)*(3/4) and (0.5*subset)*(4/4) the average strain values were calculated as 3.74, 3.67, 3.62 and 3.54 percent, respectively. Test conditions having the least error rate according to the methods specified were determined as; fine speckle, controlled illumination and high resolution. In these specimens, cracks that cannot be observed by eye counting method successfully detected by DIC method. In conclusion; this study demonstrated the superiority of DIC over traditional deformation measurement methods when the correct analysis parameters are selected.