Abstract:
This thesis investigates the comparison between the behavior of a structure modeled
in a finite element (FE) software to the behavior captured using Digital Image Correlation
(DIC). This experiment opens the gateway to quicker and easier structural assessments of
the nation’s infrastructure through the use of DIC, a research area that requires further
investigation. Through experimental analysis, the ability of DIC to be used as a finite
element calibration tool was tested. Comparisons were completed by analyzing and
comparing the behavior of a fixed-fixed aluminum frame under varying static loads at the
midspan of the beam. Two structural parameters were assessed in the comparative analysis
including the structures response strain and deflection. Through the analysis of the
deflection, an adequate comparison was concluded between DIC and the FE model, with
average absolute differences in deflection between 0.004866 and 0.007075 inches.
However, the structures strain response consisted of high error in the DIC strains resulting
in unsuccessful comparisons between DIC and the FE model. These errors are predicted to
have been caused by vibration in the system during loading, noise caused by high light
reflectivity, construction error, and primarily, speckle pattern inadequacy. Since the
deflection results were deemed feasible, Euler-Bernoulli theory was used to derive strain
equations from DIC deflections, resulting in viable strain results for the structure. Overall,
DIC’s ability to be used as a finite element model updating tool was confirmed yet requires
further investigation into optimization methods for model parameter updating. Through
this research, a better understanding of the limitations of the DIC was investigated,
providing recommendations on the method of initial calibration between DIC and FE
software.
