Evaluation of camera motion in stereo-DIC

Abstract

The full-field nature of stereo-digital image correlation (stereo-DIC) makes it a widespread technique with vast possibilities. The uncertainty quantification of the technique is however not yet fully understood due to the non-linear optical-numerical measurement chain, limiting analytical research of the technique. Camera motion is a perfect example of this; it is hard to measure and the influence of it is not yet investigated. A simulator (presented in [Balcaen et al. Finite element based image generator for stereo-DIC uncertainty quantification applied on stereo-DIC calibration, 2016] and used in [Balcaen et al. Stereo-DIC uncertainty quantification based on simulated images, 2016]) is used to study how camera motion during a test influences the resulting deformation- and strain-field. Since the exact imposed deformation-field is known in a simulator the resulting errors can be clearly identified. Camera motion between the calibration-stage and the actual measurement is not investigated since this comes down to a poorly calibrated rig. We refer to the literature for more information concerning the influence of poor calibration data on measurement results. Numerous tests indicate that small camera-rotations can introduce a linear displacement-offset and thus also a constant strain-offset, mainly depending on the focal length of the lenses and the angle of rotation. Camera translation on the other side seems to be less pernicious.