Enabling Digital Image Correlation with High-Resolution Microscopic Optics via Working Distance Automation: Advancing Resolution and Accuracy Limits

Abstract

AbstractOptical microscopy (OM) implementation of the digital image correlation (DIC) technique stands out with high practicality (e.g., no vacuum requirement and high amenability to be combined with other measurement channels). Despite its limited intragrain resolution compared to scanning electron microscopy (SEM) variants, OM-DIC provides critical identification at the interaction length scale of polycrystalline aggregates. The OM-DIC variant that is considered here (Shafaghi N, Kapan E, Aydıner CC, Exp Mech 60:735–751, 2020; Özdür NA, Üçel IB, Yang J, Aydıner CC, Exp Mech 61:499, 2020), however, further attempts to minimize the comparative intragrain resolution deficiency of OM-DIC by utilizing high-resolution [high numerical aperture (NA)] objectives. Images with high-NA objectives will typically immediately suffer defocusing for a deforming sample, given the extremely limited depth of fields (in the μm order). The technique employs continual automated working distance (WD) adjustment to fight off defocusing through custom instrumentation that also implements area scanning to expand field coverage to the mm scale. The precise WD adjustments also help to minimize WD error (a biaxial strain error in DIC measurements). While the technique has been formerly used to study highly strained polycrystalline fields (Shafaghi N, Kapan E, Aydıner CC, Exp Mech 60:735–751, 2020; Özdür NA, Üçel IB, Yang J, Aydıner CC, Exp Mech 61:499, 2020), the purpose of this study is to investigate and advance its accuracy limits. For this purpose, 40× microscopy with a high-NA objective (Özdür NA, Üçel IB, Yang J, Aydıner CC, Exp Mech 61:499, 2020) is utilized over a pure FCC nickel polycrystal with enlarged grains (average 70 μm), yielding about 5000 DIC grid points per grain. Regardless of the length scale, however, DIC offers limited sensitivity for pointwise strains (about 0.1%). While this is deemed sufficient for plasticity, elastic strains are not locally resolved. Here, we will employ small load increments in the initial elastic ramp of nickel to test the sensitivity limits of the method in a grain-resolved setting, i.e., the strain fields of individual grains are separately considered. The crystallographic orientations will be known thanks to pre-experiment electron backscatter diffraction. The trade-off between accuracy and resolution will be tested by local averages inside the grains to see whether regional DIC accuracy can be pushed toward elastic strain levels. Results of multiple grains will be compared and consistency with finite element predictions that account for the crystallite orientations will be presented.KeywordsDICPolycrystalMicroscopyHigh resolution