Stereocorrelation Formalism Considering Brightness and Contrast Effects: Application to Torsional Loadings

Abstract

Many multiaxial experimental setups rely on stereocorrelation (SC) techniques to resolve surface strains and deformations. Torsional loadings can affect the quality and accuracy of SC results due to changes in lighting conditions and the non-conservation of gray levels with the angle of twist. In this note, the gray level correction method implemented for 2D DIC cases is introduced in a 3D surface NURBS SC formalism and applied for torsional loadings. In particular, tension-torsion loading of AISI 1144 steel rods is used to demonstrate the approach. Different surface parameterizations, with and without gray level corrections, are tested in order to enhance the SC quality (defined via gray level residuals). It is found that both gray level corrections and relaxing the regularization of an overly constrained NURBS surface definition (6 vs. 36 knots), lower SC residuals.

Ambient multiaxial creep: shakedown analysis of structures using stereo digital image correlation

“…The residual map (which corresponds to the difference between the left and right images after determination of their respective projection matrices, Fig. 4f,g) shows a good calibration stage for the stereo DIC system [51]. In order to verify the quality of the DIC calibration, rigid body motions were applied using the MTS cells in displacement and angle control.…”

“…C-Mount Schneider-Kreuznach 2.8/50 lenses were used on both cameras to reach a resolution of 25 μm/pixel . Best practices reported in [51] to overcome gray level variations due to the change in lighting conditions upon torque application were applied. In particular, gray level corrections, with a set of polynomial functions up to order 2 were used to alleviate the non-conservation of the gray levels due to the angle of twist.…”

“…First a 0.080 mm axial displacement was applied while the DIC measurement detected a translation of 0.083 mm. A rotation of 1.0 degree was applied and a measurement of 0.94 degree was obtained [51]. This gives an indication of measurement uncertainty, approximately 4% in axial displacement and 6% in rotation.…”

Photometric DIC: a unified framework for global Stereo Digital Image Correlation based on the construction of textured digital twins

“…The same idea was introduced earlier by Faugeras and Keriven (1998) in CV. Because a correction on the whole ROI was not always sufficient to explain higher residual values in some areas, local corrections were introduced in SDIC on a finite-element basis (Colantonio et al 2020) for instance, or thanks to low-order polynomials (Dufour et al 2015;Charbal et al 2020).…”

On the benefits of correcting brightness and contrast in global digital image correlation: Monitoring cracks during curing and drying of a refractory castable

“…Tests performed at temperatures up to 1860°C, for which gray level inversions occurred, can also be monitored via global DIC [28]. Such procedures were extended to global stereocorrelation to account for illumination variations [34], in hybrid multiview correlation to register images of different modalities [35]. In digital volume correlation, the determination of metric and topological differences on 3D woven composite samples also required gray level corrections [36].…”