New levels of high angular resolution EBSD performance via inverse compositional Gauss–Newton based digital image correlation

Abstract

Conventional high angular resolution electron backscatter diffraction (HREBSD) uses cross-correlation to track features between diffraction patterns, which are then related to the relative elastic strain and misorientation between the diffracting volumes of material. This paper adapts inverse compositional Gauss Newton (ICGN) digital image correlation (DIC) to be compatible with HREBSD. ICGN-based works by efficiently tracking not just the shift in features, but also the change in their shape. Modeling a shape change as well as a shift results in greater accuracy. This method, ICGN-based HREBSD, is applied to a simulated data set, and its performance is compared to conventional cross-correlation HREBSD, and cross-correlation HREBSD with remapping. ICGN-based HREBSD is shown to have about half the strain error of the best cross-correlation method with a comparable computation time.

Global DIC approach guided by a cross-correlation based initial guess for HR-EBSD and on-axis HR-TKD

“…The accuracy of the initial guess directly affects the convergence characteristics of both FA-GN and IC-GN algorithms [38]. As illustrated in the context of HR-EBSD by Vermeij and Hoefnagels [32] and Ruggles et al [33] by using dynamically simulated patterns, the convergence speed decreases with increasing disorientation between the initial guess and the true solution. Divergence is likely to occur for differences over » 1°, which is of the order of the Hough-transform based indexation uncertainty.…”

“…As proven by Hardin et al [31], the stress normal to the free surface remains negligible in the absence of stress field sources located near the surface or of large error on the specimen tilt. During this last year, three independent works by Vermeij & Hoefenagels [32], Ruggles et al [33] and Shi et al [34] proposed a 'global' HR-EBSD approach. This method is an 'integrated' Digital Image Correlation (I-DIC) approach [32,34] that correlates EDP signal intensities by using a unique and large ROI.…”

“…The relative deformation of the target ROI with respect to the reference ROI is accounted for by eight degrees of freedom, which are the components of b F e . The deformation gradient is computed iteratively by means of a forward-additive (FA) [32,34] or an inverse-compositional (IC) [33] Gauss-Newton (GN) algorithm in order to solve the non-linear DIC problem, i.e. the minimization of the quadratic difference of intensities between the reference and the warped target ROI.…”

“…the minimization of the quadratic difference of intensities between the reference and the warped target ROI. The 'global' approach shows competitive performance as well as numerous advantages as compared to the 'local' one [33,34]. Bias related to the number of ROI, their location or weights is avoided.…”

“…Bias related to the number of ROI, their location or weights is avoided. The absence of ROI overlapping avoids redundant calculations resulting in similar [33] or even faster [34] analysis times. Remapping of the EDP [24,25] is included in the GN algorithm and the potential of cross-grain measurement using this technique (i.e.…”