Residual Intensity as a Morphological Identifier of Twinning Fields in Microscopic Image Correlation

2020 
In the microscopic observation of deforming metals, it is well known that crystallite defects that accommodate strain can occasionally become visible, namely, they introduce image contrast to their locality. For microscopic digital image correlation (DIC) applications, this is typically known as a disturbance. Here, we explore a potential upside of these image-intensity offsets, to present a new mode of differential imaging that exclusively displays the underlying plasticity agents. For this, the intensity-offset signal is isolated with residual intensity, essentially the differential between reference- and deformed-configuration intensity of each material point. The premise is showcased over an autocatalytic twin band in Magnesium AZ31, with an advanced DIC instrument that utilizes bright-field optical microscopy. With robust area-scanning that utilizes in-situ corrective measures, a material field of around 5000 grains (13 μm average size) is sampled with a maximal intragranular resolution (~250 data points per grain) for this technique. For added robustness against the intensity alterations, a DIC algorithm (Augmented Lagrangian DIC) that enforces global kinematic compatibility constraints is utilized. The calculated residual intensity map yields a detailed image of the twin networks that show a strong positional alignment with the strain localizations. At the band boundary, the twins (and their accompanying strain localization) protrude into the dormant material in a comb-like pattern. With a combination of high-resolution optics and defects that alter the surface topography, residual intensity presents a new in-situ microscopy mode that is tied to the DIC analysis. This principle also offers potential micro-deformation imaging capabilities for various other material-microscopy combinations.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    38
    References
    1
    Citations
    NaN
    KQI
    []