On the limits of the Inverse SURFOR Wheel in fabric analysis
Gary Solar, SUNY Buffalo State
A major component of fabric analyses is to determine the shape and orientation of the finite grain-shape ellipsoid. Several methods have been used with success. One is the Inverse SURFOR Wheel of Panozzo (JSG 1987) that uses grain boundary intercepts with a grid of parallel lines. The method was original formulated to determine a strain ellipse using 2-D imagery, and presumably an ellipsoid using mutually perpendicular images. Like many methods, the Inverse SURFOR Wheel would return a strain ellipse, at least in part, in cases where original geometries of grains are known or determined by knowing the protolith fabrics. High-T metamorphic textures, on the other hand, do not enjoy such control because recrystallization destroys fabrics progressively; these textures are the result of a concert of pre-, syn- and post-microtectonic processes. Yet, the Inverse SURFOR Wheel is still useful as a method to return fabric ellipse shapes and orientations of both the bulk and individual phase fabrics. Regardless, like all analyses, any method used to determine fabric shapes and orientations, requires understanding of the limits and uncertainties on that method. This study was devised to investigate both of these aspects using 2-D analog experiments where in each the object shape and orientation is known, and where perturbations to fabrics using such objects are also known.
The Inverse SURFOR Wheel uses grain boundary intercepts of a set of parallel lines inside a circle (the wheel). Intercepts of the grid lines with grain boundaries are recorded at each 10 degrees of rotation of the wheel through 180 degrees. A plot of counts at each step is smoothed with a sinusoidal curve. The difference in intercept counts between the crest and trough is taken as the fabric shape aspect ratio, and the coordinate orientations of those are taken as the short and long axis orientations, respectively. Thus reveals a shape and orientation of the fabric ellipse in the two dimensions of the analysis.
Experiments began with unit circles and squares as control that returned between 8 and 10 percent uncertainties, but no fabrics (as expected). Experiments of known fabrics returned good agreement between the known shape aspects and orientations, but with artifacts between the grain spacing, grain size vs. grid spacing. The results do confirm the method's reliability, and confirms the value of the fitting of the smoothed sinusoidal curve as the way to determine the ellipse shape and orientation, rather than using intercept maxima and minima. Caution must be employed however to recognize that folded and S-C fabrics do not return an ellipse unless each part is counted separately.