FABRIC ANALYSIS USING POINT COUNTING: A NEW 3-D USE FOR AN OLD 2-D TOOL?

We propose a potentially new application of point-counting using 3-D-sets of oriented thin-sections to document 3-D fabrics. The question arose from point counting of an L-S tectonite on 2 perpendicular sections, both across the foliation, one L-parallel (S2 plane), one L-perpendicular (S1). S1 results were consistent with published XRF data, but S2 yielded unreasonably high silica content. The difference is an apparent artifact by cuts of the fabric ellipsoid; longer grains in S2 causes content to deviate from true, whereas S1, with the smallest fabric ratio, shows more grains in the section. We postulate that this difference using point counting may be used in fabric analysis in two ways: (1) to quantify the fabric ellipsoid using a 3-D series of oriented thin sections, and (2) to identify the fabric axes and shape not obvious in hand specimen. To this end, we studied a suite of sections to see if point-counts produce a useful field of geochemical data between the fabric S1, S2 and S3 planes.

We performed point counts on sections cut according to fabrics, calculated bulk chemistries for each, and compared the data against XRF data from the same samples. Fabrics were quantified using standard techniques for each section, and integrated to determine the fabric ellipsoid. For robust testing, we studied two different rocks (a quartzofeldspathic gneiss and a semi-pelitic schist) to test if the method works for more than one rock type or texture. For structural control, each rock had discernable L-S fabric at the hand-specimen scale so that the orientations of the fabric elements are known, and that sections were cut in proper orientations. The systematic suite of sections for each rock began with the principal planes S1, S2 and S3, and completed with sections at each 22.5° between the principal planes (12 sections/rock).

Results show silica compositions unreasonably high, however, data appear systematically skewed. Clearly precise mineral identification must be achieved to render the method useful. A field on geochemical plots drawn between section data is different by rock specimen where a more defined field is seen for the quartzofeldspathic rock, though ~12 wt.% higher in silica v. XRF data. Normalization to XRF data suggests the S3 section is closest to silica content, but the S2 plane may be the most similar to total XRF data.