Three-dimensional grain fabric measurements using high-resolution X-ray computed tomography

High-resolution X-ray computed tomography (HRXCT) provides detailed imagery of the interiors of rocks up to hand-sample size, non-destructively and in three dimensions. New tools described here allow these data to be used for analysis of grain fabrics. Two separate sets of measurement techniques have been developed. The first concentrates on quantifying individual crystals (phenocrysts, porphyroblasts) or other discrete objects or void spaces within a sample. Quantifiable properties include location, size, shape, orientation, and contact relationships with adjacent objects. The second set of techniques performs a more general fabric analysis on any distinguishable component in a sample. A fabric tensor can be computed based on a number of metrics, including the star volume distribution (SVD), star length distribution (SLD), and mean intercept length (MIL) methods. The fabric tensors provide principal component directions and magnitudes, which in turn provide a measure of degree of anisotropy and shape indices. Because the SVD and SLD measure only the material of interest, whereas the MIL is also influenced by spatial distribution, results can be divergent for sparse phases. Also introduced is a three-dimensional rose diagram that can be viewed interactively and inspected to reveal further details about non-orthogonal directional components. These techniques are demonstrated through analysis of a garnet-kyanite schist from Mica Dam, British Columbia. (C) 2005 Elsevier Ltd. All rights reserved.