USE OF IMAGE-ANALYSIS IN MATHEMATICAL CHARACTERIZATION OF ORIENTATION DATA

The mechanical behavior of particulate media is influenced to a great extent by their discrete nature. To date, however, few attempts have been made to quantitatively relate microfabric to the stress-strain characteristics of these media. Recent advances in image analysis have opened new avenues for the use of mathematical characterization of these materials in constitutive models. A technique for it is presented. A comprehensive description of geomaterials can be achieved via two measures: a scalar measure giving the mean value of a fabric descriptor, and a tensor measure giving its distribution in space. The technique is based on extension to the classical stereo-logical principles for implementation in an image analyzer. Use of the technique is illustrated with observations of the microstructure of soils. The effect of fabric on the response of soils under one-dimensional compression is quantified on the basis of the proposed characterization. The proposed technique may also prove useful in quantifying the microstructure of many other types of particulate materials and cellular arrangements.