Image Analysis of Sand Void Fabric Based on Stereology Principle

Abstract:

Based on the theoretical framework of quantitative description for void fabric with the principle of stereology, the void fabric tensor redefined with normalized idea, and a novel description method of void fabric for sand was proposed. The porosity amplitude parameter defined by the second invariant of the plane tensor to describe the degree of anisotropy and the direction parameter defined by its component to describe its direction, the plane distribution of void fabric were described by these two scalars. The program with different image processing techniques that eliminate the confounding factors of analysis on scanning electron microscope(SEM) image of sand samples was developed to obtain a clear binary image with optimization methods. The SEM image analysis of sand sample showed that parallel test lines with different angles could measure the average porosity but could not measure the porosity distribution, while the circular test line could measure the plane distribution of void fabric. The new method with two scalar described the pore distribution on the plane, with the changed of pore geometry relationship, which degraded to the existing description format naturally. Image analysis validated the rationality of the new method.

Key words:void fabric; image processing; image analysis; quantitative analysis; sand