Research papersQuantifying representative elementary volume of connectivity for translucent granular materials by light transmission micro-tomography

•We develop a light transmission micro-tomography (LTM) based on micro-structure.•LTM is used for estimating representative element volume (REV) of translucent silica.•Variation curve of the relative gradient error facilitates identification of REV plateau.•REVs of multiple connectivity variables are estimated by relative gradient error.•REV approximate 10.0 mm is available to represent connectivity on the whole.

Aquifers composed of granular materials are major repositories of groundwater resource in which water can flow freely and be stored abundantly. Undoubtedly, exploring connectivity of granular materials is essential to understand the mechanism of water and contaminant migration in subsurface environment, while characterizing the connectivity remains a difficult task currently. This study proposes a new light transmission micro-tomography (LTM) with high resolution to address this problem. The new approach relies on scanning micro-structure by light transmission through translucent granular materials in given thickness. An experiment of light transmission through a two dimensional (2D) sandbox packed by heterogeneous translucent silica is conducted to examine the efficiency of LTM in capturing all the features of connectivity including porosity (n), density (ρ), solid phase-pores interface area (Asp), and tortuosity (τ). Considering the importance of representative elementary volume (REV) in characterizing the representativeness and reliability of connectivity, associated REV scales of characteristic variables are also estimated using a criterion of relative gradient error (εgi). Results suggest that the frequencies of minimum REV sizes of connectivity are close to Gaussian distribution in 0.0-12.0 mm and the REV size of approximately 10.0 mm is available to represent connectivity of translucent silica. Then the quantification of connectivity and the corresponding REV estimates are significant for accurate simulation of fluid migration and for associated optimal design of contaminant remediation in subsurface environment. More important, this study provides the possibility of rapid, handy and economical on-site measurements of connectivity for translucent materials.

Graphical abstractDownload high-res image (111KB)Download full-size image