An Improved Method for Determining Constriction Size Distribution Using Discrete Element Method and Its Inspiration for Corresponding Laboratory Tests

Abstract

Constriction is defined as the narrowest passages between pores, and constriction size distribution (CSD) is a fundamental property that quantifies the soil retention capability of filters. Although three types of methods have been developed to determine the CSD, that is, analytical, numerical, and experimental methods, these approaches face limitations such as potential inaccuracies or limited applicability. In this paper, the inverse method combined with discrete element method (DEM) proposed by Sjah and Vincens (2013) was improved from several facets, that is, the characterization of the mean distance between adjacent constrictions, the generation of probe particles instead of base soils, the determination of actual path length for probe particles, and the deletion of clogged particles. Then the improved method was compared with the previous one and other analytical and numerical methods. Besides, different confidence levels, that is, maximum probability of failing to traverse through constrictions continuously, were selected in the improved inverse method to investigate the effect of this value on CSD. Furthermore, the effect of probe particles path tortuosity, the clogging of probe particles within the filter, and filter particle shape on CSD were investigated. These findings provide insights for improving the accuracy of CSD determination in laboratory tests.