Unified Packing Model for Improved Prediction of Porosity and Hydraulic Conductivity of Binary Mixed Soils

Ammar El-Husseiny

doi:10.3390/w13040455

Abstract

Binary mixed soils, containing coarse sand particles mixed with variable content of fines (fine sand, silt, or clay) are important for several environmental and engineering applications. The packing state (or porosity) of such sand-fines mixtures controls several important physical properties such as hydraulic conductivity. Therefore, developing an analytical packing model to predict porosity of binary mixed soils, based on properties of pure unmixed sand and fines (endmembers), can contribute to predicting hydraulic conductivity for the mixtures without the need for extensive laboratory measurements. Toward this goal, this study presents a unified packing model for the purpose of predicting the porosity and hydraulic conductivity of binary mixed soils as function of fines fraction. The current model modifies an existing packing model developed for coarse binary mixed soils to achieve three main improvements: (1) being inclusive of wide range of binary mixed soils covering the whole range particle sizes, (2) incorporating the impact of cohesive packing behavior of the fines on binary mixture porosity, and (3) accounting for the impact of clay swelling. The presented model is the first of its kind incorporating the combined impact of all three factors: particle size ratio, fines cohesive packing and swelling, on binary mixtures porosity. The predictions of the modified model are validated using experimental published data for the porosity of sand-fines mixtures from 24 different studies. The model shows significant improvement in predicting porosity compared to existing packing models that frequently underestimate the porosity. By using the predicted porosity as an input in Kozeny–Carman formulation, the absolute mean error in predicting hydraulic conductivity, as function of fines fraction for 16 different binary mixed soils, is reduced by 50% when compared to the use of the previous packing model. The current model provides insights about the endmembers properties (porosity, hydraulic conductivity, and grain size) and fines content required to achieve a certain target desirable porosity and hydraulic conductivity of the mixed soils. This can assist the optimization of soil mixing design for various applications.

Highlights

Binary mixed soils are composed of large particles such as coarse sand, mixed with finer particles like fine sand, silt, clay, or bentonite
This study presents a unified packing model for the purpose of predicting the porosity and hydraulic conductivity of binary mixed soils as function of fines fraction
The model modifies the formulations developed originally by Zhang et al [25], to achieve three main improvements: (1) developing a correction function that is inclusive of wide range of binary mixed soils covering the whole range of d f /DL, (2) it takes into account the cohesive packing behavior of the fines, and (3) it accounts for the impact of clay swelling

Summary

Introduction

Binary mixed soils are composed of large particles such as coarse sand, mixed with finer particles like fine sand, silt, clay, or bentonite. It is very important in practice to determine the fines content that is needed to achieve certain desirable porosity (n) and hydraulic conductivity (K) values in the binary mixture This can be achieved by preparing binary mixed soils by mixing two endmembers (pure unmixed sand and pure clay for example) at variable fines content and measuring n and K of the resultant mixtures. Such approach requires extensive laboratory testing and is time consuming To overcome this issue, one would ideally like to be able to predict n and K of binary mixed soils as function of fines content based on the properties of endmembers and without the need to conduct extensive experimental work. Atfirst lowwith finesfines volume fraction (Vporosity f ), fine particles fill in the porosity increase of fines fraction 1).

Objectives

Findings

Discussion

Conclusion