Modelling the swelling and osmotic properties of clay soils. Part II: The physical approach

Abstract

The increasing use of clays, with a high montmorillonite content in their mineralogical composition, as hydraulic and contaminant barriers for landfill and soil remediation applications needs to be supported by an adequate theoretical modelling of the mechanical behaviour and transport properties, in order to assess the expected performances in the long term. The framework of the thermodynamics of irreversible processes was adopted in a companiom paper to derive phenomenological constitutive equations for a clay soil characterized by swelling and osmotic phenomena, without specifying any of the physical mechanisms that occur at the pore scale. In this paper, a physical approach is proposed in order to provide an interpretation of the phenomenological parameters, obtained from laboratory tests. The soil structure is assumed to be constituted by montmorillonite lamellae, that can be aggregated to form the so-called tactoids, which have a slit-like geometry. Chemical equilibrium is assumed to be established between the bulk electrolyte solutions and the internal pore solution at the macroscopic scale, so that the hydraulic pressure and ion concentrations can be evaluated through the Donnan equations. Water and ion transport is described at the pore scale through the generalized Navier-Stokes equation and the generalized Nernst-Planck equations, respectively. Mechanical behaviour is modelled taking into account intergranular contact stresses. The approach is applied to interpret literature experimental results, showing how it can reduce the number of tests that need to be carried out and provide insight into the physical mechanisms that determine the observed phenomena.

Critical review of coupled flux formulations for clay membranes based on nonequilibrium thermodynamics

“…(93) can be derived using a physical modeling approach to describe the phenomenological coefficients on the basis of physical and chemical interactions between the solid and liquid phases at the pore scale. As noted in a more recent study by Dominijanni and Manassero (2012b), the theoretical basis for Eq. (93) is valid under the condition that pore-scale variations in hydraulic pressure, ion concentrations, and water velocity within the membrane are negligible.…”

Membrane efficiency and diffusive tortuosity of a dense prehydrated geosynthetic clay liner

“…However, higher solute concentrations cause shrinkage of the diffuse double layers surrounding the bentonite particles and a decrease in u, such that t r / 1 as u / 0, assuming that all other potentially restrictive effects are insignificant. Both pore scale modeling (e.g., Dominijanni, 2005;Dominijanni and Manassero, 2012;Dominijanni et al, 2013) and the aforementioned experimental studies (e.g., see Malusis et al, 2015) indicate that t r for conventional GCLs may be approximated as a simple, linear function of u, as follows: Considering the higher bentonite dry density (or lower porosity) of DPH GCLs relative to conventional GCLs, higher u and lower tortuosity factors (t m and t a ) may be expected for DPH GCLs relative to conventional GCLs tested under similar conditions. However, previous experimental investigations to characterize the membrane behavior of DPH GCLs are limited (Di Emidio, 2010), and the influence of membrane efficiency on restricted salt diffusion in DPH GCLs has not yet been investigated.…”

Use of Geosynthetic Clay Liners for Pollutant Control

“…The average half spacing, b, in perfectly dispersed clays may be estimated, assuming a uniform distribution of the clay platelets in a parallel orientation (Dominijanni and Manassero) [2] from the relation:…”

“…If the clay has an aggregated structure, only the external surface of the tactoids is in contact with the mobile fluid, therefore the void space within the platelets in the tactoids should be subtracted from the total void space to obtain the micro-void space, e m , with reference to the conducting pores (Dominijanni and Manassero) [2]. If N l, AV is the average number of platelets per tactoid, the external or effective specific surface, S eff , and the internal specific surface, S n , are given by:…”

“…As shown in Dominijanni and Manassero [2,6], the framework that includes the aforementioned chemical parameters is able to link the coupled transport phenomena of water and ions by imposing the chemical equilibrium between the bulk electrolyte solutions and the internal micro-pore solution at the macroscopic scale level, through the Donnan, Navier -Stokes and Nernst-Planck equations. Moreover, also some specific aspects of the mechanical behaviour can be modelled and coupled with the chemical and transport behaviour by taking into account the different types of intergranular actions, beyond the solid contact stress and the bulk pore pressure (Terzaghi) [7], such as electromagnetic attraction/repulsion and osmotic swelling/suction forces (Mitchell and Soga) [8].…”