Modeling the arrangement of particles in natural swelling-clay porous media using three-dimensional packing of elliptic disks

Abstract

Swelling clay minerals play a key role in the control of water and pollutant migration in natural media such as soils. Moreover, swelling clay particles' orientational properties in porous media have significant implications for the directional dependence of fluid transfer. Herein we investigate the ability to mimic the organization of particles in natural swelling-clay porous media using a three-dimensional sequential particle deposition procedure [D. Coelho, J.-F. Thovert, and P. M. Adler, Phys. Rev. E 55, 1959 (1997)]. The algorithm considered is first used to simulate disk packings. Porosities of disk packings fall onto a single master curve when plotted against the orientational scalar order parameter value. This relation is used to validate the algorithm used in comparison with existing ones. The ellipticity degree of the particles is shown to have a negligible effect on the packing porosity for ratios ℓ(a)/ℓ(b) less than 1.5, whereas a significant increase in porosity is obtained for higher values. The effect of the distribution of the geometrical parameters (size, aspect ratio, and ellipticity degree) of particles on the final packing properties is also investigated. Finally, the algorithm is used to simulate particle packings for three size fractions of natural swelling-clay mineral powders. Calculated data regarding the distribution of the geometrical parameters and orientation of particles in porous media are successfully compared with experimental data obtained for the same samples. The results indicate that the obtained virtual porous media can be considered representative of natural samples and can be used to extract properties difficult to obtain experimentally, such as the anisotropic features of pore and solid phases in a system.

Nanoscale origin of the thermo-mechanical behavior of clays

“…Above the micrometer scale clay particles arrange with some orientational disorder to form a matrix in which other minerals grains are embedded (silica, carbonate...). Current reconstruction algorithms provide realistic arrangements of particles in the matrix (see for instance [22]). Such reconstructions account for the variability of particle size and shape and their orientational order parameter.…”

Influence of crystal structure defects on the small-angle neutron scattering/diffraction patterns of clay-rich porous media

“…The three-dimensional virtual porous medium, representing a particle packing of the 0.1-0.2 mm size fraction and used to calculate SANS/ND patterns, was obtained from work by Ferrage et al (2015). This medium was constructed from the successive deposition of elliptic discs in a gravitational field, similar to the algorithm proposed by Coelho et al (1997).…”

“…to slide, rotate or swivel) and the number (2400) of movement attempts. More details on the detection of contacts and interpenetration of particles can be found in the work by Ferrage et al (2015). The overall degree of anisotropy in particle orientation can be tuned by adjusting the amplitude of motions in a trial-anderror fashion (Ferrage et al, 2015).…”

“…In this study, theoretical SANS and neutron diffraction (ND) patterns are calculated for a three-dimensional virtual porous medium, representative of the packing of swelling clay particles (i.e. vermiculite; Ferrage et al, 2015). This packing accounts for the distribution in size, shape and particle orientation of bulk vermiculite samples (Reinholdt et al, 2013), for which experimental SANS/ND data were collected (Hubert et al, 2013).…”

“…(b) Sample cell designed for neutron scattering/diffraction experiments (Hubert et al, 2013). (c) Three-dimensional particle packing in a square simulation box of 1.2 mm width with a scalar order parameter S = 0.63 (Ferrage et al, 2015). (d) Comparison between experimental rocking curves (solid line) and calculated profiles (solid circles) extracted from the virtual porous medium shown in (c), where represents the angle between the normal unit vector of the flat surface of the particle and the z axis of the simulation box (Ferrage et al, 2015).…”