LBM Simulation of Self-Assembly of Clogging Structures by Evaporation of Colloidal Suspension in 2D Porous Media

Abstract

Evaporation of colloidal suspension in two-dimensional (2D) porous media leads to the formation of self-assembled clogging structures (SCS). The self-assembly pattern is studied with a hybrid two-phase lattice Boltzmann method incorporating non-isothermal phase change, particle transport and deposition models. During drying, particles accumulate along the liquid–vapor interface while colloidal suspension is evaporating. Upon reaching a certain local concentration threshold, the particles deposit and form a solid structure. The patterns formed by these structures are analyzed in different 2D porous media. In small porous systems of 4 pillars, the self-assembly of C-shaped and X-shaped structures is observed, which compares well with experimental bridge configurations. SCS in porous media of three different initial particle concentrations and of three different porosities are studied in larger porous systems. Simulated self-assembled clogging configurations show good qualitative matches with experimental configuration results. Particle concentration and porosity are both seen to affect the dynamic drying processes as well as the final self-assembled clogging configuration. The liquid configuration and the clogging structure affect each other mutually during drying. With initial concentration increasing from C0 = 0.00 to C0 = 0.16 at a given porosity $$ \phi_{0} = 0.68 $$ , the average evaporation rate and porosity decrease by 21.9% and 1.9%, respectively, due to blockage of pores. With increasing initial porosity from $$ \phi_{0} = 0.53 $$ to $$ \phi_{0} = 0.81 $$ at a given concentration of C0 = 0.16, the average evaporation rate increases by a factor of 2.9 due to larger liquid–vapor interfacial area. Also with the given concentration of C0 = 0.16, the decrease in the porosity (0.94%, 1.9% and 2.7%) is higher for higher initial porosity ( $$ \phi_{0} = 0.53, \, 0.68,{\text{ and }}0.81 $$ ), since more particles (proportional to $$ \phi_{0} \cdot C_{0} $$ ) are initially present. This work opens the door for numerically assisted design of colloid-deposition-based clogging patterns.