Modeling Suffusion in Porous Media Using a Particle Scale Linear Decay Function

Abstract Suffusion with migration of fine particles (fines) in pores is a common subsurface hydrological process in nature, which has significant impacts on breaching of dikes, dam failure, and hillslope instability. However, the effects of soil properties and hydrodynamic conditions on fines migration, and the physical implications of mass flow rates at the particle scale remain unclear. Therefore, a computational fluid dynamic model coupled with discrete element modeling is used to develop a unidirectional seepage numerical model to study the effects of the diameter ratio of the pores to the fines, the volume ratio of the fines to the pores, and hydraulic gradient on suffusion. A time‐scaling factor is proposed to better match the numerical results and physical observations of different physical sizes for the verification of numerical method. An equation is developed based on a linear decay function to calculate the mass flow rate due to suffusion. The results show that the erosion coefficient in the power law function displays a negative correlation with the volume ratio, while it shows a positive correlation with the hydraulic gradient. At the particle scale, the erosion coefficient emerges as a statistical average parabolic function that is based on the ratio of tortuosity of the migration paths and time average velocity of the fines. Moreover, the probability density distribution of the tortuosity ratio indicates that suffusion is regulated by the Weibull function. The erosion coefficient bridges microscopic particle dynamics and macroscopic variables. Model improvements lead to a better mechanistic understanding of suffusion.