Size Segregation of Bidisperse Granular Beds in Laminar Shear Flow: A CFD-DEM Investigation

Granular sediments in rivers, coasts, and pipelines often undergo particle size segregation due to the action of the carrier fluid and particle-particle interactions. This process can significantly affect geomorphology and the associated geohazards, but our mechanistic understanding of granular segregation in fluid-driven bedload transport remains elusive. In this study, a particle-scale numerical simulation based on the coupled computational fluid dynamics–discrete element method (CFD-DEM) is conducted to investigate the segregation of a bidisperse bed sheared by high-viscosity fluids. The evolution of segregation under varying shear intensities, characterized by the Shields number, is systematically analyzed in laminar flow. The results show that: (1) under various Shields numbers, the granular bed can be divided into an upper bedload layer (fluid-like, fast-moving) and a lower creep layer (solid-like, slowly moving), with the bedload layer thickening and the creep layer thinning linearly as the shear intensity increases; (2) particle segregation evolves exponentially over time, and at the same duration, the final degree of segregation for the entire bed increases linearly with the Shields number; (3) the segregation timescale shows a non-monotonic dependence on the Shields number, governed by the competing effects of increasing segregation velocity and active layer thickness as the Shields number is increased; and (4) the segregation timescale follows a power-law relationship with the shear rate in laminar flow, showing similarities to dry granular flow behavior. Future work will focus on developing a predictive model that captures the evolution of coarse and fine particle concentration profiles, thereby enhancing our modeling capabilities of granular segregation and its feedback effects on the mobility of sediment transport.