Wet granular bed eroded by a dry granular flow

TL;DR

This study uses DEM simulations to explore how wet granular beds erode under shear flow, revealing two regimes of erosion and proposing a scaling law based on inertial and cohesion parameters.

Contribution

It introduces a new framework linking inertial and capillary forces to erosion dynamics, with a scaling law for erosion rate in cohesive granular flows.

Findings

Identified slow stochastic and fast collective erosion regimes.

Developed a scaling law for erosion rate based on inertial number and cohesion index.

Determined the threshold conditions for fast erosion regime.

Abstract

Using three-dimensional Discrete Element Method (DEM) simulations, we investigate the erosion dynamics of a cohesive bed composed of wet spherical particles subjected to the shear flow of an overlying non-cohesive granular layer. Cohesion is modeled through a capillary attraction law, where the erosion process is governed by the irreversible rupture of liquid bridges at the interface. By systematically varying the liquid-vapor surface tension and the inclination angle of the bed, we analyze the influence of cohesive strength and flow intensity on the mass entrainment rate. Our results identify two distinct erosion regimes: a slow, stochastic regime driven by granular temperature fluctuations, and a fast, collective regime characterized by a global mechanical instability of the interface. We propose a robust scaling law for the surface erosion rate in the fast erosion regime, based on the interplay between two dimensionless parameters: the inertial number ($I$) and the cohesion index ($\xi$). This framework reveals that the threshold for the fast erosion regime is determined by the ratio of the geometric mean of the driving stresses (kinetic and pressure) to the cohesive resistance. These findings provide a comprehensive description of the coupling between inertial and capillary forces, offering a predictive tool for the stability of cohesive interfaces in sheared granular flows.