A fluid–solid interaction model has been implemented by coupling two numerical methods — computational fluid dynamics (CFD) and discrete element method (DEM) — that capture the mesoscale behaviors of the fluid–solid system. The model is first validated by comparing the results of simulations with two types of experiments: free settling of a single sphere in water and formation of angle of repose of particles under water, which show its capability in modeling the behaviors of both particle phase and fluid phase. The verified model is then used to study factors affecting the soil erodibility, where case studies are designed for soil particles deposited inside a pipe and subsequently subjected to water flow–induced surface erosion. Influencing factors for soil erodibility, including particle diameter and interparticle bond, are analyzed. For cohesionless soils without bond strength, the critical shear stress is found to be linearly related to particle size; while for soils with bond strength, simulation results show that interparticle bonding largely decelerates the erosion process and causes a much lower erosion rate. To further the understanding of soil surface erosion under turbulent flow, the “k–ε” turbulence model has been successfully implemented for the fluid phase. Comparison between the laminar model and the turbulence model shows turbulence significantly accelerates the erosion process.