CFD–DEM simulations of current-induced dune formation and morphological evolution

Abstract

Understanding the fundamental mechanisms of sediment transport, particularly those during the formation and evolution of bedforms, is of critical scientific importance and has engineering relevance. Traditional approaches of sediment transport simulations heavily rely on empirical models, which are not able to capture the physics-rich, regime-dependent behaviors of the process. With the increase of available computational resources in the past decade, CFD-DEM (computational fluid dynamics-discrete element method) has emerged as a viable high-fidelity method for the study of sediment transport. However, a comprehensive, quantitative study of the generation and migration of different sediment bed patterns using CFD-DEM is still lacking. In this work, current-induced sediment transport problems in a wide range of regimes are simulated, including 'flat bed in motion', 'small dune', 'vortex dune' and suspended transport. Simulations are performed by using SediFoam, an open-source, massively parallel CFD-DEM solver developed by the authors. This is a general-purpose solver for particle-laden flows tailed for particle transport problems. Validation tests are performed to demonstrate the capability of CFD-DEM in the full range of sediment transport regimes. Comparison of simulation results with experimental and numerical benchmark data demonstrates the merits of CFD-DEM approach. In addition, the improvements of the present simulations over existing studies using CFD-DEM are presented. The present solver gives more accurate prediction of sediment transport rate by properly accounting for the influence of particle volume fraction on the fluid flow. In summary, this work demon- * Corresponding author. Tel: +1 540 231 0926 strates that CFD-DEM is a promising particle-resolving approach for probing the physics of current-induced sediment transport.

Direct Numerical Simulation of Transverse Ripples: 1. Pattern Initiation and Bedform Interactions

“…These fully resolved simulations have provided a wealth of data but are computationally very expensive. Recent modeling efforts have focused on an Eulerian Lagrangian framework to model the flow and the particle bed (e.g., Nabi et al, ; Sun & Xiao, ). Other coupled flow‐particle numerical approaches resolve the flow field, through detached‐eddy, large‐eddy, or direct numerical simulation, and model the bed using the Exner equation (Chou & Fringer, ; Escauriaza & Sotiropoulos, ; Khosronejad et al, ; Sotiropoulos & Khosronejad, ).…”

On the Role of Sidewalls in the Transition From Straight to Sinuous Bedforms

“…Recently, numerical simulations have become a popular tool for studying bed formation. These include particle‐resolved direct numerical simulation (DNS; e.g., Kidanemariam & Uhlmann, , ), as well as simulations that resolve the flow but model the bed through a point‐particle approach (e.g., Sun & Xiao, ) or through the Exner equation (e.g., Chou & Fringer, ; Sotiropoulos & Khosronejad, ). Currently, fully resolved DNS or DNS with some degree of modeling for the bed have been unable to produce sinuous‐crested bedforms.…”

Realistic representation of grain shapes in CFD–DEM simulations of sediment transport with a bonded-sphere approach

“…Computational Fluid Dynamics-Discrete Element Method (CFD-DEM) has been increasingly used in the study of sediment transport (Schmeeckle, 2014;Sun and Xiao, 2016a). In this method, individual particles are tracked in a Lagrangian framework, which offers more insight compared to continuum-based descriptions of the particle phase.…”