Prediction of local scour around bridge piers under livebed regime using a CFD-DEM model

Scouring around bridge pier has been identified as one of the main causes of bridge failures in Australia and worldwide. This could lead to the loss of soil support and ultimately the collapse of the bridge structure without prior warning. The current methods for scour estimation in many practicing bridge design codes involve empirical equations which have limited applicability due to being developed based on limited ranges of flow and sediment conditions and pier geometries. Also, the empirical models are mostly overestimating and occasionally underestimating the scour depth. As an alternative to empirical equations, numerical modelling has been proposed in the literature for estimation of bridge pier scour. The interest for numerical modelling is ever increasing due to the advancement in computing technology including high speed computers and sophisticated software tools. Current numerical models proposed in literature include semi-empirical and Eulerian two-phase CFD models. However, none of those models are capable of resolving the two-way interactions between sediment particles and the flowing water. Accurate estimation of scour around bridge pier requires a comprehensive analysis of the interactions between fluid, soil and bridge structure. The most advanced numerical algorithm proposed so far to properly resolve such interactions, is the coupled CFD-DEM algorithm which couples CFD (Computational Fluid Dynamic) and DEM (Discrete Element Method) modules and accounts for two-way interactions in particulate flows. One of the major limitations of the CFD-DEM methodology is the significant computational overhead - even for a modest number of sediment particles - as it requires computation of both the turbulent flow structure and friction-dominated granular mechanics and the coupling between the two. This computational overhead is then almost prohibitive for the numbers of particles (tens of thousands to more than millions) required by a full-scale simulation of pier scour.

To properly resolve the multi-body interactions involved in scour process with reasonable computational expenses, a small-scale CFD-DEM model is proposed in this study. The model is developed with the aid of periodic boundaries to simulate an infinite river bed under live-bed condition. Periodic boundaries not only reduced the computational domain size and, consequently, the computational expenses, but also facilitated modelling the live-bed scour regime with suspended particles. The small-scale model has been validated against experimental data for both scour initiation and scour extent. Results indicate that the microscale model is reasonably capable of predicting the scour initiation as well as the equilibrium scour depth. The verified model is used to identify the governing mechanisms and controlling variables in scour process using a detailed parametric study. Also, the model provides ground for developing an upscaled model for estimation of local scour around bridge piers.

Finally, a novel upscaling methodology is proposed to upscale the results of the small-scale CFD-DEM model to predict local scour around bridge piers. Via this approach, results of the small-scale CFD-DEM model are populated into a sediment-specific scour function estimating the scour as a function of fluid shear stress and particle collision velocity. The scour function is developed for a given sediment assemblage and varying flow condition and could be developed for any sediment assemblages. This upscaling model provides a viable methodology for the macroscopic prediction of scour in engineering applications with modest computational resources.