Abstract

Turbulent flow around in-stream bridge piers in open channels is complicated, with vortices and eddy motions of various length and time scales. These vortices cause significant channel-bed erosion and sediment transport, known as the scour problem. Consequences of scour include bridge failures, which can cause losses of property and, in some cases, of human life. Sediment scour from around the bridge piers has been a primary reason for all bridge failures and has incurred costs of millions of dollars to repair and rebuild. Researchers have tried to find ways to suppress vortices and reduce scour. They have suggested active and passive methods of suppressing vortices. One of the practical passive methods is to attach a splitter plate to the pier on either the downstream or the upstream side of the pier. \par
This research uses the computational fluid dynamic (CFD) method to simulate three-dimensional turbulent flow around a cylinder with or without an attached splitter plate. The objectives are to understand how effective the attached splitter plate is in reducing vortices, Turbulence Kinetic Energy (TKE), channel-bed shear stress around the cylinder, and to explore the optimal splitter plate length and direction. So far, detailed knowledge of the above-mentioned aspects is incomplete. This study provides a comparison of performance between splitter plates attached on the downstream and upstream side and having different lengths. The CFD simulations used the mesh-based numerical method, with sufficiently fine resolutions for flow regions around the cylinder and near the channel-bed in order to capture detailed turbulence structures. The simulations solved the three-dimensional unsteady Navier-Stokes equations. The scope of work covers an assessment of the suitability of the shear-stress transport (SST) K-ω, the K-ω model, and the K-ɛ model for turbulence closure, as well as sensitivity tests to ensure the independence of numerical results on mesh configuration and time step.

The results show that a splitter plate fitted on the downstream or the upstream side of the cylinder can reduce drag and lift force coefficients. In particular, a downstream splitter plate with a longitudinal length of the pier diameter is the most desirable for minimizing the force coefficients. The downstream splitter plate can reduce TKE and channel-bed shear stress values by 38.7% and 32.81%, respectively. The downstream splitter plate shifts the peak TKE region further downstream, which has positive effect for in-stream pier stability. The force coefficients and TKE value are shown to compare well with the results from other studies reported in the literature. Splitter plates are shown to reduce cross-channel velocities in the channel-centre plane, compared to the case of no plate. The findings from this research are of practical values for the safe protection of bridge piers.