- 7th September 2018
- Posted by: Joanne Woodhouse
- Category: Projects
Bridge scour can lead to potentially catastrophic failure of bridges with a risk of loss of life if this occurs whilst users are on the bridge. For this reason, authorities responsible for bridges (Network Rail, Highways Agency, Local Authorities) have regular programmes of bridge scour assessment, which we support. These generally consist of an initial screening of potential risk, followed by a more detailed assessment at specific bridges based on average velocities.
If these traditional tools indicate scour protection should be considered, then this is an expensive undertaking and can be more cost-effective if the extent of the bridge requiring protection can be more closely constrained. This requires a more detailed hydraulic analysis of the complex flow patterns within bridges.
3D Computational Fluid Dynamics (CFD) is a cost-effective alternative to physical modelling to investigate complex flows around bridge piers and abutments. Tests have shown that it compares well to physical models in terms of water surface profile predictions and it has the advantage that the user can extract velocity and shear stress information at any location.
Shear stress predictions on the existing bed are already a powerful output: when compared to critical shear stress for particular bed material sizes (sands, gravels, cobbles), it can indicate potential zones of erosion of existing bed material and inform the choice of material for scour protection. However, we are also able to couple the hydraulic model directly to a sediment transport model to simulate likely bed evolution for a given bed material composition.
Under steady-state conditions, the bed will erode until the increased depth leads to lower velocities and the shear stress reduces below the critical value of the bed material.
The figure above shows predicted equilibrium erosion depth for a mixed sand-gravel bed, along with streamlines. This clearly shows the zones of high risk of erosion at the abutments where streamline curvature leads to high turbulent shear, and flow constriction leads to acceleration. The predicted erosion depths elsewhere can be compared to known foundation depths and scour protection can be targeted at the locations where undermining would be a risk.
It should be noted that the results are very sensitive to the assumed bed material composition and in this project, a further run assuming a uniform sand bed was also undertaken which suggested maximum scour depths could increase by 50% (from 1.4m to 2.1m). Sensitivity testing in this way can be used to give a range of likely scour depths depending on actual sub-surface bed material composition to inform decision making on the necessity of providing protection.