FLOW-INDUCED CHANGES IN STREAMBED TOPOGRAPHY UPSTREAM OF A CYLINDRICAL OBSTACLE

Riverbed geomorphology is subject to substantial changes near in-stream obstacles. Scour holes, which develop locally at the junction region between the obstacle and the streambed, result from complex flow-structure-sediment interactions. In particular, a bridge pier mounted on a loose streambed introduces pressure gradients that are not present in the unobstructed flow. Approaching fluid cannot negotiate the mounting pressure and separates to form a number of characteristic flow patterns around the pier, such as: the system of horseshoe vortices acting at its base, the downwash of fluid at the leading edge, the accelerated fluid flow at the flanks and the wake vortices. The action (and interaction) of these flow structures triggers a mechanism that mobilizes sediment at high rates. Despite a plethora of related studies, important aspects of this mechanism remain poorly understood. For example, the organized vortical motions around the pier cannot be resolved in a conclusive way using traditional flow measuring techniques. It is therefore not clear how these low frequency, energetic vortices contribute to the various phases of scouring. Furthermore, synchronized measurements of flow velocities and bed elevation over an appreciable portion of the region of interest are difficult to obtain. Motivated by these research needs, we applied the Particle Image Velocimetry (PIV) technique to simultaneously map the turbulent flow and the scour hole upstream of a cylindrical model pier. We tested clear-water scour (approach flow was free of sediments) developing on a laboratory streambed of non-cohesive sediments (crushed limestone). Emphasis was placed on the initial phases of scouring. PIV results revealed the highly unsteady behavior of the horseshoe vortex, which eluded a time-averaged characterization. We also found differences between the results of this test and those reported from studies that used “frozen” geometries of the scour hole as a means to obtain spatially-resolved velocity measurements. Finally, we combined flow dynamics with active bed topography (also extracted from PIV images) to identify cause-and-effect relationships and to highlight feedback loops between the flow and the streambed.