The influence of near bed turbulent flow structures on scour hole development at pipe culvert outlets.
This thesis presents the results of experimental measurements of scouring and turbulent
velocity measurements in and around scour holes downstream of pipe culvert outlets.
Centreline scour hole profiles have been measured at 4 flow rates and 4 tailwater depths
resulting in up to 16 different experimental conditions. These results have enabled the
maximum scour depth downstream of culvert outlets to be quantified in terms of flow rate
and tailwater depth. An equation in terms of densimetric Froude number is presented to
predict scour depth incorporating constants dependent on tailwater depth calculated from
additional equations. The prediction of scour depth for different tailwater depths is of
immediate benefit to design engineers.
Additionally, measurementosf scourh ole profiles haveb een madeu sing four experimental
facilities of different sizes. Froudian scaling protocols have been used to investigate the
effect of model scale on scour hole development. Three experimental model facilities have
been used at the University of Hertfordshire and a fourth prototype scale facility used at
HR Wallingford. Model scale has been found to have an effect with small-scale models
failing to accurately predict maximum scour depth. This is of particular significance as
previous equations predicting scour depth downstream of outlets have been derived from
studiesu sing small-scalem odel facilities.
Velocity data was collected for three tailwater depths at one flow rate using a downward
facing 3 component Acoustic Doppler Velocimeter. A method of bed fixing was used to
enable velocity measurements to be made in scour holes at different stages of development
without changes in bed form during the velocity data collection. From this data mean
velocity vectors and contours have been plotted at different stages of scour hole
development and turbulence intensities and Reynolds stresses have been determined for the
centreline profiles. Further, using the quadrant analysis technique, the near-bed bursting
events have been examined at different stages of development, which has shown that
different events occur at different locations in the scour hole.
The analysis of the turbulent flow structure in the scour hole has revealed that initial scour
is a result of high velocities exceeding the critical velocity for sediment transport. As the
scour hole develops the velocities close to the bed reduce and a gradual increase in scour
depth takes place as a result of quadrant 4 events (sweeps) impacting on the bed in this
region. Upstream of the dune the jet comes into contact with the bed and flow structures
similar to those observed downstream of backward facing steps or dunes in open channel
flow are noted. In particular it is suggested that hairpin vortices may be present in this
region and lead to further scouring.
The identification of flow structures in the scour hole may lead to the development of
computer simulations of scouring downstream of pipe culvert outlets which in the long
term could be used as a design tool.