This work describes the free surface flow and clear water scour around an elongated vertical pier.
Measurements were made using two-component Laser Doppler Anemometry, with the main objective of
studying the turbulent sediment transport around elongated piers.
The flow around a circular pier has been the subject of a large number of studies (e.g. Dargahi (1989),
Graf (1998) Roulund et al. (2005)). When the pier is mounted on a sediment bed, the tri-dimensional flow
field around the pier interacts with the bed of sediments (either cohesive or cohesionless) and originates
clear water scour. The flow around a pier is characterized by the existence of a horseshoe vortex in front
of the pier, which is responsible for the evolution of the scour process. The scour hole increases both in
depth and in extension towards upstream, while the horseshoe vortex continues to excavate the sediment
slope. As the sediments are suspended in the flow, they travel downstream and eventually fall and
sediment in regions of lower turbulence. The scour cavity dimensions increase both upstream and
downstream, until equilibrium is achieved. This phenomenon is usually studied by means of mathematical
modelling (Kirkil et al. 2008, Kirkil et al. 2009), due to the experimental complexity of the scour process.
The experimental work concerning initial conditions (Roulund et al., 2005), i.e. prior to scour cavities
developing, can however provide adequate validation of models.
Experiments were conducted in a 0.4 m wide and 16.7 m long flume, and the test section was installed
approximately 9.7 m downstream the beginning of the flume. The test section consisted of a rectangular
vertical pier with round nose shapes. The pier was 4 cm wide and 8 cm long. Two alternative
experimental configurations were studied: (i) the flow over a flat bottom and (ii) the flow over a sediment
bed. In case (i) the pier was mounted in an acrylic plate 1.8 m long, were a layer of uniform sand with
0.376 mm mean diameter was glued, in order to reproduce the natural roughness of a river stream. In case
(ii) the pier was mounted in a box, 10 cm deep and 1.8 m long, filled in with uniform sand with a mean
diameter of 0.376 mm. The flow was controlled by a gate valve and measured by an electromagnetic flow
meter (ABB, model IDE41F). Water depth was controlled by a sluice gate at the downstream extremity of
the flume.
The light source was an Argon-ion Laser (Spectra-Physics, Model 177-G0232) and the optical system
was a Dantec 60X41 FiberFlow, including a 40 MHz frequency shifter and colour separation, combined
with an 85 mm probe together with a beam expander and 500 mm front lens. A colour separator (55X35)
split the collected light into 2 wavelengths before reaching the corresponding photomultipliers (Dantec
57×18). LDA data processing and acquisition was performed using BSA F60 Flow analyser (Dantec).
Simultaneous measurements of longitudinal (horizontal) and vertical velocities were made using the
overlapped coincidence method.
The flow was studied for water depths of 5 and 15 cm and mean horizontal approach velocities equal to
0.17 and 0.25 ms-1.
Measurements of horizontal and vertical velocities in the vicinity of an elongated pier show the strong
interaction between the structure and the flow. In case (i) the flow decelerates as it approaches the pier,
and is defected towards the bottom. In the wake of the pier there is a reverse flow until one pier diameter
downstream, and then the flow starts to accelerate, as the influence of the pier disappears. In free surface
flows Reynolds stresses usually present negative values, but in the present case, as the flow approaches
the pier, Reynolds stresses become positive and increase. In case (ii) a similar flow pattern was found, but
smaller turbulence intensities were measured.