Geoscience Reference
In-Depth Information
the pneumatic valve control. Closure of the flow control valve signalled the end of
the experiment, where runs were typically of 20-60 min duration.
In a significant advance on previous studies of bedforms in closed conduits (e.g.
Ismail 1952 ; Nakagawa and Tsujimoto 1984 ), bed profiles as these developed from
plane-bed conditions for the experiments were automatically and non-intrusively
measured through the plexiglass soffit using a high-accuracy high-frequency
Keyence LB300 laser moved along the tunnel at a computer-controlled rate by a
Velmex BiSlide carriage (Fig. 4a ). Bed profiles were measured 0.08-0.10 m from
the tunnel side wall over a length of approximately 1 m. For each run, the rate of
carriage movement was selected to enable the profile length to be traversed in
typically 15-20 s. The frequency of bed elevation measurement was then defined to
give horizontal distances between successive elevation measurements of 1-2.5 mm
for each run. Because any air bubbles along the soffit during the experiment would
prevent local bed-level measurement during a run, prior to each test, a pair of
magnets collocated inside and outside of the upper tunnel surface were used to
sweep air bubbles from the flooded tunnel soffit along the line of the tunnel to be
profiled. Bed elevations, flow velocities, and pressures were measured to resolu-
tions of 0.03 mm, 1 mm/s, and 1 mm, respectively, for each run undertaken.
The experiments undertaken show that analogous to open-channel flows, seed
waves on a planar sediment bed of a closed-conduit are instigated by discontinuities
in the bed, with seed-wave lengths proposed by Coleman et al. ( 2003 ) to be given
by
175 d 0.75 as for alluvial seed waves (see above). For closed-conduit flows,
both ripples and dunes grow from these seed waves (at rates increasing with
increasing flow strength and utilising the mechanisms of bed-form speed decreasing
with size, coalescence, throughpassing, and interwave generation) to limiting
lengths, heights, steepnesses, and bed friction factors that are approximately main-
tained or possibly decrease thereafter (Coleman et al. 2003 ). Figure 5 confirms the
respective ripple and dune natures of the bed forms for the present finer-sediment
and coarser-sediment runs, where the measured data indicate that equilibrium
closed-conduit ripple and dune magnitudes can be predicted using relations derived
for equivalent open-channel flows. Limitation of free-surface deformation does
result in increased rates of bed-wave development for closed-conduit flows in
comparison to open-channel flows. The closed-conduit tests undertaken confirm,
however, that the interaction of bed waves with associated free surface waves is
principally an effect, rather than a cause, of bedform initiation and growth to
equilibrium.
l ΒΌ
2.3 PIV Measurements and Fluid-Sediment System
Instability or Turbulent Fluid Motions
In the 1980s, particle image velocimetry (PIV) emerged as a powerful tool for
simultaneously measuring velocity time series at multiple points over an extended
Search WWH ::




Custom Search