Geoscience Reference
In-Depth Information
factor in the present laboratory tests. The system also relies on good reflection from
scanned surfaces, inhibiting its use with dark-coloured sediments. The system can
include refraction correction to enable scanning through different media, although
the flume was drained prior to bed measurement for the present tests.
The flume used for the tests is a glass-sided tilting recirculating (water and
sediment) laboratory flume measuring 1.5 m
1.2 m (deep)
45 m. The flume
10 4 for the tests. A full-width erodible-bed test section
slope was fixed at 6.74
measuring 4.6 m long
0.6 m deep was located 25 m downstream of the entrance
to the flume. Immediately upstream of this section was an additional full-width
length of erodible-sediment bed measuring 1.6 m
0.02 m
that acted to provide sediment input to the test section and also limit any potential
scour-hole effects propagating into the upstream end of the test section. For a given
experiment, the entire sediment bed was smoothed, the flume was filled to the required
water depth, and then steady flow was run for 18-21 h for the four basic tests, or
for about 30 min for the additional tests of (a) dunes for high flow velocities, and (b)
measurement of a partly developed bed. A side-looking handheld ADV was used
to measure a centreline velocity profile immediately prior to the end of each test.
The flume was drained after each test and the bed surface measured. All runs
utilised a uniform white quartz sand of median size d
0.12 m and then 2 m
0.2 mm and critical
¼
shear velocity u *c ¼
0.013 m/s. The flows tested covered depths of 0.15-0.55 m
and shear velocities of u * / u *c ¼
1.9-4.3, producing ripples for the four basic tests
and dunes for the high-velocity test.
The scanner setup to measure the exposed bed surface is shown in Fig. 4c . The
wooden measuring frame was removed for flow over the bed during each test to
prevent any interference with the bed development. Owing to limitations of wand
position relative to the magnetic-field transmitter, the wand and tracker (the box
immediately above the bed in Fig. 4c ) were located approximately 0.5 and 0.2 m
above the smoothed sediment bed, respectively. From this elevation, the length of
the laser line emitted by the wand was approximately 250 mm in the streamwise
direction. The wand was moved along the supporting fixed wooden beam (Fig. 4c )
to obtain a 250 mm wide sweep of the sediment bed over a length of about 1 m
across the flume. Starting from the downstream end of the test section, the carriage
supporting the scanning system was then moved 150 mm in the upstream direction
and another transverse sweep of the bed recorded and saved. This continued until a
streamwise length of approximately 4-5 m had been measured in successive
sweeps. For a given transverse sweep, the bed surface data were output in point-
cloud format, with an estimated measurement accuracy of the order of
1 mm. The
data of each sweep were mapped onto a grid of resolutions of 2 and 5 mm in the
streamwise and lateral directions, respectively.
The metal frame and base of the flume introduced distortions in the magnetic
field used to locate the wand position for the tests. In order to correct the measured
sweeps for this distortion, a set of straight bands installed between fixed locator feet
(Fig. 4c ) were measured at each carriage position. The corrections required to
restore the recorded curves (distorted by the metal flume components) to straight
lines at each position were then imposed on the respective measured bed surfaces to
Search WWH ::




Custom Search