Geoscience Reference
In-Depth Information
influenced by event fragmentation. The first observation worth being made is that
the period of sweeps (quadrant IV) is much larger than that of the ejections
(quadrant II). This feature seems to indicate that the criterion for detecting ejections
could be further improved: they appear to be a less organized kind of bursting
motion, and there might be advantages to coalesce some small intensity events into
one large event. As in the case of the duration of the event, the period of these
events does not seem to be affected by the presence of a mobile boundary.
The profiles of the normalized momentum transported by each event are shown
in Fig.
25
. Its observation reveals that over the measured reach there seems to be
little effect of the mobile boundary (cf. Nikora et al.
2001
). However, small changes
may pass unnoticed due to data scattering. For this reason, the ratio
M
IV
/
M
II
will
be analyzed with more detail (Fig.
26
). The ratio sweep-magnitude to ejection-
magnitude depends on the roughness Reynolds number
k
s
(Nezu and Nakagawa
1993
, p. 184). Sweep events become stronger as the roughness increases. This
effect attenuates as
y
/
h
increases, a behavior related to changes in the skewness of
u
and
v
distributions.
Figure
26
shows the ratio
M
IV
/
M
II
for the totality of the experimental tests
performed. It stands clear that the sweep/ejection ratio is larger in the fixed bed
tests than in the mobile bed ones with the same
u
*
. This result indicates that the
existence of a layer of finer sediment traveling among the protuberant coarser grains
trigger a flow beahviour propper, in some aspects, of smoother boundaries. The
analysis of Fig.
26
also reveals that the ratio sweep/ejection might decrease with
increasing Froude number. In fact, the lowest values correspond to the Tests E3 and
E3D, with the largest Froude numbers.
Figure
27
shows the vertical distribution of the average maximum value of |
uv
|,
calculated as seen in Fig.
22
. A fundamental difference between mobile and fixed
beds is revealed: in the wall region the profiles clearly depart, revealing that the
maximum magnitude of the event is increased by the presence of a mobile bed. In
the outer region, the results collapse visibly into a single curve.
Analyzing this feature with the previous ones, it can be said that the presence of
the mobile bed imposes a reorganization of the events. They become “sharper”
albeit keeping the same duration at the chosen threshold level. Further investiga-
tions on the variation of this parameter with
H
should be carried out. Under the
hypothesis that the transported momentum does not vary significantly, changes on
the shape of the
u
0
v
0
series must produce effects on higher order moments, namely
the skewness and kurtosis.
The last feature worth mentioning about
uv
max
is that sweep and ejection events
are equally affected from the presence of the mobile bed. Since ejections are related
to the uplifting of low-speed flow and the occurrence of streaky structures in plan
view, this finding should encourage further study on the formation and character-
istics of such streaky structures under generalized sediment transport.
It is concluded that the structure of the profiles is much the same for fixed and
mobile bed. Yet, subtle differences can easily pass unnoticed, hidden in the data
scattering.
