Geoscience Reference
In-Depth Information
Table 2 Conditions of fixed
bed experiments
Test
E1D
E2D
E3D
(m
3
s
1
)
Q
0.0135
0.0135
0.0130
H
(m)
0.071
0.069
0.070
i
0
) 0.0026 0.0030 0.0036
u
*
(m s
1
) 0.042 0.044 0.049
Re (
) 32,452 32,452 31,250
B/h
(
) 5.6 5.8 5.7
Fr (
) 0.57 0.59 0.56
t
*84
a
(
) 0.0211 0.0224 0.0277
t
*c84
(
) 0.0318 0.0316 0.0310
d
84b
(mm) 5.19 5.32 5.41
f
LDA
(Hz) 240 250 300
a
Computed from the armored bed mixture; incorporates an
hiding-exposure coefficient determined as in Wu et al. (
2000
).
(
100
75
50
25
0
0.1
1
10
D
(mm)
E1
E2
E3
Initial
Fig. 17 Sedimentological aspects of the experimental tests. Grain-size distribution of bed-load
and of initial bed
This prediction proved wrong. A surprisingly high bed-load rate for grains larger
than the
d
84
of the initial bed was measured (see also MacAuley and Pender
1999
;
Ferreira
2005
, pp. 139-151; Ferreira et al.
2007
; among others). This size-fraction
represented 1.7%, 2.1%, and 3.1% of the total bed-load of, respectively, Tests E1,
E2, and E3. The presence of coarser grains in the bed-load is shown in Fig.
17
,
where the size distributions of the initial bed and of the bed-load for tests E1 to E3
are shown.
As a consequence of the armoring process, the nondimensional bed shear stress
decreased to values below the critical value (cf. Table
2
). One can conclude that the
observation of t
*84
and t
*c84
works well as a verification but not as a prediction of
bed-load composition and hiding functions. Later it will be discussed how the
differences in the structure of turbulence can help to explain the entrainment of
coarse grain-sizes even if its critical shear stress is larger than the applied shear stress.
