2.4 Experimental Methods and Materials
15
Table 2.1 Experimental parameters
Variable
Prototype
Model B Model Da or Model
Db
Comments
Length of the main
gully, L (m)
752
12.5
3.1
Restricted by the
experimental ground
Dam height, H (m)
*
1
60 H P
1
240 H P
Geometrical
similarity
Drainage area, A
(m 2 )
2.1 × 10 5
60.2
4.1
Geometrical
similarity
Rainfall duration, t
(min)
60–240
20
10
Froude number
similarity
Soil concentration,
C (kg/m 3 )
200
50–70
50–70
λ C = 1.15–3
(Zhang et al. 1994)
Soil loss, S (kg)
4.29 × 10 6 121.5
4.75
Distorted-scale by
Eq. (2.3)
Particle density, ρ s
(10 3 kg/m 3 )
2.65
2.56
2.56
–
*Dams 1, 2 and 8 were 18 m high, Dam 7 was 9 m high, and Dams 3, 4, 5, 6, 9, 10, 11, 12 were
5.4 m high
Spout-type Rainfall Simulator, comprising several rows of leptosomatic PVC pipes
with the spouts 1 mm in diameter, were horizontally arranged on the top of a metal
frame 4 m above the Model Db landscape. The simulator was utilized to simulate
the rainfall in the experimental plot, which measured 1.5 m × 2.7 m.
Water was pumped from a nearby reservoir to a constant head tank, and then
pumped to the rainfall simulators. A short but very intense downpour was applied,
representing the typical rainfall events on the Loess Plateau. Sprinkling intensity
was adjusted by pressure valves to maintain a constant hydraulic head for all nozzles
located at different elevations. A flow-meter and a cut-off valve for each rainfall
unit were manipulated manually to adjust the delivery rate of an immersed pump for
the desired rainfall intensity. The rainfall distribution uniformity, determined using
rain gauges spaced equidistantly over the landscape surface, was measured for three
separate 10 min periods before the experiments started. The uniformity coefficients
of the rainfall intensities for the three simulators exceeded 80% and were constant
with time, especially when the rainfall intensity was 1.2–2.5 mm/min for the Model B
and 1.3–3.5 mm/min for the Model Da and Model Db. Drops, with a median diameter
of about 1.2–2.0 mm, were produced in the simulators and measured by catching
the drops on a sheet of absorbent paper. The rainfall-erosion is associated with the
energy, intensity and uniformity of the rainfall. The average energy impacting the
ground cover per unit area and per unit time of the rainfall generated by the SX2004
Sprayer-styled Rainfall Simulator on the Model B and on Model Db was less than
that generated by the SX2002 Spout-type Rainfall Simulator on the Model Da (Xu
et al. 2006).
15
Table 2.1 Experimental parameters
Variable
Prototype
Model B Model Da or Model
Db
Comments
Length of the main
gully, L (m)
752
12.5
3.1
Restricted by the
experimental ground
Dam height, H (m)
*
1
60 H P
1
240 H P
Geometrical
similarity
Drainage area, A
(m 2 )
2.1 × 10 5
60.2
4.1
Geometrical
similarity
Rainfall duration, t
(min)
60–240
20
10
Froude number
similarity
Soil concentration,
C (kg/m 3 )
200
50–70
50–70
λ C = 1.15–3
(Zhang et al. 1994)
Soil loss, S (kg)
4.29 × 10 6 121.5
4.75
Distorted-scale by
Eq. (2.3)
Particle density, ρ s
(10 3 kg/m 3 )
2.65
2.56
2.56
–
*Dams 1, 2 and 8 were 18 m high, Dam 7 was 9 m high, and Dams 3, 4, 5, 6, 9, 10, 11, 12 were
5.4 m high
Spout-type Rainfall Simulator, comprising several rows of leptosomatic PVC pipes
with the spouts 1 mm in diameter, were horizontally arranged on the top of a metal
frame 4 m above the Model Db landscape. The simulator was utilized to simulate
the rainfall in the experimental plot, which measured 1.5 m × 2.7 m.
Water was pumped from a nearby reservoir to a constant head tank, and then
pumped to the rainfall simulators. A short but very intense downpour was applied,
representing the typical rainfall events on the Loess Plateau. Sprinkling intensity
was adjusted by pressure valves to maintain a constant hydraulic head for all nozzles
located at different elevations. A flow-meter and a cut-off valve for each rainfall
unit were manipulated manually to adjust the delivery rate of an immersed pump for
the desired rainfall intensity. The rainfall distribution uniformity, determined using
rain gauges spaced equidistantly over the landscape surface, was measured for three
separate 10 min periods before the experiments started. The uniformity coefficients
of the rainfall intensities for the three simulators exceeded 80% and were constant
with time, especially when the rainfall intensity was 1.2–2.5 mm/min for the Model B
and 1.3–3.5 mm/min for the Model Da and Model Db. Drops, with a median diameter
of about 1.2–2.0 mm, were produced in the simulators and measured by catching
the drops on a sheet of absorbent paper. The rainfall-erosion is associated with the
energy, intensity and uniformity of the rainfall. The average energy impacting the
ground cover per unit area and per unit time of the rainfall generated by the SX2004
Sprayer-styled Rainfall Simulator on the Model B and on Model Db was less than
that generated by the SX2002 Spout-type Rainfall Simulator on the Model Da (Xu
et al. 2006).