EXPERIMENTAL APPROACH ON MEASUREMENT OF IMPULSIVE

FLUID FORCE USING DEBRIS FLOW MODEL

Nobutaka Ishikawa , Ryuta Inoue , Kenjiro Hayashi , Yuji Hasegawa

and

Takahisa Mizuyama
ABSTRACT
This paper presents an experimental approach on measurement of impulsive fluid force
using several materials of debris flow model. First, the hydrodynamic test for only water
was performed by using water in stead of debris flow in order to confirm the measurement
accuracy of impulsive fluid force as a preliminary test. Second, three kinds of debris flow
model, i.e., sediment with water, gravel with sediment including water and beads with water
were used as the quasi-debris flow by using channel test with a sharp slope. Third, the
pumice stone produced at the Sakurajima volcanic mountain was used as the quasi-debris
flow by flowing it naturally. This final test resulted in showing the impulsive load time
relation by presenting the surge in front wave.
Key words: debris flow model, impulsive loading,
pumice stone, fluid force, hydrodynamic test
INTRODUCTION
Recently many sediment disasters of Sabo dam as
shown in Fig. 1 have occurred at the mountainous area
in Japan by local downpour based on the global

Fig.1: Debris flow disaster

warming (Sabo Technical Center, 2005). These

disasters may be caused by the impulsive loading of debris flow in the steep slope. In the
current design of Sabo dam, the impulsive loading of debris flow is divided into two
categories, i.e., the one is the impact load due to a huge rock based on the impact theory of
1

Professor Emeritus of National Defense Academy, Research Adviser, Society for the Study of Steel Sabo
Structures,6-20-68,Kugo-cho,Yokosuka,238-0022,Japan(e-mail;cgishikawa@m4.dion.ne.jp)
2
Civil Engineer, Kyosei-Kiko, 1-23-1 Shinjiku, Shinjiku-ku,160-0022 Japan (e-mail;inoue@kyosei-kk.co.jp)
3
Associate Professor, Department of Civil and Environmental Engineering, National Defense Academy,
1-10-20
Yokosuka, 239-8686,Japan (e-mail ;hayashik@nda.ac.jp)
4
Rearcher, Civil Engineering Research Laboratory, 904-1, Tohigashi, Tukuba-shi, Ibaraki, 300-2633,
Japan
(e-mail;
hasegawa@crl.or.jp)
5
Professor, Department of Forestry, Graduate School of Agriculture, Kyoto University,
Kitashirakawa,
Oiwake-cho, Sakyo-Ku, Kyoto,606-8502, Japan(e-mail;mizuyama@kais.kyoto-u.ac.jp)

solid body and the other is the fluid force due to the dynamic fluid theory (Mizuyama,1979).
However, the latter is acted on the Sabo dam as a static load based on the dynamic water
pressure theory. On the other hand, the dynamic response analysis for the single degree of
freedom system structure tells us that the dynamic deformation becomes two times larger
than the static one, if the fluid force acts on the structure impulsively (Ishikawa, et al. 2005).
This means that the structural response will be changed by the action of static or dynamic
loading. To this end, many studies have been made on the fluid force of debris flow based
on the dynamic fluid theory (Hirao, et al.1970, Daido,1988, Miyamoto and Daido,1983,
Mizuyama, et al. 1985, Miyoshi and Suzuki,1990, Horii, et al. 2002).
However, the measurement device with high frequency is required in order to measure the
impulsive loading of fluid force accurately. Further, it should be considered for the
occurrence device for the debris flow, the measurement of flow velocity and discharge. It
should be also properly selected for the materials of debris flow model.
In this study, the hydrodynamic test was first carried out in order to confirm the accuracy of
measurement of impulsive loading of fluid force by using only water as a preliminary test
(Ishikawa, et al. 2006). Herein, both the force component meter and the pressure sensor
were used to measure the fluid force simultaneously. Second, the hydrodynamic channel test
with a steep slope was performed to examine the fluid force-time relations of sediment with
water, gravel with sediments including water and beads with water. Third, the channel test
was also executed for the measurement of the fluid force-time relation by using the pumice
stone produced in the Sakurajima volcanic mountain. Finally, the current design load of
fluid force is compared with the peak load and the stabilized load after the peak load
obtained by the test results using the different materials of debris flow model.
PRELIMINARY
WATER

TEST

BY

The hydrodynamic channel test was set up to measure the load time relation at the instant
of impact of fluid force by using only water in stead of debris flow model as a preliminary
test. Both the force component meter and the pressure sensor were used to measure the fluid
force at the same time. The slope of channel can be changed from 1/50 to 1/5 and the
channel has the length of 12m, the width of 0.5m and height of 0.4m as shown in Fig.2. The
water was flown suddenly by taking off the stopping panel. The pressure receiving panel is
composed of the channel made by Aluminum in which the length is 100mm, the width is
100mm and the thickness is 5mm and is set up vertically as shown in Fig 2.
Measurement Items
The fluid force is measured by the force component meter (frequency is 700Hz) and the
three pressure sensors (frequency is 2.5KHz) as shown in Fig. 3. The flow velocity is

/50

component meter
and pressure receiving panel

supporting device

stopping panel

/5

200
U
U
1000

5
/

200

400
1000

str400
2 ea
0 m
0

g
.
2
:

(
a
)

H
y
d
r
o
d
y
n
a
m
i
c

1
/
5
s
l
o
p
e
c
h
a
n
n
e
l

t
e
s
t
[
m
m
]

[
m
m
]
pressure
100
20

pressure
com
sensorpone
strea
nt
mete
r

component
meter

10

20
15

stream

5
0
00

p
r
e
s
s
u
r
e

3.0

PC

10
.0
PB

2.0
5.5
PA

1.5

2.5

sum of pressure sensors

s

e
n
s
o
r
s

Sum
of
press
ure
senso
rs:

Fx
10 (2.5
PA+ 2.0 PB +
3.0 PC)

Fig.3:
Measurement
system [cm]

(b)
measuremen
t
(c) front

[N]

measured by the Laser-Doppler type meter.

plan
F
i

Accuracy of Measurement

3.5

Figure 4 (a) shows the fluid forcetime relation in case of slope 1/50
and flow velocity of
2.6m/sec.The fluid force measured
by the force component meter is
almost agreement with

the one by the sum of pressure sensors, but is a little different from after 0.55 sec.
This may be caused by no existence of pressure sensor at the upper of pressure receiving
panel. Figure 4(b) illustrates the local pressure time relation measured at the points
PA, PB, PC which are occurred from the bottom of channel in turn. The rise time (0.01 sec)
to the peak pressure measured by the pressure sensors is smaller than the one (0.13 sec )
measured by the force component meter in Fig.4 (a). This may be due to the difference
between the frequencies of pressure sensor and component meter.
Figure 5 also shows the fluid force- time relation in case of 1/5 and velocity 2.0 m/sec. It is
noted that fluid force measured by the force component meter is completely agreement with
the one by the sum of pressure sensors. This may be due to the steep slope channel and
therefore, the starting times of PA, PB, PC are almost the same. The rise time to the peak
load by the pressure sensors is quite quick (0.01 sec) in this case.
.

)
(Fl
N
ui
d
e
c
rfo
fo
d
irc
e
u
lF

Load
cell meter
Component

30
25
20
15
10

Sum
of pressure sensors
Pressure

(N

Time
Time(sec)
(sec)

Fluid force-time
(a) Fluid(a)
force
-time relation
relation

)a
(
P
Pr
e
ru
ess
s
e
ur
ePr
(P

(b)Pressure-time
Time(sec relation
)
Fluid force-time
relation
Pressure-time
relation flow velocity of
( Water, (b)
slope
of 1/50,
Fig.1 Fluid
force-time
relation
Fig.4: Fluid
force-time relation

Fig.4

(water, channel slope 1/50, flow velocity 2.6m/sec)

Sum
Sum
Fl
ui
d
fo
rc
e
(N

ofofpressure
sensors
pressure

Component
meter

-6

Fig.5

Time
(sec)
Fluid
relation
Fig.5: force-time
Fluid force-time
relation

1/5slope
(water,

1/5,2.0m/s
flow velocity

(water, channel slope 1/5, flow velocity 2.0m/sec)

40c40c40c
Sediment
(height of 40cm)
40cmmmm

Discharge 11.5.5
of 1.5 l
/s/
s/s/s
Slope of
17
Paste sand
1. of 1.4mm on floor
1.4444mmmmmmmm
Pressure receiving
panel

Boxes to examine the distribution of grain size

Fig.6
Sharp
channel
test
Fig.6:
Steep
channel
17 test

set-up

set up

STEEP CHANNEL TEST OF DEBRIS FLOW MODEL

Figure 6 shows the steep channel test set-up in which the debris flow model (sediment etc.)
is flown by taking off the stopping panel after piling up the sediment to the height of 40cm.
The channel has the slope of 17 degree, the width of 10cm and the slope length of 5m.
The discharge of water is 1.5 l /sec and 4 kinds of debris flow model are used as follows:

only water,
sediment with water,
gravel with sediment including water and
beads with water. In order to examine the distribution of grain size in the sediment, the
boxes are used at the lower channel end by running them instantly as shown in Fig.6.
Fluid Force-Time Relation
Figures 7,8,9 and 10 show the fluid force-time relations of
only water,
sediment with
water,
gravel with sediment including water and
beads with water, respectively.
Table 1 illustrates the test results.
(1) The fluid force-time relations of
only water (Fig.7) and
gravel +sediment +water
(Fig.9) show the bilinear shape with steep rise time. On the other hand, the fluid
force-time relations of
sediment + water (Fig.8) and
beads + water (Fig.10)
represent the bilinear type with slow rise time.
(2) It is considered that the latter tendency may be caused by the reason why the
consistency is not reached to the equilibrium and the head of flow becomes to the wedge
shape.

Fl
ui
d
fo
rc

Fl
ui
d
fo
rc
Time
(sec)

Time
(sec)

Fig. 7: Fluid force-time relation at sharp channel

Fig.8: Fluid force-time relation at sharp channel

(only water )

( sediment with water )

N
Fl
ui
d
for
ce
Time (sec)
Fig.9: Fluid force-time relation at steep channel
(gravel with sediment including water)

Fig.10: Fluid force-time relation at sharp channel

(bead with water )

Tab.1:
Case

Test results using water, sediment, gravel and bead

Peak

Stabilized

Ratio

Rise

Flow

Load

Load

Fmax/F0

time

Fmax N

F (N)

1(water)

60.5

50.0

2(water)

60.5

3(sediment

Flow

Design

Ratio

Ratio

Velocity depth

load

Fmax/F

F /F

t (s)

v(m/s)

h(cm)

F(N)

1.21

0.10

3.42

4.05

47.4

1.27

1.05

51.0

1.11

0.10

3.26

3.93

41.8

1.45

1.22

40.0

40.0

1.00

0.20

3.05

2.34

39.2

1.02

1.02

50.0

55.0

0.91

0.20

2.38

4.91

46.5

1.08

1.18

50.5

52.5

0.96

0.15

2.54

4.84

52.2

0.96

1.01

50.0

51.3

0.97

0.20

1.96

------

-------

--------

-------

45.0

46.0

0.98

0.20

2.21

--------

---------

---------

---------

100.0

90.0

1.11

0.40

2.33

6.98

79.6

1.13

1.13

100.0

89.0

1.12

0.50

2.40

6.91

83.6

1.08

1.065

+water)
4(gravel
+sediment
+water)
5(gravel
+sediment
+water)
6(gravel
+sediment
+water)
7(gravel
+sediment
+water)
8(bead
+water)
9(bead
+water)

Peak Load and Stabilized Load

The peak load of
beads +water in Fig.10 is the largest among all cases. Because the
impact load of fluid force may be due to the hardness of bead, although the rise time to the
peak load is the latest. The stabilized load means when the fluid force becomes constant
after the peak load. These loads of materials
,
and
are smaller than the peak load
except the material
. The reason why the stabilized load of material (cases 4-7 in
Tab.1 ) becomes larger than the peak load may be due to the effect of sedimentation of
gravel.
Rise Time to the Peak Load

The rise time to the peak load is found as shown in Tab.1 by the fluid force time relations
in Figs.7,8,9 and 10.
The rise times of materials

and

are very slow compared with

only water.

This may be the same reason as mentioned in fluid force-time relation (2).
Design Load
The design fluid force load is computed by using Eq.(1) as shown in Table 1.
2
F = Av
where, :density (g/cm ), A(= b h) : sectional area of channel (cm
),

(1)

b, h :channel width

and the average water depth, v : flow velocity (cm/sec).

It should be noted from Tab.1 that the design loads in all cases are almost smaller than the
peak loads. This may be the reason why the design loads in all cases are almost the same as
the stabilized loads after the peak and the average water depth may be estimated as the
smaller than the depth after the sedimentation.
DEBRIS FLOW MODEL TEST USING PUMICE STONE
Outline of Test
The pumice stone produced in the Sakurajima volcanic mountain was used as the debris
flow model. The slope of channel is 10 and the density of pumice stone is 1.29g/cm .
The method of flow is performed in the two ways as follows as shown in Fig.11.

component meter

stream

n=0.016

pumice

load measure panel

(a) Type A Natural flow
component meter

panel

n=0.016
load measure panel (b) Type B Washout
Fig. 11: Model test using pumice

pumice stream

Time
Fig.12:

Time

Case 1: Fluid force-time relation

(pumice Type B)

[Thick line: design load ]

Time
Fig.14: Case 3: Fluid force-time relation
(pumice ,Type A) [ Thick line: design load ]

Fig.13:

Case2: Fluid force-time relation

(pumice, Type A) [Thick line: design load ]

Time
Fig.15: Case 5: Fluid force-time relation
(pumice, Type C) [Thick line: design load ]

Furthermore, it is added to change the channel slope 17 (Type C).

(1) Type A (Natural flow, channel slope with 10 ): The pumice stones are flown naturally
without using the stopping plate.
(2) Type B (Washout, channel slope with 10 ): The pumice stones are flown by taking off
the stopping plate after sedimentation.
(3) Type C (Natural flow, channel slope with 17 ): The pumice stones are flown naturally
as the same manner as Type A by only changing the channel slope with 17 .
The flow velocities in all cases are about 1.6-1.7m/sec.
Fluid Force-Time Relations
Figures 12, 13, 14 and 15 show the fluid force-time relations in cases 1,2,3 and 5 in Tab. 2 ,
respectively.
(1) It is found that the fluid force-time relations in all cases show the bilinear behavior with

very steep slope rather than the cases of water in Fig.7 and gravel +sediment +water in
Fig.9.
(2) This tendency may be due to the reason why the head flow of pumice forms the surge
shape by coming up to the surface at the front of pumice stones.
(3) It is interested to note that the rise times in all cases are very quick rather than the cases
of sediment + water, gravel + sediment +water and bead + water. This may be caused by
the surge shape in which the front wave of pumice is flown as stepwise.
The Peak Load and Stabilized Load
Table 2 shows the peak load and the stabilized load after the peak in all cases of pumice
stones.
(1) The ratios of peak load and stabilized load (Fmax / F0 ) are almost 1.7-1.9 and this
tendency means the impulsive loading, because of forming the surge shape due to the
effect of rising up to the surface of pumice stones.
(2) However, the ratio (Fmax / F0 ) was 1.2 in case 5 of type C. This may be the reason that
the velocity of front wave increases and as such, the front pumice did not rise up to the
surface and did not represent the surge shape.
Rise Time to the Peak Load
Table 2 shows the rise time to the peak load in all cases using pumice stone.
(1) The rise times of cases 1-4 are all less than 0.1 sec except case 5. This is regarded as the
impulsive loading time relation, and the structural dynamic response will become two
times larger than the static loading, if this impulsive loading acts on the structure
(Ishikawa, N. et al. 2005).
(2) Therefore, the rise time is important factor for the judgment of impulsive loading or not,
although this value is actually compared with the natural frequency of the structure.

Design Load
The thick line in Figs.12-15 and Table 2 show the design fluid force load which is computed
by using the average water depth, the average velocity and the density of pumice stones
3

( = (1.29 0.44 + (1.00 (1 0.44)) = 1.13g / cm )), because the transportation

consistency of pumice stones is measured as 0.44.
(1) The ratios of stabilized load and design load ( F0 / F ) in all cases become almost 1.0.
This fact indicates that the design load coincides with the stabilized load.
(2) The ratios of peak load and design load ( Fmax / F ) of cases 2,3,4 are about 1.7-1.8 and as
such, the impulsive loading is 1.7-1.8 times larger than the design load. This means that
the impulsive load acts on the Sabo dam large rather than the design load.

Tab.2: Test results using pumice

Case

Peak

Stabilized

Ratio

Rise

Flow

Flow

Design

Ratio

Ratio

(Type

load

load

Fmax/F

time

velocity

depth

load

Fmax/F

F /F

Fmax N

F (N)

tr(sec)

v(m/sec)

h(cm)

F(N)

1 (B)

112.2

80

1.4

0.098

2.54

11.6

84.6

1.3

0.95

2 (A)

63.9

38

1.7

0.078

1.60

13.4

38.8

1.7

0.98

3 (A)

82.7

47

1.8

0.092

1.84

12.9

49.4

1.7

0.95

4 (A)

89.2

48

1.9

0.070

1.84

13.0

49.7

1.8

0.97

5 (C)

45.4

38

1.2

0.112

1.71

13.0

43.0

1.2

0.88

Sedimentation Profile
Figure 16 shows the sedimentation profile before the impact to the panel in Cases 1,2,3 and
5. It is noted that the front waves in Cases 2 and 3 resulted in showing the surge shape, but
the front waves in Cases 1 and 5 illustrated the wedge shape.

(a)

(c )

Case 1 Type B

Case 3 (Type A)

(b)

Case 2 (Type A)

(d)

Case 5 (Type C)

Fig.16: Sedimentation profile of pumice before impact to pannel

CONCLUSIONS
The following conclusions are drawn from this study.
(1) It is confirmed that the fluid force measured by the force component meter is almost

good agreement with the sum of pressure sensors. Therefore, the force component meter
can measure the fluid force of debris flow models, i.e., water, sediment +water, gravel
+sediment +water, bead +water and pumice +water.
(2) It is found that it is difficult to get the impulsive loading in cases of sediment +water,
gravel + sediment +water, even if the channel slope becomes steep.
(3) It is interested to note that the front wave of debris flow model using pumice stone
resulted in showing the surge shape and as such, the ratio of the peak load and the stabilized
load became quite large (1.7-1.8). This phenomenon is called as the impulsive fluid force.
(4) The rise time in fluid force-time relation using pumice stone became faster than other
debris flow model materials. This may be due to the effect of forming the surge shape.
(5)These phenomena will be simulated by using the particle method which may be used for
the Sabo dam design in the near future.
REFERENCES
Daido, J. (1988): Imapct Load of Debris Flow acting on Sabo Dam Proc. of Sabo Society
Meeting, pp.275-276.
Hirao, K., Tenda, K., Tabata, S., Matsunaga, M. and Ichinose, E. (1970): Fundamental Test
on the Impulsive Pressure of Surge(Part 1), Journal of Shin-Sabo, Vol.76, pp.11-16.
Horii, N., Toyosawa, Y., Tamate, S. and Hashizume, H. (2002): Special Research Report of
Industrial Safety Institute, No.25, pp.17-23.
Ishikawa, N., Shima, J., Yoshida, K. and Beppu, M. (2005): A Study on the Behavior of
Sabo Dam under Debris Fluid Force, Proc.of the Sabo Society Meeting, pp.224-225.
Ishikawa, N., Hayashi, K., Shima, J. and Mizuyama, T. (2006): Measurement Test of
Impulsive Fluid force acting on Sabo Dam Model, Proc. of the Sabo Society Meeting,
pp.226-227.
Mizuyama, T. (1979): Evaluation of Debris flow Impact on Sabo Dam and Its Problems,
Journal of Shin-Sabo, 112, pp.40-43.
Mizuyama, T., Shimohigashi, H., Nakanishi, H. and Matsumura, K. (1985): Experimental
Study on Debris Flow Loads for Steel Slit Type Sabo Dam, Journal of Shin-Sabo,
Vol.37, No.5, pp.30-34.
Miyamoto, K. and Daido, J. (1983): A Study on Impact Load acting on Sabo Dam (Part 1),
Memoirs of Science and Engineering Institution of Ritsumeikan University, Vol.41,
pp.61-79.
Miyoshi, I. and Suzuki, M. (1990): Experimental Study on Impact Load of Debris Flow,
Journal of Shin-Sabo, Vol.43, No.2, pp.11-19.
Sabo Technical Center (2005): Actual Conditions of Sediment Disaster.