Document Actions

ijege-13_bs-bregoli-et-alii.pdf

background image
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
115
DOI: 10.4408/IJEGE.2013-06.B-08
EXPERIMENTAL STUDIES ON 3D IMPULSE WAVES GENERATED
BY RAPID LANDSLIDES AND DEBRIS FLOWS
F
rancesco
BREGOLI
(*)
, a
llen
BATEMAN PINZÓN
(*)
,
V
icente
MEDINA IGLESIAS
(*)
& D
iego
A. GÓMEZ CORTÉS
(*)
(*)
University of Catalonia, Department of Hydraulic, Marine and Environmental Engineering, Technical, Sediment Transport
Research Group (GITS) - C/Jordi Girona 1-3, Campus Norte UPC, Edifici D1, 08034 Barcelona, Spain
Email: francesco.bregoli@gits.ws - allen.bateman@gits.ws - vicente.medina@gits.ws - diegoagomez86@gmail.com
K
ey
words
: impulse waves, debris flow, landslide
INTRODUCTION
An impulse wave is created when a sufficient quan-
tity of material, having a high velocity, enters a reser-
voir, a natural lake, a fjord or the sea. The momentum
of the sliding material is transferred to a mass of water
turning into a giant wave able to travel large distances.
That particular phenomenon, known also as land-
slide tsunami wave, can be highly destructive and un-
likely predicted. Block mass, granular mass as well as
volcanic lava can trigger such phenomenon. The present
study focuses on granular mass sliding in a water body.
That phenomenon can take place in mountain-
ous zones where slopes instability is more frequent.
A landslide or a debris flow can be triggered by
means of various behaviours as the increase of wa-
ter pore pressure in the soil due to heavy rainfall or
snow melting, earthquakes and defrosting of alpine
permafrost in between others (c
oussot
& M
eunier
,
1996; H
ungr
et alii, 2001; i
Verson
et alii, 1997).
When such sliding mass with a relative high velocity
hit a water body, a set of giant waves is generated
and propagates for a long distance. Initially that wave
can produce a very large run-up of hundreds of me-
ters destroying the shoreline or easily overtop dams
(P
anizzo
et alii, 2005).
Although having a high destructive potential,
debris flows or landslides seem to produce a tsunami
wave that rapidly decays if compared with earthquake
ABSTRACT
The input of material in a water body at high veloc-
ity, like a landslide or a debris flow, can induce a big,
abnormal wave, known as impulse wave or landslide
tsunami wave. Once the wave is triggered, the effects
on the shorelines are devastating and moreover unlikely
predicted. Disastrous past events have been extensively
analysed but remain too scarce to properly describe
the process. Experiments of impulse waves have been
carried out by various authors. The present work was
planned to fill the lack of experimental data on the ef-
fect of granular material falling in a water basin, explor-
ing new ranges of parameters governing this process.
It is introduced a new experimental set up installed
in the fluvial-morphodynamic laboratory of the GITS
team in the Technical University of Catalonia. The ex-
perimental device consists of a wheeled box sliding on
a steeply sloped flume and releasing granular material,
which ends up in a basin. The system allows reaching
a relatively high velocity of the granular mass for a
correct simulation of the process’s behaviour.
A system has been defined in order to be able
to measure the velocity of the granular material and
its depth, as well as the propagation of the waves,
with high-speed cameras and a laser grid system.
The dynamic forces of the granular mass’ impact on
different surfaces is also measured and related to the
studied phenomenon.
Several experiments’ runs have been carried out.
First results are here presented and analysed.
background image
F. BREGOLI, A. BATEMAN PINZÓN, V. MEDINA IGLESIAS & D.A. GÓMEZ CORTÉS
116
International Conference on Vajont - 1963-2013 - Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
by the considerable number of parameters involved,
and not always easy to control (Fig. 1). The param-
eters are listed in Tab. 1.
Some considerations and simplifications are ini-
tially introduced for the presented study:
tsunamis (H
eller
& H
ager
, 2010). It suggests the dif-
ferent scale of the processes and the forces involved:
crustal blocks that move during large earthquakes are
incomparably larger than any landslide.
Disastrous past events are testimonials of impulse
wave’s power. Some catastrophes are well reported by
different authors: Ariake Bay, Japan 1792 (M
iyaMoto
,
2010); Lituya Bay, Alaska 1958 (F
ritz
et alii, 2001);
Vajont Dam, Italy 1963 (D
atei
, 1968); Lake Yanahuin,
Perú 1971 (s
lingerlanD
& V
oigHt
, 1979); Stromboli
Island, Italy (various events, M
araMai
et alii, 2005;
t
inti
et alii, 2005; B
ellotti
, et alii, 2009).
Experimental studies have been carried out with a
rigid body or a sliding granular material plunging in a
straight channel (two-dimensional, 2D) or in a water
basin (three-dimensional, 3D).
Works on 2D channel are, among others: r
ussell
(1837), n
oDa
(1970), K
aMPHuis
& B
owering
(1972),
H
uBer
(1980), F
ritz
(2002), z
weiFel
(2006) and H
el
-
ler
et alii (2007).
Works on 3D basin are, among others: H
uBer
(1980), P
anizzo
et alii (2005) and D
i
r
isio
(2009).
In particular, these last two works are concerning
a rigid body entering a water basin.
In the present studies a new laboratory facility
described hereafter is set up to explore new range of
variables of the presented phenomena for a granular
sliding mass entering in a water basin. A new measur-
ing system is settled up too. The preliminary experi-
mental results are presented and discussed.
PARAMETRIZATION OF THE PHENO-
MENON
The complexity of the phenomenon is highlighted
Fig. 1 - Sketch of the phenomenon and the related governing parameters: (I) lateral view; (II) aerial view of the channel with
the coordinates system; (III) aerial view of the schematic final deposit of material
Tab. 1 - List and description of governing parameters. LS:
landslide; FD: Final Deposit
background image
EXPERIMENTAL STUDIES ON 3D IMPULSE WAVES GENERATED BY RAPID LANDSLIDES AND DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
117
slope where a metallic wheeled box is sliding along
rails with a low roughness. At the end of the channel,
a high resistance piston suddenly stops the box forcing
the opening of the door and thus the releasing of the
granular material (see Fig. 2 and Fig. 3).
The material plunges into the basin and triggers
a set of waves.
The main characteristics of the experimental
device are: (1) maximum sliding length 6.20 m, (2)
angle of impact up to 27.5º, (3) slide impact velocity
• only one type of bulk material is considered;
• the sliding mass hit the water body perpendicular to
the slope (β
=
90º);
• only constant sliding slope angle, width and length
of sliding mass are considered;
• only the principal direction (Γ
=
0º) is considered.
Nevertheless almost every parameters can be mod-
ified, according to the configuration of the experiment.
Given the previous hypothesis, the wave ampli-
tude a, as well as the wave height h, is a function of a
reduced number of variables as below
where the subscript M means the maximum value of
the wave’s amplitude and height, which are registered
near the impact.
With the objective to properly and synthetically
describe the phenomenon in act, new parameters are
defined (dimensions in square bracket):
• The momentum flux of the sediment,
• The initial water pressure,
These magnitudes can be related in a new dimen-
sionless parameter:
Other interesting parameters are the ones related
to the final deposit that can be synthesized as the el-
lipse’s area A
d
=
πa
d
b
d
which is a measure of the basal
friction the deposit has experienced.
EXPERIMENTAL SET-UP
A new experimental device has been settled in
the laboratory of Hydraulic Engineering of the Tech-
nical University of Catalonia in Barcelona. The facil-
ity was planned to study the effect of a granular slide
plunging in a basin.
The set up consists in a steep flume with a variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
Fig. 2 - Frontal view of the box sliding on the flume; on the
top of the kart, the piston that absorbs the hit
Fig. 3 - The slide is plunging in the basin after the doors
opened. In this case the mass is 120 kg, reaching
a velocity of around 5.5 m/s on a slope of 27.5º.
The frame is taken from a video of a high velocity
camera recording at 650 frames per second (f/s)
background image
F. BREGOLI, A. BATEMAN PINZÓN, V. MEDINA IGLESIAS & D.A. GÓMEZ CORTÉS
118
International Conference on Vajont - 1963-2013 - Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
wave’s height. A description of that process is given in
the Fig. 5, where it is possible to appreciate the evolu-
tion of the impulse wave. The water splash and the first
formed wave have a completely different size, the first
being much greater that the second. A similar situation
is described in F
ritz
et alii (2004). For that reason, it
is decided to differentiate between the measurement of
the maximum wave’s height and the initial splash.
RECENT ADVANCEMENTS ON DEBRIS-
FLOW IMPACT
Recently a high-frequency-response load cell has
been purchased in order to conduct a parallel research.
The study consists on the evaluation of the effect of
a debris flow hitting vertical structures. The intent is
to analyse the efficiency of various countermeasure
structures like walls and deflectors, and the stress ex-
perienced. The experiments have been run in the same
configuration as the impulse wave generation. There-
fore this research could be an opportunity to study the
pressure that a granular sliding mass can produce in
correspondence to the entrance into the basin. That
pressure can be related to the produced wave to bet-
ter understand the formation and behaviour of waves.
The load cell (sketch in Fig. 6 I) is able to meas-
up to 6 m/s, (4) sliding mass weight up to 150 kg,
(5) initial basin’s water depth from 0.2 to 0.6 m, (6)
dimension of the basin 4.10x2.45 m.
Due to the reduced dimension of the basin in re-
spect to the velocity of propagation of the produced
waves, only the first wave can be measured: the fol-
lowings are distorted by the reflection on boundaries.
In the laboratory, granular materials with different
characteristics are available. At the moment the granu-
lar material used corresponds to selected white gravel
with a mean grain diameter of 19.5 mm and a median
diameter of 16.9 mm (see Fig. 3).
MEASUREMENTS
To run a complete experiment actually is neces-
sary to repeat the same experiment twice at different
environmental condition. The first experiment (named
type a) is carried out with a high velocity camera (650
frames per second) focusing on the inlet of the basin
with a powerful illumination (Fig. 3). A calibration
grid is applied and the measurement of the geometry
and velocity of the granular slide material is achieved.
The second experiment (named type b) is carried
out to observe the propagation of the wave. Three high
definition video cameras record the basin from differ-
ent points of view, where different laser sheets project
lines that mark the free surface. The water is previ-
ously filled with a small amount of kaolin that colours
the fluid, reflecting the lasers (Fig. 4).
Using a calibration process and mathematical
transformation’s algorithm created ad-hoc, the met-
rical measurement with a high precision is achieved
(B
regoli
, 2008). The final deposit is also measured
after each run once the basin is emptied.
The evaluation of the maximum wave’s height and
amplitude needs a particular caution. The formation of
the wave is anticipated by a big splash in the zone of
impact, making difficult the choice of the maximum
Fig. 5 - Photo sequences at 26.43 f/s of a forming wave triggered by a mass of 125 kg sliding along a slope of 27.80º. The ar-
rows represent the main direction of wave’s propagation which corresponds with the direction of the entering sliding
mass. The basin’s water depth is 0.2 m. The splash (near field), the formation of the wave (transition) and a defined
wave crest travelling along the basin (propagation) are visible
Fig. 4 - Laser grid reflecting over the water surface,
filled by kaolin. The arrow represents the direc-
tion of the sliding mass entering in the basin. The
material is already plunged in the water and the
wave is forming. The green line shows the main
direction of propagation
background image
EXPERIMENTAL STUDIES ON 3D IMPULSE WAVES GENERATED BY RAPID LANDSLIDES AND DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
119
RESULTS
14 runs, corresponding to 7 complete experiments
have been carried out. The results presented, concern
the principal direction (Γ
=
0º), marked with the green
laser sheet (see Fig. 5).
A plot of the evolution of the wave for the dif-
ferent runs is presented in Fig. 7, where it can be
appreciated the decay of the waves over the time.
The decay is similar for every run. The example of
regression line for the run relative to a slide of 75 kg,
shows a strong wave’s decay of 66 mm/s, which cor-
responds to a reduction of around 50% in 1 second,
respect to the initial wave amplitude.
In the Fig. 8 the relationship between the ratio of
maximum amplitude and maximum height of wave
and the dimensionless parameter Ψ is reported. In this
case, it is difficult to recognize a pattern, due to the
limited range of parameters.
ure forces and torques along the three direction with
a frequency of 1000 Hz. The structures used for this
research are: a planar structure (that simulates a
check-dam, Fig. 6 II), a “comb” structure (that simu-
lated a filter-dam, Fig. 6 III), a deflector at different
degrees and a system to study the vertical distribu-
tion of impact’s forces of the granular slide. For the
present study, only the planar structure is taken into
account.
Fig. 6 - Experiments on debris flow’s impact on structure. (I) sketch of the load cell, able to measure forces and torques along
the 3 directions; (II) lateral view of the planar plate structure (check-dam) with the integrated load cell; (III) view of
the “comb” structure (filter-dam)
Fig. 7 - Decay of the wave crest over time. The legend
shows the different runs, where the codes corre-
spond to the mass of sliding material. The dotted
line represents the linear regression of the 75 kg’s
run, a
max
=
-66.56t
+
150.76, R
2
=
0.915
Fig. 8 - Relation between the ratio of maximum amplitude
and maximum height of wave and the dimension-
less parameter Ψ
background image
F. BREGOLI, A. BATEMAN PINZÓN, V. MEDINA IGLESIAS & D.A. GÓMEZ CORTÉS
120
International Conference on Vajont - 1963-2013 - Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
Data of the final deposit’s geometry are recollect-
ed. The basal area of the final deposit is thus related to
the momentum flux of the sediment, showing a com-
forting trend (Fig. 9).
Results of experiments on debris flow’s impact
are reported in Tab. 2. Here it can be observed that the
weight increases with increasing impact forces (maxi-
mum and medium). Fixing the weight and varying the
height of the slide, the maximum force of impact is
proportional to the height (Fig. 10). The Relation of
maximum force over the averaged force of impact and
the Froude number of the slide is shown on Fig. 11.
DISCUSSION AND CONCLUDING RE-
MARKS
An experimental device is set up to study the
phenomenon of impulse wave. The facility generat-
ing high speed granular slides is able to perform the
described behaviour correctly. The high velocity is
necessary to create a high turbulence where high Rey-
nolds number allows a certain comparison between
laboratory and reality.
The limited number of experiment and the narrow
range of variables such as vs are not always enough to
define pattern of regression. The relationships shown
in Fig. 8 should be a good way to relate consequence
(wave height) and cause (the sliding mass opposing
the water pressure), but it needs to be analysed in wid-
er ranges of parameters.
The experiments on slides impact over a planar
plate give a clear idea on the relation between the
characteristics of the slide and the triggered forces
(Fig. 10). Moreover the relationship became strong-
er when only comparing the height of the sliding
mass and the maximum force acting over the plate
(Fig. 11). This remark gives to the height of the slide
Fig. 9 - Relation between the basal friction experienced
by the deposit and the momentum sediment flux
Tab. 2 - Results of experiments of impact over a planar
structure at different loads and height of slide.
Fmax is the maximum resultant force, Fm is the
medium force, averaged in time till the stabilization
Fig. 10 - Relation between maximum force experienced by
the planar plate and the height of the slide. Linear
regression: F
max
=
17938h
s
, R
2
=
0.867
background image
EXPERIMENTAL STUDIES ON 3D IMPULSE WAVES GENERATED BY RAPID LANDSLIDES AND DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
121
an exceptional importance on the creation of impul-
sive waves, even if the comparison cannot be direct-
ly conducted because of the different behaviour in
presence: hits against a vertical wall or a water basin
differ significantly.
New experiments, exploring a wider range of
variables, will be carried out to improve the variables
correlations.
ACKNOWLEDGEMENTS
This work is founded by the project “Debris
Flow” (CGL 2009-13039) of the Ministry of Educa-
tion of Spain. The authors want to kindly thank the
team of the WAV laboratory of Zurich for the coop-
eration during the set-up of the experimental device.
Fig. 11 - Dimensionless relationship between cause (veloc-
ity and depth of granular slide) and forces (maxi-
mum and average forces). The power regression is
y
=
165.59x
-0.96
, R
2
=
0.573
REFERENCES
B
ellotti
g., D
i
r
isio
M. & D
e
g
irolaMo
P. (2009) - Feasibility of tsunami early warning systems for small volcanic islands.
Nat. Hazards Earth Syst. Sci. 9: 1911-1919.
B
regoli
F. (2008) - Messa a punto di un’ apparecchiatura sperimentale e prove preliminari per lo studio della morfodinamica di
alvei in ghiaia. Master Thesis in Civil and Environmental Engineering, University of Florence, Italy (In Italian).
c
oussot
P. & M
eunier
M. (1996) - Recognition, classification and mechanical description of debris flows. Earth-Science
Reviews 40: 209-227.
D
atei
C. (1968) - Riproduzione su modello in scala 1:500 della frana caduta nel 1959 entro il lago-serbatoio del Maé. Internal
report. Univ. df Padova. Padova, Italy (in Italian).
F
ritz
H.M., H
ager
w.H. & M
inor
H.e. (2001) - Lituya bay case: rockslide impact and wave run-up. Sci. Tsunami Haz., 19
(1): 3-22.
F
ritz
H.M. (2002) - Initial phase of landslide generated impulse waves. Ph.D Thesis dissertation, ETH Zurich, Switzerland.
F
ritz
H. M., H
ager
w. H. & M
inor
H.e. (2004) - Near field characteristics of landslide generated impulse waves. Journal of
Waterway, Port, Coastal, and Ocean Engineering, 130 (6): 287-302.
H
eller
V., H
ager
w.H. & M
inor
H.e. (2007) - Scale effects in subaerial landslide generated impulse waves. Exp Fluids,
44: 691-703.
H
eller
V. & H
ager
w.H. (2010) - Impulse product parameter in landslide generated impulse waves. Journal of Waterway, Port,
Coastal, and Ocean Engineering, 136 (3): 145-155.
H
uBer
A. (1980) - Schwallwellen in seen als folge von felsstürzen, VAW-Mitteilung, 180. D. V
iscHer
, ed., ETH Zurich (in
German).
H
ungr
o., e
Vans
s.g., B
oVis
M.J. & H
utcHinson
J.n. (2001) - A review of the classification of landslides of the flow type.
Environmental & Engineering Geoscience, 7 (3): 221-238.
i
Verson
r.M., r
eiD
M.e. & l
a
H
usen
r. g. (1997) - Debris-flow mobilization from landslides. Annu. Rev. Earth Planet. Sci.,
25: 85-138.
K
aMPHuis
J.w. & B
owering
r.J. (1972) - Impulse waves generated by landslides. Proc. of 12
th
ICCE, ASCE: 575-588.
M
araMai
a., g
raziani
l., & t
inti
s. (2005) - Tsunamis in the Aeolian Islands (southern Italy): a review. Mar. Geol., 215
(1-2): 11-21.
M
iyaMoto
K. (2010) - Numerical simulation of landslide movement and Unzen-Mayuyama disaster in 1792, Japan. Journal of
Disaster Research, 5 (3): 280-287.
n
oDa
E. (1970) - Water waves generated by landslides. Journal of the Waterways, Harbours and Coastal Engineering Division,
background image
F. BREGOLI, A. BATEMAN PINZÓN, V. MEDINA IGLESIAS & D.A. GÓMEZ CORTÉS
122
International Conference on Vajont - 1963-2013 - Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
96 (4): 835-855.
P
anizzo
a., D
e
g
irolaMo
P., & P
etaccia
a. (2005) - Forecasting impulse waves generated by subaerial landslides. J. Geophys.
Res., 110 (1): 20-25.
s
cott
-r
ussell
J. (1837) - Experimental researches into the laws of certain hydrodynamical phenomena. In: Edin. Roy Soc Trans
XIV, 1840, 47-109, plus Plates I and III.
T
inti
s., M
anucci
a., P
agnoni
g., a
rMigliato
a. & z
aniBoni
F. (2005) - The 30 December 2002 landslide-induced tsunamis in
Stromboli: sequence of the events reconstructed from the eyewitness accounts. Nat. Hazards Earth Syst. Sci., 5: 763-775.
z
weiFel
a., H
ager
w.H., M
inor
H.e. et alii (2006) - Plane impulse waves in reservoirs. Journal of Waterway, Port, Coastal, and
Ocean Engineering, 132: 358.
Statistics