Nat. Hazards Earth Syst. Sci.

, 9, 451–458, 2009
www.nat-hazards-earth-syst-sci.net/9/451/2009/ Natural Hazards
© Author(s) 2009. This work is distributed under and Earth
the Creative Commons Attribution 3.0 License. System Sciences

Review Article: Numerical analysis of the seismic behaviour of

earth dam
Y. Parish, M. Sadek, and I. Shahrour
Laboratory of Mechanics of Lille, University of Science and Technology of Lille, 59 650 Villeneuve d’Ascq, France
Received: 24 July 2008 – Revised: 5 February 2009 – Accepted: 16 March 2009 – Published: 23 March 2009

Abstract. The present study concerns analysis of the seismic 2000). Well-compacted modern dams can withstand substan-
response of earth dams. The behaviour of both the shell and tial earthquake shaking with no detrimental effects.
core of the dam is described using the simple and popular non The pseudo static approach is largely used in engineering
associated Mohr-Coulomb criterion. The use of this constitu- practice to assess the seismic stability of earth fill dams. In
tive model is justified by the difficulty to obtain constitutive this approach, the earthquake effect on a potential soil mass
parameters for more advanced constitutive relations includ- is represented by means of equivalent static horizontal force
ing isotropic and kinematic hardening. Analyses with real equal to the soil mass multiplied by a seismic coefficient.
earthquake records show that the seismic loading induces This approach is quite simplistic since it attempts to represent
plasticity in a large part of the shell and in the lower part of complex dynamic behaviour in terms of static forces. Stabil-
the core. Analysis shows that plasticity should be considered ity is expressed in terms of an overall factor of safety. The
in the analysis of the seismic response of the dam, because it implicit assumption is that the soil is rigid-perfectly plastic
leads to a decrease in the natural frequencies of the dam to- behaving as an undeformable block.
gether to energy dissipation, which could significantly affect Since the 1971 San Fernando earthquake in California
the seismic response of the dam. Plastic analysis constitutes (Ming and Li, 2003), major progress has been achieved in
also a good tool for the verification of the stability of the dam the understanding of the earthquake action on dams. Gaze-
under seismic loading. tas (1987) discussed the historical developments of theoret-
ical methods for estimating the dynamic response of earth
dams to earthquakes ground excitation. He outlined their im-
1 Introduction portant features, their advantages and limitations. Progress
in the area of geotechnical computation and numerical mod-
Seed et al. (1978, 1979) reported that the seismic perfor- eling offers interesting facilities for the analysis of the dam
mance of embankment dams has been good in general, ex- response in considering complex issues such as the soil non
cept when liquefaction or unusual circumstances have been linearity, the evolution of the pore pressure during the dam
involved. They noted that a well-built compacted embank- construction procedure and real earthquake records. De-
ment dam can withstand moderate earthquake shaking, with tailed analysis techniques include equivalent linear (decou-
peak accelerations of 0.2 g and more, with no detrimental pled) solutions, and non linear finite element and finite dif-
effects. The efficiency of modern compacted embankment ference coupled or decoupled formulations (Lin and Chao,
dams was further demonstrated in 1994 when the Los Ange- 1990; Abouseeda and Dakoulas, 1998; Cascone and Ram-
les Reservoir was severely shaken by the Northridge Earth- pello, 2003).
quake (Davis and Sakado, 1994). The seismic performance Wood (1973) showed that where the frequency at which
of embankment dams has been closely related to the nature the principal energy of the input motions approaches the
and state of compaction of the fill material USCOLD (1992, fundamental frequency of the unrestrained backfill, dynamic
amplification becomes an important factor, which is not con-
sidered in engineering approaches that assess the earth fill
Correspondence to: M. Sadek dam stability. The paper presents a numerical study of the
([email protected]) seismic behaviour of earthfill dams. It will mainly focus on

Published by Copernicus Publications on behalf of the European Geosciences Union.

 452 Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam

Table 1. Properties of foundation and earth dam soils.

Units Foundation Earth dam Shell

Core
Dry density (ρ) (kg/m3 ) 2200 1800 2000
Young’s modulus (E) (MPa) 1000 40 60
Poisson’s ratio (ν) 0.25 0.30 0.30
Elastic shear modulus (G) (MPa) 400 15.38 23.08
Bulk modulus (K) (MPa) 666.67 33.33 50.00
Cohesion (C) (kPa) – 100 0.10
Plasticity Friction angle (ø) (Degree) – 15 35
Dilation angle (ψ) (Degree) – 3 10
Fig. 1. Geometry of dam and the shape of zones (units in meter).

Figure 1. Geometry of dam and the shape of zones (units in meter).

seismic amplification at the dam crest under real earthquake discretization using the Langrangian approach (FLAC3D,
records with frequency content close to the natural frequency 2005).
of the dam. In the present study, analysis is conducted using This program is based on a continuum finite difference dis-
a 3-D finite difference modeling. It corresponds to the re- cretization using the Langrangian approach. Every derivative
sponse of the dam before water impoundment. Results are in the set of governing equations is replaced directly by an
presented first considering linear visco-elastic behaviour for algebraic expression written in terms of the field variables
the dam material; then analyses are conducted within the (e.g. stress or displacement) at discrete point in space. For
framework of plasticity in order on investigate the influence dynamic analysis, it uses an explicit finite difference scheme
of plasticity on the seismic response of the dam. Indeed, plas- to solve the full equation of motion using lumped grid point
ticity could affect both the material damping and the domi- masses derived from the real density surrounding zone. The
nant frequencies of the dam. calculation sequence first invokes the equations of motion to
The elastoplastic analysis constitutes an efficient tool for derive new velocities and displacements from stresses and
the investigation of the stability of dams under seismic load- forces. Then, strain rates are derived from velocities, and
ing. The seismic induced displacement could be used for the new stresses from strain rates. Every cycle around the loop
evaluation of the stability of the dam. Since the displacement correspond to one time step. Each box updates all of its grid
is dependant on the geometric and mechanical properties of variables from known values that remain fixed over the time
the dams as well as the frequency and amplitude of the input step being executed.
13
loading, a specific study could be conducted for the analysis Dynamic loading is applied at the base of the foun-
of the influence of the major input parameters on the stability dation layer as a velocity excitation. Kuhlemeyer and
of the dam. This analysis is not presented in this paper. Lysmer (1973) showed that for an accurate representation of
This paper does not consider the fluid-skeleton interaction, the wave transmission through the soil model, the spatial ele-
which could have a significant influence on the seismic re- ment size, 1l, must be smaller than approximately one-tenth
sponse of the dam. This issue was recently investigated by to one-eighth of the wavelength associated with the highest
the authors (Parish et al., 2008). frequency component of the input wave i.e.,

2 Problem under consideration 1l ≤ λ/10 (1)

The selected example is a simplified representation of typ- λ is the wave length associated with the highest frequency
ical earth dam geometry. The dam section assumed in the component that contains appreciable energy. The conse-
present survey is a symmetric zone section with clay core quence is that reasonable analyses may be time and mem-
and foundation as shown in Fig. 1. Geotechnical properties ory consuming. In such cases, it may be possible to adjust
used in the analyses are presented in Table 1 for foundation the input by recognizing that most of the power for the input
soil and earth dam materials. The materials properties are history is contained in lower frequency components.
chosen more close to reality. The foundation is assumed to The procedure of Free-Field Boundaries used in FLAC3D
be stiff with a Young’s Modulus E=1000 MPa. The Young’s aims absorbing outward waves originating from the struc-
modulus of the shell is equal to 60 MPa, while that of the ture. The method involves the execution of free-field calcu-
core is equal to 40 MPa lations in parallel with the main-grid analysis. The lateral
boundaries of the main grid are coupled to the free-field grid
2.1 Numerical model by viscous dashpots to simulate a quiet boundary.
Rayleigh damping Rd =5% is used in the analyses to
Numerical analyses are conducted using the finite difference compensate for the energy dissipation through the medium
program FLAC3D based on a continuum finite difference (Paolucci, 2002; Lokmer et al., 2002). The behaviour of the

Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009 www.nat-hazards-earth-syst-sci.net/9/451/2009/

Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam 453

40 50
40
1.6
30
Displacement (cm)

30

Velocity (cm/sec)
20 1.4
20
10 10
1.2

spectral velocity(m/s)
0 0
-10 -10
-20
-20 1
-30
-30 -40
0 5 10 15 20 25 30 0 5 10 15 20 25 30
0.8
Tim e(sec) Tim e(sec)
0.6
0.3 3.50E-04
0.4

spectral velocity(m/s)
0.2 3.00E-04
Acceleration (g)

2.50E-04
0.1 0.2
2.00E-04
0
1.50E-04 0
-0.1 1.00E-04 0 0.7 1.4 2.1 2.8 3.5 4.2
-0.2 5.00E-05 frequency(HZ)
-0.3
0.00E+00
0 5 10 15 20 25 30 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Tim e(sec) frequency(HZ)
Fig. 3. Response spectra of free horizontal motion at the dam crest.

Fig. 2. Kocaeli earthquake record, 1999 (Ambarli Station, Record

Figure 3. Response spectra of free horizontal motion at the dam crest.
ID P1086) (a) Displacement,(b) Velocity, (c) Acceleration, (d)
Figure 2. Spectra
Fourier Kocaeli earthquake record, 1999
of Velocity (Ambarli Station, Record ID P1086)
Component. H

a) Displacement, b) Velocity, c) Acceleration, d) Fourier Spectra of Velocity Component. h

shell and core of the dam is described using the non asso-
ciated Mohr Coulomb criterion (Table 1); in this case the
hysteretic damping is considered using Rayleigh damping
Rd =2%. The use of this model is justified by the difficulty
to obtain constitutive parameters for more advanced consti- Fig. 4. Dam deformation at the maximum of excitation (Kocaeli
tutive relations including both isotropic and kinematic hard- earthquake record) (Umax =0.30 m at the dam crest).
Figure 4. Dam deformation at the maximum of excitation (Kocaeli earthquake record)
ening.
(Umax = 0.30 m at the dam crest)

2.2 Input loading . 3 Analysis of the seismically induced response in the

dam
Dynamic loading is applied at the base of the foundation 14
layer as a velocity excitation. The earth dam is subjected to 3.1 Elastic response
earthquake loading representative of the 1999 Kocaeli earth-
quake in Turkey (Mw=7.4, Chen and Scawthorn 2003). The The response of the dam at the maximum excitation is pre-
estimated peak velocity is approximately 40 cm/s (peak ac- sented in Fig. 4. It shows an increase in the horizontal am-
celeration 0.247 g), and the duration is approximately 30 s. plification at the upper part of the dam. Figure 5 shows the
15
The record for base acceleration, velocity, and displacement velocity amplification in the axis of the dam. It can be ob-
waves are shown in Fig. 2a, b and c (records at Ambarli Sta- served that the amplification increases with the distance from
tion). Fourier analysis of the earthquake velocity record re- the foundation; it attains 3.45 at the top of the dam. Figure 6
sults in a power spectrum depicted in Fig. 2d. The veloc- shows the variation of the lateral amplification in the horizon-
ity spectrum reveals a dominant frequency of about 0.9 Hz tal direction at the middle height of the dam and the crest. In
(the second peak is observed at 1.3 Hz). Also note that most the first section, we observe a variation in the dynamic am-
of the power for the input history is contained in low fre- plification between 2 and 2.5. At the crest, we observe a
quencies. From the other hand, the natural frequencies of the uniform distribution of the amplification (close to 3.45).
foundation-dam system were determined by a Fourier anal-
ysis of the free vibration response of the dam (Fig. 3). It 3.2 Influence of Plasticity
16
shows a fundamental frequency f1 =0.7 Hz which is close to
dominant frequency of seismic loading (f =0.9 Hz); the sec- Figure 7 shows the location of the zones concerned by plastic
ond frequency is close to f2 =1.4 Hz (close to the second peak deformation at the peak of the seismic excitation. It can be
1.3 Hz) observed that plasticity is induced in a large part of the shell
and in the lower part of the core. The upper part of the core
remains in the elastic domain. Figure 8 shows the displace-
ment pattern in the axis and the middle height of the dam
at the maximum of seismic excitation. It can be observed
that the displacement in the axis of the dam is close to the
first mode of the dam; the variation of the displacement at
the middle height shows a sharp increase at the extremities,

www.nat-hazards-earth-syst-sci.net/9/451/2009/ Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009

 454 Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam

0.9

0.8

0.7

0.6

h / H 0.5
Fig. 7. Reference example: Distribution of plasticity in the dam (at maximum excitation).
0.4 Fig. 7. Reference example: Distribution of plasticity in the dam (at
0.3
(-n: the zone is
maximum at active failure now; -p: the zone has failed in the past; shear: yield criterion -
excitation).
shear failure)
(-n: the zone is at active failure now; -p: the zone has failed in the
0.2
past; shear: yield criterion -shear failure).
0.1

0 H
0 0.5 1 1.5 2 2.5 3 3.5 4 h 1 1.2
v/V
1
0.8
References

Dis (m) (m)

0.8
0.6

Displacement
Fig. 5. Velocity amplification in the dam axis (Kocaeli earthquake Abouseeda,
h/H
H. and Dakoulas P. Non-Linear Dynamic
0.6
Earth Dam-Foundation Interaction
record) (V : velocity at the base for h=0). Using0.4a BE-FE Method, Earthquake Engineering and Structural Dynamics, Vol. 27, pp. 917-
0.4

936, 1998
0.2
elocity amplification in the dam axis (Kocaeli earthquake record) 0.2

3 5 Gazetas,
0 G. : Seismic Response of Earth Dams: Some
0
Recent Developments”, Soil Dynamics
at the base for h=0) 0 0.05 0.1 0.15
Displacement
and Earthquake Engineering,
0.2 0.25 0.3
(m) 6, No. 1, pp.1-48, 1987.
Dis (m ) Vol.
0 0.2 0.4
l/L
0.6 0.8 1

2.5
4
Ohmachi T. and Kuwano
(Vertical J., dam)
axis of the Dynamic safety of earth (Middle
and rockheight
fill dams,
of the A.A.
dam) Balkema,
2
3 Rotterdam (1994), 160 p. ISBN 90-5410-265-9.
v / V 1.5 v/V

2
Fig. 8. Displacement pattern at maximum of velocity.
Figure 8. Displacement pattern at maximum of velocity.
1

1
0.5
3.3 Influence of the input loading
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
l/L l/L
This section presents an analysis of the influence of both the
(Middle height) (Dam crest) frequency content and the loading amplitude on the seismic
response of the dam.
Figure 6. Variation of the amplification in the horizontal direction.
Fig. 6. Variation of the amplification in the horizontal direction.
3.3.1 Frequency content of seismic excitation

which could indicate the imminence of soil instability in this The seismic performance of the dam is analysed through
area. The residual (permanent) displacement is presented in time-history analysis using the 1978 Tabas earthquake record
the Fig. 9 that shows the seismic induced residual displace- that has been adjusted to have a maximum amplitude veloc-
ment in the upper part of the dam and the extremities. As ity of 0.4 m/s in order to compare it with the response under
seen from this figure, whereas the residual displacement
17 al- Kocaeli earthquake. The time history of this record is illus-
most unvarying in the lower two-thirds of the crest, it dis- trated in Fig. 12 (9101 Tabas Station). Note that the dominant
plays significant increase in the upper one-third of the dam. frequency of this record is about 0.5 Hz.
In fact, previous research works indicated this behaviour Figure 13 shows a comparison between the dynamic am-
(Ohmachi and Kuwano, 1994; Ozkan et al., 2006). That’s plification profiles obtained with the elastoplastic analyses of
why in the design of embankment dams, due to the stronger the dam subjected the Kocaeli and Tabas earthquake records.
shaking at the upper parts, special attention should be given We observe an agreement between these profiles. The max-20
to the crest to avoid undesirable deformations. imum velocity due to the Tabas earthquake record is equal
Figure 10 shows a comparison between the elastic and to 1.26 m/s (v/V =2.97); this value is higher than that ob-
elastoplastic analyses at the maximum of velocity. It can tained with the Kocaeli earthquake record (Vmax =0.95 m/s,
be observed that the plastic deformation leads to a decrease v/V =2.59).
in the velocity amplification, in particular in the upper part. Figure 14a and b depict the response spectra obtained with
This reduction attains about 50%. This result could be at- 18 the Kocaeli and Tabas earthquake records together with the
tributed to the energy dissipation by plastic deformation and input motions respectively for elastic and elastoplastic mod-
to the influence of plasticity on the reduction of the funda- eling. For elastic analysis, it can be observed that maximum
mental frequencies of the dam as illustrated Fig. 11, which peak of the response to Tabas earthquake (respectively Ko-
shows the influence of plasticity on the spectral response at caeli) occurs at the first frequency of the dam f1 dam =0.7 Hz
the crest of the dam. (respectively at f2 dam =1.4 Hz) while the second peak is

Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009 www.nat-hazards-earth-syst-sci.net/9/451/2009/

Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam 455

1 1.2
1
1 Elastic-Sv max =1.95
0.9
0.8 0.8
0.8 Plastic-Sv max=1.59e-2
0.6

spectral velocity ratio

Dis (m) (m)
0.6 0.4 0.7 Kocaeli- Load - Sv max=3.19e-4

Displacement
h/H 0.2
0.6
0.4 0
0 0.2 0.4 0.6 0.8 1 0.5
-0.2

0.2 -0.4 0.4

-0.6 0.3
0 -0.8
0.2
-0.03 -0.02 -0.01 0 -1
Displacement
Dis (m )
(m) l/L 0.1

0
(Vertical axis of the dam) (Middle height of the dam) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
frequency(HZ)

Fig. 9. Residual displacement.

Figure 9. Residual displacement. Fig. 11. Reference example: Influence of plasticity on velocity
spectra at the dam crest.
Figure 11. Reference example: Influence of plasticity on velocity spectra at the dam crest.
Plastic Elastic Plastic Elastic
1 2.5
50 50
40 40
0.8 2
30
30

Displacement (cm)

Velocity (cm/sec)
20
1.5
20
0.6 10
v/V 0 10
h/H
1 -10 0
0.4
-20
-10
-30
0.2 0.5 -20
-40
-50 -30
0 0 -60 -40
0 1 2 3 4 0 0.2 0.4 0.6 0.8 1 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
v/V l/L Time (sec) Time (sec)

(Vertical axis of the dam) (Middle height of the dam) 0.6 0.012
0.5

Spectral Velocity (m/sec)

0.4 0.01
Fig. 10. Reference example: Influence of plasticity on the maxi-
Acceleration (g)

0.3
0.008
0.2
mum of velocity. 0.1
0.006
Figure 10. Reference example: Influence of plasticity on the maximum of velocity. 0
-0.1
0.004
-0.2
-0.3 0.002
obtained at the first loading frequency f1 load =0.50 Hz (re- -0.4
-0.5 0
spectively at f2 load =1.3 Hz). When plasticity is considered, 0 1 2 3 4 5 6 7 8 0 0.25 0.5 0.75 1 23
1.25 1.5 1.75
Time (sec) Frequency(HZ)
the maximum spectral velocity are obtained at the loading21
dominant frequencies (f1 load =0.50 Hz for Tabas; and f2 load
for Kocaeli). Fig. 12. Tabas earthquake record (9101 Tabas Station) (a) Displace-
ment, (b) Velocity, (c) Acceleration, (d) Spectra.
Figure 12. Tabas earthquake record (9101 Tabas Station)
3.3.2 Loading amplitude
a) Displacement, b) Velocity, c) Acceleration, d) Spectra.
displacement in the horizontal direction clearly indicates the
The influence of the loading amplitude on the dam response .
presence of instability at high amplitude of the input motion.
to the Kocaeli earthquake record was investigated through
analyses conducted for three values of amplitude of the in- 3.4 Mechanical role of Core and Shell
put motion (Va =0.40, 1.60 and 3.60 m/s). Such values are
extremely high, they are considered for parametric studies. In order to investigate the mechanical role of the core rela-
Figure 15 shows the influence of the input motion amplitude tive to the shell on the overall seismic induced response of the
on the distribution of plasticity in the dam. The plasticity22 dam. Numerical simulations are carried out with the Kocaeli
distribution increases with the loading amplitude. For both earthquake record. Analyses were conducted with different
Va =1.60 and 3.60 m/s, plasticity extends to the totality of the values of the core cohesion (100×105 kPa) and the shell co-
dam. Note that these values of ground velocity correspond to hesion (0.1 kPa×105 kPa) as described in Table 2. High val-24
extremely high-magnitude earthquakes. ues of the cohesion in are used to elucidate the influence of
Figure 16 shows the influence of the loading amplitude plasticity on the response of the dam.
on the maximum of displacement in the vertical axis and at Figure 17 presents a comparison results for different mod-
the middle height of the dam. It can be observed that the els in term of the velocity amplification in three sections of
increase in the input motion amplitude leads to an important the dam: vertical axis, middle height and the crest. It can
increase in this displacement, in particular near the lateral be seen that the variation of core cohesion doesn’t induce a
extremities of the dam, where the displacement attains 3.5 m significant change in the seismic amplification. Considera-
for Va =3.6 m/s which indicates collapse. The variation of the tion of plasticity in the shell (cohesion=0.1 kPa) leads to a

www.nat-hazards-earth-syst-sci.net/9/451/2009/ Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009

456 Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam

1
Kocaeli Tabas
2.5
Kocaeli Tabas
Block State
None
0.8 2 shear-n shear-p
shear-p
0.6 1.5
h/H v/V
0.4 1

0.2 0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 0.2 0.4 0.6 0.8 1
v/V l/L
Block State
None
(Vertical axis of the dam) (Middle height of the dam) shear-n shear-p
shear-p

Fig. 13. Comparison of the elastoplastic responses to the Kocaeli

and Tabas earthquake records.
Figure 13. Comparison of the elastoplastic responses to the Kocaeli and Tabas earthquake
records.
ELASTIC
Block State
1
Tabas - Load - SV =1.330e-2
max
Tabas - Response - SV
max
=8.70e-2 None
0.8
Kocaeli - Load - SV
max
=3.19e-4 shear-n shear-p
Kocaeli - Response - SV
max
=1.95 shear-p
Spectral velocity ratio

0.6

0.4

Fig. 15. Influence of the loading amplitude on the distribution

0.2 of plasticity (Kocaeli earthquake record) : (a) Va =0.40 m/s, (b)
Va =1.60
Figure m/s, (c)ofVthe
15. Influence a =3.6 m/s.amplitude on the distribution of plasticity (Kocaeli
loading
0 earthquake record) : a) Va = 0.40 m/s, b) Va = 1.60 m/s, c) Va = 3.6 m/s.
0 0.4 0.8 1.2 1.6 2 2.4
Frequency (Hz) . 1
4

(a) 0.9
3.5
0.8
Elastplastic
1 3
0.7
Tabas - Load - SV =1.330e-2
m ax
0.6 2.5
Tabas - Response - SV =7.67e-2
m ax
h/H
25 Ux (m)
Kocaeli - Load - SV =3.19e-4 0.5
0.8 m ax
2
Kocaeli - Response - SV =1.59e-2
m ax
0.4
Spectral velocity ratio

1.5
0.3
0.6
0.2
Va = 0.40 m/s
1
27
Va = 1.60 m/s
0.1 Va = 3.60 m/s 0.5
0.4 0
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0 0.2 0.4 0.6 0.8 1
Ux (m) l/L
0.2

(Vertical axis of the dam) (Middle height of the dam)

0
0 0.4 0.8 1.2 1.6 2 2.4
Fig. 16. Influence of the loading amplitude on the maximum of
Frequency (Hz)
displacement in the dam (Kocaeli earthquake record).
(b) Figure 16. Influence of the loading amplitude on the maximum of displacement in the dam
Figure 14. Comparison of the spectra of the elastic (a) and elastoplastic (b) responses to the (Kocaeli earthquake record).
Kocaeli
Fig. 14.andComparison
Tabas earthquake
of records (The response
the spectra of the iselastic
taken at(a)
the and
crest).elastoplas-
ral frequency of the dam and dominant frequency of loading
tic (b) responses to the Kocaeli and Tabas earthquake records (The
record. The response spectra confirms also that the variation
response is taken at the crest).
in core cohesion don’t affect the overall response of the dam
where we obtain practically the same result for model 1 and
26
decrease of about 50% in the seismic amplification. This re- 3.
sult could be attributed to the energy dissipation by plastic
deformation and to the influence of plasticity on the reduc- 4 Conclusions
tion of the fundamental frequencies of the dam as illustrated
Fig. 18, which shows the influence of plasticity on the spec- This paper included analysis of the seismic behaviour of
tral response at the crest of the dam. The Fourier spectrum earth dams. It corresponds to the response of the dam before
illustrates a large component of amplification factor at natu- water impoundment. Analyses were conducted for real

Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009 www.nat-hazards-earth-syst-sci.net/9/451/2009/28

Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam 457

Table 2. Core and shell cohesion for different models of the earth Model1 Model2 Model3 Model4

dam.
1 2.5 3.5
0.9
3
0.8 2
Model 1 Model 2 Model 3 Model 4
0.7 2.5
1.5
100×103 100 ×103
0.6
Core cohesion (kPa) 100 100 h / H 0.5 v/V v/V
2

Shell cohesion (kPa) 100×103 0.1 100×103 0.1 0.4 1 1.5

0.3 1
0.2 0.5
0.5
0.1
0 0 0
0 1 2 3 4 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
v/V l/L l/L

earthquake records. Non linear elastoplastic analyses show

(Vertical axis of the dam) (Middle height of the dam) (Dam crest)
that the seismic loading induces plasticity in a large part of
the shell and in the lower part of the core. The variation of
the displacement in the middle height shows a sharp increase Fig. 17. Comparison of velocities results in different models for
at the extremities, which could indicate the imminence of Turkey earthquake wave. (For model 1×model 4: see Table 2)
Figure 17. Comparison of velocities results in different models for Turkey earthquake wave.
instability in this area. The upper third part of the crest
(For model 1Æ model 4: see table 2)
should be designed carefully, and a design should take into 1
account dominant frequency of the load and the dam where 0.9
Kocaeli Load-max=3.19e-4
Model 1
the movement could be significantly amplified. On the other

spectral velocity ratio (m/s)

0.8
Model 2
hand, it can be observed that the plastic deformation leads 0.7
Model 3
to a decrease in the velocity amplification, in particular in 0.6
Model 4
0.5
the upper part of the dam. Parametric analysis shows that 0.4
mechanical role of the core is not significant compared to 0.3

the shell in the overall seismic induced response of the earth 0.2

dam. Analysis shows that plasticity should be considered in 0.1

0
the investigation of the seismic response of the dam, because 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

it leads to a decrease in the natural frequencies of the dam frequency(HZ)

together to an increase in energy dissipation, which could

significantly affect the seismic response of the dam. Plastic Fig. 18. Comparison of the spectra of the different models to the
analysis constitutes also a good tool for the verification of Kocaeli earthquake record.
the stability of the dam under seismic loading. Figure 18. Comparison of the spectra of the different models to the Kocaeli earthquake
record.

Edited by: M. Contadakis 29

Lin, J. and Chao, B.: Estimation of Shear Moduli and Damping Fac-
Reviewed by: A. Ouahsine and two other anonymous referees
tors of Earth Dam Materials, Earthquake Engineering and Struc-
tural Dynamics, 19, 891–910, 1990.
Lokmer, I., Herak, M., Panza, G. F., and Vaccari, F.: Amplification
References of strong ground motion in the city of Zagreb, Croatia, estimated
by computation of synthetic seismograms, Soil Dynam. Earth.
Abouseeda, H. and Dakoulas, P.: Non-Linear Dynamic Earth Dam- Eng., (22), 105–113, 2002.
Foundation Interaction Using a BE-FE Method, Earthquake En- Ming, H. Y. and Li, X. S.: Fully coupled analysis of failure and
gineering and Structural Dynamics, 27, 917–936, 1998. remediation of lower San Fernando dam, J. Geotech. Geoenvi-
Cascone, E. and Rampello, S.: Decoupled Seismic Analysis of ron.Eng., ASCE, 129(4), 336–49, 2003.
an Earth Dam, Soil Dynamics and Earthquake Engineering, 23, Ohmachi, T. and Kuwano, J.: Dynamic safety of earth and rock
349–365, 2003. fill dams, edited by: Balkema, A. A., Rotterdam, 160 p., ISBN
Chen, W. F. and Scawthorn, C.: Earthquake Engineering Handbook, 90-5410-265-9, 1994.
CRC Press LLC, 2003. Özkan, M., YÖzyazicioglu, M., and Aksar, U. D.: An evaluation of
Davis, C. A. and Sakado, M. M.: Response of the Van Norman Güldürcek dam response during 6 June 2000 Orta earthquake,
Complex to the Northridge Earthquake, Association of Dam Soil Dynamics and Earthquake Engineering, 26(5), 405–419,
Safety Officials Conference Proceedings, Boston, MA, 241–255, 2006.
30
1994. Parish, Y., Shahrour, I., and Sadek, M.: Influence of the water-
Gazetas, G.: Seismic Response of Earth Dams: Some Recent De- skeleton interaction on the seismic response of earth dams, Int. J.
velopments, Soil Dynam. Earth. Eng., 6(1), 1–48, 1987. Multiphysics, 2(3), 267–277, 2008.
FLAC3D, Itasca Consulting Group, FLAC: Fast Lagrangian Analy- Paolucci, R.: Amplification of earthquake ground motion by steep
sis of Continua, vol. I. User’s Manual, vol. II. Verification Prob- topographic irregularities, Earthquake Engineering and Struc-
lems and Example Applications, Second Edition (FLAC3D Ver- tural Dynamics, 31, 1831–1853, 2002.
sion 3.0), Minneapolis, Minnesota 55401 USA, 2005. Seed, H. B., Makdisi, F. I., and De Alba, P.: Performance of Earth
Kuhlmeyer, R. L. and Lysmer, J.: Finite Element Method Accuracy Dams During Earthquakes, J. Geotechn. Eng., American Society
for Wave Propagation Problems, J. Soil Mech. Foundations Div., of Civil Engineers, 104(GT7), 967–994, 1978.
ASCE, 99(SM5), 421–427, 1973. Seed, H. B: Considerations in the Earthquake-Resistant Design of

www.nat-hazards-earth-syst-sci.net/9/451/2009/ Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009

458 Y. Parish et al.: Numerical analysis of the seismic behaviour of earth dam

Earth and Rockfill Dams, 19th Rankine Lecture of the British Wood, J.: Earthquake-Induced Soil Pressures on Structures, Re-
Geotechnical Society, Geotechnique, 29(3), 215–263, 1979. port EERL 73-05, California Institute of Technology, Pasadena,
USCOLD (US Committee on Large Dams): Observed Performance p. 311, 1973.
of Dams during Earthquakes, Committee on Earthquakes, July,
Denver, CO, 1992.
USCOLD (US Committee on Large Dams): Observed Performance
of Dams during Earthquakes, 2, Committee on Earthquakes, Oc-
tober, Denver, CO, 2000.

Nat. Hazards Earth Syst. Sci., 9, 451–458, 2009 www.nat-hazards-earth-syst-sci.net/9/451/2009/